PHYS07200604023 Prabodh Kumar Kuiri - Homi Bhabha National ...

ENERGETIC Au IRRADIATION EFFECTS ON
NANOCRYSTALLINE ZnS FILMS DEPOSITED
ON Si AND Au NANOPARTICLES EMBEDDED
IN SILICA GLASS
By
PROBODH KUMAR KUIRI
INSTITUTE OF PHYSICS
BHUBANESWAR, INDIA
A THESIS SUBMITTED TO THE
BOARD OF STUDIES IN PHYSICAL SCIENCES
IN PARTIAL FULFILLMENT OF REQUIREMENTS
FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
OF
HOMI BHABHA NATIONAL INSTITUTE
April 2010

Statement by Author
This dissertation has been submitted in partial fulfillment of requirements for an
advanced degree at Homi Bhabha National Institute (HBNI) and is deposited in the
Library to be made available to borrowers under rules of the HBNI.
Brief quotations from this dissertation are allowable without special permission,
provided that accurate acknowledgement of source is made. Requests for permission
for extended quotation from or reproduction of this manuscript in whole or in part
may be granted by the Competent Authority of HBNI when in his or her judgment the
proposed use of the material is in the interests of scholarship. In all other instances,
however, permission must be obtained from the author.
(Probodh Kumar Kuiri)
ii

Declaration
I, hereby declare that the investigation presented in the thesis has been carried out
by me. The work is original and has not been submitted earlier as a whole or in part
for a degree / diploma at this or any other Institution / University.
(Probodh Kumar Kuiri)
iii

.
To My PARENTS
iv

Acknowledgements
I take this opportunity to express my sincer gratitude to my advisor, Prof. D. P. Mahapatra,
for his constant encouragement, support, and invaluable guidance throughout
my research period at the Institute of Physics. I learnt many things from his love for
Physics, intuative way of looking into the problems, and constant drive for perfection.
He has helped me grow both academically and personally. I am also thankful to him
for giving me complete freedom in my academic and social life during the period of
this work. Without his help this dissertation would have not been possible. I heartly
thank him and his family for making my stay valuable in the Institute.
I am grateful to Dr. D. Kanjilal and Prof. N. C. Mishra for their encouragement,
suggestions, and help. It was always a nice experience while discussing physics with
them. My sincer thanks to Dr. B. S. Acharya, Prof. S. M. Bhattacharyya, Prof.
G. Tripathy, Dr. A. Tripathi, Prof. S. K. Pal, and Dr. G. Kuri for their effective
discussions and help received at various stages of this work.
I have spent many enjoyable hours working with my fellow group members. I am
happy to thank them all, both past and present, for their support and friendship in
and out of the lab., including Dr. S. Mohapatra, Dr. B. Joseph, Mr. H. P. Lenka, Dr.
J. Ghatak, and Mrs. G. Sahu.
I take this opportunity to thank all the experimental condensed matter group
faculty members of the Institute of Physics (IOP), Bhubaneswar for their generous
support in extending different facilities. Special thanks are to Prof. P. V. Satyam,
Prof. S. N. Sahu, and Prof. S. Varma in this regard. I am grateful to Mr. A.
K. Behera, a person who possesses a vast amount of practical knowledge. I learned
something new every time we worked in lab. Sincere thanks to Mrs. R. Dash, Mr. M.
Baliarsingh, and Mr. D. B. Sahi for providing me a homely atmosphere in our lab. I
thank Dr. S. N. Saragni for doing several optical measurements for me. I thank all the
accelerator and other technical staff members of IOP and Inter University Accelerator
Centre, New Delhi for their help during carrying out different experiments. I would
v

like to specially mention the names of M. Majhi, P. K. Biswal, A. Sahu, K. C. Patra,
S. K. Choudhary, A. K. Dash, and P. C. Marandi. I acknowledge the help received
from all scholars of the experimental condensed matter group whenever I was in need.
It is my pleasure to thank my seniours, scholars, friends, and others who have
made my stay in the beautiful campus of IOP a memorable and pleasing one. Special
thanks to Sumalay-da, Jay, Prabir, Anupam, and Kuntala-di for their encouragements,
help, company, and standing by me in good and bad times. I also thank Sandip-da,
Anand, Dipak (Mishra), Raghu, Srikumar-da, Tanay-da, Manas, Sadhana, Ajay, and
Chitrasen.
I would like to thank my friends and others, outside IOP, who have helped me in
one way or other. Special thanks to my friend-cum-elder brother, Manoranjan-da for
his valuable advice, useful comments, and help whenever I needed.
My sincer regards and respect are due to all my teachers, especially Mr. B. B.
Mishra, who had taught me at various stages of my educational career. I am grateful
to late M. M. Kuiri, Dr. S. Chattopadhyay, Prof. S. K. Pradhan, and Prof. A. K.
Bhattacharyya for their encouragement to pursue research work.
I am indebted to the former director, Dr. Y. P. Viogy, the present director, Prof.
A. Khare of IOP, the principal Dr. D. N. Mahata of A. M. College, Jhalda, and all
the members (academic and non-academic) of both the institutions for their timely
help, cooperation, and support whenever needed. Special thanks are to all the people
who have maintained the excellent library and computer facilities at IOP.
Finally, my sincer gratitude to my grand mother, parents, in-laws, wife, and brothers
for their love and unwavering support. They have always been around when I
needed them and they continue to support me in my decisions. I would have never
made it this far without them.
Date : April 6, 2010
Bhubaneswar
(Probodh Kumar Kuiri)
vi

List of Publications / Preprints
In journals:
1 Low energy O induced redistribution of nanosized gold inclusions in SiO 2 grown
on Si(100),
B. Joseph, S. Mohapatra, B. Satpati, H. P. Lenka, P. K. Kuiri and D. P.
Mahapatra,
Nucl. Instrum. Methods Phys. Res. B 227, 559 (2005).
2 Size saturation in low energy ion beam synthesized Au nanoclusters and their
size redistribution with O irradiation,
B. Joseph, S. Mohapatra, H. P. Lenka, P. K. Kuiri and D. P. Mahapatra,
Thin Solid Films 492, 35 (2005).
*3 Observation of ZnS nanoparticles sputtered from ZnS films under 2 MeV Au
irradiation,
P. K. Kuiri, B. Joseph, J. Ghatak, H. P. Lenka, G. Sahu, B. S. Acharya and
D. P. Mahapatra,
Nucl. Instrum. Methods Phys. Res. B 248, 25 (2006).
*4 Effect of Au irradiation energy on ejection of ZnS nanoparticles from ZnS film,
P. K. Kuiri, J. Ghatak, B. Joseph, H. P. Lenka, G. Sahu, N. C. Mishra, A.
Tripati, D. Kanjilal and D. P. Mahapatra,
J. Appl. Phys. 101, 014313 (2007).
5 Effect of 100 MeV Au irradiation on embedded Au nanoclusters in silica glass,
B. Joseph, J. Ghatak, H. P. Lenka, P. K. Kuiri, G. Sahu, N. C. Mishra and D.
P. Mahapatra, Nucl. Instrum. Methods Phys. Res. B 256, 659 (2007).
6 Low energy C n cluster ion induced damage effects in Si(100) substrates,
H. P. Lenka, B. Joseph, P. K. Kuiri, G. Sahu and D. P. Mahapatra,
Nucl. Instrum. Methods Phys. Res. B 256 , 665 (2007).
vii

7 Synthesis of alloy metal nanoclusters in silica glass by sequential ion implantation,
B. Joseph, H. P. Lenka, P. K. Kuiri, D. P. Mahapatra and R. Kesavamoorthy,
Int. J. Nanoscience. 6, 432 (2007).
8 Anomalous diffusion of Au in mega-electron-volt Au implanted SiO 2 /Si(100),
S. Mohapatra, J. Ghatak, B. Joseph, H. P. Lenka, P. K. Kuiri and D. P.
Mahapatra,
J. Appl. Phys. 101, 063542 (2007).
9 Formation and growth of SnO 2 nanoparticles in silica glass by Sn implantation
and annealing,
P. K. Kuiri, H. P. Lenka, J. Ghatak, G. Sahu, B. Joseph and D. P. Mahapatra,
J. Appl. Phys. 102, 024315 (2007).
10 MeV Au irradiation induced nanoparticle formation and recrystallization in a
low energy Au implanted Si layer,
G. Sahu, B. Joseph, H. P. Lenka, P. K. Kuiri, A. Pradhan and D. P. Mahapatra,
Nanotechnology 18, 495702 (2007).
*11 Observation of a universal aggregation mechanism and a possible phase transition
in Au sputtered by swift heavy ions,
P. K. Kuiri, B. Joseph, H. P. Lenka, G. Sahu, J. Ghatak, D. Kanjilal and D.
P. Mahapatra,
Phys. Rev. Lett. 100, 245501 (2008).
12 Study of low energy Si − 5 and Cs − implantation induced amorphization effects in
Si(100),
H. P. Lenka, B. Joseph, P. K. Kuiri, G. Sahu, P. Mishra, D. Ghose and D. P.
Mahapatra,
J. Phys. D.: Appl. Phys. 41, 215305 (2008).
13 The mechanism of ion induced amorphization in Si,
H. P. Lenka, U. M. Bhatta, P. K. Kuiri, G. Sahu, B. Joseph, B. Satpati and
D. P. Mahapatra,
arXiv:0811.0806v4 [cond-mat.mtrl-sci] (2008).
14 Effect of oxidation temperature on the photoluminescence of Sn implanted and
annealed SiO 2 /Si,
viii

P. K. Kuiri and M Ghosh,
Adv. Sci. Lett. 2, 35 (2009).
*15 Tuning size and shape of Au nanoparticles embedded in silica glass by swift heavy
ion irradiation,
P. K. Kuiri, B. Joseph, J. Ghatak, H. P. Lenka and G. Sahu
Adv. Sci. Lett. (accepted) (2010).
*16 Modification of nanocrystalline ZnS thin films by Au ion irradiation: Role of
energy loss processes,
P. K. Kuiri, H. P. Lenka, G. Sahu, D. Kanjilal and D. P. Mahapatra,
(to be published).
In proceedings:
1 A low energy negative ion implanter for surface modification studies at IOP,
Bhubaneswar,
B. Joseph, A. K. Behera, S. Mohapatra, P. Kuiri, and D. P. Mahapatra,
Solid State Phys. (India) 46, 271 (2003).
2 Study of Au-Ag alloy nanoparticles in silica glass synthesized by sequential low
energy ion implantation,
B. Joseph, S. Mohapatra, H. P. Lenka, P. K. Kuiri, and D. P. Mahapatra,
Proceedings of the DAE Symposium on Solid State Phys., Amritsar, India, December
2004.
——————————————————————-
* indicates papers on which this thesis is based.
ix

Synopsis
A major part of the modern technology depends on materials with precise controlled
properties. Ion implantation is a favoured method to achieve controlled modification
of surface and near-surface regions of various materials [1]. It is a key doping technique
in semiconductor processing, and it is used to induce phase formation, amorphization,
grain growth, structural transformations, tailoring optical band gap or mixing in a
wide range of materials. Also energetic ion irradiation of solids leads to the ejection
of atoms, molecules, and sometimes clusters from the surface of the target. The
observation of stable, intact clusters of atoms and molecules, due to ion impact [2],
is rather surprising because the energies involved in the ion-target collisions, that
generate sputtered atoms, are typically much larger than the binding energies (∼ 1–2
eV) holding the clusters together. Hence it is difficult to understand how stable clusters
can form during ion sputtering. Eversince the first observation, the fundamental
question of the cluster emission mechanism has attracted considerable attention both
experimentally [3] and theoretically [4,5,6].
As an ion penetrates a solid, it loses energy due to two distinct interactions [1].
At low energies, it loses energy by elastic, nuclear collisions with the target atoms
described by the well known Rutherford cross sections for the Coulomb interaction
between the charged atomic nuclei. At higher energies, it loses energy inelastically
by exciting and ionizing electrons of target atoms. Depending upon the incident ion
energy and ion-solid combinations, there can be production of elastic [7] or inelastic [8]
thermal spikes leading to local temperature rise well above the melting temperature of
the materials. In fact, inelastic thermal spikes can sometimes result in the formation of
a cylindrically-shaped molten region around the ion track. These are typically several
nm in diameter [8]. This process can therefore bring about structural changes in the
target material. Here, we have utilized energetic ion beams for controlled modifications
of nanocrystalline ZnS films and Au nanoparticles (NPs). We have also looked at
ejection of NPs from these materials. Attempts have been made to understand the
x

underlying physics through a study of the size distributions of ejected particles.
We have studied the effect of Au irradiation on structural, optical, and surface
morphological properties of nanocrystalline ZnS films deposited on Si and silica glass
substrates [9], using a simple thermal evaporation of ZnS powder. The as-deposited
films were of wurtzite structures with an optical band gap of 3.56 eV. X-ray diffraction
measurements indicated the presence of compressive stress with a grain size of ∼ 17.5
nm. Irradiation at MeV energies have been found to result in a considerable grain
growth and a decrease in dislocation density. In addition, the irradiated films were
found to be under tensile stress. The tensile stress produced in the film is found to
be much higher at 100 MeV than at 2 MeV. There is also a change in optical band
gap primarily due to irradiation induced changes of the grain size and the formation
of trap levels. On the other hand no observable changes in grain size or optical band
gap could be seen following a lower energy (35 keV) Au irradiation. The samples
have also been studied using atomic force microscopy. The 35 keV and 2 MeV Au
irradiations are found to produce rough surfaces the characteristics of which have
been studied through the analysis of the power spectral density. At large spatial
frequency, q, (at small length scales) the Fourier exponent, γ (determined from the
fall off going as q −γ ) has been found to have values close to 2.80, 3.11, 3.72, and 2.85
for the as-deposited and the samples irradiated with Au at 35 keV, 2 MeV, and 100
MeV, respectively. The lower values close to 3 could be understood in a model where
evaporation and condensation dominate, the higher value corresponding to a surface
diffusion dominated process.
The effect of Au irradiation on isolated and embedded nanostructures system [10]
has also been studied. For this, spherical Au NPs (of average size 7 nm) were synthesized
in the near surface region of silica glass samples by implanting 32 keV Au − ions
to a fluence of 4×10 16 cm −2 , followed by annealing at 850 ◦ C in air for 1 h. The silica
glass samples so prepared, containing Au NPs are found to show a surface plasmon
resonance (SPR) band centered around 543 nm in the optical absorption spectrum.
The silica glass samples with embedded NPs were subjected to Au irradiations (at
room temperature) at 10 and 100 MeV. The higher energy irradiation was found to
result in an elongation of the embedded Au NPs along the beam direction. At higher
irradiation fluence there was Au loss from the silica matrix. Up to a fluence of 5×10 13
ions cm −2 , the smaller NPs (diameter < 9 nm) are found to grow in size while larger
ones deformed anisotropically along the ion beam direction. The anisotropy in the
larger NPs have been seen to increase with increase in ion fluence. Optical absorption
xi

spectra corresponding to the elongated Au NPs show two SPR bands corresponding to
transverse and longitudinal modes. Further, longitudinal SPR band has been seen to
shift towards higher wavelength with increase in fluence. This shift of the longitudinal
SPR band is a direct consequence of the further elongation of the NPs [11]. However,
in the sample irradiated at a fluence of 1×10 14 ions cm −2 , the Au NPs are seen to be
almost spherical in shape, with a large interparticle separation. The average size is
found to be ∼ 9.2 nm, which is little larger than that of the unirradiated samples. In
fact what is found is a two-dimensional array of small Au NPs, very near the silica
surface which result in an SPR band with a single peak at around 561 nm. This SPR
peak position appeared to be red-shifted as compared to that of the as-grown sample.
Using a simple simulation [12] that uses, NP size, density, and dielectric constant of
the host together with optical density of the NPs, it has been shown that the shift of
the SPR band corresponding to the spherical NPs is, mainly, the result of the change
in matrix dielectric constant, due to the dissolution of Au in the form of atom and/or
small clusters, instead of the growth of the NPs after ion irradiation.
The results on change of shape (from nearly spherical to elongated ones) from
ion irradiation, can be understood within the framework of the inelastic thermal spike
induced ion track formation [8]. Smaller NPs within the ion track completely evaporate
and collect around existing small clusters leading Oswald ripening of small NPs. In
any case they can grow to a size comparable to the track diameter, which is about
8.5 nm in silica glass in the present experimental conditions [8]. On the other hand
larger ones can undergo a melting and due to pressure imbalance can get squeezed
to elongated shapes along the ion tracks which are formed along the beam direction.
The material loss at higher fluence occurs from a squeezing out and vaporization of
NPs formed near the surface.
Following the above study, it was also necessary to see whether the material which
is ejected during ion irradiation really contained any nanoclusters. This is because
the size distribution in that case can be connected to the underlying mechanism of
emission. Therefore, in further pursuance of our investigations we have collected the
ejected particles on catcher foils placed suitably in the reflection geometry with an
aim to look at their size distribution. This has been done in both cases concerning
irradiation of ZnS films and embedded Au NPs in silica glass as mentioned earlier.
In each case the size distribution has been determined from transmission electron
microscopy.
For ZnS [13,14], with 35 keV Au irradiation no NP larger than 1 nm could be
xii

observed on the catcher foil and therefore no size distribution study could be carried
out. However, we have observed wurtzite ZnS NPs on the catcher foils due to MeV Au
irradiations. The ejected NPs have been found to have sizes (diameter) lying between 2
to 7 nm. The most probable size is seen to be around 3−3.5 nm. For particle sizes ≥ 3
nm, the distributions show a power law decay in the form of Y (n) ∼ n −δ ; n and δ being
the number of atoms in a NP and the power law exponent, respectively. In case of Au
irradiation at 2 MeV, the values of exponent, δ, are found to be around 2.60 ± 0.09,
whereas the same for 100 MeV irradiation is found to be around 3.45 ± 0.09. After
correcting the results for cluster breakup effects, the values of power law exponent
reduced to about 2.0 and 2.6 for the above two irradiation energies. A δ value close
to 2 indicates that the NPs are produced when shock waves, generated by subsurface
displacement cascades, ablate the surface [4]. On the other hand, Coulomb explosion
followed by thermal spike [15] induced vaporization of ZnS seems to be the dominant
mechanism regarding material removal at 100 MeV. In such a case the evaporated
material can cool down going into the fragmentation region ejecting clusters with a
power law exponent close to 7/3 [5]. Interestingly enough, no NP could be found
in the catcher foil even during the 100 MeV Au irradiation when the sample was
maintained at liquid nitrogen temperature. This suppression on cluster ejection at low
temperature can be explained as due to an enhanced thermal conductivity resulting
in fast heat dissipation that suppresses thermal spike effects.
However, the scenario become completely different in case of sputtering from embedded
nanostructures [16]. For 10 MeV Au irradiation of Au NPs, embedded in
silica glass, no NP could be observed on the catcher foils. However, crystalline Au
NPs of size 1−20 nm could be seen on the catcher foils for 100 MeV Au irradiation
case. Size distribution, of the ejected NPs, has been seen to follow a similar power
law decay as seen in the sputtered ZnS NPs. Here, we have seen the existence of two
power law exponents: 3/2 for smaller NPs, below 12.5 nm in size, and 7/2 for larger
NPs. The first case can be rationalized as occurring from a steady state aggregation
process [17], independent of cluster size. The later case may come from a dynamical
transition to another steady state where aggregation and evaporation rates are size
dependent. To understand the aggregation of clusters during sputtering at very high
energy (100 MeV), inelastic thermal spike [8] concept has been invoked. Production
of thermal spike in silica, due to 100 MeV Au ions, can lead to a local temperature
rise well above the vaporization temperature of Au. In such a case evaporated Au
clusters can exchange particle between nucleation sites, within the framework of a
xiii

mass-aggregation model, that takes diffusion, aggregation on contact, and dissociation.
This can result in a size independent steady state aggregation process with a δ
value of 3/2 [17]. This happens to be a very general case corresponding to a broad
class of phenomena, such as fish schooling, distribution of wealth, cloud formation,
and polymer gels. On the other hand, a phase transition can occur when the aggregation
and evaporation rates become size dependent leading to a δ value of 7/2 at the
critical point [18]. Since the system indicates a steady state scenario there is no need
to correct the exponent against any breakup effects as applicable for the ZnS NPs
ejected during Au irradiations.
REFERENCES
[1] M. Nastasi, J.W. Mayer and J.K. Hirvonen, Ion-Solid Interactions: Fundamentals and
Applications, Cambridge University Press, (1996).
[2] R.E. Honig, J. Appl. Phys. 29, 549 (1958).
[3] R. Birtcher, S.E. Donnely and K. Nordlund (E. Knystautas, Ed.), Engineering Thin Films
and Nanostructures with Ion Beams, Taylor & Francis, Florida, (2005), Ch. 1.
[4] I.S. Bitensky and E.S. Parilis, Nucl. Instrum. Methods Phys. Res. B 21, 26 (1987).
[5] H.M. Urbassek, Nucl. Instrum. Methods Phys. Res. B 31, 541 (1988).
[6] K.O.E. Henriksson, K. Nordlund and J. Keinonen, Phys. Rev. B 71, 014117 (2005).
[7] J.A. Brinkman, Am. J. Phys. 24, 246 (1956).
[8] M. Toulemonde, J.M. Costantini, Ch. Dufour, A. Meftah, E. Paumier and F. Studer,
Nucl. Instrum. Methods Phys. Res. B 116, 37 (1996).
[9] P.K. Kuiri, H.P. Lenka, G. Sahu, D. Kanjilal and D.P. Mahapatra, [to be published].
[10] P.K. Kuiri, B. Joseph, J. Ghatak, H.P. Lenka and G. Sahu, Adv. Sci. Lett. (accepted)
(2010).
[11] W. Kreibig and M. Vollmer, Optical Properties of Metal Clusters (Springer Series in
Materials Science Vol. 25) Springer, Berlin (1995).
[12] M.A. Gracia, J. Llopis and S.E. Paje, Chem. Phys. Lett. 315, 313 (1999).
[13] P.K. Kuiri, B. Joseph, J. Ghatak, H.P. Lenka, G. Sahu, B.S. Acharya and D.P. Mahapatra,
Nucl. Instrum. Methods Phys. Res. B 248, 25 (2006).
[14] P.K. Kuiri, J. Ghatak, B. Joseph, H.P. Lenka, G. Sahu, D.P. Mahapatra, A. Tripathi,
D. Kanjilal and N.C. Mishra, J. Appl. Phys. 101, 014313 (2007).
[15] E.M. Bringa and R.E. Johnson, Phys. Rev. Lett. 88, 165501 (2002).
[16] P.K. Kuiri, B. Joseph, H.P. Lenka, G. Sahu, J. Ghatak, D. Kanjilal and D.P. Mahapatra,
Phys. Rev. Lett. 100, 245501 (2008).
[17] H. Takayasu, Phys. Rev. Lett. 63, 2563 (1989).
[18] R.D. Vigil, R.M. Ziff and B. Lu, Phys. Rev. B 38, 942 (1988).
xiv

Contents
Statement by Author
Declaration
Acknowledgements
List of Publications / Preprints
Synopsis
ii
iii
v
vii
x
1 Introduction 1
1.1 Interests and applications of ion beams in materials science . . . . . . . 1
1.2 Ion-solid interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.1 Basic concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.2 Ion penetration and stopping . . . . . . . . . . . . . . . . . . . 5
1.2.3 Nuclear thermal spike . . . . . . . . . . . . . . . . . . . . . . . 8
1.2.4 Coulomb explosion and electronic thermal spike . . . . . . . . . 9
1.3 Sputtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.3.1 Sputtering mechanisms . . . . . . . . . . . . . . . . . . . . . . . 12
1.3.2 Size distribution of clusters ejected during ion irradiation . . . . 13
2 Experimental techniques 19
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.2 Sample preparation techniques . . . . . . . . . . . . . . . . . . . . . . . 19
2.2.1 Vacuum evaporation thin film deposition setup . . . . . . . . . . 19
2.2.2 Ion implantation and irradiation . . . . . . . . . . . . . . . . . . 20
2.3 Characterization techniques . . . . . . . . . . . . . . . . . . . . . . . . 23
2.3.1 Rutherford backscattering spectrometry . . . . . . . . . . . . . 23
2.3.2 X-ray diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . 27
xv

2.3.3 Transmission electron microscopy . . . . . . . . . . . . . . . . . 29
2.3.4 Atomic force microscopy . . . . . . . . . . . . . . . . . . . . . . 32
2.3.5 Optical absorption . . . . . . . . . . . . . . . . . . . . . . . . . 33
3 Modification of nanocrystalline ZnS films by Au irradiation 38
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.2 Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.3 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.3.1 Energy loss processes . . . . . . . . . . . . . . . . . . . . . . . 40
3.3.2 Thickness and composition of the ZnS films . . . . . . . . . . . 41
3.3.3 Structural properties . . . . . . . . . . . . . . . . . . . . . . . . 41
3.3.4 Optical properties . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.3.5 Surface morphology . . . . . . . . . . . . . . . . . . . . . . . . . 47
4 Ejection of ZnS nanoparticles from ZnS films by Au irradiation 52
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.2 Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.3.1 Ejection of ZnS NPs due to 2 MeV Au irradiation . . . . . . . . 54
4.3.2 Ejection of ZnS NPs due to 100 MeV Au irradiation . . . . . . . 58
4.3.3 A power law behaviour in nanoparticle size distribution . . . . . 60
4.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.4.1 NP ejection mechanisms . . . . . . . . . . . . . . . . . . . . . . 62
4.4.2 Temperature dependence of ZnS NP ejection under 100 MeV
Au irradiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
5 Anisotropic deformation of Au nanoparticles in silica glass by swift
heavy ion irradiation 67
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.2 Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.3.1 Synthesis of Au NPs in silica glass . . . . . . . . . . . . . . . . . 69
5.3.2 Shape deformation of the NPs: TEM studies . . . . . . . . . . . 71
5.3.3 Outward movement and loss of Au: RBS studies . . . . . . . . . 73
5.3.4 Shape deformation of the NPs: OA studies . . . . . . . . . . . . 74
5.3.5 Dependence of longitudinal SPR peak position on aspect ratio . 76
xvi

5.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
6 Observation of a universal aggregation mechanism in Au sputtered
by swift heavy ions 82
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
6.2 Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
6.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
6.3.1 Ejection of Au NPs: Dependence on ion fluence . . . . . . . . . 86
6.3.2 Crystal structure of the ejected NPs . . . . . . . . . . . . . . . . 87
6.3.3 Loss of Au content in the target: RBS studies . . . . . . . . . . 87
6.3.4 A power law behaviour in NP size distribution . . . . . . . . . . 88
6.4 Discussion: NP ejection mechanisms . . . . . . . . . . . . . . . . . . . 89
7 Summary and conclusions 92
xvii

List of Figures
1.1 Schematic diagram illustrating the basic ion-solid interaction processes. . . . 3
1.2 Schematic diagram illustrating the basic materials modification processes:
(a) implantation; (b) damage due to collision cascades; (c) sputtering; and
(d) atomic mixing across an interface (shown as a dashed line). . . . . . . . 4
1.3 Variation of nuclear and electronic energy losses as a function of incident ion
energy in case of Au in ZnS. All the data were extracted using SRIM code [31]. 5
1.4 Schematic presentation of the ion depth distribution. . . . . . . . . . . . . 7
1.5 Schematic diagram showing multiple vacancy production under a primary
knock-on (an atom which is struck and displaced from its lattice site by a
high energy particle) leading to a displacement spike production. . . . . . . 8
1.6 Time evolution of an ion track due to Coulomb explosion and thermal spike
production by the high energy heavy ion. This figure is adapted from Ref. [46] 10
1.7 Schematic presentation of the shock wave model with two processes: 1 – primary
compressed region, 2 – region of shock wave propagation. The particles
in the shaded region are ejected. . . . . . . . . . . . . . . . . . . . . . . 14
1.8 Schematic presentation of different sputtering scenario for cluster formation
in a p−V-diagram. A target volume is heated up by ion bombardment and
thus attains a high pressure (points 1, 2 and 3). This figure is adapted from
Ref. [23]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.1 Schematic diagram showing the thermal evaporation (resistive heating) experimental
setup. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.2 Schematic of the low energy negative ion implanter facility. BPM – beam
profile monitor, FC – Faraday cup, TMP – Turbo-molecular pump. . . . . . 21
2.3 Schematic diagram of the 3.0 MV 9SDH-2 tandem Pelletron accelerator facility
at the Institute of Physics, Bhubaneswar. . . . . . . . . . . . . . . . 22
2.4 Schematic presentation of an elastic collision between an energetic projectile
and a target atom. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
xviii

