Workshop proceeding - final.pdf - Faculty of Information and ...

The 3 rd VACPS Research Workshop
Proceedings
Victorian Association of
Chinese PhD Students and Scholars
– VACPS –
EN Building, Hawthorn Campus,
Swinburne University of Technology
24 Oct 2009

版 权 声 明
经 作 者 同 意 , 本 论 文 集 收 录 了 第 三 届 维 多 利 亚 州 中 国 博 士 生 及 其 青 年 学 者 优 秀 论 文 报
告 会 所 接 收 的 学 术 论 文 。 该 论 文 , 及 其 摘 要 、 图 、 表 和 公 式 等 , 其 版 权 属 于 作 者 所
有 。 墨 尔 本 中 国 博 士 沙 龙 等 主 办 方 对 此 版 权 不 提 出 要 求 。
未 经 作 者 本 人 许 可 , 任 何 人 不 得 局 部 或 者 全 部 变 更 、 转 载 、 复 制 、 改 动 、 散 布 、 展
示 、 发 行 该 论 文 , 否 则 主 办 方 有 权 利 依 法 追 究 法 律 责 任 。 墨 尔 本 中 国 博 士 沙 龙 对 该 版
权 声 明 具 有 最 终 解 释 权 。
特 此 声 明 。
墨 尔 本 中 国 博 士 沙 龙
Disclaimer
This proceeding is only for the archive purpose, and information included is authorized by the
authors. Rights and intellectual properties related to the manuscripts, abstracts, figures, tables,
equations and other materials are held by their respective authors. The Third VACPYS
Workshop committee claims no copyright to the content of the manuscripts and abstracts.
No part of this proceeding or the related files may be reproduced or transmitted in any form,
by any means, without the prior written permission from the respective authors.
VACPS reserves all rights of interpretation of this disclaimer.

Organizing Committee
Victorian Association of Chinese PhD Students and Scholars 2009
Mr. Erjiang (Frank) Fu, RMIT University; President of VACPS 2009
Mr. Duo (Winston) Wu, Monash University; Vice President of VACPS 2009
Mr. Peng (Peter) Hao, The University of Melbourne; Secretary General of VACPS 2009
Mr. Xiao Liu, Swinburne University of Technology; Head of Organization and Membership
Team, VACPS 2009
Mr. Haidong (Rick) Zheng, RMIT University; Head of Media Team, VACPS 2009
Dr. Gui (John) Qin, RMIT University; Head of Business Team, VACPS 2009
Ms. Ping Yu, RMIT University; Secretary and Treasurer Team, VACPS 2009
Ms. Xiaohuan (Iris) Yan, Monash University; Secretary and Treasurer Team, VACPS 2009
Mr. Wei Tao, Swinburne University of Technology; Organization and Membership Team,
VACPS 2009
Mr. Yu Zong, Victoria University; Organization and Membership Team, VACPS 2009
Mr. Jian Zhong, RMIT University; Media Team, VACPS 2009
Ms. Yini Wang, Deakin University; Media Team, VACPS 2009
Ms. Nan Fu, Monash University; Academy Team, VACPS 2009
Ms. Ying Lu, Monash University; Business Team, VACPS 2009
i

Editorial Board
Editor-in-Chief:
Nan Fu
Academy Team, VACPS 2009
Symposium Chair, Bioscience and Biotechnology, The 3 rd VACPS Workshop 2009
Editorial Board:
Erjiang Fu
President, VACPS 2009
Duo Wu
Vice President, VACPS 2009
Symposium Co-Chair, Bioscience and Biotechnology, The 3 rd VACPS Workshop 2009
Peng Hao
Secretary General, VACPS 2009
Liang Chen
Symposium Chair, Electrical and Electronic Engineering, The 3 rd VACPS Workshop 2009
Xiaohuan Yan
Symposium Co-Chair, Electrical and Electronic Engineering, The 3 rd VACPS Workshop 2009
Xiaohui Zhao
Symposium Chair, Information and Communication Technologies, The 3 rd VACPS Workshop
2009
Xiao Liu
Symposium Co-Chair, Information and Communication Technologies, The 3 rd VACPS
Workshop 2009
Haidong Zheng
Symposium Chair, Material Science, The 3 rd VACPS Workshop 2009
Wei Tao
Symposium Co-Chair, Material Science, The 3 rd VACPS Workshop 2009
Ying Lu
Symposium Chair, Business and Management, The 3 rd VACPS Workshop 2009
Gui Qin
Symposium Co-Chair, Business and Management, The 3 rd VACPS Workshop 2009
ii

