Complete Report - University of New South Wales
Complete Report - University of New South Wales
Complete Report - University of New South Wales
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
ARC Centre <strong>of</strong> Excellence for<br />
Advanced Silicon Photovoltaics<br />
and Photonics<br />
Annual <strong>Report</strong> 2005
Centre <strong>of</strong> Excellence for<br />
Advanced Silicon Photovoltaics and Photonics<br />
is a Centre <strong>of</strong> Excellence <strong>of</strong><br />
The Australian Research Council
Page<br />
1. DIRECTORS' REPORT 1<br />
2. HIGHLIGHTS 3<br />
3. STAFF LIST 9<br />
4. RESEARCH<br />
4.1 INTRODUCTION TO RESEARCH 13<br />
4.2 FACILITIES AND INFRASTRUCTURE 17<br />
4.3 FIRST GENERATION: WAFER-BASED PROJECTS 25<br />
4.4 SECOND GENERATION: THIN-FILMS 51<br />
4.5 THIRD GENERATION: ADVANCED CONCEPTS 71<br />
4.6 SILICON PHOTONICS 97<br />
4.7 COLLABORATIVE RESEARCH 117<br />
5 EDUCATION 127<br />
6. ORGANISATIONAL STRUCTURE 145<br />
7. PROGRESS AGAINST TARGETS 147<br />
8. FINANCIALS 161<br />
9. PUBLICATIONS 165
DIRECTORS' REPORT<br />
Photovoltaics involve the conversion <strong>of</strong> light,<br />
especially sunlight, directly into electricity when<br />
falling upon devices known as solar cells. Silicon is<br />
the most common material used to make these cells,<br />
as well as being the material most widely used in<br />
microelectronics.<br />
The Australian Research Council (ARC) Centre <strong>of</strong><br />
Excellence for Advanced Silicon Photovoltaics and<br />
Photonics (or, more compactly, the ARC Photovoltaics<br />
Centre <strong>of</strong> Excellence) <strong>of</strong>fi cially commenced at the<br />
<strong>University</strong> <strong>of</strong> <strong>New</strong> <strong>South</strong> <strong>Wales</strong> (UNSW) on 13th June,<br />
2003. The Centre‘s mission is to advance silicon<br />
photovoltaic research on three separate fronts, as<br />
well as to apply these advances to the related fi eld<br />
<strong>of</strong> silicon photonics. The educational activities <strong>of</strong><br />
the former Key Centre for Photovoltaic Engineering<br />
have also been successfully integrated into the new<br />
Centre.<br />
We are in a period <strong>of</strong> dynamic growth for<br />
photovoltaics, currently the world’s most rapidly<br />
growing energy source, with markets increasing at a<br />
compounded rate <strong>of</strong> 35%/year over the last decade,<br />
with the pace accelerating further over the last 2<br />
years. Most present sales are <strong>of</strong> “fi rst-generation”<br />
solar cells made from silicon wafers, similar to the<br />
wafers used in microelectronics. The Centre has a<br />
world-leading program with these “fi rst-generation”<br />
devices, holding international records for the highestperforming<br />
silicon cells in most major categories,<br />
including that for the outright highest-performing<br />
device. First-generation Centre research addresses<br />
the dual challenges <strong>of</strong> reducing cost and improving<br />
effi ciency. A key goal is to develop technology which<br />
allows silicon wafers doped with boron to be replaced<br />
by phosphorus-doped wafers. The problem with<br />
boron is that, under sunlight, it interacts with trace<br />
quantities <strong>of</strong> oxygen in the wafers, producing defects<br />
that reduce cell output. Phosphorus does not have<br />
this problem, but presents more challenges to highperformance<br />
cell fabrication. During 2005, the<br />
Centre’s progress in this area was documented<br />
by the demonstration <strong>of</strong> world-record 22.7% cell<br />
energy-conversion effi ciency using phosphorus-doped<br />
wafers.<br />
Silicon is quite a brittle material so silicon wafers<br />
have to be reasonably thick, about one-quarter <strong>of</strong> a<br />
millimetre, to be suffi ciently rugged for processing<br />
into solar cells with reasonable yields. Without this<br />
mechanical constraint, the silicon would perform<br />
well even if very thin, over 100 times thinner. Centre<br />
researchers have pioneered an approach where such<br />
very thin silicon layers are deposited directly onto a<br />
sheet <strong>of</strong> glass with the glass providing the required<br />
mechanical strength. Such a “second-generation”<br />
approach gives enormous potential cost savings<br />
since, not only are the costly processes involved in<br />
making wafers no longer required, but also there is<br />
an enormous saving in silicon material and cells can<br />
be made more quickly over the entire area <strong>of</strong> large<br />
glass sheets.<br />
The Centre is at the forefront <strong>of</strong> international<br />
research with this “second-generation”, “silicon on<br />
glass” approach. A partner in the Centre is CSG<br />
Solar, an originally Sydney-based company formed to<br />
commercialise the Centre’s initial approach. During<br />
2005, the company built the fi rst manufacturing<br />
plant for this new technology in Thalheim, Germany,<br />
with ramping-up <strong>of</strong> production scheduled for the fi rst<br />
half <strong>of</strong> 2006. Complementary work wholly within<br />
the Centre demonstrated improvements in its EVA,<br />
ALICIA and ALICE thin-fi lm solar cell concepts. A<br />
signifi cant factor behind these improvements was<br />
the acquisition <strong>of</strong> a major new processing facility, as<br />
a result <strong>of</strong> a major ARC Centre <strong>of</strong> Excellence/UNSW<br />
initiative outlined in our 2004 Annual <strong>Report</strong>.<br />
The silicon thin-fi lm approach has a large cost<br />
advantage over the wafer-based approach, due<br />
mainly to reduced material costs. However, in large<br />
enough production volumes, even these reduced<br />
material costs eventually will dominate the costs <strong>of</strong><br />
the thin-fi lm approach. This has led to the Centre’s<br />
interest in advanced “third-generation” thin-fi lm<br />
solar cells targeting signifi cant increases in energyconversion<br />
effi ciency. Higher conversion effi ciency<br />
means more power from a given amount <strong>of</strong> material,<br />
reducing costs. The Centre’s experimental program<br />
is concentrating on “all-silicon” tandem solar cells,<br />
where high energy-bandgap cells are stacked on<br />
top <strong>of</strong> lower-bandgap devices. The silicon bandgap<br />
is controlled by quantum-confi nement <strong>of</strong> carriers<br />
in small silicon quantum-dots dispersed in an<br />
amorphous matrix <strong>of</strong> silicon oxide, nitride or carbide.<br />
Progress with this approach was recognised by<br />
the award <strong>of</strong> substantial complementary funding<br />
during 2005 under the Global Climate and Energy<br />
Project managed by Stanford <strong>University</strong> and funded<br />
by a large international consortium. Other thirdgeneration<br />
approaches are being explored based<br />
on up-conversion or down-conversion <strong>of</strong> the energy<br />
<strong>of</strong> photons in the solar spectrum, to manipulate this<br />
spectrum so it is more suitable for conversion by<br />
1
the cell. Cells based on “hot” carriers are also being<br />
investigated since they <strong>of</strong>fer the potential for very high<br />
effi ciency from simple device structures. Although<br />
their implementation poses daunting challenges,<br />
systematic progress on addressing these was made<br />
during 2005.<br />
The fi nal Centre research strand involves silicon<br />
photonics where, as mentioned, the emphasis is<br />
upon using our experience with solar cells, using<br />
light to produce electricity, to the reverse problem<br />
<strong>of</strong> engineering silicon devices that use electricity<br />
to produce light. Devices <strong>of</strong> this type would be <strong>of</strong><br />
great interest in microelectronics since they would<br />
allow optical processes to be integrated with the<br />
normal electronic functions on the same silicon<br />
chip. The Centre holds the international record for<br />
the light emission performance from bulk silicon,<br />
in both electroluminescent and photoluminescent<br />
devices. Emphasis is now upon incorporating these<br />
improvements into silicon microchip-compatible light<br />
emitting devices using silicon-on-insulator wafers<br />
and upon exploiting our expertise in silicon light<br />
emission to develop new techniques for silicon wafer<br />
characterisation.<br />
In addition to these four research strands, the<br />
activities <strong>of</strong> the former Key Centre for Photovoltaic<br />
Engineering have been integrated into the ARC<br />
Centre <strong>of</strong> Excellence. The second cohort <strong>of</strong> students<br />
from the Bachelor <strong>of</strong> Engineering (Photovoltaics and<br />
Renewable Energy) program graduated during the<br />
year.<br />
This program has been enormously successful,<br />
attracting some <strong>of</strong> the best students entering the<br />
<strong>University</strong>. The Centre’s second undergraduate<br />
program, leading to a Bachelor <strong>of</strong> Engineering<br />
(Renewable Energy), completed its third year <strong>of</strong><br />
operation.<br />
Another notable achievement during 2005 was<br />
completion <strong>of</strong> the Centre’s major third year review.<br />
The Review Committee very favourably assessed<br />
the Centre on all aspects <strong>of</strong> its operations. Other<br />
highlights originated from the rapidly growing<br />
commercial interest in photovoltaics as a sustainable<br />
and increasingly marketable energy source. Centre<br />
partner, Suntech Power, a largely Australianowned<br />
and managed company operating in China,<br />
successfully fl oated on the <strong>New</strong> York Stock Exchange<br />
in December, in the largest technology fl oat <strong>of</strong> the year.<br />
A second partner and Centre “spin-<strong>of</strong>f”, CSG Solar,<br />
also announced the completion <strong>of</strong> its manufacturing<br />
facility in Germany, as previously noted, and licensee,<br />
BP Solar, announced signifi cant increases in the<br />
performance <strong>of</strong> its commercial product.<br />
We thank all those who contributed to the Centre’s<br />
success during 2005, particularly the Australian<br />
Research Council for its on-going support. The present<br />
continues to be a most exciting time for photovoltaics<br />
with time drawing closer to a future where solar cells<br />
provide a signifi cant part <strong>of</strong> the world’s energy needs,<br />
without the problems and escalating costs associated<br />
with the present mainstream options.<br />
………………………………………………<br />
Pr<strong>of</strong>essor Stuart R. Wenham,<br />
Director<br />
………………………………………………<br />
Pr<strong>of</strong>essor Martin A. Green,<br />
Executive Research Director<br />
2
HIGHLIGHTS<br />
Cell Efficiency World Record<br />
The Centre’s high-effi ciency group demonstrated an independently-confi rmed energy<br />
conversion effi ciency <strong>of</strong> 22.7% for a large area 22 cm2 cell fabricated on an n-type silicon<br />
substrate, equalling the world-record for a cell on such substrates. Other fi rst generation work<br />
within the buried-contact group resulted in further progress in developing low-cost sequences<br />
on such n-type substrates. Highlights include improved energy conversion effi ciency, opencircuit<br />
voltage and levels <strong>of</strong> surface passivation. Another highlight was the rapid development<br />
<strong>of</strong> applications <strong>of</strong> photoluminescence techniques for device and material characterisation, as<br />
a result <strong>of</strong> strong collaboration with the Centre’s silicon photonics group.<br />
Technology Commercialisation<br />
2005 was another exciting year for<br />
the photovoltaic industry with global<br />
production increasing 45%, building<br />
on the record 67% growth during<br />
2004. Annual sales <strong>of</strong> BP Solar<br />
<strong>of</strong> the Centre’s fi rst generation,<br />
buried contact technology again<br />
exceeded the $100 million mark,<br />
with the company announcing a<br />
new, higher performance module<br />
using this technology during 2005.<br />
The increasing international pr<strong>of</strong>i le<br />
<strong>of</strong> photovoltaics has resulted in an<br />
increasingly high visibility on fi nancial<br />
markets. Suntech Power, a largely<br />
Australia-owned company operating<br />
in China and a partner in the Centre<br />
partner, commenced trading on<br />
the <strong>New</strong> York Stock Exchange in<br />
December, in what became the<br />
largest technology-based initial public<br />
<strong>of</strong>fering <strong>of</strong> 2005, internationally.<br />
The Centre has been working with<br />
the company to bring high-effi ciency,<br />
fi rst generation cell concepts into<br />
production, including those based<br />
on semiconductor fi nger technology<br />
(below). A second Centre member,<br />
CSG Solar, successfully took Centre<br />
“second-generation” crystallinesilicon-on-glass<br />
technology a<br />
major step closer to large scale<br />
production, with the competition<br />
<strong>of</strong> the fi rst manufacturing plant in<br />
Germany. The fi rst product is due on<br />
the market in 2006, with production<br />
capacity also expected to approach<br />
the $100 million/year mark by the<br />
end <strong>of</strong> the year.<br />
Former student, Zhengrong Shi, rings the bell to open the NYSE after the<br />
Suntech IPO (Centre directors Stuart Wenham (right) and Martin Green<br />
(left) are in the background next to the President <strong>of</strong> the NYSE, Catherine<br />
Kinney and CEO John Thain, respectively)<br />
3
<strong>New</strong> Semiconductor Finger Technology<br />
The new semiconductor fi nger technology developed jointly between UNSW and Suntech-<br />
Power for screen-printed solar cells has been successfully implemented at the pilot production<br />
level using the same materials and equipment as for standard screen-printed solar cells. The<br />
10% performance gain is achieved without any increase in cell manufacturing cost, with the<br />
new jointly owned technology expected to be in large scale production in 2006. Based on<br />
pilot production experience, cell effi ciencies for the new technology in large scale production<br />
are expected to average about 18%. A particular strength <strong>of</strong> this new solar cell design is the<br />
ability to achieve extremely good spectral response, particularly for short wavelengths <strong>of</strong> light,<br />
compared to most commercial cell technologies. This new cell design appears to overcome<br />
the fundamental weaknesses <strong>of</strong> conventional screen-printed solar cells associated with the<br />
heavily diffused emitter, but while retaining the many benefi ts <strong>of</strong> screen-printing technology.<br />
This work won “The Best Oral Presentation” award for its area at the International PVSEC-15<br />
Conference in Shanghai in October 2005 and has been selected as an invited paper as part<br />
<strong>of</strong> the plenary session at the World Conference in Hawaii in May 2006.<br />
Image <strong>of</strong> a screen-printed<br />
metal line (foreground) in<br />
excellent contact with a<br />
heavily phosphorus-diffused<br />
laser groove running<br />
perpendicular to the metal<br />
line. The region covered with<br />
tiny pyramids is lightly diffused<br />
with phophorus.<br />
Thin-Film Silicon<br />
4<br />
Good progress has been made with the Thin-Film Group’s solar cell technologies EVA, ALICIA<br />
and ALICE. A milestone achieved in 2005 was the realisation <strong>of</strong> the Group’s fi rst poly-Si<br />
thin-fi lm solar cells on glass with voltages <strong>of</strong> over 500 mV. The best cell has a voltage <strong>of</strong><br />
517 mV obtained with the EVA structure which, to our knowledge, is a new world record<br />
for homojunction poly-Si thin-fi lm solar cells on glass. Another important milestone was<br />
the achievement <strong>of</strong> an internal quantum effi ciency <strong>of</strong> over 70%, obtained using the ALICIA<br />
structure. These advances benefi ted signifi cantly from an improved hydrogenator that became<br />
available in 2005 in the Centre’s research facility at Bay Street, Botany. Good progress was<br />
also made with optimising and upscaling <strong>of</strong> the Group’s AIT glass texturing method and its<br />
AIC seed layer process. As a result <strong>of</strong> this work, AIT-textured glass sheets and AIC-seeded<br />
glass sheets with a size <strong>of</strong> 10 x 10 cm 2 are now routinely produced by the Group. Progress<br />
with cell metallisation has improved the effi ciency <strong>of</strong> ALICIA cells to 3%. The Group was again<br />
active in generating intellectual property, with the fi ling <strong>of</strong> a provisional Australian patent on<br />
an interdigitated thin-fi lm cell metallisation method (“SAMPL”) and the entering <strong>of</strong> the national<br />
phase for two recent PCT patent applications. Furthermore, the Group has published one<br />
book chapter, eleven journal papers, and twenty one conference papers. Fifteen <strong>of</strong> these<br />
papers were presented as talks at international conferences, including four invited talks by<br />
the Group’s leader, Armin Aberle. The Group’s thin-fi lm solar cell technologies are attracting<br />
a signifi cant amount <strong>of</strong> interest from the solar industry and the venture capital sector. In<br />
response to this interest, the Group has created a commercialisation roadmap involving a<br />
business plan proposal and layouts <strong>of</strong> a pilot line and a fi rst factory. UNSW’s commercial arm,<br />
<strong>New</strong><strong>South</strong> Innovations Ltd, has entered negotiations with several interested parties.
PhD student Daniel Inns inspecting EVA poly-<br />
Si thin-fi lm solar cell material fabricated on a<br />
planar sheet <strong>of</strong> glass.<br />
Silicon Quantum Dots in Silicon Nitride<br />
The same technique previously used to make silicon quantum dots (QDs) by deposition <strong>of</strong><br />
thin Si rich layers (4nm) interspersed with thin silicon dielectric layers, used in silicon oxide<br />
based materials, has been successfully applied to silicon nitride based materials. The lower<br />
Si 3 N 4 barrier height should give a higher interdot probability for carrier transport, thus<br />
resulting in a higher conductivity material for the same QD density. The concept will next be<br />
applied to SiC with an even lower barrier height. Transfer to nitride is an important pro<strong>of</strong> <strong>of</strong><br />
concept <strong>of</strong> the transferability <strong>of</strong> the technology to different material systems. Growth has<br />
been accomplished both by the same reactive sputtering technique used for oxide and with<br />
PECVD. On annealing, both these structures produce nanocrystals <strong>of</strong> Si limited in size by the<br />
thickness <strong>of</strong> the Si rich layer – as demonstrated by the TEM images shown. Electron density<br />
contrast for the TEM images <strong>of</strong> QDs in the nitride is lower than in oxide but the crystal planes<br />
in the Si nanocrystals can still be seen. Characterisation <strong>of</strong> nanostructures is inherently<br />
diffi cult because <strong>of</strong> the small sizes involved. Hence in 2005 there has been a signifi cant effort<br />
to improve both the resolution and breadth <strong>of</strong> nanoscale characterisation techniques. This<br />
has resulted in successful XRD <strong>of</strong> Si nanostructures which nicely complements TEM data. It<br />
both corroborates the presence <strong>of</strong> Si nanocrystals and highlights some subtle differences in<br />
lattice parameter which are useful in identifying artifacts in both TEM and XRD.<br />
(a)<br />
120kV TEM image <strong>of</strong> PECVD deposited (a) Si quantum dots in 10 bi-layers <strong>of</strong> a silicon-rich-<br />
SiNx /SiNx superlattice and (b) 300kV High Resolution TEM image showing lattice planes in<br />
Si nanocrystals.<br />
(b)<br />
5
Stanford <strong>University</strong> GCEP Grant<br />
The Global Climate and Energy Project (GCEP) managed by Stanford <strong>University</strong> and fi nanced<br />
by an industrial consortium consisting <strong>of</strong> Exxon Mobil, General Electric, Sclumberger and<br />
Toyota is a major US$225 million, 10-year effort to seek new solutions to one <strong>of</strong> the major<br />
energy challenges <strong>of</strong> this century, supplying energy to meet and needs <strong>of</strong> a growing world<br />
population in a way that protects the environment. Promising areas for support have been<br />
identifi ed as photovoltaics, hydrogen, biomass, advanced combustion and CO 2 capture,<br />
separation and storage. After requesting proposals in these fi elds from selected institutions,<br />
these underwent review by two independent technology and impact review panels, as well as<br />
by a technical committee formed by the sponsors. The Centre responded to a request for<br />
a “game-changing” proposal in photovoltaics by outlining its activities in the nanostructured<br />
silicon tandem cell area (Section 4.5) and requesting support to accelerate the development<br />
<strong>of</strong> this approach. This was one <strong>of</strong> three successful projects chosen in the photovoltaics<br />
fi eld, and the only one awarded to a non-US group. Funding <strong>of</strong> US$2.4 million over 3 years<br />
has been allocated. Subsequently, Centre Executive Research Director, Martin Green, was<br />
selected as Team Leader for the overall GCEP Photovoltaic Program with responsibilities for<br />
identifying areas under-represented in the GCEP portfolio and in selecting technical experts<br />
for the evaluation <strong>of</strong> projects solicited in these areas.<br />
Other Awards<br />
Other highlights in 2005 include Mark Griffi n’s Faculty <strong>of</strong> Engineering Technical Staff Excellence<br />
Award, Axel Straub’s Graduate Student Award at the European Materials Research Society<br />
(E MRS) 2005 Spring Meeting in France, Roland Utama’s best fi nal-year thesis poster award<br />
and best overall thesis mark in Photovoltaic and Renewable Energy Engineering at UNSW in<br />
2005, and the award <strong>of</strong> a A$156,000 research infrastructure grant by UNSW to Patrick<br />
Campbell et al for a new multi-target reactive sputtering machine. Research student, Supriya<br />
Pillai, was the recipient <strong>of</strong> fi rst prize in the Faculty <strong>of</strong> Engineering’s Inaugural Dean’s Awards for<br />
Postgraduate Research, with her work selected as the best overall from about 70 applicants<br />
at a poster presentation at the Scientia in December. Papers presented by research student<br />
Chu-Wei Jiang, Deputy-Director Jianhua Zhao and Director Stuart Wenham all won best<br />
paper in their fi eld awards at the 15th International Photovoltaic Science and Engineering<br />
Conference in Shanghai in October, with Research Director Martin Green presented with a<br />
special award from Jiangsu province after his keynote presentation during the conference<br />
opening ceremony.<br />
Mark Griffi n’s winner <strong>of</strong> the Faculty <strong>of</strong> Engineering Technical<br />
Staff Excellence<br />
6
Education<br />
A major highlight <strong>of</strong> the educational activities has been the Centre’s ability to attract<br />
outstanding high school graduates to join us. The spread <strong>of</strong> their <strong>University</strong> Admissions Index<br />
values is skewed to the top end, while it would more commonly be expected to cluster towards<br />
the lower cut<strong>of</strong>f value.<br />
70<br />
UAI Distribution 2000 - 2005<br />
# Students<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
3656<br />
3655<br />
3657<br />
3642<br />
0<br />
Below 80 80-84.99 85-89.99 90-94.99 95-100<br />
<strong>University</strong> Admission Index<br />
Other highlights include a new Co-Op Scholarship for 2006 the Bachelor <strong>of</strong> Engineering in<br />
Photovoltaics and Solar Energy program, sponsored by Suntech Power Co. Ltd. <strong>of</strong> Wuxi,<br />
China. This complements the two existing scholarships, sponsored by BP Solar and CSG Solar.<br />
One <strong>of</strong> our high achieving undergraduate scholars, Nicole Kuepper, received an Australian<br />
fellowship as part <strong>of</strong> the Goldman Sachs Global Leaders Program. A new edition <strong>of</strong> the<br />
Centre’s popular text, Applied Photovoltaics, was completed in 2005 and printed in January<br />
2006. This book is destined for rapid translation to Chinese to support staff training for the<br />
booming industry in China. The Centre gained effective access in 2005 to a dedicated ro<strong>of</strong>top<br />
area for experimental activities in our education programs. Huge effort was expended by<br />
Centre teaching staff during 2006 to facilitate the introduction in 2006 <strong>of</strong> a new fl exible fi rst<br />
year program throughout the Faculty <strong>of</strong> Engineering and the associated development <strong>of</strong> a<br />
Faculty-wide fi rst year Engineering Design and Innovation course, as well as a new admissions<br />
scheme for local students.<br />
7
ARC Centre Review<br />
The Centre was reviewed by an ARC Review Panel on 3 November, 2005 with very<br />
positive outcomes. As well as recommending on-going funding <strong>of</strong> the Centre, the<br />
Review Panel’s comments included:<br />
“The Review Panel is impressed with the success <strong>of</strong> the ARC Centre <strong>of</strong> Excellence for<br />
Advanced Silicon Photovoltaics and Photonics and believes that since its inception in<br />
2003 the Centre has displayed a high level <strong>of</strong> achievement against all performance<br />
criteria….”;<br />
“The Centre is conducting world leading research in each <strong>of</strong> the four major areas with<br />
an array <strong>of</strong> corresponding performance records, the latest world record being in the area<br />
<strong>of</strong> n-type silicon solar cells, placing it at the forefront internationally”;<br />
“The Review Panel particularly noted that the Centre has proved to be highly innovative<br />
and very successful at commercialisation”;<br />
“The Centre is particularly strong at developing collaborations with industry and<br />
undertaking commercially significant research. Several <strong>of</strong> its technologies having<br />
been either commercialised or are approaching large scale manufacturing in the near<br />
future”;<br />
“The Review Panel found the Centre to be highly organised and fully functional. The<br />
Centre has successfully attracted high-quality staff including four Deputy Directors from<br />
overseas”;<br />
“Review Panel discussions with Centre postdoctoral fellows and students revealed<br />
the Centre to be a highly inclusive, supportive and open environment. Students felt<br />
encouraged and well supported in regard to equipment use and training. Students valued<br />
the Centre’s opportunities for industry and research experience as well as participation<br />
in international conferences”;<br />
“The Review Panel considered the Centre to be performing very well in meeting its<br />
performance targets in the area <strong>of</strong> international, national and regional links and<br />
networks”;<br />
“The Review Panel was very impressed with the level <strong>of</strong> support the Centre has attracted<br />
from Collaborating Organisations”;<br />
“The Review Panel concluded that the ARC Centre for Advanced Silicon Photovoltaics and<br />
Photonics has performed extremely well against all Key Result Areas and Performance<br />
measures”.<br />
8
STAFF LIST<br />
Director<br />
Stuart Ross Wenham, BE BSc PhD UNSW, FTS, SMIEEE, FIEAust<br />
(Scientia Pr<strong>of</strong>essor)<br />
Executive Research Director<br />
Martin Andrew Green, BE MEngSc Qld, PhD McMaster, FAA, FTS, FIEEE<br />
(Scientia Pr<strong>of</strong>essor and Federation Fellow)<br />
Deputy Directors<br />
Armin G. Aberle, BSc MSc PhD Freiburg, Dr habil Hannover, SMIEEE, MDPG<br />
Gavin Conibeer, BSc MSc London, PhD <strong>South</strong>ampton<br />
Jeffrey E. Cotter, BEE MSc PhD Delaware<br />
Jianhua Zhao, BE ME Nanjing I.T., PhD UNSW<br />
Business & Technology Manager<br />
Mark D. Silver, BE UNSW, GMQ AGSM<br />
Undergraduate Co-ordinator<br />
Jeffrey E. Cotter, BEE MSc PhD Delaware<br />
Postgraduate Co-ordinator<br />
Alistair B. Sproul, BSc PhD UNSW<br />
Associate Pr<strong>of</strong>essors<br />
Armin G. Aberle, BSc MSc PhD Freiburg, Dr habil Hannover, SMIEEE, MDPG<br />
Thorsten Trupke, PhD Karlsruhe<br />
Senior Lecturers<br />
Richard Corkish, BEng (Com) RMIT, PhD UNSW<br />
Jeffrey E. Cotter, BEE MSc PhD Delaware<br />
Alistair B. Sproul, BSc Sydney, PhD UNSW<br />
Ge<strong>of</strong>frey J. Stapleton, BE UNSW<br />
Muriel Watt, BSc N.E., PhD Murdoch<br />
Senior Fellow<br />
Jianhua Zhao, BE ME Nanjing I.T., PhD UNSW<br />
Research Fellows<br />
Robert Bardos, BSc (Hons) PhD Melbourne<br />
Patrick Campbell, BSc BE PhD UNSW<br />
Jiun-Hua “Allen” Guo, BSc MSc Nat.Taiwan U, PhD UNSW<br />
Aihua Wang, BE Nanjing IT, PhD UNSW<br />
Per Widenborg, BSc MSc Stockholm, PhD UNSW<br />
Post-Doctoral Fellows<br />
Gavin Conibeer, BSc MSc London, PhD <strong>South</strong>ampton<br />
Anita W.Y. Ho, BE PhD UNSW<br />
Dirk König, Dipl. Ing. (EE) Dr. rer. nat. (Ph) Chemnitz<br />
Kuo-Lung Lin, BE Tatung, MSc Liverpool<br />
Bryce Richards, BSc VUW, MEngSc PhD UNSW (until March 2005)<br />
Dengyuan Song, BE ME Hebei, PhD UNSW<br />
9
Adjunct Fellows<br />
Kylie Catchpole, BSc PhD ANU<br />
Didier Debuf, BE ME PhD UNSW<br />
Visiting Pr<strong>of</strong>essors/Fellows<br />
Roman Horacio Buitrago (UNL Santa Fe, Argentina)<br />
Rulong Chen (Suntech, China)<br />
Jean-François Guillemoles, PhD, Dr. Habil P&M Curie, Paris (IRDEP, Paris)<br />
Hans-Richard Karl Leiner (<strong>University</strong> College Konstanz, Germany)<br />
Hua Li (Suntech, China)<br />
Peter Würfel, Dipl. Physics PhD Karlsruhe (Karlsruhe <strong>University</strong>)<br />
Research Associates<br />
Peter Cousins BE PhD UNSW (until January 2005)<br />
Pr<strong>of</strong>essional Officers<br />
Malcolm Abbott, BE UNSW<br />
Ximing Dai, BSc Zhejiang, PhD UNSW<br />
Yidan Huang, BSc Anhui<br />
Tom Puzzer, BSc PhD UNSW<br />
Bernhard Vogl, BE Regensburg<br />
Guangchun Zhang, BE ME Shandong<br />
Research Assistants<br />
Frederick Bamberg<br />
Wan Lam Florence Chen<br />
Matt Edwards<br />
Peter Klement<br />
Mathieu Larose<br />
Esther Lee<br />
Ly Mai<br />
Shervin Motahar<br />
Edwin Pink<br />
Giuseppe Scardera<br />
Adriana Segovia<br />
Budi Tjahjono<br />
Stefan Vetter<br />
Vincent Vogt<br />
Juergen Weber<br />
Guoxiao Yao<br />
Taste <strong>of</strong> Research Summer Scholars<br />
Sin Von Chan<br />
Cuong Dang<br />
Alexander Fattal<br />
Brett Hallam<br />
Phillip Hamer<br />
James Rudd<br />
Samantha Wong<br />
Yicheng “Sherman” Zhou<br />
10
Technical Support Staff<br />
Gordon Bates, BA(IndDes) UTS<br />
Bruce Beilby, BSc Monash, PhD UNSW<br />
Kathryn Fisher, BE UNSW<br />
Jamie Green, BE UNSW<br />
Mark Griffi n, BE UNSW<br />
Kian Fong Chin, BE QUT, MEngSc QUT, MEngSc UNSW<br />
Robert Largent, AS USA<br />
Timothy Seary, InstrTradeAdvCertIndElectrElecTrade TAFE<br />
Nicholas Shaw, BE UNSW , PhD UNSW<br />
Jules Z.S. Yang, BSc Eastern China<br />
Alan Yee, BE UTS<br />
Casual Teaching Staff<br />
Anna Bruce, BE UNSW<br />
Robert Passey, BSc UNSW, BAppSc Murdoch, MSc Murdoch, PhD UNSW<br />
Casual Staff<br />
Anja Aberle<br />
Daniel Ball<br />
Roman Balla<br />
Andrew Barson<br />
Jonathan Bocking<br />
Nino Borojevic<br />
Damian Camporeale<br />
Susan Conyers<br />
Joel Courtney<br />
Dawei Di<br />
Brett Hallam<br />
Sue Anne Lit Hunter<br />
Zoe Jennings<br />
Anahita Karpour<br />
Daniel Kong<br />
Nicole Kuepper<br />
Daniel Lagos<br />
Tsu-Tsung Andrew Li<br />
Daniel Mansfi eld<br />
Gavin Pereira<br />
James Rudd<br />
Erdinc Saribatir<br />
Nancy Sharopeam<br />
Adeline Sugianto<br />
Fennalia Tan<br />
Owen The<br />
Belinda Thorne<br />
Roland Utama<br />
Dean Utian<br />
11
Student Administration Manager<br />
Trichelle Burns, BCom UNSW<br />
Financial Officer<br />
Julie Kwan<br />
Computer System Officer<br />
Lawrence Soria, AssocDipCompAppl W’gong<br />
Administrative Staff<br />
Lisa Cahill<br />
Danny Lin Chen, BSc UNSW<br />
Jenny Hansen<br />
Jill Lewis<br />
Ana Naumoska<br />
PhD Students<br />
Malcolm Abbott, BE UNSW<br />
Anna Bruce, BE UNSW<br />
Wan-Lam Florence Chen, BE UNSW<br />
Young Hyun Cho, BSc Hanyang U, MEngSc UNSW<br />
Matthew Edwards, BE UNSW<br />
Thipwan Fangsuwannarak, BEng V’kul, MEng C’korn<br />
Ziv Hameiri BSc BGU<br />
Daniel Inns, BE UNSW<br />
Chu-Wei “Scott” Jiang, BSEE Nat.Taiwan U, MEngSc Qld<br />
Oliver Kunz<br />
Ly Mai, BE UNSW<br />
Supriya Pillai, BTech M.G.U<br />
Giuseppe Scardera, BSc McGill , BASc Ottawa<br />
Avi Shalav, BSc MSc Massey<br />
Dengyuan Song, BSc MSc Hebei<br />
Ge<strong>of</strong>frey Stapleton, BE UNSW<br />
Axel Straub, BE ME Ulm, PhD UNSW (until July 2005)<br />
Mason Terry, BSEET Oregon IT, MSEE Washington<br />
Budi Tjahjono, BE UNSW<br />
Attachai Ueranantasun, BE KMUTT, T’land, MEngSc UNSW<br />
Timothy Walsh, BSc Melbourne, (Hons) ANU<br />
Guoxiao “James” Yao, BE UT Zhegiang, MS Sweden<br />
Masters Students<br />
Kathryn Fisher, BE UNSW<br />
Bijaya Paudyal<br />
Lei “Adrian” Shi<br />
Bernhard Vogl, BE Regensburg<br />
Juergen Weber<br />
12
4.1 INTRODUCTION TO RESEARCH<br />
Photovoltaics, the direct conversion <strong>of</strong> sunlight to electricity using solar cells, is recognised<br />
as one <strong>of</strong> the most promising options for a sustainable energy future. The ARC Photovoltaics<br />
Centre <strong>of</strong> Excellence commenced in mid-2003, drawing together previous disparate strands<br />
<strong>of</strong> work, supported under a variety <strong>of</strong> programs, into a coherent whole addressing the key<br />
challenges facing photovoltaics, as well as “spin-<strong>of</strong>f” applications in microelectronics and<br />
optoelectronics.<br />
The Centre’s photovoltaics research is divided into three interlinked strands addressing nearterm,<br />
medium-term and long-term needs, respectively. The present photovoltaic market is<br />
dominated by “fi rst-generation” product based on silicon wafers, either single-crystalline as in<br />
microelectronics, or a lower-grade multicrystalline wafer (Fig. 4.1.1). This market dominance<br />
is likely to continue for at least the next decade. First-generation production volume is growing<br />
rapidly, with the technological emphasis upon streamlining manufacturing to reduce costs<br />
while, at the same time, improving the energy conversion effi ciency <strong>of</strong> the product. Other key<br />
issues involve reducing the manufacturing spread on multicrystalline wafer lines caused by<br />
variability in wafer quality (typically 20% spread in cell output) and elimination <strong>of</strong> the effects <strong>of</strong><br />
boron-oxygen defects in both types <strong>of</strong> wafers. These defects become active under illumination<br />
and reduce the performance <strong>of</strong> most commercial modules by about 3%. They also constrain<br />
the specifi cation <strong>of</strong> the starting silicon wafer, restricting cell design possibilities. Also important<br />
is the reduction <strong>of</strong> the thickness <strong>of</strong> the starting silicon wafer without losing performance, to<br />
save on material use, and the development <strong>of</strong> low-cost techniques for reducing refl ection<br />
from multicrystalline cells.<br />
Figure 4.1.1:<br />
“First-generation” waferbased<br />
technology (BP<br />
Solar Saturn Module,<br />
the photovoltaic product<br />
manufactured in the highest<br />
volume by BP in Europe,<br />
using UNSW buried-contact<br />
technology).<br />
The Centre’s fi rst-generation research is focussed on these key issues. Major emphasis is<br />
upon the “buried-contact” solar cell, originally developed by Centre researchers, the fi rst <strong>of</strong><br />
the modern high-effi ciency cell technologies to be successfully commercialised (Fig. 4.1.1).<br />
Centre research seeks improvements to these devices to increase effi ciency, particularly<br />
for devices fabricated on thin wafers. Of key interest is the development <strong>of</strong> buried-contact<br />
sequences for substrates doped with phosphorus, rather than boron, to avoid the boronoxygen<br />
defect problem previously noted.<br />
Wafers are expensive and need quite elaborate and expensive encapsulation, since they<br />
are brittle and also thermally mismatched to the glass coversheet, making fi rst-generation<br />
technology inherently material-intensive. To avoid the associated cost penalties, several<br />
companies worldwide are commercialising “second-generation” thin-fi lm cell technology<br />
based on depositing thin layers <strong>of</strong> the photoactive material onto supporting substrates or<br />
superstrates, usually sheets <strong>of</strong> glass (Fig. 4.1.2). Although materials other than silicon are<br />
<strong>of</strong> interest for these fi lms, silicon avoids problems that can arise with these more complex<br />
compounds due to stability, manufacturability, moisture sensitivity, toxicity and resource<br />
availability issues. CSG Solar, a partner in the Centre, has commercialised an approach<br />
pioneered by Centre researchers that is unique in that it is based on the use <strong>of</strong> the same high<br />
quality silicon used for fi rst-generation production, but deposited as a thin layer onto glass.<br />
13
Figure 4.1.2:<br />
Example <strong>of</strong> “second-generation”<br />
thin-fi lm technology (module<br />
fabricated on CSG Solar’s pilotline<br />
in Sydney, based on thin-fi lms<br />
<strong>of</strong> polycrystalline silicon deposited<br />
onto glass, again UNSWpioneered<br />
technology).<br />
As well as its collaborative activities with CSG Solar, the Centre currently maintains a largely<br />
independent program addressing alternative solutions to those adopted by CSG Solar for<br />
producing high-performance “silicon-on-glass” solar cells. The main emphasis <strong>of</strong> this program<br />
is on improving the quality <strong>of</strong> the silicon fi lms by using thin crystallographic templates (formed<br />
by aluminium induced crystallisation) and on the development <strong>of</strong> lower-cost deposition<br />
approaches (evaporation rather than PECVD)<br />
At the present time, second-generation thin-fi lms are entering the market in increasing<br />
quantities. The successful commercialisation <strong>of</strong> thin-fi lm product will lead to a completely<br />
different manufacturing cost structure. However, costs again are expected increasingly to<br />
become dominated by material cost as production increases, for example, by the cost <strong>of</strong> the<br />
glass sheet on which the cells are deposited.<br />
More power from a given investment in material is possible by increasing energy-conversion<br />
effi ciency. This leads to the notion <strong>of</strong> a third-generation <strong>of</strong> solar cell distinguished by the fact<br />
that it is both high-effi ciency and thin-fi lm. To illustrate the cost leverage provided by effi ciency,<br />
Fig. 4.1.3 shows the relative cost structures <strong>of</strong> the three generations being studied by the<br />
Centre. This fi gure plots effi ciency against manufacturing cost, expressed in US$/square<br />
metre. First-generation technology has relatively high production cost per unit area and<br />
moderate likely effi ciencies (10-20%). The dotted lines in Fig. 4.1.3 show the corresponding<br />
cost/watt, the market metric. Values as low as US$1/watt may be feasible by increasing the<br />
effi ciency while reducing manufacturing cost, but this is the likely limit <strong>of</strong> the fi rst-generation<br />
approach (present manufacturing costs for large manufacturers are as low as US$2-3/<br />
watt).<br />
Second-generation thin-fi lm technology has a different cost structure as evident from this<br />
fi gure. Production costs per unit area are a lot lower, since silicon wafers are a lot more<br />
expensive than glass sheet. However, likely energy-conversion effi ciencies are lower (5-15%).<br />
100<br />
US$0.10/W<br />
US$0.20/W<br />
US$0.50/W<br />
Efficiency, %<br />
80<br />
60<br />
40<br />
20<br />
Thermodynamic<br />
Limit<br />
US$1.00/W<br />
Present limit<br />
US$3.50/W<br />
Figure 4.1.3: Effi ciency<br />
and cost projections for<br />
fi rst-, second- and thirdgeneration<br />
photovoltaic<br />
technology (wafers, thinfi<br />
lms and advanced thinfi<br />
lms, respectively).<br />
0 100 200 300 400 500<br />
Cost, US$/m2<br />
14
Overall, this trade-<strong>of</strong>f produces costs/watt estimated as about 2 to 3 times lower than<br />
those <strong>of</strong> the wafer product, in large production volumes.<br />
The third-generation is specifi ed as a thin-fi lm technology, which therefore has manufacturing<br />
costs similar to second-generation, but is based on operating principles that do not constrain<br />
effi ciency to the same limits as conventional cells (31% for non-concentrated sunlight for<br />
these). Unconstrained thermodynamic limits for solar conversion are much higher (74% for<br />
non-concentrated light), giving an indication <strong>of</strong> the potential for improvement. If a reasonable<br />
fraction <strong>of</strong> this potential can be realised, Fig. 4.1.3 suggests that third-generation costs could<br />
be lower than second-generation by another factor <strong>of</strong> 2 to 3.<br />
Of the third-generation options surveyed by Centre researchers, “all-silicon” tandem cells<br />
based on bandgap-engineering using nanostructures was selected as the most promising for<br />
implementation in the Centre’s timescale (Fig. 4.1.4). This involves the engineering <strong>of</strong> a new<br />
class <strong>of</strong> mixed-phase semiconductor material based on partly-ordered silicon quantum-dots in<br />
an insulating amorphous matrix. Photon up- and down-conversion as a way <strong>of</strong> “supercharging”<br />
the performance <strong>of</strong> relatively standard cells forms a second line <strong>of</strong> research. A third is the<br />
investigation <strong>of</strong> schemes for implementing hot-carrier cells.<br />
Figure 4.1.4: Conceptual design <strong>of</strong> an all-silicon<br />
tandem cell based on Si-SiO 2 (or Si-Si 3 N 4 or Si-SiC)<br />
quantum dot superlattices. Two solar cells <strong>of</strong> different<br />
bandgap controlled by quantum dot size are stacked<br />
on top <strong>of</strong> a third cell made from bulk silicon.<br />
15
The fourth “spin-<strong>of</strong>f” Centre strand <strong>of</strong> silicon photonics draws upon elements <strong>of</strong> all three <strong>of</strong><br />
the photovoltaic strands. The aim is to develop silicon light-emitting-diodes (LEDs) that can be<br />
integrated into silicon integrated circuits, with their output capable <strong>of</strong> modulation at gigahertz<br />
frequencies (Fig. 4.1.5). This strand is closely related to the high-effi ciency cell work in our fi rstgeneration<br />
strand <strong>of</strong> research, since the associated Centre researchers have demonstrated<br />
the highest-performing LEDs based on bulk silicon by a very clear margin, internationally. The<br />
work also draws upon the expertise in the second-generation strand since thin fi lms <strong>of</strong> silicon<br />
are integral to the silicon-on-insulator (SOI) wafers used in this work. “Light-trapping” is also<br />
very relevant to both strands. The common interest in quantum-confi nement in silicon also<br />
provides strong links to the third-generation strand. Another objective being pursued in the<br />
silicon photonic strand is the demonstration <strong>of</strong> the fi rst silicon laser. This is the “holy grail” <strong>of</strong><br />
the silicon photonics fi eld.<br />
Figure 4.1.5:<br />
One <strong>of</strong> the concepts for an<br />
integrated silicon light-emitter/<br />
modulator presently being<br />
investigated experimentally.<br />
Modulation is sought using the<br />
quantum-confi ned Stark effect<br />
in thin silicon-on-insulator (SOI)<br />
layers.<br />
16
4.2 FACULTIES AND INFRASTRUCTURE<br />
The ARC Centre <strong>of</strong> Excellence for Advanced Silicon Photovoltaics and Photonics is located at<br />
the Kensington campus <strong>of</strong> the <strong>University</strong> <strong>of</strong> <strong>New</strong> <strong>South</strong> <strong>Wales</strong>, about 6 km from the heart <strong>of</strong><br />
Sydney and close to its world famous beaches including Bondi, Coogee and Maroubra (Fig.<br />
4.2.1).<br />
Sydney CBD<br />
Kensington<br />
<strong>University</strong> <strong>of</strong><br />
<strong>New</strong> <strong>South</strong> <strong>Wales</strong><br />
Coogee<br />
Bondi<br />
Sydney Airport<br />
Botany<br />
CSG Solar Pty Ltd<br />
Botany Laboratory<br />
Maroubra<br />
Figure 4.2.1: Centre <strong>of</strong><br />
Excellence location in<br />
Sydney.<br />
Organisationally, the Centre <strong>of</strong> Excellence is located within the Centre for Photovoltaic<br />
Engineering within the Faculty <strong>of</strong> Engineering. The Centre <strong>of</strong> Excellence has a large range <strong>of</strong><br />
laboratory facilities (Fig. 4.2.2). These include the Bulk Silicon Research Laboratories, the<br />
Device Characterisation Laboratory, the Optoelectronic Research Laboratory, the Thin-Film<br />
Cell Laboratory, the Industry Collaborative Laboratory and Inkjet Processing Laboratory. Off<br />
campus the Centre has a Thin-Film Cleanroom facility at Botany, 5km south-west <strong>of</strong> the main<br />
campus. Another important resource is the Semiconductor Nan<strong>of</strong>abrication Facility jointly<br />
operated by the Faculty <strong>of</strong> Science and the Faculty <strong>of</strong> Engineering.<br />
A new <strong>of</strong>fi ce accommodation area on the fourth fl oor <strong>of</strong> the Electrical Engineering building has<br />
consolidated 12 <strong>of</strong>fi ces for up to 19 academic/senior researchers around 28 open plan desk<br />
spaces with supporting meeting rooms and tea room. Whilst not only providing fi rst class<br />
accommodation, the consolidation provides space for personnel previously accommodated in<br />
an adjacent building and permits the conversion <strong>of</strong> LG <strong>of</strong>fi ce space to a new laser laboratory<br />
for the Optoelectronic Research Laboratory.<br />
Additional equipment commonly used for solar cell work is also available on the <strong>University</strong><br />
campus. Included in this category are TEM/SEM electron and focussed ion beam (FIB)<br />
microscopes, X-ray diffraction, Raman spectroscopy, AFM and surface analysis equipment.<br />
TEM, ellipsometry and high-precision I-V equipment are also regularly accessed at Sydney<br />
<strong>University</strong>.<br />
17
4<br />
3<br />
2<br />
1<br />
G<br />
During 2005 the development and acquisition <strong>of</strong> laboratory equipment and infrastructure<br />
continued with specifi c details about signifi cant additions found under the laboratory headings<br />
which follow.<br />
4<br />
3<br />
2<br />
1<br />
G<br />
4<br />
3<br />
2<br />
1<br />
Office ARC Centre <strong>of</strong> Excellence for Advanced<br />
Silicon Photovoltaics and Photonics<br />
Buried Contact Cell Lab<br />
High Efficiency Call Lab<br />
Operations Workshop<br />
Office Accommodation<br />
Student Computer Lab<br />
G<br />
Industry Collaboation Lab.<br />
Thin-Film Cell Lab<br />
Offices for School <strong>of</strong> Photovoltaic and<br />
Renewable Energy Engineering<br />
InkJet Processing Laboratory<br />
Student Project/Seminar Room<br />
Centre for Photovoltaic Engineering Office<br />
Device Characterisation and Laser Lab<br />
Figure 4.2.2: Layout <strong>of</strong> laboratories and other facilities within the Electrical Engineering<br />
building, Kensington.<br />
The Centre <strong>of</strong> Excellence has two computer networks. One is used for research and general<br />
administrative purposes and the other by students enrolled in the Undergraduate Degree<br />
Courses in Photovoltaics and Renewable Energy. In 2005 the Research and Administrative<br />
network grew nearly 30% to consist <strong>of</strong> 3 servers, 3 Intranet servers, two Internet webservers<br />
and over 90 workstations. An additional 16 computers are dedicated to the<br />
computer control <strong>of</strong> laboratory and other equipment. The computer resources are used<br />
for modelling/simulations, equipment control, general administrative purposes, document<br />
control, laboratory support, Internet access, and maintaining the Centre’s presence on the<br />
Internet. Each postgraduate research student is allocated a dedicated personal computer in<br />
addition to access to existing shared computers.<br />
The Centre also has a computer controlled SCADA system and PLC network for equipment<br />
control and monitoring <strong>of</strong> the research laboratories and related infrastructure. The Centre is<br />
soon to install a Linux (Beowulf) cluster with an estimated computational power <strong>of</strong> about 100<br />
Gigafl ops for Density-Functional-Hartree-Fock and molecular dynamical computations.<br />
The Student Computer network comprises <strong>of</strong> two computer laboratories <strong>of</strong> total 83 m2 with<br />
4 servers, a UNSW custom CSE computer based router and 26 workstations. Students<br />
enrolled in the Undergraduate Degree Courses use these computers for computer-related<br />
coursework and Internet access. There is also an Internet capable web-server that gathers<br />
and displays data collected from solar arrays on the ro<strong>of</strong> <strong>of</strong> the building. These data can be<br />
viewed using a web browser and can be made available for Internet access.<br />
The Laboratory Development and Operations Team develops and maintains core Centre and<br />
laboratory facilities. During 2005, the team, under the leadership <strong>of</strong> Mark Silver, comprised<br />
an additional 5 equivalent full-time and 11 casual employees, including: electrical and industrial<br />
design engineers, a computer/network manager, electronic and computer technicians, and<br />
administrative staff.<br />
18
Bulk Silicon Research Laboratories<br />
The Centre houses the largest and most sophisticated bulk silicon solar cell research facility<br />
in Australia, incorporating both the High Effi ciency/LED and Buried Contact Cell Laboratories.<br />
Total laboratory processing space <strong>of</strong> over 500 m 2 is located in the Electrical Engineering<br />
Building and is serviced with fi ltered and conditioned air, appropriate cooling water, processing<br />
gas, de-ionized water supply, chemical fume cupboards and local exhausts. There is an<br />
additional 540 m 2 area immediately adjacent to the laboratories for the accommodation <strong>of</strong><br />
staff, research students, school <strong>of</strong>fi ce and laboratory support facilities. Off site, areas totalling<br />
200 m 2 are used for the storage <strong>of</strong> chemicals and equipment spare parts.<br />
The laboratories are furnished with a range <strong>of</strong> processing and characterisation equipment<br />
including 26 diffusion furnaces, 5 vacuum evaporation deposition systems, a laser-scribing<br />
machine, rapid thermal annealer, four-point sheet-resistivity probe, quartz tube washer, silver,<br />
nickel and copper plating units, visible wavelength microscopes, 3 wafer mask aligners, spinon<br />
diffusion system, automated photoresist dual-track coater, photoresist spinner, electron<br />
beam deposition system, TiO 2 spray deposition, belt furnace, manual screen printer and a<br />
laboratory system control and data acquisition monitoring system.<br />
Figure 4.2.3: CNC Laser Scribe Tool<br />
The laser scribe tool, shown in Figure 4.2.3, has a 20 watt Nd:YAG laser for infrared operation<br />
(1064 nm) and an optional frequency doubler for green operation (532 nm). The work stage<br />
is CNC controlled allowing 1 micron positional accuracy and table speeds approaching 25<br />
cm/second across an area <strong>of</strong> 15 cm by 15 cm. The tool is used primarily for Buried Contact<br />
solar cell fabrication, cutting 35-micron wide laser grooves as deep as 100 microns into<br />
silicon wafers. It can also be used to cut other suitable materials, such as stainless steel.<br />
Device Characterisation Laboratory<br />
This laboratory is located on the lower ground fl oor <strong>of</strong> the Electrical Engineering Building.<br />
Associated with it is the Optoelectronic Research Laboratory, reception area, seminar room,<br />
<strong>of</strong>fi ces for Centre staff interacting with the public and industry, including the Business &<br />
Technology Manager, External Relations Manager and Design Assistance Division Manager<br />
and computer workstations for the device modelling activities <strong>of</strong> the Centre.<br />
The Device Characterisation Laboratory houses characterisation equipment including “Dark<br />
Star”, the Centre’s station for temperature controlled dark current-voltage measurements,<br />
the Centre’s Fourier-transform infrared spectroscopy system (FTIR), admittance spectroscopy<br />
19
system, microwave carrier lifetime system, ellipsometer, Sinton photoconductance lifetime<br />
equipment, wafer probing station for SOI LED work, open circuit voltage versus illumination<br />
measurement system (Suns-Voc), infrared microscope and equipment for spectral response<br />
and related optical measurements.<br />
In 2005 a new Fourier-transform infrared spectroscopy system (FTIR) was purchased for the<br />
laboratory along with a new 4 point resistivity probing station. Planning is also underway to<br />
provide additional effective workspace by removing the viewing wall partition and redesigning<br />
the associated work bench area.<br />
Optoelectronic Research Laboratory<br />
This facility has two optical benches and several visible and near-infrared semiconductor<br />
diode lasers along with other optical and electrical instrumentation. The facility is used for<br />
photoluminescence and electroluminescence measurements in the visible and infrared<br />
spectral range up to wavelengths <strong>of</strong> 2500nm, photoluminescence (PL) excitation spectroscopy,<br />
luminescence experiments with simultaneous two-colour illumination, quasi steady state<br />
photoluminescence lifetime measurements and Sinton lifetime testing with the conventional<br />
fl ash-light replaced by a high-power light emitting diode array. An area separate from the<br />