2.5 Schematic diagram showing Rutherford backscattering spectrometry experimental
setup. PREAMP – pre-amplifier, AMP – amplifier, ADC – analog
to digital converter, MCA – multi-channel analyzer. . . . . . . . . . . . . . 26
2.6 Schematic diagram illustrating the x-ray diffractometer setup. . . . . . . . . 28
2.7 Block diagram of a transmission electron microscope. . . . . . . . . . . . . 30
2.8 Schematic diagram illustrating an atomic force microscope setup with a scanner
attached with the sample holder. . . . . . . . . . . . . . . . . . . . . . 32
2.9 van der Waals force as a function of tip to sample surface distance. . . . . . 33
2.10 Schematic diagram of the spectrophotometer used for optical absorption
measurements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.1 Variation of energy losses (S n , S e , and S T ) against ZnS depth for 100 MeV
Au ion at normal incidence. Inset: Same for 35 keV and 2 MeV Au ions. All
the energy loss data are extracted using SRIM code [31]. . . . . . . . . . . 40
3.2 RBS spectrum obtained with the as-grown ZnS film on Si(100) substrate.
The Zn, S and Si edges are marked in the figure. Inset shows the RBS
experimental geometry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.3 The XRD spectra of the as-grown and the Au irradiated ZnS films. The
spectra corresponding to irradiated samples are shifted vertically downward
for clarity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.4 Variation of the strain in as-grown and Au ion irradiated films as a function
of energy deposition in the ZnS surface. . . . . . . . . . . . . . . . . . . . 45
3.5 (a) Plots of (αhν) 2 versus photon energy, hν, for the as-grown and the Au
irradiated ZnS films. Intersections of the straight lines with the abscissa, as
shown by the arrows indicate the band gaps of the films. Inset shows the
experimental OA spectra of the samples. (b) Variation of the band gap as a
function of total energy deposition in the ZnS matrix for different irradiation
energies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.6 2×2 µm 2 three-dimensional AFM images of (a) as-grown, (b) 35 keV, (c) 2
MeV, and (d) 100 MeV Au irradiated ZnS surfaces. Vertical scale is 10 nm
per division. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.7 Variation of the PSD function with the spatial frequency of ZnS surfaces
for as-grown and Au ion irradiated ZnS films. In the high q region, the
experimental data are fitted with a function as given in Eqn. 3.6. The slopes
yield the parameter γ for each case. . . . . . . . . . . . . . . . . . . . . . 49
4.1 Schematic diagram illustrating the sputtering experiment. . . . . . . . . . . 54
xix

4.2 HRTEM micrographs of the NPs corresponding to 2 MeV Au irradiation to
the fluences of (a) 1×10 11 and (b) 5×10 13 ions cm −2 . . . . . . . . . . . . . 54
4.3 SAED patterns of the NPs corresponding to 2 MeV Au irradiation to a
fluence of 5×10 13 ions cm −2 . The diffraction rings corresponding to various
planes are indicated in the figure. . . . . . . . . . . . . . . . . . . . . . . 55
4.4 Bright-field TEM micrographs of ZnS NPs collected on catcher foils. (a),
(b), (c), and (d) correspond to the fluences of 1×10 12 , 5×10 13 , 1×10 14 , and
1×10 15 ions cm −2 with corresponding size distributions shown in (e), (f),
(g), and (h), respectively. The histograms in (e) – (h) have been fitted using
a log-normal distribution function (Eqn. 4.1), which are also plotted in the
figure as solid curves. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.5 Variation of (a) the width of the particle size distribution and (b) the catcher
foil surface coverage by the ejected NPs as a function of fluence. . . . . . . 57
4.6 HRTEM micrograph of a NP corresponding to 100 MeV Au irradiation to a
fluence of 1×10 13 ions cm −2 . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.7 (a) Bright-field planar TEM micrograph of ZnS NPs collected on catcher foil
during 100 MeV Au irradiation at room temperature. (b) shows corresponding
size distribution of the ZnS NPs. . . . . . . . . . . . . . . . . . . . . . 59
4.8 Bright-field TEM micrograph taken on the catcher foil for the 35 keV Au
irradiation to a fluence of 1×10 14 ions cm −2 . . . . . . . . . . . . . . . . . 59
4.9 Measured number of collected particles (having diameter ≥3 nm) as a function
of hemisphere volume. Experimental data are fitted, with a function
given in Eqn. 4.2, and shown in the figure as a straight line for each set of
data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.10 Snapshots from a simulation of sputtering of Au at different times for 20 keV
Xe incident on Au. The labels A and B show clusters that are fragmenting,
and C illustrates late sputtering (a cluster separating from the surface after
the displacement cascade has ended). This figure is adapted from Ref. [21]. . 61
4.11 SRIM simulated collision cascades as a function of depth for (a) 35 keV and
(b) 2 MeV Au in ZnS at an incident angle of 45 ◦ with the target normal. . . 64
5.1 Bright-filed XTEM micrographs of the Au NPs in silica glass after 32 keV
Au − ions impantation to a fluence of 4×10 16 cm −2 (a) before and (b) after
annealing in air at 850 ◦ C for 1 h. (c) and (d) show the corresponding size
distributions, respectively. Inset in (a) shows an HRTEM image of a Au NP
indicating the particles are crystalline in nature. . . . . . . . . . . . . . . . 70
xx

5.2 OA spectrum of Au implanted and annealed silica glass samples. . . . . . . 71
5.3 (a)−(d) Bright-filed XTEM micrographs of the as-grown and 100 MeV Au
irradiated samples with fluences of 2×10 13 , 5×10 13 , and 1×10 14 ions cm −2 ,
respectively; (e)−(h) show the corresponding size distributions, respectively.
Note that (b) and (c) are having same scale as in (a). Dashed lines are
indicating the surfaces. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
5.4 Bright-field planar TEM micrograph of the 100 MeV Au irradiated sample
for the fluence of 1×10 14 ions cm −2 . . . . . . . . . . . . . . . . . . . . . . 73
5.5 (a) RBS spectra taken from the as-grown samples before and after SHI irradiations.
The Au and Si RBS profiles are indicated in the figure. (b) Au
sputtered from the samples after SHI irradiation with different ion fluences. 74
5.6 OA spectra taken on the (a) as-grown and SHI irradiated samples corresponding
to the fluences of (b) 2×10 13 , (c) 5×10 13 , and (d) 1×10 14 ions
cm −2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
5.7 OA spectrum (filled circles) of the as-grown sample after 2×10 13 ions cm −2
irradiation along with the fitted (solid line) curve using two Gaussian functions
with proper background substraction (dashed lines). . . . . . . . . . . 76
5.8 Dependence of the absorption maximum of the longitudinal plasmon resonance
against the average aspect ratio as determined by XTEM. The solid
line is the linear fit to the experimental data, except the 1×10 14 ions cm −2
irradiated one. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.9 Simulated OA spectra of the Au NPs embedded in silica matrix for different
NP sizes (d) and medium dielectric constants (ǫ m ). Inset shows the variation
of SPR peak position with dielectric constant (filled circles). The solid line
is the linear fit to the simulated data. . . . . . . . . . . . . . . . . . . . . 80
6.1 Bright-field XTEM micrographs of the Au NPs in silica glass after 32 keV
Au − ions implantation to a fluence of 4×10 16 cm −2 followed by annealing
in air at 850 ◦ for 1 h. (b) show the corresponding size distribution with a
Gaussian fitting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
6.2 Schematic diagram illustrating the sputtering experiment. . . . . . . . . . . 85
6.3 Bright-field planer TEM micrographs of the ejected Au particles on the
catcher foils collected during ion irradiation with irradiation fluences of (a)
2×10 13 , (b) 5×10 13 , and (c) 1×10 14 ions cm −2 , respectively. (d) Coverage
of the catcher foils surface by the ejected NPs. Solid line is the linear fit to
the experimentally obtained data. . . . . . . . . . . . . . . . . . . . . . . 86
xxi

6.4 (a) HRTEM micrograph of a NP and (b) SAED pattern of the NPs collected
on the catcher foil corresponding to a Au irradiation fluence of 1×10 14 ions
cm −2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
6.5 The Au part of the RBS spectra as measured on the targets before and after
the irradiations. The irradiation fluences are in the unit of ions cm −2 . Inset
shows the experimental arrangement. . . . . . . . . . . . . . . . . . . . . 88
6.6 Size distribution in terms of the number of clusters of particular size (cluster
yield) plotted against the number of atoms in the cluster. Dotted, dotdashed,
and continuous lines correspond to power law decays with δ values
of 3/2, 7/3, and 7/2, respectively. . . . . . . . . . . . . . . . . . . . . . . 88
xxii

Chapter 1
Introduction
1.1 Interests and applications of ion beams in materials
science
A major part of the modern technology depends on materials with precise controlled
properties. Ion implantation is a favoured method to achieve controlled modification
of surface and near-surface regions of various materials. In fact, ion beams have proven
to be versatile tools for processing materials and synthesizing unique structures for
electronic [1], optoelectronic [2], and magnetic [3] device applications. Introduction of
ion species in the near surface of solids can be used to alter and tailor many material
properties. For instance, He implantations, with energy in the range of 10 to 180 keV,
into Ni has been found to result in an increase in hardness as much as approximately
seven times than that of unirradiated Ni [4]. Implantation of 250 keV Ge + ions into
SiC has been seen to result in a decrease in resistivity with increase in Ge fluence [5].
Co implantation at 30 keV into thin Fe 70 Co 8 Si 12 B 10 films deposited on SiO 2 has been
found to show a stress-induced magnetic anisotropy in the ferromagnetic thin film [6].
It is possible to tune the band-gap of semiconductor materials by using ion implantation.
For example, 18 keV P + implantation into InAs/InP heterostructures has been
shown to result in a shift in photoluminescence peak towards higher wavelength. This
shift in peak position increases with increase in ion fluence [7]. C implantation into
Si(111) at 1 MeV energy has been found to result in enhanced gettering of Au [8].
Recently, using ion implantation, a new route for synthesis of nanostructure materials
inside a host has been reported [9]. For example, 32 keV Au − ions implantation to
a fluence of 8×10 16 cm −2 into silica glass has been found to result in Au nanocluster
formation with an average size of about 6 nm [9].
Defect generation during ion-solid interactions adds a further dimension to this
1

Introduction 2
processing capability enabling kinetic mechanisms to synthesize metastable phases
normally unobtainable under equilibrium thermodynamic conditions. For example,
mixing of thermally immiscible layers of Fe/Ag using ion irradiation has been reported
[10]. In this case, 85 keV Ar + ions were irradiated in multilayers of Fe/Ag in the
fluence range of 5×10 14 – 3×10 16 ions cm −2 . It was seen that the mixing increased
with increase in fluence. In certain cases, involving both metals and insulating systems,
irradiation with high energy heavy ions has been found to result in the formation of ion
tracks, a few nm in diameter and few µm in length [11]. These can be used as templates
for growth of nanoparticles (NPs) inside them. In addition, anisotropic deformation
of Ag NPs, embedden in silica, has also been observed by 8 MeV Si irradiation [12].
The anisotropy in shape results in change in optical properties. Also energetic ion
irradiation of solids leads to the ejection of materials from the surface of the target.
This process of removal of materials, either atoms or chunks of atoms (clusters), is
known as sputtering [13]. Sputtering is used for cleaning solid surfaces [13], coating
thin films (where the sputtered material is deposited on another body) [14], etc. It is
also used in the analysis of materials using secondary ion mass spectroscopy [15].
Very recently, sputtering of metal as well as alloy metal NPs have been observed
[16]. This observation of stable, intact clusters of atoms and molecules, due to ion
impact, is rather surprising because the energies involved in the ion-target collisions,
that generate sputtered atoms, are typically much larger than the binding energies (∼
1 – 2 eV) holding the clusters together [17]. Hence it is difficult to understand how
stable clusters can form during ion sputtering. Eversince the first observation, the
fundamental question of the cluster emission mechanism has attracted considerable
attention both experimentally [13, 16, 18, 19] and theoretically [20, 21, 22, 23].
Due to all the above interests, during the last several decades, considerable attention
has been paid to ion-solid interactions, leading to a detailed understanding
of the important processes, such as ion penetration, energy loss, radiation damage,
sputtering of clusters, grain growth in thin films, and irradiation-induced mixing.
1.2 Ion-solid interactions
1.2.1 Basic concepts
When a beam of energetic charged particles is directed onto the surface of a solid,
the fast projectiles interact with the stationary near-surface atoms of the solid. These

Introduction 3
interactions can lead to ion penetration into the subsurface region of the solid, rearrangement
of the near-surface target atoms, and the emission of particles and other
forms of radiation from the surface [24]. When an ion penetrates a solid, it undergoes
a succession of binary collisions with target atoms and surrounding electrons, losing
energy at each encounter. The incident ion may conceivably be reflected back out of
the surface during a single “hard” collision or from many correlated collisions. However,
for ions of incident energy above a few keV, it is most probable that the ion will
finally come to rest within the solid. The transfer of energy from projectile ion to
the solid can be conveniently divided into two independent processes, namely nuclear
collisions and electronic collisions [25, 26]. The former process constitutes collisions
between the incident ion and lattice atoms where conservation of energy and momentum
apply (so-called elastic collisions). The latter electronic process depicts the
interaction of fast ions with lattice electrons (so-called inelastic collisions). Electronic
collisions result in ionization and excitation processes, which can lead to emission of
radiation in the form of characteristic x-rays, optical photons, and Auger or secondary
electrons. In Fig. 1.1, the types of emission process which can result from both nuclear
and electronic collisions are illustrated schematically.
Ion beam
Nuclear processes
Electronic processes
Backsc−
attered
atoms
Nuclear
reaction
products
Sputtered
target
atoms
Photons
Phonons
Electrons
Figure 1.1: Schematic
diagram illustrating the
basic ion-solid interaction
processes.
Solid
Electronic
and Nuclear
energy−loss
processes

Introduction 4
Many
ions
(a)
Single ion
(b)
Modified composition
(c)
Sputtering
Atomic
displacement
Surface
erosion
Collision
cascade
Figure 1.2: Schematic diagram illustrating
the basic materials modification
processes: (a) implantation;
(b) damage due to collision cascades;
(c) sputtering; and (d) atomic mixing
across an interface (shown as a dashed
line).
(d)
Thin film
Substrate
Figure 1.2 (a) shows the ion implantation process, which can lead to the build up
of a concentration profile of foreign atoms within a solid, thus altering the near-surface
composition [27]. The distribution of foreign atoms depends on the ion energy and the
stopping (nuclear and electronic) processes which determine how individual ions slow
down and where they ultimately come to rest. Figure 1.2 (b) illustrates that “hard”
nuclear collisions can result in displacement of lattice atoms from their regular sites. A
single heavy ion can lead to displacement of several tens, and even hundreds, of lattice
atoms within a volume surrounding the ion trajectory, termed a collision or displacement
cascade. As a consequence, ion bombardment can create considerable structural
damage to the material [28]. Also atoms at or near the surface may receive sufficient
energy to surmount the surface potential barrier and leave the solid. This sputtering
process [29], results in a flux of ejected atomic and/or molecular species, which can

Introduction 5
be neutral or in various charge states [13]. Schematic illustration of this process is
given in Fig. 1.2 (c) and discussed in details in Section 1.3. Finally, in Fig. 1.2 (d) we
illustrate the process of atomic mixing in which solid atoms can be transported across
an interface within dimensions of the collision cascade at temperatures below those
at which normal diffusion processes would operate. One consequence of this effect is
shown for a thin film of one material on a substrate of another: ion bombardment can
produce appreciable intermixing with the substrate material.
1.2.2 Ion penetration and stopping
Energy loss processes
As an ion penetrates a solid, it loses energy due to two distinct interactions [25,
30]. At lower energies, it loses energy by elastic, nuclear collisions with the target
atoms (nuclear stopping) described by the well-known Rutherford cross sections for
the Coulomb interaction between the screened charges of the atomic nuclei. At higher
energies, it loses energy inelastically by exciting and ionizing electrons of target atoms
(electronic stopping). In this process, incident ion transfers its energy to the lattice
electrons inelastically. It is expected that the lattice atoms are not displaced in the case
of electronic excitation. At low ion velocities, where the nuclear energy-loss dominates,
the energy loss process is accurately described by the electrostatic interaction between
the projectile and the target atoms using a Thomas-Fermi potential [25, 26]. At
higher ion velocities, the electronic energy-loss processes are better described by the
Bethe-Bloch formalism [25, 26]. Both stopping processes strongly depend on the
kinetic energy of the ion. Figure 1.3 shows a graph of the nuclear and electronic
Energy Loss [keV/nm]
30
25
20
15
10
5
Nuclear energy loss
Electronic energy loss
Figure 1.3: Variation of nuclear and
electronic energy losses as a function
of incident ion energy in case of Au in
ZnS. All the data were extracted using
SRIM code [31].
0
10 0 10 1 10 2 10 3 10 4 10 5 10 6 10 7 10 8
Energy [keV]

Introduction 6
energy losses as a function of ion energy for Au irradiation of a ZnS target. As
can be clearly seen in Fig. 1.3, the stopping for Au in ZnS matrix is dominated by
the nuclear energy loss for ion energies smaller than 1 MeV. In this region, due to
collisions with the incident ions, target atoms are displaced from their original positions
[32], leading to subsequent nuclear collision cascades. This can lead to structural
changes, such as the generation of point defects and the amorphization of crystalline
materials (and vice-versa). This happens when the Rutherford scattering cross section
is highest, i.e., for heavy incident ions at energies of typically 10−100 keV. On the
other hand, for higher ion energies (> 4 MeV in Fig. 1.3) energy loss is dominated by
electronic stopping, resulting in a rise in local lattice temperatures, as described above.
Depending upon the type of material and ion energy, this can lead to the formation of
a cylindrically-shaped molten region around the ion track, with typical diameter in the
range of few nm [33]. A fast cooling of the molten region can, therefore, bring about
structural changes in the target material, such as amorphization, defect generation,
and structural transformations. One prominent effect of electronic stopping processes
is anisotropic deformation [12], a macroscopic materials shape change that is the result
of microscopic processes occurring in the thermal spike of individual ion impacts.
Ion Range and Range distributions
An ion traversing a solid gradually loses its energy in every monolayer or so following
the above energy loss processes, finally coming to rest at a depth below the surface.
This is knows as “ion implantation” as mentioned in Section 1.2.1. Ion implantation
is widely used in materials engineering for the introduction of atoms into surface
layer of a solid substrate by bombardment of the solid with ions of desired species
in the keV to MeV energy range. Use of the implantation technique affords the
possibility of introducing a wide range of atomic species, thus making it possible to
obtain impurity concentrations and distributions of particular interest; in many cases,
these distributions would not be otherwise attainable [34].
One of the most important parameters of implantation process is the “range” of
the implanted ions which correspond to the total distance traveled by an ion before
coming into rest as given by [25]
R =
∫ E0
0
[( ) dE
+
dx
nucl.
( ) ] −1
dE
dE, (1.1)
dx
elec.
where E 0 is the incident ion energy at the target surface, x is the distance measured

Introduction 7
along the ion path, ( )
dE
and ( )
dE
are the nuclear and electronic energy losses,
dx nucl. dx elec.
respectively, of the ion at energy E. Since energy is lost by the projectile to the target
atoms and electrons in a series of discrete collisions, the energy loss per collision and
hence the total path length will have a statistical spread of values. This leads to a
near-Gaussian distribution of stopping distances. The mean length of projection of
the ion path on its incident direction is known as the projected range, R p . This is
schematically shown in Fig. 1.4. In ion implantation studies one is actually interested
Z
log scale
ion
distribution
∆Rp
Ion beam
Rp
∆Rt
X
Y
R
Figure 1.4: Schematic presentation of the ion depth distribution.
in the entire depth distribution of the implanted atoms and, therefore, a knowledge of
the range straggling (fluctuation) is also necessary. For ions of moderate energies, in
the absence of channeling, the range distribution is approximately Gaussian in shape
and can be described by a single extra parameter namely the fluctuation (standard
deviation) ∆R p , in the projected range along the direction of R p . ∆R p is called the
longitudinal range straggling and the fluctuation along the direction perpendicular
to this is known as transverse range straggling, ∆R t . A schematic diagram showing
range, R, projected range, R p , and range stragglings (∆R p and ∆R t ) of the implanted
atoms is shown in Fig. 1.4. These parameters of ion implantation are important for
device modeling and design, as well as for the process optimization and control in
device fabrication. A large amount of experimental and theoretical work has been

Introduction 8
devoted to the task of understanding the energy-loss processes that govern the range
distribution, and it is now possible to predict fairly accurately most of the factors
involved in the ion implantation process [25, 35].
1.2.3 Nuclear thermal spike
In the elastic collision regime, with few keV energy, an ion upon collision with a lattice
atom can result in displacement of the lattice atom. The minimum energy imparted
to the lattice atoms as required for this is known as the displacement threshold which
is about 10 – 50 eV [34]. At much higher energy transfers the displaced atoms can lead
to collision cascades leading to a larger number of displaced atoms. Multiple collisions
between recoiling atoms can be described by linear transport theory as independent
binary collision events as long as the density of recoiling atoms is low. However,
when the projectile mass and/or the atomic mass of the substrate are large, these
assumptions fail as nonlinear processes occur [36]. When the mean free path between
collisions is of the order of the atomic spacing, a highly disturbed volume known as
an atomic displacement spike or energy spike [Fig. 1.5] is created in less than 10 −13 s.
Figure 1.5: Schematic diagram showing
multiple vacancy production under
a primary knock-on (an atom which
is struck and displaced from its lattice
site by a high energy particle) leading
to a displacement spike production.
This is followed by a thermal spike phase (which lasts for about 10 −11 s) of a small
volume in which the energy density may be several eV/atom. Here the temperature
goes well above the melting temperature and in some cases going even beyond the
vaporization temperature of the material. This is known as elastic or nuclear thermal
spike. In addition to this there is also a corresponding pressure spike. Molecular
dynamics (MD) simulations, using 10 keV Au on Au(001) surface do indicate such a
scenario associated with cascade induced melting and viscous flow [37].

Introduction 9
1.2.4 Coulomb explosion and electronic thermal spike
However, the scenario become completely different when a high energy heavy ion (swift
heavy ion) passes through a material. The energy loss in such a situation (involving inelastic
interactions with target electrons) is given by the electronic stopping power S e .
After the initial energy-transfer by high energy ions, electrons escape the central region
of the ion path. Depending on the ionization density and on the charge-neutralization
time, the mutual repulsion of positively charged target ions may convert a significant
amount of the stored electronic potential energy into atomic motion. This is known
as the Coulomb explosion [38, 39, 40, 41]. Coulomb explosion is significant only if the
charge-neutralization time exceeds 10 −14 s for light target atoms and 10 −13 s for heavy
atoms.
In such a case the initially excited and ionized electrons transfer their energy to
the lattice atoms (in about 10 −12 s) by electron-phonon coupling. This would induce
a local rise in lattice temperature [42]. In this scenario, the electron and atom system
are not in thermal equilibrium with each other. In such a case, the space and time
evolutions of the electronic and lattice temperatures, T e and T, are governed by a set
of coupled nonlinear differential equations [43] (in cylindrical geometry) as given by
and
ρC e (T e ) δT e
δt = δ [
K e (T e ) δT ]
e
δr δr
ρC(T) δT
δt = δ [
K(T) δT ]
δr δr
+ K e(T e ) δT e
r δr − g(T e − T) + A(r), (1.2)
+ K(T) δT
r δr − g(T e − T), (1.3)
where C e , C and K e , K are the specific heat and thermal conductivity for the electronic
system and lattice, respectively, ρ is the materials density, g is the electron-phonon
coupling, A(r) is the energy brought on the electronic system in a time considerably
less than the electronic thermalization time, and r is the radius in cylindrical geometry
with the heavy ion path as axis. This description of rise of local lattice temperature
and its dissipation to surround medium is well known as inelastic or electronic thermal
spike model [43].
As a result of Coulomb explosion and inelastic thermal spike production, there can
be formation of amorphized latent tracks (of few nm in diameter and several µm in
length) [33, 44] coming from sudden cooling of the molten zone. Using MD simulations,
Bringa and Johnson have shown that the Coulomb explosion and the thermal spikes
correspond to the early and late stages of the ionization track produced in a solid by a
fast incident ion [45]. Figure 1.6 displays a schematic view of the time dependence of

Introduction 10
Figure 1.6: Time evolution of an
ion track due to Coulomb explosion
and thermal spike production by the
high energy heavy ion. This figure is
adapted from Ref. [46]
the ion-track evolution. The upper part shows the rapidly passing projectile (dashed
arrow). Once the projectile has reached its equilibrium charge state, there will be only
minor fluctuations of its internal state and it will move with constant velocity along
a straight-line trajectory until deep inside the solid. Thus, the projectile ion acts
as a well defined and virtually instantaneous source of strongly localized electronic
excitation. According to Bethe’s equipartition rule, about 50 % of the total electronic
energy is deposited inside the so-called intra-track radius of about 1 nm around the
projectile path at projectile energies of a few MeV per nucleon [46]. Excitation times
are 10 −19 to 10 −17 s for inner-shell processes and reach 10 −16 s for collective electronic
excitations (plasmon production). The lattice atoms, displaced following thermal spike
production, freeze out and may lead to permanent rearrangements. In the bulk, this
may lead to structural or chemical modifications. At the surface craters or blisters on
an atomic scale can be produced. These are shown schematically in the last part of
Fig. 1.6.

Introduction 11
1.3 Sputtering
In 1852 W. R. Grove, studying electric discharges between metal electrodes in various
gas atmospheres [47], observed that material was removed from the negative electrode
(the cathode) and got deposited on the glass walls containing the electrodes and the
gas mixture. The above phenomenon corresponds to the simplest erosion process of
a surface called sputtering [13]. In the sputtering process, atoms, molecules, and
sometimes clusters are ejected from the outer surface layers of the target. Sputtering
is quantitatively described by the sputtering yield, Y , which is defined as the average
number of sputtered target atoms per incident ion, i.e.,
Number of sputtered atoms
Y = (1.4)
Number of incident ions
The sputtering yield, Y , depends on the ion incident angle, the energy of the ion, the
masses of the ion and target atoms, the surface topography, and the surface binding
energy of atoms in the target. For a crystalline target the orientation of the crystal
axes with respect to the target surface is relevant.
Depending upon the ion-solid interactions, there can be various types of sputtering
processes. Among these, physical sputtering is well studied in the materials science.
This type of sputtering is driven by momentum exchange between the energetic ion
and atoms in the material, due to collisions [13]. This type of sputtering takes place
only when the surface atoms of the substrate receive from the incident ion an energy
which is greater than the surface binding energy. In certain cases a chemical reaction
between the ion and the target atoms may result in lowering of the binding energy
resulting in enhancement in sputtering yield [48].
Apart from these, there are another two types of sputtering, viz. electronic sputtering
and potential sputtering. The term electronic sputtering means either sputtering
induced by energetic electrons, or sputtering due to very high-energy or highly charged
heavy ions which lose energy to the solid mostly by electronic stopping processes. In
the last case electronic excitations cause sputtering [49]. On the other hand, in the
case of multiply charged projectile ions, a particular form of electronic sputtering can
take place which has been termed potential sputtering [50]. In this case a low energy
ion in a high charge state can interact with a surface forming a hollow atom with
empty inner orbitals. Impact of a hollow atom on the surface, in some cases, can
lead to sputtering of target atoms through the release of the potential energy stored.
In the present case we will confine ourselves to sputtering from nuclear or electronic
collisions as observed at low and high ion energies, respectively.