Preface
On the occasion of the publication of the Proceeding of the 3 rd Victorian Chinese PhD
students and Young Scholars Research Workshop, I would like to extend my congratulations
on the successful organization of the workshop. I would also express my congratulations on
the research achievements presented on the workshop. Personally, I was very impressed by
the quality and scope of the presentations on the workshop.
As we are living the Information Age, communications and networking are part of
important aspects of the professional career development for each of PhD students and young
scholars no matter where you are. In that sense, Victorian Association of Chinese PhD
Students and Young Scholars (VACPS) has played an important role in Australia, particular
in Victoria. I sincerely wish that VACPSYS is more successful in creating a bridge not only
among Chinese students but also between Australia and China in the exchange of science,
technology and culture in future.
Professor Min Gu
Fellow, the Australian Academy of Science
Fellow, the Australian Academy of Technological Sciences and Engineering
Pro Vice-Chancellor, Swinburne University of Technology (International Research
Collaboration)
It is our great honour to support the 3rd Victorian Chinese PhD students and Young
Scholars Research Workshop. It was a great success for this annual event held at Swinburne
University of Technology this year. Organising such a workshop requires a significant
dedication, I would congratulate all who were involved.
The series of the Victorian Chinese PhD student s and Young Scholars Research
Workshops provide an excellent platform for presenting and sharing ongoing research work
among students across a variety of disciplines. It also plays an important role for networking.
I wish and have no doubt that this event will continue to be fruitful.
Professor Yun Yang
Associate Dean (Research), Faculty of Information & Communication Technologies of
Swinburne University of Technology
iii

Organizing Committee
Editorial Board
Preface
Table of Contents
Table of Contents
Scientific Advisory Board………………………………………………………….…. 1
Acknowledgements……………………………………………………………………. 2
List of Awardees………………………………………………………………………... 3
Oral Presentation Sessions
Invited Presentations…………………………………………………………………... 4
Career planning and Job Interview in Australian ICT Companies………………. 5
Novel medical therapies to treat ectopic pregnancies……………………………. 6
The role of mitochondria in atrial fibrillation with Rheumatic heart valvular
disease………………………………………………………………………………… 7
The characteristics of the InGaAs/GaAs quantum wells and its application in
VECSEL lasers……………………………………………………………………….. 8
Electrical and Electronic Engineering………………………………………………. 9
Track-Before-Detect Procedures for Low PRF Surveillance Rader…………….. 10
Fast realization of Automatic optical and infrared image registration…………... 27
Impulsive Interference Detection Method Based on Morlet Wavelet and
Maximum Likelihood Estimation……………………………………………………. 33
Demonstration and Performance Analysis of an Uplink based on Digitized RFover-Fiber
Signal Transport…………………………………………………………. 34
Fiber Nonlinearity Compensation for CO-OFDM Systems with Periodic
Dispersion Maps……………………………………………………………………… 35
Bioscience and Biotechnology………………………………………………………. 36
Effect of mini-tyrosl-tRNA synthetase / mini-tryptophanyl-tRNA synthetase on
ischemic angiogenesis in rats with acute myocardial infarction………………… 37
Distribution of three forms alpha-melanocyte stimulating hormone in the brain
of male and female sheep by using HPLC and Mass Spectrometry Analysis… 49
Investigation of the mechanism of the cardioprotective effect of flavonols…….. 61
Efficient Ethanol production in glucose/xylose co-fermentations by a novel coculture
scheme of Zymomonas mobilis and Pichia stipitis………………………. 62
Dual mode roll-up effect in a multicomponent near adiabatic adsorption
process………………………………………………………………………………... 63
Effect of glucose and sodium chloride on the stability of aged glucose oxidase
from Aspergillus niger………………………………………………………………... 64
i
ii
iii
iv
iv