Device Characterisation Laboratory was necessary in order to meet stringent standards for<br />
avoidance <strong>of</strong> laser eye and skin exposure for users and others. It shares cryogenic cooling<br />
equipment with the Device Characterisation Laboratory.<br />
In 2005 the world’s fi rst photoluminescence imaging system was installed along with a<br />
silicon CCD camera (for sensitive PL measurements) and a second spectroscopic PL system.<br />
Demand for additional laboratory space for the Optoelectronic Research Laboratory has<br />
necessitated the conversion <strong>of</strong> adjacent lower ground fl oor <strong>of</strong>fi ce space into a future laser<br />
laboratory to accommodate experiments on an additional 4 optical benches.<br />
Figure 4.2.4: Optical characterisation bench.<br />
Thin-Film Cell Laboratory<br />
This 40 m2 laboratory is equipped with a range <strong>of</strong> equipment for thin-fi lm deposition and<br />
patterning, including a plasma-enhanced chemical vapour deposition (PECVD) system, a<br />
sputtering system, a reactive ion etcher (RIE), a resistively heated vacuum evaporator, and<br />
an optical microscope with digital image acquisition system. Also used by the laboratory is an<br />
electron-beam vacuum evaporator for silicon which is physically located within the Bulk Silicon<br />
Research Laboratory. This Si evaporator is also equipped with an ionizer unit and a sample<br />
heater, enabling fast-rate Si homoepitaxy at temperatures <strong>of</strong> about 500-600 °C by means <strong>of</strong><br />
ion-assisted deposition (IAD). Other equipment <strong>of</strong> use in thin-fi lm projects is located within the<br />
Semiconductor Nan<strong>of</strong>abrication Facility.<br />
20
The PECVD system, shown in Fig.4.2.5, has a 40x20 cm 2 process platen and can handle<br />
large-area silicon wafers as well as smaller pieces. Two types <strong>of</strong> plasma excitation (remote<br />
microwave and direct RF) are available. The machine is used for the low-temperature<br />
deposition <strong>of</strong> thin dielectric fi lms (silicon nitride, silicon dioxide) and <strong>of</strong> amorphous silicon. The<br />
dual-cylinder, remote microwave plasma source produces excellent-quality silicon nitride and<br />
silicon dioxide fi lms, with precise control over the stoichiometry at temperatures up to 500°C.<br />
Amorphous and microcrystalline silicon fi lms can also be deposited in the system.<br />
In 2005 a plasma etching machine capable <strong>of</strong> processing sample areas larger than the<br />
existing RIE was added. The sputtering machine was upgraded with mass fl ow controllers to<br />
improve process repeatability.<br />
Figure 4.2.5: Remote plasma PECVD machine.<br />
Industry Collaborative Laboratory<br />
This 120 m 2 laboratory houses equipment needed for many <strong>of</strong> the industry-collaborative<br />
research activities in the Buried-Contact Solar Cell group. The laboratory was refurbished<br />
in 1999 and several new pieces <strong>of</strong> infrastructure have been acquired or constructed since,<br />
including: a belt furnace; a state <strong>of</strong> the art laser micromachining tool; a new PECVD deposition<br />
system (located in the adjacent thin fi lm solar cell laboratory); and a TiO2 spray deposition<br />
station; a high temperature semiconductor muffl e furnace and manual screen printer. In 2005<br />
a fully automatic production scale screen printer was commissioned in the laboratory.<br />
Inkjet Processing Development Lab<br />
This laboratory, new for 2005, houses the Centre’s<br />
inkjet printing development system. The laboratory<br />
is used to develop solar processing “inks” (chemical<br />
solutions) and for printing them onto a solar wafer<br />
under computer control. The aim is to develop low cost<br />
processing techniques for creating fi ne structures in<br />
solar cells. It is anticipated that the inkjet printing is<br />
capable <strong>of</strong> replacing processes such as laser scribing<br />
and photolithography for forming fi ne patterns for<br />
contacting but at a cost <strong>of</strong> at least 10 times cheaper.<br />
Figure 4.2.6: Inkjet development system<br />
21
Semiconductor Nan<strong>of</strong>abrication Facility<br />
The Centre also owns equipment within, and has access to, the Semiconductor Nan<strong>of</strong>abrication<br />
Facility (SNF) at the <strong>University</strong>. This is a joint facility shared by the Faculties <strong>of</strong> Science and<br />
Engineering and houses a microelectronics laboratory and a nan<strong>of</strong>abrication laboratory<br />
for e beam lithography. A collaborative project continues with the SNF to e-beam pattern<br />
a photonic crystal on the surface <strong>of</strong> the Centre’s SOI LED structures for enhanced optical<br />
emission. The SNF provides an Australian capability for the fabrication <strong>of</strong> advanced nanoscale<br />
semiconductor devices and their integration with microelectronics. SNF research projects<br />
form an integrated effort to fabricate innovative semiconductor nanostructures using the<br />
latest techniques <strong>of</strong> electron beam patterning and scanning probe manipulation. A major<br />
applied objective <strong>of</strong> the SNF facility is the development <strong>of</strong> a prototype silicon nuclear spin<br />
quantum computer.<br />
Thin-Film CleanRoom facility in Bay Street, Botany<br />
During 2003 and 2004 the Centre added a 120 m 2 cleanroom facility in Bay Street, Botany to<br />
its infrastructure, greatly improving its experimental capabilities in the area <strong>of</strong> thin-fi lms. This<br />
cleanroom is equipped with several fume cupboards, two tube furnaces, an electron-beam<br />
vacuum evaporator, a thermal vacuum evaporator, a glass washing machine, a rapid thermal<br />
processing (RTP) machine, and a 5-chamber cluster tool. The cluster tool presently features<br />
four plasma-enhanced chemical vapour deposition (PECVD) chambers and one lamp-heated<br />
vacuum annealing chamber. The PECVD chambers enable the low-temperature deposition <strong>of</strong><br />
dielectric fi lms (silicon oxide, silicon nitride, etc) and amorphous silicon fi lms (either n- or p-<br />
doped or undoped). Furthermore, samples can be hydrogenated by PECVD using a hydrogen<br />
plasma at substrate temperatures <strong>of</strong> up to 480°C. During 2004 the Centre purchased a<br />
low-pressure chemical vapour deposition (LPCVD) system, an infrared NdYAG laser scriber<br />
and a box furnace for sample annealing. The LPCVD system is capable <strong>of</strong> depositing doped<br />
crystalline silicon on glass and with its additional remote plasma source is currently engaged<br />
in hydrogenation work. The NdYAG laser is used for scribing silicon fi lms and other suitable<br />
metal and dielectric materials. In 2005 a new state <strong>of</strong> the art sputtering machine was<br />
ordered and plans are underway for the installation <strong>of</strong> an additional fume cupboard.<br />
Figure 4.2.7: Bay St Clean Room Cluster Tool Controller.<br />
22
Nancy Sharopean<br />
Mark Silver<br />
and<br />
Nick Shaw<br />
Per Widenborg and Anya Aberle<br />
Jill Lewis<br />
Anita Wing Yi Ho<br />
Kian Chin<br />
Lawrence Soria<br />
Alan Yee<br />
Jules Yang<br />
Gordon Bates<br />
23
4.3 FIRST GENERATION: WAFER-BASED PROJECTS<br />
4.3.1 High Efficiency Cells<br />
<strong>University</strong> Staff:<br />
A/Pr<strong>of</strong>. Jianhua Zhao (group leader)<br />
Pr<strong>of</strong>. Martin Green<br />
Pr<strong>of</strong>. Stuart Wenham<br />
Research Fellows: Dr. Aihua Wang<br />
Research Assistant: Guangchun Zhang<br />
Technical Staff: Jules Yang<br />
During 2005, a cell effi ciency equalling the world record for n-type silicon cells has been<br />
demonstrated by a new rear emitter PERT cell structure, which was developed by the high<br />
effi ciency group at the Centre. The previous record was set more than a decade ago by<br />
Stanford <strong>University</strong>. This new cell structure is expected to produce further improvements<br />
in cell effi ciency and, importantly, is also suitable for low cost CZ silicon substrates. This<br />
research work is reported below.<br />
4.3.1.1 Rear Emitter PERT Cell Structure on N-type Silicon Substrates<br />
In 2004, a new rear-boron-emitter PERT (passivated emitter, rear totally-diffused) cell<br />
structure on n-type silicon substrates was developed by the high effi ciency cell group in the<br />
Centre. During the initial experiments, these devices demonstrated high cell effi ciencies<br />
close to 22%, and very stable cell performance under one-sun illumination and after long<br />
term storage in nitrogen. Figure 4.3.1.1 shows the structure <strong>of</strong> these n-type rear emitter<br />
PERT (re-PERT) cells.<br />
finger<br />
“inverted” pyramids<br />
Figure 4.3.1.1<br />
Rear boron emitter PERT cell<br />
structure on n-type silicon<br />
substrates<br />
double layer<br />
antireflection<br />
coating<br />
n-silicon<br />
rear contact<br />
oxide<br />
High effi ciency processing methods, similar to those used for PERL (passivated emitter, rear<br />
locally-diffused) cells, are used for processing these n-type re-PERT cells. These processing<br />
methods include lithographically defi ned regular inverted pyramids surface for excellent light<br />
trapping, TCA thermal oxide growth and aluminium anneal for high quality surface passivation,<br />
boron tribromide liquid source boron diffusion producing very low surface damage, lift-<strong>of</strong>f metal<br />
scheme combined with electro-chemical silver plating for minimum contact recombination<br />
and very little metal shading and resistance loss, and double layer antirefl ection coatings to<br />
further minimise the surface refl ection loss.<br />
25
Since the rear emitter boron diffusion was over the entire rear surface, the rear emitter cells<br />
have to be scribed into individual cells before measurement, in contrast to standard research<br />
PERL cells which are <strong>of</strong>ten measured on the whole silicon wafers without scribing. Hence,<br />
the performances <strong>of</strong> these scribed rear emitter cells are lower when compared to those unscribed<br />
small-area research cells, due to the extra edge recombination loss and increased<br />
cell area, which is non-optimal for the present silver plated metallisation method. By way<br />
<strong>of</strong> comparison, the best scribed PERL cell on p-type substrates <strong>of</strong> the same 22 cm 2 area<br />
demonstrated 23.7% effi ciency [J. Zhao et al, “20 000 PERL Silicon Cells for the ‘1996 World<br />
Solar Challenge’ Solar Car Race”, Progress in Photovoltaics, Vol. 5, pp. 269, 1997], while the<br />
best small 4 cm 2 unscribed cell had a signifi cantly higher effi ciency <strong>of</strong> 24.7%.<br />
Small-area 4 cm 2 rear emitter cells were initially investigated. However, due to the relatively<br />
larger edge scribing damage contributing to increased total recombination, these small cells<br />
had relatively low Voc around 680 mV. Hence, the present research concentrated upon<br />
large-area rear emitter cells <strong>of</strong> 22 cm 2 area in an effort to reduce the edge recombination<br />
loss, resulting in high Voc <strong>of</strong> over 700 mV.<br />
4.3.1.2 Calculated Device Performances with the<br />
Device Simulation Program PC-1D<br />
The device simulation program PC-1D was used to calculate and analyse the cell performance.<br />
The cell Voc and the IQE (internal quantum effi ciency) were the main data to be fi tted in<br />
these calculations. After comparing to the experimental device parameters, the bulk and the<br />
surface recombination parameters were selected as following:<br />
Bulk minority carrier lifetime:<br />
1.1 ms<br />
Front surface recombination velocity: 300 cm/s<br />
Rear surface recombination velocity: 400 cm/s<br />
The substrate resistivity was 0.9 Ω-cm. However, the substrate resistivity was found noncritical<br />
for the calculated device performance if the recombination parameters were not<br />
altered. However, normally the higher substrate resistivity is, the higher the minority carrier<br />
lifetime. In this sense, the substrate resistivity may be very critical to the performance <strong>of</strong><br />
the fabricated devices, particularly for these thin cells. Figure 4.3.1.2 shows the calculated<br />
variation <strong>of</strong> Jsc, Voc and cell effi ciency with the substrate thickness, simulated by PC-1D.<br />
(a) (b) (c)<br />
37.9<br />
716<br />
37.8<br />
714<br />
712<br />
37.7<br />
710<br />
37.6<br />
708<br />
37.5<br />
706<br />
37.4<br />
704<br />
37.3<br />
702<br />
700<br />
37.2<br />
698<br />
37.1<br />
696<br />
37<br />
694<br />
500<br />
400 300 200 100<br />
Cell Thickness, um<br />
0<br />
Jsc, mA/cm2<br />
500<br />
400 300 200 100<br />
Cell Thickness, um<br />
Figure 4.3.1.2: PC-1D simulated variation <strong>of</strong> (a) J sc , (b) Voc, and (c) cell effi ciency with the<br />
substrate thickness. Reducing the substrate thickness from 400 µm to 200 µm is predicted<br />
to signifi cantly increase all the three parameters.<br />
0<br />
Voc, mV<br />
500<br />
400 300 200 100<br />
Cell Thickness, um<br />
0<br />
22.4<br />
22.3<br />
22.2<br />
22.1<br />
22<br />
21.9<br />
21.8<br />
21.7<br />
21.6<br />
21.5<br />
21.4<br />
Efficiency, %<br />
26
According to the calculated data in Figure 4.3.1.2, when the substrate thickness is reduced<br />
from 400 µm to 200 µm, both Voc and Jsc are expected to increase signifi cantly, due to<br />
the higher minority carrier density at the rear emitter and an increased spectral response.<br />
However, when the substrate thickness is reduced to less than 200 µm, the quantity <strong>of</strong> silicon<br />
material becomes insuffi cient to absorb all <strong>of</strong> the incident light, resulting in a reduced IQE<br />
and current density. Hence, the effi ciency reaches a plateau at 200 µm thickness, while<br />
the calculated Voc keeps increasing and the Jsc starts to reduce for thinner substrates. In<br />
experimental devices, the cell Voc would not increase as quickly as the calculation suggested<br />
with reducing substrate thickness, since the surface passivation treatment may become<br />
more and more diffi cult to implement for extremely thin substrates.<br />
4.3.1.3 Experimental Performance <strong>of</strong> N-type Rear Emitter PERT Cells<br />
During 2005, experimental efforts were concentrated on reducing the thickness <strong>of</strong> these<br />
n-type re-PERT cells. Different n-type substrate materials with higher resistivity were also<br />
tested, even though the effect <strong>of</strong> this parameter can not be simulated clearly using PC-1D.<br />
These experiments have demonstrated further signifi cant improvements in the performance<br />
<strong>of</strong> these re-PERT cells.<br />
In 2005, two successful batches <strong>of</strong> re-PERT cells, with structures as shown in Figure 4.3.1.1,<br />
were fabricated using phosphorus doped n-type FZ substrates <strong>of</strong> 1.5 Ω-cm resistivity, 270-<br />
µm and 170-µm thickness, supplied by Wafer World. These wafers had to be etched to<br />
reduce their thickness and then mechanically polished in-house before cell processing, since<br />
there is no suitable supplier <strong>of</strong> such thin, polished silicon substrates. Table 4.3.1.1 lists the<br />
performance <strong>of</strong> these cells together with the very fi rst batch <strong>of</strong> the re-PERT cells, all measured<br />
at Sandia National Laboratories under the global AM1.5 spectrum (100 mW/cm 2 ) at 25ºC.<br />
Table 4.3.1.1. The performance <strong>of</strong> three batches <strong>of</strong> scribed 22 cm 2 large area n-type re-<br />
PERT cells as measured at Sandia National Laboratories under 100 mW/cm 2 , AM1.5 global<br />
spectrum at 25ºC (the fi rst cell, Wnrj1-1a, was fabricated during 2004).<br />
Cell ID Thickness Jsc Voc FF Effi c.<br />
(µm) (mA/cm 2 ) (mV) (%)<br />
Wnrj1-1a 400 39.1 0.696 0.774 21.0<br />
Wnrj4-3b 270 40.4 0.705 0.785 22.3<br />
Wnrj7-1a 170 40.1 0.706 0.791 22.4<br />
Wnrj7-2a 170 40.1 0.702 0.805 22.7<br />
Wnrj7-2b 170 40.1 0.701 0.802 22.6<br />
High carrier lifetime and thin substrates have helped to achieve effi cient carrier collection<br />
to the cell rear emitter, and hence improved the short circuit current density and the cell<br />
effi ciency. One <strong>of</strong> these rear emitter n-type cells demonstrated an excellent Voc <strong>of</strong> 706 mV,<br />
which is as high as the best p-type PERL cells. The current density is about 1 mA/cm 2 below<br />
the p-type PERL cells, since there is a slight loss during the long transportation path <strong>of</strong> carriers<br />
to the rear emitter and an increased edge recombination loss. High effi ciencies obtained<br />
with relatively simple processing requirements could make these cells highly attractive for<br />
commercial use. Since high carrier lifetimes <strong>of</strong> over 5 ms have been demonstrated from<br />
n-type SEH CZ substrates [Jianhua Zhao, et al, “High Effi ciency PERT Cells on N-Type Silicon<br />
Substrates”, 29th IEEE Photovoltaic Specialist Conference, <strong>New</strong> Orleans, pp.218-221, 2002],<br />
the n-type rear emitter cells fabricated using such inexpensive materials are expected to give<br />
similar performances to these n-type FZ cells.<br />
27
Figure 4.1.3.3 shows the measured I-V curve <strong>of</strong> the best rear emitter n-PERT cell, Wnrj7-2a,<br />
which has demonstrated an effi ciency <strong>of</strong> 22.7%. This equals the best reported effi ciency<br />
from n-type silicon substrates [R. King, et al, 21st IEEE PVSC, p.227, 1990, P. Verlinden, 14th<br />
European PVSEC, pp.96, 1997].<br />
1.0<br />
amps<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
0.0<br />
Cell ID: Wnrj7-2a<br />
Cell Area: 22.04 cm 2<br />
Temperature: 25.1°C<br />
Voc = 702 mV<br />
Isc = 0.884 A<br />
Jsc = 40.1 mA/cm 2<br />
FF = 0.805<br />
Eff = 22.7%<br />
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8<br />
volts<br />
Figure 4.3.1.3: I-V curve <strong>of</strong> a rear emitter<br />
n-type cell as measured at Sandia National<br />
Laboratories under 100 mW/cm 2 , AM1.5<br />
global spectrum at 25ºC.<br />
It is seen that the cells on thinner substrates demonstrate improved effi ciencies. Also, limited<br />
emitter areas were defi ned by lithography for the last batch, Wnrj7. The emitter boron<br />
diffusion area was terminated about 200 µm from the scribed cell edge. This has signifi cantly<br />
improved the cell fi ll factors to over 80% and hence improved the cell effi ciencies.<br />
4.3.1.4 Spectral Response Measurement and Analysis<br />
The spectral response <strong>of</strong> the best n-type re-PERT cell, Wnrj7-2a, was measured at Sandia<br />
National Laboratories, with results shown in Figure 4.3.1.4. This cell has internal quantum<br />
effi ciency (IQE) around 98%, compared to 100% for the standard front emitter PERL cells<br />
on p-type substrates. Hence, its current density is only about 1 mA/cm 2 lower compared<br />
to the best p-type PERL cells. The previous 400 µm thick cells had a much lower IQE <strong>of</strong> only<br />
around 93% and a short-circuit current density about 2 mA/cm 2 lower than the best PERL<br />
cells. It is clear that the 170 µm thin substrates and the modest resistivity <strong>of</strong> 1.5 Ω-cm have<br />
helped this cell to increase its IQE and Jsc. The IQE for the second batch cell, Wnrj4-3b,<br />
was very close to that from the 170 µm thin cell. Hence, the 270 µm cells and 170 µm thin<br />
cells had similar experimental short-circuit current densities. However, the long wavelength<br />
refl ection from the 170 µm thin cell had been signifi cantly increased compared to that from<br />
the thicker cells. This gives evidence that the 170 µm might be too thin already for the best<br />
light absorption.<br />
Reflectance, EQE, and IQE (%)<br />
100<br />
90<br />
80<br />
70<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
0<br />
300 400 500 600 700 800 900 1000 1100 1200<br />
Wavelength (nm)<br />
Figure 4.3.1.4: Spectral response <strong>of</strong><br />
the best n-type re-PERT cell, Wnrj7-2a,<br />
on a 1.5 µ-cm resistivity, 170 µm thin<br />
n-type substrate.<br />
28
It is seen that the measured IQE in Figure 4.3.1.4 initially increases for the longer wavelength<br />
light. This corresponds to the fact that the weakly absorbed longer wavelength light generates<br />
minority carriers closer to the rear emitter, and hence it has a lower recombination loss<br />
during carrier transportation.<br />
In this work, the device simulation program, PC-1D, was used to simulate and to fi t the<br />
performance <strong>of</strong> the re-PERT cells. The bulk and the surface recombination parameters<br />
listed in Section 4.3.1.2 are also used for this PC-1D simulation and give a close match to<br />
the experimental cell performance. The reduced IQE in the very short wavelength range<br />
below 500 nm is due largely to the absorption in the 50 nm thick ZnS in the double layer<br />
antirefl ection coating. The PC-1D simulation neglects this ZnS absorption but allows a good<br />
fi t to the rest <strong>of</strong> the IQE curves, as shown in Figure 4.3.1.5.<br />
Figure 4.3.1.5: Thickness effect on IQE<br />
simulated from PC-1D.<br />
As expected from the simulation, the 170 µm thin device gives a clear advantage in its IQE.<br />
Hence, it also demonstrated a higher short-circuit current density. However, the increased<br />
long wavelength refl ection for the 170 µm thin cell gives evidence that the material absorption<br />
is decreasing signifi cantly. Hence it is believed that the 170 µm cell is a little too thin, since<br />
the 270 µm device gave a slightly higher current density, as shown in Table 4.3.1.1. The<br />
optimum thickness may be around 200 µm.<br />
4.3.1.5 Improved Cell Stability<br />
One <strong>of</strong> the major problems with our previous front emitter n-type PERT cells was their instability<br />
under one-sun illumination, and even during storage in a nitrogen box. However, it is found<br />
that the rear emitter n-type PERT cells have very stable and even improved performance<br />
under one-sun illumination by an ELH lamp light source for many hours. Figure 4.3.1.6 shows<br />
these results.<br />
29
Voc, mV<br />
705<br />
703<br />
701<br />
699<br />
697<br />
695<br />
693<br />
691<br />
Wnrj1-2b<br />
Wnrj1-1a<br />
Wnrj1-1b<br />
Wnrj1-2a<br />
Wnrj7-1b<br />
0 5 10<br />
One - sun Illumination Time, hr<br />
Jsc, ma/cm2<br />
40.0<br />
39.8<br />
39.6<br />
39.4<br />
39.2<br />
39.0<br />
38.8<br />
Wnrj1-2b<br />
Wnrj1-1a<br />
Wnrj1-1b<br />
Wnrj1-2a<br />
Wnrj7-1b<br />
38.6<br />
0 5 10<br />
One - sun Illumination Time, hr<br />
FF, %<br />
80.0<br />
79.5<br />
79.0<br />
78.5<br />
78.0<br />
77.5<br />
77.0<br />
Wnrj1-2b<br />
Wnrj1-1a<br />
Wnrj1-1b<br />
Wnrj1-2a<br />
Wnrj7-1b<br />
Efficiency, %<br />
22.5<br />
22.0<br />
21.5<br />
21.0<br />
Wnrj1-2b<br />
Wnrj1-1a<br />
Wnrj1-1b<br />
Wnrj1-2a<br />
Wnrj7-1b<br />
76.5<br />
0 5 10<br />
One -sun Illumination Time, hr<br />
20.5<br />
0 5 10<br />
One -sun Illumination Time, hr<br />
Figure 4.3.1.6: The performance <strong>of</strong> rear boron emitter n-type PERT cells during 1–2 days<br />
one-sun illumination by ELH lamp. The edge recombination may have been reduced by the<br />
illumination process.<br />
For the cells in the fi rst batch, Wnrj1, with scribed and cleaved through emitter edges,<br />
effi ciencies improved by 0.2% to 0.4% absolute after illumination. Most <strong>of</strong> these gains<br />
came from Jsc improvements, which is a sign <strong>of</strong> reduced total recombination loss. The high<br />
recombination at the scribed edges may have been reduced during illumination, possibly due<br />
to slight oxidation <strong>of</strong> the cleaved edge surfaces.<br />
However, the cells in the latest batch, Wnrj7, did not show this increase to the same extent<br />
under one-sun illumination. Separating the emitter from the cleaved exposed edges with the<br />
limited-area emitter design has stabilised cell performance, as well as improving the cell fi ll<br />
factors. Fortunately, cleaved emitter edges for larger area commercial cells will have much<br />
less effect on the fi nal cell performance.<br />
It is clear that moving the light sensitive boron diffused emitter from the front to the rear<br />
surface has signifi cantly improved the cell stability.<br />
4.3.1.6 Recent Design Without Cell Scribing<br />
Recently, the n-type re-PERT cells have been redesigned into one cell on a single wafer<br />
structure. This will eliminate the need for cell scribing. In this case, the edge recombination<br />
loss from the scribed and the cleaved edge surfaces will be eliminated. This structure is<br />
designed to test the recombination contribution from the scribed edges. It will also allow<br />
investigation <strong>of</strong> the full potential <strong>of</strong> such n-type re-PERT cell structures. We are confi dent that<br />
such device structures with minimum edge recombination contribution will further improve<br />
the cell effi ciency to a new level above the present world record, as demonstrated by the<br />
scribed re-PERT cells.<br />
30
The re-PERT cells on n-type CZ substrates will also be investigated in the near future. Since<br />
high carrier lifetimes <strong>of</strong> over 5 ms have been demonstrated from n-type SEH CZ substrates,<br />
the n-type CZ rear emitter cells are expected to give similar performance to the n-type FZ<br />
cells.<br />
4.3.1.7 High Efficiency Group’s Journal Publications:<br />
G. Kumaravelu, M.M. Alkaisi, D. Macdonald, J. Zhao, B. Rong and A. Bittar,<br />
“Minority carrier lifetime in plasma-textured silicon wafers for solar cells”, Solar<br />
Energy Materials and Solar Cells, Vol. 87, pp. 99-106, 2005.<br />
4.3.1.8 High Efficiency Group’s Conference Papers:<br />
J. Zhao and A. Wang, “High Effi ciency Rear Emitter PERT Solar Cells on N-type FZ<br />
Single Crystalline Silicon Substrates” (oral presentation), 20th European<br />
Photovoltaic Solar Energy Conference and Exhibition, Barcelona, Spain, pp.<br />
806- 809, 6-10, June, 2005.<br />
Guoxiao Yao, Jianhua Zhao and Jeffery E. Cotter, “Hybrid stencil-print/buriedcontact<br />
solar cells” (oral presentation), 20th European Photovoltaic Solar Energy<br />
Conference and Exhibition, Barcelona, Spain, pp. 749-752, 6-10, June, 2005.<br />
Jianhua Zhao, “The foundation and the strategy <strong>of</strong> Photovoltaic Industry in China”,<br />
China Power Conference, Beijing, China, 10-12, August, 2005.<br />
J. Zhao and A. Wang, “Rear Emitter N-Type Silicon Cells with Stable High<br />
Performance” (invited talk), Technical Digest <strong>of</strong> the 15th International<br />
Photovoltaic Science and Engineering Conference, Shanghai, 10-15, October, pp.<br />
122-123, 2005.<br />
J. Zhao, “High Effi ciency Silicon Solar Cell Technologies Ready for Production”<br />
(invited talk), Technical Digest <strong>of</strong> the 15th International Photovoltaic Science and<br />
Engineering Conference, Shanghai, 10-15, October, pp. 310-311, 2005.<br />
J. Zhao, “High Effi ciency Silicon Solar Cells for Mass Production” (invited talk),<br />
Sino Euro Workshop, the 15th International Photovoltaic Science and Engineering<br />
Conference, Shanghai, 10, October, 2005.<br />
J. Zhao, “Manufacturing Technologies for High Effi ciency Crystalline Silicon Solar<br />
Cells” (invited talk), PV Industrial Workshop, the 15th International Photovoltaic<br />
Science and Engineering Conference, Shanghai, 12, October, 2005.<br />
31
4.3.2 BURIED CONTACT SOLAR CELL GROUP<br />
<strong>University</strong> Staff:<br />
Dr. Jeffrey Cotter<br />
Pr<strong>of</strong>. Stuart Wenham<br />
Project Scientists and Technicians:<br />
Dr. J. H. Allen Guo<br />
Jamie Green<br />
Kate Fisher<br />
Fennalia Tan<br />
David Di<br />
Brett Hallam<br />
Adeline Sugianto<br />
Research Associates:<br />
Dr. Robert Bardos<br />
Dr. Tom Puzzer<br />
Dr. Alistair Sproul<br />
Ted Szpitalak<br />
Postgraduate Students:<br />
Malcolm Abbott<br />
Florence Chen<br />
Matthew Edwards<br />
Kate Fisher<br />
Guoxiao “James” Yao<br />
Attachai “Tao” Ueranantason<br />
Undergraduate Thesis Students:<br />
Nino Borojevic<br />
Andrew Li<br />
Jonathon Bocking<br />
Undergraduate Summer Scholars:<br />
Brett Hallam<br />
Neil Bennett<br />
The Buried-Contact Solar Cell Group aims to develop new solar cell structures and novel<br />
process technologies specifi cally for commercially relevant silicon wafers. The group has<br />
a broad spectrum <strong>of</strong> research and development activities that address the evolving nature<br />
<strong>of</strong> commercial silicon solar cells. The Group’s main activities are focused on developing higheffi<br />
ciency, thin-wafer Buried Contact solar cells, developing <strong>of</strong> low-cost processing technologies<br />
and related device designs, and transferring BC technology to industrial and technical<br />
collaborators.<br />
4.3.2.1 N-Type vs. P-Type Wafers for High Efficiency<br />
Silicon Solar Cells<br />
N-type Czochralski and multicrystalline silicon as materials for solar cells have received<br />
considerable attention recently due to emerging evidence that they are electrically superior to<br />
p-type material. Several groups have reported very high pre- and post-processing lifetimes in<br />
both cast multicrystalline and Czochralski materials. There appears to be excellent potential<br />
for commercial grade, n-type silicon wafers as a basis for high-effi ciency commercial solar<br />
cells.<br />
33
In silicon wafers that have naturally occurring defects, such as dislocations, grain boundaries<br />
and chemical impurities, the choice <strong>of</strong> wafer doping type can play a key role in mitigating<br />
the effect that the impurities have on the electrical quality <strong>of</strong> the wafer. For example, Figure<br />
4.3.2.1a shows the predicted injection-dependent normalized recombination rate for two<br />
Shockley-Read-Hall (SRH) defect species in p-type and n-type materials. Curves A and B<br />
predict the behaviour in p-type material: (A) for a defect located at a single energy level within<br />
the bandgap and (B) the behaviour for a defect distributed over a continuum <strong>of</strong> energy levels<br />
within the bandgap. Curve C predicts the behaviour for the same defects in n-type material,<br />
while curve D predicts the behaviour <strong>of</strong> both n-type and p-type material in the absence <strong>of</strong><br />
the defect. When viewed along the 1-sun operating loci (“1-sun Vmp”), it is clear that for a<br />
given defect species, the effective recombination rate depends signifi cantly on the doping<br />
type <strong>of</strong> the wafer. In this case, the recombination in the n-type wafer is one or two orders <strong>of</strong><br />
magnitude lower than the p-type wafer.<br />
Figure 4.3.2.1: (a) Predicted injection-dependent normalized recombination rate for n-type<br />
and p-type silicon wafers in the presence or absence <strong>of</strong> SRH defects; (b) Predicted effect <strong>of</strong><br />
SRH defects on the illuminated IV curve for the cases presented in (a).<br />
The signifi cance <strong>of</strong> the SRH behaviour in n-type and p-type wafers is highlighted in Figure<br />
4.3.2.1b, which shows the predicted illuminated IV curve for the four cases shown in Figure<br />
4.3.2.1a. The strong SRH recombination behaviour degrades the IV performance in the p-<br />
type material signifi cantly, reducing the operating point voltage by as much as 120 mV. The IV<br />
performance is hardly affected in the n-type material.<br />
The DSBC solar cell design, shown in Figure 4.3.2.2a, is ideal for making a direct comparison<br />
<strong>of</strong> solar cell terminal characteristics <strong>of</strong> solar cells made side-by-side on n-type and p-type<br />
silicon wafers. The DSBC processing sequence can be applied to either p-type or n-type<br />
wafers without signifi cant changes in the processing sequence - diffusions, oxidations, even<br />
metallisation processes can be applied with good effect to both types <strong>of</strong> wafers processed in<br />
split batches. Also, the DSBC process uses a heavy boron contact diffusion that is known to<br />
introduce diffusion induced misfi t dislocations into the bulk <strong>of</strong> the wafer. Such misfi t dislocations<br />
behave similarly to the SRH defects used in the simulations <strong>of</strong> Figure 4.3.2.1a. Figure 4.3.2.1b<br />
shows a comparison <strong>of</strong> terminal characteristics <strong>of</strong> p-type and n-type DSBC solar cells that<br />
were fabricated using the nearly identical fabrication sequences and processed in a split<br />
batch.<br />
34
Figure 4.3.2.2: (a) DSBC solar cell design for p-type silicon wafers; (b) Illuminated IV curves for<br />
p-type and n-type DSBC solar cells.<br />
The characteristic behaviour <strong>of</strong> SRH recombination attributed to boron-diffusion-induced<br />
misfi t dislocations is apparent in the terminal characteristics <strong>of</strong> all three p-type solar cells,<br />
as can be seen in the degraded fi ll factor and open-circuit voltage. The n-type solar cell is not<br />
affected.<br />
The case <strong>of</strong> the 1 Ωcm p-type and 1 Ωcm n-type solar cells is interesting. While the fi ll factor<br />
<strong>of</strong> both solar cells looks comparable, the detrimental effect <strong>of</strong> the SRH recombination on the<br />
p-type solar cell is not immediately apparent by examining the fi ll factor alone. However, the<br />
voltage characteristics <strong>of</strong> this solar cell are degraded by SRH recombination - the open-circuit<br />
voltage <strong>of</strong> the p-type solar cell is 20 mV lower than its n-type counterpart, yet, based on the<br />
relative doping levels (the p-type wafer has a factor <strong>of</strong> three higher doping level), one would<br />
expect the p-type solar cell to reach open-circuit voltages about 25 mV higher than the n-type<br />
solar cell.<br />
The performance bias toward n-type wafers in the above device comparison is attributed<br />
to the electronic nature <strong>of</strong> the defect species – diffusion induced misfi t dislocations have a<br />
relatively high electron capture cross section and low hole capture cross section. Many defect<br />
studied at UNSW also display similar electronic nature, including laserinduced<br />
defects, surface defects and chemical contaminants. Figure<br />
4.3.2.3 shows a comparison <strong>of</strong> the recombination rates <strong>of</strong> a split<br />
batch <strong>of</strong> n-type and p-type wafers that were unintentionally subjected<br />
to contamination during a high temperature process sequence.<br />
All <strong>of</strong> the p-type wafers show high and varied recombination rates,<br />
compared to the n-type wafers which show much lower and less<br />
varied recombination.<br />
Figure 4.3.2.3: Photoconductance measurements <strong>of</strong> p-type and n-<br />
type silicon wafers unintentionally contaminated during a phosphorus<br />
diffusion and thermal oxidation process sequence. The dash-dot line<br />
shows the approximate one-sun operating point for a high-effi ciency<br />
solar cell.<br />
35
Based on our fi ndings, we consider n-type silicon wafers to present a signifi cant opportunity<br />
for commercial, high-effi ciency silicon solar cells. The growing list <strong>of</strong> defects that exhibit high<br />
hole lifetime and low electron lifetime, and the impact that the associated SRH recombination<br />
has on the terminal characteristics <strong>of</strong> solar cells, suggests that n-type silicon wafers are<br />
better suited for high-effi ciency, commercial silicon solar cells. Overall, in terms <strong>of</strong> the<br />
electrical requirements <strong>of</strong> a solar grade wafer for industrial high-effi ciency silicon solar cells,<br />
n-type wafers seem to hold a signifi cant advantage over p-type wafers.<br />
Given the demonstrated potential <strong>of</strong> n-type silicon wafers for high-effi ciency commercial<br />
silicon solar cells, UNSW is focusing considerable effort on developing technologies for n-type<br />
silicon solar cells, including high-effi ciency interdigitated backside buried contact silicon solar<br />
cells, screen printed n-type silicon solar cells, laser diffused n-type silicon solar cells and silicon<br />
nitride based n-type silicon solar cells.<br />
4.3.2.2 N-TYPE SILICON SOLAR CELLS<br />
High Efficiency IBBC SILICON Solar Cells<br />
In 2004, we demonstrated the use <strong>of</strong> laser grooved BC technology to fabricate high-effi ciency<br />
Interdigitated Backside Buried Contact (IBBC) solar cells, as well as the more conventional<br />
n-type Double-sided Buried Contact (DSBC) solar cells.<br />
Three key obstacles to effi ciency were overcome in the IBBC process and in device design<br />
during 2004-parasitic shunt resistance at the diffusion overlap regions, simultaneous<br />
metallization <strong>of</strong> n++ and p++ contact regions, and optimization <strong>of</strong> the diffused regions. An<br />
effi ciency <strong>of</strong> 19.2% was achieved on the n-type FZ IBBC solar cell design, demonstrating<br />
the high-effi ciency potential <strong>of</strong> buried contact silicon solar cells and <strong>of</strong> n-type silicon wafers.<br />
Moreover, all types <strong>of</strong> Buried Contact solar cells made on Czochralski and multicrystalline<br />
phosphorus-doped n-type silicon wafers showed no light-induced degradation.<br />
In 2005, we focused on the further development <strong>of</strong> the IBBC solar cell design and process,<br />
with signifi cant improved in 2005 due to developments in both processing and design aspects.<br />
Further improvements in diffusions, especially the boron emitter, has led to near elimination<br />
<strong>of</strong> the shunting in the diffusion overlap regions and improvements in open circuit voltage. A<br />
passivated junction edge along the cell perimeter <strong>of</strong> the solar cell signifi cantly reduces the<br />
edge recombination, which previously had limited fi ll factor and open circuit voltage. Combined,<br />
these developments have led to increases in fi ll factor, open circuit voltage and effi ciency.<br />
Table 4.3.2.1 summarizes a typical batch result as <strong>of</strong> the end <strong>of</strong> 2005.<br />
textured passivated AR coated<br />
n+type FSF<br />
n+type base<br />
n++-type<br />
groove<br />
p+type emitter<br />
36<br />
Figure 4.3.2.4: IBBC solar cell design for n-type wafers.
TABLE 4.3.2.1: Electrical output <strong>of</strong> the IBBC solar cells (unverifi ed) under both front and rear<br />
illumination.<br />
Front Illumination<br />
Rear Illumination<br />
Cell Voc Jsc FF η Voc Jsc FF η<br />
ID (mV) (mA/cm 2 ) (%) (mV) (mA/cm 2 ) (%)<br />
D2-4 672.9 36.99 0.7743 19.27 670.9 31.79 0.7813 16.66<br />
D2-5 674.1 37.12 0.7805 19.54 670.8 31.66 0.7890 16.76<br />
D2-8 676.5 37.12 0.7911 19.87 670.9 31.70 0.7908 16.82<br />
D2-9 672.6 37.06 0.7773 19.43 672.6 31.85 0.7875 16.87<br />
Detailed modelling indicates improvements in both the front and rear surface recombination<br />
velocities – about 200 cm/s (improved from 600 cm/s in 2004) and 6000 cm/s (improved<br />
from 10000 cm/s in 2004) respectively. Saturation current density due to edge-junction<br />
recombination (J 02 ) has been reduced by about 1 order <strong>of</strong> magnitude (3.3 pA/cm2 versus<br />
40 pA/cm2 in 2004) due to the incorporation <strong>of</strong> an edge junction passivation technique.<br />
Figure 4.3.2.5 shows the illuminated J-V curve <strong>of</strong> the best solar cell (device D2-8). Compared<br />
to the results reported in 2004, open-circuit voltage and fi ll factor are increased by 12 mV<br />
and 2.7% respectively. In these devices, short circuit current is reduced by typically 0.9 mA/<br />
cm2 due to a lower bulk lifetime, the cause <strong>of</strong> which is currently under investigation. Overall,<br />
the effi ciency best has increased by 0.65% to about 19.85%.<br />
Figure 4.3.2.5: Light J-V curve <strong>of</strong> IBBC solar cell D2-8.<br />
In 2005, we also introduced statistical process control (SPC) techniques to ensure that<br />
the process fl uctuation is well monitored and controlled and to enhance cell and process<br />
development efforts. Overall, the distribution <strong>of</strong> results from batch to batch, and across any<br />
batch are much tighter, indicating the fabrication process is under control.<br />
The effi ciency <strong>of</strong> IBBC solar cell is expected to achieve 20.3% after improving the postprocessing<br />
lifetime to currently expected values. Further research on reducing series<br />
resistance and other key loss mechanisms is currently in progress<br />
Screen Printed N-Type Silcion Solar Cells<br />
The use <strong>of</strong> n-type Czochralski silicon wafers and stencil-print metallisation has the potential<br />
to improve the performance (Cz) <strong>of</strong> commercial silicon solar cells, but little work has been<br />
done in investigation <strong>of</strong> print pastes and fi ring conditions for boron-diffused, p-type emitters.<br />
To investigate the formation <strong>of</strong> low-cost, good-quality metallization on n-type Cz silicon, several<br />
commercially available print pastes and fi ring cycles were evaluated on p-type emitters by<br />
characterizing the junction shunting and contact resistance.<br />
37
Conductive Paste Firing Tests<br />
Basic cell test structures (Figure 4.3.2.6) were fabricated with a diffused, 100 Ω/sq front<br />
boron emitter. A bare silicon surface was used to avoid added complexities from having to fi re<br />
through an insulating layer. The front grid metallisation was formed by metal stencil printing <strong>of</strong><br />
thick-fi lm metal pastes, using stainless steel foil stencil techniques developed in 2004. These<br />
stencils are capable <strong>of</strong> printing high defi nition contact patterns with fi ngers as narrow as<br />
60 µm in width and up to 30 µm in height. The stencil-printed front contacts were fi red with<br />
varying peak temperatures and durations in a four-zone belt furnace with peak temperature<br />
varied between 700°C and 900°C.<br />
Screen-printed Al contact<br />
p-type emitter<br />
n-type bulk<br />
n+ back surface field<br />
Screen-printed Ag contact<br />
Figure 4.3.2.6: Test structure used for evaluation <strong>of</strong> front contact shunt resistance.<br />
Figure 4.3.2.7a shows shunt-resistance results obtained for the case <strong>of</strong> aluminium paste, fi red<br />
with a peak time <strong>of</strong> 10 s. Results clearly show acceptable shunt resistance for a wide range<br />
<strong>of</strong> peak temperatures ranging between 720oC and 870oC. Acceptable shunt resistances<br />
were also obtained using silver paste (Figure 4.3.2.7b), however perhaps surprisingly, shunt<br />
resistances obtained for aluminium pastes were <strong>of</strong>ten higher than those in cells fabricated<br />
with silver.<br />
Figure 4.3.2.7: Shunt resistances after fi ring a) Ferro 5540 aluminium and b) Ferro 33-466<br />
silver thick-fi lm conductive paste. Peak fi ring time was 10 s.<br />
38
To measure contact resistance to the boron-diffused layer, a stencil with a specially formed<br />
contact pattern consisting <strong>of</strong> vertical lines spaced progressively further apart was fabricated,<br />
and contacts printed on similar test structures to that shown in Figure 4.3.2.6. Contact<br />
resistances were measured using the transmission-line method, and are shown in Figure<br />
4.3.2.8.<br />
(a)<br />
(b)<br />
Figure 4.3.2.8. Results <strong>of</strong> contact resistance measurements with a) Ferro 5540 aluminium<br />
and b) Ferro 33-466 silver thick-fi lm conductive paste fi ngers. Peak fi ring time was 10 s.<br />
For aluminium paste fi red at peak temperature for 10 s, area-specifi c contact resistances as<br />
low as 3.85 mΩ.cm2 were obtained, as shown in Figure 4.3.2.8. For silver-aluminium paste<br />
fi red for 10 s, area-specifi c contact resistances were as low as 13.8 mΩ.cm2. Aluminium<br />
was found to be superior in forming ohmic contact to a boron-diffused layer, however, it was<br />
observed that the benefi t <strong>of</strong> low contact resistance when using aluminium was sometimes<br />
outweighed by the higher bulk resistivity <strong>of</strong> these pastes.<br />
Paste rheology in stencil printing is critical, and it was found that the lower viscosity <strong>of</strong><br />
commercially available aluminium pastes designed for the formation <strong>of</strong> the rear contact in<br />
p-type cells led to inferior print quality compared to silver pastes, which generally have a<br />
higher viscosity. This leads to fi ll factor degradation in fi nished cells that were stencil printed<br />
with low-viscosity aluminium. Since the inferior contact resistance <strong>of</strong> silver paste also leads<br />
to fi ll factor degradation, high viscosity aluminium is an ideal choice <strong>of</strong> paste for use in stencil<br />
printed n-type solar cells.<br />
Fabrication <strong>of</strong> n-Type Solar Cells with Stencil Printed Contacts<br />
Optimised printing and fi ring parameters for aluminium stencil-printed front contacts on<br />
boron emitters were applied to full solar cell structures. Figure 4.3.2.9 shows results for<br />
a typical cell. It was found that while high fi ll factors were achievable, overall cell effi ciencies<br />
were relatively low due to poor blue response and high front surface recombination velocities.<br />
Passivation <strong>of</strong> the boron-diffused surface was found to be very diffi cult. Collaboration with<br />
the Universitat Politecnica de Cataluña (UPC) is underway to devise an effective passivation<br />
scheme for the front surface <strong>of</strong> our cells using either a silicon nitride or silicon carbide layer.<br />
39
Figure 4.3.2.9: Illuminated IV curve and quantum effi ciency characteristics for a typical 4 cm 2<br />
n-type cell fabricated with stencil printed Al contacts. The above cell has fi ll factor <strong>of</strong> 75% and<br />
effi ciency <strong>of</strong> 10.1%. Shunt resistance was 8 kΩ and series resistance was 0.38Ω.<br />
Laser Diffused N-Type Silcion Solar Cells<br />
Laser diffusion has been investigated as a technique for the formation <strong>of</strong> high-effi ciency IBBC<br />
and DSBC solar cell structures. Currently the formation <strong>of</strong> the contact regions on these<br />
devices requires an estimated 50% <strong>of</strong> the total fabrication time and requires several hightemperature<br />
processing steps. This intensive high-temperature processing is not compatible<br />
with high-throughput manufacturing or with commercial wafer materials such as CZ and<br />
mc-Si wafers. Laser diffusion <strong>of</strong>fers good potential to simplify the processing <strong>of</strong> high-effi ciency<br />
bifacial silicon solar cells.<br />
The laser diffusion performed in this work occurs through a process <strong>of</strong> laser-induced surface<br />
melting and liquid-state diffusion <strong>of</strong> dopant atoms. A thin layer <strong>of</strong> dopant fi lm is applied to<br />
the wafer surface over an existing silicon dioxide layer, then diffused by a series <strong>of</strong> laser<br />
pulses incident on the surface. Each laser pulse causes the underlying silicon to melt, then<br />
recrystallize, without any signifi cant increase in temperature away from the target area.<br />
Dopant atoms diffuse quickly in the liquid phase and are incorporated in electrically active<br />
lattice sites during recrystallization. During the laser diffusion process the silicon dioxide is<br />
locally removed, leaving a mask in the non-doped regions for subsequent electroless metal<br />
plating. The heavily diffused regions are formed selectively, at room temperature and without<br />
the need for any NaOH etching to remove laser damage or oxide deglazing to remove the<br />
silicon dioxide layer.<br />
During 2005, the diffusion process was optimised to achieve the heavy surface concentration<br />
(>10 19 ) and deep junction (> 1 µm) required to form ohmic contact between silicon and metal<br />
contacts. The diffused contact regions were then integrated into a simplifi ed version <strong>of</strong> the<br />
DSBC solar cell to investigate issues with device fabrication.<br />
Figures 4.3.2.10 shows typical images <strong>of</strong> the laser diffused contact regions. Laser diffusion<br />
was found to be compatible with the other cell fabrication processes including the electroless<br />
plating <strong>of</strong> metal contacts.<br />
40
Figure 4.3.2.10: Micrographs <strong>of</strong> laser diffused contacts; (a) plan view <strong>of</strong> intersection between<br />
contact fi nger with end <strong>of</strong> busbar; (b) plan view <strong>of</strong> contact pad area; and (c) a fi nger after<br />
electro-less copper plating.<br />
The electrical results <strong>of</strong> the best cell, shown in Figure 4.3.2.11, are very promising with the<br />
highest open-circuit voltage <strong>of</strong> 672 mV, a fi ll factor <strong>of</strong> 78% and short-circuit current <strong>of</strong> typically<br />
33 mA/cm2 demonstrated on an untextured, n-type, FZ wafer. It is concluded that laser<br />
diffusion is a promising alternative to furnace diffusions on IBBC and DSBC devices. Further<br />
work will focus on further optimisation <strong>of</strong> the process to increase the fi ll factor and integrate<br />
the laser diffusion process with IBBC cells.<br />
Figure 4.3.2.11: Illuminated IV characteristics <strong>of</strong><br />
the best n-type planar double-sided solar cell with<br />
laser doped contacts (unconfi rmed).<br />
Silicon Nitride Based n-type Silcion SolarCells<br />
It is well-established that silicon nitride (SiN) fi lms have excellent anti-refl ection and passivation<br />
properties, with a wide range <strong>of</strong> achievable refractive index, large amount <strong>of</strong> hydrogen for<br />
passivation, and positive fi xed charges that creates inversion/accumulation layer to enhance<br />
the passivation. While extensive work has been done on SiN for p-type solar cell applications (in<br />
particular, excellent passivation is demonstrated on p-type bare surface and on phosphorus<br />
diffused emitter) little literature is available on SiN for n-type solar cell applications. In 2005,<br />
we investigated the application <strong>of</strong> SiN fi lms for n-type wafers and solar cells.<br />
Our work shows that PECVD SiN deposited using a Roth&Rau AK400 PECVD system is<br />
particularly well suited for surface passivation <strong>of</strong> n-type silicon wafers and solar cells. In<br />
particular, the passivation quality on bare planar silicon surfaces, and boron diffused planar<br />
surfaces on n-type, 1 Ωcm, FZ wafers is exceptional. Furthermore, since SiN creates an<br />
accumulation layer rather than inversion layer on n-type surfaces, problems caused by the<br />
inversion layer on p-type surface such as shunting through the inversion layer are avoided.<br />
41
Silicon Nitride Passivation <strong>of</strong> Bare, n-type Silicon Surfaces<br />
Figure 4.3.2.12 shows the effective surface recombination velocity (SRV) versus injection level<br />
curves for three different refractive index fi lms on FZ wafers. For comparison purpose, the<br />
effective SRV <strong>of</strong> a sample that had a 100 ohm/sq phosphorus diffusion and oxidation are<br />
also plotted on the same graph. The two dotted lines on the graphs that cross the effective<br />
SRV curves represent the locus <strong>of</strong> the 1-sun open-circuit voltage and approximate maximumpower<br />
voltage operating points for a high-effi ciency, 200-micron thick solar cell.<br />
As illustrated in the graph, the surface passivation quality <strong>of</strong> as-deposited SiN is exceptional,<br />
achieving better recombination than the baseline phosphorus diffused and oxidized surface.<br />
In our work, silicon-rich fi lms give the best surface passivation at as-deposited conditions.<br />
Table 4.3.2.2 summarizes the extracted SRV and implied open<br />
circuit voltages for various as-deposited silicon nitride fi lms.<br />
Figure 4.3.2.12: The effective surface recombination velocity vs.<br />
minority carrier density plots <strong>of</strong> various SiN on 1 ohm.cm n-type FZ<br />
wafer.<br />
Table 4.3.2.2: The best effective SRV, effective SRV at 1-sun Voc, and<br />
implied voltage at 1-Sun Voc operating points <strong>of</strong> various SiN fi lms on<br />
planar n-type surface <strong>of</strong> a 200-micron thick wafer.<br />
Refractive index Wafer type Best SRVeff (cm/s) SRVeff at Voc (cm/s) Imp Voc (mV)<br />
2.00 n-FZ 7.01 13.34 723<br />
2.35 n-FZ 2.29 10.74 732<br />
2.75 n-FZ 1.96 9.53 739<br />
Phos and oxide n-FZ 5.23 15.48 709<br />
A voltage test structure device was prepared to further demonstrate the exceptional<br />
passivation quality provided by our SiN fi lms on n-type wafers. The device has two very small<br />
diffused contact regions, one was phosphorus diffused and the other boron diffused. A lowindex<br />
(nr = 2.00) SiN fi lm was deposited on both sides <strong>of</strong> the wafer in the as-deposited condition.<br />
Small openings made photolithographically in the contact regions allowed the measurement<br />
<strong>of</strong> the open-circuit voltage at approximately 1-sun, 25ºC <strong>of</strong> 719 mV. Further improvement<br />
is expected during 2006, with open-circuit voltages approaching 725 mV expected for test<br />
structures and 700 mV in fully fabricated silicon nitride passivated solar cells.<br />
Silicon Nitride Passivation <strong>of</strong> Boron Diffused Emitters<br />
42<br />
The ability <strong>of</strong> silicon nitride to passivate a boron-diffused emitter is important for making higheffi<br />
ciency n-type silicon solar cells. Our results show that for some diffusion sheet resistances,<br />
boron diffusion process, and subsequent thermal treatment, SiN is particularly well-suited for<br />
passivating boron-diffused surface on n-type silicon. Figure 4.3.2.13a shows the implied open<br />
circuit voltages (extracted from photoconductance measurements) <strong>of</strong> various passivation<br />
fi lms on boron-diffused emitters with sheet resistances ranging from roughly 45 ohm/sq to<br />
250 ohm/sq, including silicon dioxide and various silicon nitride fi lms. The results illustrate<br />
that at the as-deposited condition for the sheet resistance range studied in this work, SiN<br />
provides poorer passivation compared to silicon dioxide. Interestingly, it appears that, at the<br />
as-deposited condition, the three different SiN fi lms provide roughly the equivalent passivation<br />
quality and have similar trend across the range <strong>of</strong> sheet resistances studied.