Introduction 12
1.3.1 Sputtering mechanisms
Sputtering from elastic collision
For some time, the sputtering process was thought of as an evaporation process which
implied that sputtering was due to very high local temperatures created at or very
near the target surface by the bombarding ions. Later, however, measurements on
the dependence of sputtering yield on the angle of incidence of the bombarding ions
established that sputtering is really a momentum transfer process [29, 51]. There have
been a number of theories and models to explain the main features of the sputtering
process, observed at lower ion energy, but the most widely accepted one is that developed
by Sigmund [29]. The underlying concepts and the results of the Sigmund’s
theory is discussed briefly below.
As mentioned earlier, in case of low energy (keV), nuclear energy loss dominates. In
such a case, incident ion transfers its energy to the target atoms leading to generation
of collision cascades. Assuming that the cascades are produced as a result of binary
collisions, for particles incident normal to a substrate surface, the sputtering yield
at a given energy, E 0 , is described by a linear function of nuclear energy loss. This
sputtering yield, Y (E 0 ), can be written as
Y (E 0 ) = ΛF D (E 0 ), (1.5)
where F D is the energy deposition per unit length due to nuclear processes at the
surface, and Λ is a term which depends on the properties of target and the state of
the surface. Λ is given by
Λ = 3 1
4π 2 NC 0 U 0
where N is the target density, U 0 is the surface binding energy, and C 0 is a constant
defined as
C 0 = 1 ( ) (
2 πλ 0a 2 M1 2Z 1 Z 2 e 2 )
M 2 a
Here M 1 and M 2 are the ion and target atom masses respectively, Z 1 and Z 2 being
the ion and target atomic numbers, and a is the screening radius.
Electronic sputtering
Sigmund sputtering theory was quite successful in explaining the sputtering yields
for single elemental targets for various ion-target combinations. However, this theory
completely fails to explain enhanced sputtering as observed in certain cases as well

Introduction 13
as the sputtering of large clusters of atoms and molecules where nonlinear effects
of energy deposition are required to be included to explain the observed data [16,
18, 19, 20, 21]. As one enters the realm of nonlinear energy transport processes,
the mechanisms leading to sputtering become less obvious and, consequently, less
amenable to theoretical treatment discussed above. In spite of that, several nonlinear
models have been proposed [52]. Among them, the thermal spike theory is most
often used in this regime [53, 54]. This theory assumes that the moving atoms in
the target achieve, quickly, a state of local thermal equilibrium. Therefore, if the
density is assumed to remain constant, the local temperature suffices to determine the
thermodynamical state of the system. The deposited energy then relaxes according to
the heat conduction equation (Eqns. 1.2 and 1.3) and the sputtering can be readily
obtained as the flux of atoms evaporated from the hot surface. One testable outcome
of this theory is that, under quite general conditions, for large dE/dX, the sputtering
yield for a cylindrical excitation geometry is a quadratic function of the deposited
energy, i.e. Y ∼ (dE/dX) 2 [55]. Such a quadratic behaviour of Y was in fact observed
in many experiments, and it appeared to be so firmly established that a larger-thanlinear
sputtering yield often is sufficient to conclude that a thermal spike occurred.
1.3.2 Size distribution of clusters ejected during ion irradiation
Recently, several researchers have observed the ejection of large clusters in the sputtered
species during ion bombardment [16, 17, 18, 19, 20, 21]. Most of these studies,
carried out at keV energies, show a monotonically decreasing yield distribution, which
closely follows an inverse power law decay with increasing cluster size.
In the last several decades, a lot of efforts has been made to understand cluster
ejection mechanisms, resulting from energetic ion bombardment. In almost all cases
the size distributions have been found to follow power law distributions with different
decay exponents [16, 17, 18, 19]. With low energy ions, for small clusters, the decay
exponents have been found to lie between 4 to 8 [19]. However, at higher energy it
seems to be close to 2 which has been argued as coming due to shock wave induced
emissions [16]. MD simulations [20, 21] do indicate cluster emissions however, with
some breakup during the flight. On the theoretical front there are two models viz.
the one based on generation of shock waves [22] and a thermodynamic liquid-gas type
phase transition model [23] for the clustering of sputtered material. Both these models

Introduction 14
predict decay exponents close to 2 and there are no results, available in the literature,
to distinguish between the two. There are also non-equilibrium aggregation processes
that can play a role in determining the cluster size distribution. In these cases, there
is a competition between aggregation and breakup leading to entirely different size
distributions [56]. Such processes are expected to be present in sputtered material
coming from thermal spikes induced by very high energy ions. There have been no
data on this as well. Some basic features of the above mentioned models are given
below.
Shock wave model
One of the most successful model for cluster ejection is shock wave model developed
by Bitensky and Parilis [22]. For low energy heavy ion irradiation on materials, it is
plausible that the mean free displacement collision paths of cascading atoms become
of the order of the atomic spacing or less than it. In this situation, the nonlinear interactions
become significant and the available kinetic energy may be instantaneously
released locally. If this is the case, recoiling atoms in the region may be energetic
enough to form a fire ball locally. If the average velocity of these atoms exceeds the
sound velocity in the target medium, atoms of the cascade may play the role of a
“hammer”compressing the surrounding medium towards the outer direction following
a spherical geometry (Fig. 1.7). In case of high energy heavy ion irradiation, particu-
Projectiles
Solid
Casade process
1
Figure 1.7: Schematic presentation of
the shock wave model with two processes:
1 – primary compressed region,
2 – region of shock wave propagation.
The particles in the shaded region are
ejected.
2
larly of insulators, the accumulated energy is initially the potential energy of Coulomb
repulsion of the ionized atoms. In a non-metallic solid the time for its neutralization is
rather long and after 10 −12 s, this potential energy is transferred through a Coulomb

Introduction 15
explosion into kinetic energy which is spent in track creation, formation of high density
collision cascades and shock wave creation. In this case, the shock wave propagates
perpendicular to the cylindrical ion track. The necessary condition for shock wave
generation is the achievement of a critical mean kinetic energy of atoms ǫ c , which is
the free parameter of the theory [22].
Once a shock wave is formed, it propagates towards the surface. Upon reaching
the surface it gets reflected. When the pressure at this surface exceeds a critical value;
a part of the surface is unloaded (shaded region in the Fig. 1.7). The result of this
unloading gives the sputtering yield in the shock wave model. Ejection of clusters
following this model leads to an inverse power-law size distribution [22] i.e.,
Y (n) ∼ n −δ , (1.6)
where, n is the number of atoms in a cluster and δ is the exponent with a value of
5/3 or 7/3. The first exponent corresponds to the sputtering of clusters due to the
formation of collision cascades by single ion, whereas the second exponent corresponds
to the sputtering of clusters due to the formation of collision cascades by the impacts
of multiple ions.
Shock waves production has also been seen in MD simulations [37] involving the
impact of 10 keV Au on Au(001) surface. However, in that case there is a viscous flow
of molten Au atoms coming from the pressure spike generated. Cluster emission data
[16] from Au thin films, irradiated using four different ions (Ar, Ne, Kr, and Au) with
energies in the range of 400 to 500 keV, have been seen to follow the size distribution
as given in Eqn. 1.6 with δ values as given by the shock wave model of Bitensky and
Parilis [22].
Equilibrium liquid-gas type phase transition model
Another important cluster ejection model is the thermodynamic equilibrium liquidgas
type phase transition model proposed by Urbassek [23]. Main assumption of
this model is that after the high energy atoms in a collision cascade, produced by
energetic ion impact, are emitted from the near surface regions of the targets, the
energized cascade volume will begin to equilibrate thermally. We depict this in the
p−V diagram of Fig. 1.8 as a vertical transition, assuming that collisional sputtering
did not substantially alter the relevant volume. The high pressure that builds up
leads to an expansion of the heated volume into the vacuum, and its further evolution
is essentially controlled by the temperature: If the temperature T is greater than a

Introduction 16
Figure 1.8: Schematic presentation of
different sputtering scenario for cluster
formation in a p−V-diagram. A target
volume is heated up by ion bombardment
and thus attains a high pressure
(points 1, 2 and 3). This figure is
adapted from Ref. [23].
critical temperature T c , the heated material gasifies and may flow out of the target.
Moreover, in the course of its development, the system will cool down, and may enter
the coexistence region. Now it is decisive, whether it enters at the large volume side
(V > V c ) or at the low volume side (V < V c ). This is obviously to a large part
determined by T c of the system. In the first case, the gasified system may form
clusters by nucleation from the gas phase (agglomeration). In the latter case, the
liquid system may break up due to the expansion and form a vapour (fragmentation),
as a result of liquid-gas phase transition, leading to clusters ejection. This equilibrium
model predicts transitions from a exponential decay to power law decay as the phase
transition occurs closer to the critical point. The dependence of the cluster yield Y (n)
on the cluster size, n, is given by
Y (n) = Y 0 n −δ exp[(−∆Gn − 4πn 2/3 r 2 σ)/kT], (1.7)
where ∆G is the difference of the Gibbs free energies of the liquid and gas phase, k is
Boltzmann’s constant, T is the temperature of the energized region, Y 0 is the sputter
yield (a constant for a given projectile-target system), r is the cluster radius, σ is the
surface tension, and δ is the exponent. At equilibrium ∆G is zero and at the critical
point the surface tension vanishes. It follows that predicted cluster size distributions
are very similar for the shock wave model. This model also predicts a power-law size
distribution of the sputtered particles with a δ value in between 2 and 2.5. For a van
der Waals gas the exponent, δ, has a value of 7/3.

Introduction 17
Nonequilibrium aggergation mechanisms
Another one of the important cluster formation model is nonequilibrium aggregation
model [56, 57, 58, 59] which was not studied earlier in the sputtering process. In
this model, there is a competition between aggregation and breakup or evaporation, a
delicate balance between the two leading to a variety of steady state mass distributions
[56]. The nonequilibrium aggregation model was considered to be in a discretized space
and time [57]. That is, particles with integer mass are placed on each site of the twodimensional
lattice, and they aggregate by random-walk process with discrete time
steps. At the beginning of each time step, there are particles on every site of the
lattice. All of them independently jump to randomly chosen sites according to a given
probability. If two or more particles come together at a site after the jump, they are
combine to form a new particle with a conserved mass. Then a particle with unit
mass is added to every site. Thus the evolution of one time step is completed and the
process of aggregation without considering any fragmentation has been repeated. This
model then shows that the asymptotic distribution of mass always follows a power law
behaviour. Later the analysis was extended to include both positive and negative
values for the dynamical variable which was taken to be charge rather than mass [58].
In this scenario, if m(j, n) be the charge of the particle on site j at the nth time step
then the aggregation process can be represented by the following stochastic equation
for m(j, n):
m(j, n + 1) = ∑ W jk (n)m(k, n) + I(j, n), (1.8)
k
where I(j, n) denotes the charge injected at the jth site at time n, and W jk (n) is a
stochastic variable which is equal to 1 when the particle on the kth site jumps to the
jth site and which is equal to 0 otherwise. Since one particle cannot go to two different
sites in a single time step, W jk (n) must be normalized as ∑ j W jk (n) = 1. In case of
W jk (n) = 1/N with probability 1/N, where N is the total number of sites, which is
taken to be infinity in the calculation. This is what corresponds to the mean field
limit. The distribution of m can be obtained by introducing the r-body characteristic
function
Z 1 (ρ, n + 1) = Φ(ρ) ∑ a r Z 1 (ρ, n) r , (1.9)
r=0
where a r ≡ ( N r )(1/N)r (1 − 1/N) N−r . In the limit N → ∞, the steady-state solution
of Eqn. 1.9 is obtained as
Z 1 (ρ) = 1 − √ 2 < I > 1/2 |ρ| 1/2 i −1/2 + . . . for < I >≠ 0 (1.10)

Introduction 18
This equation gives the following solution for m ≫< I > > 0 (or for m ≪< I > <
0):
p(m) ∝ |m| −3/2 , (1.11)
where p(m) is the probability density.
A similar result has also been obtained by Majumdar et al. [59] for aggregation
in a mass conserving mean field type site-site interaction model. It has also been
shown that small changes in the breakup parameters do not affect the decay exponent
[60, 61]. This happens to be a very general case corresponding to a broad class of
phenomena. As shown by Bonabeau et al. [62, 60], fish schools, with breakup and
injection, show a similar aggregation, the size distribution showing an inverse power
law behaviour with an exponent of 3/2. This is seen even in economics related to
distribution of wealth [61]. Hence the nonequilibrium aggregation process is known to
be Universal in nature [63].

Chapter 2
Experimental techniques
2.1 Introduction
In the era of the advancement of electronic devices, fabrication and characterization
techniques play an important role. As the device size is shrinking to nanometer scale,
size, shape, and local environment of the building blocks (say transistors) will eventually
strongly influence the device performance. The above fact has given a big push to
research related to the synthesis of nanostructured materials, either on or embedded
in different substrates. A variety of experimental tools are being used for both the
fabrication and the characterization of nanostructured materials.
In this chapter, we provide a brief description of some of the synthesis and characterization
techniques that are used for preparation and characterization of the samples
used in the present studies. We begin with a short description of the vacuum evaporation
thin film deposition setup which was used for preparation of ZnS films. This
is followed by a description of the ion implantation/irradiation setups which were essentially
particle accelerators used for acceleration of heavy ions in the keV and MeV
energy ranges. Following this, we will discuss briefly, the principles of ion scattering, x-
ray diffraction, transmission electron microscopy, atomic force microscopy, and optical
absorption spectroscopy. These are the techniques used for sample characterization.
2.2 Sample preparation techniques
2.2.1 Vacuum evaporation thin film deposition setup
A schematic diagram of the vacuum coating unit (HINDHIVAC, Bangalore), available
at the Institute of Physics (IOP), Bhubaneswar, is shown in Fig. 2.1. This unit
19

Experimental techniques 20
Thickness monitor
Water
cooled
chamber
wall
Vent and gas
flow valves
Sample holder
Shutter
Mo boat
To vacuum
pumps
Figure 2.1: Schematic diagram
showing the thermal evaporation
(resistive heating) experimental
setup.
I
Power
supply
is equipped with three resistive heating facilities and an electron beam evaporation
facility (not shown in the figure). All these are housed in a relatively large vacuum
chamber with a base pressure in the low 10 −6 mbar range. There is a sample holder and
a quartz crystal thickness monitor (that could be placed close to the sample) mounted
inside the chamber. For typical applications, the material is loaded in a Molybdenum
(Mo) boat, which is resistively heated by passing a high current through it. During
heating, the material, to be deposited, melts and evaporates from the boat. At this
low pressure ∼ 10 −6 mbar , the mean free path of the evaporated atoms is larger
than the Mo boat to substrate distance. Therefore the evaporated particles travel in
straight lines from the evaporation source towards the substrate kept above. In the
present arrangement the sample holder is always at room temperature. A shutter,
available just below the sample holder, allows to start or stop the material deposition
on the substrates.
2.2.2 Ion implantation and irradiation
The Au implantations carried out on various samples, as discussed in this thesis,
have been performed using the low energy negative ion implanter facility at IOP,
Bhubaneswar [64]. The higher energy ion irradiation experiments, carried out on ZnS
thin films and Au implanted silica glass samples have been carried out using the implantation
beamline of the 3 MV tandem Pelletron accelerator (9SDH-2, NEC, USA)

Experimental techniques 21
at IOP [65] and the materials science beamline of the 15 MV tandem Pelletron accelerator
at the Inter University Accelerator Centre, New Delhi [66]. A brief description
of the ion implantation and the irradiation facilities are given below.
Low energy negative ion implanter facility
valve
cathode
cooling
000000000000
111111111111
000000000000
111111111111
000000000000
111111111111
000000000000
111111111111
000000000000
111111111111
000000000000
111111111111
00 11
00 11
00 11
magnet
gate Cs vapor vent cryo gate
valve pump valves
ion ion acce− pump
source −leration + tee
ionizer
power
steerer quadrupole
singlet
drift
tube
BPM
slit
FC
einzel
lens
000000000000
000000000000
111111111111
111111111111 cyro
000000000000
111111111111 pump
000000000000
111111111111
pump + tee
scanner
000000000000
111111111111
000000000000
111111111111
000000000000
111111111111 TMP rotary
000000000000
111111111111 pump
000000000000
111111111111
gate
valves
target
holder
chamber
Figure 2.2: Schematic of the low energy negative ion implanter facility. BPM – beam profile
monitor, FC – Faraday cup, TMP – Turbo-molecular pump.
A schematic of the low energy negative ion implanter facility used is shown in Fig. 2.2.
This machine is equipped with a SNICS-II ion source made by NEC, USA [67, 68].
SNICS stands for source of negative ions by cesium sputtering and the source is capable
of producing ions from almost all the elements in the periodic table, except the inert
gas elements. In SNICS, Cs vapours from a heated reservoir are thermally ionized
by a tantalum ionizer and are accelerated towards a cold cathode, in order to sputter
out the cathode material. A part of the Cs vapour also gets deposited on the cathode
surface in the form of a film. The sputtered materials while coming out of the cathode
get an extra electron while passing through the Cs film. A positive potential is applied
to the extractor and focus electrode assembly to extract the negative ions in the form
of a beam. The entire source is kept floating at a high negative potential. The ions
thus extracted get accelerated to the ground potential. If C, E, and B are the cathode,
extractor and bias potentials respectively, then the total negative potential at which
negative ions are generated is given by V 0 = C + F + B. SNICS is known to provide
singly charged ions and therefore the energy of the ions is given by V 0 eV (V 0 is usually
between 10–60 keV). To shift the beam suitably, an electrostatic X-Y steerer assembly

Experimental techniques 22
is provided. This is followed by an electrostatic quadruple singlet which focuses the
beam in the vertical plane and puts it into a 45 ◦ bending magnet. This magnet is
being used for the ion mass selection. Following the magnet, there is a drift section at
the end of which there is a beam profile monitor (BPM) and a Faraday cup followed
by an Einzel lens. The BPM provides a continuous oscilloscope display of the shape
and position of the beam, the Faraday cup is used to monitor the beam current.
After a drift section there is an Einzel lens which is used to focus the beam onto the
target inside a scattering chamber. This beam is raster scanned over the sample by
using another X-Y steerer placed before the scattering chamber. The entire system
is maintained at high vacuum (10 −6 mbar) using two cryo pumps, a turbo molecular
pump followed by a rotary pump, as shown in Fig. 2.2. This implanter facility can
provide singly charged negative ions with a maximum energy of 60 keV.
High energy ion implantation/irradiation facilities
Figure 2.3: Schematic diagram of the 3.0 MV 9SDH-2 tandem Pelletron accelerator facility
at the Institute of Physics, Bhubaneswar.
A schematic diagram of the 3 MV tandem accelerator along with various beamlines,

Experimental techniques 23
available at IOP, Bhubaneswar, is shown in Fig. 2.3. The machine is equipped with
two ion sources – one RF ion source (alphatros) used for providing inert gas ions and
the other one, a Cs sputtering ion source (a multi cathode SNICS). The negative ions
produced by the ion source are accelerated to a maximum energy of 75 keV before
entering the machine through a 90 ◦ injector magnet. This mass selected low energy
beam is focused using an Einzel lens and directed into the accelerating column of
the Pelletron machine. The Pelletron is a typical double ended Van-de-Grafff with
the belts replaced by chains of metallic pellets separated by insulating connectors.
The pellets carry the induced charge more efficiently into the terminal since charges
reside on the inner surface of each pellet.In total there are two chains carrying charges
continuously to the terminal. In the high voltage terminal of the Pelletron, the negative
ions are converted to the positive ions by making them pass through a gas-stripper.
The positive ions formed (of different charge states) get further accelerated towards
the other end of the machine kept at ground potential. After the acceleration, the ions
pass through an analyzing magnet for selecting the required ions with right energy.
Finally the switching magnet directs the beam into the desired beamline. For ion
implantation or irradiation at few MeV, the beamline at -15 ◦ , as shown in Fig. 2.3,
is used. This beamline is equipped with a raster scanner operating at 517 Hz in the
horizontal direction and 64 Hz in the vertical direction.
A similar tandem accelerator facility with a maximum terminal voltage of 15 MV,
available at the Inter University Accelerator Centre, New Delhi, has been used to
irradiate samples at much higher energy using Swift Heavy Ions (SHI) (e.g. 100 MeV
Au 8+ ). The working principle of the this machine is almost similar to the 3 MV tandem
Pelletron accelerator, described above. The implantation/irradiation chambers are
maintained at high vacuum using turbo-molecular pumps and are equipped with the
necessary electrical connections for accurate beam current measurements. The target
holders have two degrees of freedom viz. z-position and tilt angle. In each case several
samples can be loaded at a time.
2.3 Characterization techniques
2.3.1 Rutherford backscattering spectrometry
One of the most important analytical methods in material science using ion beams is
the Rutherford backscattering spectrometry (RBS) [69]. RBS is based on the hard

Experimental techniques 24
sphere collisions between atomic nuclei and derives its name from Lord Ernest Rutherford,
who in 1911 was the first to present the concept of atoms having nuclei. The
technique involves shooting a monoenergetic beam of light ions (proton, He + or He ++ )
into a target and detect the number and energy of particles in a backward direction.
With this information, it is possible to determine atomic mass of the scatterer and
elemental concentrations as a function of depth starting from the surface.
When a sample is bombarded with a beam of high energy light particles (few keV
< E < few MeV), the interaction between the projectile and the target atoms, can be
modeled accurately as an elastic collision using classical physics. There are four basic
factors that are involved in the process [69]. These are given below.
• The energy of the projectile after the collision can be related to its energy before
the collision by means of a kinematic factor. This factor brings in the much
needed target-mass dependence.
• The interaction between the projectile and the target atom can be described as
an elastic Coulomb collision between two isolated particles and expressed in term
of a scattering cross section. Knowing the cross section, the scattering yield can
be connected to the target mass concentration.
• The energy loss of the energetic particle, as it traverses the scattering medium,
is expressed in terms of the stopping cross section. Through this one determines
the energy of the ion just before it gets scattered in a given depth. It also is used
to determine the final energy at which the ion comes out losing energy along its
path.
• The energy loss is a statistical process. Mono-energetic projectiles assume an
energy distribution after penetrating a certain depth in the target. This uncertainty
in energy loss is known as energy straggling.
The above four concepts form the basis of RBS analysis. The kinematic factor provides
RBS an elemental detection capability, the scattering cross section gives it the ability
of a quantitative measurement, the stopping cross section results in the capability for
the depth analysis and energy straggling sets limits on mass and depth resolutions.
When a projectile atom with mass M 1 and energy E 0 undergoes an elastic collision
with a stationary atom of mass M 2 (M 2 > M 1 ) in a target, energy is transferred
from the projectile to the target. The interaction between two atoms can be properly
described by a simple elastic collision of two isolated particles. The interaction is

Experimental techniques 25
basically Coulomb interaction between two nuclei of charges Z 1 and Z 2 , provided
the following two conditions are satisfied: (i) the projectile energy E 0 must be much
larger than the binding energy of the atoms in the targets and (ii) nuclear reactions
and resonances must be absent. The energy transfers in the elastic collision between
two isolated particles can be obtained by applying the principles of conservation of
energy and momentum. A schematic diagram of the scattering process is shown in
Fig. 2.4. The energy (E 1 ) of the projectile after the collision with the target atom is
scattering angle
E 1, M 1
collimated
E 0, M 1 beam
M 2
/
E , M 2
Figure 2.4: Schematic presentation
of an elastic collision
between an energetic
projectile and a target atom.
related to its energy before collision, E 0 and is given by [69]
where k =
angle.
[
E 1 = k E 0 , (2.1)
] 2
M 1 cosθ+(M2 2−M2 1 sin2 θ) 1/2
M 1 +M 2
is the kinematic factor and θ is the scattering
If Q is the total number of incident particles that hit a target and dQ is the
number of backscattered particles recorded by the detector, placed subtending a solid
angle dΩ at an angle θ with the direction of incidence, then the differential scattering
cross section dσ
dσ
dΩ = 1 Nt
[ dΩ
dQ/dΩ
Q
is related to the number of detected particles dQ by the relation
]
, where N is the volume density of atoms in the target and t is its
thickness. So Nt is the number of target atoms per unit area. Thus, the number of
particles detected by a detector with 100 % efficiency, subtending a solid angle dΩ is
given by
Y = QNt dσ dΩ (2.2)
dΩ
The differential scattering cross section for an elastic collision between two atoms is
given by Rutherford’s formula
[
dσ
dΩ = Z1 Z 2 e 2 ]
[ 2
4 [1 − (
M 1
M 2
sinθ) 2 ] 1/2 + cosθ ] 2
4E 0 sin 4 [ ]
θ
1/2
(2.3)
1 − (
M 1
M 2
sinθ) 2

Experimental techniques 26
Simulations of the RBS spectra are needed for the interpretation of the experimental
results. This is because the composition in the depth direction can not be
derived, directly, from the experimental data. Several algorithms [70, 71, 72] based on
the above physical principles are available for this purpose .
RBS setup
All the ion scattering experiments as presented in this thesis have been carried out in
a multi-purpose scattering chamber, equipped with a sample manipulator under high
vacuum conditions (∼ 10 −6 Torr). This chamber sits in the 45 o beam line (schematically
shown in Fig. 2.3) of the 3 MV tandem Pelletron accelerator (9SDH-2, NEC,
USA) facility at IOP [73]. For RBS experiments pertaining to the studies presented
in this thesis, He ions of 1.35 to 3.0 MeV energy, with beam currents in the range of
1 to 15 nA have been used.
current
integrator
000 111
000 111
000 111
000 111
000 111
000 111
000 111
000 111
000 111
000 111
000 111
000 111
000 111
000 111
000 111
000 111
000 111
sample
slit 2 slit 1
mass selected mono−energetic
ion beams from accelerator
scattered ions
detector
ADC AMP PREAMP
MCA
detector
bias
Figure 2.5: Schematic diagram showing Rutherford backscattering spectrometry experimental
setup. PREAMP – pre-amplifier, AMP – amplifier, ADC – analog to digital converter,
MCA – multi-channel analyzer.
A schematic diagram of the RBS experimental setup is shown in Fig. 2.5. A well
collimated mono-energetic beam of ions (usually He + or He ++ ), obtained from the
accelerator is made to fall on a target. A silicon surface barrier detector placed in
front of the target detects the scattered ions coming at a backscattering angle. A
good setup allows for target tilt (changing the angle of incidence) and an angular
movement of the detector leading to a change in the scattering angle. The distance

Experimental techniques 27
from the sample can also be adjusted resulting in a change in solid angle (usually kept

Experimental techniques 28
sample surface at a grazing angle. Especially for the analysis of thin films, XRD
techniques have been developed for which the primary beam enters the sample with
a very small angle of incidence. In its simplest variant, this configuration is denoted
by GIXRD that stands for grazing incidence XRD. The small entrance angle causes
the path traveled by the x-rays to significantly increase and the structural information
contained in the diffractogram to stem primarily from the thin surface layer. In order
to evaluate the mean grain size (D) of the film, Scherrer’s formula [74],
D =
0.94 λ
β cos(θ B )
(2.6)
is often used, where λ, β, and θ B are incident radiation wavelength, full width at
half maximum of the diffraction peak (in radians), and the Bragg diffraction angle,
respectively.
XRD setup
Detector
Divergent
slits
Axis
X−rays
Receiving
slits
Theta
2 Theta
Figure 2.6: Schematic diagram
illustrating the x-ray
diffractometer setup.
X−ray
source
Theta
Sample
A schematic diagram of the XRD setup used in the present studies is shown in Fig. 2.6.
A well collimated monochromatic beam of x-rays (Cu K α radiation, i.e., λ=1.54
Å from a Panalytical made X ′ PERT powder x-ray diffractometer) are directed towards
the sample. The interaction of the incident rays with the sample produces
constructive interference (and a diffracted ray) when Bragg’s law, i.e., Eqn. 2.4 is
satisfied. These diffracted x-rays are then detected, processed, and counted. In case
of pollycrystalline or powder samples, by scanning through a range of 2θ angles, all
possible diffraction directions of the lattice could be included for analysis. Conversion

Experimental techniques 29
of the diffraction peaks to d-spacings allows identification of the materials because
of unique d-spacings. Typically, this is achieved by comparison of d-spacings with
standard reference patterns.
2.3.3 Transmission electron microscopy
Like XRD, transmission electron microscopy (TEM) provides informations, such as
crystal structure and microstructure of materials [75]. Apart from these, TEM observation
can provide direct imaging of the structure of the particles, from which size,
shape, and their arrangements can be obtained. This technique can be used for the
study of defects in the atomic-scale. TEM uses a high energy electron beam transmitted
through a very thin specimen to image and analyze its microstructure with atomic
scale resolution. The resolution of an optical microscope is limited by the wavelength
of light used and hence limited to the order of a micrometer. However, electrons accelerated
to hundreds of keV have wavelengths much smaller than that of light. At an
energy of 200 keV, electrons have a wavelength of 0.025 Å. However, the resolution of
a TEM is limited, by the aberrations inherent in electromagnetic lenses, to about 1 to
2 Å and hence can reveal details down to the atomic scale.
Energetic electrons incident on the specimen undergo various elastic and inelastic
scatterings due to interaction with the atoms of the specimen [75]. Most of the electrons
are elastically scattered by the nuclei of atoms in the specimen. Some electrons
are inelastically scattered by the electrons in the specimen as well. Energetic electrons
passing through the specimen near the nuclei are somewhat accelerated towards the
nuclei causing small local reductions in wavelength, resulting in a small phase change.
In this way, information about the specimen structure get transferred to the phase of
the electrons. For getting high resolution images only the elastically scattered electrons
are of importance. The inelastically scattered electrons contribute mostly to
the background. This can be removed by inserting an energy filter in the microscope
between the specimen and the recording device. The inelastically scattered electrons
can sometimes be useful. Analysis of them can be helpful in obtaining chemical information
from very small regions in the spectrum. From the energy loss spectrum of the
inelastically scattered electrons or the filtered part of the spectrum, this information
can be extracted.
Electron diffraction from a crystalline lattice can be described as a kinematic scattering
process that meets the wave reinforcement and interference conditions given

Experimental techniques 30
in the Bragg equation. The patterns are formed by diffraction of an electron beam
transmitted through the thin specimen. The electron diffraction yields spot patterns
from single crystals, ring patterns from fine-grains and randomly oriented crystallites
and superimposed ring and spot patterns from larger-grain polycrystalline films.
TEM setup
A transmission electron microscope consists of an illumination system, specimen stage
and imaging system analogous to a conventional light microscope. A schematic diagram
of a TEM is shown in Fig. 2.7. At the top, there is an electron gun, producing
filament
000 111 0000
000 111 0000
000 111 0000
000 111 0000 1111
1111
1111
1111
000 111
1100 0000 1111
000 111
0000 1111
000 111
0000 1111
000 111 0000 1111
000 111 0000 1111
wehnelt cap
(negative potenial)
space charge
gun corss over
electron beam
anode plate
(positive potenial)
000000000000000000000
111111111111111111111first condenser lens
000000000000000000000
111111111111111111111
00000000000
11111111111
00000000000
11111111111
00000000000
11111111111
00000000000
11111111111
1111100000
1111100000
second condenser lens
Condenser aperture
sample
0000 1111
00000000000
11111111111objective lens
00000000000
11111111111objective aperture
selectred area aperture
Figure 2.7: Block diagram
of a transmission electron microscope.
00000000000
11111111111first intermediate lens
00000000000
11111111111
00000000000
11111111111
00000000000
11111111111
00000000000
11111111111
second intermediate lens
00000000000
11111111111projector lens
00000000000
11111111111
000000000000
111111111111 main screen (phospher)
a stream of monoenergetic electrons. In the system we have used, the electrons were
generated using a LaB 6 electron source by thermionic emission. The stream of electrons
from the electron gun is focused to a small, thin, coherent beam by the use of
the first and the second condenser lenses. The first lens largely determines the “spot