Paper as a low-cost base material for diagnostic and environmental sensing
applications…………………………………………………………………………… 65
Information and Communication Technologies…………………………………... 66
Data Management in Cloud Scientific Workflow Systems……………………….. 67
HC_AB: A New Heuristic Clustering Algorithm based on Approximate
Backbone……………………………………………………………………………… 73
Web Page Prediction Based on Conditional Random Fields……………………. 74
Provenancing Qualifications Using MEASUR within Higher Education
Institutions: An Australian Case…………………………………………………….. 75
A Multi-modal Gesture Recognition System in a Human-Robot Interaction
Scenario……………………………………………………………………………….. 76
Distributed Agent based Interoperable Virtual EMR System for Healthcare
System Integration…………………………………………………………………… 77
Searching for Fair Joint Gains in Agent-based Negotiation……………………... 78
Material Science………………………………………………………………………… 79
Mesoporous Silica-Templated Assembly of Luminescent Polyester Particles… 80
Grain refinement of pure magnesium by back pressure equal channel angular
pressing at room temperature………………………………………………………. 91
Dielectrophoretic Separation of Carbon Nanotubes and Polystyrene
Microparticles…………………………………………………………………………. 96
Use of a Radially Polarized Beam for Ultra-low Energy Threshold for Cancer
Photothermal Therapy with Gold Nanorods……………………………………….. 112
Electrospinning Thermoplastic Polyurethane-contained Collagen Nanofibers
for Tissue Engineering Applications………………………………………………... 116
Uniform Coating of WO x on TiO 2 Nanotubes for Enhanced Electrochromic
Performance………………………………………………………………………….. 129
Nanoparticle Trapping by Surface Plasmon Resonance Assisted Thermal
Forces…………………………………………………………………………………. 130
Business and Management…………………………………………………………… 131
Union as a Social Regulator of Market Risk? --Empirical Evidence from
Professional Union Leader Program in B City…………………………………….. 132
Developing Dynamic Capability through Partnership: The Role of Capabilities. 141
Early Childhood Education Matters: Functions, Situations, and Perceptions…. 148
Five Dimensions of Entrepreneurship: A Study of First and Second
Generations Chinese Entrepreneurs in Melbourne………………………………. 153
An Exploration of Country of Origin Effect on Union Policies in Chinese
Multinational Enterprises…………………………………………………………….. 154
Poster Session………………………………………………………………………….. 155
v

Scientific Advisory Board
(listed alphabetically by last name)
Dr. Peggy Chan, Department of Chemical Engineering, Monash University
Dr. Jinjun Chen, Centre for Complex Software Systems and Services, Swinburne University
of Technology
Prof. Xiao Dong Chen, Department of Chemical Engineering, Monash Univeristy
Dr. Zhaolin Chen, Department of Electrical and Computer Systems Engineering, Monash
University
Dr. Shanfei Feng, Department of Marketing, Monash University
Dr. Wenyu Guo, Centre for Power Transformer Monitoring, Diagnostics and Life
Management, Monash University
Prof. Yuemei Guo, Faculty of Business and Economics, Monash University
Dr. Baohua Jia, Centre for Micro-Photonics, Swinburne University of Technology
Dr. Chaofeng Li, Department of Electrical Engineering, The University of Melbourne
Dr. Xiangping Li, Centre for Micro-Photonics, Swinburne University of Technology
Mr. Ke Liu, Department of Electrical and Computer Systems Engineering, Monash
University
Dr. Hongming Ma, Faculty of Education, Monash University
Prof. Paul Peiris, School of Natural Sciences, University of Western Sydney
Dr. Ramanie Samaratunge, Department of Management, Monash University
Dr. Dongxia Xu, Department of Electrical Engineering, The University of Melbourne
Dr. Guandong Xu, Centre for Applied Informatics, Victoria University
Mr. Qi Yang, Department of Electrical Engineering, The University of Melbourne
Prof. Yun Yang, Centre for Complex Software Systems and Services, Swinburne University
of Technology
Dr. Abu Zadek, School of Electrical and Computer Engineering, RMIT University
Ms. Yuan Zhou, Department of Electrical Engineering, The University of Melbourne
1