The SiN fi lms with nr = 2.35 were annealed at 450°C in an N2 ambient for a total duration <strong>of</strong><br />
240 minutes. The evolution in implied open-circuit voltage <strong>of</strong> the annealed samples is plotted<br />
in Figure 4.3.2.13b. After 5 minutes <strong>of</strong> annealing, the relationship between the implied Voc<br />
and sheet resistance <strong>of</strong> boron emitter resembles that <strong>of</strong> as-deposited condition. However,<br />
after 30 minutes <strong>of</strong> annealing, the relationship has changed, with the implied open-circuit<br />
voltage increasing with increasing boron emitter sheet resistance. It is likely that the implied<br />
open-circuit voltage at low sheet resistance range is limited by the emitter recombination,<br />
whereas, at the high sheet resistance range, the surface recombination dominates. After<br />
240 minutes <strong>of</strong> annealing, an implied open-circuit voltage <strong>of</strong> 683 and 718 mV on 135 and 215<br />
ohm/sq boron emitter respectively was achieved. Dark emitter saturation current densities<br />
as low as 5 fA/cm2 per side was achieved by fully annealing the SiN fi lms. The results also<br />
indicate that the passivation quality provided by the annealed fi lms is superior to that by oxide<br />
for sheet resistance above around 100 ohm/sq.<br />
Figure 4.3.2.13: (a) The implied open-circuit voltage <strong>of</strong> various SiN fi lms at as-deposited<br />
condition on boron emitters across range <strong>of</strong> sheet resistances; (b) the implied open-circuit<br />
voltage <strong>of</strong> the nr = 2.35 SiN fi lm after various annealing times; (c) the extracted emitter<br />
saturation current density <strong>of</strong> the fully annealed nr = 2.35 SiN fi lms.<br />
4.3.2.3. PHOTOLUMINESCENCE AND PHOTOCONDUCTANCE TECHNIQUES FOR<br />
SILICON SOLAR CELL CHARACTERIZATION AND DEVELOPMENT<br />
Photoluminescence Techniques<br />
Photoluminescence characterization techniques have recently emerged in other Centre<br />
work as an accurate, fast, powerful tool for developing high-effi ciency silicon solar cells.<br />
The related PL techniques – spatially averaged PL lifetime, Suns-PL and PL imaging – <strong>of</strong>fer<br />
unprecedented quality and quantity <strong>of</strong> information that is benefi cial for a wide spectrum<br />
<strong>of</strong> solar cell development and fabrication purposes. As PL techniques are fast (especially<br />
imaging), contact-less and provide complementary spatial and injection level dependent<br />
information about recombination, they are very powerful techniques relevant to both well<br />
controlled laboratories and high-throughput solar cell factories.<br />
In 2005, we demonstrated the application <strong>of</strong> multiple PL techniques to several important<br />
aspects <strong>of</strong> high-effi ciency solar cell development at UNSW: wafer handling, furnace<br />
contamination, process-induced defects, cell design and cell process monitoring. In only a few<br />
months, photoluminescence (PL) techniques have greatly enhanced our solar cell fabrication<br />
sequences and designs, identifying problems that, in some cases, existed in our fabrication<br />
sequences for over ten years. With measurement times <strong>of</strong> 1 second per wafer (regardless<br />
<strong>of</strong> area) the power <strong>of</strong> this technique is as relevant to high-throughput mass-production as it is<br />
to laboratory cell research and development.<br />
43
Wafer Handling<br />
PL imaging was used to characterize a long-standing traditional technique <strong>of</strong> wafer drying<br />
at UNSW. In practice, wafers were dried using pressurized N 2 while holding the wafer by its<br />
edges with cleaned plastic tweezers. PL imaging quickly revealed the impact <strong>of</strong> this drying<br />
technique on electrical quality. Figure 4.3.2.14 shows PL images <strong>of</strong> two wafers that have been<br />
dried by two different techniques prior to high-temperature processing. In the images, bright<br />
areas correspond to a strong PL signal, indicating regions <strong>of</strong> good local lifetime. The dark<br />
areas indicate regions <strong>of</strong> poor lifetime where recombination has been locally increased. The<br />
image in Figure 4.3.2.14a shows four very dark spots on the four edges <strong>of</strong> the wafer, where<br />
it was held with the tweezers whilst drying with a nitrogen gun. The image in Figure 4.3.2.14b<br />
shows a wafer that was dried in a furnace in nitrogen ambient with no wafer handling by the<br />
edges, indicating no such dark spots on the edges.<br />
Figure 4.3.2.14: 1. sun photoluminescence<br />
image <strong>of</strong> an n-type CZ wafer that was processed<br />
in two successive high-temperature diffusion/<br />
oxidation processes: (a) Dried using tweezers<br />
to hold the wafer by the edges; (b) dried in a<br />
furnace with no edge handling. The location <strong>of</strong><br />
tweezers during the drying process is marked<br />
as (X) before the fi rst high-temperature step<br />
and (+) before the second high-temperature<br />
step.<br />
Furnace Contamination<br />
Contamination during high-temperature furnace processes can reduce bulk lifetime and<br />
lead to poor electrical properties <strong>of</strong> fi nished cells. PL monitoring can be used to identify an<br />
individual contaminated wafer and identify furnace equipment that requires cleaning. Figure<br />
4.3.2.15 demonstrates the case with which PL is able to detect furnace contamination. The<br />
wafer top left (labeled a) was contaminated (unintentionally) during the preparation for high<br />
temperature processing. This contamination was able to spread to the adjacent wafers<br />
(labeled b and c) in the direction <strong>of</strong> the gas fl ow, whereas other wafers in the same furnace<br />
tube (d, e and f), but upstream <strong>of</strong> the gas fl ow were unaffected by the contamination.<br />
Figure 4.3.2.15. 1 Sun photoluminescence<br />
images <strong>of</strong> samples after phosphorus diffusion:<br />
a) CZ sample; with worst contamination; b) CZ<br />
sample in slot next to contaminated sample, c) CZ<br />
sample further down furnace tube in direction <strong>of</strong><br />
gas fl ow; ; d-f) FZ samples successively upstream<br />
from the contaminated wafer.<br />
44
Solar Cell Design<br />
PL was used to characterize a new method <strong>of</strong> edge isolation for our n-type cells that involves<br />
a laser formed isolation trench through the emitter and around the solar cell’s edge. Several<br />
test wafers were prepared, half including the new edge isolation technique and the other<br />
half employing the traditional technique <strong>of</strong> laser-cleaving the edge. Figure 4.3.2.16 shows<br />
the results <strong>of</strong> Suns-PL and PL imaging used to analyze the effectiveness <strong>of</strong> the edge isolation<br />
technique. The image in Figure 4.3.2.16a shows the sample in high injection, revealing where,<br />
in order to create a recombination active area, we have deliberately scratched the surface<br />
outside the isolation trench with a diamond pen. Figure 4.3.2.16b shows the same sample in<br />
lower injection, demonstrating the effectiveness <strong>of</strong> the new process in providing isolation <strong>of</strong><br />
the main solar cell area from the edges and scratches.<br />
Figure 4.3.2.16: Photoluminescence images <strong>of</strong><br />
1 Ωcm, 240µm thick, n-type wafers with phosphorus<br />
diffusion on the rear side and boron diffusion on<br />
the front side with isolation trench after various<br />
processing steps: a) at ~ 1 Sun ; b) ~ 0.1 Suns<br />
Figure 4.3.2.17 shows the pseudo dark IV and dark local ideality factor <strong>of</strong><br />
the samples shown in Figure 4.3.2.16. The curves were generated using the<br />
Suns-PL technique. All <strong>of</strong> the curves labeled a) have the new trench isolation<br />
technique applied. These curves compare to the non-isolated cases b), c)<br />
and d) in which the effect <strong>of</strong> edge enhanced junction recombination and edge<br />
shunting are apparent. In this case, PL techniques provided almost immediate<br />
feedback in assessing the effectiveness <strong>of</strong> the new edge isolation process.<br />
Figure 4.3.2.17: (top) Pseudo dark I-V curves obtained from QSS-PL<br />
measurements on a 1 Ωcm, 240µm thick, n-type wafer with phosphorus<br />
diffusion on the rear side and boron diffusion on the front side after various<br />
processing steps (see text for details); (bottom) corresponding local ideality<br />
factor.<br />
Silicon Nitride – Wafer Handling and Preparation<br />
PL imaging is particularly powerful for developing silicon nitride as a passivation layer because,<br />
unlike thermal oxide fi lms, silicon nitride fi lms show a rich variety <strong>of</strong> spatially resolved effects,<br />
even in samples with exceptional quality passivation layers.<br />
In 2005, we investigated the application <strong>of</strong> PL imaging to the development <strong>of</strong> improved quality<br />
silicon nitride fi lms. We found that PL imaging can be used to understand the deposition<br />
process, to identify and address key manual handling issues, to characterize thermal annealing<br />
processes, and to study the effect <strong>of</strong> post-deposition chemical processing.<br />
PL imaging can be used to help identify how sample handling can affect the passivation<br />
quality <strong>of</strong> the SiN fi lms. Figure 4.3.2.18 shows a few PL images that highlight handling related<br />
issues.<br />
45
(a) (b) (c)<br />
Figure 4.3.2.18: SiN samples showcasing poor sample handling: a) sample with micro-cracks;<br />
b) sample with micrometer marks; c) sample with intentionally induced scratches.<br />
Figure 4.3.2.18a shows how PL imaging can detect micro-cracks that are not visible to the<br />
naked eye. The PL image in Figure 4.3.2.18b reveals that the micrometer used to measure<br />
the wafer thickness, either before or after the SiN deposition, can also lead to localized regions<br />
<strong>of</strong> higher recombination. Note the four perfectly round dark circles in Figure 4.3.2.18b. Figure<br />
4.3.2.18c illustrates the impact <strong>of</strong> scratches on passivation quality. In this case, two very light<br />
scratches were deliberately made on the sample, one before and one after the SiN deposition.<br />
The image suggests that even very light scratches, both before and after deposition, can be<br />
resolved by the PL imaging technique.<br />
Silicon Nitride – Process Monitoring<br />
At UNSW, the passivation quality <strong>of</strong> the samples in the same deposition run is occasionally<br />
inconsistent from sample to sample. Occasionally, a particularly bad sample may be included<br />
in a batch run. Since the PL imaging technique is fast (typically 1 second for the samples<br />
studied here), it is a powerful tool for process monitoring and control in cases like these.<br />
Figure 4.3.2.19 shows two samples prepared identically in the same deposition runs. One<br />
sample is clearly better passivated than the other, as indicated by a comparison <strong>of</strong> the PL<br />
intensity <strong>of</strong> both samples.<br />
Figure 4.3.2.19: PL images <strong>of</strong> two<br />
samples prepared identically in<br />
the same batch and deposition<br />
run.<br />
(a)<br />
(b)<br />
This observation from PL imaging also agree with QSS-PC measurement, as the sample on<br />
the left hand side has an extracted implied open-circuit voltage <strong>of</strong> 670 mV and the one on the<br />
right hand side has an implied open-circuit voltage <strong>of</strong> 725 mV.<br />
46
Photoconductance Techniques<br />
During 2005, we developed a new LED-based light source for our heavy use photoconductance<br />
testing system. It uses seven high-luminosity light emitting diode (LED) arrays, consisting <strong>of</strong><br />
over 400 individual LEDs, to illuminate wafers instead <strong>of</strong> the traditional fl ash lamp. This LED<br />
array is PC controlled and can generate arbitrary waveforms over more than 4 orders <strong>of</strong><br />
magnitude and optical intensities <strong>of</strong> up to 10 suns. The control system allows the intensity<br />
and shape <strong>of</strong> the light pulse be be very accurately controlled to induce near perfect steady<br />
state or transient conditions. More importantly, the control system, along with computer<br />
data acquisition electronics, allows the system to repeat a measurement thousands <strong>of</strong> times<br />
to obtain high quality data that is free from spurious noise. Figure 4.3.2.20 shows the LED<br />
assembly.<br />
Figure 4.3.2.20: LED assembly above a test wafer<br />
on a Sinton Consulting Photoconductance Tester.<br />
The control s<strong>of</strong>tware is written in LabView.<br />
It presents a user-friendly graphical user<br />
interface which allows the users to easily alter<br />
all the settings to suit their needs. As results are<br />
acquired, they are displayed on the screen in real<br />
time. Once fi nished and results are averaged, a<br />
series <strong>of</strong> theoretical curves can be overlayed on<br />
the graph to determine key wafer parameters.<br />
One <strong>of</strong> the most useful features <strong>of</strong> this s<strong>of</strong>tware<br />
is that it supports a database that allows users<br />
to store the results <strong>of</strong> a single wafer test, or the<br />
results <strong>of</strong> a multitude <strong>of</strong> tests made on several wafers in a batch tested at multiple stages<br />
<strong>of</strong> fabrication, in a single database fi le. The database also stores all measurement settings<br />
for every test, allowing users to recall and retest wafers with ease. This database is stored<br />
in an excel workbook fi le, which can easily be accessed by the user for editing or to create<br />
customized graphs and tables. The user interface is shown in Figure 4.3.2.21.<br />
Figure 4.3.2.21: LabView user interface<br />
for the upgraded photoconductance<br />
tester.<br />
The system will be upgraded in early<br />
2006, doubling the number <strong>of</strong> LEDs and<br />
optimising their arrangement in order to<br />
reach intensity levels <strong>of</strong> up to 30 suns.<br />
47
4.3.2.4 Process Engineering – Contact Silicide Formation<br />
Contact resistance and contact adhesion is important for the formation <strong>of</strong> high-effi ciency<br />
silicon solar cells. In 2005, we studied the formation <strong>of</strong> nickel silicide for n-type silicon solar<br />
cells. High effi ciency n-type solar cells use both heavy boron contact diffused (BCD) and heavy<br />
phosphorus contact diffused (PCD) regions to form the metal semiconductor regions. Effective<br />
metallization requires the simultaneous deposition <strong>of</strong> electroless nickel and formation <strong>of</strong> nickel<br />
silicide in both BCD and PCD regions.<br />
In this study, we form arrays <strong>of</strong> small-area contacts, formed by laser drilling a small hole<br />
in an oxidized silicon wafer. The resulting inverted, square-based pyramid “pits” are then<br />
diffused with either a BCD or PCD, which are subsequently deglazed with an HF-based etch.<br />
The diffused pits are then subjected to an electroless nickel deposition process, followed by a<br />
350°C sintering process in nitrogen. The small pits afford a better view <strong>of</strong> the nucleation and<br />
sintering processes compared to grooves.<br />
Figure 4.3.2.22 shows two diffused and nickel plated pits that were formed with insuffi cient<br />
diffusion glass deglazing. Figure 4.3.2.22a shows poor deposition in a poorly deglazed PCD<br />
contact pit, while Figure 4.3.2.22b shows poor nucleation on a BCD pit.<br />
Figure 4.3.2.22: Nickel plated to a pits in a silicon wafer<br />
with (a) PCD and (b) BCD pits.<br />
(a)<br />
(b)<br />
An improved deglazing process leads to improved plating<br />
in the PCD pit, exhibiting a smooth, conformal layer <strong>of</strong><br />
nickel plated over the entire contact surface, as shown<br />
in Figure 4.3.2.23a. The nucleation in the BCD is still poor, as shown in Figure 4.3.2.23b, due<br />
to the poor affi nity <strong>of</strong> the alkali-based nickel solution to plate p-diffused surfaces.<br />
Figure 4.3.2.33: Properly deglazed pits with a heavy a)<br />
PCD and b) BCD.<br />
(a)<br />
(b)<br />
The samples in Figure 4.3.2.22 were sintered in a<br />
furnace at 350°C for 10 minutes in nitrogen to form<br />
a nickel silicide at the interface between the nickel and<br />
the heavily diffused silicon. The un-reacted nickel was<br />
completely removed by etching and the surface <strong>of</strong> the pits examined (Figure 4.3.2.24). There<br />
is clearly interaction between the silicon and the nickel as is evidenced by the roughening <strong>of</strong><br />
the surface on the PCD pit. The nickel was so badly nucleated on the BCD pit that the Si has<br />
been drawn out from the interface such that it almost fi lls the whole pit, forming a poor-quality<br />
physical metal-semiconductor interface.<br />
Figure 4.3.2.24: Pits on the same sample as that shown<br />
in Figure 4.3.2.20 after sintering and nickel removal.<br />
(a)<br />
(b)<br />
Palladium activation is effective in increasing the<br />
nucleation density <strong>of</strong> electroless nickel plating on both<br />
p++ and n++ diffused silicon surfaces. Figure 4.3.2.25<br />
shows PCD and BCD pits that have been well deglazed,<br />
activated with a palladium solution and subsequently plated with electroless nickel. The<br />
electroless nickel plate is more uniform in the palladium activated BCD pit compared to<br />
without palladium activation.<br />
48
Figure 4.3.2.25: Properly deglazed pits with palladium<br />
activation: a) PCD and b) BCD.<br />
(a)<br />
(b)<br />
The palladium activated pit samples were sintered at<br />
350°C for 10 minutes and the un-reacted nickel was<br />
removed. Figure 4.3.2.26 shows a much improved<br />
interface in the BCD pits with palladium activation, and the roughening <strong>of</strong> the surface indicates<br />
that a thin, fairly uniform layer <strong>of</strong> silicide has been formed at the interface. The morphology <strong>of</strong><br />
the PCD interface is also improved by the inclusion <strong>of</strong> the palladium activation process.<br />
Figure 4.3.2.26: Pits on the same sample as that shown<br />
in Figure 4.3.2.23 after sintering and nickel removal.<br />
(a)<br />
(b)<br />
49
4.4 SECOND GENERATION: THIN-FILMS<br />
<strong>University</strong> Staff:<br />
A/Pr<strong>of</strong>. Armin Aberle (group leader)<br />
Dr. Alistair Sproul<br />
Technical Staff:<br />
Mark Griffi n<br />
Dr. Bruce Beilby (part-time)<br />
Jürgen Weber (part-time)<br />
Research Fellows/Associates:<br />
Frederick Bamberg (since 11/05)<br />
Dr. Patrick Campbell (part-time)<br />
Peter Klement (since 08/05)<br />
Dr. Dengyuan Song<br />
Dr. Per Widenborg<br />
Visiting Research Fellows:<br />
Pr<strong>of</strong>. Roman Buitrago (08/05 to 11/05)<br />
Postgraduate Research Students:<br />
Daniel Inns (PhD)<br />
Oliver Kunz (PhD)<br />
Lei “Adrian” Shi (Masters, since 08/05)<br />
Mason Terry (PhD)<br />
Timothy Walsh (PhD)<br />
Jürgen Weber (Masters, since 07/05)<br />
Axel Straub (PhD, until 07/05)<br />
Ferdi Ferdiansjah (Masters, until 12/05)<br />
Research Assistants:<br />
Anja Aberle<br />
Daniel Ball<br />
Sin Von Chan<br />
Suzie Hunter<br />
Daniel Kong<br />
Undergraduate Thesis Students (4th year):<br />
Sin Von Chan (since 07/05)<br />
Robyn Hyslop<br />
Matthew Linney<br />
Roland Utama<br />
Visiting Undergraduate Thesis Students (final year):<br />
Stefan Vetter (10/05 to 03/06)<br />
Mathieu Larose (06/05 to 09/05)<br />
Vincent Vogt (02/05 to 09/05)<br />
Shervin Motahar (until 03/05)<br />
Undergraduate Project Students (2nd year):<br />
Peter Gress<br />
James Holgate<br />
Tokushin Tanaka<br />
Chee Hong Thoo<br />
Undergraduate Taste <strong>of</strong> Research Summer Scholars (3rd year):<br />
Sin Von Chan (2004/05)<br />
Cuong Dang (2005/06)<br />
Alexander Fattal (2005/06)<br />
Yicheng “Sherman” Zhou (2005/06)<br />
51
4.4.1 Summary<br />
The primary aim <strong>of</strong> the Centre’s Thin-Film Group (or “Second-Generation Group”) is to<br />
develop polycrystalline silicon (poly-Si) thin-fi lm solar cells on glass, an approach that is widely<br />
recognised as being a pathway towards substantially lowering the cost <strong>of</strong> photovoltaic (PV)<br />
solar electricity. Previous work by the Group has led to the invention <strong>of</strong> three poly-Si thinfi<br />
lm solar cells on glass (ALICIA, EVA, ALICE), a glass texturing method (AIT), and a solar cell<br />
metallisation and interconnection method. Based on these earlier results, the Group’s main<br />
areas <strong>of</strong> research in 2005 have been:<br />
• Optimisation and upscaling <strong>of</strong> the Aluminium-Induced Texture (AIT) glass texturing<br />
method<br />
• Optimisation and upscaling <strong>of</strong> poly-Si seed layers formed on glass by the Aluminium-<br />
Induced Crystallisation (AIC) method<br />
• Improvement <strong>of</strong> the open-circuit voltages <strong>of</strong> the Group’s solar cells<br />
• Optimisation <strong>of</strong> optical absorption in our solar cells<br />
• Optimisation <strong>of</strong> the Group’s cell metallisation and interconnection methods<br />
• Improvement <strong>of</strong> the effi ciency <strong>of</strong> the Group’s solar cells<br />
Voc [mV]<br />
600<br />
500<br />
CSG<br />
400<br />
7%<br />
8%<br />
4%<br />
300<br />
200<br />
EVA<br />
100<br />
ALICIA<br />
0<br />
Jan-00 Jan-01 Jan-02 Jan-03<br />
Year<br />
9%<br />
ALICE<br />
Jan-04 Jan-05 Jan-06<br />
Progress with the voltages <strong>of</strong> the Group’s<br />
thin-fi lm solar cells during 2005 has been<br />
very good, as shown in Figure 4.4.1. A<br />
momentous milestone that was achieved in<br />
2005 was the realisation <strong>of</strong> our fi rst poly-Si<br />
thin-fi lm solar cells on glass with voltages<br />
<strong>of</strong> over 500 mV. The best cell has a voltage<br />
<strong>of</strong> 517 mV which, to our knowledge, is a<br />
new world record for homojunction poly-Si<br />
thin-fi lm solar cells on glass. This result was<br />
obtained with the EVA structure. Another<br />
important milestone was the achievement<br />
<strong>of</strong> an internal quantum effi ciency <strong>of</strong> over<br />
70%, obtained using the ALICIA structure.<br />
These advances benefi ted signifi cantly<br />
from an improved hydrogenator (remote<br />
plasma machine) that became available<br />
in early 2005 in the Centre’s cleanroom<br />
facility at Botany.<br />
Figure 4.4.1: Evolution <strong>of</strong> the open-circuit voltages Voc <strong>of</strong> the poly Si thin-fi lm solar cells on<br />
glass (ALICIA, EVA, ALICE) under development in the Centre’s Thin-Film Group. Also shown, for<br />
comparison, is the Voc <strong>of</strong> CSG Solar’s poly-Si on glass technology.<br />
Good progress was also made in 2005 with optimising and upscaling <strong>of</strong> both our AIT glass<br />
texturing method and our AIC seed layer process. As a result <strong>of</strong> this work, AIT-textured glass<br />
sheets and AIC-seeded glass sheets with a size <strong>of</strong> 10x10 cm 2 are now routinely produced by<br />
the Group.<br />
Compared to solar cells on planar glass superstrates, the use <strong>of</strong> AIT-textured glass has been<br />
shown in 2005 to drastically improve the absorption <strong>of</strong> infrared light. This is due to two<br />
effects, reduced refl ection losses at the illuminated cell surface and trapping <strong>of</strong> infrared light<br />
due to the oblique path <strong>of</strong> light within the cells. The current gain possible from this improved<br />
optical absorption is larger than 30%.<br />
52
Another highlight <strong>of</strong> the Group in 2005 was the establishment <strong>of</strong> the Device Fabrication<br />
Team. This step has resulted from the good progress with device voltages and the growing<br />
need for fully metallised solar cells. This has helped to improve the effi ciency <strong>of</strong> ALICIA solar<br />
cells to 3% in 2005.<br />
The Group was again active in generating intellectual property, with the fi ling <strong>of</strong> provisional<br />
Australian patent 2005901285 on an interdigitated thin-fi lm cell metallisation method<br />
(“SAMPL”) and the entering <strong>of</strong> the national phase for two recent PCT patent applications.<br />
Furthermore, the Group has published one book chapter, 11 journal papers, and 21<br />
conference papers in 2005. Fifteen <strong>of</strong> these papers were presented as talks at international<br />
conferences, including four invited talks by group leader Armin Aberle.<br />
The Group’s thin-fi lm solar cell technologies are attracting a signifi cant amount <strong>of</strong> interest<br />
from the solar industry and the venture capital sector. In response to this interest, the Group<br />
has created a commercialisation roadmap in 2005 involving a business plan proposal and<br />
layouts <strong>of</strong> a pilot line and a fi rst factory. UNSW’s commercial arm, <strong>New</strong><strong>South</strong> Innovations Ltd,<br />
has entered negotiations with several interested parties.<br />
4.4.2 Background Information on Thin-Film PV<br />
It is widely agreed that thin-fi lm photovoltaics has the potential to generate solar electricity at<br />
much lower cost ($/kWh) than is possible with a silicon wafer based technology. Silicon thin<br />
fi lms on foreign supporting materials are particularly interesting due to their non-toxic nature,<br />
the abundance <strong>of</strong> silicon, the huge technological experience with silicon in the microelectronics<br />
industry, and silicon’s well-matched bandgap energy with respect to photovoltaic conversion<br />
<strong>of</strong> sunlight. Various high-temperature as well as low-temperature approaches are currently<br />
being investigated internationally to explore how suitable silicon thin-fi lms can be fabricated<br />
on foreign supporting materials [4.4.1 - 4.4.5] or can be transferred onto such materials<br />
[4.4.6 - 4.4.8]. The silicon thin-fi lm materials presently being explored for PV applications<br />
include amorphous silicon (a-Si:H), microcrystalline silicon (µc-Si:H; grain size g < 1 µm),<br />
polycrystalline silicon (poly-Si; g > 1 µm) and single crystalline silicon. Glass is a particularly<br />
promising supporting material as it is cheap, transparent (enabling superstrate and/or<br />
bifacial modules), long-term stable, and readily available. The drawback with glass sheets<br />
is their limited thermal stability, excluding the use <strong>of</strong> lengthy high-temperature (> 650°C)<br />
processing steps during solar cell fabrication.<br />
The Thin-Film Group’s work focuses on polycrystalline silicon fi lms because we believe that, on<br />
a life-cycle basis, PV modules using this material will, in the not-too-distant future, produce the<br />
cheapest PV electricity. This assessment is based on the assumption that poly-Si thin-fi lm PV<br />
modules will have the same excellent long-term stability as today’s standard Si wafer-based<br />
PV modules. One <strong>of</strong> the advantages <strong>of</strong> poly-Si over a Si:H and µc-Si:H materials is the much<br />
higher (orders <strong>of</strong> magnitude) lateral conductance <strong>of</strong> (heavily) doped thin fi lms. Therefore,<br />
no transparent conductive oxide (TCO) is required on the illuminated solar cell surface for<br />
cell metallisation, enabling potentially very inexpensive metallisation and interconnection<br />
methods.<br />
53
The most effi cient poly-Si thin-fi lm solar cells made as yet at low temperature on a foreign<br />
supporting material have an effi ciency <strong>of</strong> 9% and were made by solid phase crystallization<br />
(SPC) <strong>of</strong> PECVD-deposited amorphous silicon [4.4.9, 4.4.10]. The CSG technology [4.4.10] <strong>of</strong><br />
UNSW spin-<strong>of</strong>f company CSG Solar is particularly attractive as it uses glass as the supporting<br />
material and has already been tested extensively at the pilot line level. Pilot sized CSG modules<br />
(aperture area <strong>of</strong> about 600 cm 2 ) have effi ciencies <strong>of</strong> over 8% and open-circuit voltages <strong>of</strong><br />
close to 500 mV/cell. The CSG devices have excellent light trapping properties, enabling<br />
25 mA/cm 2 <strong>of</strong> short-circuit current with 1.5 µm thin Si fi lms [4.4.10]. This shows that poly-<br />
Si thin-fi lm solar cells on glass merely require a minority carrier diffusion length <strong>of</strong> a few<br />
microns in the absorber region to achieve over 8% effi ciency. The CSG technology is ready<br />
for mass production and the fi rst factory has its <strong>of</strong>fi cial opening in March 2006 in Thalheim,<br />
Germany.<br />
4.4.3 Motivation Behind Group’s Work<br />
The Thin-Film Group’s research aims at the realisation <strong>of</strong> novel poly-Si thin-fi lm solar cells on<br />
glass. One aim is to establish potentially cheaper Si deposition methods than PECVD ( ~ 30<br />
€/m 2 for 1.5 µm thick SPC poly-Si fi lms [4.4.10]) for such cells. Another aim is to realise poly-<br />
Si fi lms with better electronic quality (larger grain size, longer diffusion length, etc) compared<br />
to the existing technologies, enabling higher cell effi ciency. Cell thickness is generally in the<br />
range 1-3 µm.<br />
4.4.4 Challenges with C-SI Thin-film Cells<br />
The main challenges associated with crystalline silicon thin-fi lm solar cells on glass are:<br />
• Glass is the ideal supporting material (cheap, long-term stable, transparent), but<br />
doesn’t permit lengthy high-temperature steps. Thus, a low-thermal-budget Si<br />
formation method is required that is cheap but nevertheless gives good-quality<br />
c-Si.<br />
• Glass is amorphous and hence c-Si on glass is polycrystalline. Thus, a cheap and<br />
effi cient method for passivating grain boundaries (and bulk defects) is required.<br />
• Crystalline silicon is a rather poor absorber <strong>of</strong> infrared light. To maintain a high<br />
current in thin cells, an excellent light trapping scheme is required.<br />
• The fabricated thin-fi lm material needs to be scribed into individual cells and<br />
these need to be metallised and interconnected. Thus, an inexpensive cell<br />
metallisation and interconnection method is required.<br />
4.4.5 How We Tackle These Challenges<br />
While tempered soda lime glass is the cheapest PV-compatible glass (about 10 €/m 2 ), we<br />
use a more expensive fl oat glass (Bor<strong>of</strong>l oat33 from Schott AG, Germany, about 25 €/m 2<br />
[4.4.11]) because this glass has a better thermal stability and a thermal expansion coeffi cient<br />
that is better matched to silicon. As a consequence, cell processing can occur at higher<br />
temperatures and/or with higher thermal budget, providing scope for signifi cantly improved<br />
Si material quality. To achieve light trapping, we texture the Si-facing surface <strong>of</strong> the glass plate<br />
with the AIT (Aluminium-Induced Texture) method recently invented by us (patent application<br />
WO 2004/089841 A1).<br />
Low-temperature poly-Si formation on glass is realised using three different methods we<br />
have developed. The resulting solar cells are called ALICIA, EVA, and ALICE. The fi rst two cells<br />
are made by vacuum evaporation <strong>of</strong> Si in a non-ultra-high vacuum (non-UHV) environment,<br />
54
whereas ALICE cells can be made by either vacuum evaporation or PECVD. Evaporation<br />
in a non-UHV environment is interesting for PV as it is a particularly fast and potentially<br />
inexpensive Si deposition method. It also has an excellent Si source material usage and avoids<br />
the use <strong>of</strong> toxic gases. Because our Si evaporator is not equipped with a vacuum annealer,<br />
no evaporated ALICE cells have been made in 2005 and instead PECVD was used for ALICE<br />
work. For comparative reasons, we are also making solar cells by SPC <strong>of</strong> PECVD-deposited<br />
a-Si (“PLASMA” cells). These are structurally identical to EVA cells, however, they are made by<br />
PECVD instead <strong>of</strong> evaporation.<br />
Rapid thermal annealing is used to<br />
reduce the density <strong>of</strong> point defects in<br />
our cells and to increase the fraction <strong>of</strong><br />
electrically active dopants. Grain boundary<br />
and bulk defect passivation is realised by<br />
hydrogen passivation, using either direct<br />
plasma excitation (parallel-plate reactor)<br />
or remote plasma excitation.<br />
A range <strong>of</strong> methods is being explored<br />
by us for cell metallisation and interconnection,<br />
including shadow mask<br />
evaporation <strong>of</strong> aluminium and the use <strong>of</strong><br />
photolithographic lift-<strong>of</strong>f processes.<br />
Figure 4.4.2 shows an overview <strong>of</strong> the<br />
fabrication sequence <strong>of</strong> the thin-fi lm<br />
solar cells on glass investigated by us in<br />
2005.<br />
a-Si<br />
EVA<br />
PLASMA<br />
glass<br />
SiN<br />
Crystallisation<br />
RTA<br />
a-Si<br />
ALICE<br />
Seed<br />
(AIC)<br />
c-Si<br />
(by IAD)<br />
ALICIA<br />
Figure 4.4.2: Process sequence <strong>of</strong><br />
the poly-Si thin-fi lm solar cells on glass<br />
investigated by the Thin-Film Group<br />
in 2005. All cells are designed for<br />
the superstrate confi guration, i.e. the<br />
sunlight enters the solar cells through<br />
the glass.<br />
Hydrogenation<br />
Metallisation<br />
4.4.6 Glass Texturing<br />
We are aiming at producing polysilicon thin-fi lm solar cells on a glass superstrate, with<br />
evaporation as a possible silicon deposition method. Thus, a glass texturing method that<br />
produces a surface morphology that is compatible with “line-<strong>of</strong>-sight” silicon deposition is an<br />
advantage. Use <strong>of</strong> a textured conducting oxide is ruled out because <strong>of</strong> the high temperatures<br />
involved in our cell fabrication process. We are using Schott Bor<strong>of</strong>l oat33 glass because<br />
its thermal expansion matches that <strong>of</strong> silicon quite well. Recently we have invented a glass<br />
texturing process called Aluminium-Induced Texture (AIT), in which an Al fi lm evaporated onto<br />
planar glass is annealed in an inert atmosphere. High temperature transfers oxygen in a<br />
redox reaction from the glass, largely as SiO 2 , to the aluminium. A surface texture is imparted<br />
to the glass according to the nucleation conditions provided during this anneal. Careful control<br />
<strong>of</strong> the HF-nitric acid based oxidation-dissolution process used to remove the reduced silicon<br />
also affects texture morphology, since dissolution consumes both the freshly oxidised silicon<br />
and SiO 2 in the glass.<br />
55
During 2005, we have devoted signifi cant effort to optimising and upscaling <strong>of</strong> the AIT glass<br />
texturing process. We found that considerable freedom is available for varying conditions<br />
used in the AIT process, and that these conditions affect the morphology <strong>of</strong> the resulting<br />
texture. Prime determinants are the Al layer thickness, anneal temperature pr<strong>of</strong>i le, as well<br />
as HF-nitric acid ratio and immersion time. For example, Fig. 4.4.3 shows FIB microscope<br />
images <strong>of</strong> the surface (plus cross section) <strong>of</strong> 2.3 µm thick SPC poly-Si fi lms deposited by<br />
PECVD onto three different AIT-textured glass sheets. All were AIT-annealed at a maximum<br />
temperature <strong>of</strong> 630°C, with differing temperature pr<strong>of</strong>i les. The Al thickness and HF-nitric<br />
mixture are varied as follows: 200 nm and 1:20 for sample (a), 400 nm and 1:1 for sample<br />
(b), 600 nm and 1:20 for sample (c). The HF-nitric immersion time was kept to the minimum<br />
needed to remove the silicon residue.<br />
All samples had a similar Voc <strong>of</strong> about 400 mV after the same SPC crystallisation, rapid<br />
thermal anneal (RTA) and hydrogenation steps, despite the large variation in surface<br />
morphology. In another experiment, AIC seed layers were prepared on the above three AIT<br />
textures. The Al was evaporated, whereas the a-Si was deposited by sputtering. Only sample<br />
(b) was free <strong>of</strong> pinholes, indicating its promising potential for general use with evaporated<br />
silicon solar cells.<br />
Sample (a) Sample (b) Sample (c)<br />
Figure 4.4.3: Microscope images (with milled trenches to reveal the samples’ cross sections)<br />
<strong>of</strong> 2.3 µm thick PECVD-based SPC poly-Si fi lms on three AIT-textured glass sheets. The<br />
samples’ structure is: 3.3 mm glass /70 nm SiN /100 nm n++ /2100 nm p- /100 nm p+.<br />
Figure 4.4.4 shows the total measured optical absorption 1-R-T (where R is the sample’s<br />
refl ection and T is its transmission) for the three poly-Si samples, along with that for a planar<br />
PECVD-based poly-Si fi lm with the same layer structure and treatment, in the superstrate<br />
confi guration. These curves are compared with a randomising texture with zero refl ection<br />
at the glass-silicon interface and no back refl ector (top curve; calculated with the s<strong>of</strong>tware<br />
WVASE). In these measurements, light was collected by an integrating sphere and the<br />
Edwards mount was used. In this arrangement, the sample is placed in the centre <strong>of</strong> the<br />
integrating sphere and the beam is collimated to illuminate only a small, central area <strong>of</strong> the<br />
sample. All light emerging from the sample is collected, including R, T, and light from the<br />
edges <strong>of</strong> the glass (the latter can be signifi cant for textured samples at weakly absorbed<br />
wavelengths, particularly with thick glass). Because edge leakage reduces coupling <strong>of</strong> light<br />
with the sample, the absorption measured at these wavelengths for the textured samples<br />
underestimates what would be absorbed in a larger solar cell module. Thus, we can say that<br />
the centremount-derived curves in Fig. 4.4.4 represent a conservative estimate <strong>of</strong> the true<br />
absorption for the textured samples.<br />
56
To illustrate the signifi cance <strong>of</strong> edge leakage, an absorption curve is also shown in Fig. 4.4.4<br />
for sample (a), based on separate R and T measurements with the beam similarly illuminating<br />
a small central area only. Here, edge leakage is not collected by the sphere and absorption is<br />
overestimated. The true value <strong>of</strong> absorption therefore lies between these two limits.<br />
Figure 4.4.4: Measured absorption<br />
in the three 2.3 µm thick AITtextured<br />
poly-Si samples shown in<br />
Figure 4.4.3. Absorption attained<br />
in decreasing order in the midband<br />
range (around 800 nm) is for<br />
sample (a) (separate R and T scans),<br />
then (a), (c) and (b) (combined R+T<br />
scans). Also shown is absorption<br />
<strong>of</strong> a planar control sample<br />
(lowest curve, combined R+T<br />
scan) and calculated absorption<br />
with random scatter (top curve).<br />
Absorption<br />
1<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
planar<br />
0<br />
400 500 600 700 800 900 1000 1100 1200<br />
Wavelength (nm)<br />
Random scattering<br />
(calculated)<br />
Comparisons <strong>of</strong> absorption in the measured samples are only relevant where the Si fi lm<br />
absorbs signifi cantly more than the glass sheet. In the random model, the silicon and glass<br />
absorb equally at around 950 nm. This means that comparisons <strong>of</strong> absorption for each<br />
texture are best limited to the band 500-900 nm. Here, our AIT-textured samples absorb<br />
midway between the planar and random values, see Fig. 4.4.4. Although absorption can be<br />
seen to increase with sample roughness, the difference is not very large. This is important<br />
because crystallisation tends to favour a low level <strong>of</strong> roughness with evaporated material. At<br />
long wavelengths, the textured samples appear to absorb more than the random model. We<br />
assume this is due to the WVASE modelling relying upon a T scan which includes only a small<br />
amount <strong>of</strong> absorption at these wavelengths. This high uncertainty in absorption measurement<br />
is passed on to the k value <strong>of</strong> the silicon, and this uncertainty is further magnifi ed by pathlength<br />
enhancement in the random model. Long-wavelength absorption measured in the samples<br />
is attributed to either defects and grain boundaries, or modifi cation <strong>of</strong> the glass properties<br />
during RTA and/or hydrogenation.<br />
As an overall estimate <strong>of</strong> the benefi t from texturing, we have calculated the 1-Sun Jsc available<br />
from the absorption in the wavelength range 400-900 nm. For the planar sample in Fig. 4.4.4<br />
this is 17.3 mA/cm 2 , whereas sample (b) provides 22.4 mA/cm 2 , a boost <strong>of</strong> 30%. It should<br />
be noted that these estimations <strong>of</strong> current densities are obtained without a back refl ector.<br />
In summary, good progress has been obtained in 2005 with fi nding AIT process parameters<br />
that provide a texture topography that is well suited for both PECVD-based and evaporationbased<br />
poly-Si fi lms. Furthermore, good progress has been made with upscaling <strong>of</strong> the process<br />
to larger glass sheets. As a result <strong>of</strong> this work, AIT-textured glass sheets with a size <strong>of</strong> 10x10<br />
cm 2 are now routinely produced by the Group.<br />
57
4.4.7 Optimisation <strong>of</strong> AIC Seed Layers<br />
One method to produce a poly-Si seed layer on glass that has received a lot <strong>of</strong> attention<br />
recently is the UNSW-developed Aluminum-Induced Crystallization (AIC) method [4.4.12,<br />
4.4.13], where Al is used as a catalyst for a low-temperature crystallization <strong>of</strong> amorphous<br />
silicon (a-Si). Our AIC bi-layer process on glass is performed by vacuum evaporating Al onto<br />
SiN-coated glass, followed by an exposure to air for oxide and hydroxide growth on the Al<br />
surface. Amorphous silicon is then deposited by sputtering or PECVD onto the oxidized Al<br />
layer. The sample is then annealed for about 12 hours at about 450°C, followed by a patented<br />
surface preparation method that removes the Al, the Al oxide/hydroxide, and the island-like<br />
silicon regions formed on top <strong>of</strong> the continuous AIC poly-Si fi lm.<br />
In 2005, we performed a detailed optimisation study <strong>of</strong> AIC seed layers formed from PECVDdeposited<br />
a-Si:H. We found that a pure silane plasma for a-Si:H deposition results, during<br />
the AIC anneal, in blisters in the a-Si:H fi lm and localised peeling-<strong>of</strong>f <strong>of</strong> the fi lm. The problem<br />
with blistering in a-Si:H fi lms, believed to be connected to hydrogen out-gassing during<br />
crystallisation, was minimised by the use <strong>of</strong> an argon-diluted silane plasma.<br />
Figure 4.4.5 shows an AIC poly-Si fi lm deposited with an argon-diluted silane plasma, after Al<br />
removal and surface preparation. A grain-induced contrast is visible in this FIB microscope<br />
image, indicating a grain size in the order <strong>of</strong> 20 µm. Figure 4.4.6 displays the UV refl ectance<br />
measured on this sample, together with that <strong>of</strong> a polished single crystalline Cz silicon wafer.<br />
[Remark: Optical simulations show that for a poly-Si fi lm thicker than 70 nm, the interference<br />
effect will not affect the measurements at 367 nm, the position <strong>of</strong> the E1 peak in c-Si.] This<br />
poly-Si fi lm has the highest UV refl ectance fi gure <strong>of</strong> merit ever reported for an AIC poly-Si fi lm<br />
on glass (98.9%, compared to 100% for perfect c-Si material).<br />
Furthermore, good progress has been made with upscaling <strong>of</strong> the AIC process to larger glass<br />
sheets. As a result <strong>of</strong> the work in 2005, high-quality AIC seeded glass sheets with a size <strong>of</strong><br />
10x10 cm 2 are now routinely produced by the Group.<br />
Figure 4.4.5: FIB microscope image <strong>of</strong> a state-<strong>of</strong>the-art<br />
AIC poly-Si seed layer on glass. The sample<br />
structure is glass / SiN / AIC poly-Si (100 nm).<br />
75<br />
Reflectance (%)<br />
70<br />
65<br />
60<br />
55<br />
Cz Si wafer<br />
AIC poly-Si<br />
Figure 4.4.6: Measured UV refl ectance <strong>of</strong><br />
the AIC poly-Si sample <strong>of</strong> Fig. 4.4.5. Also<br />
shown for comparison is the measured UV<br />
refl ectance <strong>of</strong> a polished Cz Si wafer.<br />
50<br />
45<br />
210 240 270 300 330 360 390<br />
Wavelength (nm)<br />
58
4.4.8 Optimisation <strong>of</strong> Evaporated Solar Cells<br />
A. EVA solar cells<br />
EVA solar cells are made by “solid phase crystallisation <strong>of</strong> EVAporated silicon”. The deposition<br />
<strong>of</strong> the amorphous silicon occurs at about 200°C in a non-ultra-high vacuum environment<br />
( ~ 10-7 Torr) onto silicon nitride (SiN) coated Bor<strong>of</strong>l oat33 glass sheets. To realise the<br />
desired dopant pr<strong>of</strong>i les, phosphorus, boron and/or gallium are added in-situ (i.e., during the<br />
deposition) via high-temperature effusion cells. After the silicon evaporation, the amorphous<br />
structure is crystallised at 600°C in a nitrogen-purged tube furnace. The grain size <strong>of</strong> the<br />
resulting cells is in the order <strong>of</strong> 1 µm, and the Si thickness is in the range <strong>of</strong> 1.5 to 2 µm.<br />
A schematic <strong>of</strong> a crystallised EVA cell is shown in Fig. 4.4.7. A rapid thermal anneal (RTA)<br />
process (at about 900°C for several minutes) activates dopants and anneals point defects,<br />
and a hydrogen passivation (at about 600°C for about 15 minutes) decreases the defect<br />
density drastically both inside the grains and at grain boundaries, and thereby enhances the<br />
electrical performance signifi cantly (see Section 4.4.10 for more details).<br />
p+<br />
p-<br />
n+<br />
SiN<br />
Glass<br />
Fig. 4.4.7: Structure <strong>of</strong> an EVA solar cell on glass (not to scale).<br />
Fig. 4.4.8: PhD student Daniel Inns<br />
inspecting EVA poly-Si thin-fi lm solar cell<br />
material fabricated on a planar sheet <strong>of</strong><br />
glass.<br />
In 2005, the voltage <strong>of</strong> EVA solar cells has<br />
been improved by optimisation <strong>of</strong> several<br />
process variables. First, an increase <strong>of</strong><br />
the substrate temperature during the<br />
hydrogenation step to about 600°C has<br />
led to higher hydrogen concentrations in<br />
the fi nished fi lms, which clearly increases<br />
the voltages. Second, boron was used for<br />
the base doping instead <strong>of</strong> gallium. Third,<br />
preliminary experiments were performed<br />
towards improved crystal growth via precrystallisation<br />
treatments.<br />
59
As a result <strong>of</strong> these modifi cations, the open-circuit voltage <strong>of</strong> EVA cells has been improved to<br />
517mV in 2005. This is believed to be a new world record for homojunction poly-Si thin-fi lm<br />
solar cells on glass. The light intensity versus open-circuit voltage characteristics (“Suns-Voc<br />
curve” [4.4.14]) <strong>of</strong> the 517-mV poly-Si thin-fi lm solar cell is displayed in Fig. 4.4.9 (left). A twodiode<br />
model fi t (oblique straight lines in the graph) reveals that the cell is clearly limited by n=2<br />
recombination. The Suns-Voc curve can be converted into a “pseudo current-voltage curve”,<br />
as shown in the graph on the right hand side <strong>of</strong> Fig. 4.4.9. This curve has a good fi ll factor FF<br />
<strong>of</strong> 0.71.<br />
Fig. 4.4.9: Suns-Voc curve <strong>of</strong> the 517-mV EVA solar cell (left) and the equivalent pseudo<br />
current-voltage curve (right).<br />
The dopant concentration <strong>of</strong> the a-Si in our research-style silicon evaporator depends<br />
strongly on the location <strong>of</strong> the sample within the evaporator during the deposition. This would<br />
be undesirable in solar cell manufacturing lines, but can be exploited in the laboratory for<br />
optimising the dopant concentrations in the fi nished solar cells. As an example, Fig. 4.4.10<br />
compares the contour plots <strong>of</strong> the emitter dopant concentration (left) and the measured<br />
open-circuit voltage <strong>of</strong> fi nished EVA cells (right). The black squares indicate the positions where<br />
measurements were taken. A clear correlation can be seen, suggesting that the emitter<br />
doping level in these EVA solar cells needs to be reduced to enhance the voltage <strong>of</strong> the cells.<br />
This is very likely due to poorer crystallisation, since high dopant concentrations are known to<br />
cause an increased nucleation rate and thus a smaller grain size. A corresponding analysis<br />
<strong>of</strong> the boron base doping has revealed that the best Voc performance was obtained for the<br />
highest boron doping investigated in this set <strong>of</strong> experiments.<br />
position y<br />
30<br />
a<br />
b c d<br />
20 0.8<br />
1.0 2.0 3.0<br />
3.0<br />
10<br />
3.4<br />
h<br />
f<br />
e<br />
g<br />
0<br />
-10<br />
i j k l<br />
-20<br />
m n o p<br />
-30<br />
-30 -20 -10 0 10 20 30<br />
position x<br />
position y<br />
30<br />
20<br />
10<br />
0<br />
-10<br />
-20<br />
470<br />
440<br />
-30<br />
-30 -20 -10 0 10 20 30<br />
position x<br />
410<br />
380<br />
Fig. 4.4.10: Contour plots <strong>of</strong> (left) the relative emitter doping density (as determined by SIMS<br />
and sheet resistance measurements) and (right) the measured Voc <strong>of</strong> the resulting EVA solar<br />
cells.<br />
60
Secondary ion mass spectroscopy (SIMS) measurements were performed externally to<br />
investigate the dopant and impurity pr<strong>of</strong>i les in a fi nished EVA cell, see Fig. 4.4.11. The dopant<br />
pr<strong>of</strong>i les for both boron and phosphorus are satisfying; however, despite the good voltage<br />
<strong>of</strong> the cell, the impurity levels <strong>of</strong> carbon (C) and oxygen (O) are surprisingly high. Tails <strong>of</strong><br />
contaminants close to the air-cell interface indicate high carrier recombination in this region,<br />
and thus a small diffusion length. This will most severely affect the minority carrier collection<br />
and hence the short-circuit current <strong>of</strong> the cell. More work is required to fi nd and eliminate the<br />
source <strong>of</strong> these contaminants, and to thereby further improve these solar cells.<br />
1E21<br />
Cell-SiN interface<br />
Fig. 4.4.11: SIMS analysis <strong>of</strong> the<br />
dopants (B, P) and the contaminants<br />
O, C, and N in a recent EVA solar<br />
cell.<br />
concentration (atoms/cm -3 )<br />
1E+2<br />
1E+1<br />
1E+1<br />
1E+1<br />
N<br />
C<br />
O<br />
B. ALICIA Solar Cells<br />
B<br />
P<br />
1E16<br />
0 0.2. 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8<br />
ALICIA stands for “ALuminium-Induced Crystallisation Ion-Assisted deposition” [4.4.15]. The<br />
idea behind the patented ALICIA solar cell technology is to directly (i.e., epitaxially) grow the<br />
crystalline absorber layer on a hydrogen-terminated seed layer made by Aluminium-Induced<br />
Crystallisation (AIC) on glass. The AIC precursor structure is glass / SiN (100 nm) / Al<br />
( ~ 200 nm, evaporated) / a-Si ( ~ 300 nm, sputtered). This structure is annealed for 12 hours<br />
at about 450°C in a tube furnace at atmospheric pressure and then the Al and the excess Si<br />
are removed. The resulting average grain size is about 10 to 20 µm. For low-temperature Si<br />
epitaxy we use IAD (Ion-Assisted Deposition) because this method is capable <strong>of</strong> high-rate Si<br />
growth at low (i.e., borosilicate glass compatible) temperatures <strong>of</strong> about 600°C. Figure 4.4.12<br />
schematically shows a Mesa-type ALICIA solar cell. The two IAD-grown poly-Si layers have<br />
a combined thickness <strong>of</strong> about 2 µm and are deposited in less than 30 minutes (this time<br />
includes sample heating, Si deposition, cooling and unloading <strong>of</strong> the sample). In recent years<br />
we have developed an IAD process that is capable <strong>of</strong> achieving good-quality epitaxial Si in a non<br />
-UHV environment [4.4.16]. This makes IAD Si epitaxy potentially suitable for the PV industry.<br />
p+ layer<br />
Rear metal<br />
n+ layer<br />
Absorber<br />
layer<br />
Seed layer<br />
Glass<br />
Sunlight<br />
V, I<br />
Front metal<br />
SiN<br />
One limitation <strong>of</strong> the ALICIA cell to date has been poor<br />
light absorption. Due to the fl at surfaces <strong>of</strong> the planar<br />
glass superstrates used so far, a great deal <strong>of</strong> light<br />
refl ects from the front surface <strong>of</strong> the cell, and this is<br />
a severe limitation for the cell’s effi ciency. In addition,<br />
a large fraction <strong>of</strong> low-energy light which does enter<br />
the cell reaches the back side, and can be transmitted<br />
through. An aluminium fi lm has been used to date to<br />
refl ect this light back into the cell. This should allow light<br />
which is usually transmitted to refl ect back through<br />
the cell before being lost, giving an increase in light<br />
absorption and cell current.<br />
Figure 4.4.12: Schematic <strong>of</strong> a Mesa-type ALICIA solar<br />
cell realised by shadow-mask epitaxy (not to scale).<br />
61
In order to trap additional light, texturing is required. This enables light to be scattered so<br />
that it travels diagonally through the cell, giving a longer pathlength before and after the light<br />
hits the back surface. The AIT glass texturing method (see Section 4.4.6) has been used for<br />
making textured glass substrates. The rough surface <strong>of</strong> the resulting textured AIC seed layer<br />
made on such glass can be seen in fi gure 4.4.13(A). Also shown (see plot B) is a cross section<br />
<strong>of</strong> the fi nished ALICIA cell, which was grown on top <strong>of</strong> the textured seed layer.<br />
Using the textured glass sheets gives a large decrease in refl ection, particularly for infrared<br />
light, as shown in the left graph <strong>of</strong> fi gure 4.4.14. Considering that the rear surface <strong>of</strong> both<br />
cells is completely covered by Al, no light can be transmitted through the cell. The increase<br />
in external quantum effi ciency (EQE) from the texture is shown in the right graph <strong>of</strong> fi gure<br />
4.4.14, where the longer-wavelength light is collected better by the textured cell. Figure 4.4.14<br />
also shows that the short-wavelength EQE is better for planar cells – this is due to the planar<br />
cell having a thinner emitter. In this case, the texture added 12% to the short-circuit current<br />
compared to the planar device. Modelling suggests that if the emitter <strong>of</strong> the textured cell<br />
were as thin as that <strong>of</strong> the planar cell, this texture would have given a current improvement<br />
<strong>of</strong> 35%. A diffuse refl ector at the rear <strong>of</strong> the cell would boost the current further, as the<br />
pathlength <strong>of</strong> long-wavelength light would be increased further.<br />
Future work will continue to focus on the use <strong>of</strong> textures with other light trapping features like<br />
diffuse back refl ectors.<br />
Figure 4.4.13 A) Microscope<br />
image (with cross section) <strong>of</strong> a<br />
textured AIC poly-Si seed layer<br />
on glass. The viewing angle is<br />
about 45°. The SiN fi lm glows<br />
white, and the grey silicon fi lm is<br />
seen to follow the rough glass<br />
surface. The texture dimples can<br />
be seen on the top surface. B)<br />
Microscope image <strong>of</strong> the cross<br />
section <strong>of</strong> the resulting textured<br />
ALICIA cell.<br />
Figure 4.4.14: Comparison<br />
<strong>of</strong> the refl ection and<br />
external quantum effi ciency<br />
(EQE) <strong>of</strong> a planar and a<br />
textured ALICIA cell. The<br />
texture reduces refl ection<br />
at the illuminated surface<br />
and boosts light trapping in<br />
the Si fi lm. As a result, the<br />
long-wavelength EQE <strong>of</strong> the<br />
textured cell is better than<br />
that <strong>of</strong> the planar cell.<br />
Reflection (%)<br />
100<br />
80<br />
60<br />
40<br />
20<br />
0<br />
Planar<br />
Textured<br />
400 600 800 1000 1200<br />
Wavelength (nm)<br />
EQE (%)<br />
50<br />
40<br />
30<br />
20<br />
10<br />
0<br />
Textured<br />
Planar<br />
400 500 600 700 800 900 1000 1100<br />
Wavelength (nm)<br />
62<br />
In addition to the work on textured surfaces, variation <strong>of</strong> the doping in the absorber region<br />
<strong>of</strong> ALICIA cells has been investigated in 2005. The natural doping <strong>of</strong> the absorber region <strong>of</strong><br />
these cells was found to be n-type, due to a background phosphorus and oxygen content in the<br />
silicon evaporator. Boron has been added to reduce the n-type base doping in the device, and<br />
the effi ciency <strong>of</strong> ALICIA cells has been improved to 3% using this approach. The measured<br />
current-voltage and power-voltage curves <strong>of</strong> this device are shown in fi gure 4.4.15. The best<br />
voltage <strong>of</strong> ALICIA cells has been improved to 441 mV in 2005.
Current Density (mA/cm 2 )<br />
16<br />
14<br />
12<br />
10<br />
8<br />
6<br />
4<br />
2<br />
Jsc = 14.8mA/cm 2<br />
Voc = 403mV<br />
FF = 0.51<br />
Eff. = 3.0%<br />
4<br />
Power density (mW/cm 2 )<br />
3<br />
2<br />
1<br />
Figure 4.4.15: Current-voltage<br />
and power-voltage curves <strong>of</strong><br />
the best ALICIA solar cell made<br />
as yet. The glass substrate is<br />
planar.<br />
0<br />
0<br />
0 100 200 300 400 500<br />
Voltage (mV)<br />
4.4.9 Optimisation <strong>of</strong> PECVD Alice Solar Cells<br />
ALICE stands for “ALuminium-Induced Crystallisation solid-phase Epitaxy”. The idea behind the<br />
patented ALICE solar cell technology is to deposit the absorber onto an H-terminated AIC<br />
seed layer at very low temperature ( ~ 200°C) as amorphous material and then to crystallise<br />
the amorphous material in a thermal annealing step at elevated temperature (570-600°C).<br />
This method is related to solid-phase crystallisation (SPC), however, because <strong>of</strong> the presence<br />
<strong>of</strong> a crystalline seed layer, it is actually a solid-phase epitaxy (SPE) process. The key feature<br />
in the ALICE process is a crystallographic transferral <strong>of</strong> information during a thermal anneal<br />
from the seed layer into the crystallising a-Si overlayer [4.4.17]. The SPE process is a twostage<br />
process where the fi rst step is an a-Si deposition at around 200-300°C substrate<br />
temperature onto a hydrogen-terminated AIC poly-Si seed layer, followed by the second step<br />
which is a thermal anneal at about 570°C substrate temperature during which the a-Si<br />
crystallises via solid phase epitaxy. It seems possible that, compared to SPC solar cells (which<br />
do not feature a seed layer), the epitaxial growth feature in SPE solar cells improves the<br />
crystal quality and hence the effi ciency <strong>of</strong> the resulting solar cells.<br />
During 2005, we investigated the structural quality as well as the diode properties <strong>of</strong> poly-Si<br />
thin-fi lm solar cell fabricated by the SPC and SPE methods, utilizing PECVD a-Si:H precursor<br />
material. Two post-deposition treatments, rapid thermal annealing (RTA) and hydrogenation,<br />
were used to increase the voltages <strong>of</strong> the cells. The poly-Si fi lms were characterized with<br />
Raman, UV refl ectance, and Suns-Voc measurements. All samples were deposited onto 3<br />
mm thick Bor<strong>of</strong>l oat33 glass substrates (Schott AG). A conventional 13.56 MHz parallelplate<br />
PECVD machine was used for the SiN and the doped a-Si:H layers. Phosphine and<br />
diborane-doped silane gas was used as the doping gas for the deposition <strong>of</strong> n-type and p-<br />
type a-Si:H fi lms, respectively. SPC and SPE crystallisation anneals were performed ex-situ<br />
in a conventional nitrogen-purged atmospheric-pressure tube furnace. The SPC anneal was<br />
performed iso-thermally at about 600°C for 15 hours, whereas the SPE anneal had a 2-step<br />
anneal (15 hours at 570°C, followed by 6 hours at 600°C). RTA was done at 900°C for 4<br />
minutes. Hydrogenation was performed in a parallel-plate reactor at 13.56 MHz, using a<br />
hydrogen plasma for 1 hour at a glass temperature <strong>of</strong> about 470°C. Table 4.4.1 summarizes<br />
the different sample structures used in this study.<br />
Table 4.4.1: Structures <strong>of</strong> the investigated samples.<br />
Sample Final structure<br />
SPE-A Glass / ~ 70 nm SiN / 200 nm p++ AIC poly-Si / 1200 nm n- SPE poly-Si /100 nm n+ SPE poly-Si<br />
SPE-B Glass / ~ 70 nm SiN / 200 nm p++ AIC poly-Si / 1200 nm p- SPE poly-Si / 100 nm n++ SPE poly-Si<br />
SPC Glass / ~70 nm SiN / 200 nm n++ SPC poly-Si / 1200 nm p- SPC poly-Si / 100 nm p+ SPC poly-Si<br />
63
The main Raman peak at 520 cm -1 broadens in damaged and disordered crystalline silicon<br />
due to relaxation <strong>of</strong> the momentum conservation in the Raman scattering process. Hence, the<br />
full width at half maximum (FWHM) <strong>of</strong> the main Raman peak at around 520 cm -1 is a quantity<br />
that measures the amount <strong>of</strong> crystal disorder in our poly-Si fi lms. Raman measurements<br />
were performed with a red laser (λ = 633 nm) on samples SPE-A and SPC. As a reference,<br />
a polished Si wafer was also analysed. The Raman peak at ~ 520 cm -1 was fi tted with a<br />
mixed Lorentzian-Gauss function. The measured FWHM values are displayed in Fig. 4.4.16.<br />
This shows that our poly-Si material grown on an AIC seed layer has a signifi cantly improved<br />
crystal quality compared to SPC poly-Si fi lms grown on SiN-coated glass.<br />
5.5<br />
5.0<br />
FWHM [1/cm]<br />
4.5<br />
4.0<br />
3.5<br />
Figure. 4.4.16: Raman measurements on a Si<br />
wafer, on sample SPE-A and on a SPC sample.<br />
The error bars are calculated with 95%<br />
confi dence.<br />
3.0<br />
Si wafer SPE - PECVD SPC -PECVD<br />
UV refl ectance can be used for evaluating the crystal quality. The total hemispherical<br />
refl ectance was measured with our double-beam spectrophotometer (Varian Cary 5G) and<br />
an integrating sphere, in the wavelength range 220–400 nm. The reproducibility <strong>of</strong> the UV<br />
measurements was very good (± 0.1 absolute %). As can be seen in Fig. 4.4.17, SPE poly-Si<br />
has a signifi cantly better UV refl ectance response than SPC poly-Si. Hence Raman and UV<br />
refl ectance are in agreement with respect to the Si fi lm quality.<br />
75<br />
Si Wafer<br />
70<br />
SPE<br />
Reflection (%)<br />
65<br />
60<br />
55<br />
50<br />
SPC<br />
Figure 4.4.17: UV refl ectance <strong>of</strong> a polished<br />
singlecrystalline Cz Si wafer (top curve),<br />
sample SPE-A (middle curve), and sample<br />
SPC (lowest curve).<br />
45<br />
210 240 270 300 330 360 390<br />
Wavelength (nm)<br />
Suns-Voc measurements were performed on samples SPE-A, SPE-B and SPC at room<br />
temperature, using illumination <strong>of</strong> the samples from the glass side (“superstrate confi guration”)<br />
for sample SPE-A and SPC and air-side illumination for sample SPE-B. The measured Suns-Voc<br />
characteristics were fi tted with a 2-diode model (using fi xed diode ideality factors n <strong>of</strong> 1 and<br />
2) and a shunt resistance (Rsh). For each sample, the shunt resistance was high enough to<br />
not affect the measured Suns-Voc curve. Figure 4.4.18 shows the measured Suns-Voc curve<br />
for sample SPE-A, together with the 2-diode model analysis. We found that the 1-Sun Voc <strong>of</strong><br />
SPE-A and SPE-B are both dominated by the n=1 diode, whereas the Voc <strong>of</strong> sample SPC also<br />
has some infl uence from the n=2 diode. As displayed in Table 4.4.2, the SPC solar cell has a<br />
signifi cantly better 1 Sun-Voc (463 mV) than the SPE solar cells (423 and 393 mV). The 1-<br />
Sun Voc <strong>of</strong> the SPE diodes are completely dominated by the n=1 diode.<br />
64
The fact that the SPE diodes are n=1 limited strongly suggests that the voltage problem is<br />
not associated with grain boundaries or the p-n junction depletion region. Investigation <strong>of</strong><br />
the active doping levels in the base and emitter regions <strong>of</strong> the SPE cells is in progress, but it<br />
should be noted that the 393-mV SPE diode had no intentionally different doping level from<br />
that <strong>of</strong> the 463-mV SPC diode.<br />
From the results obtained in this study it is clear that SPE poly-Si material grown on AICseeded<br />
glass has a signifi cantly better crystal quality compared to SPC poly-Si material grown<br />
on bare (i.e., non-seeded) glass. So far, this advantage has not been translated into superior<br />
photovoltaic performance. Analysis <strong>of</strong> Suns-Voc curves indicates that the problem lies in the<br />
absorber region and/or the emitter region <strong>of</strong> the SPE diodes.<br />
10..0<br />
2-diode model<br />
Light Intensity (suns)<br />
1.0<br />
0.1<br />
n=1<br />
Exp. data<br />
n=2<br />
Figure 4.4.18: Suns-Voc data<br />
(symbols) and 2-diode model analysis<br />
(solid lines) <strong>of</strong> sample SPE-A.<br />
0.0<br />
0.36<br />
0.38 0.40 0.42 0.44 0.46 0.46<br />
Voltage (V)<br />
Table 4.4.2: Suns-Voc data <strong>of</strong> samples SPE-A, SPE-B and SPC.<br />
Sample p-n junction location 1-Sun Voc n=1 Voc n= 2 Voc<br />
SPE-A glass side 423 424 561<br />
SPE-B air side 393 401 499<br />
SPC glass side 463 474 523<br />
4.4.10 Optimisation <strong>of</strong> the Rapid Thermal Annealing and Hydrogenation<br />
Processes<br />
Due to the fabrication process and structure <strong>of</strong> poly-Si thin-fi lm solar cells, defects are<br />
present in the material. These defects can drastically reduce the open-circuit voltage (Voc)<br />
and the short-circuit current (Jsc) <strong>of</strong> the device. As such, the removal <strong>of</strong> these defects is<br />
essential and is typically done via thermal annealing and subsequent hydrogen passivation. An<br />
additional concern is achieving complete activation <strong>of</strong> dopants within the device.<br />
Thermal annealing is well known to dramatically reduce defects and also to activate dopants.<br />
Traditionally this annealing has been done in a tube furnace. Rapid thermal annealing (RTA)<br />
has replaced the tube furnace in many applications. RTA processes are ideally suited for<br />
poly-Si thin-fi lm solar cells on glass substrates [4.4.18 - 4.4.21] as precise control <strong>of</strong> the<br />
thermal pr<strong>of</strong>i le the device receives can be achieved. It is well known that hydrogen passivation<br />
(hydrogenation) is crucial in passivating defects. In thin-fi lm devices this hydrogenation is<br />
typically done using an RF or microwave plasma. The Voc is typically more than doubled and<br />
the Jsc is tripled by the hydrogenation step [4.4.18 - 4.4.21]. The combination <strong>of</strong> these two<br />
processes is critical in drastically improving poly-Si thin-fi lm solar cells.<br />
65
In 2005, an optimisation <strong>of</strong> the RTA and hydrogenation processes was performed for three<br />
related poly-Si thin-fi lm solar cells, using planar glass. These cell technologies are EVA, PLASMA<br />
(“solid-phase crystallised PLASMA-enhanced chemical vapor deposited a-Si”), and ALICIA. The<br />
fi rst two cell types are closely related, yet use different a-Si deposition methods. Both have a<br />
grain size <strong>of</strong> about 1.5 µm, as seen in Figure 4.4.19. The third cell type features a seed layer<br />
formed via aluminium-induced crystallization [4.4.12, 4.4.13]. The remaining cell structure is<br />
grown on this seed layer via ion-assisted deposition. The grain size <strong>of</strong> ALICIA cells is typically<br />
about 20 µm. Figure 4.4.20 shows a cross section <strong>of</strong> an ALICIA solar cell as measured with<br />
an electron microscope.<br />
Figure 4.4.19: Transmission electron<br />
microscope image <strong>of</strong> an EVA cell showing the<br />
crystallographic structure and grain size.<br />
Figure 4.4.20: Transmission electron<br />
microscope image <strong>of</strong> an ALICIA cell showing<br />
the crystallographic structure.<br />
Figures 4.4.21 and 4.4.22 show the change in Voc following RTA and hydrogenation as a<br />
function <strong>of</strong> RTA time for EVA and PLASMA at 900°C plateau temperature and ALICIA at<br />
1000°C plateau temperature, respectively. As expected, the Voc increases with longer times,<br />
reaches a maximum and then falls <strong>of</strong>f. This fall-<strong>of</strong>f in Voc is believed to be due to drastic<br />
smearing <strong>of</strong> the junction space charge region due to excessive dopant diffusion along grain<br />
boundaries.<br />
Figure 4.4.21: Voc as a function <strong>of</strong> time<br />
at 900°C plateau temperature for EVA<br />
and PLASMA solar cells following RTA and<br />
hydrogenation.<br />
Figure 4.4.22: Voc as a function <strong>of</strong> time at<br />
1000°C plateau temperature for ALICIA solar<br />
cells following RTA and hydrogenation.<br />
66
The current response <strong>of</strong> the cell as a function <strong>of</strong> the wavelength <strong>of</strong> the incident light (internal<br />
quantum effi ciency, IQE) provides valuable information on the performance <strong>of</strong> the solar cell.<br />
Figures 4.4.23 and 4.4.24 illustrate the effect <strong>of</strong> the variation in RTA times at 1000°C and<br />
hydrogenation on the IQE, respectively. With a 15-sec 1000°C RTA the peak IQE and Jsc<br />
increase from 16% to 40% and 1.92 mA/cm 2 to 6.11 mA/cm 2 , respectively. Following<br />
hydrogenation the peak IQE and Jsc increase further to 70% and 10.95 mA/cm 2 , a total<br />
gain <strong>of</strong> factors 4.4 and 5.7, respectively. Hydrogenation alone without an RTA results in a gain<br />
in peak IQE and Jsc to 41% and 6.43 mA/cm 2 , respectively.<br />
The maximum Voc achieved in this RTA and hydrogenation study were 491 mV, 436 mV<br />
and 435 mV for EVA, PLASMA and ALICIA, respectively. Improved understanding <strong>of</strong> defect<br />
removal and passivation paves the way towards all three cells having a Voc well over 500 mV<br />
and a Jsc greater than 15 mA/cm 2 on planar glass.<br />
Figure 4.4.23: Measured IQE <strong>of</strong> ALICIA<br />
cells before and after the 1000°C RTA<br />
process step. RTA plateau times are 10<br />
and 15 sec.<br />
4.4.11 Cell Metallisation<br />
Figure 4.4.24: IQE <strong>of</strong> hydrogenated<br />
ALICIA cells as a function <strong>of</strong> the 1000°C<br />
RTA plateau time. RTA plateau times are<br />
1, 10, 15 and 30 sec. Also shown, for<br />
comparison, are the IQE <strong>of</strong> the as-grown<br />
cell and a hydrogenated cell that did not<br />
receive a RTA step.<br />
During 2005, the Thin-Film Group’s metallisation efforts took major strides forward with the<br />
formation <strong>of</strong> the Device Fabrication Team (DFT) within the Thin-Film Group. The DFT’s purpose<br />
is the development and implementation <strong>of</strong> metallisation and interconnection schemes for the<br />
Group’s thin-fi lm poly-Si solar cells.<br />
In 2005, our patented self-aligning maskless photolithography (SAMPL) technique [4.4.22]<br />
has been applied to fabricate large-area ( ~ 4 cm 2 ) interdigitated solar cell structures (see Fig.<br />
4.4.25). The process used to make the large-area solar cells is as follows: First, a comb-like<br />
air-side electrode is deposited on the surface <strong>of</strong> the solar cell by evaporation or sputtering<br />
<strong>of</strong> metal (aluminium) through a shadow mask. Next, the regions <strong>of</strong> silicon fi lm which are not<br />
covered by the air-side electrode are etched away by a plasma process. Finally, the comb-like<br />
glass-side electrode is formed in the etched regions using the SAMPL process.<br />
Large-area (4 cm 2 ) thin-fi lm solar cells with reasonable effi ciencies have been fabricated with<br />
this method. In addition, research is underway to improve the effi ciency <strong>of</strong> these devices by<br />
optimising the electrode geometry. This is being realised by narrowing the glass-side fi ngers<br />
by defi ning the air-side electrode using a photolithographic technique.<br />
67
Figure 4.4.25: Schematic representation <strong>of</strong> a metallised interdigitated solar cell<br />
(not to scale).<br />
Furthermore, we have developed a method for interconnecting these large-area solar<br />
cells to form mini-modules. The method is schematically illustrated in Figure 4.4.26. The<br />
interconnection <strong>of</strong> the large-area solar cells is achieved by using a simple stencil-printing<br />
technique: First, adjacent cells are separated by scribing through the glass-side busbar with<br />
a laser. Next, a layer <strong>of</strong> isolating polymer is printed across the edge <strong>of</strong> the busbar. Finally, a<br />
layer <strong>of</strong> conductive silver paste is printed over the insulating polymer to connect the glass-side<br />
busbar <strong>of</strong> each cell with the air-side busbar <strong>of</strong> the adjacent cell. Three-cell series-connected<br />
poly-Si thin-fi lm mini-modules with reasonable effi ciency were fabricated in this way in 2005.<br />
glass - side<br />
contact<br />
air - side<br />
conta ct<br />
series<br />
connection<br />
printed busbar<br />
~ 4 cm<br />
cells isolated<br />
by laser<br />
cutting<br />
p +<br />
p -<br />
n +<br />
printed busbar<br />
isolation<br />
layer<br />
p<br />
+<br />
p -<br />
n +<br />
glass<br />
glass - side glass - side<br />
contact contact<br />
Cell n<br />
Cell n+1<br />
4 mm<br />
air -side<br />
contact<br />
Si<br />
Figure 4.4.26: Schematic<br />
diagram <strong>of</strong> series connection<br />
<strong>of</strong> thin-fi lm cells (not to scale).<br />
Three interdigitated solar cells<br />
are connected in series (left,<br />
top view). The cross section<br />
on the right shows a detailed<br />
view <strong>of</strong> the interconnection<br />
region.<br />
68
4.4.12 References for Section 4.4<br />
4.4.1 H. Keppner, J. Meier, P. Torres, D. Fischer, and A. Shah, Appl. Phys. A 69, 169-177<br />
(1999).<br />
4.4.2 T. Kieliba, S. Bau, D. Osswald, and A. Eyer, Proc. 17th European Photovoltaic Solar<br />
Energy Conference, Munich, 2001, p. 1604.<br />
4.4.3 K. Yamamoto, A. Nakajima, M. Yoshimi, T. Sawada, S. Fukuda, K. Hayashi, T. Suezaki,<br />
M. Ichikawa, Y. Koi, M. Goto, H. Takata and Y. Tawada, Proc. 29th IEEE Photovoltaic<br />
Specialists Conference, <strong>New</strong> Orleans, 2002, p. 1110.<br />
4.4.4 P.A. Basore, Proc. 3rd World Conference on Photovoltaic Solar Energy Conversion,<br />
Osaka, 2003, p. 935.<br />
4.4.5 A.G. Aberle, P.I. Widenborg, A. Straub, and N. P. Harder, Proc. 3rd World<br />
Conference on Photovoltaic Energy Conversion, Osaka, 2003, p. 1194.<br />
4.4.6 C. Berge, R.B. Bergmann, T.J. Rinke, and J.H. Werner, Proc. 17th European<br />
Photovoltaic Solar Energy Conference, Munich, 2001, p. 1277.<br />
4.4.7 K. Feldrapp, R. Horbelt, R. Auer and R. Brendel, Progress in Photovoltaics 11, 105-<br />
112 (2003).<br />
4.4.8 A. Blakers, K. Weber, V. Everett, S. Deenapanray, J. Babaei, and M. Stocks, Proc.<br />
19th European Photovoltaic Solar Energy Conference, Paris, 2004, p. 431.<br />
4.4.9 T. Matsuyama, N. Terada, T. Baba, T. Sawada, S. Tsuge, K. Wakisaka and S. Tsuda, J.<br />
Non-Cryst. Solids 198-200, 940 (1996).<br />
4.4.10 P.A. Basore, Proc. 19th European Photovoltaic Solar Energy Conference, Paris,<br />
2004, p. 455.<br />
4.4.11 http://www.us.schott.com/hometech/english/products/bor<strong>of</strong>l oat/<br />
4.4.12 O. Nast, T. Puzzer, L.M. Koshier, A.B. Sproul, and S.R. Wenham, Applied Physics<br />
Letters 73, 3214-3216 (1998).<br />
4.4.13 P.I. Widenborg and A.G. Aberle, Journal <strong>of</strong> Crystal Growth 242, 270-282 (2002).<br />
4.4.14 R.A. Sinton, A. Cuevas, Proc. 16th European PV Solar Energy Conference, Glasgow,<br />
2000, p. 1152.<br />
4.4.15 A.G. Aberle, A. Straub, P.I. Widenborg, A.B. Sproul, Y. Huang, and P. Campbell,<br />
Progress in Photovoltaics 13, 37-47 (2005).<br />
4.4.16 A. Straub, N.-P. Harder, Y. Huang and A.G. Aberle, Journal <strong>of</strong> Crystal Growth 268,<br />
41-51 (2004).<br />
4.4.17 P.I. Widenborg, A. Straub and A.G. Aberle, Journal <strong>of</strong> Crystal Growth 276, 19-28<br />
(2005).<br />
4.4.18 M.L. Terry, A. Straub, D. Inns, D. Song, and A.G. Aberle, Applied Physics Letters 86,<br />
172108 (2005).<br />
4.4.19 M.L. Terry, P.I. Widenborg, O. Kunz, and A.G. Aberle, Tech. Digest 15th International<br />
Photovoltaic Science and Engineering Conference, Shanghai, 2005, p. 905.<br />
4.4.20 M.J. Keevers, A. Turner, U. Schubert, P.A. Basore and M.A. Green, Proc. 20th<br />
European Photovoltaic Solar Energy Conference, Barcelona, 2005, p. 1305.<br />
4.4.21 A. Straub, D. Inns, M.L. Terry, Y. Huang, P.I. Widenborg, and A.G. Aberle, Journal <strong>of</strong><br />
Crystal Growth 280, 385-400 (2005).<br />
4.4.22 T.M. Walsh, D. Song, S. Motahar, and A.G. Aberle, Tech. Digest 15th International<br />
Photovoltaic Science and Engineering Conference, Shanghai, 2005, p. 706.<br />
69
4.5 THIRD GENERATION STRAND: ADVANCED CONCEPTS<br />
<strong>University</strong> Staff:<br />
Dr. Gavin Conibeer (group leader)<br />
Dr. Richard Corkish<br />
Pr<strong>of</strong>. Martin Green<br />
Research and Postdoctoral Fellows:<br />
Dr. Ximing Dai<br />
Dr. Tammy Humphrey (part time to Feb 2005)<br />
Dr. Dirk König (from June, 2005)<br />
Dr. Kuo-Lung Lin (to Jan 2005)<br />
Dr. Tom Puzzer (part time)<br />
Dr. Bryce Richards (to March 2005)<br />
Research Associates:<br />
Yidan Huang<br />
Edwin Pink<br />
Dr. Didier Debuf (adjunct fellow)<br />
Higher Degree Students:<br />
Young Hyun Cho<br />
Thipwan Fangsuwannarak<br />
Chu-Wei “Scott” Jiang<br />
Supriya Pillai<br />
Giuseppe Scardera<br />
Avi Shalav<br />
Visiting Researchers:<br />
Dr. Jean-Françoise Guillemoles<br />
(IRDEP: joint EDF/CNRS/ENSCP lab, Paris, April to Oct 2005)<br />
Pr<strong>of</strong>. Peter Würfel (Karlsruhe <strong>University</strong>, Oct to Dec 2005)<br />
Visiting Students:<br />
James Rudd (U/grad thesis student 2005)<br />
Theo Comino (U/grad thesis student 2005)<br />
Owen The (U/grad thesis student 2005)<br />
In 2005, work in the Third Generation Strand has continued to build on the work <strong>of</strong> 2004,<br />
2003 and <strong>of</strong> the earlier Third Generation Special Research Centre. The concept <strong>of</strong> third<br />
generation photovoltaics is to signifi cantly increase device effi ciencies whilst still using thin<br />
fi lm processes and abundant non-toxic materials. This can be achieved, for example by using<br />
multiple energy threshold approaches.<br />
In 2005, each <strong>of</strong> the three principal project areas has shown signifi cant development: with<br />
transfer <strong>of</strong> technology to nitride dielectrics in Si nanostructures; further quantum effi ciency<br />
improvements in Up-conversion and application <strong>of</strong> carrier cooling theory in Hot Carrier Cells<br />
to include sensitivity to some <strong>of</strong> the non-idealities found in real materials. Also there have been<br />
advances in subsidiary project areas, with demonstrated increased luminescence effi ciency<br />
using Surface Plasmon coupling; development <strong>of</strong> more robust windows based limiting effi ciency<br />
programs; and a small resurgence <strong>of</strong> theoretical work on Quantum Antennas.<br />
71
The parallel project with Toyota on high effi ciency thermoelectrics has made good progress,<br />
with important developments in modelling and characterisation, which also link to the work on<br />
Hot Carrier cells. In addition, a very signifi cant development is the successful proposal to the<br />
Global Climate and Energy Program (GCEP) at Stanford <strong>University</strong>, for a project to develop the<br />
Si nanostructure work to a prototype Si based tandem cell over the next three years.<br />
4.5.1 Third Generation Photovoltaics<br />
“Third generation” approaches aim to achieve high effi ciency for photovoltaic devices but<br />
still use “thin fi lm” second generation deposition methods. The concept is to do this with only<br />
a small increase in areal costs and hence reduce the cost per Watt peak [4.5.1]. Also, in<br />
common with the silicon based second generation thin fi lm technologies, abundant and nontoxic<br />
materials will be used. Thus these “third generation” technologies will be compatible with<br />
large scale implementation <strong>of</strong> photovoltaics. The approach differs from “fi rst generation”<br />
fabrication <strong>of</strong> high quality and hence low defect single crystal photovoltaic devices, which have<br />
high effi ciencies, approaching the limiting effi ciencies for single band gap devices, but which<br />
use energy and time intensive techniques. Third Generation aims to decrease costs to below<br />
US$0.50/W, potentially to US$0.20/W or better, by dramatically increasing effi ciencies<br />
but maintaining the economic and environmental cost advantages <strong>of</strong> thin fi lm deposition<br />
techniques (see Figure 4.1.3). To achieve such effi ciency improvements, the targetted devices<br />
aim to circumvent the Shockley-Queisser limit for single band gap cells that limits effi ciencies<br />
to the “Present limit” indicated in Fig. 4.1.3 <strong>of</strong> between 31% to 41% (for one sun and maximum<br />
concentration respectively). One approach multiple energy threshold devices such as the<br />
tandem or multi-colour solar cell. We are investigating a number <strong>of</strong> approaches to achieve<br />
such multiple energy threshold devices. [4.5.1, 4.5.2]<br />
The two most important power loss mechanisms in single-bandgap cells arise from the inability<br />
to absorb photons with energy less than the bandgap (1 in Figure 4.5.1) and thermalisation<br />
<strong>of</strong> photon energy exceeding the bandgap, (2 in Figure 4.5.1). These two mechanisms alone<br />
amount to the loss <strong>of</strong> about half <strong>of</strong> the incident solar energy in solar cell conversion to electricity.<br />
Multiple threshold approaches can utilise some <strong>of</strong> this lost energy. Such approaches avoid<br />
the Schockley-Queisser limit, by the exploitation <strong>of</strong> more than one energy level - in some form<br />
– for which the limit does not apply. The limit which does apply is the thermodynamic limit<br />
shown in Figure 4.1.3, <strong>of</strong> 67% to 86.8% (again depending on concentration).<br />
There have been proposed three families <strong>of</strong> approaches for avoiding the Schockley-<br />
Queisser limit [4.5.2]: (a) increasing the number <strong>of</strong> bandgaps; (b) capturing carriers before<br />
thermalisation; and (c) multiple carrier pair generation per high energy photon or single carrier<br />
pair generation with multiple low energy photons. Of these, tandem cells, an implementation<br />
<strong>of</strong> strategy (a), are the only ones which have as yet been realised with effi ciencies exceeding<br />
the Shockley-Queisser limit.<br />
Figure 4.5.1: Loss processes in<br />
a standard solar cell: (1) nonabsorption<br />
<strong>of</strong> below band gap<br />
photons; (2) lattice thermalisation<br />
loss; (3) and (4) junction and contact<br />
voltage losses; (5) recombination<br />
loss.<br />
72
In the Third Generation Strand, we are implementing strategy (a) by fabricating a tandem<br />
cell based on silicon and its oxides, nitrides and carbides using reduced dimension silicon<br />
nanostructures to engineer the band gap <strong>of</strong> an upper cell material. We are also tackling<br />
strategy (b) by investigating the “Hot Carrier solar cell” in which carrier cooling is slowed such<br />
that carriers can be extracted before thermalisation. This requires both an absorber with<br />
slowed carrier cooling properties and collection <strong>of</strong> carriers over a limited range <strong>of</strong> energies,<br />
such that cold carriers in the external contacts do not cool the hot carriers . Finally we are<br />
investigating implementation <strong>of</strong> strategy (c) by up-conversion in a layer behind the Si cell. Rare<br />
earth doped phosphors in the up-converter absorb below band gap photons and up-convert<br />
two or more to above band gap photons which are then incident on the Si cell.<br />
Hence all Third Generation approaches are based on tackling one or both <strong>of</strong> the “below band<br />
gap” or “thermalisation” loss mechanisms mentioned above, as detailed below.<br />
4.5.2 Si-Based Nanostructure tandem cells<br />
Researchers:<br />
Young Cho, Gavin Conibeer, Martin Green, Thipwan Fangsuwannarak, Chu-Wei Jiang, Dirk<br />
König, Tom Puzzer, Giuseppe Scardera<br />
Tandem cells are stacks <strong>of</strong> individual cells with different energy thresholds each absorbing<br />
a different band <strong>of</strong> the solar spectrum, usually connected together in series. We aim to<br />
engineer a new silicon-based material to form a top cell above a silicon cell, see Fig. 4.1.4.<br />
This material is ‘engineered’ using a quantum dot nanostructure <strong>of</strong> silicon in a silicon based<br />
dielectric matrix. The confi ned energy levels in the quantum dots will increase the lowest<br />
absorption edge <strong>of</strong> the material compared to bulk silicon. If the quantum dot density is high<br />
enough, the wavefunctions <strong>of</strong> the quantum dots will overlap to create true superlattice<br />
minibands and increase the effective band gap <strong>of</strong> the material.<br />
The main challenge for a nanostructure engineered material for a tandem cell is to achieve<br />
suffi cient carrier mobility and hence a reasonable conductivity. For a nanostructure, this<br />
generally requires formation <strong>of</strong> a superlattice with overlap <strong>of</strong> the wavefunction for adjacent<br />
quantum wells or quantum dots; which in turn requires either close spacing between QWs or<br />
QDs or low barrier height. Another requirement for a tandem cell element is the presence<br />
<strong>of</strong> some form <strong>of</strong> junction for carrier separation. This can either be a grown or diffused p-n<br />
junction or a p-i-n junction formed in the superlattice as the i-region.<br />
As reported elsewhere [4.5.3, 4.5.4], such Si quantum dots have been fabricated by cosputtering<br />
deposition <strong>of</strong> silicon rich layers interspersed with stoichiometric SiO 2 layers (an<br />
application <strong>of</strong> the method <strong>of</strong> Zacharias et al. [4.5.5] to photovoltaics). On annealing at 1100°C<br />
silicon precipitates from the super-saturated solid solution to form nanocrystals. For layers<br />
<strong>of</strong> thickness less than about 4nm, the precipitation enters a regime <strong>of</strong> 2D diffusion in which<br />
the dot size is accurately controlled by the layer thickness.<br />
The increased strength <strong>of</strong> optical processes in confi ned structures due to localisation <strong>of</strong><br />
electrons and holes is highly advantageous for solar cell structures, as it means that only<br />
very thin layers <strong>of</strong> these 100nm dots are required for strong absorption. With an appropriate<br />
light-trapping scheme, thin 100nm fi lms <strong>of</strong> unconfi ned bulk Si to give high solar absorption.<br />
Quantum dot devices <strong>of</strong> only 100nm thickness seem feasible given the increased optical<br />
strength <strong>of</strong> the confi ned processes.<br />
73
4.5.2.1 Alternative matrices for Si quantum dots:<br />
Transport properties are expected to depend on the matrix in which the silicon quantum dots<br />
are embedded. As shown in Figure 4.5.2, different matrices produce different transport<br />
barriers between the Si dot and the matrix, with tunnelling probability heavily dependent on<br />
the height <strong>of</strong> this barrier. Si 3 N 4 and SiC give lower barriers than SiO 2 allowing larger dot<br />
spacing for a given tunnelling current.<br />
Figure 4.5.2: Bulk band alignments<br />
between crystalline silicon and its<br />
carbide, nitride and oxide.<br />
The barrier to tunnelling between<br />
quantum dots is reduced for a lower<br />
barrier height material. Considering<br />
the tunnelling probability for electrons<br />
in a square potential well (e.g. [4.5.6]<br />
p244):<br />
1.9 eV<br />
c-Si<br />
0.5 eV<br />
1.1 eV c-Si 1.1 eV c-Si<br />
0.9 eV<br />
2.3 eV<br />
3.2 eV<br />
1.1 eV<br />
4.7 eV<br />
where m* is the bulk effective mass in the conduction band <strong>of</strong> the matrix, d is the spacing<br />
between dots and ∆E is the energy difference between the conduction band and the fi rst<br />
confi ned minband energy level in the quantum dot superlattice. Hence the important<br />
parameter in determining the degree <strong>of</strong> interaction between quantum dots is m*∆Ed 2 . As<br />
barrier height decreases, the barrier thickness for a given tunnelling probability increases,<br />
thus requiring lower dot density for a given conductivity or higher conductivity for a given<br />
dot density. (In addition as dot size decreases ∆E also decreases, thus increasing T e and<br />
enhancing the effect further for smaller quantum dots) [4.5.7]. A simialr argument applies<br />
for tunnelling probability <strong>of</strong> holes in the valence band.<br />
Hence, transport between dots can be signifi cantly increased as the barrier height decreases<br />
with alternative matrices. The spacing <strong>of</strong> dots would have to be closest in the oxide, nitride and<br />
carbide, in that order. Similar deposition and quantum dot precipitation approaches should<br />
work for all three material systems.<br />
4.5.2.2 Si quantum dots in a nitride matrix<br />
For the above reason, we have explored application <strong>of</strong> similar techniques to those described<br />
in Section 4.5.2 for SiO 2 , to the growth <strong>of</strong> Si nanocrystals in silicon nitride by both sputtering<br />
and PECVD [4.5.8]. For sputtering, growth parameters are very similar to the oxide but<br />
with co-sputtering from a Si and Si 3 N 4 target rather than from an oxide one. Results from<br />
HRTEM show crystalline nanocrystals in the nitride matrix.<br />
74<br />
In addition two separate PECVD machines have been used for the growth <strong>of</strong> Si nanocrystals<br />
in nitride, with a regime similar to that used in oxide, <strong>of</strong> growing alternate Si rich and<br />
stoichiometric nitride layers. Annealing is carried out again at 1100°C but with a pre-anneal<br />
at 500°C to drive <strong>of</strong>f hydrogen incorporated from the PECVD process. Again HRTEM<br />
images showing even clearer nanocrystals have been obtained, Figure 4.5.5 [4.5.8]. The<br />
two machines are a non-automated single plasma cluster tool and an automated dual mode<br />
RF/microwave plasma Roth and Rau machine, thus demonstrating the transferability <strong>of</strong> the<br />
approach both to other material systems and to other thin fi lm processes.