Experimental techniques 31
size”; the general size range of the final beam spot that strikes the sample. The second
lens is used for varying the size of the spot on the sample, changing it from a wide
dispersed spot to a pinpoint beam. The beam is restricted by the condenser aperture
(user selectable), knocking out high angle electrons (those far from the optic axis).
The beam strikes the specimen and part of it is transmitted. This transmitted
portion is focused by the objective lens into an image. Optional objective and selected
area metal apertures can restrict the beam. The objective aperture enhances
the contrast by blocking out high-angle diffracted electrons, the selected area aperture
enabling the user to examine the periodic diffraction of electrons by ordered
arrangements of atoms in the sample. The image is passed down the column through
the intermediate and projector lenses, being enlarged all the way. Finally the image
strikes the phosphor screen allowing the user to see the image. The darker areas of
the image represent those areas of the sample that fewer electrons were transmitted
through (they are thicker or denser). The lighter areas of the image represent those
areas of the sample through which more electrons were transmitted through (they are
thinner or less dense). A detailed description of the transmission electron microscopy
technique is given in Ref. [75].
Sample preparation for TEM
Due to the strong interaction between electrons and matter, the specimens have to
be rather thin (

Experimental techniques 32
energy ion beam at grazing incidence. Ar ions of energy 1.5 keV are used for this
purpose.
2.3.4 Atomic force microscopy
Atomic force microscopy (AFM) is an extensively used techniques in the field of fundamental
research, technology and bio-medicine [76]. It does not require the sample
to be electrically conductive unlike scanning electron microscopy. AFM is applicable
to any type of materials – insulators, semiconductors as well as conductors. Because
it uses a probe which is very small and sharp, AFM can be used to acquire highresolution
real-space topographic images of various sample surfaces. Hence it is a very
useful technique to monitor or characterize the morphology and topological changes of
various ion irradiated surfaces [76]. A schematic diagram of AFM is shown in Fig. 2.8.
The AFM probes the surface of a sample with a sharp tip, a couple of µm long and
Display and
feedback
electronics
Sample surface
Position
sensitive
photodiode
Laser
Cantilever
& Tip
Figure 2.8: Schematic diagram illustrating
an atomic force microscope
setup with a scanner attached with the
sample holder.
PZT Scanner
often less than 10 nm in diameter. The tip is located at the free end of a cantilever,
that is 100 to 200 µm long. Forces between the tip and the sample surface cause the
cantilever to bend or deflect. A detector measures the cantilever deflection as the tip
is scanned over the sample or the sample is scanned under the tip. The measured
cantilever deflection allow a computer to generate a map of surface topography.
An AFM can be used to study insulators and semiconductors as well as conductors
as mentioned above. Several forces typically contribute to the deflection of an AFM
tip and hence the cantilever. The force most commonly associated with AFM is the
interatomic van der Waals force. The dependence of the van der Waals force on the

Experimental techniques 33
Tip is in hard contact
with the surface −
repulsive regime
Tip is far from the surface
− no deflection
Force
0
Figure 2.9: van der Waals
force as a function of tip to
sample surface distance.
Tip is pulled towards the
surface − attractive regime
Probe distance from sample (z distance)
distance between the tip and the sample surface is schematically shown in Fig. 2.9.
Two distinct regimes are labeled in the figure namely attractive and repulsive regimes
(contact and non-contact regimes). In the contact regime, the cantilever is held less
than a few Å from the sample surface and the interatomic force between the cantilever
and sample is repulsive. In the non-contact regime, the cantilever is held above the
surface at about tens to hundreds of Å from the sample surface. The interatomic force
between the cantilever and the sample, in this mode, is attractive (largely a result of
the long-range van der Waals interaction).
In addition to the contact and the non-contact mode there is a third mode of
operation corresponding to the tapping mode which is an intermittent-contact mode.
In this case the vibrating cantilever is brought closer to the sample surface so that
at the bottom of its travel just barely hits or “taps” the sample. This mode is a
compromise between the contact and non-contact modes where the cantilever is made
to oscillate so that the tip remains very close to the sample for a short time and then
going far away for a short time. The compromise between the two allows one to scan
soft adsorbate on a substrate with better resolution than in the non-contact mode but
with a small interaction (and consequently less modification of the samples) between
the tip and adsorbate as in the non-contact mode.
2.3.5 Optical absorption
Optical absorption by metal particles
In the beginning of the 20 th century, Gustav Mie solved Maxwell’s equations for an
electromagnetic light wave interacting with a metal [77]. When applying the correct

Experimental techniques 34
boundary conditions for a metallic sphere, the calculations gave a series of multipole
oscillations (dipole, quadrupole, etc.) for the extinction and scattering cross sections
of the particle as a function of particle radius. Using the macroscopic, frequencydependent
complex dielectric constant of the bulk material of the metal sphere, he
was able to explain the beautiful colors of stained glass first systematically studied by
Michel Faraday in 1857 [78].
For very small metal particles in an optical medium, the fields can be considered
homogeneous (constant) over the particle, where only the low-order modes contribute
to the extinction. Then, the particles can be imagined to be placed in a uniform static
electric field (quasistatic regime). When there is negligible scattering, extinction cross
section is same as the absorption cross section and the extinction cross section, under
the dipolar approximation, is given by [79]
σ ext = 12πR3 ωǫ 3/2
m
c
ǫ 2 (ω)
[ǫ 1 (ω) + 2ǫ m ] 2 + ǫ 2 (ω)
(2.7)
In the above expression R is the radius of the metal particle, ω is the angular frequency
of the incident light, c, is the speed of light in vacuum, ǫ 1 (ω) and ǫ 2 (ω) are the
real and imaginary parts of the metal dielectric function, ǫ, and ǫ m is the dielectric
function of the matrix. A resonance in the spectrum of σ ext will occur whenever
the condition ǫ 1 (ω) = -2ǫ m is satisfied. This resonance is well known as the surface
plasmon resonance (SPR) which occurs due to a collective oscillation of the conduction
electrons of the metallic system in response to an exciting electromagnetic radiation.
As can be seen from Eqn. 2.7, the extinction cross section depends on the dielectric
function, ǫ, of the metal particles. However, no analytical theory gives a correct
description of the variation of the dielectric function with frequency, ω. Although the
Drude model and the Drude-Sommerfield model [80] are effective in explaining many
of the metal properties, a correct description of the dielectric constant is still lacking.
The main weakness of the “free” or “nearly free electron model” [80] is its inability to
account for the intraband transitions. These intraband transitions are very important
in case of Au- and Ag-like noble metals. In addition, for small particles (nm size),
some extra effects, such as quantum confinement [81] can play a significant role.
Therefore, in case of very small particles (sizes ≤ 50 nm), to get the correct size
dependence of the absorption cross section, one needs to use the dielectric function for
the metal, corrected for the variation of electron mean free path with the radius.(For
example, electron mean free path of bulk Ag is around 40 nm [82].) A phenomenological
formula for the complex dielectric function, incorporating such a correction is

Experimental techniques 35
given by [83]
ǫ(ω, R) = ǫ bulk (ω) +
ωp
2 ωp
2 −
ω 2 + iωγ 0 ω 2 + iω(γ 0 + Av F /R)
(2.8)
In the above equation, R is the radius of the nanocluster, ǫ bulk (ω) = ǫ 1 (ω) + iǫ 2 (ω) is
the bulk dielectric function of the metal particle, ω p is the plasma frequency, A is a
constant usually taken as 1.0, and v F is the Fermi velocity. The factor γ 0
is defined
as v F /l ∞ , where l ∞ is the mean free path of the electrons in the bulk material. Using
Eqn. 2.8 in Eqn. 2.7, one can estimate the optical properties of the particles of a
given size. For calculations involving Cu, Ag, and Au, the bulk dielectric constants,
as measured experimentally by Jonhson and Christy [82], are used.
Optical absorption by semiconductors
When a semiconductor is illuminated, photons can make the valence electrons of an
atom go to a higher electronic energy levels in the conduction band. In this process,
the incident photon is absorbed by an electron. The fundamental absorption process
is strongly affected by whether the gap between the valence band maximum and the
conduction band minimum (i.e. optical band gap) is a direct or an indirect [80] one.
In a direct gap semiconductor, an electron executes a vertical transition at the band
gap. Energy is conserved according to
¯hω = E f − E i , (2.9)
where, E f and E i are the final and initial state energies, respectively, of the material
and ¯hω is photon energy. For an indirect gap material (like Si), there is a requirement
of an additional momentum to reach the conduction band minimum at a non-zero wave
vector. This momentum comes from interaction with a phonon. Then the statement
of energy conservation then becomes
¯hω = E f − E i ± ¯hΩ, (2.10)
where, ¯hΩ is the energy of the phonon, and the plus and minus signs correspond
to phonon emission or absorption, respectively. Phonon emission dominates at low
temperatures. The need for an additional interaction with the phonon makes indirect
absorption far less probable than direct absorption which means that is the indirect
absorption is a weaker process.
The absorption coefficient for a direct or indirect band-gap semiconductor can be
calculated from
α = C ∑ n i n f P if , (2.11)

Experimental techniques 36
where, C is a constant, P if is the probability of the transition from the initial to the
final state, n i is the electron density in the initial state and n f is the density of the
empty states. The summation is over all states separated by the photon energy. For
a direct gap semiconductor, this relation gives an absorption coefficient [84] as
α = A(¯hω − E gap ) p , (2.12)
where, A is a constant independent of ω (frequency of the photon) and the exponent
p is 1/2 or 3/2 depending on quantum selection rules for the particular material. For
example, the value of p for ZnS is 1/2 [84]. Thus using Eqn. 2.12, one can find out
the band gap of a semiconductor by measuring the optical absorption coefficient, α.
This has been used in our analysis.
Optical absorption spectroscopy
000 111
000 111
000 111
000 111
000 111
000 111
m1
m7
00 11
00 11
00 11
00 11
00 11
00 11
00 11
00 11
00 11
00 11
00 11
00 11
m8
000 111
000 111
000 111
m9
sample
compartment
w3 w2
w3
sample
reference
w2
000 111
000 111
000 111
000 111
000 111
00 11
00 11
00 11
00 11
00 11
m3
m4
m5
00 11
00 11
00 11
00 11
00 11
C
000 111
000 111
000 111
000 111
000 111
000 111
m6
s3
00 11
00 11
00 11
00 11
00 11
00 11
m2
00 11
00 11
00 11
00 11
00 11
s2
G1
00 11
00 11
00 11
00 11
00 11
G2
s1
w1
0000 1111
0000 1111
0000 1111
0000 1111
source
0000 1111
0000 1111
0000 1111
0000 1111
0000 1111
0000 1111
0000 1111
detector
w1−w3 window plate
G1−G2 grating
C − chopper mirror
s1−s3 slits
m1−m9 mirrors
Figure 2.10: Schematic diagram of the spectrophotometer used for optical absorption measurements.
Optical spectroscopy is related to the absorption and emission of light by matter
[85]. A spectrophotometer can be used for studying the optical absorption properties
of embedded metal NPs and semiconductor thin films prepared in and the surface of
transparent media. In the present case, as will be shown in the subsequent chapters, we

Experimental techniques 37
are concerned with Au particles embedded in silica glass and ZnS thin films deposited
on silica glass. For studies related to these samples a Shimadzu UV-3101PC UV-VIS-
NIR dual beam spectrophotometer was used. A schematic diagram of the optics of
the spectrometer is shown in Fig. 2.10. This spectrometer uses a deuterium or halogen
lamp as the photon source. Different frequencies of the light are selected using a grating
assembly. In one of the arms of the spectrophotometer, the sample is kept while the
other is loaded with a reference sample. In studies involving metal nanoparticles in
silica glass or thin film deposited on silica glass, we have used a pristine sample (silica
glass piece from the sam batch) without any film deposited or ion implanted into it
as the reference. Depending on the energy either a photomultiplier tube or a PbS
detector is used for the detection of the outcoming photons. The spectrophotometer
is connected to a personal computer equipped with the control and data acquisition
software. Before starting the measurements one needs to make sure that the two
identical reference samples show zero absorption (which is carried out through the
baseline correction and auto zero options available with the control software).

Chapter 3
Modification of nanocrystalline ZnS
films by Au irradiation
3.1 Introduction
ZnS has been extensively investigated as an important wide-band-gap semiconductor
(E g ∼ 3.6 eV at room temperature [86]). It is one of the oldest and probably one of the
most important materials in the electronic industry with a wide range of applications
in light-emitting diodes, photovoltaic cells, dielectric filters, and efficient phosphors in
flat-pannel display. It is an excellent material for band-gap engineering. At ambient
conditions, ZnS shows two structural polymorphs: cubic sphalerite and hexagonal
wurtzite [87]. Synthesis of nanoscale ZnS, due to its marked varied properties, which
come from quantum-confinement effect, has received a lot of attention in the recent
years [86, 88, 89, 90]. To prepare ZnS thin films, a variety of techniques, such as
molecular beam epitaxy [91], pulsed laser deposition [92], metalorganic chemical vapor
deposition [93], chemical bath deposition [94], and thermal evaporation [95, 96] have
been utilized. Among these, the technique of thermal evaporation is a very simple
one suitable for good crystalline thin film preparation. This technique also provides a
possibility to grow thin films with grain sizes in the nanoscale regime.
As far as the ZnS is concerned, there are not many studies in literature that have
investigated ion irradiation induced modifications. In a recent work, a change in
morphology, of ZnS colloidal particles, from spherical to prolate shape under 4 MeV
Xe irradiation has been reported [97]. Chowdhury et al. [98] have carried out such a
study, on ZnS nanoparticles embedded in a polymer matrix, using Ni ions . They have
observed an increase in size of the crystallites and a reduction in optical band gap with
38

Modification of nanocrystalline ZnS films by Au irradiation 39
increase in ion fluence. In an another study, Mohanta et al. have observed a tunable
surface state emission from Mn-doped ZnS nanoparticles due to O irradiation [99].
For both the studies, very high irradiation energies viz. 160 and 80 MeV, respectively,
have been used. Since such studies are very limited it was necessary to carry out a
systematic study of ion irradiation effects on ZnS over a wide range of energies. Results
of this study on modifications in structural as well as optical properties and surface
morphology are presented in this chapter [100]. Details regarding the experiment and
results are given in the following sections.
3.2 Experiments
ZnS films (thickness ∼ 600 nm) were deposited, at room temperature, on cleaned
Si(100) (with native oxide) and silica glass substrates by thermal evaporation of 99.99
at. % pure ZnS powder (Materials Research Corporation, New York) using a vacuum
coating unit at a working base pressure of ∼ 7×10 −6 Torr. The evaporant (ZnS) was
placed in a molybdenum (Mo) boat at a distance of about 15 cm from the substrates.
The thickness and the rate of deposition of the films were measured in-situ using a
quartz crystal thickness monitor. The rate of deposition of the film was maintained
at about 1.5 nm s −1 . This way prepared films were irradiated, at room temperature,
with 35 keV Au − , 2 MeV Au 2+ , and 100 MeV Au 8+ ions at normal incidence with
the sample surface. In each case the ion fluence was always kept fixed to a value of
1×10 14 ions cm −2 with a maximum beam current density of ∼ 0.04 µA cm −2 to avoid
the beam induced heating effects. The ions beam were raster scanned over an area of
1×1 cm 2 on the sample surfaces for uniform irradiation.
Rutherford backscattering spectrometry (RBS), using 3 MeV He 2+ ions was used
to determine the composition and thickness of the as-grown and the irradiated films.
RBS was carried out with a surface barrier detector at a backscattering angle of 160 ◦ .
Crystal structure of the films were analyzed, before and after ion irradiations, using
a Panalytical make X ′ PERT powder x-ray diffractometer with CuK α radiation (λ =
0.154 nm) over a 2θ scan range of 10 ◦ −65 ◦ . Optical absorption (OA) measurements
were carried out on the as-grown and the irradiated ZnS films, deposited on silica
glass substrates, in the wavelength range of 250−800 nm using a Shimadzu UV 3101
PC UV-VIS-NIR dual spectrophotometer. The measurements were carried out in
the transmission mode with a blank silica glass sample (from the same batch) in the
reference line. Irradiation induced changes in the surface morphology of the ZnS/Si

Modification of nanocrystalline ZnS films by Au irradiation 40
samples were studied using a Nanoscope IIIA atomic force microscope (AFM) with a
silicon nitride cantilever operated in tapping mode. All the measurements were carried
out at room temperature.
3.3 Results and Discussion
3.3.1 Energy loss processes
During irradiation, the incident ions impart their energy to the target material through
both electronic and nuclear energy loss processes. The energy lost to the film results in
modifications in the properties and it is important to estimate the relative importance
of the energy loss processes. Using this in mind we have carried out a calculation
of the energy loss by Au in ZnS using SRIM [31] at 35 keV, 2, and 100 MeV. The
results, as a function of depth, are shown in Fig. 3.1. In the figure, nuclear energy
1600
600
o
Energy Loss [eV A -1
]
1400
1200
1000
800
600
400
200
S
S n + S e
n
Se
E. L. [eV A ]
-1
o
400
200
35 keV
0
0 1000 2000 3000 4000
o
Depth [A]
0
0 20000 40000 60000 80000 100000
o
Depth [A]
2 MeV
Figure 3.1: Variation of
energy losses (S n , S e , and
S T ) against ZnS depth for
100 MeV Au ion at normal
incidence. Inset: Same for
35 keV and 2 MeV Au ions.
All the energy loss data are
extracted using SRIM code
[31].
loss (S n ), electronic energy loss (S e ), and sum of the both (S T = S n + S e ) are shown
as a function of ZnS depth for the above three energies. One can see the value of S T ,
in case of 35 keV Au irradiation, rapidly decreases with increase in depth. Here the
major contribution is from S n . However, S T decreases slowly up to a depth of about
8000 nm for 100 MeV Au irradiation. Beyond ∼ 8000 nm, S T is nearly constant. In
this case, the major contribution comes from S e . On the other hand, for 2 MeV Au
irradiation, S T is almost constant up to a depth of 300 nm. Beyond this S T decreases
rapidly and finally Au atoms stop at a depth of about 400 nm. Here, both S n and

Modification of nanocrystalline ZnS films by Au irradiation 41
S e are contributing to S T significantly. Therefore any change that takes place in the
material properties is due to both electronic and nuclear energy deposition processes.
3.3.2 Thickness and composition of the ZnS films
Figure 3.2 shows the RBS spectrum of the as-grown ZnS film deposited on Si(100).
Composition and thickness of the film were determined by simulating the RBS spectrum
using GISA simulation code [72]. A good fit to the observed RBS spectrum
could be obtained considering a single film with Zn to S concentration ratio as 1:1
with a total thickness of around 615 nm. The above fit to the data is also shown in
Fig. 3.2. This indicates that the as-grown film was stoichiometric in nature.
Backscattered Yield (arb. unit)
2000
1500
1000
500
. . . . Experimental
Gisa fitted
Si
0
200 300 400 500 600 700 800
S
Si ZnS
Channel Number
3 MeV He 2+
Zn
o
20
Detector
Figure 3.2: RBS spectrum obtained
with the as-grown ZnS film on
Si(100) substrate. The Zn, S and Si
edges are marked in the figure. Inset
shows the RBS experimental geometry.
3.3.3 Structural properties
Figure 3.3 shows the XRD spectra of the as-grown and the Au irradiated ZnS films
for various Au irradiation energies viz. 35 keV, 2, and 100 MeV, respectively. The
spectrum corresponding to the as-grown film exhibits a peak at 2θ = 28.58 ◦ with a
very narrow full width at half maximum (FWHM). This peak is identified as either of
a cubic crystal structure with a preferred (111) orientation or a wurtzite structure with
a preferred (002) orientation. This is because the reflections from these two planes
cannot be resolved, using our x-ray diffractometer, due to their similar inter-planner
spacing [101]. The presence of only one peak, indicating a preferential orientation,
could be due to the growth of crystallites perpendicular to the substrate surface resulting
from controlled nucleation process associated with the slow deposition rate.
The observe narrow FWHM also indicates that the ZnS films were having good crystalline
quality. At this point it is also important to estimate the mean grain size, D,

Modification of nanocrystalline ZnS films by Au irradiation 42
as-grown
Intensity [arb. units]
35 keV
2 MeV
Figure 3.3: The XRD spectra of the
as-grown and the Au irradiated ZnS
films. The spectra corresponding to irradiated
samples are shifted vertically
downward for clarity.
100 MeV
10 20 30 40 50 60
2θ o
from the FWHM of the observed peak. This can be done using Scherrer’s formula
[74], as given by
D =
0.94 λ
β cos(θ β ) , (3.1)
where λ, β, and θ β are CuK α radiation wavelength, the FWHM of the diffraction
peak (in radians), and the Bragg diffraction angle, respectively. The estimated grain
size of the as-grown film is found to be 17.4 nm (Table 3.1) which indicates that
the crystallites were very small in size. It must be mentioned that the FWHM of
the observed peak has been corrected due to instrumental broadening of ∼ 0.06 ◦ for
determination of the mean grain size.
Table 3.1: Parameters obtained form the XRD spectra for the as-grown and the Au irradiated
ZnS films.
Irradiation Peak Grain size Grain size
energy position before correction after correction
(2θ ◦ ) (nm) (nm)
as-grown 28.58 17.2 17.4
35 keV 28.58 17.2 17.4
2 MeV 28.44 27.9 28.3
100 MeV 28.31 46.7 49.5
From Fig. 3.3, it is also observed that there is almost no change in intensity and
FWHM of the XRD peak after an irradiation at 35 keV. On the other hand, compared
to the as-grown case, there is a small increase in intensity and decrease in FWHM for

Modification of nanocrystalline ZnS films by Au irradiation 43
the 2 MeV irradiation case. In case of the 100 MeV Au irradiated sample the peak
intensity is seen to increases further with a further decreases in FWHM as compared to
the other cases. The increasing peak intensity and narrowing of the peak is indicative
of irradiation induced grain growth. The estimated grain sizes of the irradiated films,
as obtained using Scherrer’s formula, are shown in Table 3.1. A direct correlation
between an increase in grain size with irradiation energy is clearly seen.
One can see from Fig. 3.1 that for 35 keV Au ion in ZnS, the S n and the S e values
at the surface are 2.91 and 0.24 keV nm −1 , respectively. For 2 MeV Au ion, they
are 3.66 and 1.91 keV nm −1 , respectively. The fact that we could not see any grain
growth at 35 keV while there are grain growth for the Au irradiation at 2 MeV makes
one feel that S e plays a dominant role regarding the observed grain growth. At both
35 keV and 2 MeV although around 3 keV nm −1 , due to S n , is deposited in the ZnS
film, in case of the lower ion energy, the ion range is limited to only 20 nm. Therefore,
all energy is deposited in 20 nm where grain growth may be expected. However, this
depth is too small compared to the total film thickness which is seen by the x-rays.
Therefore the observed x-ray signal is mostly that coming from the film not affected
by the irradiation. That is why the signal from the 32 keV Au irradiated film is
almost identical to that obtained from the as-grown sample. However, in case there
are thermal spikes produced solely from nuclear energy loss [37], there would be some
melting associated with flow of materials towards the surface. This would result in an
enhancement in surface roughness as compared to the as-grown sample. As would be
shown later this is indeed true. Compared to this, at 2 MeV, both S e and S n play a
role although S n dominates. Because of this, one would expect a larger material flow
resulting in a higher surface roughness. At MeV energies, electrons in a cylindrical
region around the ion path get excited and ionized. In case of ZnS, which has a low
hole mobility [102], there is a possibility of Coulomb explosion coming from a higher
density of positive charges in the ionized region. If the hole mobility is higher then
the positive charges can get neutralized quickly not effecting a Coulomb explosion
and a subsequent thermal spike [45]. However, there is an increase in grain growth
coming from localized melting are recrystallization. Here energy deposition from both
electronic excitations and nuclear collisions contribute.
At 100 MeV with a much higher value of S e (∼ 15.8 keV nm −1 ), one would expect
Coulomb explosion to take place followed by an inelastic thermal spikes [45] resulting in
localized melting and recrystallization. Obviously, at 100 MeV one would have melting
over bigger zones as compared to those produced at 2 MeV. In case of inelastic thermal

Modification of nanocrystalline ZnS films by Au irradiation 44
spikes, produced by swift heavy ions (energy 1 MeV/amu), it has been shown that
the ion track diameter over which melting occurs increases with ion energy [33]. This
can result in larger grains over the entire film thickness. But with much reduced S n
value ∼ 0.4 keV nm −1 , one would not expect any material flow towards the surface
coming from nuclear collision effects. The surface roughness is thus expected to be
comparable to that of the as-grown sample.
From a closer look at Fig. 3.3, one can also see that there is a shift of the XRD
peak towards lower values of 2θ with increase in irradiation energy from 35 keV to
2 and 100 MeV. Further the XRD peak for the as-grown (or the 35 keV Au irradiated)
sample is seen to be at a higher value of 2θ as compared to bulk ZnS [101]. In
general, a shift in the peak position from that of its corresponding bulk material indicates
the development of stress in the film [103] during deposition and post-deposition
treatments. The observed strain can be due to several origins: (a) from the strained
regions in the films (e.g. grain boundaries, dislocations, voids, impurities, etc.), (b)
from the film-substrate interface (due to difference in thermal expansion coefficients
between the film and the underlying substrate), (c) from the film-vacuum interface
(surface stress, adsorption, etc.) or (d) from dynamic processes (e.g. recrystallization,
interdiffusion, etc.) [104]. In the present case the strain is mostly due to point (a)
as mentioned above. The strain, ǫ, developed in the films can be estimated using the
relation [105]
ǫ = d − d 0
d 0
× 100%, (3.2)
where d is the lattice-spacing of the ZnS film and d 0 is the unstrained bulk latticespacing.
A shift of the diffraction peaks to lower angles indicates the presence of
tensile strain whereas a shift towards higher angles indicates a compressive strain
[74, 106]. Based on this, the as-grown and the 35 keV Au irradiated films are found
to have a compressive strain. This is because the observed XRD peak positions of
the ZnS films are slightly higher than that of the bulk ZnS [101]. This compressive
strain is modified into tensile strain after MeV Au irradiations. Further, the tensile
strain, observed in the MeV irradiated films is seen to increase with increase in MeV
energy. The modification of internal compressive stress to tensile stress is mainly due
to the irradiation induced grain growth and annihilation of preexisting defects, such
as dislocations in the as-grown films. That an increase in grain size, does result in a
changeover from compressive strain to tensile strain, has been seen in ZnS thin films
grown using the thermal evaporation technique [96]. The variation of the strain in

Modification of nanocrystalline ZnS films by Au irradiation 45
going from the as-grown to the irradiated ZnS films with energy deposited in the film
surface is shown in Fig. 3.4.
0.8
0.6
(d)
Strain [%]
0.4
(c)
0.2
(a) as-grown
0.0
(b) 35 keV
(a)
-0.2
(c) 2 MeV
(b)
(d) 100 MeV
-0.4
0 4 8 12 16
Energy Loss at Surface [keV/nm]
Figure 3.4: Variation of the strain in
as-grown and Au ion irradiated films
as a function of energy deposition in
the ZnS surface.
3.3.4 Optical properties
Based on the above XRD results it was not possible to conclude about the crystal
structure of the ZnS film. It is also well known that cubic and wurtzite ZnS have
different band gaps. For cubic ZnS the band gap is 3.66 eV [86] while for wurtzite
structure it is 3.54 eV [94]. That prompted us to carry out an OA measurement on
this system. The absorption spectra of the as-grown and the irradiated ZnS films
are shown in inset of Fig. 3.5(a). The irradiation caused a tailing of the absorbance
towards higher wavelength region, which also increases with increase in energy. At 35
keV, we have not seen any observable change in the OA spectrum as compared to the
as-grown film and, therefore, the corresponding results are not shown in the figure.
To estimate the band gap of the films, we have used the standard expression for direct
transition between two parabolic bands as given by [84]
α 2 = C(hν − E g ), (3.3)
where α, C, hν, and E g are absorption coefficient, a constant, incident photon energy,
and the optical energy gap, respectively. The value of optical band gap, E g , is determined
from the intercept on the energy axis, in the plot of (αhν) 2 versus hν. The data
for various films are shown in Fig. 3.5(a). For the estimation of the value of α, we
have made use of the relation α = 2.303 A/t [107], where A and t correspond to the
measured absorbance and the film thickness, respectively. The estimated band gap
for the as-grown film is found to be about 3.56 eV, which compares favorably with

Modification of nanocrystalline ZnS films by Au irradiation 46
( α h ν ) 2 -1 2
(10 eV cm )
4
8
6
4
Absorbance [a. u.]
0.4
0.3
0.2
0.1
0
300 400 500
Wavelength [nm]
3.35
2 as-grown
3.30
2 MeV
100 MeV
100 MeV (a) 0.0 3.0x10 6 6.0x10 6 9.0x10 6
0
2 2.5 3 3.5 4 4.5
Energy Deposition [eV]
Photon Energy, h ν [eV]
Band Gap [eV]
3.60
3.55
3.50
3.45
3.40
as-grown
35 keV
2 MeV
Figure 3.5: (a) Plots of (αhν) 2 versus photon energy, hν, for the as-grown and the Au
irradiated ZnS films. Intersections of the straight lines with the abscissa, as shown by the
arrows indicate the band gaps of the films. Inset shows the experimental OA spectra of the
samples. (b) Variation of the band gap as a function of total energy deposition in the ZnS
matrix for different irradiation energies.
(b)
the literature value of 3.54 eV for wurtzite ZnS at room temperature [94]. Therefore,
together with the XRD data, we conclude that the ZnS films were having wurtzite
crystal structure with the diffraction peak, as seen in Fig. 3.3, at (2θ =) 28.58 ◦ coming
from the (002) reflection.
Figure 3.5(b) shows the variation of the band gap, E g , of the irradiated films as
a function of total energy deposition in the ZnS matrix as obtained from the SRIM
simulation [31], for the present experimental conditions. One can see, there is no
change in E g in case of a 35 keV Au irradiated film as compared to the as-grown film.
On the other hand the band gap, E g , decreases to a value as small as about 3.0 eV
for a 100 MeV Au irradiated film. The figure also shows a consistent decrease of E g
with increase in total energy deposition or in other words band gap is red-shifted with
increase in irradiation energy.
It must be mentioned here that in chalcogenide materials the lone pair orbital forms
the valence band, while the conduction band is formed by the antibonding orbital. It
is known that ion irradiation produces lattice damage in the form of point defects such
as vacancies, antisite defects, and interstitials. In such a situation, the reduction in
the band gap with increase in irradiation energy, may arise due to the effect of band
tailing, owing to the defects produced during irradiation. In the high-energy regime,
i.e., at 100 MeV the production of defects is dominated by electronic excitations only

Modification of nanocrystalline ZnS films by Au irradiation 47
and the influence of nuclear energy loss is insignificant. Therefore, the incident ions
excite the electrons from the lone pair and bonding states to higher energy states.
Vacancies created in these states are immediately filled by the outer electrons with
Auger processes that in turn induce more holes in the lone pair and bonding orbital
leading to a vacancy cascade process. In this process, bond breaking or ionization of
atoms is easier to occur which leads to a change in the local structure order causing a
decrease in the optical band gap [108].
3.3.5 Surface morphology
(a)
(b)
(c)
(d)
Figure 3.6: 2×2 µm 2 three-dimensional AFM images of (a) as-grown, (b) 35 keV, (c) 2
MeV, and (d) 100 MeV Au irradiated ZnS surfaces. Vertical scale is 10 nm per division.