Acknowledgements
We gratefully acknowledge the support from the following organizations:
中 国 驻 墨 尔 本 总 领 事 馆 教 育 组
Education Office, The Consulate General of The P. R. China in Melbourne
斯 文 本 科 技 大 学 中 国 学 生 学 者 联 合 会
Chinese Students and Scholars Association, Swinburne University of Technology
斯 文 本 科 技 大 学 信 息 与 通 信 学 院 微 光 电 子 研 究 中 心
Faculty of ICT & Centre of Micro-Photonics, Swinburne University of Technology
We gratefully appreciate the following sponsors:
2

List of Awardees
Best Presentations
Bioscience and Biotechnology
Dr. Rui Zeng, “Effect of Mini-tyrosl-tRNA Synthetase / Mini-tryptophanyltRNA
Synthetase on Ischemic Angiogenesis in Rats with Acute Myocardial Infarction”, Sichuan
University & Monash University
Business and Management
Henri Lee, “Five Dimensions of Entrepreneurship: A Study of First and Second
Generations Chinese Entrepreneurs in Melbourne”, Swinburne University of Technology
Electrical and Electronic Engineering
Xiaobo Deng, “Track-Before-Detect Procedures for Low PRF Surveillance Rader”,
University of Electronic Science and Technology of China & The University of
Melbourne
Information and Communication Technologies
Minyi Li, “Searching for Fair Joint Gains in Agent-based Negotiation”, Swinburne
University of Technology
Material Science
Chen Zheng, “Dielectrophoretic Separation of Carbon Nanotubes and Polystyrene
Microparticles”, RMIT University
Best Posters
Wei Tao, Bao Hongchun, Dru Morrish and Min Gu, “Supercontinuum generation for fiberoptic
nonlinear microscopy”, Swinburne University of Technology
Gang Li, Ranjeet Singh, Dan Li, Chunxia Zhao, Liying Liu, and Paul A. Webley, “Synthesis
of biomorphic zeolite honeycomb monoliths with 16000 cells per square inch”, Monash
University
3

Career planning and Job Interview in Australian ICT Companies
* Invited speaker
1. Telstra
Victor Li 1*
In this talk, Victor Li will share his experiences on career planning and some practical
tips on job interview and writing resumes.
BIO: Victor Li got his bachelor and master degree in electronic engineering in Beijing
Institute of Technology, and PhD in Computer and Software Engineering in the University of
Melbourne. He is currently the Group Manager for Enterprise Solutions in Telstra Network
and Technology. He is the Chief Architect in Telstra network operations, a well known expert
in streaming media and SOA. He has been working in ICT area for over 15 years. He used to
work in Institute of Computing Technology, Chinese Academy of Science, Motorola
Software Lab Centre and Falcom in Germany. The annual budgets for his projects are over
$100M in the past three years. The company founded by him in China was successfully
floated on New York Stock Exchange in 2007.
5