Further optical and electrical characterisation is underway on these nitride materials.<br />
Other authors [4.5.9] have published data on Si QD in Si 3 N 4 demonstrating crystallinity and<br />
increased PL energy but not using the close control <strong>of</strong> QD size that we have employed using<br />
the layering method. To the best <strong>of</strong> our knowledge this is the fi rst time layered Si QDs in nitride<br />
samples have been prepared using either type <strong>of</strong> PECVD growth. The automated system<br />
also allows the possibility <strong>of</strong> growing large numbers <strong>of</strong> multilayers, which is important for<br />
several characterisation techniques. The technique is also being investigated for large area<br />
uniformity.<br />
4.5.3 Characterisation<br />
Researchers:<br />
Young Cho, Gavin Conibeer, Martin Green, Thipwan Fangsuwannarak, Yidan Huang, Chu-Wei<br />
Jiang, Dirk König, Edwin Pink, Tom Puzzer, Giuseppe Scardera, Thorsten Trupke<br />
In 2005 several new techniques have been implemented to improve characterisation <strong>of</strong><br />
Si nanostructures. Complementary information from a range <strong>of</strong> techniques is necessary<br />
because the very small sizes <strong>of</strong> the nanostructures involved push the resolution <strong>of</strong> the<br />
techniques to the limit. Also complementary information on physical structure, optical and<br />
electronic properties is required in order to optimise processing.<br />
4.5.3.1 Physical characterisation<br />
Recent work has continued characterisation using high resolution transmission electron<br />
microscopy (HRTEM) and compared it with the complementary technique <strong>of</strong> secondary ion<br />
mass spectrometry (SIMS). The latter is carried out at ANSTO, Lucas Heights, NSW, as part<br />
<strong>of</strong> a grant supplied by AINSE.<br />
Figure 4.5.3a is a medium resolution TEM at 120kV showing the multilayer Si nanostructure<br />
in SiO 2 . Figure 4.5.3b is a HRTEM at 300kV showing the crystal planes in Si nanocrystals.<br />
Figure 4.5.4 is a SIMS plot <strong>of</strong> the Si nanostructure with detail <strong>of</strong> the Si and O traces on<br />
the right. It clearly shows the alternation <strong>of</strong> the Si and O concentrations in the multilayer<br />
structure.<br />
Figure 4.5.3: a) TEM image <strong>of</strong> multilayer Si<br />
nanostructure in a-SiO2 matrix; b) HRTEM<br />
image <strong>of</strong> Si quantum dots - highlighted by<br />
ellipses.<br />
Figure 4.5.4: SIMS pr<strong>of</strong>i le <strong>of</strong><br />
SiQDs/SiO2 multilayer structure<br />
on quartz.<br />
75
4.5.3.1.1 TEM <strong>of</strong> SiQDs in nitride<br />
HRTEM images showing even clearer nanocrystals have been obtained, as shown in Figure<br />
4.5.5.<br />
(a)<br />
(b)<br />
Figure 4.5.5: TEM image <strong>of</strong> PECVD deposited (a) Si quantum dots in 10 bi-layers <strong>of</strong> SRN/SiNx<br />
superlattices and (b) HRTEM lattice image.<br />
4.5.3.1.2 TEM Sample Preparation<br />
Transmission Electron Microscopy (TEM) remains a crucial characterisation technique for<br />
nanostructures. TEM sample preparation, however, has traditionally been diffi cult and time<br />
consuming. Great efforts have been made at the Centre to improve the turn around time for<br />
TEM sample preparation. A very promising technique being explored involves a dual beam<br />
Focused Ion Beam (FIB) system. Figure 4.5.6 shows a cross sectional TEM specimen prepared<br />
using a FEI Nova Nanolab 200 Dualbeam FIB, the fl agship instrument at UNSW’s Electron<br />
Microscopy Unit. This same specimen was used to produce the image shown in Figure 4.5.7.<br />
The sample preparation time for this sample was approximately one hour, signifi cantly faster<br />
than traditional techniques.<br />
Figure 4.5.6: TEM Si QD specimen prepared using a dual beam FIB.<br />
Figure 4.5.7: TEM image <strong>of</strong> Si QDs in a nitride matrix grown by dual-mode PECVD.<br />
76
4.5.3.1.3 Grazing Incidence X-Ray Diffraction (GIXRD)<br />
During 2005 X-ray diffraction (XRD) analysis has been developed for the Si nanostructure<br />
samples. This gives information on crystallinity, nanocrystal size and crystal fraction and<br />
potentially on nanocrystal size distribution, shape and any preferred orientation. The<br />
technique is complementary to TEM analysis and has the advantage that very little sample<br />
preparation is required.<br />
Two approaches have been used: analysis using X-ray tube diffractometers in the School <strong>of</strong><br />
Materials Science at UNSW and analysis using synchrotron XRD at the Australian Nuclear<br />
Beamline Facility, Tsukuba, Japan.<br />
In both cases the experiments aimed to measure XRD from the thin-fi lms at grazing<br />
Incidence (GIXRD). Figure 4.5.8 shows a result using X-ray tube for a sample with 62 bilayers<br />
<strong>of</strong> alternating 4.5nm SRO/SiO 2 layers (sample DK24) [SRO refers to silicon rich oxide]. It<br />
confi rms that (at least for this sample):<br />
• The lattice constant given by the {111} peak does not vary from the bulk silicon value<br />
by more than 0.2%.<br />
• The average Si nanocrystal size is 4.5nm - thus for this sample, the SRO layer<br />
thickness controlled the size <strong>of</strong> the nanocrystals.<br />
• There was no measurable microstrain present in the nanocrystals (i.e. no small<br />
variations <strong>of</strong> lattice constant from dot to dot).<br />
• The crystal orientations <strong>of</strong> the dots were perfectly random.<br />
Figure 4.5.8: X-ray tube<br />
XRD <strong>of</strong> a 62 bi-layer,<br />
4.2nm, SRO/SiO2<br />
sample.<br />
An important aspect <strong>of</strong> the analysis is comparison <strong>of</strong> X-ray tube with synchrotron XRD data<br />
and comparison <strong>of</strong> both with information from TEM. Figure 4.5.9 shows two-theta plots <strong>of</strong><br />
the same Si QD in nitride sample for which the TEM is shown in Figure 4.5.7. The sample has<br />
20 bilayers <strong>of</strong> 10nm Si rich nitride layers. It shows that:<br />
• The average Si nanocrystal size is 4.3nm – thus for this sample the SRO layer<br />
thickness was too thick to control the nanocrystal size.<br />
• Lattice parameter has less than 0.1% deviation from bulk value.<br />
• Deviation <strong>of</strong> intensities from random orientation are less than 10%.<br />
The presence <strong>of</strong> crystalline Si is clear from these plots. In terms <strong>of</strong> quantitative analysis, the<br />
viability <strong>of</strong> both techniques for extracting average size, strain and lattice parameters <strong>of</strong> the<br />
nanocrystals is being examined.<br />
77
Figure 4.5.9: Two Theta Plots<br />
<strong>of</strong> the Si QDs in nitride sample<br />
shown in Figure 4.5.7, using (a)<br />
synchrotron radiation and (b) a<br />
conventional x-ray tube source.<br />
The peak positions and shapes<br />
are very similar between a) and<br />
b) in Figure 4.5.9 (except for the<br />
peak at about 50°, which is due<br />
to substrate), indicating that information that can be extracted from X-ray tube measurements<br />
is similar to synchrotron, at least at present. This is particularly true for the fi rst peak at<br />
{111}. Information on average nanocrystal size is contained in the integral breadth or FWHM<br />
<strong>of</strong> the peaks. The less rapidly decreasing intensity with 2θ for the synchrotron is because<br />
synchrotron X-rays are polarised. The usefulness <strong>of</strong> this for detailed analysis <strong>of</strong> high index<br />
planes is being investigated.<br />
There are still issues that need to be resolved in order to extract better information from<br />
XRD. Approximate corrections to data were made to remove the effects <strong>of</strong> instrumental<br />
FWHM broadening and other effects. However future work will involve de-convolution (the<br />
mathematically correct procedure) <strong>of</strong> the diffractogram to remove machine-induced effects<br />
and the subsequent Fourier transformation to obtain the full Radial Scattering Distribution<br />
Function (RDF) for both the dots and the amorphous matrix. This represents the most<br />
complete information that can be obtained from XRD for small crystals and disordered<br />
materials such as amorphous silicon dioxide.<br />
4.5.3.2 Optical Characterisation – Photoluminescence (PL)<br />
Photoluminescence (PL) is a key technique in assessing the energy levels within quantum dot<br />
structures. The highest PL intensities will be from the lowest radiative recombination states<br />
within the illuminated area close to the surface. Thus it is a tool which can assess whether<br />
quantum confi ned energy levels, greater than the band gap energy, are present. The fact that<br />
it can be applied to samples without contacts and with minimal preparation lends it to a quick<br />
turnaround feedback technique for processing. However, additional information with more<br />
careful analysis and at reduced temperatures can also be obtained. This can yield specifi c<br />
detailed information on the stability <strong>of</strong> energy levels with temperature and their relative<br />
luminescence intensities.<br />
4.5.3.2.1 PL on Si QDs In Si 3 N 4<br />
Figure 4.5.10 is a comparison <strong>of</strong> PL from Si QDs in a SiO 2 matrix for various layer thicknesses<br />
and hence QD size with Si QDs in a Si 3 N 4 matrix for a fi xed layer thickness <strong>of</strong> 4nm but<br />
with varying numbers <strong>of</strong> layers. PL measurements were carried out at room temperature<br />
(300 K) using excitation by a laser operating at 405nm and 50mW with detection by a<br />
thermoelectrically cooled Si CCD detector.<br />
78
These results are discussed further elsewhere [4.5.8] and provide tentative evidence for<br />
quantum confi ned energy levels in Si QDs in nitride at a higher energy than in equivalent sized<br />
QDs in oxide. However these are as yet early data. PL on different sized QDs in nitride has<br />
not yet been obtained. These would show whether the 1.8eV peak increases or not with<br />
decreasing dot size, an increase indicating a true confi ned energy level, no increase indicating<br />
a fi xed defect in the interface or matrix. In addition, the intensity <strong>of</strong> the nitride peaks is<br />
not high, probably indicating a low radiative transition probability. Nonetheless, the fact that<br />
intensity increases with the number <strong>of</strong> SiQD layers is encouraging. Further data are being<br />
generated.<br />
Figure 4.5.10: Comparison <strong>of</strong> the PL spectra <strong>of</strong><br />
Si QDs in nitride and oxide (300K). Peaks on<br />
the left-hand side are in the nitride matrixes and<br />
higher peaks on the right-hand side are from the<br />
oxide matrixes.<br />
4.5.3.2.2 PL on Si QDs in SiO 2 - Analysis<br />
PL spectra have been used to identify radiative<br />
recombination energy levels in Si QD in SiO 2<br />
nanostructures. In the model used, radiative<br />
contributions are considered to come from one<br />
<strong>of</strong> three sources: from quantum confi nement<br />
(Q); from SiQD/SiO 2 interfaces (I) and from<br />
amorphous Si clusters (α). The relative strengths <strong>of</strong> these assumed processes have been<br />
determined by deconvoluting the spectral peaks. The 3 Gaussian deconvolutions <strong>of</strong> the PL<br />
spectra for Q, I and α are shown in Figure 4.5.11. Initial comparison with similar measurements<br />
on other samples suggests that the position <strong>of</strong> the Q peak depends on the size <strong>of</strong> the QDs<br />
whereas the position <strong>of</strong> the I peak seems to be independent <strong>of</strong> QD size. Possible mechanisms<br />
for these transitions are being assessed, including a possible recombination due to an oxygen<br />
defi ciency that is at 6.3eV above the valence band <strong>of</strong> SiO 2 or at 1.65-1.70eV above the<br />
conduction band <strong>of</strong> the SiQD in the ground state, see Figure 4.5.12.<br />
Figure 4.5.11: Gaussian deconvolution <strong>of</strong><br />
PL spectrum for a SiQD/SiO2 sample:<br />
peaks labelled Q, I and α are attributed to<br />
quantum confi nement, interface states and<br />
a-Si clusters, respectively.<br />
Figure 4.5.12: Schematic model <strong>of</strong> the Q<br />
and I radiative mechanisms.Also shown are<br />
possible hopping conduction mechanisms.<br />
Additional data on the change <strong>of</strong> the ratio <strong>of</strong> Q:I peak intensity with the degree <strong>of</strong> excess<br />
Si in SRO samples and with anneal time are being assessed and promise more detailed<br />
information on these mechanisms.<br />
79
4.5.3.3 Electrical characterisation<br />
In 2005 (as in 2004), electrical characterisation concentrated on lateral conductivity<br />
measurements. Although vertical measurements are the more relevant for structures <strong>of</strong><br />
interest, lateral measurements are easier to model because they involve transport only<br />
within layers containing Si nanocrystals whereas vertical measurements include both intralayer<br />
and inter-layer transport across the intervening oxide layers.<br />
Contacting for lateral conductivity measurement has been achieved by using a technique <strong>of</strong><br />
using Al contacts to ‘punch’ through the surface oxide, by annealing evaporated Al at 500°C<br />
for 15 hr. This induces non-uniform spiking through the upper layers <strong>of</strong> oxide and provides<br />
a very large surface area contact to the QD layers, as shown in Figure 4.5.13. This in turn<br />
results in a voltage drop across the contact about a factor <strong>of</strong> 10 lower than the voltage drop<br />
across the active region <strong>of</strong> the SiQD SLs, thus resulting in a small relative contact resistance,<br />
Rc .<br />
AL electrode<br />
SiQDs/Sio 2 SL layers<br />
Keithley 617<br />
Electrometer<br />
w=200-400mm<br />
Quartz substrate<br />
Figure 4.5.13: Schematic <strong>of</strong> sample<br />
structure for lateral conduction<br />
measurements.<br />
Figure 4.5.14 shows current densities vs. applied voltage in the dark for a range <strong>of</strong> Si<br />
concentrations in the SRO layers. It shows that conductivity generally increases as Si<br />
concentration increases, probably because <strong>of</strong> an increase in carrier transport as the average<br />
distance between neighbouring Si QDs decreases. Such transport would either be via a direct<br />
tunnelling mechanism through the oxide or by hopping via defects in the oxide, as shown<br />
schematically in Figure 4.5.12. This has been investigated further by measuring conductivity<br />
against temperature. The slopes <strong>of</strong> the data in Figure 4.5.15 show activation energies in the<br />
range 0.31 to 0.64eV. The variation <strong>of</strong> conductivity with Si concentration is more diffi cult to<br />
determine. Pre-exponential factors seem to vary from 0.01 to 100Scm -1 , but extrapolation<br />
<strong>of</strong> the slopes is not very conclusive. Work using this technique is continuing.<br />
Figure 4.5.14:<br />
J-V characteristics <strong>of</strong> SiQDs<br />
superlattice samples with<br />
various %Si concentrations.<br />
Al contact spacing is 300 µm.<br />
(All samples had 20 bilayers<br />
<strong>of</strong> 5nm SRO and 10nm SiO2<br />
layers.)<br />
Figure 4.5.15:<br />
Plot <strong>of</strong> conductivity vs.<br />
temperature showing activation<br />
energy and pre-exponential<br />
factors for another set <strong>of</strong><br />
samples with nominally the<br />
same parameters.<br />
80
4.5.4 Modelling<br />
Modelling is important for the interpretation <strong>of</strong> characterisation data and for designing new<br />
structures and experiments. In 2005 modelling techniques have been developed further,<br />
with additional improvements expected to continue.<br />
4.5.4.1 Modelling <strong>of</strong> Si nanostructures<br />
Researchers:<br />
Martin Green, Thipwan Fangsuwannarak, Chu-Wei Jiang, Dirk Kőnig<br />
Effective mass approximation approaches are being used to model expected confi ned energy<br />
levels in experimental structures. However as the QD size decreases towards the Bohr<br />
radius, the effective mass approximation will become increasingly inaccurate. Work is<br />
underway to calculate the energy levels using ab-initio calculations based on density functional<br />
theory (DFT). These will be compared to experimental data and effective mass solutions. The<br />
ab-initio calculations are expected to be a powerful tool in modelling a wide range <strong>of</strong> potential<br />
structures.<br />
4.5.4.2 Simulation <strong>of</strong> HRTEM and Electron Diffraction <strong>of</strong> Si nanoparticales<br />
Researchers:<br />
Tom Puzzer, Giuseppe Scardera<br />
Nanoscale structured materials and systems play a central role in the development <strong>of</strong><br />
Third Generation solar cell structures. The physical parameters determining the ultimate<br />
nanoparticle properties are the elemental composition, size, external shape and internal<br />
structures. Characterising these materials and their nanostructures is an essential part <strong>of</strong><br />
the research project. Although other techniques, such as X-ray diffraction and selected-area<br />
electron diffraction, can provide accurate structural information (averaged over the excited<br />
sample volume), they are incapable <strong>of</strong> providing information on the structure <strong>of</strong> individual<br />
nanoparticles. The information contained in the diffraction pattern, however, extends much<br />
further in reciprocal space than is accessible to HRTEM. Since the various types <strong>of</strong> microscope<br />
aberrations do not infl uence electron diffraction, high resolution electron diffraction data can<br />
be obtained without imaging artefacts.<br />
Despite the diffi culties involved with HRTEM, it remains the most powerful technique to<br />
examine these materials at sub-nanometre resolution. It is important to remember,<br />
however, that the image formation system <strong>of</strong> the TEM is not perfect. The contrast<br />
transfer function contains zeros and inversions meaning that some spatial frequencies<br />
present in the sample are not observed in the images and other spatial frequencies<br />
are imaged with reversed contrast. Therefore, there is a problem with unambiguous<br />
interpretation <strong>of</strong> individual high-resolution images and only under well-defi ned conditions<br />
may high-resolution images be naively interpreted in terms <strong>of</strong> the projected potential <strong>of</strong><br />
the atomic columns aligned along the beam direction.<br />
In the past, the standard approach to extracting structural information from HRTEM<br />
images <strong>of</strong> periodic objects is to match a series <strong>of</strong> experimental and simulated images,<br />
usually as a function <strong>of</strong> the objective lens focus setting – a through-focus series.<br />
Computer simulations, based on the multislice algorithm [4.5.10], have been developed<br />
to a relatively mature level and now may be routinely used to simulate images <strong>of</strong> largearea<br />
(semi-infi nite) uniform crystalline material for comparison with experimentally<br />
acquired images. The calculation has two main parts, a determination <strong>of</strong> the electron<br />
wave-function at the exit-surface <strong>of</strong> the specimen followed by propagation through the<br />
imaging system to the fi nal image plane. 81
Generating a description <strong>of</strong> the sample in a suitable form for the multislice algorithm<br />
can be the most diffi cult part <strong>of</strong> the simulation problem. The scattering potential <strong>of</strong><br />
many crystalline materials, oriented along a high-symmetry direction with respect to<br />
the electron beam, can be described as a repetitive sequence <strong>of</strong> a small number <strong>of</strong><br />
identical layers and is a comparatively straightforward problem. The problem is more<br />
diffi cult for nanoparticles, as the scattering potential is no longer strictly periodic in the x-<br />
y plane and the number <strong>of</strong> atoms in each slice varies along the z-, or beam direction. The<br />
particle also needs to be embedded in a suffi ciently large supercell to avoid the effects<br />
<strong>of</strong> wraparound errors. Similar problems are encountered when simulating images <strong>of</strong><br />
defects and crystal surfaces.<br />
To examine the image simulation process, a large number <strong>of</strong> images <strong>of</strong> particles <strong>of</strong><br />
various sizes, shapes and orientations have been performed and a library <strong>of</strong> images is<br />
being generated. This will form the basis <strong>of</strong> an image-matching system which will be able<br />
to identify particular crystal structures found in experimentally acquired images. As an<br />
example, a series <strong>of</strong> images along high-symmetry orientations has been simulated for a<br />
32Å wide truncated cubeoctahedral Si nanoparticle containing 1251 atoms. The atomic<br />
coordinates are based on a bulk Si lattice which has been truncated along 8 sets <strong>of</strong> {111}<br />
planes and 6 sets <strong>of</strong> {100} planes. For the (110) and (111) orientations, the particle<br />
is rotated about the z-axis, followed by a tilt around an axis perpendicular to the z-axis.<br />
Coordinate data is generated using Mathematica [4.5.11] and centred in a 100Å x<br />
100Å x 60Å supercell, hardsphere<br />
images are rendered<br />
using the VMD package<br />
[4.5.12] and the simulated<br />
images are generated<br />
using the Kirkland Multislice<br />
code [4.5.10] at electron<br />
beam energy <strong>of</strong> 200kV<br />
and Cs=1mm for defocus<br />
settings <strong>of</strong> 613 (Scherzer),<br />
900 and 1200Å. The effects<br />
<strong>of</strong> beam divergence, beam tilt<br />
and defocus spread are not<br />
included. A contrast reversal<br />
occurs between 613Å and<br />
900Å and the particle<br />
appears to be at least 1<br />
fringe larger at 1200Å than<br />
at 613Å. The apparent<br />
particle size is approximately<br />
10% larger than the actual<br />
size.<br />
Figure 4.5.16: The highsymmetry<br />
orientations <strong>of</strong><br />
the truncated cuboctahedral<br />
Si nanoparticle. The fi rst<br />
panel shows a hard-sphere<br />
structural image in real space<br />
<strong>of</strong> a particle viewed along<br />
the (100), (110) and (111)<br />
directions. The second panel<br />
shows the corresponding<br />
simulated HRTEM images.<br />
82
4.5.5 Up-conversion<br />
Researchers:<br />
Avi Shalav, Bryce Richards, Thorsten Trupke, Peter Würfel, Martin Green<br />
A photovoltaic Up-conversion (PV-UC) device is shown schematically in Figure 4.5.17. The<br />
concept aims to absorb sub-bandgap photons in an Up-converting layer behind a bifacial solar<br />
cell. This layer has the property <strong>of</strong> radiatively absorbing two or more long wavelength photons<br />
and emitting a photon <strong>of</strong> higher energy above the bandgap <strong>of</strong> the bifacial solar cell. Thus the<br />
current in the device is boosted by photons which would not normally be absorbed. It tackles<br />
loss (1) in Figure 4.5.1 and strategy (a) discussed above in Section 4.5.1).<br />
Figure 4.5.17: Schematic <strong>of</strong> an<br />
Up-conversion layer, designed to<br />
capture sub-bandgap light, behind<br />
a bifacial solar cell.<br />
Incident light<br />
bifacial solar cell<br />
up-converter<br />
reflector<br />
The response <strong>of</strong> such a device<br />
depends primarily on the<br />
luminescent quantum effi ciency<br />
(LQE) <strong>of</strong> the up-conversion (UC)<br />
phosphors. Using EQE measurements, the average response between 1480 and 1580nm is<br />
about a third the maximum response at 1523nm. The best case scenario for a PV-UC device<br />
is to assume there are minimal losses. In this case, the measured EQE will be approximately<br />
equal to the LQE (at sub-bandgap wavelengths). Using these idealised values for the EQE, it<br />
is possible to calculate the near infraredNIR effi ciency enhancement η uc <strong>of</strong> the PV-UC device<br />
at NIR wavelengths under AM1.5 conditions. The total effi ciency for such a device will be<br />
the energy conversion effi ciency <strong>of</strong> the solar cell, η, plus that <strong>of</strong> the UC enhancement η uc .<br />
Figure 4.5.18 shows the calculated η uc values using the average value for the LQE from 1480-<br />
1580nm for various solar AM1.5 concentrations. One <strong>of</strong> the primary reasons for this low<br />
effi ciency enhancement is because only a narrow range <strong>of</strong> wavelengths are utilised (1480-<br />
1580nm). Even if the PV-UC device had a LQE <strong>of</strong> 100% for this wavelength range, the value<br />
<strong>of</strong> η uc would only be 2.4%. A number <strong>of</strong> methods to improve the spectral sensitivity (or<br />
robustness) <strong>of</strong> PV-UC devices are being investigated. The methods include co-doped rareearth<br />
systems and quantum dots.<br />
4.0<br />
0.10<br />
Figure 4.5.18: Average LQE<br />
and NIR PV-UC effi ciency<br />
enhancement, η uc , versus<br />
AM1.5G solar concentration.<br />
The solar concentration was<br />
calculated using the broad<br />
absorption properties <strong>of</strong> the<br />
phosphors (1480-1580nm).<br />
LQE [%]<br />
3.5<br />
3.0<br />
2.5<br />
2.0<br />
1.5<br />
theoretical limit (46200x)<br />
0.09<br />
0.08<br />
0.07<br />
0.06<br />
0.05<br />
0.04<br />
0.03<br />
1.0<br />
0.02<br />
0.5<br />
0.01<br />
0.0<br />
0<br />
0.00<br />
20000 40000 60000 80000 100000<br />
number <strong>of</strong> suns (concentration)<br />
83
Many lanthanide energy levels are below about 10000 cm -1 (with emission wavelengths less<br />
than about 1000nm) making them potential UC luminescent centres for silicon PV-UC devices.<br />
Unfortunately, closely spaced energy levels result in very large non-radiative relaxation.<br />
Investigation <strong>of</strong> the Dieke diagram suggests that the only promising co-doped ETU system<br />
that may improve the spectral robustness <strong>of</strong> a silicon PV-UC device is one based on Dy 3+<br />
and Er 3+ ions. This is illustrated in Figure 4.5.19. The 6 H 7/2 and 6 F 9/2 levels <strong>of</strong> Dy 3+ can<br />
be excited using 1100-1120nm photons, the 6 H 9/2 and 6 F 11/2 by 1290-1320nm photons,<br />
the 6 H 9/2 by 1700-1730nm photons. These excited ions quickly non-radiatively relax down<br />
to the 6 H 13/2 energy level. This level can also be excited by 2800-2900nm photons, but the<br />
number <strong>of</strong> photons at these wavelengths from the solar spectrum is negligible. Nevertheless,<br />
the 6 H 13/2 level becomes highly populated from the non-radiative relaxation from the higher<br />
levels. This energy level is resonant with the 4 I 11/2 - 4 I 13/2 <strong>of</strong> Er 3+ . This implies that energy<br />
transfer up-conversion (ETU) from the 4 I 13/2 to the 4 I 11/2 from the Er 3+ ions is likely if the<br />
Er 3+ are in their fi rst excited state. This fi rst excited state can occur with 1480-1580 nm<br />
photons (in the NaYF 4 host). Due to the rapid non-radiative relaxation <strong>of</strong> higher energy levels<br />
in the Dy 3+ ions, and non-resonant conditions between the Dy3+ and Er3+ ions, higher order<br />
UC processes or cross-relaxation are unlikely. A signifi cant amount <strong>of</strong> the sub-bandgap NIR<br />
spectral irradiance can be utilised in this co-doped system.<br />
Figure 4.5.19: ETU between Dy 3+ and<br />
Er 3+ ions. Solid up arrows represent<br />
GSA. The wavy lines represent rapid nonradiative<br />
relaxation. The dotted arrows<br />
represent the ETU mechanism. The solid<br />
down arrow represents the resulting UC<br />
luminescence.<br />
In the above case, the Dy 3+ ions ‘downshift’<br />
higher energies to lower ones which<br />
better match the resonant conditions <strong>of</strong><br />
the Er 3+ ions. Although this is an energy<br />
loss mechanism, improved UC responses<br />
will most likely result.<br />
Quantum dots (QDs) could also be used to down-shift higher energy photons to lower energy<br />
photons which are within the absorption range <strong>of</strong> the UC phosphors. ZnS-passivated CdSe<br />
and PbSe quantum dots have ideal absorption and emission properties for PV-UC QD devices<br />
and are readily commercially available (for example from Evident Technologies). These<br />
quantum dots have large EQE’s (over 50% for the CdSe) and high indices <strong>of</strong> refraction<br />
compared to the phosphors, making them better matched for Si devices. Figure 4.5.20<br />
shows typical absorption and emission spectra for PbSe quantum dots. The energy transfer<br />
will probably occur by radiative emission (from the quantum dot) followed by absortpion by the<br />
UC phosphor. Direct coupling between the PbSe and UC phosphors resulting in ETU may also<br />
occur and may prove to be a dominant UC mechanism.<br />
84
absorption<br />
emission<br />
absorption, emission [au]<br />
Figure 4.5.20:<br />
Absorption/emission<br />
from three different PbSe<br />
quantum dot samples. The<br />
emission <strong>of</strong> the quantum<br />
dots coincide with the<br />
absorption <strong>of</strong> Er 3+ doped<br />
phosphors.<br />
1200 1300 1400 1500 1600 1700<br />
wavelength [nm]<br />
4.5.6 Hot Carrier Cell<br />
The concept underlying the hot carrier solar cell is to slow the rate <strong>of</strong> photoexcited carrier<br />
cooling, caused by phonon interaction in the lattice, to allow time for the carriers to be<br />
collected whilst they are still at elevated energies (“hot”) and thus allow higher voltages to be<br />
achieved from the cell [4.5.13]. This approach thus tackles the major PV loss mechanism <strong>of</strong><br />
thermalisation <strong>of</strong> carriers (loss (2) in Figure 4.5.1 and strategy (b) discussed above in Section<br />
4.5.1). In addition to an absorber material that slows the rate <strong>of</strong> carrier relaxation, a hot<br />
carrier cell must allow extraction <strong>of</strong> carriers from the device through contacts which accept<br />
only a very narrow range <strong>of</strong> energies (selective energy contacts).<br />
Within the Third Generation Strand, the parallel project, funded by Toyota, on thermoelectric<br />
and energy conversion devices has strong parallels with the Hot Carrier cells project. They<br />
both require the selective energy contacts discussed below but in addition the Hot Carrier<br />
cell requires an absorber in which carrier thermalisation is very signifi cantly slowed from the<br />
picosecond timescale <strong>of</strong> normal semiconductors, to closer to the nanosecond timescale <strong>of</strong><br />
carrier recombination.<br />
Experimental work in the Third Generation Strand is currently focussed on selective energy<br />
contacts using resonant tunnelling structures based on a similar Si quantum dot approach<br />
to that discussed in Section 4.5.2. Theoretical work is directed at investigating mechanisms<br />
for slowing carrier cooling by modifying phononic band structures in nanostructure<br />
superlattices.<br />
85
4.5.6.1 Selective Energy Contacts<br />
Researchers:<br />
Dirk König, Chu-Wei Jiang, Gavin Conibeer, Martin Green<br />
For the hot carrier cell to be effective “hot” carriers, as described above, must be collected<br />
from the absorber over a very small energy range with selective energy contacts. This is so<br />
that only a small amount <strong>of</strong> their excess energy above the band edge is lost in cooling to the<br />
contact temperature. In thermodynamic terms, the carriers are thus collected with a very<br />
small increase in entropy. Ideally this collection would be isoentropic using mono-energetic<br />
contacts.<br />
Double barrier resonant tunnelling contacts are being investigated as selective energy<br />
contacts, with quantum dots providing a discrete energy level between two insulating barriers.<br />
This will give conduction strongly peaked at the discrete energy level. The total energy fi ltering<br />
<strong>of</strong> a quantum dot based structure is required <strong>of</strong> a selective energy contact rather than 1D<br />
energy fi ltering because the 1D energy fi ltering in a quantum well resonant tunnelling device<br />
is only effective for carriers with momenta entirely perpendicular to the plane <strong>of</strong> the well<br />
[4.5.14]. Carriers with components <strong>of</strong> momenta away from this normal can be transmitted<br />
even if their total energy is outside the range <strong>of</strong> the energy fi lter. Hence work in the Centre<br />
is focussed on resonant tunnelling structures using quantum dots or other discrete total<br />
energy confi ned centres as the resonant centres. These would give the required total energy<br />
fi ltering. Such a fi lter should exhibit negative differential resistance (NDR).<br />
A signifi cant pro<strong>of</strong> <strong>of</strong> concept was achieved in 2004 (as reported in the 2004 annual<br />
report) with negative differential resistance observed at room temperature for a double<br />
barrier quantum dot structure [4.5.15]. The device structure is based on our Si quantum dot<br />
technology discussed previously.<br />
In addition, modelling <strong>of</strong> the 1D resonant tunnelling has been carried out with a model<br />
developed for perturbations in the position <strong>of</strong> a defect in an oxide barrier [4.5.16]. The theory<br />
is being extended to a 2D model.<br />
4.5.6.1.1 Conductive atomic force microscopy<br />
In 2005, further characterisation <strong>of</strong> the resonant tunnelling structures has been carried out<br />
using conductive atomic force microscopy (CAFM) through a collaboration with Dr Chris Pakes<br />
at Melbourne <strong>University</strong>. CAFM is a powerful tool for investigating the transport properties<br />
<strong>of</strong> nanostructures. Data are obtained by modifying the local band pr<strong>of</strong>i le - on a scale <strong>of</strong> a few<br />
nanometres - using an applied bias at the conductive probe tip. This technique can be used to<br />
map the current-voltage characteristics <strong>of</strong> a single layer <strong>of</strong> Si quantum dots, below an oxide<br />
layer <strong>of</strong> a few nanometres, at a very high areal resolution. Hence it can be applied to the<br />
mapping <strong>of</strong> resonant tunnelling sites on a sample. The results so far indicate that CAFM is<br />
useful for verifying the location and density <strong>of</strong> quantum dots non-destructively. Samples with<br />
thinner oxide layers are now being produced to increase the tunnelling current to allow lower<br />
bias CAFM scans and hence to estimate the resonant energy levels more accurately.<br />
4.5.6.1.2 Optically assisted I-V<br />
Conventionally, tunnelling IV measurements on a MIS structure involve a strong bias fi eld so<br />
as to obtain hot electrons at the Selective Energy Contact / Si interface, which increases<br />
the tunnelling current density such that it can be detected. For a resonant structure, the<br />
resulting dI/dV characteristic would show a blurred negative differential resistance (NDR)<br />
feature.<br />
86
However with a monochromatic optical excitation <strong>of</strong> high intensity, this shortcoming can be<br />
circumvented. If hot carriers can be generated near to the barrier, a hot carrier population is<br />
created at the Selective Energy Contact. This also requires that the optical generation rate<br />
within the semiconductor is high enough and/or that the current is measured within a few<br />
picoseconds <strong>of</strong> the massive generation <strong>of</strong> hot carriers. Using a tuneable monochromatic<br />
light source the energy range <strong>of</strong> the hot carrier population can be chosen for probing the<br />
relevant energetic portion <strong>of</strong> the Selective Energy Contact. In addition, in order to collect the<br />
electrons on the metal electrode, a weak bias fi eld may be necessary. By keeping this fi eld<br />
weak, the band bending <strong>of</strong> the barrier is minimized, resulting in a maximum overlap <strong>of</strong> the<br />
subband DOS on both sides <strong>of</strong> the barrier.<br />
As a pro<strong>of</strong> <strong>of</strong> concept, measurement has so far only been carried out with a broad band<br />
Xenon light source. It comprises:<br />
• A quartz halogen illumination source.<br />
• Short pass fi lter (= 1000 nm): cutting <strong>of</strong>f IR in order to minimize heating <strong>of</strong> sample<br />
• Long pass fi lter ( = 495 nm): preventing optically assisted thermionic emission <strong>of</strong><br />
electrons over the oxide barrier (which would falsify the IV measurements)<br />
• Plane-convex lens<br />
• Faraday box used for IV measurements with a lid with a quartz window.<br />
Figure 4.5.21: Optically assisted<br />
I-V measurement apparatus.<br />
A mesa sample with a single layer <strong>of</strong> Si QDs, with no p-n junction, was measured, as shown<br />
in Figure 4.5.22. The area <strong>of</strong> each mesa = 1/16 cm 2 and this contains about 10 10 Si QDs.<br />
The sample is nominally the same as those used to obtain negative differential resistance in<br />
normal I-V during 2004 (see 2004 Annual report).<br />
Figure 4.5.22: Sample<br />
structure for optically<br />
assisted I-V measurements<br />
(the same as that<br />
used in earlier dark I-V<br />
measurements).<br />
The enhanced current at all biases shown in Figure 4.5.23a for optically assisted<br />
measurements as compared to dark I-V, is evidence for an enhanced hot carrier population<br />
near the contact. The increase cannot be explained purely by an increase in Fermi-distribution<br />
with temperature as the temperature increase above room temperature is calculated to be<br />
only a few tens <strong>of</strong> degrees, suggesting a non-equilibrium hot carrier population exists. There is<br />
also tentative evidence for a resonance in the data in the dark at a bias <strong>of</strong> about 3.5V (Figure<br />
4.5.23b) which, under illumination, both increases in magnitude and decreases in required<br />
87
ias to about 2.2V (Figure 4.5.23c). This is further tentative evidence for the collection <strong>of</strong><br />
hot carriers, as the energies appropriate for collection by the contact are generated at lower<br />
bias voltages with optical assistance. However it should be pointed out that measurements<br />
are made at present without the presence <strong>of</strong> additional fi lters to control the spectrum <strong>of</strong><br />
light illuminating the sample. In subsequent measurements, such fi lters - either providing a<br />
narrow range <strong>of</strong> energies or a variable sharp long wavelength pass cut-<strong>of</strong>f – will allow more<br />
distinct quantitative data to be collected.<br />
Figure 4.5.23: a) Dark and illuminated, or optically assisted, I-V measurements.<br />
Detail <strong>of</strong> b) dark and c) illuminated I-V traces.<br />
4.5.6.2 Slowed carrier cooling<br />
Researchers:<br />
Gavin Conibeer, Jean-Francois Guillemoles, Martin Green<br />
Slowing <strong>of</strong> carrier cooling has been observed in bulk semiconductors by several authors, but<br />
only at very high illumination intensities – typically equivalent to 250 to 2500 suns [4.5.17].<br />
However, the reduced dimensionality in nanostructures <strong>of</strong>fers a promising route for such a<br />
role as this restricts the number <strong>of</strong> acoustic phonon modes available. because <strong>of</strong> the Brillouin<br />
zone folding that occurs as a result <strong>of</strong> the periodicity <strong>of</strong> the superlattice. Slowed carrier<br />
cooling in QW superlattices has been demonstrated experimentally [4.5.17, 4.5.18] and is<br />
described as an enhancement <strong>of</strong> the phonon bottleneck effect. Some <strong>of</strong> the current authors<br />
describe the likelihood <strong>of</strong> an increased effect in QD superlattices elsewhere [4.5.19]. This<br />
work is principally theoretical at present.<br />
4.5.6.2.1 Modelling <strong>of</strong> phononic band structures<br />
Investigation in this area is currently focussed on obstructing the phonon decay mechanisms<br />
for carrier cooling. This potentially has relevance to any device which relies on reducing<br />
thermal transport, including thermoelectrics.<br />
The current work [4.5.19] stems from the realisation that hot carriers lose their energy by<br />
scattering with optical phonons which in turn build up a hot non-equilibrium population. Thus,<br />
the critical feature in the overall carrier cooling is the rate <strong>of</strong> decay <strong>of</strong> the hot optical phonons.<br />
It has been identifi ed by other workers that the principal and perhaps only mechanism for this<br />
is the decay <strong>of</strong> an optical phonon into two longitudinal acoustic (LA) phonons <strong>of</strong> energy half<br />
that <strong>of</strong> the optical phonon and <strong>of</strong> equal and opposite momenta.<br />
In some bulk semiconductors, with a large difference in their anion and cation masses, there<br />
can be a large gap between the highest acoustic phonon energy and the lowest optical phonon<br />
88
energy. In a few cases – such as InN – this gap can be larger than the highest acoustic energy<br />
and hence prevent operation <strong>of</strong> the optical phonon decay mechanism described above, see<br />
Figure 4.5.24.<br />
Figure 4.5.24: Phonon dispersion for InN<br />
[4.5.20] in which E LO >2E LA such that LO<br />
→ 2LA (only) is forbidden. (E LO ; LA ) denote<br />
longitudinal optical and acoustic phonon<br />
energies.)<br />
The current work is investigating achieving a similar effect in less exotic materials possibly<br />
based on silicon, by exploiting the refl ection that occurs at the mini-Brillouin zone boundaries<br />
<strong>of</strong> nanostructure superlattices. This opens up mini-gaps in the phonon dispersion for those<br />
acoustic phonon energies which satisfy the Bragg condition. Figure 4.5.25 and Figure 4.5.26<br />
show phonon dispersions modelled using a nearest neighbour elastic continuum model<br />
based on Rytov calculations, for quantum well (QW) and quantum dot (QD) superlattices<br />
respectively.<br />
Figure 4.5.25: Acoustic phonon dispersion<br />
and DOS calculated using an elastic<br />
continuum model based on the Rytov<br />
formulation for QW and barrier thicknesses<br />
<strong>of</strong> 5 lattice spacings and ratio <strong>of</strong> acoustic<br />
impedances <strong>of</strong> 4.<br />
Figure 4.5.26: Acoustic phonon dispersion<br />
for multiple QD array showing a folded<br />
dispersion with mini-gaps in all directions<br />
in real and reciprocal space, hence giving<br />
true mini-gaps in the DOS.<br />
89
The phonon dispersions show that if the superlattice structure is engineered correctly, the<br />
mini-gaps can in principle be arranged such that they prevent the decay <strong>of</strong> optical phonons by<br />
interrupting the same mechanism. However the density <strong>of</strong> states (DOS) in Figure 4.5.25 and<br />
Figure 4.5.26 indicate that complete gaps in the DOS for all directions in reciprocal space<br />
only occur for the QD superlattice. Hence this mechanism <strong>of</strong> slowing carrier cooling should<br />
only be present in QD structures. Experimental evidence for this in the literature is being<br />
assessed and effects on effi ciencies calculated. However, to achieve such effects in practice,<br />
the structure would need to be fi nely tuned.<br />
A further effect under investigation is that <strong>of</strong> the QD interface with the matrix, on phonon<br />
dispersion. Figure 4.5.27 shows that with a s<strong>of</strong>t interface the mini-gaps group into doublets<br />
with a larger gap in the centre. This would improve the ease <strong>of</strong> fi ne tuning the QD superlattice<br />
such as to block the optical phonon decay. Further work on modelling such interfaces is<br />
underway.<br />
Figure 4.5.27: Dispersion curve <strong>of</strong><br />
acoustic phonons for 1nm3 QDs in a<br />
superlattice with 1nm spacing between<br />
QDs – incorporating s<strong>of</strong>t interface modes<br />
between QDs and matrix.<br />
Finally, some progress has been made<br />
towards predicting the effi ciencies that a<br />
Hot Carrier cell might achieve using the<br />
non-ideal properties <strong>of</strong> existing materials<br />
[4.5.21]. Preliminary computations show<br />
that, at the illumination level used for the<br />
experiments in [4.5.17, 4.5.18] (equivalent to about 2500 suns), effi ciencies <strong>of</strong> 50% could be<br />
achievable. This is in spite <strong>of</strong> thermal losses in state <strong>of</strong> the art materials such as the multiple-<br />
QW structures discussed elsewhere [4.5.17, 4.5.18], even though these have a non optimal<br />
band gap <strong>of</strong> 1.5 eV. This result compares to a calculated 54% effi ciency at these concentrations<br />
without any thermalisation losses at all [4.5.13]. Hence, at these concentrations the loss to<br />
thermalisation for existing materials is not large. However, these fi gures assume perfect<br />
energy selection at the contacts. Therefore, as operation at lower concentrations would be<br />
preferable and because <strong>of</strong> the diffi culties in the fabrication <strong>of</strong> optimal contacts, it is desirable<br />
to search for further reduction in carrier cooling and/or a reduction in the threshold injection<br />
levels at which this occurs. Phononic band gap engineering in superlattices seems a potential<br />
route to achieving this.<br />
4.5.7 Supporting project areas<br />
There are three projects associated with Third Generation Strand work but not directly<br />
concerned with either <strong>of</strong> Si nanostructures, Up-converters or Hot Carrier cells. These are<br />
continuing work on limiting effi ciency computer programs, coupling <strong>of</strong> light with surface<br />
plasmons and quantum antennae.<br />
90
4.5.7.1 Limiting efficiency programs<br />
Researchers:<br />
James Rudd, Gavin Conibeer, Jean-Françoise Guillemoles<br />
In the 2004 annual report progress on development <strong>of</strong> these programs was reported to the<br />
stage <strong>of</strong> initial fabrication <strong>of</strong> a graphical user interface (GUI). In 2005 this GUI was developed<br />
further and tested for a range <strong>of</strong> cell types. This work was carried out in a Taste <strong>of</strong> Research<br />
project and in an undergraduate thesis project. Screenshots <strong>of</strong> the resulting GUI are shown<br />
in Figure 4.5.28.<br />
Figure 4.5.28: GUI for limiting effi ciency programs.<br />
The programs were originally written in C++. The thesis work has transferred this to C.net<br />
and tidied up much <strong>of</strong> the code. The program works on a particle balance approach which<br />
equates the number <strong>of</strong> incident photons with the electron current plus the number <strong>of</strong> emitted<br />
photons. These calculations are performed using the generalised Planck expression for<br />
photon fl uxes:<br />
Solving this equation requires use <strong>of</strong> numerical integration methods as the -1 term in the<br />
denominator introduces a singularity preventing the use <strong>of</strong> simple integration routines, such<br />
as Simpson’s rule [4.5.22].<br />
91
In addition the thesis project investigated ways to automate the input parameters to the<br />
program – a signifi cant challenge for multi-energy level cells – and developed empirical<br />
relationships based on previously run scenarios. The formulae used to describe these<br />
relationships are in turn giving indications <strong>of</strong> the physical processes behind the calculations.<br />
The program is now at the stage <strong>of</strong> testing by members <strong>of</strong> the Third Generation strand.<br />
4.5.7.2 Surface Plasmons<br />
Researchers:<br />
Supriya Pillai, Kylie Catchpole, Thorsten Trupke<br />
Surface plasmons are the oscillation <strong>of</strong> conduction electrons in metal particles. The excitation<br />
<strong>of</strong> surface plasmons is a promising way <strong>of</strong> increasing light absorption in thin-Si based devices.<br />
According to the generalised Kirch<strong>of</strong>f’s law, enhanced absorption can also lead to enhanced<br />
emission. Thus the work is also <strong>of</strong> interest in Si light emission for effi cient light emitting diodes<br />
and is discussed in Section 4.6.<br />
Silver particles are formed on the surface by evaporation followed by annealing. As reported<br />
in Section 4.6, the seven fold enhancement in electroluminescence and spectral response<br />
(Figure 4.5.29 and Figure 4.5.30 respectively) is very promising for signifi cant enhancement<br />
<strong>of</strong> absorption in the visible range. Work is continuing on optimisation <strong>of</strong> silver particle size<br />
and shape and <strong>of</strong> shifting the peak <strong>of</strong> absorption to more appropriate wavelengths, using an<br />
overlayer <strong>of</strong> a high refractive index layer such as ZnS.<br />
EL / Photocurrent Enhancement<br />
8<br />
7<br />
6<br />
5<br />
4<br />
3<br />
2<br />
1<br />
EL<br />
SR<br />
0<br />
300 400 500 600 700 800 900 1000 1100 1200 1300<br />
Wavelength(nm)<br />
Enhancement<br />
8<br />
7<br />
6<br />
5<br />
4<br />
3<br />
2<br />
1<br />
with Ag<br />
with Ag+ZnS<br />
0<br />
900 1000 1100 1200 1300<br />
Wavelength (nm)<br />
Figure 4.5.29: Enhancement plots showing absorptance and emission from 95nm thin SOI<br />
devices.<br />
Figure 4.5.30: EL enhancement plot showing the shift in the peak using ZnS overcoating.<br />
Enhancement <strong>of</strong> absorption is likely to be <strong>of</strong> key importance in Si based nanostructure devices.<br />
This is because the devices will be thin and hence adequate absorption will be challenging<br />
even if the materials behave as direct bandgap devices, which is likely due to localisation <strong>of</strong><br />
electrons and holes due to quantum confi nement. Hence the approach is currently being<br />
investigated for Si nanostructure materials.<br />
92
4.5.7.3 Antenna Collection <strong>of</strong> Solar Energy<br />
Researchers:<br />
Richard Corkish, Samantha Wong, Gavin Conibeer<br />
Research on the potential use <strong>of</strong> antennae for conversion <strong>of</strong> solar energy to direct current<br />
electricity was revived during 2005, with a focus on the ultimate effi ciencies that could be<br />
achieved.<br />
Coherent, polarised, narrow-band radio waves may already be received and converted to<br />
direct current by rectifying antennae, or “rectennas” but the conversion <strong>of</strong> broad spectrum,<br />
incoherent, unpolarised solar radiation is more challenging. This work has continued to derive<br />
insights from the fi eld <strong>of</strong> radio astronomy, where the accurate detection <strong>of</strong> such radiation is<br />
well established. In particular, we have used the model <strong>of</strong> the “Dicke radiometer” to show that<br />
the electrical noise output <strong>of</strong> an antenna receiving solar radiation can be represented as that<br />
<strong>of</strong> a resistor at the sun’s black body temperature.<br />
The effi ciency <strong>of</strong> conversion <strong>of</strong> the noise power from a resistor has been studied already<br />
by others but our attempts to reproduce published results in a key paper repeatedly failed<br />
and discussions with the author led to a correction (http://summa.physik.hu-berlin.de/<br />
~ sokolov/).<br />
That work investigated the rectifi cation by a diode, at one temperature, <strong>of</strong> the electrical noise<br />
from a resistor at a higher temperature. Using the detailed balance method, the calculation<br />
showed that the effi ciency <strong>of</strong> a system with a diode rectifi er as a heat engine must always be<br />
less than the Carnot effi ciency. The reason for this thermodynamic irreversibility is that even<br />
when no work is done, heat transport takes place between the heat baths.<br />
The Carnot effi ciency increases with temperature. For an assumed black body temperatures<br />
<strong>of</strong> 6000K for the sun and 300K for the receiver, the Carnot effi cieny equals 0.95. Figure<br />
4.5.31 is a plot <strong>of</strong> the effi ciency <strong>of</strong> an ideal rectenna against the current drawn from the<br />
device, with a peak effi ciency at 48.81%. (Dimensionless current is shown as this does not<br />
depend on geometrical factors.)<br />
Camot Efficiency = 0.95<br />
Figure 4.5.31: Rectenna effi ciency vs.<br />
dimensionless current for a Carnot effi ciency<br />
<strong>of</strong> 95%.<br />
0.5<br />
0.4<br />
While signifi cant practical challenges<br />
remain for the possible implementation <strong>of</strong><br />
solar rectennas, the Centre’s work has now<br />
developed a model to predict their ultimate<br />
effi ciency.<br />
Efficiency<br />
0.3<br />
0.2<br />
-0.1<br />
0<br />
-4.5 -4 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 -0<br />
Normalised Current (dimensionless)<br />
93
4.5.8 Concluding remarks for the Third Generation section<br />
The preceding sections show the signifi cant advance in all the Third Generation projects in<br />
2005. Advances include the successful transfer <strong>of</strong> Si nanostructure technology from Si QDs<br />
in oxide matrix to a nitride matrix – both for sputtering and for PECVD - and preparation for<br />
transfer to a silicon carbide matrix. This should facilitate signifi cant improvements in electrical<br />
transport bringing the goal <strong>of</strong> a wider band gap Si based material much closer. This work has<br />
also been instrumental in leveraging signifi cant additional funding via the Global Climate and<br />
Energy Project, which started at the end <strong>of</strong> 2005.<br />
Up-conversion has moved beyond the pro<strong>of</strong> <strong>of</strong> concept <strong>of</strong> 2004, to demonstrate improved<br />
quantum effi ciencies and progress towards sensitising the devices to photons between the Si<br />
bandgap energy and the absorption energy <strong>of</strong> erbium at 1500nm. Pro<strong>of</strong> <strong>of</strong> concept <strong>of</strong> these<br />
is expected in 2006 as are moves towards other Up-conversion approaches.<br />
Hot Carrier cells have advanced in the characterisation <strong>of</strong> selective energy contacts using<br />
conductive AFM and optically assisted I-V. This work has been in parallel with the Toyotafunded<br />
thermoelectric and energy conversion project. There has also been development <strong>of</strong><br />
more sophisticated theory on reduction <strong>of</strong> carrier cooling in nanostructure superlattices.<br />
This work has started to fi nd application in prediction <strong>of</strong> hot carrier cell effi ciencies in real<br />
non-ideal material systems.<br />
The associated projects have continued to develop with the limiting effi ciency computer<br />
programs advancing to the testing stage, seven fold enhancement <strong>of</strong> electroluminescence<br />
from surface plasmons and an ideal effi ciency for quantum antennae <strong>of</strong> 48% predicted.<br />
94
References for Section 4.5<br />
4.5.1 M. A. Green, “Third Generation Photovoltaics: Ultra-High Effi ciency at Low Cost”<br />
(Springer-Verlag, 2003).<br />
4.5.2 J. Nelson, “The Physics <strong>of</strong> Solar Cells”, Imperial College press, 2003.<br />
4.5.3 E-C. Cho, Y.H. Cho, T. Trupke, R. Corkish, G. Conibeer, M.A. Green, Proc. 19th<br />
European Photovoltaic Solar Energy Conference (Paris, June 2004) 235.<br />
4.5.4 E-C. Cho, Y.H. Cho, R. Corkish, J. Xia, M.A. Green, D.S. Moon, Asia-Pacifi c<br />
Nanotechnology forum, (Cairns, 2003).<br />
4.5.5 M. Zacharias, J. Heitmann, R. Scholz, U. Kahlere, M. Schmidt, J. Blasing, Appl. Phys.<br />
Lett. 80 (2002) 661.<br />
4.5.6 K. Boer, “Survey <strong>of</strong> Semiconductor Physics”, van Nostrand Reinhold, 1990.<br />
4.5.7 M.A. Green, E-C. Cho, Y. Cho, Y. Huang, E. Pink, T. Trupke, A. Lin, T. Fangsuwannarak,<br />
T. Puzzer, G. Conibeer, R. Corkish, Proc. 20th European PVSEC, p3 (Barcelona, June<br />
2005).<br />
4.5.8 Y. Cho, E-C. Cho, Y. Huang, C-W. Jiang, G. Conibeer, M.A. Green, Proc. 20th<br />
European PVSEC, p47 (Barcelona, June 2005).<br />
4.5.9 B-H. Kim et al, Appl Phys Lett, 86 (2005) 091908.<br />
4.5.10 Kirkland, E. J., “Advanced Computing in Electron Microscopy” Plenum Press, <strong>New</strong><br />
York (1998).<br />
4.5.11 Wolfram Research, Inc., Mathematica, Version 5.2, Champaign, IL (2005).<br />
4.5.12 Humphrey, W., Dalke, A. and Schulten, K., “VMD- Visual Molecular Dynamics” J.<br />
Molec. Graphics 14 (1996) 33-38.<br />
4.5.13 P. Würfel, “Solar Energy Conversion with Hot Electrons from Impact Ionisation”, Sol.<br />
Energy Mats. and Sol. Cells. 46 (1997) 43.<br />
4.5.14 G.J. Conibeer, C-W. Jiang, M.A. Green, N. Harder, A. Straub, Proc. 3rd World Conf.<br />
on Photovoltaic Solar Energy Conversion, Osaka, Japan (2003).<br />
4.5.15 C-W. Jiang, E-C. Cho, G. Conibeer, M.A. Green, Proc. 19th European Photovoltaic<br />
Solar Energy Conference (Paris, June 2004) 80.<br />
4.5.16 C-W. Jiang, M.A. Green, E-C. Cho, G. Conibeer, J Appl Phys, 96 (2004) 5006.<br />
4.5.17 A.J. Nozik, C.A. Parsons, D.J. Dunlary, B.M. Keyes, R.K. Ahrenkiel, Solid State<br />
Comm. 75 (1990) 247.<br />
4.5.18 D.J. Westland et al, Solid Sate Eelectronics, 31(1988)431<br />
4.5.19 G.J Conibeer, J-F. Guillemoles, M.A. Green, “Phononic band gap eng. for Hot Carrier<br />
solar cell absorbers”, Proc. 20th EPVSEC (Barcelona, June 2005) p35.<br />
4.5.20 Davydov et al., Appl Phys Lett, 75 (1999) 3297.<br />
4.5.21 J-F. Guillemoles, G.J. Conibeer, M.A. Green, “Phononic Engineering with<br />
Nanostructures for Hot Carrier Solar Cells”,Proc. 15th PCSEC (Shanghai,<br />
Oct 2005).<br />
4.5.22 J. Rudd, G. Conibeer, J-F. Guillemoles, “User Friendly S<strong>of</strong>tware for Calculation<br />
<strong>of</strong> Limiting Cell Effi ciencies for Research on Multijunction Solar Cells”, Proc. “Solar<br />
2005” Australia and NZ Solar Energy Society conf. (Dunedin NZ, Nov 2005).<br />
95
4.6 SILICON PHOTONICS RESEARCH GROUP<br />
<strong>University</strong> Staff<br />
A/Pr<strong>of</strong>. Jianhua Zhao (group leader: integrated LEDs)<br />
A/Pr<strong>of</strong>. Thorsten Trupke (group leader: silicon laser)<br />
Pr<strong>of</strong>. Martin Green<br />
Pr<strong>of</strong>. Stuart Wenham<br />
Research Fellows<br />
Dr. Aihua Wang<br />
Adjunct Fellow<br />
Dr. K. Catchpole<br />
Research Staff<br />
Dr. Robert Bardos<br />
Dr. Tom Puzzer<br />
Research Assistant<br />
Guangchun Zhang<br />
Research Student<br />
Supriya Pillai<br />
Technical Staff<br />
Jules Yang<br />
Collaborators<br />
Susan Angus, Dr. Fay Hudson, A/Pr<strong>of</strong>. Andrew Dzurak, Centre <strong>of</strong> Excellence for Quantum<br />
Computing, UNSW<br />
Pr<strong>of</strong>. Peter Würfel, <strong>University</strong> <strong>of</strong> Karlsruhe<br />
4.6.1 Thick SOI LED Devices<br />
Research on Silicon On Insulator (SOI) Light Emitting Diodes (LEDs) continued in the Centre<br />
during 2005. Originally most <strong>of</strong> these SOI LEDs were fabricated on thin silicon layers aiming<br />
for a few nanometre silicon thickness to achieve quantum confi nement effects. The benefi t<br />
<strong>of</strong> such quantum confi nement is the increased light emission and absorption effi ciencies, and<br />
the ability to emit shorter wavelength light than the band-edge emission. However, no such<br />
quantum confi nement has been demonstrated to date due to the diffi culties in precise control<br />
<strong>of</strong> the processing conditions.<br />
In 2005, emphasis was changed to the fabrication <strong>of</strong> SOI LEDs on thicker silicon layers <strong>of</strong><br />
about 1.3 µm thickness. This was expected to result in better electrical performance for<br />
these LEDs, and also increased electroluminescence (EL) effi ciency. Figure 4.6.1 shows a<br />
schematic diagram <strong>of</strong> the cross sectional view <strong>of</strong> the light emitting diode design.<br />
97
Figure 4.6.1:<br />
Cross sectional view <strong>of</strong> the SOI<br />
light emitting diode.<br />
These thick silicon LEDs gave stronger EL signals, which allow more accurate measurement<br />
due to the limited sensitivity <strong>of</strong> our infrared sensors. Images <strong>of</strong> these SOI LEDs were taken<br />
using the Olympus microscope in combination with a CCD camera, in which the silicon CCD<br />
can catch the light up to 1 µm wavelength, enough to see the short wavelength end <strong>of</strong> the<br />
emitted infrared light from the SOI LED. From these images, we observe the EL distribution<br />
and the direction <strong>of</strong> light travel in these SOI LEDs.<br />
Figure 4.6.2(a) is the illuminated top view <strong>of</strong> a thick SOI LED, LED016-1-B2-D1, which has a<br />
small 300 µm by 24 µm SOI silicon island area with about 1.3 µm thick active silicon layer. The<br />
device is formed by the middle rectangular silicon island (dark) and the two aluminium contact<br />
pads (bright) apparent in Figure 4.6.2(a). The gate aluminium line was deliberately removed<br />
for the EL observation. Figure 4.6.2(b) shows the infrared image <strong>of</strong> the EL distribution <strong>of</strong> the<br />
same device. As seen in Figure 4.6.2(b), the light is totally blocked in areas which are covered<br />
by aluminium fi lm, seen in Figure 4.6.2(a) as the white areas.<br />
(a)<br />
(b)<br />
Figure 4.6.2: (a) the illuminated top view <strong>of</strong> a SOI LED and (b) the infrared images <strong>of</strong> the EL<br />
distribution <strong>of</strong> the same LED at 190 mA bias current.<br />
98<br />
Due to the resistance in the narrow aluminium lines towards both ends <strong>of</strong> the silicon island,<br />
the current density was reduced in these areas. This resulted in a uniform bright centre<br />
luminescence and gradually darker areas toward both ends as seen in Figure 4.6.2(b).<br />
However, it is also observed that one end <strong>of</strong> LED is brighter than the rest <strong>of</strong> the areas in<br />
this fi gure. The other end <strong>of</strong> the silicon island is shaded by an aluminium pad, and hence is<br />
completely dark in Figure 4.6.2(b). It is clear that some light travels along the device and is<br />
emitted out at the end <strong>of</strong> the device. The light travelling towards the end but not coming out <strong>of</strong><br />
the top surface <strong>of</strong> the silicon island, it should be mostly travelling in a direction parallel to the<br />
surface. However, some <strong>of</strong> the light parallel to the surface may be scattered up towards the<br />
observer at the end <strong>of</strong> the silicon island. It is also seen that the end light spot is stronger than<br />
the centre part. This indicates that the light parallel to the surface has a higher concentration<br />
than the vertically emitted light, since light is accumulated along the silicon island. The device<br />
seems to be a good emitter for light parallel to the surface.