Modification of nanocrystalline ZnS films by Au irradiation 48
Figure 3.6 shows the three-dimensional AFM images of the as-grown and the Au irradiated
surfaces of ZnS films deposited on Si(100) substrates. In the image corresponding
to the as-grown film [Fig. 3.6(a)], one can see an almost uniform distribution of small
structures on the surface. Three other images corresponding to the surface topography
of the Au (35 keV, 2, and 100 MeV) irradiated films are shown in Figs. 3.6 (b), (c),
and (d), respectively. One can see there is clear variation in the film surface roughness.
The 35 keV Au irradiated surface shows a clear growth in surface structures with
respect to the unirradiated sample. The surface features in the form of nano-hillocks
are seen to be the largest in the 2 MeV irradiation case. Au irradiation at 100 MeV
is seen to result in smaller structures as compared to the 2 MeV irradiation case.
Usually the features on as-grown and irradiated surfaces are described through a
height-height correlation function which contains three important roughness parameters:
(i) the vertical correlation length σ rms , (ii) the lateral correlation length η 0 , and
(iii) the roughness exponent α [109]. The lateral correlation length, η 0 , describes the
lateral characteristics of the surface, the roughness exponent, α, describing the static
scaling properties. The most commonly reported parameter of surface roughness i.e.,
σ rms or the root-mean-square (rms) roughness, characterizes the surface, only along
the vertical direction. This is defined as standard deviation of the surface height profile,
h(x, y), at each point (x, y) of a reference surface plane from the mean height
(< h >) as given by,
[ 1
σ rms =
N
] 1/2 N∑
(h i − < h >) 2 , (3.4)
i=1
where N is the number of pixels, h i = h(x, y) being the height at the i th pixel. The σ rms
values have been found to be 0.75, 0.96, 1.49, and 0.91 nm, for the surfaces obtained
with as-grown and Au irradiations as 35 keV, 2, and 100 MeV, respectively. At
lower energies there will be elastic thermal spike induced by nuclear collision cascades,
leading to shock wave [22, 110, 111] and material flow [112, 113] towards the surface.
If F d is the damage energy deposited per unit length and ǫ m is the average energy per
atom at the melting point, then as shown by Averback and Ghaly [114], the parameter
(F D /ǫ) 2 determines whether there would be melting or not. In the present case with
2 MeV Au, F D (can be taken as S n ) has a value close to 3.66 keV nm −1 . Taking the
melting point of ZnS to be 2023 K, one can estimate ǫ m to be 0.5 eV. Thus (F D /ǫ m ) 2
turns out to be 6000 2 which is quite high indicating surface melting. At 100 MeV,
the electronic energy loss dominates leading to inelastic thermal spikes produced in a
cylindrical region around the ion track [33, 115]. This occurs when the energy taken

Modification of nanocrystalline ZnS films by Au irradiation 49
by the electrons goes to the lattice through electron-phonon coupling. In case the
system has lower thermal conductivity as in ZnS there can be localized melting with
evaporation and/or plastic flow [115, 116]. However, this is seen to result in much
reduced surface damage.
However, σ rms is not sufficient to provide the complete information of the surface
modifications. This is because of σ rms gives information along the vertical direction
only, and hence, morphology parallel to surface plane is left out. In view of this, a
power spectral density (PSD) analysis is often used to look at surface features and
their possible origin [117]. The PSD analysis is accomplished by radially averaging
the square magnitude of the coefficients of the two-dimensional Fourier transform of
the digitized surface profile and is defined by [118]
C(q) = 1 L 2 ∫ ∫ d 2 r
2π e−iq.r 〈h(r)〉 2 , (3.5)
where L 2 , q, and h(r) are the scanned area of length L, the spatial frequency, and the
height at the position r = (x, y), respectively.
0.1
PSD [A
4
]
o
0.01
0.001
■
●
▲
▲
as-grown
35 keV
2 MeV
100 MeV
1 10 100
Spatial Frequency [ m ]
µ -1
Figure 3.7: Variation of the PSD
function with the spatial frequency of
ZnS surfaces for as-grown and Au ion
irradiated ZnS films. In the high q region,
the experimental data are fitted
with a function as given in Eqn. 3.6.
The slopes yield the parameter γ for
each case.
The PSD curves of as-grown and irradiated ZnS surfaces are shown in Fig. 3.7.
The variations in magnitude of the PSD function indicate the perturbations on the ion
irradiated surface which increase up to the irradiation energy of 2 MeV and decrease
for 100 MeV. The surface corrugation defined as the slope of a line connecting two
points on the surface, becomes small for points separated by a length longer than
the correlation length, η 0 (= 1/q 0 ). For lengths larger than correlation length the
surface can be considered to be flat. Thus, PSD is expected to be independent of q
for q < q 0 . For lengths smaller than η 0 (i.e. q > q 0 ), corrugations become significant

Modification of nanocrystalline ZnS films by Au irradiation 50
and no special value of q is characteristic of the surface morphology, and hence, the
PSD curve displays a power-law dependence [109]:
C(q) = A q −γ , (3.6)
where A is a constant and γ is the power-law exponent.
The power spectra as shown in Fig. 3.7 can be divided into two distinct regions:
the low frequency part resembles the uncorrelated white noise, while the straight line
high frequency part represents the correlated surface features. The estimated values of
γ, obtained by fitting the experimental data using Eqn. 3.6, are found to be 2.80±0.03,
3.11±0.04, 3.72±0.06, and 2.85±0.03 for as-grown, 35 keV, 2, and 100 MeV Au ion
irradiated ZnS films, respectively. The value of γ is seen to be the highest for the 2
MeV irradiation case decreasing to a lower value for the 100 MeV irradiation.
The evolution of surface morphology during ion irradiation is the result of a balance
between multiple roughening and smoothing processes. There is a stochastic
roughening process due to random arrival of the ions on the surface. The smoothing
mechanisms on the surface include volume diffusion, viscous flow, sputter redeposition,
and sputter removal affected by shadowing processes [109]. However, to know the
dominant mechanism behind the observed morphological changes in the ZnS surfaces,
due to ion irradiation, one can use the values of γ.
Using a simple linear dimensional analysis, Herring [119] has shown that γ has
values of 1, 2, 3, and 4 representing four modes of surface transport viz. viscous flow,
evaporation-condensation, volume diffusion, and surface diffusion, respectively. In the
case of single predominant process, the process can be easily identified. The index, γ,
is further related to the roughness scaling exponent, α, through the relation γ = 2(α
+ 1) [109].
For the as-grown ZnS film, the value of γ is about 2.8 which is close to 3. This
means the surface morphology of the as-grown film is governed by mainly volume
diffusion. This is expected as the films were deposited using thermal evaporation
of ZnS powder. During continuous film growth ZnS molecules evaporated from ZnS
powder seem to be diffusing into the earlier deposited film showing features of volume
diffusion. The 35 keV Au irradiated sample is also expected to show volume diffusion
coming from nuclear collision cascades formed near the surface. For this γ has a value
of 3.11 again close to 3. On the other hand the 2 MeV Au irradiated sample shows a
γ value of 3.7 which is close to 4, indicating surface diffusion. In fact this is expected
since there is a spreading of material on the surface coming from underneath by the

Modification of nanocrystalline ZnS films by Au irradiation 51
elastic thermal spikes. A trace of volume diffusion is also expected here which probably
is the reason why γ is slightly less than 4. On the other hand at 100 MeV, the ZnS
film surface temperature, in a localized region around the ion trajectory, may go well
above the vaporization temperature because of the inelastic thermal spike production
[33]. This would lead to surface evaporation which seems to produce smaller surface
features as compared to the 2 MeV irradiation case. This may be the reason why
the corresponding surface roughness indicate a γ value of 2.8 (close to 3). However,
as can be seen for all the four cases, involving one un-irrdaiated and three irradiated
surfaces the γ values have been found to range between 2.8 to 3.7. This implies the
surface scaling exponent, α has a value lying between 0.4 and 0.86. This indicate all
the surfaces to be self-affine.

Chapter 4
Ejection of ZnS nanoparticles from
ZnS films by Au irradiation
4.1 Introduction
Energetic ion irradiation of solids leads to ejection of atoms, molecules, and/or clusters.
This phenomenon, known as sputtering, takes place due to the energy transfer
to the atoms in the target by the incident ions through electronic interactions as well
as nuclear collision cascades produced in the target. Simple sputtering of atoms coming
from nuclear collision cascades is well described by Sigmund’s linear theory [29].
Recently, enhanced sputtering by ion irradiation has been reported by several groups
[16, 18, 111, 120, 121, 122, 123, 124, 125, 126, 127]. Sigmund’s theory fails to explain
these results and nonlinear effects were considered to explain the cluster ejection processes.
Most of these studies show a monotonically decreasing yield distribution, which
closely follows an inverse power law with increasing cluster size. At present there are
two models that predict power law cluster yield distributions. One of them is based
on shock wave model [22] while the other is based on a thermodynamic equilibrium
description of the cluster formation process [23]. The first one predicts a power law
with an exponent, δ, equal to 5/3 or 7/3, whereas the second one predicts a δ value
close to 7/3. Molecular dynamics (MD) simulations carried out in metals with ion energies
in the keV range do show emission of clusters in the sputtered species following
an inverse power law in size distribution [20, 21]. Simulations also show a break up
in the larger clusters during their flight. The exponent of the final (or metastable)
clusters size distribution has been found to be 1 to 1.7 times the same for the nascent
(or new born) clusters, i.e., the clusters existing just after ejection [21]. Correcting
52

Ejection of ZnS nanoparticles from ZnS films by Au irradiation 53
the results for such an effect the nascent power law exponent comes down closer to 2.
Because of the fact δ is either close to 2 or lower, the shock wave induced emission
model is preferred over the thermodynamic model which predicts a higher value.
Energetic ion irradiation of a non-elemental compound material is expected to
result, with a certain probability, in the formation of localized spikes that would result
in throwing out materials with the same atomic coordination as in the target. To check
for this we have carried out a study of sputtering from ZnS films using 35 keV, 2, and
100 MeV Au ions [111, 120]. In this chapter we report on the observation of ZnS
nanoclusters, collected on Cu grids coated with amorphous carbon films (hereafter
referred as catcher foil), placed in the reflection geometry.
4.2 Experiments
ZnS thin films (thickness of ∼ 600 nm) were deposited on Si(100) substrates (with native
oxide) by thermal evaporation of 99.99 at.% pure ZnS powder (Materials Research
Corporation, New York) using a vacuum coating unit at a working base pressure of
∼7×10 −6 Torr. The thickness and the rate of deposition of the films were measured
in situ using a quartz crystal thickness monitor. This way prepared films were having
wurtzite (002) crystal structure with an optical band gap of 3.56 eV. The thickness
and composition of the as-grown film were measured using Rutherford backscattering
spectrometry, which showed the film were having ∼ 615 nm thick with Zn and
S concentration ratio as one. Details of these studies were discussed in Chapter 3.
After the deposition, the as-grown ZnS films on Si(100) were irradiated with Au at
three different energies viz. 35 keV (Au − ), 2 MeV (Au 2+ ), and 100 MeV (Au 8+ ) at
an incidence angle of ≈ 45 ◦ . In all the cases, the irradiation fluence was varied from
1×10 11 to 1×10 15 ions cm −2 at a beam current density of ∼ 0.04 µA cm −2 . Au irradiations
at 35 keV and 2 MeV were carried out at room temperature, and 100 MeV Au
irradiations were carried out at two different temperatures viz. room temperature and
liquid nitrogen temperature. The particles sputtered during irradiation were collected
on catcher foils, commonly used for transmission electron microscopy (TEM) studies.
The separations between the catcher foils and the sample surfaces were ∼ 0.5 cm for
35 keV as well as for 2 MeV. However, for 100 MeV Au irradiation the separation
between the catcher foil and the sample surface was ∼ 1 cm. In each of the cases the
catcher foil was placed making an angle of ∼ 30 ◦ (with sample surface) in a reflection
geometry as shown in Fig. 4.1. TEM measurements were carried out to study the size

Ejection of ZnS nanoparticles from ZnS films by Au irradiation 54
Au beam
45 0
ZnS film
Catcher
foil
Sputtered
particles
Figure 4.1: Schematic diagram illustrating
the sputtering experiment.
Si substrate
distributions of the sputtered nanoparticles (NPs) collected on the catcher foils using
a JEOL 2010 UHR TEM (operating at 200 kV). Also high resolution TEM (HRTEM)
and selected area electron diffraction (SAED) techniques were employed to determine
the crystal structure of the sputtered NPs.
4.3 Results
4.3.1 Ejection of ZnS NPs due to 2 MeV Au irradiation
Crystal structure of the ejected NPs
a
b
Figure 4.2: HRTEM micrographs of the NPs corresponding to 2 MeV Au irradiation to the
fluences of (a) 1×10 11 and (b) 5×10 13 ions cm −2 .
Figure 4.2(a) shows an HRTEM micrograph of a large nanocluster on catcher foil,
obtained with a sample irradiated with 2 MeV Au 2+ ions to a fluence of 1×10 11 cm −2 .
For the above low fluence we have observed only very few NPs on the catcher foil.

Ejection of ZnS nanoparticles from ZnS films by Au irradiation 55
The HRTEM image clearly shows the crystalline rows of particles with an inter-planar
spacing of 0.307±0.006 nm which closely agrees with the bulk ZnS lattice spacing for
both the cubic (111) and the hexagonal (002) planes. A typical HRTEM micrograph
taken from a NP corresponding to a higher fluence of 5×10 13 ions cm −2 is shown in
Fig. 4.2(b). The image shows, again, the crystalline rows of the particles with an
inter-planar spacing of 0.310±0.006 nm.
However, to further check the structure, an SAED measurement was carried out on
the catcher foil. The obtained SAED pattern taken on the catcher foil corresponding
to a fluence of 5×10 13 ions cm −2 is shown in Fig. 4.3. The image shows spots and
Figure 4.3: SAED patterns of the NPs
corresponding to 2 MeV Au irradiation
to a fluence of 5×10 13 ions cm −2 . The
diffraction rings corresponding to various
planes are indicated in the figure.
rings which have been identified as corresponding to the (002), (101), (110) and (300)
planes of hexagonal ZnS. These are indicated in the figure. The diffraction rings from
(002) and (101) planes are seen to overlap in the form of a broad ring because of very
close inter-planar spacings. In HRTEM we did not see (300) plane due to very small
d-spacing (0.11 nm) which is smaller than the lattice resolution (0.14 nm) of our TEM.
But in diffraction measurement the spots corresponding to (300) plane are very clearly
seen. Thus the NPs on the catcher foil were unambiguously identified as ZnS having
a wurtzite structure.
Size distribution of the ejected NPs
Plane-view bright-field TEM micrographs of particles collected on the catcher foils,
during Au irradiations of ZnS/Si samples, are shown in Fig. 4.4, where (a), (b), (c),
and (d) correspond to the fluences of 1×10 12 , 5×10 13 , 1×10 14 , and 1×10 15 ions cm −2 ,
respectively. The corresponding size distributions are shown in same figure in the right
panels (e), (f), (g), and (h), respectively. In all cases the particle size is seen to vary

Ejection of ZnS nanoparticles from ZnS films by Au irradiation 56
Frequency (%)
40
30
20
10
(e)
0
2 3 4 5 6 7
Particle Size / nm
Frequency (%)
40
30
20
10
(f)
c
Frequency (%)
0
2 3 4 5 6 7
Particle Size / nm
40
30
20
10
(g)
d
Frequency (%)
0
40
30
20
10
2 3 4 5 6 7
Particle Size / nm
(h)
0
2 3 4 5 6 7
Particle Size / nm
Figure 4.4: Bright-field TEM micrographs of ZnS NPs collected on catcher foils. (a),
(b), (c), and (d) correspond to the fluences of 1×10 12 , 5×10 13 , 1×10 14 , and 1×10 15 ions
cm −2 with corresponding size distributions shown in (e), (f), (g), and (h), respectively. The
histograms in (e) – (h) have been fitted using a log-normal distribution function (Eqn. 4.1),
which are also plotted in the figure as solid curves.

Ejection of ZnS nanoparticles from ZnS films by Au irradiation 57
from 2 to 7 nm. For a quantitative comparison, the particle size distributions have
been fitted with a log-normal distribution function of the form
f(x) =
( ) 2
1
√ e − ln(x/xc)
√
2w
, (4.1)
2πxw
where x c and w are the most probable size and the width of the size distribution,
respectively.
From the left panels of Fig. 4.4, the particle number density in a given frame
is seen to increase with increase in fluence. For fluences, at and below 5×10 13 ions
cm −2 , there is practically no overlap between observed NPs. The corresponding size
distributions are seen to remain almost the same with a mean size of 3.26 nm and a
width of ∼ 0.35 nm. With increase in fluence, beyond 5×10 13 ions cm −2 , a slight shift
in the mean towards a higher value of ∼ 3.63 nm with a corresponding increase in the
width has been observed. The variation of the width of the particle size distributions
with fluence is shown in Fig. 4.5(a).
The width remains practically the same up to
Width of the particle
size distribution (nm)
0.55
0.5
0.45
0.4
0.35
0.3
(a)
1 10 100 1000
12
Ion Fluence (10 ions cm
-2
)
Surface coverage (%)
80
60
40
20
0
(b)
0 250 500 750 1000
12
Fluence (10 ions cm -2 )
Figure 4.5: Variation of (a) the width of the particle size distribution and (b) the catcher
foil surface coverage by the ejected NPs as a function of fluence.
a fluence of 5×10 13 ions cm −2 , beyond which there is a noticeable rise. The rise in
width and mean size appear to be primarily due to overlapping of sputtered particles
(atoms, molecules and clusters) on the catcher foil. Birtcher et al. [16], in their in-situ
TEM studies of Au NPs, have seen mass flow and growth in size of sputtered NPs
under continuous irradiation of Au thin films with 400 keV Xe ions. At this point it is
also important to note that for fluences in the range of 1×10 12 to 1×10 14 ions cm −2 ,
the percentage of the surface (on the catcher foil) covered by the NPs has been found
to grow linearly from a value of about 5 % to about 32 % [Fig. 4.5(b)]. A similar
behavior has also seen by Birtcher et al. [16]. However, there is a deviation at the

Ejection of ZnS nanoparticles from ZnS films by Au irradiation 58
highest fluence considered where the NPs have been found to cover about 67 % of the
catcher foil area. We argue, the above deviation is due to some changes in surface
topography of the irradiated ZnS film coming from continuous irradiation. In fact, we
have seen a change in surface morphology leading to an increase in surface roughness
of the as-deposited ZnS film after 2 MeV Au irradiation to a fluence of 1× 10 14 ions
cm −2 [Fig. 3.6 (c)]. Such changes are expected to affect the sputtering yield [22].
For example, let us imagine a situation where a high density cascade is formed in a
certain depth under a cone causing the emergence of a shock wave. As it propagates
towards the vortex, the cumulation effect increases the wave energy density. When
it reaches a certain magnitude, the cone top breaks off and flies away leading to a
different sputtering yield as compared to that in case of a plane surface [22].
4.3.2 Ejection of ZnS NPs due to 100 MeV Au irradiation
Crystal structure of the ejected NPs
To check the crystal structures of the ZnS NPs ejected during 100 MeV Au irradiation
at room temperature, we have also carried out some HRTEM measurements.
Figure 4.6 shows an HRTEM micrograph of a NP, deposited on the catcher foil. The
HRTEM image clearly shows lattice planes with an inter-planar spacing of 0.311±0.006
nm. This together with earlier data clearly indicates that the ejected NPs are having
wurtzite ZnS structure.
0.311 nm
Figure 4.6: HRTEM micrograph of a NP
corresponding to 100 MeV Au irradiation
to a fluence of 1×10 13 ions cm −2 .
Size distribution of the ejected NPs
Figure 4.7(a) shows a bright-field planar TEM micrograph of the sputtered particles
on the catcher foil due to 100 MeV Au ion irradiation to a fluence of 1×10 13 ions
cm −2 at room temperature. The observed particles were 2−5 nm in size with a most

Ejection of ZnS nanoparticles from ZnS films by Au irradiation 59
a
Frequency (%)
40
30
20
10
(b)
0
1 2 3 4 5 6
Particle size / nm
Figure 4.7: (a) Bright-field planar TEM micrograph of ZnS NPs collected on catcher foil
during 100 MeV Au irradiation at room temperature. (b) shows corresponding size distribution
of the ZnS NPs.
probable size of about 3 nm [Fig. 4.7(b)]. It was observed that the average number
of NPs per frame (i.e., number density) was about 61. However, the average number
of NPs per frame was about 57 in case of 2 MeV Au irradiation to a fluence of
5×10 13 ions cm −2 . Therefore, a comparison of the TEM images of NPs taken from
the catcher foils corresponding to 2 MeV Au ion irradiation with a fluence of 5×10 13
ions cm −2 [Fig. 4.4(b)] and 100 MeV Au ion irradiation with a fluence of 1×10 13 ions
cm −2 [Fig. 4.7(a)] showed that the particle (number) densities were nearly same even
though the distance between the catcher foil and the sample surface for later case was
two times that in case of 2 MeV.
Figure 4.8: Bright-field TEM micrograph
taken on the catcher foil for
the 35 keV Au irradiation to a fluence
of 1×10 14 ions cm −2 .
It is important to mention here is that we have not observed any NP larger than
1 nm to be present on the catcher foil after 100 MeV Au irradiation of ZnS films kept
at liquid nitrogen temperature. In addition, for the 35 keV Au irradiation case, we
did not observe any NP to be present on the catcher foil. A typical bright-field planar

Ejection of ZnS nanoparticles from ZnS films by Au irradiation 60
TEM images taken on the catcher foil for the above case is shown in Fig. 4.8.
4.3.3 A power law behaviour in nanoparticle size distribution
In order to infer about possible mechanisms behind the NPs ejection, it was decided to
look at the size distributions of the ejected particles. For this, many TEM micrographs
of the same frame size (as shown in Fig. 4.4 and Fig. 4.7) were analyzed for each fluence.
The number of visible particles of different sizes, for each frame, was determined using
the ImageJ software [128]. At 2 MeV, Au irradiation was carried out at four different
fluences, F1, F2, F3, and F4 corresponding to 1×10 12 , 5×10 13 , 1×10 14 , and 1×10 15
ions cm −2 , respectively. Corresponding to these four fluences we have analyzed a total
of 52, 342, 455, and 392 particles of different sizes for obtaining the corresponding size
distributions. However, there is only one fluence, F5, which is 1×10 13 ions cm −2 for
the 100 MeV irradiation in which case there were a total of 223 particles observed in
different frames. The results for various irradiations at different fluences, as mentioned
above, were grouped into 0.5 nm bins to get the yield, Y , as a function of NP volume.
Figure 4.9 shows the results, where the size distributions are plotted taking the
100
Frequency [%]
10
1
0.1
▼ 100 MeV
2 MeV
12 2
◆ 1x10 /cm
13 2
■ 5x10 /cm
14
▲
2
1x10 /cm
15
● 1x10 /cm 2
10 100
Hemisphere Volume [nm 3 ]
Figure 4.9: Measured number of
collected particles (having diameter
≥3 nm) as a function of hemisphere
volume. Experimental data
are fitted, with a function given in
Eqn. 4.2, and shown in the figure as
a straight line for each set of data.
frequency (in terms of % of the measured number of particles) as a function of NP
volume. For the volume calculation all the particles were assumed to be hemispherical
in nature. To avoid any overlapping among the data points the size distributions for
fluences F1, F3, F4 and F5 were shifted in the frequency axis. Following earlier works
on cluster size distributions [16, 18, 21, 22, 23, 121], we have tried to fit our data on

Ejection of ZnS nanoparticles from ZnS films by Au irradiation 61
cluster yield, Y , as obtained for larger sizes, to power law distributions of the form
Y (n) = Y 0 n −δ , (4.2)
where n is the number of atoms in a cluster; Y 0 and δ being fitting parameters.
The power law fits to our experimental data (for the particles of diameter ≥3 nm)
yield 2.56, 2.70, 2.59, 2.14, and 3.45 as the values for the exponent δ for fluences
F1, F2, F3, F4 at 2 MeV and fluence F5 corresponding to irradiation at 100 MeV,
respectively. The errors on the estimated δ values are found to be ≤ ± 0.09. As has
been mentioned in the introduction (Section 4.1), MD simulations of clusters emission
do show fragmentation of clusters during their flight path [21]. Figure 4.10 shows
Figure 4.10: Snapshots from a simulation of sputtering of Au at different times for 20 keV
Xe incident on Au. The labels A and B show clusters that are fragmenting, and C illustrates
late sputtering (a cluster separating from the surface after the displacement cascade has
ended). This figure is adapted from Ref. [21].
snapshots of a MD simulation of sputtering of Au at different times (between 16 and

Ejection of ZnS nanoparticles from ZnS films by Au irradiation 62
19 ps) for 20 keV Xe incident on Au. The fragmentation of nascent clusters has
been shown to have an important effect on the size distribution. It is to be noted
that the experimentally observed power law exponents are valid only for the final
clusters, formed after the break up of nascent clusters in the flight. Therefore it is
inappropriate to compare these experimental values of the exponent δ, as obtained for
the final clusters, to those predicted by theory. However, there is no way to detect
clusters or carryout measurements at the rapid time scale of cluster formation within
few ps of the ion impact. In such a scenario the only way out is to use a correction
factor from MD simulations.
In the MD simulation [21] as mentioned earlier, it has been shown that the ratio,
r, of the exponent δ final for the final clusters size distribution to δ nascent that for the
nascent clusters is equal to 1.2±0.2 for 15 keV Xe ion irradiation on Ag. The same
has been found to be 1.3±0.4 for 20 keV Xe ion irradiation on Au. In yet another
simulation study, carried out using the embedded atom method (EAM) for Ar ion
irradiation on fcc metals with energies between 100 eV and 5 keV, the same ratio r
has been estimated to be 1.3 and 1.4 [20]. Following the assumption made in ref.
[21], that the ratios r have a real physical basis and are transferable to other ion and
substrate types, we have estimated the nascent exponents for the present experiment
taking r=1.3. The estimated values of the nascent exponents are then found to be 1.97,
2.08, 1.99, 1.65, and 2.66 for the cases F1, F2, F3, F4, and F5 respectively. One can
see that both the final and the nascent exponents for the ion fluence F4, corresponding
to 1×10 15 Au cm −2 , are slightly smaller than those for the other fluences at the same
incident ion energy of 2 MeV. This is mainly due to the overlapping of the emitted
NPs on the catcher foil [120]. It is to be noted that the exponent for 2 MeV Au
irradiation, with the exception of the case with the highest fluence, are all close to 2
while that for 100 MeV Au irradiation is distinctly higher.
4.4 Discussion
4.4.1 NP ejection mechanisms
Sputtering from compound targets during ion bombardment can take place in the
form of individual atoms, molecules, and sometimes small clusters consisting of a
few atoms. It has been shown that ion bombardment induces compositional change
in the sputtered surface mainly due to preferential sputtering of one element over