Novel medical therapies to treat ectopic pregnancies
UW Nilsson 1 a , YE Gao 1,4 a * , TG Johns 2 , E Dimitriadis 3 , E Menkhorst 3 , Joanne Mockler 1 ,
BRG Williams 2 , S Tong 1
* Invited speaker
a These authors contributed equally to this work.
1. Centre for Women’s Health Research, Monash Institute of Medical Research, Victoria
2. Centre for Cancer Research, Monash Institute of Medical Research, Victoria
3. Prince Henry’s Institute, Victoria, Australia
4. Department of Obstetrics and Gynecology, the Second Affiliated Hospital, School of
Medicine, Xi'an Jiaotong University, Xi’an, China
Introduction: Ectopic pregnancies are those implanting outside the uterus and they can
be life-threatening emergencies. Most are treated surgically. Methotrexate (MTX) is
occasionally used, buts has limited efficacy. Hence, it is only used for small ectopic
pregnancies. Epidermal growth factor receptor (EGFR) inhibitors are used clinically to treat
cancers. They come in tablet form and have few side-effects. Given that the placenta is
dependent on EGFR signalling, EGFR inhibition may be a novel therapeutic approach to cure
un-ruptured ectopic pregnancies of any size.
We set out to develop a novel medication based treatment to cure ectopic pregnancy by
combining MTX and EGFR tyrosine kinase inhibitor.
Methods/Results: In vitro testing: Firstly, we confirmed the expression of EGFR in
JEG-3 cells and human 1st trimester placental tissue by immunostaining. Next, we added
various treatment to placenta-derived cells, BeWo (± syncytialisation) cells, JEG-3 cells and
primary 1st trimester extravillous cytotrophoblast cells, and assayed viability 72 hours later
using the CellTiter-Blue® Cell Viability Assay (Promega). We also dynamic monitored MTX
interaction with JEG-3 cells using xCELLigence real-time cell analyzing system. EGFR
inhibitor alone did not cause trophoblast cell death compared to control. MTX alone
decreased cell viability in a dose-responsive manner. Interestingly, combination of MTX and
EGFR inhibitor are supra-additive in decreasing placental tissue viability in vitro.
In vivo testing: JEG-3 xenografted in SCID mice (n=3-5 per group) was used as an in
vivo model to test effects of MTX and EGFR inhibitor. Compared with vehicle-treated mice,
MTX or EGFR inhibitor can inhibit JEG-3 xenograft growth, and are supra-additive when
combined.
Conclusion: Combination methotrexate (MTX) and EGFR inhibition can potently
regress placental tissue. It may be a novel therapeutic approach to medically cure ectopic
pregnancy, potentially replacing surgery with tablets.
BIO: Yan’e Gao is an Associate Professor in the Department of Obstetrics & Gynecology, 2 nd
Affiliated hospital, Xi’an Jiaotong University School of Medicine. She is a specialist in the
maternity clinical practice and has worked as an attending physician for more than 15 years.
She got her PhD in Anytomy, Histology & Embryology in Xi’an Jiantong University School
of Medicine in 2003, and she is now an active researcher in the area of ectopic pregnancies
and uterine cell in vitro culture.
6

The role of mitochondria in atrial fibrillation with Rheumatic heart
valvular disease
Runwei Ma 1,2 *
* Invited speaker
1. Department of Cardiovascular Surgery, The First Hospital Affiliated Kunming Medical
College, Yunnan 650032, P.R.China
2. Department of Cardiothoracic Surgery, Monash Hospital, Clayton, Vic 3168, Australia
Rheumatic heart disease (RHD) is a common disease in developing country. Most of the
disease is concomitant with Atrial Fibrillation (AF). The mechanism of AF in RHD is
different with other AF. We know that the function of mitochondria is associated with
cardiomyocyte necrosis and apoptosis which will remodel the structure of atrium, especial
fibrosis. Tissue fibrosis results from an accumulation of fibrillar collagen deposits, occurring
most commonly as a reparative process to replace degenerating myocardial parenchyma with
concomitant reactive fibrosis, which causes interstitial expansion. Fibrosis is a hallmark of
arrhythmogenic structural remodeling. In addition, there is great difference in heart function
of patients with same detriment of valves, as well as AF. We postulate that there is a
difference in mitochondria’s function which is influenced by gene mutation or difference of
mitochondria which caused by inheritance or disease. We plan to use DNA chip technique to
study the difference of mitochondrial gene in RHD patients with and without AF.
BIO: Runwei Ma is the head of the Department of Cardiovascular Surgery in the First
Affiliated Hospital of Kunming Medical College, and is responsible for the Cardiovascular
Surgeon. He has successfully applied a number of novel techniques in surgical operations. Dr
Runwei got his MD degree in Kunming Medical College in 2007 and now he is a visiting
scholar in the Medical Center Monash University. His research focuses on the improvement
of cardiovascular conditions using both traditional drug treatment and genetic therapy.
7