Some <strong>of</strong> these SOI LEDs were designed into larger 1 mm2 devices for easier detection by<br />
the infrared sensor. Figure 4.6.3 shows the EL distribution from one <strong>of</strong> such large device,<br />
LED016-1-E1-D4. Unfortunately, this design had a little too narrow line width and line spacing.<br />
This resulted in very narrow and uneven metal lines in the fabricated devices, which increased<br />
resistance <strong>of</strong> the metal lines. Hence, the EL distribution is rather non-uniform as apparent in<br />
Figure 4.6.3. Some <strong>of</strong> the metal lines were even broken, and resulted in the completely dark<br />
spots in the centre <strong>of</strong> the device.<br />
Figure 4.6.3: EL distribution <strong>of</strong> a 1 mm2 large-area LED at<br />
100 mA bias current.<br />
The EQE (external quantum<br />
effi ciency) <strong>of</strong> the SOI LEDs with<br />
the thick silicon layer were<br />
measured with a Ge photodiode.<br />
The Ge sensor has 0.25 cm<br />
diameter and was placed 1.5<br />
cm away from the SOI LED. The<br />
highest EQE in fi ve measured<br />
devices was about 1.9x10-5 for the small LEDs. The 1 mm2 large LEDs were designed for<br />
higher and more accurate EQE measurement. Unfortunately, the above-mentioned processing<br />
non-uniformity problem caused lower average EL effi ciency.<br />
Figure 4.6.4:<br />
Measured EQE<br />
<strong>of</strong> SOI LEDs.<br />
8.0E-05<br />
LED016-1-E1-D4<br />
Current(A)<br />
7.0E-05<br />
6.0E-05<br />
5.0E-05<br />
4.0E-05<br />
3.0E-05<br />
2.0E-05<br />
1.0E-05<br />
0.0E+00<br />
0.00<br />
0.10 0.20 0.30 0.40 0.50 0.60 0.70<br />
Vo(V)<br />
Figure 4.6.5: Illuminated I-V <strong>of</strong> large<br />
LED under one sun.<br />
4.6.2 Illuminated I-V Characteristics <strong>of</strong> SOI<br />
LEDs<br />
The illuminated I-V characteristics <strong>of</strong> some<br />
1 mm2 large SOI LEDs were tested on the<br />
Centre’s I-V measurement system, Dark-Star,<br />
under standard one sun illumination conditions.<br />
A typical measured I-V curve from a large SOI<br />
LED is shown in Figure 4.6.5.<br />
Our standard high effi ciency PERL solar cells<br />
have very low metal contact areas <strong>of</strong> only 0.25% and low metal grid<br />
shading areas <strong>of</strong> 2% <strong>of</strong> the total cell area. However, the large SOI<br />
LEDs have 20% metal contact areas and 50% metal shading loss.<br />
Hence, it was not a surprise that the SOI LED had relatively low Voc<br />
<strong>of</strong> 582 mV due to the signifi cantly increased contact recombination<br />
and shading losses. However, the SOI LED had a rather high shortcircuit<br />
current density <strong>of</strong> 7 mA/cm 2 , for a 1.3 µm thin and fl at silicon<br />
layer without any light trapping and antirefl ection coating.<br />
99
4.6.3 Design <strong>of</strong> SOI Silicon Waveguide<br />
One <strong>of</strong> the goals at the LED group is to demonstrate light propagation within an all-silicon<br />
circuit by using a SOI silicon waveguide. The design <strong>of</strong> such a SOI waveguide is shown in Figure<br />
4.6.6. The purple T shaped structure in Figure 4.6.6 is the silicon waveguide with 6 µm silicon<br />
waveguide width. At each end <strong>of</strong> the T-shaped waveguide, there is a SOI LED diode, with the<br />
same structure as the ones shown in Figures 4.6.1 and 4.6.2.<br />
1<br />
2<br />
3<br />
4<br />
5<br />
6<br />
7<br />
8<br />
9<br />
Silicon Island<br />
Boron Diffision<br />
Phosphorus Diffusion<br />
Silicon Thinning<br />
Contact Holes<br />
Metalization<br />
Metalization<br />
Silicon formation<br />
Phosphorus isolation lines<br />
Figure 4.6.6: SOI silicon waveguide<br />
design. The purple line represents<br />
the silicon layer. One <strong>of</strong> the LEDs is<br />
used to emit light into the T-shaped<br />
silicon waveguide, while the other 2<br />
diodes are used to detect the light<br />
received. The dimensions are in<br />
microns.<br />
4010<br />
As shown in Figure 4.6.6, one LED is used to emit light into the silicon waveguide, while the<br />
other 2 diodes are used to detect the light transmitted through the silicon waveguide. Other<br />
controlling structures are also incorporated along the SOI waveguide, such as control gates<br />
and pn-junction isolation rings. These devices are designed to cut-<strong>of</strong>f electrical signals to<br />
make sure that any signal detected is optical.<br />
The fi rst silicon waveguide devices were fabricated in 2005. Figure 4.6.7 shows an image<br />
<strong>of</strong> the T-shaped silicon waveguide and the LED or the detector at the end <strong>of</strong> the waveguide.<br />
Each <strong>of</strong> these LED and detector devices had good electrical performance if independently<br />
measured. However, when the LED is forward biased and emitting light, no signal could<br />
be detected by the reverse-biased photo detector at the end <strong>of</strong> the waveguide. Given the<br />
relatively low effi ciency <strong>of</strong> light emission and <strong>of</strong> light collection as well as the high absorption in<br />
the silicon waveguide for the emitted infrared light, it is believed that the signal arriving at the<br />
detector may be too small, in the pA range, to be measured by our equipment at present.<br />
100
(a)<br />
(b)<br />
Figure 4.6.7: (a) Top view <strong>of</strong> the T-shaped intersection <strong>of</strong> silicon waveguide; (b) top view <strong>of</strong> a<br />
LED (or a detector) at one end <strong>of</strong> the waveguide.<br />
Again the infrared CCD camera was used to take photos <strong>of</strong> the light pattern in this waveguide<br />
system. Figure 4.6.8 shows some <strong>of</strong> these photos. Here, the LED in the waveguide unit was<br />
also forward biased. Similarly to the image in Figure 4.6.2(b), Figure 4.6.8(a) shows infrared<br />
light emission from the left-side LED. However, no light was observed within the waveguide<br />
in this photo.<br />
(a)<br />
(b)<br />
Figure 4.6.8: The emitted infrared light images (a) for the LED at 100 mA forward bias; (b)<br />
for the entire silicon waveguide system with prolonged signal accumulation, and with a low<br />
external illumination to show the aluminium patterns.<br />
In Figure 4.6.8(b), a very long accumulation period for the image was selected, which gave<br />
a very brightly glowing LED emitter on the left. A low bias illumination was also added in to<br />
show the aluminium patterns, which indicate the location <strong>of</strong> the LED and detectors. In Figure<br />
4.6.8(b), the light from LED travelling inside the silicon waveguide could not be observed.<br />
However, a bright spot at the T-intersection <strong>of</strong> the waveguide is observed. This indicates that<br />
the light is indeed travelling along the silicon waveguide.<br />
In future research, we plan to improve the light emission and detection effi ciencies for these<br />
LED and detectors. These may be achieved using photonic crystals or plasmon structures.<br />
The light absorption in the waveguide should also be reduced. This may be achieved by using<br />
SiNx as the waveguide material rather than absorptive silicon.<br />
101
4.6.4 Photonic Crystals in SOI LEDs<br />
Photonic crystals are geometric structures in which the refractive index varies periodically<br />
on a length scale that is comparable to the wavelength <strong>of</strong> light, i.e. on the order <strong>of</strong> several<br />
hundreds <strong>of</strong> nanometers. One dimensional (1D), two-dimensional (2D) and three dimensional<br />
(3D) photonic crystals are distinct in the number <strong>of</strong> directions the refractive index varies<br />
periodically. An example <strong>of</strong> a very simple 1D photonic crystal is a Bragg mirror, <strong>of</strong>ten used<br />
in laser cavities, which is a one dimensional periodic structure <strong>of</strong> laterally extended layers <strong>of</strong><br />
different refractive indices.<br />
The photonic density <strong>of</strong> states in a vacuum is continuous, similar to the electronic density <strong>of</strong><br />
states in a vacuum. In a photonic crystal structure, the photonic density <strong>of</strong> states is modifi ed,<br />
resulting in allowed and forbidden energy bands, respectively, similar to the well known<br />
electronic bands in a semiconductor. In this sense photonic crystals can be considered the<br />
optical analogues to semiconductors. The modifi ed photonic density <strong>of</strong> states has a dramatic<br />
infl uence on the optical properties <strong>of</strong> the photonic crystal, i.e. on its absorptance, refl ectivity<br />
and emissivity. In the SOI LED project we aim to utilise the properties <strong>of</strong> photonic crystal to<br />
enhance the light emission effi ciency from SOI light emitting diodes.<br />
Due to the thickness <strong>of</strong> these devices <strong>of</strong> typically only a few hundred nanometers, 2D photonic<br />
crystals are the method <strong>of</strong> choice. The idea is to texture the active area <strong>of</strong> SOI LEDs with a<br />
regular array <strong>of</strong> holes in order to enhance the photonic density <strong>of</strong> states in the wavelength<br />
range around 1150nm, (where spontaneous emission from silicon is most effi cient), which<br />
should enhance the spontaneous emission probability. Combined with scattering effects, this<br />
should lead to a dramatically enhanced external luminescence quantum effi ciency <strong>of</strong> these<br />
devices. Figure 4.6.4.1 shows a theoretically calculated photonic band structure for such an<br />
array <strong>of</strong> air holes in silicon. Whether the enhanced photonic density <strong>of</strong> states falls within the<br />
emission band <strong>of</strong> silicon thus critically depends on the size <strong>of</strong> the holes and the periodicity<br />
<strong>of</strong> the photonic crystal lattice. Experimentally a seventy fold enhancement <strong>of</strong> the emission<br />
from 2D photonic crystal structured SOI layers has been demonstrated at low temperatures<br />
(T=80K) by Zelsmann et al. [Zelsmann, M., et al., Appl.Phys.Lett. 83, 2542 (2003)]. We aim<br />
to demonstrate a similar improvement <strong>of</strong> our light emitting diodes at room temperature.<br />
Fig. 4.6.4.1: Theoretical photonic band structure<br />
for a two dimensional array <strong>of</strong> air holes in silicon<br />
with a lattice parameter a [after Zelsmann et<br />
al.].<br />
102
The photonic crystal processing <strong>of</strong> SOI LEDs is carried out in collaboration with the Centre<br />
<strong>of</strong> Excellence for Quantum Computing, UNSW, where a fi rst photonic crystal array was<br />
successfully produced on the surface <strong>of</strong> a bulk silicon wafer. In that process a PMMA photo<br />
resist is deposited on the surface <strong>of</strong> an oxidised polished silicon wafer. Electron Beam<br />
Lithography (EBL) was used to write the desired holes pattern into the PMMA resist and the<br />
pattern was then transferred successively into the top SiO2 layer end then into the silicon<br />
using Reactive Ion Etching (RIE). The etched holes were more than 200nm deep. An AFM<br />
image and an SEM image <strong>of</strong> this test structure are shown in Fig.4.6.4.2 and Fig.4.6.4.3,<br />
respectively.<br />
Fig. 4.6.4.2: AFM image <strong>of</strong> a photonic<br />
crystal test structure on the surface <strong>of</strong><br />
a polished bulk silicon wafer.<br />
Fig. 4.6.4.3: SEM image <strong>of</strong> the photonic<br />
crystal test structure on the surface <strong>of</strong><br />
a polished bulk silicon wafer.<br />
For initial photoluminescence measurements on these test structures, larger patterned<br />
areas <strong>of</strong> deeper holes are ideally required. Attempts to create larger areas highlighted some<br />
practical experimental problems with the EBL patterning process as a result <strong>of</strong> the so called<br />
proximity effect. The latter is a result <strong>of</strong> the increased effective dose that high density features<br />
receive due to back-scattering <strong>of</strong> the incident beam. The negative impact <strong>of</strong> the proximity<br />
effect on large area test structures is shown in Fig 4.6.4.4: the circles on the edge <strong>of</strong> the<br />
patterned area are clearly distinct features but due to an increased dose the features have<br />
joined together in the centre <strong>of</strong> the pattern. The proximity effect may be alleviated by using<br />
lower accelerating voltages during the EBL. Figure 4.6.4.5 shows an example <strong>of</strong> a larger<br />
patterned area written with a lower (5kV) accelerating voltage.<br />
103
Additional experimental diffi culties were encountered in the attempt to increase the depth<br />
<strong>of</strong> the holes to about 1µm, which is the thickness <strong>of</strong> the thickest and currently most effi cient<br />
SOI LEDs. Different attempts are currently being made to achieve structures with a higher<br />
aspect ration. In this context a number <strong>of</strong> different RIE recipes have been investigated and<br />
alternatively the possibility to use a chromium mask in order to transfer the EBL pattern<br />
into to the silicon layer has been investigated with some success. First photoluminescence<br />
studies on SOI layers textured with photonic crystals are planned for the near future.<br />
A further interesting aspect <strong>of</strong> the photonic crystal work within the SOI group is the recent<br />
observation by S.G. Cloutier et al. [S.G. Cloutier et al.,Nature Materials 4, 887 (2005)] that<br />
optical gain associated with a so called A-centre could be observed at a wavelength <strong>of</strong><br />
1278nm at lower temperatures from silicon on insulator samples patterned with photonic<br />
crystal structures. Given the signifi cance that a successful demonstration <strong>of</strong> optical gain in<br />
an electrically pumped silicon device would have, we will carefully investigate whether this<br />
emission is also observed from photonic crystal structured SOI LEDs.<br />
Fig. 4.6.4.4: Negative impact <strong>of</strong> the proximity<br />
effect on larger area photonic crystal<br />
structures.<br />
Fig. 4.6.4.5: Improved periodicity<br />
and reduced proximity effects due to<br />
reducing the acceleration voltage during<br />
the EBL patterning <strong>of</strong> the PMMA.<br />
104
4.6.5 Improved SOI LEDs Using Surface Plasmons.<br />
A very effective way to enhance the external luminescence quantum effi ciency from silicon light<br />
emitting diodes is the application <strong>of</strong> light trapping schemes. Light trapping is conventionally<br />
used in photovoltaic devices to increase the average optical path for incident light, resulting<br />
in better absorption <strong>of</strong> weakly absorbed light. According to Kirchh<strong>of</strong>f’s law, an enhanced<br />
absorption corresponds to an enhanced emission. Light that is trapped within a planar device<br />
due to total internal refl ection can be coupled out after multiple paths through a textured<br />
device. The benefi cial effect <strong>of</strong> light trapping schemes on LED performance has been<br />
demonstrated in high effi ciency bulk silicon light emitting diodes [M.A. Green et al., Nature<br />
412, 805 (2001)].<br />
However, for thin silicon on insulator devices, traditional light trapping methods such as<br />
texturing with pyramids, cannot be used. The excitation <strong>of</strong> surface plasmons is a promising<br />
alternative way <strong>of</strong> increasing light absorption and emission from thin-Si based devices. An<br />
eighteen fold absorption enhancement at 800nm wavelength <strong>of</strong> silicon on insulator (SOI)<br />
devices was reported by Stuart and Hall [H.R. Stuart and D.G. Hall, Appl.Phys.Lett. 73, 3815<br />
(1998)]. Our work aims at increasing the absorption / emission form thin fi lm Si devices at<br />
longer wavelengths.<br />
Surface Plasmon effects are observed on thin silicon layers covered with an array <strong>of</strong><br />
nanometer sized metal islands deposited on the surface and separated from the active silicon<br />
layer by a very thin ( ~ 30nm) spacer layer (SiO2). A simple method <strong>of</strong> silver deposition followed<br />
by annealing is used to deposit Ag nanoparticles on silicon-on-insulator (SOI) light emitting<br />
devices. The fact that the entire island deposition is a low temperature process allows metal<br />
islands to be deposited onto fi nished fully processed devices. Fig.4.6.5.1 shows a typical SEM<br />
image <strong>of</strong> the nanometer sized silver islands.<br />
Figure 4.6.5.1: SEM photo <strong>of</strong> the deposited<br />
nanometre silver islands for the surface plasmon<br />
effect.<br />
Changes in the optical device performance were<br />
monitored experimentally by measuring the<br />
spectral electroluminescence (EL) emission and<br />
the spectral response (SR) <strong>of</strong> SOI LEDs before and<br />
after island deposition. The optical enhancement<br />
is calculated as the ratio <strong>of</strong> the SR and EL<br />
signals, respectively after and before deposition<br />
<strong>of</strong> the islands. This combination <strong>of</strong> EL and SR is<br />
ideal because it provides information about the<br />
optical path lengths enhancement in different<br />
spectral ranges above (SR) and near and below<br />
(EL) the band-gap <strong>of</strong> silicon. As reported earlier,<br />
our initial pro<strong>of</strong> <strong>of</strong> concept studies showed only<br />
a 30% enhancement in the electroluminescence<br />
measurements from our Silicon-on-insulator<br />
devices.<br />
105
With further optimisation <strong>of</strong> the experimental conditions, specifi cally a much thinner isolation<br />
layer between the islands and the silicon layer, a 7 fold enhancement in the electroluminescence<br />
and spectral response measurement has been demonstrated. Fig.4.6.5.2 shows the spectral<br />
emission (EL) from the same SOI LED before and after island deposition.<br />
10<br />
9<br />
8<br />
with Ag islands<br />
without Ag islands<br />
EL Signal (a.u)<br />
7<br />
6<br />
5<br />
4<br />
3<br />
2<br />
1<br />
0<br />
900 950 1000 1050 1100 1150 1200 1250 1300<br />
Wavelength (nm)<br />
Figure 4.6.5.2: Electroluminescence<br />
spectra from the same 95nm thin<br />
SOI LED measured under identical<br />
experimental conditions before (red)<br />
and after (blue) deposition <strong>of</strong> the silver<br />
islands.<br />
The resulting EL enhancement is shown together with the SR enhancement <strong>of</strong> the same<br />
device in Fig.4.6.5.3. The enhanced emission intensity due to surface plasmon effects is also<br />
very clearly visible in an electroluminescence image taken with a very sensitive silicon CCD<br />
camera (see Fig. 4.6.5.4) from a device which was partially covered with silver islands on<br />
the left hand side. The image clearly shows the much brighter emission from that half <strong>of</strong> the<br />
device.<br />
Figure 4.6.5.3: Enhancement <strong>of</strong> the<br />
spectral electroluminescence intensity<br />
(red curve), calculated as the ratio <strong>of</strong><br />
the two spectra shown in Fig.4.6.5.2.<br />
The enhancement <strong>of</strong> the photocurrent<br />
response is shown for comparison (black<br />
curve).<br />
EL / Photocurrent Enhancement<br />
8<br />
7<br />
6<br />
5<br />
4<br />
3<br />
2<br />
1<br />
EL<br />
SR<br />
0<br />
300 400 500 600 700 800 900 1000 1100 1200 1300<br />
Wavelength(nm)<br />
Figure. 4.6.5.4: Electroluminescence image <strong>of</strong> a ~ 95nm<br />
thick 1mm x 1mm SOI LED under forward bias. The left<br />
hand side <strong>of</strong> the device was coated with silver islands.<br />
Clear evidence for enhanced emission from the coated<br />
area is observed.<br />
106
Another interesting aspect <strong>of</strong> the Plasmon work is the shifting <strong>of</strong> the enhancement peak by<br />
using a layer <strong>of</strong> high refractive index material (ZnS) overcoating on the metal nanoparticles,<br />
as shown in Fig.4.6.5.5. Our results show a 70nm shift in the enhancement peak at longer<br />
wavelengths as calculated from the results <strong>of</strong> the EL measurements. The enhancement in EL<br />
could have signifi cant benefi ts in increasing the effi ciency <strong>of</strong> Si LEDs and the additional shift<br />
observed would lend more tunability to the operating wavelength <strong>of</strong> the LEDs. Absorption in<br />
metals is a source <strong>of</strong> concern for these applications, but the results presented here already<br />
account for the absorption in metals and hence look encouraging. One <strong>of</strong> the major advantages<br />
<strong>of</strong> this approach is that the Ag islands can be deposited at the fi nal stage <strong>of</strong> device processing<br />
and hence no processing incompatibilities are involved.<br />
Enhancement<br />
8<br />
7<br />
6<br />
5<br />
4<br />
3<br />
2<br />
1<br />
with Ag<br />
with Ag+ZnS<br />
0<br />
900 1000 1100 1200 1300<br />
Wavelength (nm)<br />
Figure 4.6.5.5: EL enhancement plot<br />
showing the shift in the enhancement<br />
peak due to application <strong>of</strong> a ZnS<br />
overcoating.<br />
107
108
4.6.6 Photoluminescence-Based Characterisation <strong>of</strong> Silicon<br />
Researchers:<br />
Robert Bardos, Thorsten Trupke<br />
Collaborators (internal):<br />
Malcolm Abbott, Florence Chen, Jeff Cotter, Aihua Wang, Jianhua Zhao<br />
Collaborators (external):<br />
Daniel MacDonald, Andres Cuevas, ANU, Canberra<br />
Adam Lorenz, Evergreen Solar<br />
Thomas Roth, Martin Schubert, Stefan Glunz, Wilhelm Warta, ISE, Freiburg Germany<br />
Background<br />
The determination <strong>of</strong> the injection level dependent effective excess carrier lifetime τ eff (∆n)<br />
by photoconductance (PC) measurements is an important experimental technique that is<br />
used at our Centre and in most photovoltaics research centres around the world for the<br />
characterization <strong>of</strong> silicon samples. Recently, it has been established theoretically and<br />
experimentally that the PC technique can be subject to signifi cant experimental artefacts<br />
especially at low excess carrier concentrations. Such artefacts can result from excess<br />
charges accumulated in space charge regions via the so called depletion region modulation<br />
(DRM) effect, by temperature variations upon intense illumination and by minority carrier<br />
trapping. These effects can be so dominant, that PC measurements have been claimed to<br />
be unsuitable for the characterization <strong>of</strong> the effective lifetime e.g. in thin fi lm silicon samples.<br />
In silicon wafers these effects generally make the accurate determination <strong>of</strong> the effective<br />
lifetime at low to moderate injection levels impossible.<br />
In our photoluminescence (PL) lifetime project we aim to demonstrate that quasi steady state<br />
photoluminescence (QSS-PL) lifetime measurements are advantageous in many regards and<br />
therefore represent a viable alternative to more established techniques such as QSS-PC.<br />
QSS-PL measurements are insensitive to the above mentioned artefacts and also much more<br />
sensitive at low excess carrier concentrations. QSS-PL is thus the most suitable technique to<br />
measure the minority carrier lifetime at low excess carrier densities.<br />
In a separate recent development we have demonstrated that PL imaging is an extremely<br />
sensitive and fast technique that provides spatial information about the material quality in<br />
large area silicon wafers. The response from internal and collaborating external research<br />
groups and from industry partners indicates that within a very short timeframe this technique<br />
is set to revolutionize the characterisation <strong>of</strong> silicon samples in the PV industry.<br />
Theory<br />
Photoluminescence is the emission <strong>of</strong> light as a result <strong>of</strong> an external optical excitation. In<br />
a semiconductor, luminescence is emitted when electrons and holes recombine radiatively,<br />
i.e. by transferring the excess energy released in the recombination process to an emitted<br />
photon (Fig.4.6.6.1). Luminescence is thus the reverse process to photogeneration. In a<br />
simplifi ed model the measured PL signal I PL,rel is given by the product <strong>of</strong> electron and hole<br />
concentrations, n e and n h respectively and can be written:<br />
I<br />
2<br />
,<br />
= A ⋅ B ⋅ n ⋅ n = A ⋅ B ⋅ n ⋅exp(<br />
∆η<br />
/ kT)<br />
PL rel<br />
i<br />
e<br />
h<br />
i<br />
i<br />
with B the radiative recombination coeffi cient and Ai a scaling factor, which takes into account<br />
that in most cases PL signals are measured in relative units only. On the right hand side, the<br />
luminescence signal is expressed as a function <strong>of</strong> the separation <strong>of</strong> the quasi Fermi energies<br />
∆η which, in a good solar cell, is directly linked to the cell voltage V by eV=∆η. This direct<br />
relation between the PL signal and the voltage <strong>of</strong> a solar cell allows PL measurements to be<br />
interpreted as contact-less voltage (also referred to as implied voltage) measurements (see<br />
also section on Suns-PL further below).<br />
109
Figure 4.6.6.1: Luminescence<br />
is emitted by semiconductors if<br />
electrons from the conduction<br />
band and holes from the valence<br />
band recombine radiatively, i.e. by<br />
giving up the excess energy to an<br />
emitted photon (γ).<br />
Experimental<br />
A very sensitive QSS-PL lifetime system was established in the laser controlled area LG24.<br />
A large area luminescence sensor was incorporated into a commercial Sinton Consulting<br />
PC bridge next to the PC coil, which allows QSS-PL and QSS-PC measurements to be carried<br />
out on the same wafer and under identical experimental conditions and thereby a direct<br />
quantitative comparison <strong>of</strong> these techniques. A high power LED array, that replaces the<br />
fl ash lamp commonly used in PC testers, is used to illuminate the samples. A photo <strong>of</strong> that<br />
set-up is shown in Fig.4.6.6.2. Upgrade to a 16-bit data acquisition system allows a large<br />
dynamic range <strong>of</strong> light intensities (and thus injection levels) to be scanned with one single<br />
shot. Convenient data acquisition and analysis are accomplished by custom built LabView<br />
s<strong>of</strong>tware. The raw experimental data from a typical quasi steady state photoluminescence<br />
experiment are shown in Fig. 4.6.6.3. The graph highlights two <strong>of</strong> the main advantages <strong>of</strong><br />
using a solid state light source (i.e. an LED array). Firstly, waveforms can be arbitrarily long,<br />
up to tens <strong>of</strong> seconds, thereby allowing true quasi steady state conditions to be reached even<br />
in high quality material with long carrier lifetimes. Secondly, arbitrary temporal light intensity<br />
pr<strong>of</strong>i les can be generated (in this case a triangular waveform), which is particularly important<br />
for the self consistent methods to be described further below. Another advantage <strong>of</strong> using<br />
a computer controlled solid state light source is that it allows automated repetitive signal<br />
averaging, which results in a better signal to noise ratio.<br />
Figure 4.6.6.2: Experimental set up for<br />
combined PL and PC measurements. The<br />
PL sensor is incorporated into a commercial<br />
photoconductance set up which allows<br />
convenient calibration <strong>of</strong> relative PL signals<br />
against absolute PC signals.<br />
Figure 4.6.6.3: Typical data set from a PL<br />
lifetime experiment. The PL signal (red) and the<br />
incident light intensity (black) are measured<br />
simultaneously.<br />
110
Surprisingly high sensitivity <strong>of</strong> PL lifetime measurements<br />
Using PL as a sensitive technique to determine small excess carrier concentrations in silicon<br />
may at fi rst sight appear diffi cult given silicon’s indirect band structure and the resulting low<br />
radiative recombination coeffi cient. However, the automated signal averaging, the geometry<br />
<strong>of</strong> our PL setup and the use <strong>of</strong> a home built, low noise, high gain preamplifi er allowed effective<br />
lifetimes to be determined down to injection levels < 10 9 cm -3 in 1 Ωcm p-type wafers (see<br />
Fig. 4.6.6.5) with data acquisition times <strong>of</strong> only a few seconds [4.6.1], which is orders <strong>of</strong><br />
magnitude below the detection limit <strong>of</strong> a QSS-PC set-up used at our Centre. Using that QSS-<br />
PC set-up, lifetime measurements become extremely noisy and inaccurate at injection levels<br />
< 10 13 cm -3 . Due to this somewhat unexpected sensitivity, PL lifetime measurements can<br />
also be used to determine very low effective lifetimes in quasi steady state mode and are by<br />
no means restricted to high quality fl oat zone silicon.<br />
The development <strong>of</strong> a computer controlled solid state light source for lifetime measurements<br />
has also improved the sensitivity and accuracy <strong>of</strong> PC measurements, because the same<br />
averaging techniques as used for PL can also be applied to PC measurements. While these<br />
improvements render PC measurements at low injection levels less noisy they cannot avoid<br />
some more fundamental limitations <strong>of</strong> PC measurements to be discussed now.<br />
Depletion region modulation effect<br />
In QSS-PC measurements at low to moderate injection conditions, the so-called depletion<br />
region modulation (DRM) effect leads to a sharp artifi cial increase <strong>of</strong> the apparent minority<br />
carrier lifetime on samples with a diffused or induced junction. Similar effects are expected<br />
in any other experimental technique in which the measured signal is determined by the sum<br />
<strong>of</strong> the minority and majority carrier concentrations. In contrast, theoretical analysis shows<br />
that QSS-PL measurements should be largely unaffected by the DRM effect [4.6.2]. This was<br />
proven experimentally by investigation <strong>of</strong> two 1Ωcm p-type wafers, one with Phosphorous<br />
diffusions the other one with Boron diffusions on both sides [4.6.1]. Theoretically the artefacts<br />
due to the DRM effect are expected only for the Phosphorus-diffused sample. Fig.4.6.6.4<br />
shows the lifetimes <strong>of</strong> the Boron diffused sample determined from PC and from PL. Very good<br />
agreement is observed between PL and PC with no indication in both measurements <strong>of</strong> an<br />
infl uence <strong>of</strong> the DRM effect. In contrast, the PC lifetime from the Phosphorus doped sample<br />
shown in Fig.4.6.6.5 increases steeply at injection levels < 5x10 12 cm -3 (up to an apparent<br />
lifetime <strong>of</strong> 13ms, not shown in Fig.4.6.6.5). This increase is consistent with the theoretically<br />
predicted effects resulting from the DRM effect. As theoretically predicted, the PL lifetime is<br />
unaffected by these effects.<br />
Figure 4.6.6.4: Effective lifetime from<br />
QSS-PL (red) and from QSS-PC (black)<br />
for a 1Ωcm p-type wafer with Boron<br />
diffusions on both sides. Very good<br />
agreement is observed between the<br />
lifetimes from PL and from PC over a<br />
wide injection level range.<br />
111
Figure 4.6.6.5: Effective lifetime<br />
from PL (red and pink) and from PC<br />
(black) for a 1 Ωcm p-type wafer with<br />
phosphorus diffusions on both sides.<br />
The PC data are heavily infl uenced by<br />
the DRM effect at injection levels <<br />
10 13 cm -3 . Very small injection levels<br />
Fig.4.6.6.6: Analysis <strong>of</strong><br />
experimental (coloured curves)<br />
and corresponding simulated<br />
(black lines) QSS-PL lifetime data<br />
for different calibrations <strong>of</strong> the<br />
relative PL signal into absolute<br />
values for ∆n(t). Only the correct<br />
calibration (red curve) gives<br />
consistent results for the rising<br />
and the falling branches <strong>of</strong> the<br />
waveform, thereby allowing a very<br />
accurate calibration <strong>of</strong> ∆n(t).<br />
Variation <strong>of</strong> a calibration constant in the analysis <strong>of</strong> experimental data until hysteresis<br />
effects vanish in τ eff (∆n) plots are thus a very convenient way to calibrate QSS-PL lifetime<br />
measurements. The self consistent calibration approach is particularly useful for QSS-PL<br />
measurements, where the calibration <strong>of</strong> the generation rate G(t) is fairly straightforward<br />
because a spectrally narrow light source is used. The self consistent calibration <strong>of</strong> relative<br />
QSS-PL data into absolute ∆n(t) now allows QSS-PL measurements to be carried out<br />
independently from other experimental techniques. The previous requirement <strong>of</strong> a separate<br />
QSS-PC calibration measurement is thereby avoided. Our work on self consistent calibration<br />
methods has also improved the accuracy <strong>of</strong> PC lifetime measurements.<br />
Suns-PL a contactless alternative to Suns-Voc<br />
Because the luminescence signal is directly linked to the open circuit voltage via the separation<br />
<strong>of</strong> the quasi Fermi energies, PL measurements can be interpreted as a contactless voltage<br />
measurement. We have shown that based on this idea simultaneous measurements <strong>of</strong> the<br />
incident light intensity and <strong>of</strong> the PL signal yield the equivalent <strong>of</strong> Suns-Voc measurements,<br />
with the advantage that these so called Suns-PL measurements can be performed on wafers<br />
at any stage <strong>of</strong> the processing sequence and are not restricted to solar cell structures. To<br />
demonstrate the validity <strong>of</strong> this approach, we investigated a shunted bifacial double sided<br />
buried contact silicon solar cell and simultaneously measured the open circuit voltage, the<br />
PL and the varying incident light intensity [4.6.6]. Fig.4.6.6.7 shows the resulting Suns-Voc<br />
and Suns-PL curves, i.e. the incident light intensity plotted as a function <strong>of</strong> the open circuit<br />
voltage (Suns-Voc) <strong>of</strong> the cell and as a function <strong>of</strong> the separation <strong>of</strong> the quasi Fermi energies<br />
(the latter determined from PL). The two prominent features from the Suns-Voc curve, i.e.<br />
the shunt at low voltages and a kink in the curve at larger voltages are both reproduced in<br />
the Suns-PL measurement. Very good overall agreement between the curves is observed,<br />
which demonstrates the feasibility <strong>of</strong> the method. In a recent project the Suns-PL technique<br />
was applied extensively during the development and the testing <strong>of</strong> an edge isolation method<br />
for small area solar cells [4.6.7] highlighting the strong infl uence <strong>of</strong> edge recombination<br />
especially at low voltages.<br />
113
Fig 4.6.6.7: Comparison <strong>of</strong> a<br />
Suns-Voc measurement with<br />
a Suns-PL measurement on<br />
a bifacial silicon solar cell. The<br />
good agreement between<br />
the curves demonstrates the<br />
feasibility <strong>of</strong> using the Suns-<br />
PL technique to obtain implied<br />
IV curves on unfi nished solar<br />
cells.<br />
PL imaging<br />
QSS-PL or QSS-PC lifetime<br />
measurements only provide<br />
information about the injection<br />
level dependence and/or about<br />
the temperature dependence<br />
<strong>of</strong> the minority carrier lifetime in a silicon wafer with no information about spatial variations<br />
within the wafer. Photoluminescence imaging is a recent and very exciting development at<br />
UNSW that overcomes this limitation. The idea here is to take a laterally resolved “photograph”<br />
<strong>of</strong> the luminescent emission from silicon wafers under homogeneous intense illumination<br />
[4.6.8, 4.6.9] in order to obtain an image <strong>of</strong> local variations in the material quality. The local<br />
luminescence intensity refl ects the local carrier concentration and is directly correlated with<br />
the local minority carrier lifetime. Importantly , the above mentioned robustness <strong>of</strong> QSS-<br />
PL measurements against artifacts and the high sensitivity <strong>of</strong> PL, which where previously<br />
demonstrated only for spatially averaged QSS-PL lifetime measurements, apply to PL imaging<br />
in the same way. The worldwide fi rst PL imaging set up suitable to measure large area silicon<br />
wafers was established and successfully demonstrated at UNSW during the second half <strong>of</strong><br />
2005. Fig.4.6.6.8 shows a PL image taken on a 8.5x8.5 cm 2 multicrystalline silicon wafer,<br />
measured with a data acquisition time <strong>of</strong> only one second. In Fig.4.6.6.8, the local PL intensity<br />
was calibrated into an effective minority carrier lifetime. Comparison with a lifetime image<br />
from a Carrier Density Image measured at the Fraunh<strong>of</strong>er Institute in Freiburg showed very<br />
good agreement between the two independently calibrated measurements. Such quantitative<br />
PL imaging will allow local IV curves to be calculated in analogy to Suns-PL measurements<br />
and it also allows local defect rich areas to be identifi ed. The data also shows that typical<br />
photovoltaic material with lifetime variations on the order <strong>of</strong> a few tens to hundreds <strong>of</strong><br />
microseconds can be imaged in only one second.<br />
Fig 4.6.6.8: Example <strong>of</strong> a PL image <strong>of</strong><br />
a 302µm thick polished 1.2Ωcm SiN<br />
passivated p-type multi-crystalline<br />
silicon wafer. The colour bar gives the<br />
local effective minority carrier lifetime<br />
in µs. This image was taken with a data<br />
acquisition time <strong>of</strong> only one second!<br />
114
In qualitative PL imaging, only the relative variation <strong>of</strong> the PL intensity across the sample is<br />
investigated. Such qualitative imaging was applied extensively by the buried contact group <strong>of</strong> the<br />
Centre and within a short period PL imaging was established as a standard characterisation<br />
technique. In fact the initial PL imaging set-up has been in such heavy demand that setting up<br />
<strong>of</strong> a second PL imaging system is planned for the fi rst half <strong>of</strong> 2006. A further improvement<br />
<strong>of</strong> the data acquisition speed by a factor 50 is anticipated for this new system. Within the<br />
fi rst few months <strong>of</strong> applying PL imaging as a process monitoring tool especially the buried<br />
contact group identifi ed various serious problems with standard processing sequences,<br />
that had previously been used for years [4.6.10, 4.6.11]. The fact that PL imaging is fast,<br />
simple to use and applicable to unfi nished cells turned out to be an invaluable combination<br />
in this context. Some examples <strong>of</strong> process induced defects identifi ed with PL images are<br />
shown in Fig.4.6.6.9. Various collaborative research projects with industry partners (see for<br />
example [4.6.12]) aim at demonstrating the capabilities <strong>of</strong> the system and at increasing the<br />
awareness in the PV community regarding the potential <strong>of</strong> this new technique. A mid-term<br />
goal is the demonstration <strong>of</strong> PL imaging as an industrial in-line process monitoring tool. The<br />
data acquisition speed in PL imaging is several orders <strong>of</strong> magnitude higher than in comparable<br />
imaging or mapping methods, making PL Imaging the only experimental technique that can<br />
provide high resolution images <strong>of</strong> the material quality from large area wafers in a matter <strong>of</strong><br />
seconds.<br />
Figure 4.6.6.9: Examples <strong>of</strong> effective process monitoring using PL imaging. Problems with<br />
standard processing sequences that could be identifi ed include amongst various others (from<br />
left to right): wafer handling using tweezers, furnace contamination; cracks. Each PL image<br />
is taken with a data acquisition time <strong>of</strong> only one second. Some more examples <strong>of</strong> process<br />
induced defects that were identifi ed with PL imaging are shown in Section 4.3.2.<br />
Theoretical work<br />
One complication in the analysis <strong>of</strong> QSS-PL lifetime measurements arises from the fact<br />
that at high carrier densities (>10 16 cm -3 ) the PL intensity no longer follows the product<br />
<strong>of</strong> electron and hole concentrations as given by Eq.1. It is well established and theoretically<br />
understood that the Auger coeffi cients in silicon decrease as a function <strong>of</strong> injection level due<br />
to the reduced Coulomb attraction between electrons and holes at high carrier densities.<br />
The reduced attraction <strong>of</strong> electrons and holes at high carrier densities also affects radiative<br />
recombination, which leads to an injection level dependence <strong>of</strong> the radiative recombination<br />
coeffi cient B(T,∆n). While we had earlier described the temperature dependence <strong>of</strong> B(T) at<br />
low injection levels [4.6.13], a detailed analysis <strong>of</strong> the injection level dependence has only been<br />
carried out very recently by P.P. Altermatt in collaboration with researchers from this Centre<br />
[4.6.14]. The injection level dependence <strong>of</strong> B is now used in our QSS-PL lifetime system to<br />
correctly interpret experimental photoluminescence data.<br />
115
Another problem in QSS-PL lifetime measurements can be caused by reabsorption<br />
<strong>of</strong> spontaneously emitted photons on the way to the surface. It has been argued that, in<br />
cases where the effective lifetime various strongly with injection level, resulting in strong<br />
variations <strong>of</strong> the carrier density pr<strong>of</strong>i le, the reabsorption can cause artefacts in PL lifetime<br />
measurements. A detailed theoretical analysis <strong>of</strong> this problem reveals that even in the worst<br />
case the artefacts that can be expected as a result <strong>of</strong> these effects can generally be expected<br />
to be small [4.6.15]. That analysis also revealed various experimental procedures to avoid<br />
even the small artefacts resulting from reabsorption.<br />
4.6.7 References for Section 4.6<br />
116<br />
4.6.1 Trupke, T. and R.A. Bardos, Photoluminescence: A surprisingly sensitive lifetime<br />
technique. 31st IEEE PVSC, Orlando, USA, 2005.<br />
4.6.2 Trupke, T., R.A. Bardos, F. Hudert, P. Würfel, A. Wang, J. Zhao and M.A. Green,<br />
Effective excess carrier lifetimes exceeding 100ms in fl oat zone silicon determined<br />
from photoluminescence. 19th EPVSC, Paris, France, 2004.<br />
4.6.3 Bardos, R.A., T. Trupke, M.C. Schubert and T. Roth, Trapping artifacts in quasi<br />
steady state photoluminescence and photoconductance lifetime measurements<br />
on silicon wafers. Applied Physics Letters, 2006. 88: p. 053504.<br />
4.6.4 Trupke, T. and R.A. Bardos, Self-consistent determination <strong>of</strong> the generation rate<br />
from photoconductance measurements. Applied Physics Letters, 2004. 85(16): p.<br />
3611-3613.<br />
4.6.5 Trupke, T., R.A. Bardos and M.D. Abbott, Self Consistent Calibration <strong>of</strong><br />
Photoluminescence and Photoconductance Lifetime Measurements.<br />
Applied Physics Letters, 2005. 87: p. 184102.<br />
4.6.6 Trupke, T., R.A. Bardos, M.D. Abbott and J.E. Cotter, Suns-photoluminescence:<br />
Contactless determination <strong>of</strong> the current voltage characteristics <strong>of</strong> silicon wafers.<br />
Applied Physics Letters, 2005. 87: p. 093503.<br />
4.6.7 Abbott, M.D., J.E. Cotter, T. Trupke and R.A. Bardos, Investigation <strong>of</strong> edge<br />
recombination effects in silicon solar cell structures using photoluminescence.<br />
Appl. Phys. Lett. 88, 114105, 2006.<br />
4.6.8 Trupke, T., R.A. Bardos, M.C. Schubert and W. Warta, Photoluminescence imaging<br />
<strong>of</strong> silicon wafers. Applied Physics Letters, submitted, 2006.<br />
4.6.9 T. Trupke, R.A. Bardos, M.D. Abbott, F.W. Chen, J.E. Cotter and A. Lorenz, Fast<br />
photoluminescence imaging <strong>of</strong> silicon wafers. WCPEC-4, Hawaii, USA, May 2006,<br />
accepted for oral presentation, 2006.<br />
4.6.10 Chen, F.W., J.E. Cotter, T. Trupke and R.A. Bardos, Characterisation <strong>of</strong> PECVD<br />
silicon nitride passivation with photoluminescence imaging. WCPEC-4, Hawaii, USA,<br />
May, 2006.<br />
4.6.11 Abbott, M.D., J.E. Cotter, T. Trupke, K. Fischer and R.A. Bardos, Application <strong>of</strong><br />
Photoluminescence to High-effi ciency Silicon Solar Cell Fabrication. WCPEC-4,<br />
Hawaii, USA, May, 2006.<br />
4.6.12 Bardos, R.A., J.E. Cotter, T.Trupke and A. Lorenz, Comparative analysis <strong>of</strong> n-type<br />
versus p-type string ribbon silicon wafers. WCPEC-4, Hawaii, USA, May, 2006.<br />
4.6.13 Trupke, T., M.A. Green, P. Würfel, P.P. Altermatt, A. Wang, J. Zhao and R. Corkish,<br />
Temperature dependence <strong>of</strong> the radiative recombination coeffi cient <strong>of</strong><br />
intrinsic crystalline silicon. Journal <strong>of</strong> Applied Physics, 2003. 94(8): p. 4930-4937.<br />
4.6.14 Altermatt, P.P., F. Geelhaar, T. Trupke, X. Dai, A. Neisser and E. Daub, Injection<br />
dependence <strong>of</strong> spontaneous radiative recombination in crystalline silicon:<br />
experimental verifi cation and theoretical analysis. Appl. Phys. Lett., submitted.,<br />
2006.<br />
4.6.15 Trupke, T., Infl uence <strong>of</strong> photon reabsorption on quasi steady state<br />
photoluminescence measurements on crystalline silicon. Journal <strong>of</strong> Applied<br />
Physics, in press, 2006.