Ejection of ZnS nanoparticles from ZnS films by Au irradiation 63
another and radiation induced segregation [129, 130]. In such a case, one can expect
dissimilar sputtering of constituent atoms, mainly due to their mass and binding
energy differences. In addition, there may be a possibility of having a thermodynamic
reaction between the sputtered atoms on the catcher foils leading to the formation of
non-stoichiometric clusters. In contrast to this, our HRTEM micrographs and SAED
patterns show that the larger NPs correspond to wurtzite ZnS which is stoichiometric.
As has been shown in Fig. 4.2 (a) crystalline NPs of size ∼ 5 nm could be seen
on the catcher foil even at a low irradiation fluence of 1×10 11 ions cm −2 . We must
also mention that the ion current density of 0.04 µA cm −2 amounts to the incidence of
about 10 −3 Au 2+ ions nm −2 sec −1 . This means the observed effect is only due to single
ion impact. At a fluence of 1×10 11 ions cm −2 we would expect much less sputtering of
individual atoms, molecules or small clusters which would be also spatially separated.
Therefore one would not expect any reaction taking place on the catcher foil leading
to the formation of larger ZnS NPs with sizes ∼ 5 nm. In fact, at that fluence we
did not find many of those to get a meaningful size distribution. Further, the ejection
of small clusters can be understood based on a statistical model in which cluster
emission is treated as a more or less uncorrelated ejection of constituent atoms. Since
the statistical probability for independent ejection of many atoms into essentially the
same phase space interval is exceedingly small, it is apparent that clusters containing
more than a few atoms cannot be formed in such a way [18].
Although experimental results regarding cluster ejection, in some cases, could be
quantitatively reproduced by MD simulations, a rigorous theoretical explanation of the
observed power law drop in the size distributions is yet to be found. Further, there
is a strong variation in the exponent, δ, for small and large clusters which cannot be
accounted for in any (presently) available theoretical models.
It must be added here that for 35 keV Au ions in ZnS the nuclear and the electronic
energy losses, S n and S e , respectively, as determined from SRIM simulation [31] are
2.91 and 0.24 keV nm −1 , respectively. For 2 MeV Au ions they are 3.66 and 1.91 keV
nm −1 , respectively. The fact that we could not get any NP on the catcher foil at 35
keV while there were NPs for Au irradiations at 2 MeV means S e plays a dominant role
regarding clusters ejection in ZnS. This is particularly so because S n values are not too
far different in the two cases. The SRIM calculations also show that the displacement
cascades in the ZnS for 35 keV Au ions, incident at 45 ◦ , are very close (≤10 nm) to
the target surface [Fig. 4.11(a)]. This would also imply that within the Bitensky and
Parilis model “spherical” shock waves, formed purely from nuclear collision cascades,

Ejection of ZnS nanoparticles from ZnS films by Au irradiation 64
(a)
35 keV Au in ZnS
(b)
2 MeV Au in ZnS
incident ion trajectory
displaced Zn atoms
displaced S atoms
0
10 20
Target Depth [nm]
0
200 400
Target Depth [nm]
600
Figure 4.11: SRIM simulated collision cascades as a function of depth for (a) 35 keV and
(b) 2 MeV Au in ZnS at an incident angle of 45 ◦ with the target normal.
would originate at depths close to 10 nm or less. Not finding any NP at 35 keV would
also mean these shock waves, if produced, die down almost totally while traversing a
material. On the other hand, with 2 MeV Au ions dense displacement cascades coming
from nuclear collisions will be produced in depths greater than 50 nm [Fig. 4.11(b)].
This does not mean that a 2 MeV Au ion would not undergo a collision generating a
collision cascade in the near surface region. All that it means is that denser collision
cascades, leading to sizable displacements which can result in a shock waves, are
mostly produced at larger depths (> 50 nm). The corresponding shock waves would
certainly not be able to reach the surface. However, 2 MeV Au ions lose about 2 keV
nm −1 through electronic excitations and ionization. With low hole mobility this can
result in Coulomb explosion and “cylindrical” shock waves production in a transverse
direction (around the ion track) in the near surface region. This would happen in the
Bitensky and Parilis model particularly when the incident ion falls at an acute angle
with the sample surface.
At 2 MeV, the Coulomb explosion may result in the production of a thermal spike
[45] resulting in local melting and mass flow onto the surface. If the temperature goes
above that required for vaporization then material in vapor phase can be ejected. Upon
cooling down there may be cluster formation and in such a case the thermodynamic
model of Urbassek would be valid leading to a δ value close to 7/3. Since all values
of δ nascent as obtained at 2 MeV, for various fluences, are just about 2 it is unlikely
that the thermodynamic model is valid. However, at 100 MeV incident energy, S e
is about 16 keV nm −1 . The thermal spike produced because of Coulomb explosion

Ejection of ZnS nanoparticles from ZnS films by Au irradiation 65
at this irradiation energy would result in vaporization of a large chunk of material
around the ion track. The evaporated material can thermalize and undergo cooling
resulting in cluster formation. The δ final value observed in this particular case has
been found to be about 3.45. If we take an r value of 1.5, which can be the case
within errors quoted [21], the δ nascent value can be shown to be almost equal to 7/3.
Therefore the NPs emitted from 100 MeV Au irradiation do seem to go in line with
the thermodynamic equilibrium liquid-gas-type phase transition model rather than
the Bitensky and Parilis model. As mentioned earlier, with a dose 1/5 times that at
2 MeV and the catcher foil set at twice the distance, the emitted particle flux from
the surface at 100 MeV was almost the same as that at 2 MeV. This means at the
same fluence, irradiation at 100 MeV results in an order of magnitude higher flux of
particle emission, as compared to that at 2 MeV. So much of material liberated from
the irradiated surface is expected to form a gas that may expand, cool and condense,
finally fragmenting into clusters which get deposited on the catcher foil.
4.4.2 Temperature dependence of ZnS NP ejection under 100
MeV Au irradiation
The difference in the NPs sputtering with 100 MeV Au ions at two different irradiation
temperatures can be understood using a two component inelastic thermal spike model
(see Section 1.2.4 of Chapter 1). This model predicts a local temperature rise, within
a very short time, during the transfer of energy from the excited electrons to the
lattice atoms via electron-phonon coupling, g. The coupling constant g is inversely
proportional to the thermal conductivity, K(T), and is given by [125]
g = π4 (k B z n l v s ) 2
, (4.3)
18K(T)
where k B is the Boltzman constant, v s the sound velocity, n l the atomic density, and z
the number of electrons participating in the thermal spike. It is observed that thermal
conductivity strongly depends upon temperature [131]. In the case of wurtzite ZnS,
at 80 and 300 K, the values of the thermal conductivities are about 0.95 and 0.27 W
cm −1 K −1 , respectively [131]. Therefore, the thermal conductivity at 80 K, which is
mainly due to electrons, is around 3.5 times that of the same at 300 K i.e.at room
temperature. It therefore means that g, the electron-phonon coupling parameter, will
have lower value at 80 K than the same at room temperature. This means at 80 K
only a small amount of energy will be given to a localized region in the lattice, most

Ejection of ZnS nanoparticles from ZnS films by Au irradiation 66
of the energy being taken away by the surrounding medium. On the other hand,
at room temperature more energy is given to the lattice which stays confined for a
relatively longer time because of lower thermal conductivity resulting in the production
of a thermal spike. This would result in vaporization of material which can result in
clusters ejection following the thermodynamic equilibrium model of liquid-gas type
phase transition.

Chapter 5
Anisotropic deformation of Au
nanoparticles in silica glass by swift
heavy ion irradiation
5.1 Introduction
Metallic nanoparticles (NPs) embedded in silica are interesting materials because of
their nonlinear optical properties with potential technological applications in optoelectronics.
The optical properties of these NPs are mainly determined by their surface
plasmon resonance (SPR) [132]. When metallic NPs are excited by electromagnetic
radiation, their conduction electrons exhibit collective oscillations known as localized
SPRs. The wavelength λ max corresponding to the optical absorption maximum of the
SPR is highly dependent on the size, shape, dielectric properties of the metallic NP,
and their surrounding. Because of the SPR, Ag and Au NPs embedded in optical
media have been found to result in an enhancement of the electric field near their
vicinity, that can be used for pattern transfer and nanolithography. In certain cases,
illuminations with unfocused light allows optically addressing particles either individually
or specific controlled configuration, depending upon wavelength, polarization or
the incident angle of exciting radiation [133].
For anisotropically shaped metal NPs embedded in optical media there is a splitting
of the SPR band with increase in aspect ratio (AR) [12, 134, 135]. For a large
enough AR, one of the plasmon bands, in certain cases, have been shown to shift well
into the infrared, thus enabling the system to be important in applications involving
telecommunication [136]. There is also an enhancement in nonlinear optical response
67

Anisotropic deformation of Au NPs by SHI irradiation 68
when compared to spherically shaped NPs [137].
However, it is difficult to synthesize NPs, embedded in a matrix, with the desired
AR using conventional chemical and physical routes. This is because of the requirement
of minimum surface energy for stability, which is true for spherical particle.
In such a scenario, MeV ion irradiation provides a unique possibility for shaping of
NPs with the desired AR. In these applications, it is usually the electronic energy
loss which results in inelastic thermal spikes that can create channels in the dielectric
medium containing the NPs. In case the spikes are formed in the NPs they
may melt and get squeezed into the ion tracks formed resulting in elongated structures.
In view of this, in recent times, a lot of attention has been paid to study
swift heavy ion (SHI) (E/amu ≥ 1 MeV) irradiation induced modification of NPs
[12, 115, 134, 138, 139, 140, 141, 142, 143]. For example, embedded spherical Co NPs
were found to be elongated under SHI irradiation resulting in anisotropic magnetic
properties [115]. 30 MeV Se irradiation of metallic dielectric colloids with an Au core
and silica shell structure has been found to result in different effects on the core and
the shell. The irradiation turned the spherical silica shells into oblate ellipsoids while
the spherical metal cores transformed into prolate ellipsoids [134]. Similarly, 30 MeV
Cu irradiation has been found to induce an anisotropic plastic deformation in colloids
consisting of a silica core (diameter 300–500 nm) and Au shell (thickness 20–60 nm)
structure. The shape of the spherical colloids changed into oblate ellipsoids, with
the final degree of anisotropy depending upon the ion fluence [142]. Also, embedded
spherical Au NPs were found to be deformed anisotropically under 120 MeV self-ion
irradiation [138]. The elongated Au NPs showed a splitting up of the SPR bands.
In most of these studies, anisotropic deformation of NPs have been reported, without
any growth of the particles. In a very recent work, Mishra et al. have shown
growth of Au NPs due to 90 MeV Ni irradiation [144]. Their study showed that the
size of Au NP increased with increase in fluence. However, the authors did not observe
any shape deformation of the NPs. In contrast to these, in the present study,
we demonstrate that size and shape of Au NPs in silica matrix can be controlled,
simultaneously, using SHI irradiations.
5.2 Experiments
Au NPs were synthesized, inside high purity silica glass, by implantation of 32 keV
Au − ions to a fluence of 4×10 16 cm −2 . The low energy implantations were performed

Anisotropic deformation of Au NPs by SHI irradiation 69
at normal beam incidence, at room temperature, using the low energy negative ion
implantation facility available at Institute of Physics (IOP), Bhubaneswar [64]. A set
of as-implanted samples were annealed in air at 850 ◦ C for 1 h for obtaining embedded
Au NPs of a certain reasonable size. Hereafter, the annealed samples will be referred
to as the “as-grown” samples. Three of the as-grown samples were irradiated, at room
temperature, using 100 MeV Au 8+ ions at normal incidence to fluences F2, F5, and F10
with values of 2×10 13 , 5×10 13 , and 1×10 14 ions cm −2 , respectively. This was done
using the 15 MV Pelletron accelerator at Inter University Accelerator Centre, New
Delhi. For comparison, an as-grown samples was irradiated with 10 MeV Au 4+ ions to
a fluence of F10 at IOP, Bhubaneswar. All the samples were subjected to transmission
electron microscopy (TEM) measurements to image the Au NPs inside silica matrix
using a JEOL 2010 UHR TEM (operating at 200 kV) at IOP, Bhubaneswar. The TEM
samples were prepared by the standard preparation techniques as described in Section
2.3.3 in Chapter 2. The Au content in silica, before and after SHI irradiation, was
estimated by Rutherford backscattering spectrometry (RBS) with 1.35 MeV 4 He + ions
using the 3 MV Pelletron accelerator at IOP, Bhubaneswar. The detector was placed
at 135 ◦ with respect to the incident beam direction. Such a glancing angle geometry
was used to improve the surface sensitivity. Optical absorption (OA) measurements
were carried out on the as-grown and the SHI irradiated samples using a Shimadzu
PC3101 UV-VIS-NIR dual beam spectrophotometer, available at IOP. OA spectra
were recorded in the transmission mode with a blank silica glass in the reference line.
5.3 Results
5.3.1 Synthesis of Au NPs in silica glass
A typical bright-field cross-sectional TEM (XTEM) micrograph and the corresponding
NP size distribution for the sample implanted with 4×10 16 ions cm −2 , are shown in
Fig. 5.1(a) and (c), respectively. NPs with larger size (∼ 6 nm) are found to be closer
to the surface, with smaller clusters lying in the deeper region. High resolution TEM
(HRTEM) image of one of the particles, shown in the inset of Fig. 5.1(a) shows that
the NPs are crystalline in nature. The lattice spacing in the NP has been found to be
about 2.2 Å, which is very close to (111) inter-planar spacing of the fcc Au crystal.
Annealing the as-implanted samples in air at 850 ◦ C for 1 h has been found to increase
the NP size, which is shown in Fig. 5.1(b). The corresponding NP size distribution is

Anisotropic deformation of Au NPs by SHI irradiation 70
5 nm
Surface
(a)
20
(c)
Frequency [%]
15
10
5
0.22 nm
0
1 2 3 4 5 6 7 8
Diameter [nm]
(b)
15
(d)
Surface
Frequency [%]
10
5
0
2 4 6 8 10 12 14 16
Diameter [nm]
Figure 5.1: Bright-filed XTEM micrographs of the Au NPs in silica glass after 32 keV Au −
ions impantation to a fluence of 4×10 16 cm −2 (a) before and (b) after annealing in air at
850 ◦ C for 1 h. (c) and (d) show the corresponding size distributions, respectively. Inset in
(a) shows an HRTEM image of a Au NP indicating the particles are crystalline in nature.
shown in Fig. 5.1(d). The XTEM micrograph also shows that there is redistribution
of the NP size and size distribution resulting in a Guassian-like distribution. Fitting
of a Gaussian function to the size distribution leads to a mean size of 6.6±0.2 nm with
a standard deviation of 2.4±0.1 nm. The growth of the NP size is attributed as the
Ostwald ripening process which occurs at the expense of the smaller particles.
As has been mentioned earlier, metal NPs can be probed by the OA spectroscopy.
This is because the OA technique is a very effective tool to measure the metal NP size
and shape very accurately [134, 135, 143] through the SPR. Figure 5.2 shows the OA
spectra for the Au implanted glass samples before and after annealing. The spectrum
corresponding to the as-implanted sample reveals the presence of Au NPs in silica glass
[143] with an SPR peak around 543 nm, which is due to the free electron oscillations in
small metal particles when excited by electromagnetic radiation [145]. Annealing the

Anisotropic deformation of Au NPs by SHI irradiation 71
absorbance (arb. units)
as-implanted
annealed
Figure 5.2: OA spectrum of
Au implanted and annealed silica
glass samples.
320 400 480 560 640 720
wavelength (nm)
as-implanted sample leads to an increase in the SPR peak intensity with a noticeable
reduction in the full width at half maximum (FWHM) of the SPR peak as shown in
Fig. 5.2. Such an enhancement in the SPR peak with a corresponding reduction in its
FWHM is a clear indication of the growth in size of NPs inside the matrix [145].
5.3.2 Shape deformation of the NPs: TEM studies
Figure 5.3(b) shows the bright-field XTEM micrograph of the as-grown sample irradiated
with 100 MeV Au 8+ ions at a fluence F2. Clearly, after irradiation, there is an
elongation of Au NPs along the ion beam direction (with a prolate shape). The image
also shows only bigger NPs are elongated, whereas the smaller ones remain spherical
in shape. For a comparison, XTEM micrograph of Au NPs and their size distribution
corresponding to the as-grown sample is also shown in Fig. 5.3(a) and (e), respectively.
Increasing the fluence to F5 (two and half time that of F2) is seen to result in a further
elongation of the NPs (with increased anisotropy). This is shown in Fig. 5.3(c). The
estimated average ARs of the elongated NPs are about 1.21 and 1.55 for irradiations
with fluences F2 and F5, respectively. The TEM images also show a growth of NPs
with increase in fluence. However, in case of the sample irradiated with the highest
fluence F10, the TEM image shows only spherical NPs with a very large interparticle
distance [Fig. 5.3(d)]. A Gaussian fitting of the corresponding size distribution

Anisotropic deformation of Au NPs by SHI irradiation 72
15
(e)
Frequency [%]
10
5
20 nm
(a)
0
2 4 6 8 10 12 14 16
Diameter [nm]
30
40
30
(f)
Frequency [%]
20
10
Frequency [%]
20
10
(b)
(c)
(d)
Frequency [%]
0
4 8 12 16 20 24
Long Axis Length [nm]
30
25
20
15
10
5
0
8 12 16 20 24
Long Axis Length [nm]
40
Frequency [%]
0
4 8 12 16 20
Short Axis Length [nm]
40
30
20
10
(g)
0
4 8 12 16 20
Short Axis Length [nm]
(h)
10 nm
Frequency [%]
30
20
10
0
4 6 8 10 12 14 16
Diameter [nm]
Figure 5.3: (a)−(d) Bright-filed XTEM micrographs of the as-grown and 100 MeV Au
irradiated samples with fluences of 2×10 13 , 5×10 13 , and 1×10 14 ions cm −2 , respectively;
(e)−(h) show the corresponding size distributions, respectively. Note that (b) and (c) are
having same scale as in (a). Dashed lines are indicating the surfaces.

Anisotropic deformation of Au NPs by SHI irradiation 73
[Fig. 5.3(h)] gives a mean value of 9.2±0.1 nm with a standard deviation of 0.98±0.09
nm. Note that this average Au NP size is slightly larger than that of NPs in the the
as-grown (unirradiated) sample which is about 6.6 nm. Also, for the highest fluence
(F10) irradiated sample, a very narrow distribution is found. This corresponds to a
standard deviation of 0.98 nm, which is much lower than that for NPs in the as-grown
sample which is about 2.4 nm. Interestingly, at the irradiation fluence F10, what has
been seen is a single layer, (i.e., a two-dimensional array) of well separated Au NPs
in the matrix, just below the silica glass surface. This is seen very clearly in a planar
TEM micrograph taken of this sample as shown in Fig. 5.4. The image clearly shows
well separated NPs. The image also clearly shows a contrast difference among the
NPs. The larger NPs are noticeably darker than the smaller NPs. This indicates that
the Au NPs are present almost in the same plane.
20 nm
Figure 5.4: Bright-field planar TEM micrograph
of the 100 MeV Au irradiated
sample for the fluence of 1×10 14 ions
cm −2 .
It is important to mention here that the XTEM images show a decrease in Au
content after SHI irradiation and the Au loss increases with increase in ion fluence. It
is also important to mention here that no observable modification has been seen in the
Au NPs after an irradiation of the as-grown sample by 10 MeV Au ions to fluence of
1×10 14 ions cm −2 (F10). This loss in Au which occurs through an outward movement
and subsequent evaporation is discussed in the following section.
5.3.3 Outward movement and loss of Au: RBS studies
Variation of Au concentration in the depth direction inside silica glass at different
SHI irradiation fluences were determined through an analysis of RBS data. Figure
5.5(a) shows RBS spectra of the samples before and after 100 MeV Au irradiations.
To compare the data for different samples, all the spectra have been normalized using
the silicon profile which is marked in the figure. The figure clearly shows a decrease

Anisotropic deformation of Au NPs by SHI irradiation 74
Normalized Yield [counts]
250 (a)
200
150
100
50
Si
■ ■ ■
▲ ▲ ▲
● ● ●
▼ ▼ ▼
as-grown
13
2x10
13
5x10
14
1x10
Au
0
400 500 600 700 800
Channel Number
Au Sputtered [%]
100
80
(b)
60
40
20
0
0 2 4 6 8 10
13
Fluence [10 ions cm
-2
]
Figure 5.5: (a) RBS spectra taken from the as-grown samples before and after SHI irradiations.
The Au and Si RBS profiles are indicated in the figure. (b) Au sputtered from the
samples after SHI irradiation with different ion fluences.
of backscattered yield corresponding to the Au signal with increase in fluence. The
spectra also show a shift in the Au signal, coming from deep inside the matrix, towards
higher channel number with increase in fluence. The glancing angle geometry for the
RBS measurements helped to observe the small shift in channel number. Decrease
and shift, towards higher channel number, of backscattered yield corresponding to
Au indicate reduction and outward movement of Au towards the silica glass surface.
Figure 5.5(b) shows the variation of Au content inside the silica matrix as a function
of fluence. As can be seen from the figure, there is a gradual increase in the amount of
sputtered Au with increase in fluence. For a SHI irradiation fluence as given by F10,
it is found that almost 90 % of the Au earlier present in the as-grown sample is lost.
5.3.4 Shape deformation of the NPs: OA studies
Figure 5.6 (a), (b), (c), and (d) show the OA spectra corresponding to the unirradiated
(as-grown) and the SHI irradiated samples for the fluences F2, F5, and F10, respectively.
One can clearly see an SPR peak which is very prominent in the unirradiated
sample. From Fig. 5.6, one can see a reduction of SPR peak intensity with increase
in fluence. It is also important to note, from XTEM and RBS data [Figs. 5.3 and 5.5,
respectively], that there is a reduction of Au content in the matrix after irradiation
which increases with fluence. This reduction of the SPR peak intensity is mainly due
to the reduction in the Au content in the matrix. Clearly, OA spectrum corresponding
to a fluence F5 shows two distinct SPR peaks. In fact, the spectrum for fluence F2 (as
will be shown later), is a combination of two peaks which are found to be separated

Anisotropic deformation of Au NPs by SHI irradiation 75
(a)
(b)
Absorbance [arb. unit]
(c)
(d)
Figure 5.6: OA spectra
taken on the (a)
as-grown and SHI irradiated
samples corresponding
to the fluences
of (b) 2×10 13 , (c)
5×10 13 , and (d) 1×10 14
ions cm −2 .
400 500 600 700 800 400 500 600 700 800
Wavelength [nm]
at fluence F5. On the other hand the OA spectrum corresponding to a fluence F10
shows a single peak (corresponding to nearly spherical NPs) of very low intensity. For
quantitative analysis, we have fitted all the OA spectra using Gaussian function(s)
riding over a linear background. One such fitting as obtained for the sample irradiated
with Au fluence F2, is shown in Fig. 5.7. The fitting parameters, viz. number of
SPR peaks and peak positions together with their FWHMs for different samples are
listed in Table 5.2.
Table 5.2: Fitting parameters obtained from the optical absorption spectra for as-grown
and irradiated Au NPs in silica glass.
Fluence Peak position FWHM
(ions cm −2 ) (nm) (nm)
as-grown 543 81.2
2×10 13 532 569 73.5 129.9
5×10 13 542 608 65.1 40.9
1×10 14 561 78.3
The deconvoluted spectrum for sample with fluence F2, as shown in Fig. 5.7 show
two peaks. The one at higher wavelength is found to be redshifted with increase in SHI

Anisotropic deformation of Au NPs by SHI irradiation 76
Absorbance [arb. unit]
● ● ● Experimental
Fitted
Figure 5.7: OA spectrum (filled
circles) of the as-grown sample after
2×10 13 ions cm −2 irradiation
along with the fitted (solid line)
curve using two Gaussian functions
with proper background substraction
(dashed lines).
0
400 450 500 550 600 650 700 750
Wavelength [nm]
fluence (showing two clear peaks at F5). There has also been an earlier SHI irradiation
study on Au-core silica-shell particles using 30 MeV Cu ions with a fluence of 2×10 14
cm −2 [134]. The measured optical extinction spectrum showed a broad peak in the
long-wavelength region (600−650 nm). The two peak structure in OA spectrum has
been shown to be a characteristic of elongated Au NPs. Similar two peak structures
in OA spectrum have also been seen in chemically prepared elongated Au NPs [135].
It is also well known that Au nanorods have two plasmon resonance absorptions, one
due to the transverse oscillation of electrons regardless of the aspect ratio, the other
absorption coming from the longitudinal oscillation of the electrons, whose wavelength
maximum depends strongly on the aspect ratio. The present results are in agreement
with this.
5.3.5 Dependence of longitudinal SPR peak position on aspect
ratio
Figure 5.8 shows the variation of the absorption maximum of longitudinal SPR as a
function of average AR. The linear fit of the experimental data (excluding the data
obtained from the sample irradiated with a fluence of F10) is obtained for the following
equation:
λ max = 120.48AR + 422.33, (5.1)
where λ max is the absorption maximum of the longitudinal SPR. In an earlier work,
Link et al. have shown a linear dependence of the absorption maximum of longitudinal

Anisotropic deformation of Au NPs by SHI irradiation 77
λ max [nm]
610
600
590
580
570
560
550
540
13
5x10
1x10 14
2x10
13
as-grown
1 1.1 1.2 1.3 1.4
Aspect Ratio
1.5 1.6
Figure 5.8: Dependence of the absorption
maximum of the longitudinal plasmon
resonance against the average aspect
ratio as determined by XTEM.
The solid line is the linear fit to the
experimental data, except the 1×10 14
ions cm −2 irradiated one.
SPR of Au nanorods as a function of AR, and is given by [135]
λ max = (53.71AR − 42.29)ǫ m + 495.14, (5.2)
where ǫ m is the medium dielectric constant. Taking into account the medium dielectric
constant ǫ m = 2.26 of the surrounding medium (silica glass [146]), Eqn. 5.2 gives
λ max = 121.39AR + 399.57. One can very clearly see that the slope of the linear fit to
the experimental data agrees very well with that obtained from the theoretical estimation.
The shift/intercept can be taken care of through a single point normalization.
5.4 Discussion
From the XTEM and OA data, it is clearly seen that the Au NPs are deformed in
prolate spheroids with their major axis along the ion-beam direction for the lowest
and intermediate fluences (F2 and F5), whereas there are spherical Au NPs in silica
glass after Au irradiation to a fluence of F10. It seems the highest fluence irradiated
sample have gone through all the intermediate stages of the Au NPs as in the F2 and
F5 irradiated samples and finally left out with spherical NPs. Notice that this effect
has never been observed previously. On the other hand, anisotropic deformation has
been seen for Au and Co NPs embedded in SiO 2 after irradiation with Au [138] and
In [115] ions, respectively. Recently, deformation mechanism has also been observed
in Ag NPs implanted in SiO 2 films after irradiation with Si [12]. However, in contrast
with the above we have seen an outward movement and loss of Au in the present
experiment.
In most of these studies, inelastic thermal spike concepts [125] were invoked to

Anisotropic deformation of Au NPs by SHI irradiation 78
explain the observed results. In the thermal spike model, due to the passage of a SHI,
the electronic subsystem is assumed to be excited first which thermalizes in a very short
time and then the excitation energy is transferred to the lattice by electron-phonon
coupling (Section 1.2.4 of Chapter 1). Due to a large energy deposition in a short
time, the material undergoes deformation leading to chemical and physical changes in
nm sized regions. However, the deformation mechanism in all the above cases is not
well understood. For many years, the plastic flow model had been successfully used
to understand the anisotropic deformation of amorphous materials (such as silica)
subjected to ion irradiation [116, 141]. It is shown in the previous works [12, 140]
that the plastic flow model cannot be applied directly to the metallic NPs. In fact,
in the present study, it is expected that each individual Au ion impact in silica leads
to the formation of thermal spike which results in plastic flow of silica [141]. This
plastic flow induces an in-plane stress perpendicular to the ion beam direction [116].
The combined effect of stress [116] and thermal spike [115] in Au-silica nanocomposite
leads to the elongation of Au NPs. If the elongation is only due to the pressure effect,
then the smaller NPs should also elongate, which has not been observed in present
case. Recently, an important refinement to the plastic flow model was proposed by
D’Orléans and co-workers [115]. These authors have invoked the mechanism of track
formation combined with the plastic flow model, assuming that the NPs melts and
flow into the ion track.
The passage of SHI (100 MeV Au) through thin silica layer containing the gold
NPs (the composition of Au, Si, and O has been obtained to be 13%, 29%, and 58%
by fitting the RBS spectrum, corresponding to the as-grown sample, using GISA3.9
code [72]) deposits the electronic energy of 13.5 keV nm −1 [31]. It is known from the
work done by Toulemonde et al. [147] that an electronic energy loss of ∼ 13.5 keV
nm −1 in silica creates latent track with a diameter ∼ 8.5 nm. We, therefore, expect
that 100 MeV Au ions in the present case create the track of diameter about 8.5
nm. The material within the ion tracks goes to the molten state due to thermal spike
produced by the energetic ion for a short duration of time (∼ 10 −12 s). The movement
of Au atoms and smaller particles towards bigger particles leads to Ostwald ripening
of Au NPs in the latent track around the ion path. Normally, it is expected that
the growth of particles should not exceed the track diameter, whereas in the present
case the particle growth is observed to be occurring with fluence little larger than the
track diameter. This can be explained by the fact that the temperature of silica in the
annular region around the ion track is high enough to melt the Au particles transiently.

Anisotropic deformation of Au NPs by SHI irradiation 79
The high temperature of silica leads to Ostwald ripening of Au NPs. As the fluence
increases, the overlapping of latent tracks occurs, which leads to further agglomeration
of Au NPs within overlapped ion tracks. To understand the shape deformation, as
suggested earlier by D’Orléans et al. [115], the molten material can be squeezed out
due to the difference in volume expansion and compressibility between liquid metal
and matrix. This would result in deformation in shape. Therefore, it is more likely
that the NPs grow when their size is smaller than the track size and elongate when
their size is larger than track size. The observed movement of the Au NPs towards
the silica glass surface by the SHI cannot be understood within the framework of
a thermal spike model alone. This is because of the thermal spike model does not
include density and pressure effects [125]. One needs to consider the pressure pulse
coming from fast density changes in the ion track to explain the observed movement of
Au towards silica glass surface. Similar movement of Au, towards the sample surface,
during 90−350 MeV Ar-, Kr-, Xe- and Au-ions irradiation of Au marker layer in NiO
has also been reported [148].
The process of elongation depends on the product of S e Φ, where S e and Φ are
electronic energy loss and fluence, respectively. The ARs achieved in the previous
studies were ∼ 3.5 and ∼ 8.4 for 120 MeV Au (S e =14.1 keV nm −1 and Φ=3×10 13
ions cm −2 ) [138] and 30 MeV Se (S e =6.2 keV nm −1 and Φ=2×10 14 ions cm −2 ) [134]
irradiations of silica-Au nanocomposite, respectively. It is clear from these data that
when the product of S e Φ changes by approximately three times, the AR changes by a
similar factor. In the present case, the values of S e Φ for the fluences of F2 and F5 are
13.5 keV nm −1 ×2×10 13 ions cm −2 and 13.5 keV nm −1 ×5×10 13 ions cm −2 , respectively.
These S e Φ have resulted in the average ARs of 1.21 and 1.55, respectively, which are
quite smaller than that one would expect from the earlier studies [134, 138]. In fact, in
the present case, increasing the product S e Φ by a factor of 2.5 did not lead to increase
the AR by similar factor. We argue that these deviations are mainly due to the loss
of Au from the silica matrix. Increase in Au loss with increase in fluence resulted in
observed lower values of ARs. If there was no Au loss, in the present case, we might
have also been observing a linear relation between S e Φ and AR.
In case of highest fluence, F10, irradiated sample, Au NPs are spherical in shape
instead of anisotropic. These observations can be understood as follows. Continuous
irradiation of SHI leads to reduction of Au content inside the silica matrix (Fig. 5.5),
which resulted in very less amount of Au available to form NPs. It is known that the
elongation of the NPs is observed only when the NP size become larger than the track

Anisotropic deformation of Au NPs by SHI irradiation 80
diameter in the matrix [115]. The average NP size (∼ 9 nm) observed in case of the
highest fluence is comparable with the track size (∼ 8.5 nm), created by 100 MeV
Au ions inside silica glass matrix, and hence NP deformation has not been observed.
Rather, the small amount (∼ 10 % of the implanted material) of Au form the spherical
NPs, from its melt phase, with average size of 9.2 nm.
At a SHI irradiation fluence of F10, a redshift of the SPR peak position has been
observed by about 18 nm as compared to the as-grown sample. Similar redshift of
the SPR band corresponding to Au NPs embedded in silica has also been reported,
recently, by Mishra et al. [144]. Such a redshift in its position with decrease in
FWHM is known to be due to an increase in the size of the NP [143, 135]. In fact, size
distributions of the XTEM data [Figs. 5.3(e) and 5.3(h)] clearly show slight growth
(2.6 nm) of the average size of as-grown Au NPs after SHI irradiation to a fluence of
F10. But such a small growth in size of the Au NPs alone can not account for an
SPR redshift of ∼ 18 nm. This can be easily seen from a simple simulation of the
OA spectra for the Au NPs embedded in silica glass. This simulation were carried
out using the Eqn. 2.7 that uses, NP size, dielectric constant of the host, and the
dielectric constant of the NPs. Figure 5.9 shows the simulated OA spectra of the Au
Absorbance [arb. units]
■ ■
▲ ▲
● ●
SPR peak [nm]
630
600
570
540
2 3 4 5 6
Dielectric constant
d = 9.2 nm, εm= 3.25
d = 9.2 nm, εm= 2.26
d = 6.6 nm, ε m= 2.26
500 550 600 650 700
Wavelength [nm]
Figure 5.9: Simulated OA spectra of
the Au NPs embedded in silica matrix
for different NP sizes (d) and medium
dielectric constants (ǫ m ). Inset shows
the variation of SPR peak position
with dielectric constant (filled circles).
The solid line is the linear fit to the
simulated data.
NPs embedded in silica glass for different NP sizes and medium dielectric constants,
keeping all the other parameters fixed. Details of the calculation is reported elsewhere
[149]. For ǫ m = 2.26, about 2 nm redshift of SPR peak position is seen due to a growth
of NP from 6.6 to 9.2 nm. However, for the NPs of size 9.2 nm, a change in ǫ m from
2.26 to 3.25 leads to a large redshift in SPR peak position. In case of ǫ m = 3.25, the
SPR peak appears at about 561 nm as observed in the present experiment for the

Anisotropic deformation of Au NPs by SHI irradiation 81
fluence of F10. The inset of Fig. 5.9 shows the variation of the SPR peak position
with the dielectric constant for the NPs of size 9.2 nm. A linear increase of SPR
peak position with increase in dielectric constant is seen. Therefore, we attribute the
observed redshift of SPR peak position as a result of the change in dielectric constant
of the matrix after Au irradiation to a fluence of F10. We argue that the dissolution
of Au (in the form of atoms and/or small clusters containing only few atoms) in silica
glass resulted in a change of medium dielectric constant after such high fluence SHI
irradiation. This argument is supported by the XTEM [Figs. 5.3(c) and (d)] and
RBS [Fig. 5.5(b)] data. RBS data showed nearly 10 % of Au lost from the matrix
by increasing the fluence from F5 to F10. However, it seems from XTEM data that
Au loss is more as compared to that obtained by RBS studies for the same increase
in fluence. This is because of RBS is an absolute measurement that includes all the
Au regardless of cluster size. These observations suggested that some of Au is in the
form of atoms and/or small clusters in the matrix which could not be imaged by our
TEM. It is also important to mention here that the highest fluence SHI irradiation
can result in compaction of the silica leading to an increase in density and hence an
increase in dielectric constant of the matrix [150]. This can result in a redshift of the
SPR bands of Au NPs embedded in silica glass.

Chapter 6
Observation of a universal
aggregation mechanism in Au
sputtered by swift heavy ions
6.1 Introduction
The observation of emission of clusters of few atoms and molecules from energetic ion
impact was reported way back in 1958 [17]. This observation was quite surprising
since the cluster binding energies are of the order of 1−2 eV, too small compared
to the energies of the colliding ions. Since then there has been a lot of studies on
the subject with an aim to understand the basic mechanisms behind cluster emission
[16, 18, 19, 111, 120, 123].In most cases, involving low energy ion impact [16, 18, 19],
the cluster yield, Y (n), as a function of number of constituent atoms, n, has been
found to follow an inverse power law, Y (n) ∼ n −δ , δ being a decay exponent. For
small clusters with very few atoms, detected using time-of-flight, δ has been found to
lie between 4 and 8 [19]. However, using transmission electron microscopy (TEM), it
has recently been possible to study the size distribution of large clusters, collected on
catcher foils, placed suitably during ion irradiations [16, 111, 120].
As has been mentioned earlier, on the theoretical side there is a shock-wave model
[22] where overlapping collision cascades, from low energy heavy ion impact, can result
in the production of shock waves in a medium. These can propagate to the surface. In
some cases they can result in emission of a chunk of material with essentially the same
atomic coordination as the target. This mechanism, yields a value of δ close to 2. It
has been argued to be a valid one where irradiation of Au thin films by four different
82

Observation of a universal aggregation mechanism in sputtered Au 83
ions, viz. Ne, Ar, Kr, and Au at energies between 400 and 500 keV has been found
to yield a δ value close to 2, independent of the sputtering yield [16]. There is also
a thermodynamic model [23] where the cascade of atomic displacements, produced
in the near surface region of a target, can thermalize and expand into vacuum. The
temperature is supposed to go beyond a critical temperature, T C . Upon expansion
and cooling the material can undergo a liquid-gas phase transition. A large liquid
drop thus formed can fragment producing a mass distribution with a δ value close to
7/3.
As compared to continuous media (films or bulk), irradiation of nm sized metal
islands or metal nanoparticles (NPs), embedded in a matrix, with a possibility of
melting and evaporation, form a different class of systems. In this chapter, we show
that the size distribution of clusters emitted from such a system, under swift heavy
ion (SHI) irradiation, falls under a universal class of aggregation [151]. These systems
possess non-equilibrium steady state solutions of mass distributions in the form of
inverse power laws. In all such cases there is a competition between aggregation and
breakup or evaporation, a delicate balance between the two leading to a variety of
steady state mass distributions [56] with exponent, δ = 3/2 [57] or 7/2 [63]. Using a
simple two-dimensional lattice model with jump between nearest sites and aggregation,
Takayasu et al. [57] have shown that the asymptotic distribution of mass or size always
follows a power law only with the injection of a unit mass (monomer) at each site.
Without the injection of monomers the solution, in the infinite time limit, corresponds
to an aggregate with all the particles sticking together. Later the analysis was extended
to include both positive and negative values for the dynamical variable which was taken
to be charge rather than mass [58]. The model included both injection and evaporation
(through pair creation injection of unit positive and negative charges). The kinetics of
aggregation were studied using a mean field theory. The system was found to reach a
steady state with a charge distribution following a power law with δ = 3/2 [58]. This
happens to be a very general case corresponding to a broad class of phenomena. As
shown by Bonabeau et al. [60, 62], fish schools, with breakup and injection, show a
similar aggregation, the size distribution showing a decay with δ = 3/2. This is seen
even in economics related to distribution of wealth [61]. Such a result has also been
obtained by Majumdar et al. [59] for aggregation in a mass conserving mean field type
site-site interaction model. It has also been shown that small changes in the breakup
parameters do not affect the decay exponent [60, 61]. There is however a cutoff size
which depends upon the competition between aggregation and breakup, and the time

Observation of a universal aggregation mechanism in sputtered Au 84
scales associated with them.
Here we show, Au atoms, sputtered from vaporized Au NPs, follow a steady state
aggregation as mentioned above. The observed mass distribution is found to be in the
form of a truncated power law with a decay exponent of 3/2. Beyond a critical size
there is a drop off which appears to be again in the form of a power law with a much
larger exponent. Under the present irradiation conditions, the temperature of the NPs
is known to go well above the vaporization temperature. But the mass distribution
shows power law decays quite different from what has been suggested for a liquid-gas
type phase transition [23]. Since the system indicates a steady state scenario there
is no need to correct the mass distribution against any breakup effects as applicable
for cluster emission at lower irradiation energies [21]. The results also indicate the
thermal spike production from electronic energy loss to be an essential requirement
for the present observations.
6.2 Experiments
For the present study, nearly spherical Au NPs were synthesized in silica glass, very
near to the matrix surface, by implanting 32 keV Au − ions to a fluence of 4×10 16
cm −2 followed by annealing in air at 850 ◦ for 1 h. The Au implantation was carried
out using a low energy negative ion implantation facility at the Institute of Physics
(IOP), Bhubaneswar. A bright-field cross-sectional TEM micrograph of the silica
glass sample containing Au NPs is shown in Fig. 6.1(a). One can see a buried layer of
(a)
Surface
Frequency [%]
15
10
5
(b)
0
2 4 6 8 10 12 14 16
Diameter [nm]
Figure 6.1: Bright-field XTEM micrographs of the Au NPs in silica glass after 32 keV Au −
ions implantation to a fluence of 4×10 16 cm −2 followed by annealing in air at 850 ◦ for 1 h.
(b) show the corresponding size distribution with a Gaussian fitting.

Observation of a universal aggregation mechanism in sputtered Au 85
nearly spherical NPs with particle size decreasing from about 15 to 2 nm, progressively
with depth. Figure 6.1(b) shows the corresponding size distribution. Hereafter, these
samples will be referred to as the “target”. Three of the targets were irradiated, at
room temperature, using 100 MeV Au 8+ ions at normal incidence to fluences F2, F5,
and F10 with values of 2×10 13 , 5×10 13 , and 1×10 14 ions cm −2 , respectively. This was
done using the 15 MV Pelletron accelerator at Inter University Accelerator Centre,
New Delhi. For comparison, a target was irradiated with 10 MeV Au 4+ ions to a
fluence as given by F10 using the 3 MV Pelletron accelerator at IOP, Bhubaneswar.
During the irradiation of the targets, sputtered particles were collected using catcher
foils (in the form of carbon coated Cu grids usually used for TEM studies) placed at a
distance of ∼ 1 cm in front of the target at an angle of ∼ 15 ◦ with the sample surface.
The schematic diagram of the sputtering experiment is shown in Fig. 6.2. In each case,
Sputtered
particles
Au beam
Catcher
foil
Au NPs in
silica glass
Silica glass
Figure 6.2: Schematic diagram illustrating
the sputtering experiment.
both involving 10 and 100 MeV Au irradiations, the ion beams were raster scanned over
an area of 1×1 cm 2 for uniform irradiation. All the Au implantation and irradiations
were carried out at room temperature. The the catcher foils with collected particles
were imaged using a JEOL 2010 UHR TEM operating at 200 kV. High resolution
TEM (HRTEM) and selected area electron diffraction (SAED) measurements were
also carried out to infer about the crystal structure of the NPs. The Au content in the
targets were checked before and after SHI irradiation using Rutherford backscattering
spectrometry (RBS) employing with 1.35 MeV 4 He + ions using the 3 MV Pelletron
accelerator at IOP, Bhubaneswar. A surface barrier detector was placed at an angle of
135 ◦ with respect to the incident beam direction to collect the backscattered particles.

Observation of a universal aggregation mechanism in sputtered Au 86
(a)
(b)
(c)
Surface Coverage (%)
8
7
6
5
4
3
2
(d)
2 4 6 8 10
13
Fluence [10 ions cm
-2]
Figure 6.3: Bright-field planer TEM micrographs of the ejected Au particles on the catcher
foils collected during ion irradiation with irradiation fluences of (a) 2×10 13 , (b) 5×10 13 , and
(c) 1×10 14 ions cm −2 , respectively. (d) Coverage of the catcher foils surface by the ejected
NPs. Solid line is the linear fit to the experimentally obtained data.
6.3 Results
6.3.1 Ejection of Au NPs: Dependence on ion fluence
Bright-field planer TEM micrographs of the particles collected on the catcher foils
following SHI irradiations to the fluences of 2×10 13 , 5×10 13 , and 1×10 14 ions cm −2
are shown in Figs. 6.3(a), (b), and (c), respectively. The larger particles are noticeably
darker in the micrograph indicating them to be three-dimensional entities. The
particles are seen to have sizes ranging from about 1 to 20 nm. Figures 6.3(a)−(c)
also show an increase in NP number density with increase in fluence. Using different
frames and knowing the size of the NPs it is possible to determine the fraction of the
frame area covered by the observed NPs. Figure 6.3(d) shows this surface coverage in
terms of percentage of area covered in a given frame, by the ejected NPs. The data
shown for each fluence also correspond to an average over several frames. This figure

Observation of a universal aggregation mechanism in sputtered Au 87
shows a linear increase in the coverage with increase in irradiation fluence. In fact,
the surface coverage of Au is seen to increase from about 3% to about 8% in going
from the lowest to the highest irradiation fluence. It must be mentioned here, we have
not observed any NP on the catcher foil following the 10 MeV Au irradiation where
inelastic thermal spike induced sputtering of Au is found to be much reduced.
6.3.2 Crystal structure of the ejected NPs
a
0.219 nm
b
(111)
(002)
Figure 6.4: (a) HRTEM
micrograph of a NP and (b)
SAED pattern of the NPs
collected on the catcher foil
corresponding to a Au irradiation
fluence of 1×10 14
ions cm −2 .
Figure 6.4(a) shows an HRTEM micrograph of an ejected NP on the catcher foil corresponding
to a fluence of 2×10 13 ions cm −2 . The HRTEM image clearly shows the
crystalline nature of the particles with an inter-planar spacing of 0.219±0.004 nm
which closely agrees with Au lattice spacing for cubic (111) plane. Similar HRTEM
measurements on the NPs collected on the catcher foils for other fluences also indicated
crystalline nature of the NPs. However, to further check the structure, SAED
measurements were carried out on the same catcher foil. A typical SAED micrograph
taken on the catcher foil is shown in Fig. 6.4(b). The image shows ring patterns with
dots indicating the crystalline nature of the NPs present. The dots are identified as
corresponding to the (111) and (002) planes of face-centered-cubic Au. These are
indicated in the figure.
6.3.3 Loss of Au content in the target: RBS studies
Figure 6.5 shows the RBS spectra taken from the targets before and after Au irradiations.
To compare the data for different samples, all the spectra have been normalized
using the silicon profile. The figure clearly shows almost no loss of Au in the target
following 10 MeV Au irradiation. However, a decrease of backscattered yield is seen,
in the RBS spectra corresponding to the targets after 100 MeV Au irradiations, with

Observation of a universal aggregation mechanism in sputtered Au 88
Normalized Yield [counts]
250
200
150
100
50
● ●
unirradiated
10 MeV
◆ ◆ 1x10 14
100 MeV
▲ ▲ 13
2x10
■ ■ 13
5x10
▼ ▼ 1x1014
SiO
2
+
He ions
0
725 750 775 800 825
Channel Number
increase in fluence. Decrease of backscattered yield indicates loss of Au in the target.
The Au loss is seen to increase with increase in ion fluence.
6.3.4 A power law behaviour in NP size distribution
Cluster Yield
10
10
10
10
3
2
1
0
Fluence (ions cm -2)
13
■ 2x10
●
13
5x10
▲ 1x10
14
104 5
10
Number of Atoms
45 o
Detector Figure 6.5: The Au part of the
Embedded Au RBS spectra as measured on the
targets before and after the irradiations.
The irradiation fluences
are in the unit of ions cm −2 . Inset
shows the experimental arrangement.
Figure 6.6: Size distribution in
terms of the number of clusters
of particular size (cluster yield)
plotted against the number of
atoms in the cluster. Dotted,
dot-dashed, and continuous
lines correspond to power
law decays with δ values of 3/2,
7/3, and 7/2, respectively.
The size distributions of the NPs collected on the catcher foils, for various SHI fluences,
are shown in Fig. 6.6 where the number of observed Au NPs of different sizes, Y (n),
has been plotted against n, the number of atoms in the n−atom cluster or NP. To
generate these data, many TEM micrographs of the same frame size [as shown in Figs.
6.3(a)−6.3(c)] were analyzed taking each NP to be spherical in shape. The number
of atoms in an n−atom NP is estimated multiplying the volume of the NP by the

Observation of a universal aggregation mechanism in sputtered Au 89
number density of atoms in bulk Au. A total of 1313, 2009, and 2047 particles were
considered for SHI fluences of F2, F5, and F10 respectively. Superimposed on the data
is a straight line representing a power law distribution with a δ value of 3/2 (dotted
line). This is seen to agree with the data very well up to a cutoff size of about 12.5 nm
(corresponding to about 62000 atoms) beyond which evaporation or breakup effects are
dominant. This region corresponding to larger clusters and is affected by fluctuations
because of progressively small number of particles observed. For this region we have
also shown a power law decay with a δ value of 7/2 (continuous line). The data
points seem to follow this behaviour. What is more important is that the data for
three different fluences, taken on three different samples, show the same behavior.
As shown in Fig. 5.3, SHI irradiation results in a change in the size distribution of
the embedded Au NPs. But this does not seem to affect the size distribution of the
sputtered Au NPs. In the following section we present a discussion on various aspects
of the observed phenomenon and the conditions under which such aggregation takes
place.
6.4 Discussion: NP ejection mechanisms
The first and formost requirement is the ejection of Au atoms from embedded Au NPs
under SHI irradiation. This can happen due to the formation of localized inelastic
thermal spikes [33] produced in the NPs resulting in their vaporization. As has been
mentioned earlier, in the inelastic thermal spikes model, passage of a SHI results in
excitation and ionization of electrons in a cylindrical region around the ion path. This
energy at first gets distributed into the electronic system through electron-electron
interactions in a time scale of ∼ 10 −13 s. Electron-lattice interactions cause a part of
this energy to go to lattice atoms resulting in a temperature rise. The temperature
of the thermal spike thus generated depends upon (i) the volume in which energy
imparted by the SHI diffuses due to mobility of the hot electron gas and (ii) the
strength of electron-phonon coupling that determines the efficiency of the transfer of
energy from electronic system to the lattice. Scattering of the excited electrons from
the NP surface and matrix-NP interface will reduce the mobility of the electrons and
will increase the electron-phonon coupling. Further, in small nanostructures density
of structural defects is expected to be very much higher [153] as compared to the bulk
material, which will further contribute to the reduction of the electron mobility in
NPs. Therefore, as compared to the bulk material, the temperature of the thermal

Observation of a universal aggregation mechanism in sputtered Au 90
spike is expected to be much higher in the case of NPs, thus enhancing the effects
of electronic energy loss, S e , leading to a large temperature rise which sometimes can
go beyond the evaporation temperature of the materials. There is also no dissipation
of heat into the insulating surrounding matrix. Simulations indicate this to be true
leading to vaporization of smaller particles [115], some of which get attached to other
NPs leading to Ostwald ripening. Cross-sectional TEM images taken on the targets,
after SHI irradiations at various fluences, do show this (Fig. 5.3). Such a phenomenon
does not seem to happen with 10 MeV Au ions where electronic stopping (2.5 keV
nm −1 ) is way below that at 100 MeV (13.5 keV nm −1 ) [31]. The vaporized material
must also come out of the matrix for the observed aggregation to take place. In silica
glass SHI irradiation can result in a melting of a cylindrical zone around the beam
path which, because of the pressure imbalance [115], can result in a squeezing out of
the evaporated Au atoms (Fig. 5.3).
In such a case, it is not obvious that the expanding system could be in a thermalized
state, with a temperature above a critical value, T C , before going through
a liquid-gas phase transition as proposed by Urbassek [23]. In case it goes through
such a transition, through the liquid gas co-existence region, there would be droplet
formation. The size distribution of the sputtered clusters is then expected to show a
power law decay with a δ value as given by Fisher’s critical exponent, τ, which lies
between 2 and 2.5. For a Van der Waal’s gas the exponent τ has a value of 7/3.
Compared to this, in the present case a δ value of 3/2 has been obtained for smaller
clusters which changes over to 7/2 for larger ones. In fact there is no indication of an
exponent close to 7/3 which is also shown in Fig. 6.6 (dot-dashed line) for comparison.
This rules out, any possible liquid-gas type phase transition taking place in the
vaporized Au system resulting from SHI irradiation. On the other hand, exchange of
particles between nucleation sites, within the framework of a mass-aggregation model,
that takes diffusion, aggregation on contact, and dissociation, can result in a steady
state aggregation process with a δ value of 3/2 [59]. What is more interesting is that
the exponent 3/2 appears in a variety of systems, constituting a universal class where
mass plays the role of a control parameter. Mass conservation with steady injection
of monomers is all that is required with other details of aggregation or breakup not
being important. These results are also in disagreement with simulation results on
cluster emission from Au NPs under self-ion irradiation at lower energies [154] where
electronic stopping effects are neglected.
It is also important to understand the reason behind the cutoff observed in the

Observation of a universal aggregation mechanism in sputtered Au 91
cluster size distribution at about 12.5 nm and the change over taking place at that
point. Such a cutoff can come from splitting of larger clusters [62]. But as long as
aggregation and evaporation rates are independent of mass or size, the δ value, up
to the cutoff, remains 3/2. On the other hand a change over can occur when the
aggregation and evaporation rates become mass dependent. As shown by Vigil et
al. [63], for a critical value of the ratio of the two rates, there can be a transition
between a steady state mass distribution and geletion (infinite mass aggregate). A
power law decay, with a δ value of 7/2, has been shown to occur at the transition
point. It appears, in the present case, at cluster sizes greater than 12.5 nm, somehow
both the aggregation and evaporation rates become mass dependent leading to a phase
transition from one steady state behaviour with a δ value of 3/2 to another with a δ
value of 7/2. In the absence of such a transition the mass distribution, beyond the
cutoff, would have followed the straight line for the power lay decay with a δ value of
3/2. This is expected in a model which does not include splitting [62]. Instead it has
been found to follow a new decay line. At the moment it is not clear as to how such
a transition takes place.
However, based on the present results, it is difficult to rule out the occurrence
of a liquid-gas type phase transition in sputtered material for any general ion-target
combination where the ion energy also plays a crucial role. In fact some experimental
data do exist in support of the above model [155]. What has been shown here is the
existence of a new mechanism of aggregation in sputtered particles, not shown earlier.