4.7 COLLABORATIVE RESEARCH<br />
<strong>University</strong> Nacional del Litoral, Argentina<br />
Pr<strong>of</strong>. Roman Buitrago from the <strong>University</strong> Nacional del Litoral in Santa Fe, Argentina, visited<br />
the Centre for three months from August to November 2005. He was principally involved in<br />
work on solid-phase crystallisation <strong>of</strong> evaporated amorphous silicon, but also provided useful<br />
contributions to other areas <strong>of</strong> Second Generation Strand activities.<br />
Materials Science and Engineering, UNSW<br />
A very successful collaboration was initiated with Dr Yu Wang, Materials Science and<br />
Engineering, UNSW. Dr Wang contributed signifi cantly to another trip to the Photon Factory<br />
synchrotron facility at Tsukuba, Japan in December 2005. This trip was funded by AINSE<br />
under the Australian Nuclear Beamline Facility project. Dr Wang has also provided facilities<br />
and expertise for X-ray tube X-ray diffraction. The result <strong>of</strong> both these experiments has been<br />
very useful and has provided complementary structural information from X-ray diffraction on<br />
the silicon quantum dot nanostructure work.<br />
Aachen <strong>University</strong>, Germany<br />
A collaboration was established with Dr Andreas Nositschka and Robert Rolver at Aachen<br />
<strong>University</strong> on time resolved photoluminescence. Initial data on PL on Si nanostructure<br />
samples was very promising. This will be built upon to develop time resolved data which<br />
should elucidate decay mechanisms in the material.<br />
Rudjer Boskovic Institute, Zagreb, Croatia<br />
Dr Branco Pivac and Ivana Kovačević from Zagreb carried out Grazing Incidence Small Angle<br />
X-ray Scattering (GISAXS) measurements on Si nanostructure samples at the synchrotron<br />
facility in Trieste. This gave information on nanocrystal size which complemented TEM data<br />
on the same samples and also gave information on nanocrystal shape – confi rming that they<br />
are spherical.<br />
Fraunh<strong>of</strong>er-Institute for Solar Energy Systems, Freiburg, Germany<br />
<strong>University</strong> <strong>of</strong> Bern, Switzerland<br />
Collaboration has continued on the Up/Down conversion project, see Up/Down-conversion in<br />
Third Generation Strand section. H.U. Güdel and K. Krämer from Bern have provided further<br />
Er doped phosphors and helped to interpret results <strong>of</strong> absorption and PL measurments. S.<br />
Glunz from ISE, has again provided some <strong>of</strong> the bifacial cells used in the Up-conversion device<br />
experiments.<br />
School <strong>of</strong> Physics, <strong>University</strong> <strong>of</strong> Sydney<br />
Further collaboration has been established with Dr Nicholas Ekin-Daukes at Sydney <strong>University</strong><br />
targetting carrier cooling measurements using a femtosecond pulsed laser, which is soon<br />
to be established. This is part <strong>of</strong> a successful LIEF proposal which also included pulsed DC<br />
sputtering with Pr<strong>of</strong>. Marcella Bilek at Sydney and pulsed laser deposition with Dr Nagy<br />
Valanoor at UNSW.<br />
117
Magnetic Materials Laboratory, School <strong>of</strong> Physics, UNSW<br />
Dr Jack Cochrane, Physics, UNSW has been instrumental in providing access and in assisting<br />
in measurements for Hall effect and resistivity vs. temperature measurements. These<br />
measurements are important in elucidating electrical transport mechanisms.<br />
Australian Nuclear Science and Technology Organisation, Lucas Heights, NSW<br />
The Centre was yet again successful in an application to AINSE for time on Secondary Ion<br />
Mass Spectrometry at ANSTO in 2005. Again the grant also included time on Rutherford<br />
Backscattering at ANSTO. Characterisation primarily involved Si nanostructures with high<br />
levels <strong>of</strong> depth resolution achieved for SIMS, this resolution exceeded that <strong>of</strong> last year and is<br />
now down to 1.5nm (as independently measured by TEM). RBS was used for calibration <strong>of</strong><br />
these data and in addition RBS channelling was used for further structural characterisation<br />
<strong>of</strong> Si nanostructure samples used in synchrotron experiments.<br />
IRDEP (joint EDF/CNRS/ENSCP lab.), Paris<br />
<strong>University</strong> <strong>of</strong> Karlsruhe<br />
Collaboration with the above was by way <strong>of</strong> prolonged visits by researchers from these<br />
institutions to work on projects in the Third Generation strand.<br />
Dr Jean-Françoise Guillemoles, from the IDREP, Paris, visited the Centre for six months from<br />
March to October, 2005. He was principally involved in work on Hot Carrier cells but also<br />
provided useful contributions to other areas <strong>of</strong> Third Generation Strand activities.<br />
Pr<strong>of</strong>. Peter Würfel, from Karlsruhe <strong>University</strong>, visited the Centre from October 2005<br />
to January 2006 having visited previously in 2000 and 2003. He contributed to several<br />
areas <strong>of</strong> Third Generation work, including establishment <strong>of</strong> an electro-luminescence mapping<br />
teachnique.<br />
Global Climate and Energy Project (GCEP), Stanford Univeristy, California, USA<br />
The Centre was successful in a proposal to the GCEP for work on Si nanostructures to<br />
develop a wide band gap Si nanostructure based tandem cell element over three years. This<br />
builds on the Si quantum dot nanostructure work in the Third Generation strand and will link<br />
to work in the Thin Film strand. The project <strong>of</strong>fi cially started in September 2005 and initial<br />
work involved recruitment <strong>of</strong> three new postdoctoral fellows and two new research fellows to<br />
lead the strands on materials processing and characterisation/modelling.<br />
Toyota Central Research Laboratories, Nagoya, Japan and Toyota Future Project<br />
Division, Higashi-Fuji, Japan<br />
The project on Energy Conversion Devices concentrating on high effi ciency thermoelectrics<br />
and application <strong>of</strong> selective energy contacts started in December 2004. It continued<br />
throughout 2005 with appointment <strong>of</strong> a postdoctoral fellow. Signifi cant progress was made<br />
on thermoelectric and selective energy contact theory and on experimental measurement<br />
<strong>of</strong> selective energy contacts including initiation <strong>of</strong> a new technique <strong>of</strong> optically assisted I-V<br />
measurement. Mr Nagashima, Dr Motohiro, Dr Takeda and Dr Itoh from Toyota visited in<br />
December 2005 for an annual meeting and progress report.<br />
118
UNSW Centre for Energy and Environmental Markets (CEEM)<br />
The PV Centre is one <strong>of</strong> the founding members <strong>of</strong> a new UNSW Centre for Energy and<br />
Environmental Markets. CEEM involves collaboration between the Faculty <strong>of</strong> Engineering, the<br />
Faculty <strong>of</strong> Commerce and Economics, and the Australian Graduate School <strong>of</strong> Management.<br />
It undertakes interdisciplinary research in the design, analysis and performance <strong>of</strong> energy<br />
and environmental markets and their associated policy frameworks. It was established in<br />
2004, largely as a response to government initiatives to reorganise infrastructure industries<br />
(such as electricity, gas, water and telecommunications) and to rely increasingly on markets<br />
in tradable environmental instruments as a form <strong>of</strong> environmental regulation.<br />
CEEM brings together researchers in UNSW and partner organisations, contributing to<br />
UNSW’s ability to provide world-class research, advice and education. Dr Muriel Watt is<br />
involved with the research strand focussing on “Policy framework and policy instruments<br />
in energy and the environment”. Other strands are Energy market design, Derivative<br />
markets, Experimental markets, Applications <strong>of</strong> artifi cial intelligence and Economic valuation<br />
methodologies and their application to energy and environmental issues.<br />
Australian Business Council for Sustainable Energy (BCSE)<br />
PV Centre staff liaise regularly with the BCSE on PV industry development and policy issues.<br />
Dr Muriel Watt and Dr Richard Corkish regularly attend the BCSE PV Directorate meetings.<br />
A number <strong>of</strong> consultancies and research projects have also been funded and longer term<br />
collaborative arrangements are now under discussion.<br />
CSG Solar Pty. Ltd, Botany, Sydney<br />
CSG Solar AG, Thalheim, Germany<br />
Strong collaboration continued with CSG Solar (formerly Pacifi c Solar), a Centre “spin-<strong>of</strong>f”.<br />
In 2004, the Centre purchased and began using an operational cleanroom facility at Bay<br />
Street, Botany, formerly owned by CSG Solar. CSG Solar provides cleanroom services (gases,<br />
exhausting, gas monitoring, OH&S , etc.) on a user-pays basis, with CSG Solar sharing use<br />
<strong>of</strong> some <strong>of</strong> the Centre-owned cleanroom equipment on the same basis. During 2005, the<br />
Centre provided characterisation services for CSG Solar, while CSG Solar provided thin-fi lm<br />
material for the Centre’s metallisation and cell interconnection project. Centre Executive<br />
Research Director, Martin Green, also acted as a consultant to the company during 2005<br />
and served on the German company’s Supervisory Board.<br />
BP Solar Espana, Tres Cantos, Spain<br />
In 2005, UNSW and BP Solar, Pty concluded a major Australian Research Council Linkage<br />
project that spanned 2002 to mid 2005. The project’s aimed to improve the effi ciency <strong>of</strong> p-<br />
type double sided buried contact solar cells. The main outcomes <strong>of</strong> the project were a detailed<br />
understanding <strong>of</strong> the formation <strong>of</strong> boron diffusion induced defects and their impact on the<br />
voltage and fi ll factor <strong>of</strong> solar cells, the role <strong>of</strong> oxidation induced stress on the recombination<br />
in textured silicon wafers, the fi rst conclusive experimental demonstration <strong>of</strong> Depletion<br />
Region Modulation, an artefact that affects photoconductance measurements and a detailed<br />
accounting <strong>of</strong> the key loss mechanisms in the double sided solar cell. The project resulted in<br />
5 major journal articles and 7 conference papers.<br />
119
Close collaboration between UNSW and BP Solar Espana continued during 2005 with an 8-<br />
week Postgraduate Industrial Training visit by Malcolm Abbott to the Tres Cantos production<br />
facility. This visit focused on developing BP Solar’s capabilities in advanced characterisation<br />
techniques, especially dark IV and local ideality factor analysis and photoconductance<br />
techniques. Malcolm gained invaluable industrial experience while spending 8 weeks living in<br />
nearby Madrid and absorbing everything Spanish.<br />
Suntech-Power Company, China<br />
Project 1: Innovative Emitter Design and Metal Contact for Screen-printed Silicon<br />
Solar Cells<br />
Research Team<br />
Pr<strong>of</strong>. Stuart Wenham (UNSW – Team Leader)<br />
Anita Ho (Postdoc Fellow – UNSW)<br />
Ly Mai (PhD student – UNSW)<br />
Tjahjono (PhD student – UNSW)<br />
Roland Utama (PhD student – UNSW)<br />
Philip Hamer (undergraduate research scholarship)<br />
Dr Z. Shi (Suntech)<br />
Dr Jingjia Ji (Suntech)<br />
A Zhu (Suntech)<br />
Li Hua (Suntech)<br />
Victor Chen (Suntech)<br />
Aim<br />
The broad aim <strong>of</strong> this work is to develop the next generation <strong>of</strong> screen-printed solar cell<br />
for implementation on the Suntech-Power production line. In particular, the fundamental<br />
limitations <strong>of</strong> the conventional screen-printed solar cell that have limited its performance<br />
for the last 30 years have been identifi ed, and innovative approaches to redesigning the<br />
emitter and front metal contact have been devised and are being developed and analysed<br />
in this work. More specifi cally, the initial aim in this work is to develop and demonstrate an<br />
innovative emitter design for the screen-printed solar cell that overcomes the current and<br />
voltage limitations imposed by the standard design shown in Figure 4.7.1, while retaining<br />
compatibility with existing equipment and infrastructure currently used for the manufacture<br />
<strong>of</strong> screen-printed solar cells.<br />
The primary aim during 2006 will be to complete the pilot line evaluation <strong>of</strong> the new technology<br />
in readiness for large scale production by the end <strong>of</strong> 2006. In particular, effi ciencies in excess<br />
<strong>of</strong> 18% in large scale production will be targeted following the achievement <strong>of</strong> effi ciencies<br />
in the range <strong>of</strong> 18-18.5% in pilot production in 2005 using all the same materials, wafers,<br />
equipment and processes.<br />
Background<br />
Screen-printed solar cell technology dominates commercial photovoltaic manufacturing, with<br />
well over 50% share <strong>of</strong> international markets. Despite the dominance <strong>of</strong> this technology, the<br />
solar cell design shown in Figure 4.7.1, has signifi cance performance limitations that constrain<br />
the cell effi ciencies to well below those achievable in research laboratories around the world.<br />
In particular, the front surface screen-printed metallisation necessitates a heavily diffused<br />
emitter to achieve low contact resistance and also to achieve adequate lateral conductivity<br />
in the emitter since the metal lines need to be widely spaced compared to laboratory cells<br />
120
to avoid excessive shading losses. Such cells therefore typically have emitters with sheet<br />
resistivities in the range <strong>of</strong> 40-50 ohms per square, which inevitably give signifi cantly degraded<br />
response to short wavelength light. To raise this sheet resistivity to above 100 ohms per<br />
square as required for near unity internal quantum effi ciencies for short wavelength light,<br />
serious resistive losses are introduced, both in the emitter and the contact resistance at the<br />
metal to n-type silicon interface.<br />
Furthermore, the conventional design <strong>of</strong> Figure 4.7.1 has quite poor surface passivation in<br />
both the metallised and non-metallised regions. Even if good ohmic contacts could be made<br />
to more lightly doped emitters, the large metal/silicon interface area would signifi cantly limit<br />
the voltages achievable due to the high levels <strong>of</strong> recombination in these regions and hence<br />
contribution to the device dark saturation current. These voltage limitations are not <strong>of</strong> major<br />
signifi cance at the moment due to the limitations imposed by the substrates. However, In the<br />
future as wafer thicknesses are reduced to improve the device economics, the cells will have<br />
the potential for improved open circuit voltages, but only provided the surfaces, including<br />
under the metal, are well passivated.<br />
150-200<br />
p+<br />
n++<br />
p-type<br />
3mm<br />
patterned metal contact<br />
phosphorus<br />
bulk <strong>of</strong> wafer<br />
rear metal contact<br />
Figure 4.7.1: Conventional screenprinted<br />
solar cell with heavily<br />
diffused emitter and large metal/<br />
silicon interface area for the front<br />
surface metallisation.<br />
metal<br />
Innovative Emitter Design<br />
The new emitter design is shown in Figure 4.7.2. The top surface is diffused to 100 ohms per<br />
square, while the heavily diffused grooves act as semiconductor fi ngers to carry the current<br />
to the screen printed silver fi ngers that run perpendicular to the grooves as shown in Figure<br />
4.7.3. Also not shown in Fig 4.7.2 is the surface passivating dielectric that not only passivates<br />
the lightly diffused surface so as to give near unity internal quantum effi ciencies for short<br />
wavelength light as shown in Figure 4.7.4, but it also isolates the metal from these same<br />
regions to minimise the device dark saturation current. The metal only contacts the silicon<br />
within the heavily diffused grooves, therefore giving a low area contact while still achieving low<br />
contact resistance. Fill factors <strong>of</strong> 79-80% have been demonstrated with this structure on<br />
large area devices <strong>of</strong> approximately 150cm 2 , area verifying the effectiveness <strong>of</strong> this contacting<br />
scheme. Effi ciencies as high as 18.3% (inhouse<br />
measurements) have been achieved,<br />
30 microns<br />
contributed to signifi cantly by the high short<br />
circuit currents gained through the excellent<br />
100 / n-type<br />
spectral response <strong>of</strong> Figure 4.7.4 and the low<br />
metal and top surface shading losses.<br />
45 microns<br />
Figure 4.7.2: Cross-section <strong>of</strong> the<br />
investigated innovative emitter design using<br />
semiconductor fi ngers, developed to address<br />
the fundamental limitations <strong>of</strong> screen-printed<br />
metal contacts with their inability to produce<br />
fi ne lines and make ohmic contact to lightly<br />
diffused emitters.<br />
5 / n++<br />
p-type silicon<br />
n++<br />
rear surface <strong>of</strong> solar cell<br />
121
Figure 4.7.3 : Screen-printed fi ngers running<br />
perpendicular to the heavily diffused grooves<br />
where electrical contact is made. A dielectric/<br />
AR coating passivates the top surface and<br />
isolates the metal from the lightly diffused top<br />
surface.<br />
%<br />
100<br />
90<br />
80<br />
70<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
S6A-33452<br />
EQE<br />
Reflection<br />
IQE<br />
0<br />
300 400 500 600 700 800 900 1000 1100 1200<br />
Wavelength (nm)<br />
Figure 4.7.4: Excellent IQE achieved through<br />
the use <strong>of</strong> the innovative emitter design<br />
and low refl ection losses due to AR coated<br />
textured surface and low metal shading<br />
The grooves are typically spaced less<br />
than a millimetre apart so as to minimise<br />
resistive losses within the lightly diffused<br />
emitter, while the screen-printed metal<br />
lines can be spaced signifi cantly further<br />
apart than in normal screen-printed cells <strong>of</strong> Figure 4.7.1 due to the comparatively excellent<br />
lateral conductivity <strong>of</strong> the emitter achieved by the very heavy doping within the grooves. This<br />
concept <strong>of</strong> semiconductor fi ngers does not appear to have ever been used in commercial<br />
solar cells, and has considerable appeal as it facilitates good conductivity within the emitter.<br />
This is acheived without the normal trade-<strong>of</strong>f found in screen printed cells where such regions<br />
<strong>of</strong> good emitter conduction are located at the top surface and therefore degrade the cell<br />
spectral response and current generating capability, due to the corresponding extremely<br />
short minority carrier diffusion lengths in such regions.<br />
Fig. 4.7.5 shows an SEM <strong>of</strong> a screen-printed metal line crossing one <strong>of</strong> the heavily phosphorus<br />
diffused laser grooves. The silicon is only exposed within the grooves, with the screen-printed<br />
metal having been shown to make excellent ohmic contact to the heavily phosphorus<br />
diffused silicon in these regions. Both thick oxides and silicon nitride layers, when used with<br />
appropriate pastes, appear to provide adequate protection to the lightly diffused surface<br />
regions, preventing the screen-printed metal from contacting the silicon.<br />
A simplifi cation <strong>of</strong> the proposed emitter design is to apply a phosphorus doped passivating<br />
dielectric after lightly diffusing the top surface. The laser scribing conditions for groove<br />
formation are then modifi ed so as to melt the silicon rather than ablate it, thereby allowing<br />
large amounts <strong>of</strong> phosphorus to penetrate into the molten silicon, producing heavily doped<br />
channels rather then grooves. This avoids the need for etching the grooves and subsequently<br />
122
diffusing the groove walls. The performance <strong>of</strong> these devices however does not currently<br />
match that <strong>of</strong> the devices produced using the emitter design <strong>of</strong> Figure 4.7.2. A further<br />
simplifi cation for producing the semiconductor fi ngers that appears likely to produce the<br />
best results <strong>of</strong> all three approaches and is arguably the simplest and lowest cost approach,<br />
is under development using ink jet technology. Patent protection is currently being sought for<br />
this work.<br />
Fig. 4.7.5: SEM <strong>of</strong> a screen-printed<br />
metal line making excellent electrical<br />
contact to one <strong>of</strong> the heavily<br />
phosphorus diffused laser grooves<br />
running perpendicular to the metal<br />
line. The dark regions are textured<br />
and lightly diffused to 100 Ω/square<br />
and well passivated by a thick<br />
thermally grown oxide<br />
Financial Support<br />
Suntech-Power has been particularly generous, not only funding all the collaborative research<br />
conducted in China, but also fully funding two research positions at UNSW for a period <strong>of</strong> 12<br />
months through to April 2005. The latter alone is equivalent to a cash contribution in excess<br />
<strong>of</strong> $50k, while the inkind contribution through the former has been valued in excess <strong>of</strong> $300k.<br />
In addition, Suntech has provided a cash contribution <strong>of</strong> approximately $200k in support<br />
<strong>of</strong> the collaborative research, with a commitment for this to increase to $500k for 2006.<br />
Furthermore, Suntech funded all the living, travel and accommodation expenses for the 12<br />
staff and students who worked at Suntech on the collaborative research.<br />
Suntech Project 2 : High Efficiency cells based on ink jet technology<br />
Research Team<br />
Pr<strong>of</strong> Stuart Wenham (UNSW – Team Leader)<br />
Pr<strong>of</strong> Martin Green (Executive Research Director – UNSW)<br />
Anita Ho (Postdoc Fellow – UNSW)<br />
Ly Mai (PhD student – UNSW)<br />
Tjahjono (PhD student – UNSW)<br />
Roland Utama (PhD student – UNSW)<br />
Nicole Kuepper (research assistant – UNSW)<br />
Alison Lennin (PhD student – UNSW)<br />
Philip Hamer (undergraduate research scholarship – UNSW)<br />
An exciting new research area was established at UNSW during 2005 based on ink jet<br />
technology. Ink jet printing has been applied to a wide range <strong>of</strong> applications and research<br />
areas in recent years, with CSG Solar pioneering the application <strong>of</strong> such technology to thinfi<br />
lm solar cell fabrication. Independent research activities have been established at UNSW to<br />
develop new approaches and technologies based on ink jet technology, primarily for application<br />
to wafer based technologies. Several high effi ciency technologies are being developed with<br />
a key factor being the avoidance <strong>of</strong> any photolithographic or masking processes. It appears<br />
effi ciencies on fl oatzone material above 23% are achievable while more interestingly<br />
effi ciencies above 20% appear feasible with solar grade p-type CZ wafers and likely to peak at<br />
effi ciencies above 22% for CZ n-type wafers.<br />
123
Figure 4.7.6 shows an example <strong>of</strong> one <strong>of</strong> the particularly versatile cell structures being<br />
developed by the ink jet printing team for wafer based devices. This work was initially reported<br />
in mid 2005 although limited details are available as patent protection is being sought. This<br />
device structure is <strong>of</strong> particular interest to Suntech as it can be used as a contacting scheme<br />
for a new thin-fi lm crystalline silicon solar cell technology, or can be applied to either n or p type<br />
wafers for rear contacting, or perhaps most importantly, can be used as an improved rear<br />
surface passivation and contacting approach for existing commercial technology through<br />
the use <strong>of</strong> a rear fl oating junction by eliminating the n-type contact. This latter rear surface<br />
structure, when applied to the standard PERL front surface, has held the world record for<br />
silicon solar cell voltages for most <strong>of</strong> the last decade with independently confi rmed opencircuit<br />
voltages in excess <strong>of</strong> 720mV. The new inkjet printing development system plus several<br />
<strong>of</strong> the ink jet printing team at UNSW are shown in Figure 4.7.7<br />
dielectric<br />
n- or p-type<br />
p- or n-type<br />
resin<br />
+ or - contact - or + contact<br />
Figure 4.7.6 : Versatile device<br />
contacting scheme under<br />
development using ink jet<br />
technology and applicable to<br />
wafer based devices, thin-fi lm<br />
technologies and for improved<br />
rear surface passivation and<br />
metal contacting for current<br />
commercial cell technologies.<br />
Figure 4.7.7 : <strong>New</strong> Ink jet printing<br />
development system at UNSW<br />
Centre for Quantum Computer Technology, Melbourne <strong>University</strong><br />
A collaboration has been established with Dr Chris Pakes at Melbourne <strong>University</strong>. Dr Pakes<br />
has used conductive atomic force microscopy (CAFM) on our Si QD resonant tunnelling<br />
structures. CAFM has proved to be a powerful tool for mapping transport properties at a<br />
very high areal resolution.<br />
In 2005 more detailed analysis has been carried out using this technique <strong>of</strong> the SiO 2 /Si QD/<br />
SiO 2 double barrier resonant tunnelling structures discussed in Section 4.5.6.1. Conductive<br />
atomic force microscopy (CAFM) at Melbourne <strong>University</strong> has been used to map the<br />
conductivity <strong>of</strong> single Si QD layer mesa samples containing about 1010 5nm QDs per mesa<br />
or 10,000 QDs per µm 2 . Figure 4.7.8 shows a CAFM scan at a constant voltage <strong>of</strong> 10V.<br />
The white spots are areas <strong>of</strong> high conductivity, indicating the high tunnelling probability at a<br />
resonant QD. Figure 4.7.8 also shows a line scan across two <strong>of</strong> these white spots and shows<br />
the sharp change in conductivity when resonance is achieved.<br />
124
Thus this represents a QD areal mapping technique. However, the voltage used at present is<br />
high because the oxide layers are quite thick at 5nm. Hence the indicated tunnelling regions<br />
may be either from very small QDs or from higher confi ned energy levels within QDs and only<br />
indicate a lower limit on the QD density. Pro<strong>of</strong> that these do indeed indicate QD locations was<br />
obtained by a subsequent anneal <strong>of</strong> the sample in an ordinary ambient. A subsequent CAFM<br />
scan showed no white regions, thus showing that the Si QDs had been oxidised in the anneal.<br />
Further experiments will use thinner oxide layers and should give a more complete map <strong>of</strong> the<br />
QD structure. Scans <strong>of</strong> such structures at various biases may give further information on<br />
QD size and size distribution by effectively mapping the confi ned energy levels at a very high<br />
areal resolution.<br />
Figure 4.7.8: (a) Current image obtained at 10V applied voltage (probe tip grounded); the<br />
white dots indicate areas <strong>of</strong> higher current conduction (up to 800 pA); (b) line scan <strong>of</strong> current<br />
pr<strong>of</strong>i le in region shown by red arrow in (a).<br />
125
126
Summary<br />
The Centre for Advanced Silicon Photovoltaics and Photonics incorporates the activities <strong>of</strong><br />
the former Key Centre for Photovoltaic Engineering. The former Key Centre started in 1999,<br />
after the award <strong>of</strong> special funding from the Australian Government to promote teaching<br />
and research in the area <strong>of</strong> photovoltaics. One <strong>of</strong> the primary new initiatives <strong>of</strong> this Key<br />
Centre was the establishment <strong>of</strong> the world’s fi rst undergraduate degree in Photovoltiacs and<br />
Solar Energy. This was one <strong>of</strong> only eight such Key Centres awarded Australia-wide across all<br />
disciplines, demonstrating the importance the government placed on this new and exciting<br />
fi eld <strong>of</strong> photovoltaics. In 2003 the Key Centre and its activities were incorporated into the<br />
new Centre <strong>of</strong> Excellence, awarded to the same team at the <strong>University</strong> <strong>of</strong> <strong>New</strong> <strong>South</strong> <strong>Wales</strong>.<br />
The Centre <strong>of</strong>fers undergraduate, postgraduate and research programs encompassing a<br />
range <strong>of</strong> aspects relating to the photovoltaic and renewable energy industries. These programs<br />
have been developed in consultation with representatives from industry to ensure graduates<br />
are appropriately qualifi ed to enter the fi eld upon completion <strong>of</strong> their studies. Since 2003,<br />
the Centre has seen a total <strong>of</strong> 40 students graduate from its undergraduate programs, and<br />
23 students from its postgraduate degrees. These graduates are now taking advantage <strong>of</strong><br />
a range <strong>of</strong> opportunities being created in the booming photovoltaics and renewable energy<br />
industries.<br />
During 2005 there was a signifi cant amount <strong>of</strong> curriculum development and revision due<br />
to new initiatives <strong>of</strong> the Faculty <strong>of</strong> Engineering which aimed at adding value to engineering<br />
programs at UNSW. Throughout the year the Centre continued to attract bright students<br />
to all our educational programs, and we were very excited to be able to <strong>of</strong>fer a new Co-<br />
Op Scholarship sponsored by Suntech Power Co. Ltd. School Offi ce staff also concentrated<br />
on reinvigorating existing promotional material, and utilising new promotional activities to<br />
increase the awareness <strong>of</strong> Photovoltaics and Renewable Energy Engineering at UNSW.<br />
At the end <strong>of</strong> 2005 there were 137 undergraduate students, 7 postgraduate coursework<br />
students, and 23 research students enrolled in the Centre’s educational programs.<br />
5.1 Undergraduate Educational Programs<br />
The Centre <strong>of</strong>fers two undergraduate coursework<br />
engineering programs, which can be taken in<br />
combination with a Bachelor <strong>of</strong> Arts, Bachelor <strong>of</strong><br />
Science or a Bachelor <strong>of</strong> Commerce. These degrees<br />
aim to develop well-educated graduates with the<br />
basic skills, attributes and knowledge required to<br />
practise as pr<strong>of</strong>essional engineers in the growing<br />
photovoltaics and renewable energy industries.<br />
The photovoltaics program draws on the expertise<br />
<strong>of</strong> Centre staff in this fi eld and was the fi rst <strong>of</strong><br />
its kind internationally. The program focuses on<br />
photovoltaic devices, and includes training in<br />
technology development, manufacturing, quality<br />
control, reliability and life cycle applications, system<br />
design, maintenance and fault diagnosis, marketing,<br />
policy development and the use <strong>of</strong> other renewable<br />
energy technologies. It has two unique features:– a<br />
project in the second year <strong>of</strong> the program, and a ‘strand’, a second area <strong>of</strong> specialisation. Since<br />
2003 a total <strong>of</strong> 40 students have graduated from the program, as previously mentioned, and<br />
another 10 are expected to graduate in March 2006 after completing their studies at the<br />
end <strong>of</strong> 2005.<br />
127
The second broader degree in Renewable Energy Engineering was introduced in 2003 and<br />
will see the fi rst students graduate in 2006. At the end <strong>of</strong> 2005 a total <strong>of</strong> 42 students were<br />
majoring in this area. In addition to photovoltaic devices, students in this program study solar<br />
architectural technologies, wind energy, biomass, solar thermal, and renewable energy policy<br />
as part <strong>of</strong> their core curriculum.<br />
Students generally enrol in these programs because they are<br />
interested in working with cutting edge technology, and they are<br />
enthusiastic about making a personal contribution to society and the<br />
environment. Total enrolments increased slightly in 2005 with an<br />
increase in students electing to enrol in combined degree programs.<br />
This can be attributed to the increase in range <strong>of</strong> combined degree<br />
programs <strong>of</strong>fered by the Centre which, from 2005, includes a<br />
Bachelor <strong>of</strong> Engineering / Bachelor <strong>of</strong> Commerce, and a Renewable<br />
Energy Engineering / Bachelor <strong>of</strong> Science program. The new<br />
combined Commerce program aims to enhance the technical and<br />
scientifi c expertise <strong>of</strong> Engineering graduates with an understanding<br />
<strong>of</strong> commercial and fi nancial principles which will assist engineers to<br />
manage projects, obtain fi nance for the commercialisation <strong>of</strong> new<br />
products, create and manage companies, manage the manufacturing<br />
and marketing <strong>of</strong> products, and accurately predict and monitor<br />
market and economic trends. The Renewable Energy / Science<br />
combined program allows students to expand their understanding<br />
<strong>of</strong> scientifi c concepts which form the basis <strong>of</strong> engineering principles.<br />
Students installing a<br />
solar-powered clean water<br />
system in Sri Lanka<br />
At the end <strong>of</strong> 2005, 27% <strong>of</strong> enrolled students were women. This is<br />
unusual when compared to other Engineering programs which typically<br />
have a participation rate by women, <strong>of</strong> about half this. The women<br />
enrolled in our undergraduate degree programs are more inclined to<br />
enrol in a single degree program, and in particular the Photovoltaics and<br />
Solar Energy program. Women who elect to enrol in a combined degree,<br />
are more likely to choose a combined Photovoltaics and Solar Energy /<br />
Arts program.<br />
Students using a<br />
computer lab<br />
Graduates <strong>of</strong> our undergraduate programs are gaining employment<br />
within a variety <strong>of</strong> organisations, and increasingly in the area <strong>of</strong> energy<br />
effi ciency and sustainable design. Each year a number <strong>of</strong> graduates elect<br />
to expand upon their undergraduate degree by undertaking a research<br />
program with our internationally recognised research groups, whilst<br />
others are employed in industry by solar cell and equipment manufacturers, system<br />
design and integration companies, electricity utilities and major end users <strong>of</strong> products<br />
(such as sustainable design organisations).<br />
5.1.1 The Strand<br />
The strand is unique to the Photovoltaics and Solar<br />
Energy program, and is essentially a second area <strong>of</strong><br />
specialisation which complements a student’s study<br />
<strong>of</strong> photovoltaics. Students can take strands covering<br />
a variety <strong>of</strong> areas including computing, electronics,<br />
mathematics, physics, mechanical engineering,<br />
civil engineering, and architecture. The aim <strong>of</strong> the<br />
128
strand is to provide students with broader engineering backgrounds important for the<br />
cross-disciplinary nature <strong>of</strong> photovoltaic applications. UNSW’s Sunswift Solar Car project<br />
is one example <strong>of</strong> the cross-disciplinary nature <strong>of</strong> photovoltaic applications, requiring the<br />
skills <strong>of</strong> mechanical engineers, biomedical engineers, and electrical engineers in addition<br />
to photovoltaic engineers. Since the commencement <strong>of</strong> the program, the Electronics and<br />
Physics strands have been the most popular, with Architecture, Mechanical Engineering and<br />
Computing also popular choices among students.<br />
5.1.2 Goldman Sachs Recipient<br />
One <strong>of</strong> the Centre’s outstanding undergraduate students,<br />
Nicole Kuepper, was one <strong>of</strong> eight students in 2005 to receive an<br />
Australian fellowship as part <strong>of</strong> the Goldman Sachs Global Leaders<br />
Program. Nicole received a <strong>New</strong> <strong>South</strong> Scholarship for achieving<br />
a UAI <strong>of</strong> 100, and is heavily involved with community activities<br />
including volunteer projects with indigenous Australians, teaching<br />
English in the Phillipines and fundraising events. She is enrolled in<br />
the Photovoltaics and Solar Energy program and hopes to bring<br />
sustainable improvement to disadvantaged communities through<br />
engineering and education.<br />
The Goldman Sachs Leaders Program, which is run by the Institute<br />
<strong>of</strong> International Education and the Goldman Sachs Foundation,<br />
acknowledges the outstanding achievements and abilities <strong>of</strong><br />
students across a range <strong>of</strong> disciplines. Nicole joins a group <strong>of</strong> 100<br />
outstanding students from around the world to be awarded the<br />
fellowship, which recognises academic performance and community<br />
contribution.<br />
Nicole Kuepper with English<br />
student in the Phillipines<br />
5.2 Undergraduate Student Projects<br />
In the second year <strong>of</strong> the Photovoltaics and Solar Energy program, students have the<br />
opportunity <strong>of</strong> undertaking a year long project in the photovoltaic or renewable energy areas.<br />
The main emphasis <strong>of</strong> the second year project course is hands-on “projecteering”, or project<br />
engineering. The course has a lecture component covering projecteering skills and practice,<br />
and each project has a research component, a planning component, a hands-on component<br />
and a reporting component. This course helps to prepare students for their fourth year<br />
thesis, which is generally undertaken by all students enrolled in both the Photovoltaics and<br />
Solar Energy and Renewable Energy Engineering undergraduate programs. While in the past<br />
projects in developing countries have been the most popular among students, in 2005 there<br />
was a swing in popularity towards laboratory-based projects with the introduction in 2005 <strong>of</strong><br />
a new project on Screen Printed Solar Cells.<br />
5.2.1 Biodiesel Trailer<br />
The Biodiesel Project aims to construct a small-scale, portable chemical factory to convert<br />
used vegetable oil, from UNSW’s food outlets, into diesel fuel for UNSW’s fl eet <strong>of</strong> diesel cars.<br />
The project has now been running for four years and is at a stage where completion is in<br />
sight. In 2005, four second-year students who undertook the Biodiesel Project implemented<br />
two key subsystems:- the pneumatic fl uid handling mixing subsystem, and the fl uid handling<br />
129
system for the solar thermal hot water. One <strong>of</strong> the key challenges was handling potentially<br />
explosive vapours and fl ammable liquids, and this necessitated a completely pneumatic fl uid<br />
mixing and pumping system instead <strong>of</strong> the more common electrical pumping and mixing<br />
systems. Another challenge was investigating, designing and implementing the hot water<br />
system, which required extensive research into codes and best practices for the plumbing.<br />
The students spent several full days on the construction effort on campus, learning the<br />
not-so-easy art <strong>of</strong> bending and fi tting rigid stainless steel and copper tubing. The hot water<br />
system was completed, but not tested. The pneumatic pumps and motors were fi tted and the<br />
stainless steel plumbing was started, but more work is needed.<br />
5.2.2 Developing Countries<br />
The Centre has been involved<br />
with projects in the developing<br />
countries for the last fi ve years.<br />
The application <strong>of</strong> photovoltaics<br />
and other renewable energy<br />
technologies can make the<br />
greatest difference to people’s<br />
lives and living standards in<br />
these places. Students involved<br />
with this project in the past have<br />
installed photovoltaic systems in<br />
villages in Nicaragua and Nepal.<br />
Students installing a solar-powered clean water system in Sri Lanka<br />
A student investigating solar-powered water<br />
pumping<br />
In 2005, this second-year project group <strong>of</strong> six students<br />
worked on a variety <strong>of</strong> technologies which included<br />
assembling a micro-hydro testing system, a water<br />
purifi cation system, a small solar powered system and<br />
low cost passive solar design for houses in developing<br />
countries. The culmination <strong>of</strong> this work resulted in some<br />
students travelling to Sri Lanka over the summer break<br />
to install PV powered clean water systems in a small<br />
village called Seenigama, which was badly affected by<br />
the Tsunami that struck the region in 2005. Skills that<br />
the students had developed throughout the year were<br />
put to good use in the fi eld. Invaluable lessons were<br />
gained regarding project management <strong>of</strong> activities<br />
in developing countries, as students had to face and<br />
overcome many technical and non-technical issues in<br />
implementing the project.<br />
130
5.2.3 Screen-printed Solar Cells<br />
Screen-printed solar cells are the workhorse <strong>of</strong> today’s solar cell industry and will likely remain<br />
the dominant technology for the foreseeable future. Screen-printed modules are bringing<br />
renewable electricity to people in both developing countries and the developed world. UNSW<br />
researchers are developing new printed solar cell technologies to improve effi ciency and<br />
cost <strong>of</strong> commercial technologies. This project aims to give undergraduate students broad<br />
exposure to the fi eld by engaging them in laboratory infrastructure development projects.<br />
In 2005, students undertook a variety <strong>of</strong> engineering projects around key equipment and<br />
infrastructure, including designing and implementing a shuttering system for our CNC laser<br />
machine, installing and commissioning a thermal muffl e furnace, repairing and upgrading<br />
a quartz furnace stack and receiving, commissioning and testing a new production screen<br />
printer. After completing this project, the student’s interest in laboratory research,<br />
development and infrastructure activities was such that they were placed for summer R&D<br />
employment at UNSW (as part <strong>of</strong> the Taste <strong>of</strong> Research Summer Scholarship program), at<br />
BP Solar (as part <strong>of</strong> UNSWs Co-op Scholarship Placement) or at Suntech Solar (as part <strong>of</strong> an<br />
ongoing summer exchange program). This highlights the success <strong>of</strong> this project at engaging<br />
students in laboratory or production activities.<br />
5.2.4 Thin Film Solar Processing Equipment<br />
This project aims to give students a broad experience in establishing and maintaining a worldclass<br />
thin-fi lm solar cell research laboratory. This is achieved through a two-stage process.<br />
Firstly students develop a base level understanding <strong>of</strong> equipment and systems within the<br />
thin-fi lm lab, which include vacuum systems, gas delivery systems, leak testing equipment<br />
and plasma processing. The second stage involves the students integrating what they have<br />
learnt by applying it to a specifi c project in the thin fi lm laboratory. In 2005, four secondyear<br />
students undertook this project. They were split into two groups and worked on two<br />
independent equipment upgrade projects. One group converted a decommissioned PECVD<br />
machine into a fully working plasma etching system, while the other group developed a control<br />
interface for a reactive ion etcher. Both groups undertook leak testing work on their system<br />
to achieve a better base pressure, and improved existing equipment safety and operation<br />
documentation.<br />
5.2.5 Fourth Year Thesis<br />
The thesis is generally carried out in the last two sessions <strong>of</strong> an<br />
engineering undergraduate student’s studies. Students undertake<br />
directed laboratory and research work on an approved subject under<br />
guidance <strong>of</strong> the Centre’s academic staff. Typically, the thesis involves<br />
the design and construction <strong>of</strong> experimental apparatus together<br />
with practical tests. Each student is required to present a seminar,<br />
submit a written report, and present a poster as part <strong>of</strong> an Open Day.<br />
It is a chance for students to demonstrate what they have learned<br />
throughout their studies.<br />
Theses submitted during 2005 were <strong>of</strong> a very high quality. Contrary<br />
to 2004 where application-based photovoltaic and renewable energy<br />
topics were the most popular with students, in 2005 laboratorybased<br />
research topics were the most popular. Theses in this area<br />
covered topics such as the “Saturn 7” light source for solar cell<br />
measurements, the quantum effi ciency <strong>of</strong> polycrystalline thin-fi lm<br />
solar cells, texturing <strong>of</strong> GaAs solar cells by nanosphere lithography,<br />
2005 Thesis Poster-Presentation<br />
Prize winners receiving their award<br />
from Dr Richard Corkish, Head <strong>of</strong><br />
School. Session 1 winner: Andrew<br />
C<strong>of</strong>fey; Session 2: Roland Utama<br />
131
and the determination <strong>of</strong> dopant pr<strong>of</strong>i les in crystalline silicon using wet chemical layer<br />
removal technique. However, students did investigate other topics such as life cycle analysis<br />
<strong>of</strong> alternative fuels for transport, energy and water effi ciency in schools, sustainability in<br />
commercial <strong>of</strong>fi ce buildings, a solar thermal hot water heater, solar home systems in Sri<br />
Lanka, and renewable energy policy.<br />
One outstanding thesis submitted received both the 2005 Photovoltaics Thesis Prize as well<br />
as the Poster Presentation Prize for session 2. This thesis by Roland Utama titled ‘Capacitance<br />
Voltage Analysis <strong>of</strong> Polycrystalline Silicon Thin Film Solar Cell’, investigates the applications<br />
<strong>of</strong> a novel capacitance-voltage measurement using impedance analysis to determine the<br />
active doping concentration <strong>of</strong> boron-doped ALICIA solar cells. Roland examined the infl uence<br />
<strong>of</strong> defects and two post-deposition treatments (RTA and hydrogenation) on the active base<br />
doping concentration, as well as the effect <strong>of</strong> varying boron doping concentration on the base<br />
doping type. His thesis has also been nominated for the 2005 Wal Read Memorial Prize<br />
presented by the Australia <strong>New</strong> Zealand Sustainable Energy Society (ANZSES).<br />
5.3 Postgraduate Educational Programs<br />
The Centre for Photovoltaic Engineering <strong>of</strong>fers one postgraduate coursework<br />
program, a Master <strong>of</strong> Engineering Science, and three research programs:- a<br />
Master <strong>of</strong> Philosophy, Masters by Research and a Doctor <strong>of</strong> Philosophy. These<br />
degrees are intended to provide students with an exceptional basis in advanced<br />
concepts and research in the photovoltaic area.<br />
The one-year coursework program was developed to build on the previous<br />
engineering education <strong>of</strong> engineers from other engineering disciplines who are<br />
currently being attracted to the photovoltaics and renewable energy industries.<br />
Students study courses chosen from the areas <strong>of</strong> photovoltaic devices, photovoltaic<br />
systems and applications, and renewable energy technologies. Enrolments in the<br />
program have been steady with total enrolments <strong>of</strong> 7, and 2 graduates in 2005.<br />
At the end <strong>of</strong> 2005, the Centre had 23 students enrolled in postgraduate research<br />
degrees, 18 in a Doctor <strong>of</strong> Philosophy, and 5 in a Masters by Research. Research<br />
topics available to these students cover the entire photovoltaic sector, but with<br />
greatest emphasis on device theory, device and module design, balance <strong>of</strong> system<br />
components, and photovoltaic systems and applications. Research students are<br />
involved across all the Centre’s research groups and play an important role in the<br />
Centre’s activities. In 2005 the Centre saw 6 research students graduate, the<br />
largest number so far to graduate from these research degrees.<br />
5.3.1 Dean’s Award for Excellence in Postgraduate<br />
Research<br />
One <strong>of</strong> the Centre’s research students was the recipient<br />
<strong>of</strong> fi rst prize in the Faculty <strong>of</strong> Engineering’s inaugural<br />
Dean’s Awards for Postgraduate Research. Supriya<br />
Pillai’s work was selected as the best overall from about<br />
70 applicants at a poster presentation at the Scientia<br />
in December.<br />
Supriya Pillai - First prize winner <strong>of</strong> the Dean’s<br />
Award<br />
132
5.4 Scholarships<br />
The undergraduate programs at the Centre attract very bright students from across<br />
Australia. Even though the UAI cut <strong>of</strong>f for the programs has been approximately 80 over<br />
the last few years, almost half the students have a UAI over 90 with 40% <strong>of</strong> these students<br />
having a UAI over 95. The Co-Op Scholarship Program and the Faculty <strong>of</strong> Engineering’s Rural<br />
Scholarship Program help to attract these bright students, while the Taste <strong>of</strong> Research<br />
Summer Scholarship Program provides bright students with experience in the Centre’s<br />
laboratories and encourages them to pursue research careers. The Centre’s very high<br />
international pr<strong>of</strong>i le also allows it to attract high quality research students evidenced by the<br />
number <strong>of</strong> doctoral students undertaking research programs at the Centre who have been<br />
awarded either an APA or EIPRS Scholarship.<br />
5.4.1 Co-Op Scholarship Program and <strong>New</strong> Industry Sponsor<br />
The Co-Op Program is an industry-linked scholarship program where students obtain a year <strong>of</strong><br />
work experience with industry sponsors as part <strong>of</strong> their undergraduate studies. In addition to<br />
their outstanding academic achievements, students are selected based on their involvement<br />
in school and community activities, their demonstrated leadership skills and their ability to<br />
communicate. Participation in this program enables students to apply the knowledge they<br />
have gained during their studies. This program is also benefi cial to industry sponsors, who<br />
have access to excellent students, can observe these students in their workplace, and have<br />
the fi rst opportunity <strong>of</strong> recruiting the best new graduates.<br />
Two organisations, BP Solar and CSG Solar, have sponsored scholarships for the<br />
undergraduate Photovoltaics and Solar Energy program since 2001. However, in 2005<br />
the Centre was fortunate enough to obtain a 3rd industry sponsor, Suntech Power Co. Ltd.<br />
Suntech is a partly Australian-owned company located in China and, as a result, this becomes<br />
the fi rst international Co-Op Scholarship at the <strong>University</strong> <strong>of</strong> <strong>New</strong> <strong>South</strong> <strong>Wales</strong>. Suntech<br />
is an extremely successful and rapidly growing manufacturer <strong>of</strong> multicrystalline and single<br />
crystalline silicon solar cells, and this scholarship has further strengthened the existing<br />
research and educational relationship between this organisation and the Centre.<br />
In 2005, the Centre also experienced an increase in good quality applicants applying for the<br />
scholarship program. In 2004 industry sponsors and academics interviewed 17 students,<br />
however, in 2005 they interviewed 24. Feedback received from applicants regarding the<br />
interview process was also very positive to the effect that the interviews were the best<br />
organised that they had experienced as the students knew what was going on, where it was<br />
happening, and who could help.<br />
5.4.2 Rural Scholarship Program<br />
The Faculty <strong>of</strong> Engineering established the Rural<br />
Scholarship Program in 2001 to encourage highachieving<br />
students living in rural and isolated areas<br />
to study engineering. The scholarships are valued<br />
at approximately $8,500 per annum for four years<br />
<strong>of</strong> full-time study which eases the fi nancial hardship<br />
<strong>of</strong> relocating to and living in Sydney. Due to the high<br />
standard <strong>of</strong> applicants in 2005, the Centre funded<br />
an extra scholarship in addition to the Faculty-funded<br />
scholarship. These students will commence studying<br />
with the Centre in 2006.<br />
David Uncle is a rural scholar who commenced his studies in 2005. In March he formally<br />
received his scholarship from the Dean <strong>of</strong> Engineering, Pr<strong>of</strong>essor Parker.<br />
133
5.4.3 Taste <strong>of</strong> Research Summer Scholarship Program<br />
The Taste <strong>of</strong> Research Summer Scholarship Program is primarily for high achieving 3rd<br />
year students and, in exceptional cases, 2nd year students may be considered. As part <strong>of</strong><br />
the program, engineering schools <strong>of</strong>fer 10 week projects for students to complete during<br />
their summer break. These projects provide students with scholarship support to gain<br />
experience working as part <strong>of</strong> a research team, for example in the world class laboratories<br />
at the Centre. From December 2005 to March 2006, the Centre hosted seven students<br />
working on projects, three with the thin-fi lm group on crystalline silicon thin-fi lm solar cells<br />
on glass, two with the buried contact group on high effi ciency n-type silicon solar cells, one<br />
on innovative research with industry to improve commercial solar cell performance, and<br />
another on antenna collection <strong>of</strong> solar energy. David Bennett, an undergraduate student<br />
attracted to the Centre by the Taste <strong>of</strong> Research program from ANU, was awarded one <strong>of</strong><br />
the six Engineering Undergraduate Research Prizes for the best poster and presentation at<br />
the end <strong>of</strong> the program.<br />
Participation in these projects helps students further develop their technical skills as well<br />
as their written and oral communication as students have to write a report and present a<br />
poster on the outcomes <strong>of</strong> their research.<br />
5.4.4 Research Scholarships (APA and EIPRS)<br />
Through hands-on laboratory-based projects and theses, and scholarship programs such as<br />
the Taste <strong>of</strong> Research Summer Scholarship program, the Centre aims to encourage good<br />
quality undergraduate students into research. Being an internationally recognised research<br />
organisation, the Centre attracts very high quality students to its research programs where<br />
students have the opportunity to work with leaders in this fi eld. The fact that 11 <strong>of</strong> our<br />
doctoral students are in receipt <strong>of</strong> very competitive APAs (Australian Postgraduate Award)<br />
or an EIPRS (Endeavour International Postgraduate Research Scholarship) is testament to<br />
the quality <strong>of</strong> these students. To be awarded an APA the student must be a local student<br />
and ranked near the top <strong>of</strong> the fi rst class honours graduates, while the EIPRS is available to<br />
international students and appear to demand an even higher standard than the APAs.<br />
.<br />
5.5 Curriculum Development<br />
134<br />
Throughout 2005 students indicated through<br />
UNSW’s course evaluation process, CATEI, that<br />
they were satisfi ed with the Centre’s courses.<br />
On average, at least 88% <strong>of</strong> students were<br />
satisfi ed or very satisfi ed with the content,<br />
structure, and teaching <strong>of</strong> our courses. Whilst<br />
no new courses were <strong>of</strong>fered by the Centre, the<br />
Faculty undertook several initiatives in 2005 to<br />
attract undergraduate students to Engineering<br />
at UNSW, each <strong>of</strong> which involved Centre staff.<br />
These include the introduction <strong>of</strong> a new fl exible<br />
fi rst year program, the development <strong>of</strong> a Facultywide<br />
fi rst year Engineering Design and Innovation<br />
course, and a new admissions scheme for local<br />
students. Furthermore, in order to try to increase<br />
the number <strong>of</strong> courses <strong>of</strong>fered each year to<br />
students, the Centre trialled a semi-distance<br />
mode <strong>of</strong> delivery for a core course on Sustainable<br />
and Renewable Energy Technologies.
5.5.1 <strong>New</strong> Flexible First Year<br />
In 2005 the Faculty undertook a revision <strong>of</strong> all Engineering programs to increase fl exibility for<br />
students in their fi rst year, and ensure a broad base in mathematics, science and engineering<br />
fundamentals. Prospective students had indicated that there was value in delaying their choice<br />
<strong>of</strong> specialisation, and this new program exposes students to the various engineering schools<br />
and enables them to make an informed choice at the end <strong>of</strong> their fi rst year. However, students<br />
who already know which fi eld <strong>of</strong> engineering they would like to enter, can enter directly into<br />
that specialisation in their fi rst year. The Centre’s educational programs were involved with<br />
this revision and academic and administrative staff were involved with the process that will<br />
see these new style programs effective from 2006.<br />
5.5.2 <strong>New</strong> Engineering Design and Innovation Course<br />
Along with the rationalisation <strong>of</strong> fi rst year courses for all engineering programs for the fl exible<br />
fi rst year, a new common course on Engineering Design and Innovation was created. Staff<br />
members from each School met periodically throughout the year to establish the aims and<br />
structure <strong>of</strong> the course, and to write a course outline. The course aims to give students fi rst<br />
hand experience with designing and building creative solutions to problems which are limited<br />
by budget, time and resources. The key engineering skills it aims to develop in students include<br />
concept development, critical thinking and evaluation skills, clear communication, research<br />
and information literacy skills, and teamwork. Dr Jeff Cotter is one <strong>of</strong> the leaders in the<br />
development and implementation <strong>of</strong> the new course, and is teaching the course in 2006.<br />
5.5.3 First Year Course on Sustainable Energy<br />
With the introduction <strong>of</strong> the new fl exible fi rst year program and the opportunity for fi rst year<br />
students to take electives, the Centre decided to <strong>of</strong>fer a new fi rst year course on Sustainable<br />
Energy as part <strong>of</strong> the elective pool. This new course will introduce students to the concept<br />
<strong>of</strong> energy in its different forms through a range <strong>of</strong> demonstrations which will introduce<br />
the concepts <strong>of</strong> energy storage, energy effi ciency, energy conversion and sustainability. An<br />
overview <strong>of</strong> renewable and nuclear energy sources will be given with particular emphasis<br />
on photovoltaic technology including interesting case histories and fascinating mistakes<br />
and disasters. One lecture each week will be given by guest lecturers who are experts from<br />
industry, end-user groups, research and government. The course will fi rst be <strong>of</strong>fered in 2006<br />
to all students enrolled in the Faculty <strong>of</strong> Engineering.<br />
5.5.4 Multiple Criteria Entry for Undergraduate Programs<br />
Generally, admission to undergraduate programs for local Australian citizens and permanent<br />
residents is through the Universities Admissions Centre based solely on a prospective<br />
student’s UAI result as part <strong>of</strong> the Higher School Certifi cate. However, in 2005 the Faculty<br />
<strong>of</strong> Engineering introduced the Multiple Criteria Entry admissions scheme which recognises<br />
that there are a number <strong>of</strong> factors not measured in high school assessment, which indicate<br />
the potential <strong>of</strong> a student to succeed in engineering programs. In addition to their UAI, these<br />
factors include a student’s attitude, commitment, interest in engineering and the type <strong>of</strong><br />
subjects taken for their HSC. Students who anticipated receiving a UAI less than 85, but<br />
higher than 75, were encouraged to apply for an interview with members <strong>of</strong> the Faculty’s<br />
academic staff. At the interview students have the opportunity to exchange information,<br />
demonstrate their suitability to engineering and ask questions <strong>of</strong> Engineering staff.<br />
135
The main round <strong>of</strong> interviews for this admissions scheme was held during the week<br />
<strong>of</strong> the 26th to the 30th September. Five <strong>of</strong> the Centre’s academic staff members<br />
were involved with interviews during this week. In addition to this, one academic<br />
conducted tours <strong>of</strong> our school for prospective photovoltaic and renewable energy<br />
engineering students, and one general staff member coordinated the regional<br />
interviews in Canberra on behalf <strong>of</strong> the Faculty. This program was deemed to be<br />
very successful and will be further refi ned and administered again in 2006.<br />
5.5.5 Semi-distance Delivery Mode Trial<br />
The Centre has recognised that by being a small School, it is limited in the number<br />
<strong>of</strong> courses that it can teach each year. In an attempt at increasing the number <strong>of</strong><br />
yearly course <strong>of</strong>ferings, a second year core course, Sustainable and Renewable<br />
Energy Technologies, was delivered in a semi-distance mode in 2005. This course<br />
was available to students who had previously failed the course, missed taking the<br />
course in the previous year, or were on the Co-Op Scholarship. Lecture material<br />
was delivered via the web, and students attended a one hour tutorial one night a<br />
week, submitted assignments, and sat an examination at the end <strong>of</strong> the session.<br />
Students appreciated the evening classes, and having the opportunity to take the<br />
course in a year when it wasn’t typically <strong>of</strong>fered. The Centre is investigating this<br />
mode <strong>of</strong> delivery with more courses in 2006.<br />
5.6 Educational Resources<br />
5.6.1 <strong>New</strong> Edition <strong>of</strong> Applied Photovoltaics<br />
In 2005, revisions to the Centre’s book titled Applied Photovoltaics were fi nalised<br />
and it was re-printed ready for distribution in 2006. This book was originally<br />
published by Stuart Wenham, Martin Green, and Muriel Watt in 1995, and the<br />
revisions were the result <strong>of</strong> experience teaching a course on applied photovoltaics<br />
as part <strong>of</strong> the Centre’s undergraduate and postgraduate programs. The book<br />
examines the characteristics <strong>of</strong> sunlight, how photovoltaic devices work, and<br />
various applications for the devices including stand-alone photovoltaic systems,<br />
grid-connected photovoltaic systems, remote area power supply, and water<br />
pumping.<br />
5.6.2 Virtual Survivor game<br />
The online Virtual World Solar Challenge game,<br />
developed Jeff Cotter in 1999, is a highly successful<br />
game-based learning tool (http://wsc.pv.unsw.edu.<br />
au/) that has been played over 50,000 times by highschool<br />
students and solar racing car enthusiasts<br />
across Australia and around the world. It has been<br />
used numerous times by the School for Photovoltaic<br />
Engineering and the Faculty <strong>of</strong> Engineering for<br />
promotion and education purposes since 1999. This<br />
successful game-based learning tool grew out <strong>of</strong> a<br />
simple on-paper exercise developed for the general<br />
education course Solar Racing Cars.<br />
136<br />
The Virtual Survivor game-based learning tool comes from similar origins – an on-paper<br />
exercise developed for SOLA2053 - Sustainable and Renewable Energy Engineering. The<br />
on-paper exercise was very popular and engaging with students, and it is presently under<br />
development as a web-based tool.
The main purpose <strong>of</strong> this tool is to expose students to the issues related to sustainable<br />
development, including management <strong>of</strong> dynamic resources, Tragedy <strong>of</strong> the Commons, the<br />
Precautionary Principle, negotiation and policy, and more. When completed, the learning tool<br />
will consist <strong>of</strong> an online, multiplayer simulator and a series <strong>of</strong> supporting activities to enhance<br />
learning in sustainable development.<br />
In short, each <strong>of</strong> many students (as independent agents) inhabits a small Easter-Island-like<br />
island, playing the role <strong>of</strong> Chief <strong>of</strong> a small tribe. They can harvest the natural resources and<br />
expand their individual tribes. They climb and fall in rank ordering among all tribes, based on<br />
population, collected commodities, and the number <strong>of</strong> stone-head statues they construct.<br />
However, they do not know in advance that their environment has limited capacity to sustain<br />
growth and expansion. Occasionally, droughts, famine, confl ict, external trade fl uctuations and<br />
other factors will strain the islands ability to sustain its inhabitants. In this setting, students<br />
will struggle to sustain their virtual society and, especially, learn and debate about issues<br />
confronting their virtual society – resource management, sustainable development, policy<br />
development and more. To make things more interesting, external factors will be applied, for<br />
example, trade fl uctuations, war with neighbouring islands, drought, etc.<br />
The game will eventually feature, among other things:<br />
A Polynesian setting – limited resources, isolation, external factors affect sustainability<br />
Multiplayer mode – each student plays as an independent agent<br />
Dynamic resource models – natural resources behave like dynamic systems<br />
Internal factors – confl ict, competition for resources, communication by in-game email<br />
External factors – such as trade fl uctuations, war, drought, pestilence, famine<br />
Online help<br />
Game history – s<strong>of</strong>tware will “write, record and publish” the history <strong>of</strong> every game played<br />
5.6.3 <strong>New</strong> Outdoor Experimental Area<br />
During 2005, with support from the Faculty <strong>of</strong> Engineering, the Centre gained effective<br />
access to a reserved section <strong>of</strong> new ro<strong>of</strong> over an extension <strong>of</strong> the fourth fl oor <strong>of</strong> the Electrical<br />
Engineering building. This concrete ro<strong>of</strong> area, comprising approximately 425 m 2 , has<br />
been specially engineered to allow attachments without compromising the integrity <strong>of</strong> the<br />
waterpro<strong>of</strong> membrane and is already the home for a 38 m 2 battery room and a mounting<br />
frame for more than 4 kW <strong>of</strong> solar photovoltaic modules. Modules that were previously<br />
mounted on the UNSW property at Little Bay have been moved to the new ro<strong>of</strong> for installation.<br />
That array will be used both as a grid connected solar power station and an experimental<br />
facility and will be operated in collaboration with staff from the School <strong>of</strong> Electrical Engineering<br />
and Telecommunications.<br />
The outdoor area will be used for<br />
students’ experiments and projects that<br />
require solar access and for regular<br />
community outreach activities such as<br />
the Indigenous Australian Engineering<br />
Summer School that is hosted by the<br />
UNSW Faculty <strong>of</strong> Engineering each<br />
January. Work is proceeding on the<br />
design <strong>of</strong> a dedicated experimental jig<br />
to teach about the effects <strong>of</strong> partial<br />
shading <strong>of</strong> arrays and modules and the<br />
effects <strong>of</strong> blocking and bypass diodes in<br />
photovoltaic systems.<br />
A picture <strong>of</strong> this new area on top <strong>of</strong> the Electrical<br />
Engineering Building.<br />
137
Staff and students enjoying the<br />
PVSOC annual dinner.<br />
5.7 PVSOC<br />
One characteristic <strong>of</strong> the Centre<br />
which has led to the success <strong>of</strong> the educational programs is the<br />
friendly atmosphere that is engendered by being a small school<br />
with highly motivated academic and general staff. Students appreciate being able to form<br />
friendships and support networks with fellow students, as well as feeling comfortable and<br />
familiar with academic and administrative staff. PVSOC is a social committee established by<br />
the students which fosters this atmosphere with organised social events and activities to<br />
encourage student interaction. In 2005 the committee organised barbeques, pubcrawls, a<br />
soccer match, student participation in Solar House Day (organised by ANZSES), and fi nally the<br />
PVSOC Annual Dinner. More than 120 students and staff attended this dinner in November<br />
2005 where fi nal year students were farewelled, staff thanked, and the 2006 PVSOC<br />
committee was elected. This dinner epitomised the sense <strong>of</strong> community that exists between<br />
students and staff within the Centre.<br />
5.8 Promotional Activities<br />
The Centre regularly participates in promotional activities<br />
organised by the Faculty <strong>of</strong> Engineering and Student<br />
Recruitment at UNSW. These events are important<br />
for increasing awareness and interest in the Centre’s<br />
educational programs. In 2005 in addition to these annual<br />
events, staff in the School Offi ce also undertook a revision<br />
<strong>of</strong> promotional material for prospective students, and ran<br />
an Australia-wide Virtual World Solar Car Challenge for<br />
high school students.<br />
Two concepts used in the revision <strong>of</strong><br />
promotional material for the Centre’s<br />
educational programs.<br />
5.8.1 Print Media<br />
In order to increase the awareness <strong>of</strong> the Centre’s educational programs, in 2005, the Centre<br />
increased the number <strong>of</strong> advertisements in relevant Sydney Morning Herald publications.<br />
Colour advertisements were placed in the Earth and Environmental Science HSC Study Guide,<br />
the HSC Survival Magazine, the HSC Change <strong>of</strong> Preference Guide, and the HSC Universities<br />
Advisory and Information Day Guide.<br />
5.8.2 Promotional Material<br />
Recognising that the Centre’s printed promotional material was inappropriate for attracting<br />
the attention <strong>of</strong> prospective students in the competitive tertiary marketplace, School<br />
Offi ce staff developed new promotional booklets which contained information pertinent to<br />
undergraduate, postgraduate, and research prospective students. These new colourful and<br />
eye-catching booklets were developed in consultation with current students, and contain<br />
student and graduate pr<strong>of</strong>i les, information on the Centre, its research and educational<br />
activities, and photovoltaic specifi c cartoons developed by a current research student. The<br />
feedback from current and prospective students on this promotional material has been very<br />
positive.<br />
138
5.8.3 UNSW Information Day<br />
Local undergraduate students must apply for admission to UNSW programs through UAC,<br />
and the 5th January 2005 was the last day students could change their preferences for<br />
university degrees. Therefore the university hosted an information day on the 5th January<br />
to assist students obtain information to fi nalise their preferences. Centre administrative<br />
and academic staff attended this event and talked to many prospective students who were<br />
unsure <strong>of</strong> their career direction. Students received information on our programs and had the<br />
opportunity to ask questions <strong>of</strong> staff.<br />
5.8.4 UNSW Courses and Careers Day<br />
UNSW Courses and Careers Day is the annual information day for prospective students<br />
to obtain information about programs and student life at the university. As part <strong>of</strong> this<br />
day academic and administrative staff from the Centre attended information desks in the<br />
Roundhouse to provide advice, and information to prospective students. During the day Dr<br />
Alistair Sproul also presented two lectures about the Centre’s programs as part <strong>of</strong> the<br />
lecture series organised centrally by the university.<br />
In order to increase the awareness <strong>of</strong> photovoltaic<br />
technology the Centre organised another couple<br />
<strong>of</strong> activities on the day. BP Solar was kind enough to<br />
lend the Centre a Solar Powered Ice-Cream Cart for<br />
the day which was used to distribute free ice-blocks to<br />
members <strong>of</strong> the public. Alongside the ice-cream cart<br />
the Centre displayed a solar cooker and a solar-powered<br />
water pump. Students staffed this display and talked<br />
about our educational programs. In addition to this the<br />
Centre, in collaboration with the Energy Management<br />
Unit at UNSW, also conducted tours <strong>of</strong> the inverter<br />
room for the recently installed photovoltaic array on<br />
the Quadrangle Building.<br />
The solar-powered ice-cream cart and solar-powered<br />
water pump display at Courses and Careers Day.<br />
The student-built solar cooker baking pumpkin and<br />
potato using only the sun and engineering.<br />
139
5.8.5 Faculty <strong>of</strong> Engineering Information Day for High School Students<br />
Each year the Faculty <strong>of</strong> Engineering organises an information day to give high school students<br />
an opportunity to learn about engineering and the programs <strong>of</strong>fered at UNSW. As part <strong>of</strong> the<br />
day students visit three engineering schools <strong>of</strong> their choice and engage in interactive activities<br />
aimed at demonstrating the relevant engineering area. In 2005 the demand to attend this<br />
event was so great that two High Schools Days were coordinated, one in April and another<br />
in October. In 2003 59 students attended the Centre, in 2004, 40 attended, however in<br />
2005 over the two days 114 students visited. 26% <strong>of</strong> the students attending the day visited<br />
Photovoltaic Engineering which was an increase from 18% in the previous year. Due to this<br />
large number <strong>of</strong> students, the Centre ran two tours for the day in April and 3 on the day in<br />
October when only one was conducted in prior years. As part <strong>of</strong> these tours students used the<br />
solar-powered water pumps in the Centre’s teaching laboratory, watched the charging and<br />
discharging <strong>of</strong> a capacitor with solar energy, and gained exposure to the Virtual Production<br />
Line, a computer simulation <strong>of</strong> a real life photovoltaic manufacturing line.<br />
Students using the solar-powered water pumping equipment in the Centre’s teaching<br />
laboratory.<br />
5.8.6 Honeywell Engineering Summer School<br />
The Honeywell Engineering Summer School is an event held in December and is conducted by<br />
Engineers Australia. As part <strong>of</strong> the summer school high school students from across NSW and<br />
the ACT about to enter their fi nal year take part in a week <strong>of</strong> activities which involves industry<br />
visits and lectures / demonstrations at a number <strong>of</strong> universities. In 2005 approximately<br />
30 students visited the Centre where they constructed a capacitor using PET bottles, salt<br />
water and aluminium foil under the supervision <strong>of</strong> Mr Rob Largent. At the conclusion <strong>of</strong> Rob’s<br />
demonstration <strong>of</strong> charging and discharging the student-built capacitors, the students were<br />
issued with information on the Centre’s educational programs.<br />
5.8.7 Sunsprint and Minisprint<br />
140<br />
In conjunction with the UNSW Faculty <strong>of</strong> Science, the Centre for Photovoltaic<br />
Engineering co-sponsored the 2005 UNSW Sunsprint Model Solar Car<br />
Challenge. This was the NSW state competition <strong>of</strong> the annual Australian<br />
International Model Solar Challenge and involved 15 schools and nearly<br />
100 high school students. The top teams from this competition went on to<br />
compete for the national title in Adelaide. Participants spent, on average,<br />
four months designing and building their model solar racecar.