Chapter 7
Summary and conclusions
This thesis deals with two set of studies using energetic Au beams. The first set of
studies involve the modification of nanocrystaline ZnS films and ejection of nanoclusters
from them during ion irradiation. The second set involves the modification of
silica embedded Au nanoparticles (NPs) and ejection of nanoclusters from these NPs.
In both cases ejection of NPs have been observed and attempts have been made to understand
the underlying mechanism of emission using the size distribution of emitted
particles.
The first study involves changes in structural, optical, and surface morphological
properties of nanocrystalline ZnS films induced by 35 keV, 2 MeV, and 100 MeV
Au ions. In this, irradiation induced modifications in properties of ZnS thin films,
deposited through a simple thermal evaporation technique, have been studied using X-
ray diffraction (XRD), optical absorption (OA), and atomic force microscopy (AFM).
XRD and OA measurements on the as-grown film indicated the film to have a wurtzite
crystal structure with a grain size and an optical band gap of about 17.5 nm and 3.56
eV, respectively. MeV Au irradiation has been found to result in an increase in grain
size and a decrease in band gap. There is also a change over from compressive stress in
the strating film to tensile stress after MeV Au irradiation. The effects of irradiation
are seen to be stronger at 100 MeV than that at 2 MeV. However, no change in grain
size and band gap have been observed in following irradiation at the keV energy. The
roughness produced on the sample surface have been studied through an analysis of
the power spectral density data as determined from AFM topographies. In the large
spatial frequency, q, (at small length scales) the Fourier exponent, γ, has been found to
have values 2.80, 3.11, 3.72, and 2.85 for the as-deposited and the samples irradiated
with Au at 35 keV, 2 MeV, and 100 MeV, respectively. The values close to 3 could be
92

Summary and conclusions 93
understood in a model where evaporation and condensation dominate the higer value
corresponding to a surface diffusion dominated process.
The effect of Au irradiation on isolated and embedded nanostructures system has
also been studied. For this, spherical Au NPs (of average size 7 nm) were synthesized
in the near surface region of silica glass samples by implanting 32 keV Au − ions to
a fluence of 4×10 16 cm −2 , followed by annealing at 850 ◦ C in air for 1 h. The silica
glass samples so prepared, containing Au NPs are found to show a surface plasmon
resonance (SPR) band centered around 543 nm in the optical absorption spectrum.
The silica glass samples with embedded NPs were subjected to Au irradiations (at
room temperature) at 10 and 100 MeV. The higher energy irradiation was found to
result in an elongation of the embedded Au NPs along the beam direction. At higher
irradiation fluence there was Au loss from the silica matrix. Up to a fluence of 5×10 13
ions cm −2 , the smaller NPs (diameter < 9 nm) are found to grow in size while larger
ones deformed anisotropically along the ion beam direction. The anisotropy in the
larger NPs have been seen to increase with increase in ion fluence. Optical absorption
spectra corresponding to the elongated Au NPs show two SPR bands corresponding to
transverse and longitudinal modes. Further, longitudinal SPR band has been seen to
shift towards higher wavelength with increase in fluence. This shift of the longitudinal
SPR band is a direct consequence of the further elongation of the NPs. However, in
the sample irradiated at a fluence of 1×10 14 ions cm −2 , the Au NPs are seen to be
almost spherical in shape, with a large interparticle separation. The average size is
found to be ∼ 9.2 nm, which is little larger than that of the unirradiated samples. In
fact what is found is a two-dimensional array of small Au NPs, very near the silica
surface which result in an SPR band with a single peak at around 561 nm. This SPR
peak position appeared to be red-shifted as compared to that of the as-grown sample.
Using a simple simulation that uses, NP size, density, and dielectric constant of the
host together with optical density of the NPs, it has been shown that the shift of the
SPR band corresponding to the spherical NPs is, mainly, the result of the change in
matrix dielectric constant, due to the dissolution of Au in the form of atom and/or
small clusters, instead of the growth of the NPs after ion irradiation.
The results on change of shape (from nearly spherical to elongated ones) from ion
irradiation, can be understood within the framework of the inelastic thermal spike
induced ion track formation. Smaller NPs within the ion track completely evaporate
and collect around existing small clusters leading Oswald ripening of small NPs. In
any case they can grow to a size comparable to the track diameter, which is about

Summary and conclusions 94
8.5 nm in silica glass in the present experimental conditions. On the other hand
larger ones can undergo a melting and due to pressure imbalance can get squeezed
to elongated shapes along the ion tracks which are formed along the beam direction.
The material loss at higher fluence occurs from a squeezing out and vaporization of
NPs formed near the surface.
Following the above study, it was also necessary to see whether the material which
is ejected during ion irradiation really contained any nanoclusters. This is because
the size distribution in that case can be connected to the underlying mechanism of
emission. Therefore, in further pursuance of our investigations we have collected the
ejected particles on catcher foils placed suitably in the reflection geometry with an
aim to look at their size distribution. This has been done in both cases concerning
irradiation of ZnS films and embedded Au NPs in silica glass as mentioned earlier.
In each case the size distribution has been determined from transmission electron
microscopy.
For ZnS, with 35 keV Au irradiation no NP larger than 1 nm could be observed on
the catcher foil and therefore no size distribution study could be carried out. However,
we have observed wurtzite ZnS NPs on the catcher foils due to MeV Au irradiations.
The ejected NPs have been found to have sizes (diameter) lying between 2 to 7 nm.
The most probable size is seen to be around 3−3.5 nm. For particle sizes ≥ 3 nm, the
distributions show a power law decay in the form of Y (n) ∼ n −δ ; n and δ being the
number of atoms in a NP and the power law exponent, respectively. In case of Au
irradiation at 2 MeV, the values of exponent, δ, are found to be around 2.60 ± 0.09,
whereas the same for 100 MeV irradiation is found to be around 3.45 ± 0.09. After
correcting the results for cluster breakup effects, the values of power law exponent
reduced to about 2.0 and 2.6 for the above two irradiation energies. A δ value close
to 2 indicates that the NPs are produced when shock waves, generated by subsurface
displacement cascades, ablate the surface. On the other hand, Coulomb explosion
followed by thermal spike induced vaporization of ZnS seems to be the dominant
mechanism regarding material removal at 100 MeV. In such a case the evaporated
material can cool down going into the fragmentation region ejecting clusters with
a power law exponent close to 7/3. Interestingly enough, no NP could be found
in the catcher foil even during the 100 MeV Au irradiation when the sample was
maintained at liquid nitrogen temperature. This suppression on cluster ejection at low
temperature can be explained as due to an enhanced thermal conductivity resulting
in fast heat dissipation that suppresses thermal spike effects.

Summary and conclusions 95
However, the scenario become completely different in case of sputtering from embedded
nanostructures. For 10 MeV Au irradiation of Au NPs, embedded in silica
glass, no NP could be observed on the catcher foils. However, crystalline Au NPs of
size 1−20 nm could be seen on the catcher foils for 100 MeV Au irradiation case. Size
distribution, of the ejected NPs, has been seen to follow a similar power law decay as
seen in the sputtered ZnS NPs. Here, we have seen the existence of two power law
exponents: 3/2 for smaller NPs, below 12.5 nm in size, and 7/2 for larger NPs. The
first case can be rationalized as occurring from a steady state aggregation process,
independent of cluster size. The later case may come from a dynamical transition
to another steady state where aggregation and evaporation rates are size dependent.
To understand the aggregation of clusters during sputtering at very high energy (100
MeV), inelastic thermal spike concept has been invoked. Production of thermal spike
in silica, due to 100 MeV Au ions, can lead to a local temperature rise well above the
vaporization temperature of Au. In such a case evaporated Au clusters can exchange
particle between nucleation sites, within the framework of a mass-aggregation model,
that takes diffusion, aggregation on contact, and dissociation. This can result in a
size independent steady state aggregation process with a δ value of 3/2. This happens
to be a very general case corresponding to a broad class of phenomena, such as fish
schooling, distribution of wealth, cloud formation, and polymer gels. On the other
hand, a phase transition can occur when the aggregation and evaporation rates become
size dependent leading to a δ value of 7/2 at the critical point. Since the system
indicates a steady state scenario there is no need to correct the exponent against any
breakup effects as applicable for the ZnS NPs ejected during Au irradiations.

Bibliography
[1] T.W. Kim, D.C. Choo, J.H. Shim and S.O. Kang, Appl. Phys. Lett. , 80 2168 (2002).
[2] Y.K. Kim, A.J. Danner, J.J. Raftery (Jr.) and K.D. Choquette, IEEE J. Quantum
Elect. 11, 1292 (2005).
[3] B.N. Dev, S. Bera, B. Satpati, D.K. Goswami, K. Bhattacharjee, P.V. Satyam, K.
Yamashita, O.M. Liedke, K. Potzger, J. Fassbender, F. Eichhorn, R. Gröetzschel, Microelectron.
Eng. 83, 1721 (2006).
[4] J.A. Knapp, D.M. Follstaedt and S M. Myers, J. Appl. Phys. 103, 013518 (2008).
[5] M.S. Rafique and N. Tahir, Vacuum 81, 1062 (2007).
[6] N. Martin, J. McCord, A. Gerber, T. Strache, T. Gemming, I. Mönch, N. Farag, R.
Schöfer, J. Fassbender, E. Quandt and L. Schultz, Appl. Phys. Lett. 94, 062506 (2009).
[7] B. Salem, V. Aimez, D. Morris, A. Turala, P. Regreny and M. Gendry, Appl. Phys.
Lett. 87, 241115 (2005).
[8] S. Mohapatra, B. Joseph, D.P. Mahapatr and P. Chakraborty, Nucl. Instrum. Methods
Phys. Res. B 217, 578 (2004).
[9] B. Joseph, S. Mohapatra, B. Satpati, H.P. Lenka, P.K. Kuiri, D.P. Mahapatra, Nucl.
Instrum. Methods Phys. Res. B 227, 559 (2005).
[10] S. Amirthapandian, B.K. Panigrahi, S. Rajagopalan, A. Gupta, K.G.M. Nair, A.K.
Tyagi and A. Narayanasamy, Phys. Rev. B 69, 165411 (2004).
[11] K. Awazu, S. Ishii, K. Shima, S. Roorda and J. L. Brebner, Phys. Rev. B 62, 3689
(2000).
[12] A. Oliver, J.A. Reyes-Esqueda, J.C. Cheanf-Wong, C.E. Román-Velázquez, A. Crespo-
Sosa, I. Redríguez-Fernández, J.A. Seman and C. Noguez, Phys. Rev. B 74, 245425
(2006).
96

Bibliography 97
[13] R. Behrisch (Ed.), Sputtering by Particle Bombardment I, II & III, (Springer, Berlin,
Heidelberg, 1983).
[14] J.E. Greene, A. Rockett, and J.-E. Sundgren, Beam-Solid Interactions and Transient
Processes Symposium (Mater. Res. Soc., Pittsburgh, Pennsylvania, U.S.A., 1987), pp.
5973.
[15] A. Benninghoven, F. G. Rüdenauer and H.W. Werner, Secondary Ion Mass Spectrometry:
Basic Concepts, Instrumental Aspects, Applications, and Trends (Wiley, New
York, 1987).
[16] R. Birtcher, S.E. Donnely and K. Nordlund (E. Knystautas, Ed.), Engineering Thin
Films and Nanostructures with Ion Beams (Taylor & Francis, Florida, 2005) Ch. 1.
[17] R.E. Honig, J. Appl. Phys. 29, 549 (1958).
[18] C. Staudt and A. Wucher, Phys. Rev. B 66, 075419 (2002).
[19] C. Staudt, R. Heinrich and A. Wucher, Nucl. Instrum. Methods Phys. Res. B 164/165,
677 (2000).
[20] A. Wucher, Nucl. Instrum. Methods Phys. Res. B 83, 79 (1993).
[21] K.O.E. Henriksson, K. Nordlund and J. Keinonen, Phys. Rev. B 71, 014117 (2005).
[22] I.S. Bitensky and E.S. Parilis, Nucl. Instrum. Methods Phys. Res. B 21, 26 (1987).
[23] H.M. Urbassek, Nucl. Instrum. Methods Phys. Res. B 31, 541 (1988).
[24] J.S. Williams, Rep. Prog. Phys. 49, 491 (1986).
[25] J.F. Ziegler, J.P. Biersack and U. Littmark, The Stopping and Range of Ions in Matter
(Pergamon Press, New York, 1985).
[26] J. Linhard, M. Scharff, Phys. Rev. 124, 128 (1961); J. Linhard, M. Scharff and H.E.
Schiott, Mat. Fys. Medd. Dan. Vid. Selsk 33, 14 (1963).
[27] J.W. Mayer, L. Eriksson and J.A. Davis, Ion implantation in semiconductors (Academic
Press, New York, 1970).
[28] G. Dearnaley, J.H. Freeman, R.S. Nelson and J. Stehan, Ion Implantation, (North-
Holland, Amsterdam, 1973).
[29] P. Sigmund, Phys. Rev. 148, 383 (1969); Phys. Rev. 187, 187 (1969).

Bibliography 98
[30] R. Spohr, Ion Tracks and Microtechnology, principles and applications (Friedr. Vieweg
& Sohn Verlagsgesellschaft mbH, Braunschweig, 1990).
[31] SRIM 2006: a version of TRIM program. This package is available at www.srim.org.
The physical principles of the program and the basic fortran code are given in Ref. [25].
[32] The target atom is displaced in case the transferred energy is larger than the displacement
energy of the atom: Threshold displacement energies in typical solids are of the
order of 10 – 50 eV (see Ref. [34]).
[33] M. Toulemonde, C. Trautmann, E. Balanzat, K. Hjort, A. Weidinger, Nucl. Instrum.
Methods Phys. Res. B 216, 1 (2004).
[34] M. Nastasi and J.W. Mayer, Ion Implantation and Synthesis of Materials, (Springer,
2006).
[35] K. Nordlund, Comp. Mat. Sci. 3, 448 (1995).
[36] J.A. Brinkman J. Appl. Phys. 25, 961 (1954); Am. J. Phys. 24, 246 (1956).
[37] M. Ghaly and R. S. Averback, Phys. Rev. Lett. 72, 364 (1994).
[38] R.L. Fleischer, P.B. Price and R.M. Walker, Nuclear Tracks in Solids (University of
California Press, Berkeley, CA, 1975).
[39] R.E. Johnson and W.L. Brown, Nucl. Instrum. Methods Phys. Res. B 198, 103 (1982).
[40] A. Akkermann, J. Levinson, D. Ilberg and Y. Lifshitz ( R. Baragiola Ed.), in Ionization
of Solids by Heavy Particles (NATO Advanced Study Institutes Series 306, Plenum
Press, New York, 1992) p. 431.
[41] D. Lesueur and A. Dunlop, Radiant. Eff. Deffects Solids 126, 163 (1993).
[42] M. Toulemonde, C. Dufour, and E. Paumier, Phys. Rev. B 46 14362 (1992).
[43] J.G. Fujimoto, J.M. Liu, E.P. Ippen and N. Bloembergen, Phys. Rev. Lett. 53, 1837
(1984).
[44] H.D. Mieskes, W. Assmann, F. Grüner, H. Kucal, Z.G. Wang, M. Toulemonde, Phys.
Rev. B 67, 155414 (2003); M. Toulemonde, W. Assmann, C. Trautmann, F. Grüner,
H.D, Mieskes, H. Kucal, Z.G. Wang, Nucl. Instrum. Methods Phys. Res. B 212, 346
(2003).
[45] E.M. Bringa and R.E. Johnson, Phys. Rev. Lett. 88, 165501 (2002).

Bibliography 99
[46] G. Schiwietz, K. Czerski, M. Roth, F. Staufenbiel and P.L. Grande, Nucl. Instrum.
Methods Phys. Res. B 225, 4 (2004).
[47] W.R. Grove, Phil. Trans. Roy. Soc. London 142, 87 (1852).
[48] W. Jacob, C. Hopf and M. Schlüter Phys. Scr. T124, 32 (2006).
[49] T. Schenkel1, M.A. Briere, H. Schmidt-Böcking, K. Bethge and D.H. Schneider, Phys.
Rev. Lett. 72, 2481 (1997).
[50] T. Neidhart, F. Pichler, F. Aumayr, H.P. Winter, M. Schmid and P. Varga, Phys. Rev.
Lett. 74, 5280 (1995).
[51] M. W. Thompson, Phys. Mag. 4, 139 (1959).
[52] C.T. Reimann, Mat. Fys. Medd. K. Dan. Vidensk. Selsk. 43, 351 (1993) and references
therein.
[53] R.E. Johnson, M. Pospieszalka and W.L. Brown, Phys. Rev. B 44, 7263 (1991).
[54] M.A. Jakas and E. M. Bringa, Phys. Rev. B 62, 824 (2000).
[55] M. M. Jakas, Radiat. Eff. Defects Solids 152, 157 (2000).
[56] W. White, J. Colloid Interface Sci. 87, 204 (1982).
[57] H. Takayasu, I. Nishikawa and H. Tasaki, Phys. Rev. A 37, 3110 (1988).
[58] H. Takayasu, Phys. Rev. Lett. 63, 2563 (1989).
[59] S.N. Majumdar, S. Krishnamurthy and M. Barma, Phys. Rev. Lett. 81, 3691 (1998).
[60] E. Bonabeau and L. Dagorn, Phys. Rev. E 51, R5220 (1995).
[61] H. Yamamoto, T. Ohtsuki, A. Fujihara and S. Tanimoto, J. Phys. Conference Series
31, 59 (2006).
[62] E. Bonabeau, L. Dagorn and P. Fréon, Proc. Natl. Acad. Sci. USA 96, 4472 (1999).
[63] R.D. Vigil, R.M. Ziff and B. Lu, Phys. Rev. B 38, 942 (1988).
[64] B. Joseph, A.K. Behera, S. Mohapatra, P.K. Kuiri, D.P. Mahapatra, Solid State Physics
(India) 46, 271 (2003).
[65] K. Sekhar, P.V. Satyam, G. Kuri, D.P. Mahapatra and B.N. Dev, Nucl. Instrum. Methods
Phys. Res. B 73, 63 (1993).

Bibliography 100
[66] D. Kanjilal, S. Chopra, M.M. Narayanan, Indira S. Iyer, Vandana Jha, R. Joshi and
S.K. Datta, Nucl. Instrum. Methods Phys. Res. A 328, 97 (1994).
[67] H. Zhang, Ion sources (Science Press, Beijing, Springer-Verlag, Berlin, 1999).
[68] Details regarding SNICS source can be found in: R. Middleton, Nucl. Instr. Methods
Phys. Res. 214, 139, (1983) or SNICS ion source manual, NEC, USA.
[69] W.K. Chu, J.W. Mayer, M.A. Nicolet, Backscattering spectrometry (Academic Press,
New York, 1978).
[70] M. Mayer, AIP Conference Proceedings, 475, 541 (1999).
[71] RUMP: A simulation package for backscattering spectra for ion beam analysis with
MeV ions, J.W. Mayer, L.R. Doolittle, M.O. Thompson and R.C. Cochran of Cornell
University, developed the original FORTRAN code.
[72] J. Saarilahti and E. Rauhala, Nucl. Instrum. Methods Phys. Res. B 64, 734 (1992).
[73] K. Sekhar, G. Kuri, P.V. Satyam, B. Sundaravel D.P. Mahapatra and B.N. Dev Indian
J. Phys. 684, 1 (1994).
[74] B.D. Cullity, Elements of X-ray Diffraction (Addison-Wesley, Palp Alto, 1978).
[75] D.B. Williams and C.B. Carter, Transmission Electron Microscopy, Vol. I-IV (Plenum
Press, New York, 1996).
[76] S.N. Magonov and M.H. Whangbo, Surface analysis with STM and AFM : Experimental
and theoretical aspects of image analysis (VCH, New York, 1996).
[77] G. Mie, Ann. Phys. 25, 377 (1908).
[78] M. Faraday, Philos. Trans. R. Soc. 147, 145 (1857).
[79] C.F. Bohren and D.R. Huffman, Absorption and Scattering of Light by Small Particles
(Wiley Interscience, New-York, 1983).
[80] N.W. Ashcroft and N.D. Mermin, Solid State Physics (Harcourt College Publishers,
1976).
[81] H. Fröhlich, Physica 4, 406 (1937)
[82] P.B. Johnson and R.W. Christy, Phys. Rev. B 6, 4370 (1972).
[83] H. Hovel, S. Fritz, A. Hilger, U. Kreibig and M. Vollmer, Phys. Rev. B 48, 18178 (1993).

Bibliography 101
[84] F.A. Davis and N.F. Mott, Philos. Mag. 22, 903 (1970).
[85] N.V. Tkachenko, Optical Spectroscopy, Methods and Instrumentation (Elsevier, Amsterdam,
2006).
[86] R.N. Bhargava, D. Gallagher, X. Hong and A. Nurmikko, Phys. Rev. Lett. 72, 416
(1994).
[87] B. Gilbert, B.H. Frazer, H. Zhang, F. Huang, J.F. Banfield, D. Haskel, J.C. Lang, G.
Srajer and G. De Stasio, Phys. Rev. B 66, 245205 (2002).
[88] V.V. Nikesh, A. Dharmadhikari, H. Ono, S. Nozaki, G.R. Kumar and S. Mahamuni,
Appl. Phys. Lett. 84, 4602 (2004).
[89] D. Haranath, N. Bhalla, H. Chander, Rashmi, M. Kar and R. Kishore, J. Appl. Phys.
96, 6700 (2004).
[90] I. Sarkar, M.K. Sanyal, S. Kar, S. Biswas, S. Banerjee, S. Chaudhuri, S. Takeyama, H.
Mino and F. Komori, Phys. Rev. B 75, 224409 (2007).
[91] K. Ichino, T. Onishi, Y. Kawakami and S. Fujita, J. Cryst. Growth 138, 28 (1994).
[92] S. Yano, R. Schroeder, B. Ullrich and H. Sakai, Thin Solid Films 423, 273 (2003).
[93] Z.Z. Zhang, D.Z. Shen, J.Y. Zhang, C.X. Shan, Y.M. Lu, Y.C. Liu, B.H. Li, D.X. Zhao,
B. Yao and X.W. Fan, Thin Solid Films 513, 114 (2006).
[94] J. Cheng, D. Fan, H. Wang, B. Liu, Y. Zhang and H. Yan, Semicond. Sci. Technol. 18,
676 (2003).
[95] V. Dimitrova and J. Tate, Thin Solid Films 365, 134 (2001).
[96] P. Prathap, N. Revathi, Y.P.V. Subbaiah and K.T.R. Reddy, J. Phys.: Condens. Matter
20, 035205 (2008).
[97] E. Snoeks, A. van Blaaderen, T. van Dillen, C.M. van Kats, K. Velikov, M.L.
Brongersma and A. Polman, Nucl. Instrum. Methods Phys. Res. B 178, 62 (2001).
[98] S. Chowdhury, A.M.P. Hussain, G.A. Ahmed, F. Singh, D.K. Avasthi and A. Choudhury,
Mater. Res. Bull. 43, 3495 (2008).
[99] D. Mohanta, G.A. Ahmed, A. Choudhury, F. Singh and D.K. Avasthi, J. Nanopart.
Res. 8, 645 (2006).

Bibliography 102
[100] P.K. Kuiri, H.P. Lenka, G. Sahu, B.S. Acharya, D. Kanjilal, N.C. Mishra and D.P.
Mahapatra, (to be published) (2010).
[101] Joint Committee of Powder Diffraction Standards (JCPDS), International Center for
Diffraction Data (1601 Park Lane, Swarthmore, PA 1994) p. 19081.
[102] O. Madelung, U. Rössler and M. Schulz (Landolt-Börnstein, Ed. ), Group III Condensed
Matter Numerical Data and Functional Relationships in Science and Technology
Volume 41 (Springer-Verlag, 1999).
[103] R. Sathyamoorthy, S. Chandramohan, P. Sudhagar, D. Kanjilal, D.Kabiraj, and K.
Asokan, Sol. Energy Mater. Sol. Cells 90, 2297 (2006).
[104] D. Bhattacharyya and M.J. Carter, Thin Solid Films 288, 176 (1996).
[105] H.C. Ong, A.X.E. Zhu and G.T. Du, Appl. Phys. Lett. 80, 941 (2002).
[106] L.I. Maissel and R. Glang, Handbook of the Thin Film Technology (Mc Graw Hill, New
York, 1970).
[107] A.A. Kaplanski and R.M. Macforland, Spectroscopy of Solids Containing Rare Earth
Ions (Elsevier Publishers, Amsterdam 1987); A.A. Kaminskii, Crystalline Laser (CRC,
New York 1996).
[108] M.S. Kamboj, G. Kaur, R. Thangaraj and D.K. Avasthi, J. Phys. D: Appl. Phys. 35,
477 (2002).
[109] E.A. Eklund, E.J. Snyder and R.S. Williams, Surf. Sci. 285, 157 (1993).
[110] L.E. Rehn, R.C. Birtcher, S.E. Donnelly, P.M. Baldo and L. Funk, Phys. Rev. Lett.
87, 207601 (2001).
[111] P.K. Kuiri, J. Ghatak, B. Joseph, H.P. Lenka, G. Sahu, D.P. Mahapatra, A. Tripathi,
D. Kanjilal and N.C. Mishra, J. Appl. Phys. 101, 014313 (2007).
[112] K. Nordlund, J. Keinonen, M. Ghaly and R.S. Averback, Nature (London) 398, 49
(1999).
[113] R.C. Birtcher and S.E. Donnelly, Phys. Rev. Lett. 77, 4374 (1996).
[114] R.S. Averback and M Ghaly J. Appl. Phys. 76, 3908 (1994).
[115] C. D’Orléans, J. P. Stoquert, C. Estournès, C. Cerruti, J.J. Grob, J.L. Guille, F. Hass,
D. Muller and M. Richard-Pouet, Phys. Rev. B 67, 220101(R) (2003).

Bibliography 103
[116] H. Trinkaus and A.I. Ryazanov, Phys. Rev. Lett. 74, 5072 (1995).
[117] S.G. Mayr and R.S. Averback, Phys. Rev. B 68, 075419 (2003).
[118] R. Petri, P. Brault, O. Vatel, D. Henry, E. Andre, P. Dumas and F. Salvan, J. Appl.
Phys. 75, 7498 (1994).
[119] C. Herring, J. Appl. Phys. 21, 301 (1950).
[120] P.K. Kuiri, B. Joseph, J. Ghatak, H.P. Lenka, G. Sahu, B.S. Acharya and D.P. Mahapatra,
Nucl. Instrum. Methods Phys. Res. B 248, 25 (2006).
[121] B. Satpati, J. Ghatak, P.V. Satyam and B.N. Dev, J. Appl. Phys. 98, 064904 (2005).
[122] R. Heinrich and A. Wucher, Nucl. Instrum. Methods Phys. Res. B 140, 27 (1998).
[123] B. Satpati, J. Ghatak, B. Joseph, T. Som, D. Kabiraj, B.N. Dev and P.V. Satyam,
Nucl. Instrum. Methods Phys. Res. B 244, 278 (2005).
[124] H.H. Andersen, A. Brunelle, S. Della-Negra, J. Depauw, D. Jacquet, Y. Le Beyec, J.
Chaumont and H. Bernas, Phys. Rev. Lett. 80 5433 (1998).
[125] H.D. Mieskes, W. Assmann, F. Gruner, H. Kucal, Z.G. Wang and M. Toulemonde,
Phys. Rev. B 67, 155414 (2003).
[126] M. Toulemonde, W. Assmann, C. Trautmann and F. Gruner, Phys. Rev. Lett. 88,
057602 (2002).
[127] J. Eriksson, P. Demirev, P. Hakansson, R.M. Papaleo and B.U.R. Sundqvist, Phys.
Rev. B 54, 15025 (1996).
[128] Available from .
[129] Y. Yuba, S. Hazama and K. Gemo, Nucl. Instrum. Methods Phys. Res. B 206, 648
(2003).
[130] M. Tanemura, M. Ukita and F. Okuyama, Surf. Sci. 426, 141 (1999).
[131] G.A. Slack, Phys. Rev. B 6, 3791 (1972).
[132] W.L. Barnes, A. Dereux and T.W. Ebbesen, Nature 424, 824 (2003).
[133] A.F. Koenderink, J.V. Hernández, F. Robicheaux, L.D. Noordam and A. Polman,
Nano Lett. 7, 745 (2007).

Bibliography 104
[134] S. Roorda, T. van Dillen, A. Polman, C. Graff, A. van Blaaderen and B. J. Kooi, Adv.
Mater. 16, 235 (2004).
[135] S. Link and M.A. El-Sayed, J. Phys. Chem. B 103, 8410 (1999); J. Phys. Chem. B
109, 10531 (2005).
[136] V.A. Podolskiy, A.K. Sarychev and V.M. Shalaev, J. Nonlinear Opt. Phys. Mater. 11,
65 (2002).
[137] K.P. Yuen, M.F. Law, K.W. Yu and P. Sheng, Phys. Rev. E 56, R1322 (1997).
[138] Y.K. Mishra, F. Singh, D.K. Avasthi, J.C. Pivin, D. Malinovska and E. Pippel, Appl.
Phys. Lett. 91, 063103 (2007).
[139] P.K. Kuiri, B. Joseph, J. Ghatak, H.P. Lenka and G. Sahu, Adv. Sci. Lett. (accepted)
(2010).
[140] J.-M. Lamarre, Z. Yu, C. Harkati, S. Roorda and L. Martinu, Thin Solid Films 479,
232 (2005).
[141] T. van Dillen, A. Polman, P.R. Onck and E. van der Giessen, Phys. Rev. B 71, 24103
(2005).
[142] J.J. Penninkhof, C. Graf, T. van Dillen, A.M. Vredenberg, A. van Blaaderen and A.
Polman, Adv. Mater. 17, 1484 (2005).
[143] B. Joseph, J. Ghatak, H.P. Lenka, P.K. Kuiri, G. Sahu, N.C. Mishra and D.P. Mahapatra,
Nucl. Instrum. Methods Phys. Res. B 256, 659 (2007).
[144] Y.K. Mishra, D.K. Avasthi, P.K. Kulriya, F. Singh, D. Kabiraj, A. Tripathi, J.C.
Pivin, I.S. Bayer and A. Biswas, Appl. Phys. Lett. 90, 073110 (2007).
[145] U. Kreibig and M. Vollmer, Optical Properties of Metal Clusters, Springer Series in
Materials Science Vol. 25 (Springer, Berlin, 1995)
[146] E.D. Palik (Ed.), Handbook of Optical Constants of Solids (Academic Press Inc,
Florida, 1985)
[147] M. Toulemonde, J.M. Costantini, Ch. Dufour, A. Meftah, E. Paumier and F. Studer,
Nucl. Instrum. Methods Phys. Res. B 116, 37 (1996).
[148] W. Bolse and J. Wiesner, Nucl. Instrum. Methods Phys. Res. B 218, 479 (2004).

Bibliography 105
[149] B. Joseph, S. Mohapatra, H.P. Lenka, P.K. Kuiri and D.P. Mahapatra, Thin Solid
Films 493, 35 (2005).
[150] P.D. Townsend, P.J. Chandler and L. Zhang, Optical Effects of Ion Implantation (Cambridge
University Press, Cambridge, 1994).
[151] P.K. Kuiri, B. Joseph, H.P. Lenka, G. Sahu, J. Ghatak, D. Kanjilal and D.P. Mahapatra,
Phys. Rev. Lett. 100, 245501 (2008).
[152] M. Mayer, SIMNRA User Guide Technical Report IPP 9/113 Max Plank Institut Fur
Plasmaphysik, Garching, Germany.
[153] P.K. Kuiri, H.P. Lenka, J. Ghatak, G. Sahu, B. Joseph and D.P. Mahapatra, J. Appl.
Phys. 102, 024315 (2007).
[154] R. Kissel and H.M. Urbassek, Nucl. Instrum. Methods Phys. Res. B 180, 293 (2001).
[155] H.T. Jonkman and J. Michl, J. Am. Chem. Soci. 103, 733 (1981); R.G. Orth, H.T.
Jonkman, D.H. Powell and J. Michl, J. Am. Chem. Soci. 103, 6026 (1981).