For the fi rst time in 2005, UNSW Minisprint<br />
was also held alongside Sunsprint and involved<br />
24 model solar cars. Minisprint is especially<br />
designed for primary school students.<br />
Where a competitive model solar car for the<br />
high school competition takes four months to<br />
design and build, the Minisprint model solar<br />
racecar is bult from raw materials supplied<br />
as a kit and can be assembled by primary school<br />
students, with a bit <strong>of</strong> help, in an afternoon.<br />
The thrill <strong>of</strong> the competition is tremendous for<br />
these students and the feedback from teachers<br />
regarding teaching outcomes and student<br />
motivation shows the effectiveness <strong>of</strong> this<br />
competition.<br />
5.8.8 Virtual World Solar Car Challenge<br />
Every two years a World Solar Car Challenge is held where teams race solar cars from<br />
Darwin to Adelaide. This event was held from the 25th September to the 2nd October 2005<br />
and in association with this the Centre conducted a Virtual World Solar Car Challenge, using<br />
a computer program that had been developed as part <strong>of</strong> the fi rst year teaching curriculum.<br />
As part <strong>of</strong> the Virtual World Solar Car Challenge, students design a virtual solar car within<br />
budgetary constraints and race it in a simulated race from Darwin to Adelaide. This<br />
competition was conducted throughout September and marketed to all high schools across<br />
Australia. Over the month the website experienced 5621 hits from 1440 students from high<br />
schools in every state and territory <strong>of</strong> Australia. The most number <strong>of</strong> entries came from QLD<br />
and WA and the Centre was contacted by several high school teachers seeking permission to<br />
use the game as part <strong>of</strong> their Engineering Studies curriculum. At the end <strong>of</strong> the competition<br />
the Centre presented a winner and runner-up prize, with Dr Corkish attending Normanhurst<br />
Boys High to present the winner Edmund Tse with his plaque and winner’s cheque.<br />
The fl yer sent to every high school around<br />
Australia promoting the Virtual World Solar<br />
Car Challenge.<br />
Edmund Tse receiving his plaque from<br />
Dr Richard Corkish<br />
141
5.9 Educational Collaboration<br />
A range <strong>of</strong> collaborations have been established between the Centre and other educational<br />
institutions and organisations. These collaborations involve the development and<br />
implementation <strong>of</strong> educational programs and courses, the provision <strong>of</strong> support for student<br />
projects and theses, and the exchange <strong>of</strong> students and staff.<br />
5.9.1 BP Solar Sydney<br />
BP Solar manufactures, designs, markets and installs solar systems for residential,<br />
commercial and industrial sectors. The Australian branch <strong>of</strong> BP Solar is located at Sydney<br />
Olympic Park and staff from this company kindly donated the use <strong>of</strong> their Solar Powered Ice-<br />
Cream Cart to the Centre for use on Courses and Careers Day to increase the awareness <strong>of</strong><br />
photovoltaics in a novel manner.<br />
5.9.2 Edwards Hot Water<br />
Edwards Hot Water produces hot water systems for domestic and commercial purposes<br />
using a variety <strong>of</strong> energy sources which includes solar. This organisation has supported the<br />
second year biodiesel trailer over the last few years and provided ongoing advice and support<br />
to students working in the project in 2005.<br />
5.9.3 The Foundation <strong>of</strong> Goodness<br />
Based in tsunami-affected Seenigama, Sri Lanka, The Foundation <strong>of</strong> Goodness is a nongovernment<br />
organisation which assists villagers in Seenigama and neighbouring areas with rebuilding<br />
housing, infrastructure and livelihoods, and providing medical care, welfare, education<br />
and skills development. The organisation has been active for 6 years and students and staff<br />
involved with the second year developing countries project liaised with this organisation in<br />
preparation for travel to Sri Lanka in 2006 to install photovoltaic powered water pumping<br />
and water fi ltration systems.<br />
5.9.4 Fowlers Gap UNSW Arid Zone Research Station<br />
In 2005 a fourth year thesis student collaborated with staff<br />
at UNSW’s Fowlers Gap Research Station, 110 kilometres<br />
north <strong>of</strong> Broken Hill in the state’s far west. The aim <strong>of</strong> the<br />
project was to test and maintain an existing photovoltaic<br />
pumping system, and to design and install a new water<br />
pumping system suitable for stock watering. The student’s<br />
work was deemed a success in that repairs she made to<br />
the array increased output by more than 33 percent which<br />
provided the homestead with suffi cient water and hence a<br />
new and expensive array was not required.<br />
Shelley Bambrook at Fowlers Gap Station<br />
5.9.5 Kyocera Solar<br />
Kyocera Solar produces and supplies solar energy products and in 2005 donated mounting<br />
frames for the fourth year thesis conducted in collaboration with the UNSW Fowler’s Gap<br />
Arid Zone Research Station.<br />
142
5.9.6 Murdoch <strong>University</strong> Western Australia<br />
Scientia Pr<strong>of</strong>essor Wenham was awarded an adjunct appointment at Murdoch <strong>University</strong> to<br />
help the university implement a new undergraduate degree in Renewable Energy Engineering<br />
which commenced in 2001. Pr<strong>of</strong> Wenham still holds this appointment and is also a member <strong>of</strong><br />
the Murdoch <strong>University</strong> Renewable Energy Engineering Board. This collaboration is important<br />
for the exchange <strong>of</strong> educational material for both institutions.<br />
5.9.7 Queensland Sustainable Energy Industry Development Group (QSEIDG)<br />
The Queensland Sustainable Energy Industry Development Group was formed in 2004 to<br />
enhance collaboration between government, education, industry, and community organisations<br />
involved in sustainable energy and energy effi cient projects in Queensland. This organisation<br />
has developed mobile training facilities to raise awareness <strong>of</strong> sustainable energy. Each trailer<br />
has working examples <strong>of</strong> photovoltaic systems, wind energy systems, solar hot water heating<br />
and water pumping. The organisation has allowed the Centre to use their design for this<br />
trailer to build our own mobile display unit suitable for the annual Courses and Careers Day.<br />
5.9.8 Randwick City Council<br />
The Centre is in collaboration with the Randwick City Council on a fourth year thesis which<br />
is investigating the potential application <strong>of</strong> high-pr<strong>of</strong>i le, public renewable energy projects in<br />
the Randwick City Council area. Feasibility studies are to be conducted as part <strong>of</strong> this project<br />
which commenced in July 2005 and will be completed mid-2006.<br />
5.9.9 Suntech Power Corporation<br />
Suntech, a partly Australian-owned company operating in China, has been actively involved<br />
in assisting UNSW with the development, testing and evaluation <strong>of</strong> educational material in<br />
the PV area such as in the development, testing and evaluation <strong>of</strong> a s<strong>of</strong>tware package called<br />
the Virtual Production Line. Signifi cant numbers <strong>of</strong> UNSW undergraduate students also<br />
participate in industrial training at Suntech. These students fi nd their time in China an extremely<br />
valuable educational and cultural experience. This collaboration has also resulted in members<br />
<strong>of</strong> Suntech visiting the Centre and obtaining experience in the Centre’s laboratories.<br />
143
Dean <strong>of</strong> Engineering<br />
Other Schools<br />
School <strong>of</strong> Photovoltaic and Renewable<br />
Energy Engineering<br />
Head: Dr Richard Corkish<br />
School Administration<br />
Trichelle Burns<br />
Centre Excellence in<br />
Advanced Silicon Photovoltaics<br />
and Photonics<br />
Undergraduate<br />
Education<br />
Advisory Committee<br />
Director: Pr<strong>of</strong>. Stuart Wenham<br />
Res. Director: Pr<strong>of</strong> Martin Green<br />
Management Committee<br />
Administrative<br />
Support<br />
Jenny Hansen<br />
Research Programs<br />
Business and<br />
Operations Support<br />
Mark Silver<br />
Photovoltaics<br />
and Solar Energy<br />
Renewable<br />
Energy<br />
Engineering<br />
Postgraduate Education<br />
- Coursework and Research<br />
Dr Jeff Cotter & Dr Alistair Sproul<br />
Wafer-Based Projects<br />
Manager: Dr Jeff Cotter<br />
Silicon Thin Films<br />
Manager: A/Pr<strong>of</strong>. Armin Aberle<br />
Third Generation<br />
Manager: Dr Gavin Conibeer<br />
Silicon Photonics<br />
Manager: A/Pr<strong>of</strong>. Jianhua Zhao<br />
144
The Management structure <strong>of</strong> the Centre and how it fi ts into the <strong>University</strong> system is shown<br />
below. When the Key Centre for Photovoltaic Engineering was awarded in 1999, it was<br />
established as an autonomous Centre within the Faculty <strong>of</strong> Engineering <strong>of</strong> UNSW, becoming<br />
independent from the School <strong>of</strong> Electrical Engineering within which the photovoltaic activities<br />
had been previously located. At that time, the Key Centre was given the same operational<br />
independence and rights as the other Engineering schools, with the Director Stuart Wenham<br />
being given the same status and authority as other Heads <strong>of</strong> Schools. With the awarding <strong>of</strong><br />
the Centre <strong>of</strong> Excellence in 2003, the <strong>University</strong> made the commitment that the Key Centre<br />
for Photovoltaic Engineering would become the School <strong>of</strong> Photovoltaic Engineering, within<br />
which the ARC Photovoltaics Centre <strong>of</strong> Excellence would be located. This transition occurred<br />
in January 2006.<br />
Within the Centre <strong>of</strong> Excellence, Deputy Directors have been appointed for each <strong>of</strong> the major<br />
strands <strong>of</strong> research as shown. Each Deputy Director is <strong>of</strong> the highest calibre, and would<br />
make worthy Directors or Research Directors in almost any major research Centre in this<br />
fi eld internationally. In fact when establishing the Centre <strong>of</strong> Excellence, it was noted that three<br />
<strong>of</strong> the four Deputy Directors, along with the Centre’s Director and Research Director, have<br />
all held prior positions as Directors or Deputy Directors <strong>of</strong> ARC Special Research Centres.<br />
All four Deputy Directors were originally from overseas (USA, Germany, China and the UK),<br />
being attracted to Australia by the opportunity <strong>of</strong> being involved in the photovoltaic research<br />
activities at UNSW.<br />
The Management Committee <strong>of</strong> the Centre comprises the six Directors and Deputy Directors,<br />
along with the Head <strong>of</strong> School, the Head <strong>of</strong> Characterisation and the Business and Operations<br />
Manager. This committee meets fortnightly on the 1st and 3rd Fridays <strong>of</strong> each month, with<br />
each Deputy Director giving a report on the activities in his or her area over the preceding<br />
fortnight. This committee takes responsibility for decision making within the Centre that<br />
affects the Centre as a whole, while the individual Deputy Directors receive their own annual<br />
budgets to allow them to make and implement decisions that impact only their own laboratory<br />
areas and research activities.<br />
The Advisory Committee for the Centre comprises the Centre Directors, the Head <strong>of</strong> School<br />
for Photovoltaic Engineering, leading academics from other institutions, industry leaders such<br />
as CEO’s <strong>of</strong> various companies involved in the fi eld, and research leaders. This committee<br />
provides high level advice, feedback and recommendations in relation to the Centre’s<br />
activities and their relevance. The Advisory committee meets annually, taking into account the<br />
geographical separation <strong>of</strong> its members (Australia, USA, China, Italy, Germany), although more<br />
frequent correspondence takes place when necessary with individual committee members<br />
either by email or telephone. The membership <strong>of</strong> the Advisory Committee includes:<br />
145
• Pr<strong>of</strong>essor Stuart Wenham, Centre Director<br />
• Pr<strong>of</strong>essor Martin Green, Centre Executive Research Director<br />
• Associate Pr<strong>of</strong>essor Armin Aberle, Deputy Director <strong>of</strong> the Centre (2nd Generation<br />
PV research)<br />
• Associate Pr<strong>of</strong>essor Jianhua Zhao, Deputy Director <strong>of</strong> the Centre (Photonics)<br />
• Dr Jeff Cotter, Deputy Director <strong>of</strong> the Centre (1st Generation PV Research)<br />
• Dr Gavin Conibeer, Deputy Director <strong>of</strong> the Centre (3rd Generation PV Research)<br />
• Pr<strong>of</strong>essor Allen Barnett: world leading academic in the fi eld; recipient <strong>of</strong> many<br />
prestigious international prizes such as the William Cherry Award; former<br />
President <strong>of</strong> AstroPower, one <strong>of</strong> the world’s largest solar cell manufacturers,<br />
before being purchased by GE.<br />
• Dr Zhengrong Shi: CEO <strong>of</strong> Suntech-Power, another <strong>of</strong> the world’s largest solar cell<br />
manufacturers; recipient <strong>of</strong> various industry prizes; former Deputy Research<br />
Director <strong>of</strong> Pacifi c Solar; major collaborator <strong>of</strong> the Centre.<br />
• David Hogg: CEO <strong>of</strong> CSG Solar AC, Germany; former CEO <strong>of</strong> Pacifi c Solar, Australia;<br />
• Mr. David Jordan, Manager <strong>of</strong> <strong>New</strong> Technology, BP Solar International;<br />
• Pr<strong>of</strong>. Andres Cuevas, ANU;<br />
• Pr<strong>of</strong>. Peter Würfel, Institut fur Angewandte Physik der Universität Karlsruhe;<br />
• Dr Francesca Ferrazza, Chief Scientifi c Offi cer, Enitechnologie.<br />
Also within the Management structure and a member <strong>of</strong> the Centre’s Management Committee<br />
is the Centre’s Business, Technology and Operations Manager Mark Silver, given the extensive<br />
involvement <strong>of</strong> Centre staff with industry, large number <strong>of</strong> collaborative research projects,<br />
and the high level <strong>of</strong> success <strong>of</strong> the Centre in generating, marketing and commercialising<br />
technology. The Head <strong>of</strong> Administrative Support for the Centre is Jenny Hansen. Not shown<br />
in the Management Structure is the Financial Offi cer Julie Kwan who looks after the Centre<br />
Accounts, interfaces with the <strong>University</strong> Financial System, advises the Centre Directors and<br />
Deputy Directors on budgeting and fi nancial matters, and generally assists with sourcing<br />
<strong>of</strong> funding, purchasing, reporting, writing grant applications, and advising on business<br />
opportunities.<br />
A primary responsibility <strong>of</strong> the Director in the Centre is to ensure the Management structure<br />
as shown functions effi ciently and effectively, with appropriate support and resources to<br />
facilitate the Centre’s achievement <strong>of</strong> its milestones and performance targets. An important<br />
aspect <strong>of</strong> this is managing the Centre’s fi nances to ensure suitable levels <strong>of</strong> funding are<br />
made available for supporting the four major research strands, while simultaneously<br />
ensuring adequate funding is available for repair and maintenance <strong>of</strong> existing facilities and<br />
the purchasing <strong>of</strong> appropriate new infrastructure and equipment. Another important aspect<br />
relates to how the Centre has been assimilated and functions within the university system, and<br />
how the Centre interacts with industry, other research and teaching institutions, the various<br />
levels <strong>of</strong> government and the broader community. The Director also has a signifi cant role in<br />
the appointment <strong>of</strong> new staff, the establishment and operation <strong>of</strong> the Centre’s Management<br />
Committee, and the establishment and operation <strong>of</strong> the Advisory Committee to facilitate<br />
quality feedback and guidance for the Centre’s Management Committee.<br />
The Director works closely with the Centre’s Executive Research Director to assist in setting<br />
high level research direction and priorities for the Centre’s programs. In the teaching area,<br />
with the Key Centre for Photovoltaic Engineering and its activities being incorporated into<br />
the Centre <strong>of</strong> Excellence, the Director also takes signifi cant responsibility in ensuring the<br />
successful development and implementation <strong>of</strong> the Centre’s educational programs, both<br />
at undergraduate and postgraduate levels. In this area, the Director works closely with Dr<br />
Richard Corkish, the Head <strong>of</strong> School.<br />
146
7.1 PROGRESS AGAINST CENTRE DESIGNATED MILESTONES<br />
In its initial application, the Centre proposed 18 challenging research milestones for its<br />
5-year program, as well as 4 teaching milestones, 4 “linkage” milestones and 5 governance<br />
milestones. Progress towards these milestones is reported below.<br />
Research Milestones<br />
A.1<br />
Efficiency improvement <strong>of</strong> standard p-type silicon cells to 26% or n-type cell to 24%.<br />
With the growing industry interest in n-type wafers for solar cell manufacturing, the Centre<br />
has focused on developing its highest performance solar cells to suit this type <strong>of</strong> wafer.<br />
Excellent progress has been made, particularly in recent months, with the Centre equalling<br />
the world record <strong>of</strong> 22.7% effi ciency for this type <strong>of</strong> wafer, held for more than a decade by<br />
Stanford <strong>University</strong>.<br />
Importantly, the Centre’s new high effi ciency designs for the n-type wafers are still far from<br />
optimised with the expectation that the Centre will set more world records in this area during<br />
the next couple <strong>of</strong> years, thereby establishing a clear lead over competitors internationally.<br />
A.2<br />
Efficiency improvement <strong>of</strong> Si cell boosted by an up- or down- converter to 28%.<br />
This milestone is linked to A10. An increase in spectral response in the range 1480 to<br />
1580nm (i.e. well below the band edge <strong>of</strong> Si) for a bifacial Si cell has beendemonstrated<br />
representing a very small increase in current, which nonetheless represents a pro<strong>of</strong> <strong>of</strong><br />
concept.<br />
Further work on increasing the sensitivity to light between 1250 & 1480nm which should in<br />
turn increase this up-conversion current, is being carried using two approaches (see Section<br />
4.5.5). PbSe quantum dots absorb photons between 1250 & 1480nm, which then undergo<br />
non-radiative transitions and are re-emitted at longer wavelengths, inside the absorption<br />
range <strong>of</strong> the erbium (Er) in the Up-conversion phosphors. Alternatively, addition <strong>of</strong> dysprosium<br />
(Dy) to the erbium.doped phosphors, provides energy levels for absorption <strong>of</strong> photons between<br />
1250 & 1480nm. Again these decay non-radiatively in the Dy atoms to be re-emitted within<br />
the Er absorption band.<br />
A patent application has been fi led on a new up-conversion approach.<br />
A.3<br />
Demonstration <strong>of</strong> large-area buried contact silicon solar cell <strong>of</strong> efficiency above 21% and/or<br />
multicrystalline buried contact solar cell above 18%.<br />
The link between boron diffusion induced dislocations and a key recombination mechanism<br />
limiting solar cell performance for p-type buried contact silicon solar cells has been<br />
identifi ed.<br />
The boron diffusion induced dislocation has been identifi ed as one <strong>of</strong> the major effi ciency<br />
limitations <strong>of</strong> double-sided buried contact solar cells, which have now reached effi ciencies as<br />
high as 18.6% on FZ(Boron) and 18.6% on Cz(Ga) despite this limitation.<br />
The signifi cance <strong>of</strong> defects in n-type and p-type silicon wafers using side by side comparison <strong>of</strong><br />
similar n-type and p-type solar cells has been demonstrated.<br />
We have achieved cell effi ciencies as high as 19.2% on textured FZ(Phosphorus), and<br />
performance levels almost as high on Cz(Phosphorus) n-type wafers.<br />
147
A.4<br />
A.4.1<br />
Demonstration <strong>of</strong> at least three new promising technologies for use in conjunction<br />
with commercial buried contact solar cells.<br />
Stencil printed solar cells<br />
We have developed and demonstrated single and double level stencils for the printing <strong>of</strong> fi neline<br />
metallisation lines on silicon solar cells, which can be used to print the metallisation for<br />
buried contact solar cells.<br />
We have demonstrated precision alignment <strong>of</strong> both the stencil and the surface grooves <strong>of</strong><br />
a selectively diffused emitter for a buried contact solar cell. This type <strong>of</strong> buried contact solar<br />
cell captures many <strong>of</strong> the benefi ts <strong>of</strong> printed solar cells yet has the potential to improve both<br />
current and voltage <strong>of</strong> these cells.<br />
We have already demonstrated solar cell effi ciencies <strong>of</strong> over 16% with the stencil print<br />
technology, comparable with the best commercially manufactured screen-printed solar cells<br />
fabricated using the conventual structure (typically 16.0-16.5% effi ciency).<br />
A.4.2<br />
Hybrid buried contact/screen-printed solar cells<br />
An innovative emitter structure for screen-printed solar cells incorporating semiconductor<br />
fi ngers, has been developed in conjunction with Suntech-Power company and is described in<br />
the section on industry collaborative research.<br />
Improvements during 2005 have increased effi ciencies from about 15% in 2004 to above<br />
18% effi ciency, now well ahead <strong>of</strong> conventional screen-printed solar cells that are typically<br />
16-16.5% effi ciency.<br />
This new design addresses the fundamental limitations <strong>of</strong> screen-printed metallisation<br />
schemes that have limited the performance <strong>of</strong> screen-printed solar cells for more than 30<br />
years. The approach captures the performance advantages <strong>of</strong> buried-contact solar cells with<br />
a selective emitter, low top surface shading losses, and a well passivated and lightly diffused<br />
top surface, while simultaneously retaining the simplicity and manufacturability <strong>of</strong> screenprinted<br />
solar cells including the use <strong>of</strong> all the same equipment and infrastructure.<br />
A.4.3<br />
Micro-contact Silicon Solar Cell<br />
An improved rear surface design for the buried contact solar cell that minimises the electrode<br />
diffusion/contact area, has been developed and demonstrated. This new technology appears<br />
to have the potential for high effi ciency.<br />
Test devices have demonstrated a large reduction in the recombination associated compared<br />
to the standard laser-groove contact used in the buried contact process, with open-circuit<br />
voltages approaching 685 mV achieved. This indicates the potential for achieving high<br />
effi ciency once the remainder <strong>of</strong> the structure is optimised to minimise optical and parasitic<br />
resistive losses.<br />
A.5<br />
A.5 Demonstration <strong>of</strong> optical pathlength enhancement > 30 in a silicon thin-film.<br />
Recent collaborative work with CSG Solar has resulted in optical pathlength enhancement<br />
above 20 in the thin crystalline layers deposited onto textured glass. Further improvement is<br />
expected in the near future.<br />
148
In parallel, the AIT glass texturing method (AIT = Aluminium-Induced Texture) is being<br />
developed at UNSW. Patent protection has been sought for the AIT method (Glass texturing,<br />
International PCT patent application PCT/AU2004/000339, priority date 7/04/2003).<br />
A.6<br />
Demonstration <strong>of</strong> series interconnection <strong>of</strong> silicon thin-film solar cells without the use<br />
<strong>of</strong> metal interconnects and improved durability and performance for such devices.<br />
Good progress has been made in developing new cell designs that minimise the requirement<br />
for metal contacts. Conventional cell designs require a metal grid on the light receiving<br />
surface and then another metal contact on the opposite surface, to collect current and<br />
conduct it parallel to the cell surface to a point or region where the current is subsequently<br />
transported to the adjacent solar cell. The new cell design completely alleviates the need for<br />
such metal, with the only remaining metal being used to conduct the generated current a<br />
short distance <strong>of</strong> only 40 microns across the isolation region produced by the laser between<br />
series connected cells. This cell interconnection approach is achieved with the SAMPL, a<br />
self-aligned maskless photolithography method. Patent protection has been sought for this<br />
cell interconnection method (Thin-fi lm solar cell inter¬connection, International PCT patent<br />
application PCT/AU2005/000734, priority date 4/06/2004).<br />
The latest cell designs will aim to eliminate the fi nal metal components by utilising doped<br />
semiconductor material to conduct the current the short distance <strong>of</strong> 40 microns between<br />
juxtaposed cells.<br />
A.7<br />
Demonstration <strong>of</strong> 11% efficient silicon thin-film cell deposited onto glass.<br />
Work towards this goal progresses along two paths. One is the involvement <strong>of</strong> senior Centre<br />
staff as Consultants for UNSW’s spin-<strong>of</strong>f company CSG Solar Pty Ltd (formerly Pacifi c Solar<br />
Pty Ltd), aiming at the improvement <strong>of</strong> the company’s CSG (crystalline silicon on glass)<br />
technology from 8.2% effi ciency in 2003 to 11% by the end <strong>of</strong> 2007. Good progress is being<br />
made through this collaboration, with the best CSG Solar modules now having effi ciencies <strong>of</strong><br />
over 9%.<br />
The second path is the independent research performed in the Centre’s Thin-Film Group. As<br />
outlined in Section 4.4, three novel poly-Si thin-fi lm solar cells on glass are being developed<br />
(EVA, ALICIA, ALICE). The focus in 2005 was on improving the voltages and the quantum<br />
effi ciency <strong>of</strong> these cells, and on making functioning mini-modules. Good progress has been<br />
made with all three cell technologies, with open-circuit voltages now in the range 440-517<br />
mV and internal quantum effi ciencies <strong>of</strong> up to 70%. The 517-mV result is, to our knowledge,<br />
a new world record for homojunction poly-Si thin-fi lm solar cells on glass. Simple solar cells<br />
(“Mesa cells”) with effi ciencies <strong>of</strong> up to 3% have so far been made from these novel solar<br />
cell materials. First functioning mini-modules have also been realised. The present research<br />
focuses on boosting the short-circuit current <strong>of</strong> these cells to above 20 mA/cm2 and the<br />
cell effi ciency to over 5%.<br />
A.7<br />
Demonstration <strong>of</strong> 10% photoluminescent efficiency in silicon and above 98% in GaAs,<br />
sufficient to produce device refrigeration.<br />
The fi rst part <strong>of</strong> this milestone was achieved in 2003. In photoluminescence (PL) testing<br />
using oxide passivated FZ silicon wafers with textured front surfaces, we have demonstrated<br />
EQE (external quantum effi ciency) up to 10.2% at T = 130K and 6.1% at room temperature.<br />
[T. Trupke, J. Zhao, A. Wang, R. Corkish and M.A. Green, Appl. Phys. Lett. 82, 2996, (2003)].<br />
The IQE (internal quantum effi ciency) is also estimated to be over 20%. The results <strong>of</strong> this<br />
work were also used to measure the radiative recombination coeffi cient B(T) [T. Trupke et<br />
al. J.Appl.Phys. 94, 4930 (2003)]. Further improvement in the PL effi ciency is expected<br />
from thinning substrates with textured surfaces, and improving the light-trapping surface<br />
structures.<br />
149
Progress towards the second milestone has been good with 96% effi ciency for GaAs devices<br />
demonstrated in the Centre’s fi rst year <strong>of</strong> operation (Section 5.5 <strong>of</strong> 2003 report). PL EQE<br />
measurements were carried out on an 800 nm sample <strong>of</strong> GaAs, sandwiched in a double GaInP<br />
heterojunction structure so that carriers are confi ned in the GaAs. Two measurements were<br />
used: a calibrated PL technique which gave 90% EQE for planar samples and a combined<br />
thermal/PL measurement that gave 92% ± 3% EQE (a reasonable agreement). Further<br />
calibrated PL measurements with the ZnSe dome to couple light out <strong>of</strong> the sample gave an<br />
increase from 90% to 96% EQE.<br />
Further progress predicted to increase the EQE towards the 98% for GaAs was achieved in<br />
2004, with development <strong>of</strong> the texturing technique required for the back surface. Modelling in<br />
2004 showed that this would bring the 96% with a ZnSe dome to a 97.5% external quantum<br />
effi ciency – remarkably close to the 98% fi gure calculated to be required for PL cooling.<br />
As the primary aim <strong>of</strong> this task was to improve insight into the physics and technology <strong>of</strong><br />
improved radiative emission, rather than to achieve radiative cooling itself, no further work<br />
has been carried out or is planned in this milestone area.<br />
A.9<br />
Efficiency improvement for a silicon based tandem cell over a single cell baseline.<br />
Signifi cant progress has been made in demonstrating an enhanced band gap material for<br />
the top cell for a Si based tandem (see Section 4.5.2). Both Si/SiO2 quantum-well (QW)<br />
and quantum-dot (QD) structures have been fabricated. These have also been grown in the<br />
multilayer structures that will be required for the superlattice structures <strong>of</strong> an enhanced<br />
band-gap material.<br />
Photoluminescence (PL) at enhanced energies has been demonstrated. The enhancement<br />
in PL energy is in general agreement with quantum confi nement calculations based on the<br />
dimensions for QW and QD structures observed in TEM (within the approximately 20%<br />
errors in the size measurements). X-ray diffraction and work on the artefacts present in TEM<br />
<strong>of</strong> such small structures is improving the quantitative validity <strong>of</strong> these calculations.<br />
A PL energy at 1.7 eV has been observed for 1-2 nm QDs; this being the ideal band gap for<br />
an upper cell material on Si. Further evidence for quantum confi nement in the Si QDs comes<br />
from the greatly enhanced PL intensity indicative <strong>of</strong> localisation in the QDs.<br />
This work includes initial measurement <strong>of</strong> the electrical properties, with promising, although<br />
still high, resistivities. It has been demonstrated that these can be reduced to about 103<br />
Ω.cm with hydrogen passivation. Further information on resistivity variation with temperature<br />
has been obtained in 2005 which is giving an indication <strong>of</strong> the activation mechanisms.<br />
Furthermore, there has been success with the fabrication <strong>of</strong> the analogous structure <strong>of</strong><br />
Si QDs in Si3N 4 , with TEM evidence <strong>of</strong> nano-crystals by both sputtering and PECVD. The<br />
advantage <strong>of</strong> the nitride matrix is that the lower barrier height should improve conductivity<br />
for a given QD density, although optical and electrical properties are still being tested. PL<br />
evidence for enhanced energy levels in these Si QDs in Si3N4 matrix has been achieved with<br />
preliminary evidence for higher energies than for similar sized dots in oxide – although this<br />
remains to be confi rmed.<br />
This technique is now starting to be applied to Si QDs in silicon carbide (SiC). SiC having an<br />
even lower barrier height and hence expected higher tunnelling probability between QDs and<br />
hence higher conductivity.<br />
150
A.10<br />
Demonstration <strong>of</strong> photon down-conversion quantum efficiency above 50% and/or<br />
up-conversion efficiency above 10% and demonstration <strong>of</strong> performance improvement<br />
in a baseline cell by either approach.<br />
Work to date has concentrated on up-conversion (see Section 4.5). For a luminescent<br />
phosphor, NaYF4: 20% Er 3+ supplied by the <strong>University</strong> <strong>of</strong> Bern, an external quantum effi ciency<br />
(EQE) <strong>of</strong> more than 3.5% has been measured. Furthermore small improvements in below<br />
band gap photoresponse have been measured for bifacial cells with luminescent phosphors<br />
on the underside. Although only very small current increases are observed, these are very<br />
promising results for non-optimised materials. A baseline test cell has been established using<br />
bifacial cells for up-conversion comparative tests.<br />
Properties <strong>of</strong> down-conversion luminescent materials have been surveyed in the literature<br />
and form the subject <strong>of</strong> a review paper by authors from the Centre [Richards BS, Solar<br />
Energy Materials (2005) submitted]. For some mechanisms these can be up to 1.2 EQE,<br />
i.e., 1.2 low energy photons out for one high energy photon in, and hence useful for effi ciency<br />
increase. However, in such materials there are many other mechanisms operating in parallel<br />
which bring the overall EQE to about 0.9. Nonetheless these fi gures are promisingly close to<br />
an EQE <strong>of</strong> 1, i.e., 50% <strong>of</strong> the ideal performance and hence close to the threshold useful for<br />
increased effi ciency from down-conversion.<br />
Ideal quantum effi ciency for a down-conversion process is 2, i.e., two low energy photons out<br />
(above the cell band gap) for each high energy photon in. The milestone refers to achieving<br />
50% <strong>of</strong> this ideal performance, this being the effi ciency above which real gains in PV effi ciency<br />
are achieved.<br />
A.11<br />
Demonstration <strong>of</strong> two key elements required for hot carrier cell implementation:<br />
energy selective contacts and enhancement <strong>of</strong> radiative recombination rates relative<br />
to relaxation rates <strong>of</strong> photogenerated carriers.<br />
The luminescence from QDs mentioned under Milestone A.9, is consistent with enhanced<br />
band gap emission from QD superlattices. There has also been experimental evidence <strong>of</strong><br />
resonant tunnelling in these structures. The radiative effi ciency improvements for GaAs<br />
mentioned under Milestone A.8 are also relevant for improved understanding <strong>of</strong> hot carrier<br />
absorbers.<br />
Experimental progress has been made towards selective energy contacts. This has been<br />
achieved with the demonstration <strong>of</strong> negative differential resistance (NDR) in samples<br />
containing a single layer <strong>of</strong> quantum dots sandwiched between two oxide layers (see Fig. 4.102<br />
in 2004 Annual <strong>Report</strong>). This indicates resonance in the tunnelling current across the two<br />
oxide barriers – this being a necessary pre-requisite for selective energy contacts. Further<br />
work on improving the quality factor <strong>of</strong> this NDR response is in progress. In addition. in 2005,<br />
characterisation <strong>of</strong> these structures has been improved with two additional techniques.<br />
Firstly conductive atomic force microscopy (CAFM) in collaboration with Melbourne <strong>University</strong>,<br />
gives much greater spatial resolution <strong>of</strong> resonant tunnelling centres (Section 4.5.6.1.1). The<br />
technique requires thinner oxide layers than at present but has already demonstrated detailed<br />
mapping <strong>of</strong> a resonant tunnelling structure. Secondly, a technique <strong>of</strong> optically assisted I-V has<br />
been established, which uses optical excitation <strong>of</strong> carriers retaher than a bias fi eld (Section<br />
4.5.6.1.2). This has already demonstrated more direct evidence for hot carrier generation<br />
and collection.<br />
151
The theory <strong>of</strong> hot carrier absorbers has been extended to include a greater understanding <strong>of</strong><br />
phononic transport and a recognition <strong>of</strong> similarities to the required properties <strong>of</strong> thermoelectric<br />
materials (see “Hot Carrier Absorbers” in Section 5.5 in 2003 Annual <strong>Report</strong>). In particular<br />
the specifi c mechanisms which can slow cooling by enhancing the “phonon bottleneck effect”<br />
in QW and QD superlattices are being identifi ed. Experimental evidence <strong>of</strong> such reduced<br />
cooling in QW superlattices has been identifi ed in the literature. (See section “Hot Carrier<br />
Absorbers” in section 4.5 in 2004 Annual <strong>Report</strong>). This work has been extended further in<br />
2005 with modelling <strong>of</strong> the specifi c phonon mechanisms and quantum dot conditions that<br />
can give an enhanced ‘phonon bottleneck effect’ in QD nanostructured materials.<br />
A.12<br />
Demonstration <strong>of</strong> the quantum confined Stark effect in silicon as the basis <strong>of</strong> a highspeed<br />
silicon modulator).<br />
Devices <strong>of</strong> an appropriate structure were fabricated during 2003, but silicon layers needed<br />
to be thinner to demonstrate this effect. Progress was made in 2004 and 2005 in reducing<br />
the thickness <strong>of</strong> these layers, that are now close to the desired value.<br />
A.13<br />
Evaluation <strong>of</strong> the optical constants <strong>of</strong> quantum confined silicon SOI devices as a<br />
function <strong>of</strong> layer thickness.<br />
Relevant data have been reported in section 5.5 <strong>of</strong> the 2003 Annual <strong>Report</strong> and section 4.6<br />
<strong>of</strong> the 2004 report. Photoluminescence properties have been characterised. Conversion<br />
to optical constants requires accurate knowledge <strong>of</strong> the dimensions <strong>of</strong> the confi ned material.<br />
Measurement <strong>of</strong> these produces some experimental challenges, presently being addressed.<br />
The most accurate method seems to be by identifying photoluminescence features<br />
corresponding to regions differing by the presence <strong>of</strong> one additional atomic plane in the<br />
region under test. However, fi nding an independent technique that is able to confi rm this<br />
interpretation <strong>of</strong> the data has provided a challenge.<br />
Other work quantifying the evidence for QDs and their dimensions is continuing. This includes<br />
analysis <strong>of</strong> HRTEM, high resolution SIMS and synchrotron XRD.<br />
A.14<br />
Improvement <strong>of</strong> light emission efficiency <strong>of</strong> silicon diodes above levels already<br />
demonstrated, to maintain Australian leadership and as a test bed for integrated<br />
devices.<br />
Australia retains leadership in the bulk silicon luminescence fi eld, although efforts during the<br />
last 12 months have concentrated on integrated devices.<br />
A.15<br />
Demonstration <strong>of</strong> high light emission efficiency in silicon LEDs integrated into<br />
microelectronic chips.<br />
Effi ciencies <strong>of</strong> 2x10 -6 have been demonstrated, although further improvements are expected<br />
as devices become thinner.<br />
A.16<br />
Demonstration <strong>of</strong> high speed modulation <strong>of</strong> output <strong>of</strong> silicon light emitting diodes.<br />
Modulation has been demonstrated, but not yet at high speed.<br />
152
A.17<br />
Demonstration <strong>of</strong> operational microelectronic circuits communicating at high data<br />
rates using silicon light emission and detection.<br />
This is a target for the fi nal year <strong>of</strong> the program. To date, we have demonstrated the targeted<br />
communication between bulk silicon devices, probably for the fi rst time, but data rates<br />
are slow, as expected. Test structures for demonstrating this at the chip level have been<br />
designed.<br />
A.18<br />
Investigate the feasibility <strong>of</strong> obtaining lasing action in both bulk and quantum-confined<br />
silicon and, if feasible, demonstrate the first laser in these materials.<br />
Several new mechanisms for demonstrating lasing in bulk silicon continue to be investigated.<br />
Teaching Milestones<br />
C.1<br />
At least 50 honours students completing their degrees by 2007<br />
The Centre <strong>of</strong> Excellence has performed particularly well in the educational areas with 31<br />
students having completed their degrees and graduated with honours in the last 2 years.<br />
The Centre has already achieved, throughout its duration, an average <strong>of</strong> more than the<br />
required 10 honours graduates per year, despite the time-lag associated with producing the<br />
fi rst graduates.<br />
Based on the performance and the numbers <strong>of</strong> students enrolled in the Centre’s degree<br />
programs, there will be a further 30-40 students graduating with honours by the end <strong>of</strong><br />
2007. This will therefore give a total <strong>of</strong> 60-70 students graduating with honours, exceeding<br />
the Centre’s target number <strong>of</strong> 50.<br />
C.2<br />
At least 20 new postgraduate students involved by 2007<br />
Nineteen new postgraduate research students have enrolled in the Centre since 2003,<br />
making the Centre well on track towards achieving this milestone.<br />
Importantly, the rate <strong>of</strong> enrolments have been steadily increasing since the establishment<br />
<strong>of</strong> the ARC Photovoltaics Centre <strong>of</strong> Excellence with 4 new postgraduate research students<br />
enrolling in 2003, 7 in 2004, followed by 8 in 2005.<br />
C.3<br />
The involvement <strong>of</strong> at least 10 postdoctoral research fellows by 2007<br />
The Centre <strong>of</strong> Excellence has already achieved this 5-year target with the involvement <strong>of</strong> 15<br />
postdoctoral research fellows since 2003. These include Dr. Robert Bardos, Dr. Andrew<br />
Brown, Dr. Kylie Catchpole, Dr. Eun-Chel Cho, Dr. Gavin Conibeer, Dr. Didier Debuf (Adjunct<br />
Fellow), Dr. Tammy Humphrey, Dr. Dirk Kőnig, Dr. Kuo-lung Lin, Dr. Bryce Richards, Dr. James<br />
Xia. Dr. Per Widenborg, Dr. Thorsten Trupke , Dr.Anita Ho and Dr Peter Cousins.<br />
C.4<br />
The development <strong>of</strong> at least 4 courses on photovoltaics suitable for internet-based<br />
delivery with at least 1 such internet-based course <strong>of</strong>fered each year<br />
Two courses have now been developed for internet delivery, with both being run during session<br />
2, 2005.<br />
153
The fi rst course, based on the postgraduate course Photovoltaics SOLA9001, has been<br />
completed and successfully run on multiple occasions, with course enrolments from many<br />
countries.<br />
The second is based on the undergraduate course Sustainable and Renewable Energy<br />
Engineering SOLA2053. The latter is currently being <strong>of</strong>fered electronically on-line via Web CT<br />
for local students enrolling during session 2, 2005, to demonstrate suitability for internetbased<br />
delivery internationally.<br />
A third course is under development for on-line delivery and is based on the course Solar Cell<br />
Technology and Manufacturing SOLA9006. A CD has been produced that contains all the<br />
necessary teaching material, assignments and simulation s<strong>of</strong>tware for project work, with this<br />
likely to be <strong>of</strong>fered over the internet internationally during 2006.<br />
“Linkage” Milestones<br />
G.1<br />
Total Average Cash and equipment Contribution <strong>of</strong> at least $600,000 p.a. from<br />
UNSW<br />
As documented in the fi nancial section within Centre Annual <strong>Report</strong>s for 2003, 2004 and<br />
2005, UNSW has made cash contributions to the Centre <strong>of</strong> approximately $1 million each<br />
year, well above the target allocation <strong>of</strong> $600k.<br />
In addition, with the Centre receiving “school” status, EFTSU income has been awarded to the<br />
Centre with additional earnings <strong>of</strong> close to $3 million.<br />
G.2<br />
Allocation <strong>of</strong> at least 70m 2 <strong>of</strong>fice space and 180m 2 laboratory space by UNSW<br />
The Centre’s space needs for new <strong>of</strong>fi ces have been signifi cantly greater than expected due<br />
to the growth in numbers <strong>of</strong> PhD students, post doctoral researchers and visiting Fellows/<br />
Academics associated with the Centre <strong>of</strong> Excellence. UNSW has responded well as the need<br />
has arisen, with 385m 2 having now been made available.<br />
Regarding Laboratory space, an additional 200m 2 <strong>of</strong> space has been made available for the<br />
Centre <strong>of</strong> Excellence, primarily to accommodate the new state-<strong>of</strong>-the-art facilities at Bay St,<br />
Botany.<br />
G.3<br />
Support for at least 4 academic staff affiliated with the centre from EFTSU or related<br />
income<br />
UNSW has again exceeded its targeted support during 2004 through the provision <strong>of</strong> staff<br />
funded from EFTSU generated income or related sources. The staff supported either partially<br />
or fully are listed in the Annual <strong>Report</strong> and include: Dr J. Cotter; Dr A. Sproul; Pr<strong>of</strong>essor S.<br />
Wenham; Dr R. Corkish; Dr M. Watt; Dr B. Richards; Dr A. Ho; Ms A. Bruce; Dr G. Stephenson;<br />
and M. Kobacker.<br />
G.4<br />
Average <strong>of</strong> at least 3 weeks per year <strong>of</strong> interchange between collaborating<br />
organisations<br />
This target has been comfortably exceeded as documented in the section on collaborative<br />
research in Section 4.7.<br />
154
Governance Milestones<br />
H.1<br />
Establishment <strong>of</strong> Advisory and Management Committees during 2003.<br />
As was reported in the 2003 Annual <strong>Report</strong>, this milestone has been achieved.<br />
The Advisory Committee <strong>of</strong> the ARC Centre <strong>of</strong> Excellence has met in June 2004 and June<br />
2005.<br />
The Management Committee meets fortnightly. The membership <strong>of</strong> the latter includes the<br />
Centre Director (S. Wenham), the Executive Research Director (M. Green), the 4 Deputy<br />
Directors (G. Conibeer, J. Zhao, J. Cotter and A. Aberle), the Head <strong>of</strong> Characterisation (A.<br />
Sproul), the Head <strong>of</strong> School for the Centre for Photovoltaic Engineering (R. Corkish) and the<br />
Centre’s Business and Technology Manager (M. Silver).<br />
H.2<br />
Formation <strong>of</strong> the School <strong>of</strong> Photovoltaic and Renewable Energy Engineering and the<br />
appointment <strong>of</strong> a Head <strong>of</strong> School during 2003.<br />
Richard Corkish <strong>of</strong>fi cially appointed as Head <strong>of</strong> School in August 2003, indicating impending<br />
recognition <strong>of</strong> School status<br />
UNSW Council meeting in July 2005 approved the <strong>of</strong>fi cial formation on 1 January 2006 <strong>of</strong><br />
the School <strong>of</strong> Photovoltaic and Renewable Energy Engineering<br />
H.3<br />
Establishment <strong>of</strong> the Management Structure <strong>of</strong> Figure H.1 <strong>of</strong> the initial application<br />
by June, 2003, with appropriate documentation <strong>of</strong> responsibilities <strong>of</strong> all staff and<br />
managers.<br />
The Management structure <strong>of</strong> Section 6 was successfully established in early 2003 as was<br />
reported in the 2003 Annual <strong>Report</strong>. Since initially being established, the following changes<br />
have been made:<br />
Dr J. Cotter has taken over as Manager <strong>of</strong> the wafer-based projects in C. Honsberg’s<br />
absence, with his automatic inclusion in the Management Comittee<br />
Head <strong>of</strong> School, Richard Corkish, was appointed Aug. 2003<br />
G. Conibeer replaced R. Corkish as manager <strong>of</strong> the Third Generation strand.<br />
T. Burns took over responsibility for School Administration<br />
The roles and responsibilities for most staff are defi ned, in most cases, in their <strong>of</strong>fi cial Position<br />
Descriptions, held by Human Resources Department. These are routinely reviewed and<br />
revised at times <strong>of</strong> appointments and reappointments.<br />
In special cases, such as for the Head <strong>of</strong> School who is responsible to the Dean <strong>of</strong> Engineering,<br />
additional documentation exists: http://www.hr.unsw.edu.au/poldoc/hos.htm. Management,<br />
staff and student responsibilities for occupational health and safety are documented by the<br />
UNSW Risk Management Unit<br />
H.4<br />
Consolidation <strong>of</strong> <strong>New</strong> Centre Strategic Plan in <strong>New</strong> Document by June, 2003.<br />
Following the achievement <strong>of</strong> this milestone in 2003, the strategic plan is regularly reviewed<br />
in consultation with the Head <strong>of</strong> School and the Dean <strong>of</strong> Engineering.<br />
A copy <strong>of</strong> the Centre’s Strategic Plan is held by the School <strong>of</strong> Photovoltaic Engineering<br />
155
H.4<br />
H.5 Document Management and Advisory Committee Meetings at least monthly<br />
and annually, respectively.<br />
Advisory Committee meetings have been held annually in June 2004 and June 2005.<br />
Management meetings are scheduled fortnightly. At the latter, reports have been written for<br />
each meeting to document progress in each <strong>of</strong> the research areas.<br />
Monthly overview meetings are also held for the entire Centre for information dissemination<br />
and the presentation <strong>of</strong> brief (6 minute) reports from each area <strong>of</strong> activity including the Lab<br />
Operations Team, External Affairs and Teaching.<br />
In addition to the overview meetings, members <strong>of</strong> the Management team also hold<br />
meetings with the respective staff and students from their particular area in relation to the<br />
corresponding research and other activities. Such meetings are normally held either on a<br />
fortnightly or monthly basis, except for teaching meetings which involve most staff and are<br />
only held every few months.<br />
8.2 Progress Against ARC Designated Performance Indicators<br />
Subsequently to the initial application, the ARC also designated an additional set <strong>of</strong> performance<br />
indicators. Progress against these is documented below.<br />
Quality <strong>of</strong> publications – the Centre has achieved its target <strong>of</strong> 30% <strong>of</strong> its publications in<br />
refereed journals each year, with 31.5% in 2003 and 35.6% in 2004 and 33.6% in 2005.<br />
The Centre’s 2005 publications are listed towards the end <strong>of</strong> this report.<br />
Number <strong>of</strong> publications – the Centre has easily achieved its target <strong>of</strong> 50 publications per year<br />
with 89 in 2003, 101 in 2004 and 113 in 2005.<br />
Number <strong>of</strong> patents per year – the Centre’s programs have led to more than the expected 1<br />
new patent per year. In fact there were 2 in 2003, 4 in 2004 and 7 in 2005. The patents are<br />
reasonably well distributed between the three generations <strong>of</strong> PV technology with 4 in the 1st<br />
generation PV area, 5 in the 2nd generation PV area, 3 in the 3rd generation PV area and 1<br />
in the characterisation area.<br />
Invitations to address and participate in International Conferences – the Centre has easily<br />
achieved its target <strong>of</strong> 2 per year with more than 40 <strong>of</strong> such invitations having been received<br />
since 2003.<br />
Invitations to visit leading international laboratories – quite a few Centre staff regularly receive<br />
invitations to visit leading international laboratories with almost 30 such invited visits taking<br />
place. This is well above the target number <strong>of</strong> 2/year.<br />
Number and nature <strong>of</strong> commentaries about the Centre’s achievements – the target number<br />
<strong>of</strong> 2 media articles per year has been well exceeded with more than 50 examples <strong>of</strong> such<br />
media interaction documented.<br />
Number <strong>of</strong> postgraduates recruited – the Centre has recruited 25 postgraduate students<br />
since 2003, well ahead <strong>of</strong> the target <strong>of</strong> 4 per year. The new enrolments include 19 postgraduate<br />
research students and 14 postgraduate coursework students.<br />
156
Number <strong>of</strong> postgraduate completions – the Centre has achieved 23 postgraduate completions,<br />
comprising 14 postgraduate research students and 9 postgraduate coursework students.<br />
Most <strong>of</strong> the former commenced their research degrees before the Centre commenced,<br />
with completion numbers expected to increase over the next couple <strong>of</strong> years due to the<br />
increased numbers <strong>of</strong> enrolments since being awarded the Centre. Despite this, the numbers<br />
<strong>of</strong> completions are well ahead <strong>of</strong> the target value <strong>of</strong> 3 per year.<br />
Number <strong>of</strong> honours students – despite the challenging target <strong>of</strong> producing 10 honours<br />
students per year and the delay in producing the fi rst graduates, the success <strong>of</strong> the Centre’s<br />
new undergraduate program has enabled it to comfortably exceed this average with 32<br />
honours graduates to date primarily in 2004 and 2005.<br />
Number <strong>of</strong> pr<strong>of</strong>essional courses – the Centre on average has run more than 8 such courses<br />
per year, ahead <strong>of</strong> the target fi gure <strong>of</strong> 6 per year.<br />
Participation in pr<strong>of</strong>essional courses – there has been 25 pr<strong>of</strong>essional courses run by the<br />
Centre since 2003. The target fi gure <strong>of</strong> 20 enrolments per course has been comfortably<br />
achieved with the course average being above 50.<br />
Number and level <strong>of</strong> undergraduate and high school courses in the priority area(s) – the<br />
Centre has run a large number <strong>of</strong> undergraduate courses in the priority area. The average<br />
number run each year exceeds 16, exceeding the target number <strong>of</strong> 12 per year.<br />
Number <strong>of</strong> international visitors – the Centre has received approaching 100 international<br />
visitors in the last 3 years well ahead <strong>of</strong> the target number <strong>of</strong> 10 per year.<br />
Number <strong>of</strong> national and international workshops – the Centre has achieved its target <strong>of</strong> 3<br />
workshops per year with participation in 16 such workshops internationally during the period<br />
2003-2005.<br />
Number <strong>of</strong> visits to overseas laboratories – these were listed in chronological order in the<br />
documentation produced for the ARC during the Centre’s review in late 2005. The total<br />
number comfortably exceeds the target number <strong>of</strong> 10 per year.<br />
Examples <strong>of</strong> relevant Social Science and Humanities research supported by the Centre – as<br />
described above, staff and students from the Centre have been actively involved in developing<br />
countries projects involved in providing renewable energy power sources to materially<br />
disadvantaged people who have in many cases never previously had access to electricity. Six<br />
Centre staff and more than 30 Centre students have participated in these projects including<br />
travelling to locations such as Nepal, Sri Lanka and Nicaragua to live briefl y with the locals,<br />
provide technical expertise to local technicians, design and install PV systems, and train locals<br />
in the use and maintenance <strong>of</strong> the new technology. This work is usually done in conjunction<br />
with Aid Organisations or local institutions.<br />
Number and nature <strong>of</strong> commercialisation activities – the Centre has engaged in 15 technology<br />
transfers which have in general been to companies who have licensed or purchased Centre<br />
technology, in some cases prior to the establishment <strong>of</strong> the Centre <strong>of</strong> Excellence in 2003. In<br />
addition, with the rapid growth <strong>of</strong> the PV industry, Centre staff are <strong>of</strong>ten also heavily involved<br />
in assisting companies in expanding or setting up new manufacturing facilities. These activities<br />
have exceeded the Centre’s target <strong>of</strong> 4 such technology transfers per year.<br />
157
Number <strong>of</strong> government, industry and business briefi ngs – the Centre target <strong>of</strong> 2 such<br />
briefi ngs has been exceeded with more than 20 such briefi ngs since the commencement<br />
<strong>of</strong> the Centre. The high fi gure has resulted from the extraordinary rate <strong>of</strong> growth in the<br />
industry (67% sales growth in 2004; 45% in 2005), with relatively large numbers <strong>of</strong> business<br />
briefi ngs by Centre staff in particular being sought by the industry<br />
Number <strong>of</strong> Centre associates trained/ing in technology transfer and commercialisation –<br />
although the target number for this was set at 5, the diversity <strong>of</strong> Centre technology coupled<br />
with the recent high demand from the rapidly growing industry, has necessitated the Centre<br />
train additional staff and associates in the various areas where there is commercial interest<br />
in the Centre’s technologies. The Centre now has 15 staff and/or research students at a<br />
stage where they feel confi dent in their training for technology transfer. Those trained in<br />
technology transfer <strong>of</strong>ten become particularly valuable to industry and are <strong>of</strong>ten <strong>of</strong>fered<br />
good employment opportunities. For example, two <strong>of</strong> the staff listed below have within recent<br />
months accepted <strong>of</strong>fers to work for newly established companies wishing to set up new<br />
manufacturing.<br />
Number and nature <strong>of</strong> Public Awareness programs – the Centre conducts extensive public<br />
awareness programs, particularly amongst high school students. This has been particularly<br />
important with the Centre developing and establishing the world’s fi rst undergraduate<br />
engineering degree in Photovoltaics and Solar Energy which is only able to attract students<br />
provided students become aware <strong>of</strong> its existence and relevance. There have been well over<br />
50 examples <strong>of</strong> activities in which the Centre has been engaged that increase the public<br />
awareness <strong>of</strong> the Centre’s program and <strong>of</strong> the photovoltaics fi eld in general.<br />
Annual cash contributions from Collaborating Organisations – the target <strong>of</strong> $0.5 million per<br />
year has been easily exceeded as documented in the Financial section<br />
Annual in-kind contributions from Collaborating Organisations – the target <strong>of</strong> $0.5 million<br />
per year has been easily exceeded with most collaborating organisations and companies<br />
preferring to make signifi cantly larger in-kind contributions than cash contributions.<br />
Number <strong>of</strong> new organisations recruited to or involved in the Centre – more than 20 new<br />
collaborators have been recruited to the Centre, well ahead <strong>of</strong> the target <strong>of</strong> 2 per year.<br />
Level and quality <strong>of</strong> infrastructure provided to the Centre – the Centres infrastructure and<br />
facilities were already world class at the time the Centre was awarded. Since this time the<br />
Centre has further developed its infrastructure and capabilities through acquiring state-<strong>of</strong>the-art<br />
facilities from Pacifi c Solar, using funding <strong>of</strong> well over $1 million from the <strong>University</strong><br />
and State Government <strong>of</strong> NSW.<br />
Breadth and experience <strong>of</strong> the members <strong>of</strong> the Advisory Board – the Centre’s Advisory Board<br />
comprises members recognised as industry leaders including manufacturers, end-users,<br />
academia and research institutes. In particular it includes several Heads <strong>of</strong> organisations<br />
such as CEO’s <strong>of</strong> major international companies. The membership is listed in Section 6 on<br />
Organisational Structure.<br />
158<br />
Frequency and Effectiveness <strong>of</strong> Advisory Board Meetings – the Advisory Board has met<br />
annually as planned. To be effective, a particular challenge for such an Advisory Board, with<br />
membership from 6 different countries, is to achieve adequate meeting attendance. The last<br />
meeting was held in Barcelona in conjunction with one <strong>of</strong> the major international conferences,<br />
with members from all six countries being in attendance with only two Australian members<br />
being unable to be present. Feedback and advice from Advisory Board members have been<br />
valuable for the Centre’s Management Team in effecting improvements in some areas,<br />
particularly with regard to content in teaching programs and commercial signifi cance <strong>of</strong><br />
some <strong>of</strong> the research.
Quality <strong>of</strong> the Centre Strategic Plan – the Centre’s strategic plan is <strong>of</strong> high quality following<br />
several iterations.<br />
The adequacy <strong>of</strong> the Centre’s Key Performance Indicators – the Centre’s performance<br />
criteria/milestones outlined in the original grant application are proving to be effective in<br />
guiding the research and its focus. Few, if any, have needed modifi cation. In general they<br />
are quite challenging and ambitious. For example, the Centre has made excellent progress<br />
in developing new high effi ciency solar cell technology to suit n-type wafers, with the recent<br />
achievement <strong>of</strong> equalling the world record with 22.7% effi ciency. Despite this success, the<br />
aim is to achieve the ambitious 24% effi ciency target by the end <strong>of</strong> 2007. Regarding the ARC<br />
designated performance indicators, their breadth and their direct links to the objectives <strong>of</strong><br />
the ARC Centres <strong>of</strong> Excellence program ensure the Centre adequately addresses all areas<br />
<strong>of</strong> importance and signifi cance in contributing to the achieving <strong>of</strong> the program’s aims and<br />
objectives.<br />
Measures <strong>of</strong> expansion <strong>of</strong> Australia’s capability in the priority area – the growth rate <strong>of</strong><br />
manufacturing capacity in Australia has greatly outstripped the targeted 15% per annum in<br />
recent years, with the largest manufacturer BP Solar signifi cantly expanding their production<br />
capability while Origin Energy is establishing new manufacturing capacity in Adelaide. The<br />
increase in capacity has been so great that locally manufactured product now exceeds<br />
demand in Australia by a factor <strong>of</strong> 5 (i.e. Australia now exports 80% <strong>of</strong> its manufactured<br />
product in this industry).<br />
Case studies <strong>of</strong> economic, social, cultural or environmental benefi ts – the Centre has met its<br />
target <strong>of</strong> one such case study per year. In 2003, Centre staff were involved in a case study<br />
on Policy Support for Renewables in Australia from the perspective primarily <strong>of</strong> the social and<br />
environmental impact <strong>of</strong> photovoltaics. This case study was published in January 2004:<br />
M. Watt, Hugh Outhred and Iain MacGill, “Policy Support for Renewables in Australia<br />
with PV as a Case Study”, Technical Digest, 14th International Photovoltaic Science and<br />
Engineering Conference (PVSEC-14), Bangkok, Thailand, January, 2004, p.1061-1062.<br />
A second case study in which Centre staff have been involved, focused on where PV is heading<br />
in the future, and considered economic, employment, markets, social and environmental<br />
benefi ts. This study culminated in producing for Australia a PV Roadmap with the following<br />
publication stemming from the work in 2004:<br />
M. Watt (2004) “Development <strong>of</strong> the Australian PV Roadmap”, presented on behalf <strong>of</strong><br />
the Australian Business Council for Sustainable Energy at the IEA-PVPS Workshop for<br />
Industry at the 19th European Photovoltaic Solar Energy Conference and Exhibition,<br />
Paris, France, 7-11 June, 2004.<br />
As described elsewhere, the Centre runs projects involving the use <strong>of</strong> PV in developing<br />
countries. These projects directly impact the economic, social, cultural and environmental<br />
well-being <strong>of</strong> those being helped. For example, the Solar Sisters program specifi cally aims to<br />
provide PV power to facilitate new opportunities for women such as for educational radio/<br />
television, lighting at night to engage in craft work to earn income, etc. Studies have been done<br />
to identify the social and economic benefi ts <strong>of</strong> these projects. In 2005, these case studies<br />
are being extended to the indigenous population in Australia for inclusion in the Centre’s<br />
“Developing Countries Projects”. Thesis student Gavin Periera under the supervision <strong>of</strong> Dr<br />
Richard Corkish, is currently conducting this case study.<br />
Another study being conducted at present by the Centre is a review by Centre staff <strong>of</strong><br />
the Federal Government organisation Bushlight. Bushlight is also involved in studying and<br />
capitalising on the social, cultural and economic benefi ts able to be <strong>of</strong>fered by PV and other<br />
renewable energy technologies in providing for the needs <strong>of</strong> our indigenous communities. At<br />
completion, thesis student Long Seng To, under the supervision <strong>of</strong> Dr Muriel Watt, will publish<br />
material from this case study.<br />
159
160
Centre 2005 Income<br />
The total income for 2005 for the Australian Research Council (ARC) Centre <strong>of</strong> Excellence<br />
in Advanced Silicon Photovoltaics and Photonics was $4.67 million. This does not include<br />
funding approaching $1 million from the Host Institution based on EFTSU and related income.<br />
This latter income has been generated by the Centre through its educational activities by<br />
virtue <strong>of</strong> the fact that it is now equivalent to a School within the Faculty <strong>of</strong> Engineering. This<br />
income has been used for the development <strong>of</strong> new courses and teaching materials and to<br />
fund the salaries <strong>of</strong> most <strong>of</strong> the academic staff associated with the Centre.<br />
ARC Centre Grant<br />
Other ACR grants<br />
State Government Funds<br />
Industry/Private Funds<br />
Host Institution Support<br />
Other Income<br />
Sources/Interest<br />
Figure 8.1: Centre income by source <strong>of</strong> income.<br />
The largest component <strong>of</strong> the Centre’s income was from the ARC totalling $2.6 million, while<br />
the second largest cash component <strong>of</strong> income was Host Institution support, with UNSW<br />
contributing $0.9 million excluding EFTSU and related income. This was well above the<br />
targeted $600k, and excludes the direct funding <strong>of</strong> seven full-time or part-time academic<br />
staff who have been actively involved in the Centre’s activities.<br />
The next largest cash component <strong>of</strong> income was $304k from the State Government through<br />
the Department <strong>of</strong> State and Regional Development (DSRD). This funding is particularly<br />
important for the purchase <strong>of</strong> new equipment and the development <strong>of</strong> facilities, and has<br />
contributed signifi cantly to the acquisition and development <strong>of</strong> the new state-<strong>of</strong>-the-art<br />
research facilities at Botany.<br />
161
The 2005 income from industry grew sharply over 2004 levels with cash contributions<br />
totalling $850k and even larger in-kind contributions for collaborative research. The largest<br />
overall contributor in 2005, as in the previous two years, was Suntech-Power Company who,<br />
in addition to their cash contribution <strong>of</strong> almost $200k, continued to fund two full-time research<br />
positions in the Centre’s laboratories for several months, while simultaneously supporting the<br />
collaborative research at their own facilities. Including the funding <strong>of</strong> collaborative research<br />
costs, scholarships, royalties, consulting and rental <strong>of</strong> facilities, the major cash contributors<br />
to the Centre’s 2005 income from industry were BP Solar with $306k, CSG Solar with a<br />
contribution <strong>of</strong> $262k, Suntech-Power with $187k and Toyota Motor Corporation with $90k.<br />
Other cash or in-kind contributions have been made by the Max Plank Institute, FOM Institute,<br />
Eurosolare, ANSTO, Ian Potter Foundation, the Australian CRC for Renewable Energy, the<br />
Australian Business Council for Solar Energy, the Fraunh<strong>of</strong>er Institute, SunPower, ENEA, the<br />
Paul Scherrer Institute and the Australian National <strong>University</strong>. These organisations contribute<br />
to the Centre through the expertise and experience they <strong>of</strong>fer in collaborative research areas<br />
and the in-kind support they provide through access to equipment, facilities and personnel not<br />
available at UNSW.<br />
CSG Solar<br />
BP Solar<br />
Suntech - Power<br />
Other Industry/Private<br />
Funds<br />
Figure 8.2: Industry contributions<br />
Other Centre contracts, consulting work and technology transfers are conducted through<br />
<strong>New</strong><strong>South</strong> Innovations, the commercial arm <strong>of</strong> the <strong>University</strong>. These are handled on behalf <strong>of</strong><br />
the Centre <strong>of</strong> Excellence and its staff through the <strong>New</strong><strong>South</strong> Innovations accounts and are<br />
not included in this fi nancial report.<br />
The Centre also earns income through the sale <strong>of</strong> educational CDs, books and computer<br />
s<strong>of</strong>tware, with a combined income for 2005 <strong>of</strong> $24k.<br />
162
Centre 2005 Expenditure <strong>of</strong> ARC Grant<br />
The total expenditure <strong>of</strong> ARC grant monies during 2005 was $2.56 million, close to the<br />
$2.60 million awarded. By far the largest component <strong>of</strong> this expenditure was on salaries,<br />
with the employment <strong>of</strong> 18 full-time staff and a further 53 part-time or casual staff, including<br />
student scholarships. The combined expenditure on salaries and scholarships during 2005<br />
was $1.71 million. This included 11 full-time researchers: D. König, A. Ho, M. Abbott, K. Fisher,<br />
A. Wang, G. Zhang, R. Bardos, G. Conibeer, X.M. Dai, K. Lin and B. Richards. To support the<br />
research activities, 7 technical support staff were employed by the Centre during 2005,<br />
namely M. Silver, M. Griffi n, L. Soria, A. Yee, J. Green, J. Yang and R. Largent. One parttime<br />
and one full-time administrative staff member were also employed by the Centre during<br />
2005, J. Hansen and J. Kwan. The latter is the Financial Offi cer for the Centre <strong>of</strong> Excellence,<br />
while the former carries out secretarial, administrative and receptionist duties.<br />
Equipment expenditure from the Centre grant was kept relatively small for 2005 at $209k<br />
due to the support from the State Government through DSRD and the host institution UNSW.<br />
These provided signifi cant levels <strong>of</strong> funding specifi cally targeting equipment and infrastructure<br />
development. The majority <strong>of</strong> equipment expenditure from these latter sources was for a<br />
sophisticated sputtering machine and in equipping the new Botany facilities with a range <strong>of</strong><br />
processing capabilities. This freed-up Centre funds so that they could be used primarily for<br />
small laboratory equipment items such as hotplates, temperature controllers, a small muffl e<br />
furnace, tuneable laser, pumps, a small processing furnace, specialised computing facilities,<br />
test equipment and a range <strong>of</strong> spare parts.<br />
The total expenditure from the ARC grant on consumables and maintenance for 2005 was<br />
$472k. These costs are strongly dominated by laboratory consumables to support the device<br />
research, such as high purity gases, chemicals and general laboratory supplies.<br />
The expenditure on travel during 2005 from the ARC grant was $163k. The primary purposes<br />
for this travel were for staff and students participating in collaborative research projects<br />
with overseas industry partners such as BP Solar in the USA, UK and Spain and Suntech-<br />
Power in China, and for the attendance at international conferences such as the PVSEC-15<br />
Conference in Shanghai, China and the IEEE Photovoltaic Specialists Conference in Orlando<br />
in the United States.<br />
Salaries & Scholarships<br />
Equipment<br />
Accommodation<br />
Travel<br />
Materials/Consumables<br />
Figure 8.3: Expenditure <strong>of</strong> ARC income by expenditure category.<br />
163
164
BOOKS<br />
Stuart R. Wenham, Martin A. Green, Muriel E. Watt, Richard P. Corkish, “Applied Photovoltaics”<br />
(Second Edition), Centre for Photovoltaic Engineering, 2005 (printed February 2006 and to<br />
be translated into Spanish and Chinese).<br />
BOOK CHAPTERS<br />
A.G. Aberle, “Progress in crystalline silicon thin-fi lm solar cells”, in: Progress in Solar Energy<br />
Research (Nova Science Publishers, Hauppauge, NY, USA), in press (invited).<br />
R. Corkish, “Photovoltaic Materials”, in “Encyclopedia <strong>of</strong> Chemical Processing”, Vol. 4, DOI:<br />
10.1081/E-ECHP-120007896, Taylor & Francis, 2006, p.2129-2138 (invited).<br />
R. Corkish, “Howard keeps us stuck in the coal age”, in “Renewable Energy”, Volume 229<br />
in the Spinney Press series on social issues, “Issues in Society”, ed. Justin Healey, ISBN 1<br />
920801 39 1, for Australian secondary school students (http://www.spinneypress.com.<br />
au/229_book_desc.html). (Reprinted from Sydney Morning Herald, 16 June 2004, p. 15.)<br />
M.A. Green, “Photovoltaic Applications <strong>of</strong> Nanostructures”, in A.A. Baladin and K.L. Wang<br />
(eds.), “Handbook <strong>of</strong> Semiconductor Nanostructures and Devices”, 2005 (invited).<br />
M.A. Green, “High Effi ciency Silicon Solar Cell Concepts” in “Solar Cells: Materials, Manufacture<br />
and Operation”, T. Markvart and L. Castener (eds.), Elsevier, Oxford, 2005, p.190-214<br />
(invited).<br />
M.A. Green, “Eureka!” in “The Ideas Book”, L. Carroli (ed.), Brisbane, <strong>University</strong> <strong>of</strong> Queensland<br />
Press, 2005.<br />
M.A. Green, “Very High-Effi ciency in Silico Photovoltaics” in “Artifi cial Photosynthesis: From<br />
Basic Biology to Industrial Application”, A.F. Collings and C. Critchley (eds.), Wiley-VCH,<br />
Weinheim, 2005, p.169-185 (invited).<br />
M.A. Green, “Futuristic Photovoltaics”, in “Renewable Energy 2005”, D. Flin (Ed.), Sovereign<br />
Publications, London, 2005, p.52-55, ISBN 19 03605 652 (invited).<br />
M.A. Green, “Materials for Solar Cells” in “Materials for Energy Conversion Devices”, C.C.<br />
Sorrel, S. Sugihara and J. Notvotny (Eds.), Woodhead Publishing, Cambridge, 2005, pp.1-62<br />
(invited).<br />
M.A. Green, “Third Generation Photovoltaics” in “Solar Photovoltaic Batteries”, K. Misawa<br />
(Ed.), Nikkan Kougyo Simbun Ltd, Japan, 2006, p.186-196 (invited; translated into Japanese).<br />
M.A. Green, “Photovoltaics in Deserts”, in “Global Environment Outlook -Deserts”, Global<br />
Environment Outlook Series, United Nations Environment Programme, 2006.<br />
B.S. Richards and M.A. Green, “Photovoltaic Cells”, in “Encyclopaedia <strong>of</strong> Biomedical Engineering”,<br />
Wiley, 2005 (invited).<br />
165
PATENTS, PATENT APPLICATIONS<br />
A.G. Aberle, T. Walsh, and D Inns, “Photolithography method for contacting thin-fi lm<br />
semicon¬ductor structures”, Australian Provisional Patent Application 2005901285 (fi led<br />
on 16 Mar 2005), International PCT patent application in preparation.<br />
M.A. Green, “Up- and Down-Conversion using Quantum Dot Arrays”, August 2005.<br />
M.A. Green, “Resonant Defect Enhancement <strong>of</strong> Current Transport in Semiconducting<br />
Superlattices”, Australian Provisional Patent Application No. 2005902712, (fi led on May<br />
2005).<br />
T. Trupke and R. Bardos, “Method and System For Inspecting Indirect Bandgap Semiconductor<br />
Structure”, September 2005.<br />
T. Walsh, A.G. Aberle and S.R. Wenham, “Thin-fi lm solar cell interconnection”, International<br />
PCT patent application PCT/AU2005/000734 (fi led on 23 May 2005, priority date 04 June<br />
2004), international publication number WO 2005/119782 A1, international publication date<br />
15 Dec 2005). National patent applications fi led in key countries for photovoltaic technology.<br />
S.R. Wenham and M.A. Green, “Ink Jet Printing for High Performance Solar Cells”, October<br />
2005.<br />
S.R. Wenham and L. Mai, “Low Area Screen Printed Metal Contacts”, October 2005.<br />
S.R. Wenham, L. Mai and B.S. Tjahjono, “Transparent Conductors for Silicon Solar Cells”, June<br />
2005.<br />
PAPERS IN SCIENTIFIC JOURNALS<br />
M.D. Abbott and J.E. Cotter “Optical and Electrical Properties <strong>of</strong> Laser Texturing for High<br />
Effi ciency Solar Cells,” Progress in Photovoltaics: Research and Applications, Vol. 14(3)<br />
p. 225-235, 2006.<br />
M.D. Abbott, J.E. Cotter, T. Trupke, R.A. Bardos, “Investigation <strong>of</strong> Edge Recombination Effects in<br />
Silicon Solar Cell Structures using Photoluminescence”, Applied Physics Letters (in press).<br />
A.G. Aberle, “Progress with polycrystalline silicon thin-fi lm solar cells on glass at UNSW”,<br />
Journal <strong>of</strong> Crystal Growth, Vol. 287, p.386-390, 2006.<br />
A.G. Aberle, “Fabrication and characterisation <strong>of</strong> crystalline silicon thin-fi lm materials for solar<br />
cells”, Thin Solid Films (in press).<br />
A.G. Aberle, A. Straub, P.I. Widenborg, A.B. Sproul, Y. Huang and P. Campbell, “Polycrystalline<br />
Silicon Thin-Film Solar Cells on Glass by Aluminium-Induced Crystallisation and Subsequent Ion<br />
Assisted Deposition (ALICIA)”, Progress in Photovoltaics, Vol. 13, p.37-47, 2005.<br />
P.P. Altermatt, F. Geelhaar, T. Trupke, X. Dai, A. Neisser and E. Daub, “Injection dependence <strong>of</strong><br />
spontaneous radiative emission in crystalline silicon: Experimental verifi cation and theoretical<br />
analysis”, Applied Physics Letters (submitted).<br />
166
R.A. Bardos, T. Trupke, M. Schubert, T. Roth, “Trapping artefacts in quasi steady state<br />
photoluminescence and photoconductance lifetime measurements on silicon wafers”, Applied<br />
Physics Letters, Vol. 88, 053504, 2006.<br />
G. Conibeer, M. Green, R. Corkish, Y. Cho, E.-C. Cho, C.-W. Jiang, T. Fangsuwannarak, E. Pink, Y.<br />
Huang, T. Puzzer, T. Trupke, B. Richards, A. Shalav, K.-L. Lin, “Silicon nanostructures for third<br />
generation photovoltaic solar cells”, Thin Solid Films, 2005 (submitted, refereed).<br />
R. Corkish, S.R. Wenham, M.A. Green, A.B. Sproul, J. Cotter, A.G. Aberle, A. Bruce, “Materials<br />
Engineering Education in Two <strong>New</strong> Engineering Degree Programs at the Centre for Photovoltaic<br />
Engineering”, Journal <strong>of</strong> Materials Education (accepted).<br />
J.E. Cotter, J.H. Guo, P.J. Cousins, M.D. Abbott, F.W. Chen and K.C. Fisher, “p-type vs. n-type<br />
Silicon Wafers: Prospects for High-effi ciency, Commercial Silicon Solar Cells”, IEEE Transactions<br />
on Electron Devices (submitted 2005).<br />
P.J. Cousins and J. Cotter, “Minimising lifetime degradation associated with thermal oxidation<br />
<strong>of</strong> upright randomly textured silicon surfaces”, Solar Energy Materials and Solar Cells<br />
(accepted April 2005).<br />
P.J. Cousins, A. Guo and J. Cotter, “Infl uence <strong>of</strong> diffusion-induced misfi t dislocations on high<br />
effi ciency silicon solar cells”, IEEE Transactions on Electron Devices (accepted May 2005).<br />
P.J. Cousins, N. Mason and J. Cotter, “Loss Analysis <strong>of</strong> double-sided buried contact solar cells<br />
on FZ(B), MCZ(B), CZ(Ga) and CZ(B) wafers”, IEEE Trans. Electron Devices. (accepted August<br />
2005).<br />
L.V. Dao, X. Wen, M.T.T. Do, P. Hannaford, E.-C. Cho, Y.H. Cho, Y. Huang, “Time-Resolved and<br />
Time-Integrated Photoluminescence for Studies <strong>of</strong> State Filling and Quantum Confi nement <strong>of</strong><br />
Silicon Quantum Dots”, Journal <strong>of</strong> Applied Physics, Vol. 97(1), 013501, 2005.<br />
M.A. Green, K. Emery, D.L. King, S. Igari and W. Warta, “Solar Cell Effi ciency Tables (Version<br />
25)”, Progress in Photovoltaics: Research and Applications, Vol. 13(1), p.49-54, 2005.<br />
M.A. Green, “Price/Effi ciency Correlations for 2004 Photovoltaic Modules”, Progress in<br />
Photovoltaics: Research and Applications, Vol. 13(1), p.85-87, 2005.<br />
M.A. Green, K. Emery, D.L. King, S. Igari and W. Warta, “Solar Cell Effi ciency Tables (Version<br />
26)”, Progress in Photovoltaics: Research and Applications, Vol. 13(5), p.387-392, 2005.<br />
M.A. Green, “Silicon Photovoltaic Modules: A Brief History over the First 50 Years”, Progress<br />
in Photovoltaics, Vol. 13, p.447-455, 2005.<br />
M.A. Green, “Improved Estimates for Te and Se Availability from Cu Anode Slimes and Recent<br />
Price Trends”, Progress in Photovoltaics (in press).<br />
M.A. Green “Consolidation <strong>of</strong> Thin-Film Photovoltaics Technology: The Coming Decade <strong>of</strong><br />
Opportunity”, Progress in Photovoltaics (invited).<br />
M.A. Green, K. Emery, D.L. King, S. Igari and W. Warta, “Solar Cell Effi ciency Tables (Version<br />
27), Progress in Photovoltaics, Vol. 14, pp. 45-51, 2006.<br />
167
J.H. Guo and J.E. Cotter, “Metallization Improvement on Fabrication <strong>of</strong> Interdigitated Backside<br />
and Double-sided Buried Contact Solar Cells,” Solar Energy Materials and Solar Cells, Vol.<br />
86(4), p.485-498, 2005.<br />
J.H. Guo and J.E. Cotter, “Optimizing the Diffused Regions <strong>of</strong> Interdigitated Backside Buried<br />
Contact Solar Cells,” Solar Energy Materials and Solar Cells (in review).<br />
J.H. Guo, P.J. Cousins, and J.E. Cotter, “Investigations <strong>of</strong> Parasitic Shunt Resistance in N-type<br />
Buried Contact Solar Cells,” Progress in Photovoltaics: Research and Applications, Vol. 14,<br />
p.95-105, 2006.<br />
A.W.Y. Ho and S.R. Wenham, “Innovative Contacting Scheme for Electron Devices”, IEEE<br />
Transactions on Electron Devices (submitted October 2005).<br />
A.W.Y. Ho, S.R. Wenham, “Fabrication <strong>of</strong> Silicon Solar Cells with Rear Pinhole Contacts”,<br />
Progress in Photovoltaics: Research & Applications (submitted).<br />
C.-W. Jiang and M.A. Green, “Silicon Quantum Dot Superlattices: Modelling <strong>of</strong> Energy Bands,<br />
Densities <strong>of</strong> States and Mobilities for Silicon Tandem Solar Cell Applications”, Journal <strong>of</strong><br />
Applied Physics, JR05-2773. (submitted August 2005).<br />
D. König, R. Scholz, D.R.T. Zahn, G. Ebest, “Band diagram <strong>of</strong> the AlF_3/SiO_2/Si system”,<br />
Journal <strong>of</strong> Applied Physics, Vol. 97, 093707, 2005<br />
G. Kumaravelu, M.M. Alkaisi, D. Macdonald, J. Zhao, B. Rong and A. Bittar, “Minority carrier<br />
lifetime in plasma-textured silicon wafers for solar cells”, Solar Energy Materials and Solar<br />
Cells, Vol. 87(1-4), Vol. 14(2), p.99-106, 2005.<br />
K.R. McIntosh, J.E. Cotter and D. Swanson, “A Simple Ray Tracer to Compute the Concentration<br />
<strong>of</strong> a Photovoltaic Module”, Progress in Potovoltaics: Research and Applications, p.167-177,<br />
2006<br />
S. Pillai, K. Catchpole , T. Trupke , G. Zhang , J. Zhao , M.A. Green, “Enhanced Emission from Si<br />
based LEDs using Surface Plasmons”, Applied Physics Letters (submitted December 2005).<br />
B.S. Richards, “Luminescent layers for enhanced silicon solar cell performance: Down<br />
conversion”, Solar Energy Materials and Solar Cells (in press).<br />
B.S. Richards and G. Conibeer, “A Comparison <strong>of</strong> Hydrogen Storage Technologies for Solar-<br />
Powered Stand Alone Power Supplies”, Special Issue <strong>of</strong> Int. Journal <strong>of</strong> Hydrogen Energy<br />
(submitted).<br />
B.S. Richards and A. Shalav, “The Role <strong>of</strong> Polymers in the Luminescence Conversion <strong>of</strong> Sunlight<br />
for Enhanced Solar Cell Performance”, Synthetic Metals, Vol. 54(22), Issues 1-3, p.61-64,<br />
2005.<br />
B.S. Richards, A. Shalav, K.W. Kramer and H.U. Gudel, “Up-conversion Photovoltaic Devices:<br />
Enhancing the Near Infrared Spectral Response <strong>of</strong> Silicon Solar Cell Response”, IEEE Electron<br />
Device Letters (submitted December 2005).<br />
A. Shalav, B.S. Richards, T. Trupke, K.W. Kramer and H.U. Gudel, “The Application <strong>of</strong> NaYF4:<br />
Er3+ Up-Converting Phosphorus for Enhanced Near Infrared Silicon Solar Cell Response”,<br />
Applied Physics Letters, Vol. 86(1), 013505, 2005.<br />
168
D. Song, D. Inns, A. Straub, M.L. Terry, P. Campbell, and A.G. Aberle, “Solid phase crystallized<br />
polycrystalline thin-fi lms on glass from evaporated silicon for photovoltaic applications”, Thin<br />
Solid Films (in press).<br />
A.B. Sproul, “Solar Geometry using Vector Analysis”, Renewable Energy (submitted).<br />
J. Stradal, G. Scholma, H. Li, C.H.M. van der Werf, J.K. Rath, P.I. Widenborg, P. Campbell, A.G.<br />
Aberle and R.E.I. Schropp, “Epitaxial Thickening by Hot Wire Chemical Vapor Deposition <strong>of</strong><br />
Polycrystalline Silicon Seed Layers on Glass”, Thin Solid Films (in press).<br />
A. Straub, R. Gebs, H. Habenicht, S. Trunk, R.A. Bardos, A.B. Sproul, and A.G. Aberle, “Impedance<br />
analysis - A powerful method for the determination <strong>of</strong> the doping concentration and built-in<br />
potential <strong>of</strong> nonideal semiconductor p-n diodes”, Journal <strong>of</strong> Applied Physics, Vol. 97, 083703,<br />
2005.<br />
A. Straub, D. Inns, M.L. Terry, R. Gebs, and A.G. Aberle, “The infl uence <strong>of</strong> defects and postdeposition<br />
treatments on the free carrier density in lightly phosphorus-doped large-grained<br />
polycrystalline silicon fi lms”, Journal <strong>of</strong> Applied Physics, Vol. 98(2), 023507, 2005.<br />
A. Straub, D. Inns, M.L. Terry, Y. Huang, P.I. Widenborg, and A.G. Aberle, “Optimisation <strong>of</strong> lowtemperature<br />
silicon epitaxy on seeded glass substrates by ion-assisted deposition”, Journal<br />
<strong>of</strong> Crystal Growth, Vol. 280(1-2), p.385-400, 2005.<br />
A. Straub, D. Inns, M.L. Terry, Y. Huang, P.I. Widenborg, and A.G. Aberle, “Homoepitaxial silicon<br />
growth in a non-ultra-high vacuum environment by ion-assisted deposition on Si wafer and<br />
seeded glass substrates”, Thin Solid Films (in press).<br />
M.L. Terry, A. Straub, D. Inns, D. Song, and A.G. Aberle, “Large open-circuit voltage improvement<br />
by rapid thermal annealing <strong>of</strong> evaporated solid-phase-crystallized thin-fi lm silicon solar cells on<br />
glass”, Applied Physics Letters, Vol. 86, 172108, 2005.<br />
T. Trupke, “Infl uence <strong>of</strong> reabsorption on quasi steady state photoluminescence measurements<br />
on crystalline silicon”, Journal <strong>of</strong> Applied Physics (in press).<br />
T. Trupke, R.A. Bardos et al, “Photoluminescence imaging <strong>of</strong> silicon wafers”, Applied Physics<br />
Letters (submitted).<br />
T. Trupke, R.A. Bardos, M.D. Abbott, “Self-Consistent calibration <strong>of</strong> photoluminescence and<br />
photocoductance lifetime measurements”, Applied Physics Letters, 87, 184102, 2005.<br />
T. Trupke, R.A. Bardos, M.D. Abbott and J.E. Cotter, “Suns-photoluminescence: Contact-less<br />
determination <strong>of</strong> current-voltage characteristics <strong>of</strong> silicon wafers”, Applied Physics Letters,<br />
Vol. 87(9), 093503, 2005.<br />
T. Trupke, A. Shalav, B.S. Richards, P. Würfel, M.A. Green, “Effi ciency enhancement <strong>of</strong> solar cells<br />
by luminescent up-conversion <strong>of</strong> sunlight”, Solar Energy Materials & Solar Cells (in press).<br />
P.I. Widenborg, A. Straub and A.G. Aberle, “Epitaxial thickening <strong>of</strong> AIC poly-Si seed layers on<br />
glass by solid phase epitaxy”, Journal <strong>of</strong> Crystal Growth, Vol. 276, p.19-28, 2005.<br />
P. Wuerfel, A.S. Brown, T.E. Humphrey and M.A. Green, “Particle Conservation in the Hot-<br />
Carrier Solar Cell”, Progress in Photovoltaics: Research and Applications, Vol. 13, p.277-285,<br />
2005.<br />
169
CONFERENCE PROCEEDINGS<br />
M.D. Abbott, P.J. Cousins, F.W. Chen and J.E. Cotter, ”Laser-induced defects in crystalline<br />
Silicon solar cells’’, 31st IEEE Photovoltaic Specialists Conference, Orlando, USA, 3-7 January<br />
2005, p.1241-1244.<br />
M.D. Abbott, J.E. Cotter and K. Fisher, “N-Type Bifacial Solar Cells with Laser Doped Contacts”,<br />
IEEE 4th World Conference on Photovoltaic Energy Conversion (WCPEC-4), Hawaii, 7-12 May<br />
2006 (accepted).<br />
M.D. Abbott, J.E. Cotter, T. Trupke, K. Fisher and R.A. Bardos, “Application <strong>of</strong> Photoluminescence<br />
to High-effi ciency Silicon Solar Cell Fabrication”, IEEE 4th World Conference on Photovoltaic<br />
Energy Conversion (WCPEC-4), Hawaii, 7-12 May 2006 (accepted).<br />
A.G. Aberle, “Progress in Evaporated Crystalline Silicon Thin-Film Solar Cells on Glass”, IEEE<br />
4th World Conference on Photovoltaic Energy Conversion (WCPEC-4), Hawaii, 7-12 May<br />
2006 (accepted).<br />
A.G. Aberle, “Crystalline silicon thin-fi lm solar cells on glass - Cheap electricity from the Sun?”,<br />
Proc. 16th Australian Institute <strong>of</strong> Physics Congress, Canberra, February 2005, see Book <strong>of</strong><br />
Abstracts (CD), p.237 (invited).<br />
A.G. Aberle, P.I. Widenborg, D. Song, A. Straub, M.L. Terry, T. Walsh, A. Sproul, P. Campbell,<br />
D. Inns, B. Beilby, M. Griffi n, J. Weber, Y. Huang, O. Kunz, R. Gebs, F. Martin-Brune, V. Barroux,<br />
and S.R. Wenham, “Recent advances in polycrystalline silicon thin-fi lm solar cells on glass at<br />
UNSW”, 31st IEEE Photovoltaic Specialists Conference, Orlando, USA, 3-7 January 2005,<br />
p.877-882 (invited review talk).<br />
A.G. Aberle, “Crystalline Silicon Thin-Film Solar Cells: Where are We? Where to Go from<br />
Here?”, 20th European Photovoltaic Solar Energy Conference, Barcelona, Spain, 6-10 June<br />
2005, p.1108-1111.<br />
P.P. Altermatt, F. Geelhaar, T. Trupke, X. Dai, A. Neisser and E. Daub, “Injection dependence <strong>of</strong><br />
spontaneous radiative recombination in c-Si: Experiment, theoretical analysis and simulation”,<br />
5th International Conference on Numerical Simulation <strong>of</strong> Optoelectronic Devices, NUSOD,<br />
Berlin, September 2005.<br />
R.A. Bardos and T. Trupke, “Self consistent determination <strong>of</strong> the generation rate in<br />
photoluminescence and photoconductance lifetime measurements”, 31st IEEE Photovoltaic<br />
Specialists Conference, Orlando, USA, 3-7 January 2005, p.899-902.<br />
R.A. Bardos, J.E. Cotter, A.Lorenz, T. Trupke, “Comparison <strong>of</strong> n and p type ribbon grown multi<br />
crystalline silicon wafers using photoluminescence imaging”, IEEE 4th World Conference on<br />
Photovoltaic Energy Conversion (WCPEC-4), Hawaii, 7-12 May 2006 (accepted).<br />
N. Borojevic, F.W. Chen, J.E. Cotter, “Properties <strong>of</strong> PECVD Silicon Nitride Films for Antirefl ection<br />
Coatings”, 43rd Australian and <strong>New</strong> Zealand Solar Energy Society (ANZSES) Annual<br />
Conference, Dunedin, <strong>New</strong> Zealand, 28-30 November 2005.<br />
P. Campbell, P.I. Widenborg, and A.G. Aberle, “Surface textures for large-grained poly-silicon<br />
thin-fi lm solar cells on glass using the AIT method”, 15th International Photovoltaic Science<br />
and Engineering Conference (PVSEC-15), Shanghai, China, 10-15 October 2005, p.859-860.<br />
170
F.W. Chen, J.E. Cotter, A. Cuevas, S. Winderbaum, K. Roth, “Anomalous Thermal Behaviour <strong>of</strong><br />
Surface Passivation by PECVD Silicon Nitride on P-type Silicon”, 20th European Photovoltaic<br />
Solar Energy Conference, Barcelona, Spain, 6-10 June 2005, p.1419-1422.<br />
F.W. Chen, J.E. Cotter, T. Trupke, R.A. Bardos, “Characterization <strong>of</strong> PECVD Silicon Nitride<br />
Passivation with Photoluminescence Imaging”, IEEE 4th World Conference on Photovoltaic<br />
Energy Conversion (WCPEC-4), Hawaii, 7-12 May 2006 (accepted).<br />
F.W. Chen, T.A. Li, J.E. Cotter, “PECVD Silicon Nitride Surface Passivation for High-Effi ciency<br />
N-type Silicon Solar Cells”, IEEE 4th World Conference on Photovoltaic Energy Conversion<br />
(WCPEC-4), Hawaii, 7-12 May 2006 (accepted).<br />
Y.H. Cho, M.A. Green, E.-C. Cho, Y. Huang, T. Trupke, G. Conibeer, “Silicon Quantum Dots in<br />
SiNx Matrix for Third Generation Photovoltaics”, 20th European Photovoltaic Solar Energy<br />
Conference, Barcelona, Spain, 6-10 June 2005, p.47-50.<br />
N. Chuangsuwanich, P. Campbell, P.I. Widenborg, A. Straub, and A.G. Aberle, “Light trapping<br />
properties <strong>of</strong> evaporated poly-silicon fi lms on AIT-textured glass substrates”, 31st IEEE<br />
Photovoltaic Specialists Conference, Orlando, USA, 3-7 January 2005, p.1161-1164.<br />
G. Conibeer, M.A. Green, R. Corkish, Y. Cho, E.-C. Cho, T. C.-W. Jiang, T. Fangsuwannarak, E. Pink,<br />
Y. Huang, T. Puzzer, T. Trupke, B. Richards, A. Shalav, K.-L. Lin, “Silicon nanostructures for third<br />
generation photovoltaic solar cells”, Symposium on Thin Film and Nanostructured Materials<br />
for Photovoltaics THINC-PV2, Strasbourg, France, May / June 2005 (invited).<br />
G.J. Conibeer, J.F. Guillemoles and M.A. Green, “Phononic Band Gap Engineering for Hot Carrier<br />
Solar Cell Absorbers”, 20th European Photovoltaic Solar Energy Conference, Barcelona,<br />
Spain, 6-10 June 2005, p.35-38.<br />
G.J. Conibeer, M.A. Green, R. Corkish, Y. Cho, E-C. Cho, C-W. Jiang, E. Pink,B. Richards, A.<br />
Shalav, T. Fangsuwannarak, Y. Huang, T. Puzzer, T. Trupke, “Third Generation Photovoltaics:<br />
routes to high effi ciency that retain the advantages <strong>of</strong> thin fi lm”, 43rd Australian and <strong>New</strong><br />
Zealand Solar Energy Society (ANZSES) Annual Conference, Dunedin, <strong>New</strong> Zealand, 28-30<br />
November 2005.<br />
R. Corkish, S.R. Wenham, A.B. Sproul, J. Cotter, C.B. Honsberg, M.A. Green, A.G. Aberle, G.<br />
Conibeer, “Education and Research at the Centre for Photovoltaic Engineering, <strong>University</strong> <strong>of</strong><br />
<strong>New</strong> <strong>South</strong> <strong>Wales</strong>”, International Rio 5 World Climate and Energy Event, Rio de Janeiro, Brazil,<br />
15-17 February 2005, p. 481-487.<br />
R. Corkish, S.R. Wenham, M.A. Green, J.E. Cotter , A. Sproul, A.B. Aberle, J. Zhao, G. Conibeer,<br />
G. Stapleton, M. Watt, E. Spooner, A. Bruce, A.W.Y. Ho, M. Kobacker, “Education and Research<br />
at the Centre for Photovoltaic Engineering, UNSW”, 15th International Photovoltaic Science<br />
and Engineering Conference (PVSEC-15), Shanghai, China, 10-15 October 2005, p.455-456.<br />
R. Corkish, G. Conibeer, M. A. Green, Y. Cho, E.-C. Cho, T. Fangsuwannarak, E. Pink, Y. Huang,<br />
T. Puzzer, T. Trupke, A. Shalav, K.-L. Lin, “Silicon Nanostructures for Third Generation<br />
Photovoltaics”, Material Research Society (MRS) Fall Meeting, Boston, USA, 28 Nov-2 Dec<br />
2005.<br />
171
R. Corkish, S.R. Wenham, M.A. Green, A.B. Sproul, J. Cotter, A.G. Aberle, A. Bruce, “Materials<br />
Engineering Education in Two <strong>New</strong> Engineering Degree Programs at the Centre for<br />
Photovoltaic Engineering”, Material Research Society (MRS) Fall Meeting, Boston, USA, 28<br />
Nov-2 Dec 2005 (invited).<br />
J.E. Cotter, “RaySim 6.0 - A Free Geometrical Ray Tracing Program for Silicon Solar Cells”,<br />
31st IEEE Photovoltaic Specialists Conference, Orlando, USA, 3-7 January 2005, p.1165-<br />
1168.<br />
J.E. Cotter, G. Yao and B. Eggleston, “Laser-formed Stencils for Printed Silicon Solar Cells”,<br />
31st IEEE Photovoltaic Specialists Conference, Orlando, USA, 3-7 January 2005, p.1169-<br />
1172.<br />
J.E. Cotter, J.H. Guo, P.J. Cousins, M.D. Abbott, F.W. Chen and K.C. Fisher, “p-type vs. n-type<br />
Silicon Wafers: Prospects for High-effi ciency, Commercial Silicon Solar Cells”, 15th workshop<br />
on Crystalline Silicon Solar cells & Modules: Materials and Processes, Vail, USA, 7-10 August<br />
2005 (invited).<br />
P.J. Cousins and J.E. Cotter, “Misfi t dislocations generated during non-ideal boron and<br />
phosphorus diffusion and their effect on high-effi ciency silicon solar cells”, 31st IEEE<br />
Photovoltaic Specialists Conference, Orlando, USA, 3-7 January 2005, p.1047-1050.<br />
P.J. Cousins, N.B. Mason and J.E. Cotter, “Manufacturing and design issues for thin silicon<br />
solar cells on FZ(B), MCZ(B), CZ(Ga) and CZ(B) wafers”, 31st IEEE Photovoltaic Specialists<br />
Conference, Orlando, USA, 3-7 January 2005, p.987-990.<br />
P. J. Cousins, M. D. Abbott and J. E. Cotter, “Impact <strong>of</strong> Furnace Transients on the Lifetime<br />
<strong>of</strong> Upright Randomly Textured Wafers Evaluated using Photoconductance Lifetime<br />
Measurements” 20th European Photovoltaic Solar Energy Conference, Barcelona, Spain, 6-<br />
10 June 2005, p.1182-1185.<br />
A. Cuevas, F.W. Chen, J. Tan, S. Winderbaum, K. Roth, “FTIR Analysis <strong>of</strong> Microwave Silicon<br />
Nitride”, IEEE 4th World Conference on Photovoltaic Energy Conversion (WCPEC-4), Hawaii,<br />
7-12 May 2006 (accepted).<br />
D. Debuf, “Multiple-level-defect measurements using Fundamental Frequency Deep Level<br />
Transient Spectroscopy: A new measurement technique”, 20th European Photovoltaic Solar<br />
Energy Conference, Barcelona, Spain, 6-10 June 2005, p.132-135.<br />
M.B. Edwards, G. Yao, F. W. Chen, J. E. Cotter, “Metallisation<br />
for n-type, stencil-printed silicon solar cells”, 15th International Photovoltaic Science and<br />
Engineering Conference (PVSEC-15), Shanghai, China, 10-15 October 2005, p.710-711.<br />
T. Fangsuwannarak, E. Pink, Y. Huang, Y. H. Cho, G. Conibeer, T. Puzzer and M.A. Green,<br />
“Conductivity <strong>of</strong> self-organized silicon quantum dots embedded in silicon dioxide”, Proceeding<br />
<strong>of</strong> SPIE 2005 International Symposium on Microelectronics, MEMS and Nanotechnology,<br />
Brisbane, Australia, 11-14 December 2005, Vol.6037, 60370T.<br />
K.C. Fisher, J.E. Cotter, “Investigation <strong>of</strong> Low Molarity Alkaline Texturing Solutions”, 15th<br />
International Photovoltaic Science and Engineering Conference (PVSEC-15), Shanghai, China,<br />
10-15 October 2005, p.622-623.<br />
172
M.A. Green, “Photovoltaics – Electricity’s Future?”, Local Government Manager, Accent on<br />
Energy, Dec 2004-Jan 2005, p.8 (invited).<br />
M.A. Green, “Renewable Energy and Its Future”, International Rio 5 World Climate and Energy<br />
Event, Rio de Janeiro, Brazil, 15-17 February 2005.<br />
M.A. Green, “State-<strong>of</strong>-the-Art <strong>of</strong> Solar Cell Development”, International Rio 5 World Climate<br />
and Energy Event, Rio de Janeiro, Brazil, 15-17 February 2005.<br />
M.A. Green, “Status <strong>of</strong> Silicon on Glass, a <strong>New</strong> Production Technology”, 2nd Photovoltaic<br />
Science Applications and technology Conference, Loughborough, U.K., April 2005 (invited).<br />
M.A. Green, E.-C. Cho, Y. Cho, Y. Huang, E. Pink, T. Trupke, A. Lin, T. Fangsuwannarak, T. Puzzer, G.<br />
Conibeer, R. Corkish, “All-Silicon Tandem Cells based on “Artifi cial” Semiconductor Synthesised<br />
using Silicon Quantum Dots in a Dielectric Matrix”, 20th European Photovoltaic Solar Energy<br />
Conference, Barcelona, Spain, 6-10 June 2005, p.3-7 (opening plenary).<br />
M.A. Green, “Third Generation Photovoltaics and Feasibility <strong>of</strong> Realisation”, 15th International<br />
Photovoltaic Science and Engineering Conference (PVSEC-15), Shanghai, China, 10-15<br />
October 2005, p.7-10 (invited opening ceremony paper).<br />
M.A. Green, “Bulk Silicon for Photonics Applications”, IEEE Laser and Electro-Optics Society<br />
(LEOS), Sydney, Australia, 23-27 October 2005 (invited talk).<br />
M.A. Green, “All-Silicon Tandem Solar Cells”, Joint Scientifi c Workshop on the Utilization <strong>of</strong> the<br />
Solar Spectrum, Ispra, Italy, 23-24 November 2005 (invited).<br />
MA Green, “Thin-Film Solar Cells: Review <strong>of</strong> Technologies and Commercial Status”, 43rd<br />
Australian and <strong>New</strong> Zealand Solar Energy Society (ANZSES) Annual Conference, Dunedin,<br />
<strong>New</strong> Zealand, 28-30 November 2005.<br />
M.A. Green, “Solar and Wind: An Update”, Boat House III - Technology Futures, Paris, France,<br />
3-5 April 2006 (invited paper).<br />
M.A. Green, “Nanostructure Silicon Quantum Dot Tandem and Hot Carrier Solar Cells”,<br />
Materials Research Society (MRS) Spring Symposium, San Francisco, California, 17-21 April<br />
2006 (invited paper).<br />
M.A. Green, “Recent Developments and Future Prospects for Third Generation and Other<br />
Advanced Solar Cells”, IEEE 4th World Conference on Photovoltaic Energy Conversion<br />
(WCPEC-4), Hawaii, 7-12 May 2006 (invited keynote paper).<br />
M.A. Green, “Composite Semiconductors Based on Silicon Quantum Dots”, ACUN-5<br />
International Composites Conference: Advanced, Infrastructure, Natural and Nanocomposites,<br />
Sydney, Australia, 11-14 July 2006 (invited keynote paper).<br />
J.-F. Guillemoles, G. Conibeer, M. Green, “Phononic Engineering with Nanostructures for Hot<br />
Carrier solar cells” , 15th International Photovoltaic Science and Engineering Conference<br />
(PVSEC-15), Shanghai, China, 10-15 October 2005, p.375-376.<br />
J.-H. Guo, B.S. Tjahjono, and J.E. Cotter, “19.2% Effi ciency N-type Laser-grooved Silicon Solar<br />
Cells,” 31st IEEE Photovoltaic Specialists Conference, Orlando, USA, 3-7 January 2005,<br />
p.983-986.<br />
173
A.W.Y. Ho, S.R. Wenham, “Investigation <strong>of</strong> the Surface Passivation Quality <strong>of</strong> Various Novel<br />
Low Temperature Rear Localised Contacting Schemes for PV Devices”, Proc. Of 31st IEEE<br />
Photovoltaic Specialists Conference, Orlando, USA, 3-7 January 2005, p.1189-1192.<br />
D. Inns, A. Straub, M. Terry, Y. Huang, and A.G. Aberle, “Impact <strong>of</strong> growth and hydrogenation<br />
temperature on the voltage <strong>of</strong> ALICIA thin-fi lm silicon solar cells on glass”, 15th International<br />
Photovoltaic Science and Engineering Conference (PVSEC-15), Shanghai, China, 10-15<br />
October 2005, p.774-775.<br />
C.-W. Jiang, M.A. Green and G. Conibeer, “Effective Mass Calculations <strong>of</strong> Energy Bands for<br />
Silicon Quantum Dot Solar Cells in Oxide, Nitride and Carbide Matrices”, 15th International<br />
Photovoltaic Science and Engineering Conference (PVSEC-15), Shanghai, China, 10-15<br />
October 2005, p.158-159.<br />
O. Kunz, A. Straub, A. Sproul and A.G. Aberle, “Determination <strong>of</strong> the Two-Diode Model<br />
Parameters <strong>of</strong> Non-Ideal Solar Cells”, 20th European Photovoltaic Solar Energy Conference,<br />
Barcelona, Spain, 6-10 June 2005, p.226-229.<br />
L. Mai, S.R. Wenham, B. Tjahjono, J. Ji, Z. Shi, “<strong>New</strong> Emitter Design and Metal Contact for<br />
Screen-Printed Solar Cell Front Surfaces”, IEEE 4th World Conference on Photovoltaic Energy<br />
Conversion (WCPEC-4), Hawaii, 7-12 May 2006 (invited paper).<br />
R. Passey, I. MacGill and M. Watt, “Some options for State-based Renewable Obligations<br />
in Australia”, 43rd Australian and <strong>New</strong> Zealand Solar Energy Society (ANZSES) Annual<br />
Conference, Dunedin, <strong>New</strong> Zealand, 28-30 November 2005.<br />
S. Pillai, K.R. Catchpole and A. Shalav, “Targeting Better Absorption at Longer Wavelengths<br />
Using Surface Plasmons”, 31st IEEE Photovoltaic Specialists Conference, Orlando, USA, 3-7<br />
January 2005, p.171-174.<br />
S. Pillai, K.R Catchpole, T. Trupke, M.A. Green “Application <strong>of</strong> Surface Plasmons for Increased<br />
Light Emission from Silicon Based LEDs”, 20th European Photovoltaic Solar Energy Conference,<br />
Barcelona, Spain, 6-10 June 2005, p.369-372.<br />
S. Pillai, K.R. Catchpole, T. Trupke, G. Zhang, J. Zhao, M.A. Green, “Effects <strong>of</strong> Dielectric<br />
Overcoating on the Absorption Enhancement <strong>of</strong> SOI LEDs with metal island fi lms”, SPIE<br />
International Symposium on Microelectronics, MEMS, and Nanotechnology, Queensland Univ.<br />
<strong>of</strong> Technology, Brisbane, Australia, 11–15 December 2005.<br />
M. Pop, M.E Watt, J. Rivier and A. Birch, “Tariff Implications for the Value <strong>of</strong> PV to Residential<br />
Customers”, 43rd Australian and <strong>New</strong> Zealand Solar Energy Society (ANZSES) Annual<br />
Conference, Dunedin, <strong>New</strong> Zealand, 28-30 November 2005.<br />
J.W. Rudd, G.J. Conibeer, J-F. Guillemoles, “User Friendly S<strong>of</strong>tware for Calculation <strong>of</strong> Limiting<br />
Cell Effi ciencies for Research on Multijunction Solar Cells”, 43rd Australian and <strong>New</strong> Zealand<br />
Solar Energy Society (ANZSES) Annual Conference, Dunedin, <strong>New</strong> Zealand, 28-30 November<br />
2005.<br />
A. Shalav, B.S. Richards, K.W. Cramer and H.U. Gudel, “Improvements <strong>of</strong> an Up-Conversion<br />
NaYF4:Er3+ Phosphor-Silicon Solar Cell Device for an Enhanced Response in the Near-<br />
Infrared”, 31st IEEE Photovoltaic Specialists Conference, Orlando, USA, 3-7 January 2005,<br />
p.114-117.<br />
174
A. Shalav, B. Richards, G. Conibeer and M.A. Green, “Two colour excitation up-conversion<br />
effi ciency enhancement for a silicon photovoltaic device using Er3+-doped up-converting<br />
phosphors”, IEEE 4th World Conference on Photovoltaic Energy Conversion (WCPEC-4),<br />
Hawaii, 7-12 May 2006 (accepted).<br />
D. Song, P.I. Widenborg, A. Straub, P. Campbell, N. Chuangsuwanich, Y. Huang and A.G. Aberle,<br />
“EVA polycrystalline silicon thin-fi lm solar cells on textured glass”, 31st IEEE Photovoltaic<br />
Specialists Conference, Orlando, USA, 3-7 January 2005, p.1217-1220.<br />
D. Song, P.I. Widenborg, A. Straub, D. Ins, M. Terry and A.G. Aberle, “Impact <strong>of</strong> Glass Substrate<br />
Topography on EVA Polycrystalline Silicon Thin-Film Solar Cells on Glass”, 20th European<br />
Photovoltaic Solar Energy Conference, Barcelona, Spain, 6-10 June 2005, p.1104-1107.<br />
D. Song, P.I. Widenborg, M.L. Terry, and A.G. Aberle, “Impact <strong>of</strong> a SiN barrier layer on EVA<br />
polycrystalline silicon thin-fi lm solar cells”, 15th International Photovoltaic Science and<br />
Engineering Conference (PVSEC-15), Shanghai, China, 10-15 October 2005, p.186-187.<br />
D. Song, T.M. Walsh, and A.G. Aberle, “Monolithically Integrated Polycrystalline Silicon Thin-<br />
Film Mini-module on Glass”, IEEE 4th World Conference on Photovoltaic Energy Conversion<br />
(WCPEC-4), Hawaii, 7-12 May 2006 (accepted).<br />
A.B. Sproul, A.W.Y. Ho, A. Bruce and S.R. Wenham, “Student Installation <strong>of</strong> a PV Lighting<br />
System in a Rural Health Clinic”, 43rd Australian and <strong>New</strong> Zealand Solar Energy Society<br />
(ANZSES) Annual Conference, Dunedin, <strong>New</strong> Zealand, 28-30 November 2005.<br />
J. Stradal, G. Scholma, H. Li, C.H.M. van der Werf, J.K. Rath, P.I. Widenborg, P. Campbell, A.G.<br />
Aberle, and R.E.I. Schropp, “Hot Wire CVD for epitaxial thickening <strong>of</strong> polycrystalline silicon<br />
seed layers made by AIC on glass”, 31st IEEE Photovoltaic Specialists Conference, Orlando,<br />
USA, 3-7 January 2005, p.1540-1543.<br />
A. Straub, D. Inns, O. Kunz, M.L. Terry, P.I. Widenborg, A.B. Sproul, and A.G. Aberle, “Towards<br />
400 mV ALICIA thin-fi lm silicon solar cells on glass”, 31st IEEE Photovoltaic Specialists<br />
Conference, Orlando, USA, 3-7 January 2005, p.1221-1224.<br />
A. Straub, R. Gebs, A.B. Sproul, R. Bardos, D. Inns, M.L. Terry and A.G. Aberle, “C-V<br />
Measurements by Impedance Analysis on ALICIA Polycrystalline Silicon Thin-Film Solar Cells<br />
on Glass”, 20th European Photovoltaic Solar Energy Conference, Barcelona, Spain, 6-10 June<br />
2005, p.128-131.<br />
Y. Takeda, T. Ito, T. Motohiro, T. Nagashima, D. König, G. Conibeer, “A novel method based<br />
on oblique depositions to fabricate quantum dot arrays”, IEEE 4th World Conference on<br />
Photovoltaic Energy Conversion (WCPEC-4), Hawaii, 7-12 May 2006 (accepted).<br />
Y. Takeda, T. Ito, T. Motohiro, T. Nagashima, D. König, G. Conibeer, “Formation <strong>of</strong> quantum dot<br />
arrays utilizing oblique depositions”, 2006 Spring the Japan Society <strong>of</strong> Applied Physics annual<br />
meeting, Tokyo, Japan, March 2006 (accepted).<br />
M.L. Terry, A. Straub, D. Inns, D. Song, and A.G. Aberle, “Voc improvement <strong>of</strong> evaporated SPC<br />
thin-fi lm Si solar cells on glass by rapid thermal annealing”, 31st IEEE Photovoltaic Specialists<br />
Conference, Orlando, USA, 3-7 January 2005, p.971-974.<br />
175
M.L. Terry, P.I. Widenborg, O. Kunz, and A.G. Aberle, “Effects <strong>of</strong> rapid thermal annealing and<br />
hydrogen passivation on crystalline silicon thin-fi lm solar cells on glass made by PECVD solidphase<br />
crystallization”, 15th International Photovoltaic Science and Engineering Conference<br />
(PVSEC-15), Shanghai, China, 10-15 October 2005, p.905-906.<br />
M.L. Terry, D. Inns and A.G. Aberle, “Massive improvement through rapid thermal annealing<br />
and hydrogen passivation <strong>of</strong> poly-Si thin-fi lm solar cells on glass based on aluminum induced<br />
crystallization”, IEEE 4th World Conference on Photovoltaic Energy Conversion (WCPEC-4),<br />
Hawaii, 7-12 May 2006 (accepted).<br />
T. Trupke and R.A. Bardos, “Photoluminescence: A surprisingly sensitive lifetime technique”,<br />
31st IEEE Photovoltaic Specialists Conference, Orlando, USA, 3-7 January 2005, p.903-<br />
906.<br />
T. Trupke and R.A. Bardos, “Photoluminescence: A versatile characterization technique for<br />
crystalline silicon”, 15th workshop on Crystalline Silicon Solar cells & Modules: Materials and<br />
Processes, Vail, USA, 7-10 August 2005.<br />
T. Trupke, R.A. Bardos, M.D. Abbott, F.W. Chen, J.E. Cotter, A. Lorenz, ”Fast photoluminescence<br />
imaging <strong>of</strong> silicon wafers”, IEEE 4th World Conference on Photovoltaic Energy Conversion<br />
(WCPEC-4), Hawaii, 7-12 May 2006 (accepted).<br />
T.M. Walsh, S.R. Wenham, and A.G. Aberle, “Novel method for the interconnection <strong>of</strong> thin-fi lm<br />
silicon solar cells on glass”, 31st IEEE Photovoltaic Specialists Conference, Orlando, USA, 3-7<br />
January 2005, p.1229-1232.<br />
T.M. Walsh, D. Song, S. Motahar, and A.G. Aberle, “Self-aligning maskless photolithography<br />
method for metallising thin-fi lm crystalline silicon solar cells on transparent supporting<br />
materials”, 15th International Photovoltaic Science and Engineering Conference (PVSEC-15),<br />
Shanghai, China, 10-15 October 2005, p.706-707.<br />
M. Watt, “Status <strong>of</strong> Australia’s PV Market”, Sustainable Energy 2005, the BCSE Annual<br />
Conference, 27-29 April 2005, Melbourne.<br />
M. Watt, “National Survey <strong>Report</strong> <strong>of</strong> PV Power Applications in Australia – 2004”, Australian<br />
Photovoltaic Power Systems (PVPS) Consortium, May 2005.<br />
M. Watt, S. Healy and R. Passey, “Some considerations when designing a retail energy<br />
market that works for Energy Effi ciency and distributed generation”, BCSE Energy Effi ciency<br />
Conference, Melbourne, Australia, 22-23 November 2005 (invited).<br />
S.R. Wenham, L. Mai, B. Tjahjono, J.Jia and Z. Shi, “Innovative emitter design and metal contact<br />
for screen printed solar cells”, 15th International Photovoltaic Science and Engineering<br />
Conference (PVSEC-15), Shanghai, China, 10-15 October 2005, p.805-805.<br />
P.I. Widenborg, T. Puzzer, J. Stradal, D.H. Neuhaus, D. Inns, A. Straub, and A.G. Aberle,<br />
“Structural quality <strong>of</strong> smooth AIC poly-Si fi lms on glass substrates”, 31st IEEE Photovoltaic<br />
Specialists Conference, Orlando, USA, 3-7 January 2005, p.1031-1034.<br />
P.I. Widenborg, J. Weber, M.L. Terry and A.G. Aberle, “Solid Phase Epitaxy <strong>of</strong> PECVD a-Si:H n<br />
AIC Poly-Si Seeded Glass Substrates”, 20th European Photovoltaic Solar Energy Conference,<br />
Barcelona, Spain, 6-10 June 2005, p.1000-1003.<br />
176
P.I. Widenborg, S.V. Chan, D. Inns, and A.G. Aberle, “Fabrication <strong>of</strong> poly-Si seed layer by<br />
aluminium-induced crystallisation <strong>of</strong> a-Si deposited by PECVD or sputtering”, 15th International<br />
Photovoltaic Science and Engineering Conference (PVSEC-15), Shanghai, China, 10-15<br />
October 2005, p.814-815.<br />
Per I. Widenborg, J.W. Weber, M.L. Terry, and A.G. Aberle, “A comparative study <strong>of</strong> poly-Si fi lms<br />
fabricated on glass by crystallisation <strong>of</strong> PECVD a-Si:H via the SPC and SPE methods”, 15th<br />
International Photovoltaic Science and Engineering Conference (PVSEC-15), Shanghai, China,<br />
10-15 October 2005, p.420-421.<br />
G. Yao, J. Zhao and J.E. Cotter, “Metal Stencil-Printed Selective Emitter Silicon Solar Cell<br />
Fabrication on CZ Wafer”, 20th European Photovoltaic Solar Energy Conference, Barcelona,<br />
Spain, 6-10 June 2005, p.749-752.<br />
G. Yao, S.R. Wenham and J.E. Cotter, “Comparison <strong>of</strong> Stencil and Screen Printed Solar Cell”,<br />
15th International Photovoltaic Science and Engineering Conference (PVSEC-15), Shanghai,<br />
China, 10-15 October 2005, p.299-300.<br />
J. Zhao and A. Wang, “High Effi ciency Real Emitter PERT Solar Cells on n-type FZ Single<br />
Crystalline Silicon Substrates”, 20th European Photovoltaic Solar Energy Conference,<br />
Barcelona, Spain, 6-10 June 2005, p.806-809.<br />
Jianhua Zhao, “The Foundation and the Strategy <strong>of</strong> Photovoltaic Industry in China”, China<br />
Power Conference, Beijing, China, 10-12 August 2005.<br />
J. Zhao, “High Effi ciency Silicon Solar Cell Technologies Ready for Production”, 15th<br />
International Photovoltaic Science and Engineering Conference (PVSEC-15), Shanghai, China,<br />
10-15 October 2005, p.310-311 (invited talk).<br />
J. Zhao and A. Wang, “A Novel Rear Emitter N-Type PERT Silicon Cells with Stable High<br />
Performance”, 15th International Photovoltaic Science and Engineering Conference (PVSEC-<br />
15), Shanghai, China, 10-15 October 2005, p.122-123 (invited talk).<br />
J. Zhao, “High Effi ciency Silicon Solar Cells for Mass Production”, Sino-Euro Workshop, 15th<br />
International Photovoltaic Science and Engineering Conference, Shanghai, 10 October 2005<br />
(invited talk).<br />
J. Zhao, “Manufacturing Technologies for High Effi ciency Crystalline Silicon Solar Cells”, PV<br />
Industrial Workshop, 15th International Photovoltaic Science and Engineering Conference,<br />
Shanghai, 12 October 2005 (invited talk).<br />
177
Press<br />
Richard Corkish, “Fusion fallacies”, (letter to editor re nuclear fusion) <strong>New</strong> Scientist, No. 2528,<br />
p.24, 3 Dec 2005.<br />
Other <strong>Report</strong>s<br />
R. Corkish, “Photovoltaics and Renewable Energy Education and Research at the <strong>University</strong><br />
<strong>of</strong> <strong>New</strong> <strong>South</strong> <strong>Wales</strong>”, presentation to NSW Department <strong>of</strong> State and Regional Development,<br />
UNSW, 12 April 2005 (unpublished).<br />
R. Corkish, “Photovoltaics & Silicon Photonics Research at UNSW”, presentation at UNSW-<br />
Singapore Research Collaboration Workshop, A*STAR Institute <strong>of</strong> Materials Research and<br />
Engineering, Singapore 28 February – 2 March 2005 (invited).<br />
R .Corkish, S.R. Wenham, A.B. Sproul, J. Cotter, C.B. Honsberg, M.A. Green, A.G. Aberle, G.<br />
Conibeer, “Photovoltaics and Renewable Energy Education and Research at the <strong>University</strong> <strong>of</strong><br />
<strong>New</strong> <strong>South</strong> <strong>Wales</strong>”, presentation at The Joint Graduate School <strong>of</strong> Energy and Environment,<br />
King Mongkut’s <strong>University</strong> <strong>of</strong> Technology Thonburi, Bangkok, 25 February 2005. (invited).<br />
R .Corkish, S.R. Wenham, A.B. Sproul, J. Cotter, C.B. Honsberg, M.A. Green, A.G. Aberle, G.<br />
Conibeer, “Photovoltaics and Renewable Energy Education and Research at the <strong>University</strong><br />
<strong>of</strong> <strong>New</strong> <strong>South</strong> <strong>Wales</strong>”, presentation at Department <strong>of</strong> Physics, Imperial College <strong>of</strong> Science,<br />
Technology and Medicine, London, 22 February 2005. (invited).<br />
M.A. Green, “Research Opportunities with CSG Solar Material”, Forschungszentrum Jülich,<br />
Germany, February 2005.<br />
R. Passey, M. Watt and K. Nolles, CEEM Submission on the Green Power Options Paper, June<br />
2005.<br />
Dr R. Passey, I. MacGill, H. Outhred, R. Corkish, Submission on “Proposal to increase MEPS<br />
for Room Air Conditioners and harmonise MEPS for single and three-phase units”, Regulatory<br />
Impact Statement, Draft fi nal report – 21 June 2005, Australian Greenhouse Offi ce, July<br />
2005. http://www.ceem.unsw.edu.au/documents/ceem-cpveacmepssubmission.pdf<br />
M. Watt, M. Oliphant, H. Outhred, I. MacGill, E. Spooner and S. Partlin, “Photovoltaics and Peak<br />
Electricity Loads, Summer 2003-04”, Final <strong>Report</strong> for the BCSE, May 2005.<br />
178
This report is available from:<br />
ARC Centre <strong>of</strong> Excellence<br />
for Advanced Silicon Photovoltaics<br />
and Photonics<br />
<strong>University</strong> <strong>of</strong> <strong>New</strong> <strong>South</strong> <strong>Wales</strong><br />
Sydney NSW 2052, Australia<br />
Phone:<br />
+61 2 9385 4018<br />
Fax:<br />
+61 2 9662 4240<br />
Email:<br />
s.wenham@unsw.edu.au<br />
Web:<br />
http://www.pv.unsw.edu.au<br />
Director:<br />
Stuart Wenham<br />
Executive Research Director:<br />
Martin Green<br />
<strong>Report</strong> Coordinators:<br />
Martin Green, Stuart Wenham<br />
and Jenny Hansen<br />
Graphic Artist:<br />
Roman Balla<br />
rballa@bigpond.net.au<br />
Photos, figures and graphs:<br />
Courtesy <strong>of</strong> Centre Staff