The characteristics of the InGaAs/GaAs quantum wells and its application
in VECSEL lasers
* Invited speaker
1. Beijing University of Technology, China
Yanrong Song 1*
We introduce a novel kind of laser-vertical external cavity surface emitting laser which
combinds the advantages of solid state lasers and semiconductor lasers. The design of the gain chip,
microavity and the whole setup are given and a comprehensive modeling of vertical-external-cavity
surface-emitting laser, reflective spectrum, edge-emitting photoluminescence spectrum, surfaceemitting
photoluminescence spectrum, material gain spectrum and the output characteristics are
calculated. The experiments about CW laser, double frequency output, mode locking laser are
demonstrated.
BIO: Prof. Yanrong Song works in College of Applied Sciences, Beijing University of
Technology, China, and now works as a visiting scholar in Centre for Micro-Photonics,
Swinburne University of Technology. She got her MS. Degree and Ph.D degree in Optics in
Shanxi University. She received the Bachelor degree in Semiconductor Physics and devices
from Tianjin University.
Her insterested research areas include in ultrafast processing and mode-locked lasers,
Biology sensor with nano-material, vertical external cavity of surface emitting semiconductor
lasers, ultra-fast Ti: Sipphire lasers and amplifiers.
8

Track-Before-Detect Procedures for Low PRF Surveillance Rader
Xiaobo Deng 1 * , Yiming Pi 1 , Mark Morelande 2 , Bill Moran 2
* Presenter
1. University of Electronic Science and Technology of China, Chengdu, 610054, China
2. The University of Melbourne, Parkville, VIC 3010, Australia
In this paper we present a dynamic programming (DP) based tracking-beforedetect (TBD)
procedures with reference to a low pulse repetition frequency (PRF) surveillance radar
framework. In order to avoid nonlinear transformation and meanwhile preserve the
ambiguous Doppler information which is eliminated in most of the literature, we model the
target dynamics in the range-Doppler domain. The target state evolutions in the ambiguous
range-Doppler domain are considered as a hybrid system with the ambiguous number deemed
as mode variable. Ambiguity number (or mode) transitions are governed by guard conditions
that are state-dependent. We present a DP based method for joint maximum a posteriori
(MAP) estimation of target’s trajectory in the ambiguous range-Doppler maps and the
corresponding ambiguity sequence, both assumptions of known and unknown nuisance
parameters are considered. The detection and tracking performance of the proposed procedure
is investigated under several system settings. The effect of the prior uncertainty of the
nuisance parameters (target power and noise variance) on the performance is also studied.
1. Introduction
Early detection and trajectory estimation of moving targets from remote surveillance radars is a
very challenging problem. The detection and tracking strategy should be power efficient to deal with
the targets in low signal-to-noise ratio (SNR), while the complexity should not hamper the process of
early decisions. The stringent detection specification is well met by track-before-detect (TBD)
approaches. TBD procedures allow simultaneous detection and tracking using unthresholded data and,
shows superior detection performance over the conventional methods.
Early studies on TBD methods mainly focused on its application in optical and infrared sensors as
in [1-5]. Extensions of TBD methods to pulse-Doppler radar system have been considered in [6-8].
Unlike in the infrared and optical scenarios, TBD approaches operate on a sequence of twodimensional
grey-scale images, in radar scenario, TBD approaches might apply to a sequence of threedimensional
data sets, defined by azimuth, range and Doppler (radial velocity), making the complexity
requirement much more stringent. Another difference is that, in radar applications, the inherent
properties of the azimuth- range-Doppler data space can be fully exploited to improve detection and
tracking performance. For example, in [6], a method for using TBD procedure in pulse-Doppler radars
is described, wherein the author exploits the Doppler information to define which range cell a target
would have been in the previous frame, and thus greatly simplify the search process involved in the
Viterbi algorithm. However, there is a drawback with this method if the observed Doppler is
ambiguous, which is the case when the radar under our consideration works with a low pulse
repetition frequency (PRF). In [7], the authors maximize the data set over the observed apparent
Doppler shifts, and then employ a Viterbi-like tracking algorithm in the range-azimuth domain. In [8],
the authors have also suggested taking advantage from the structure of the range-Doppler domain to
efficiently perform Viterbi-like tracking of multi-target. However, since the observed Doppler is
ambiguous, it’s also eliminated as in [7].
In order to exploit the Doppler information more effectively, we treat it in a more subtle way. The
observed Doppler is a modulo-PRF version of the true Doppler, and thus we represent the Doppler
information using observed apparent Doppler and the corresponding ambiguity number. The
ambiguity number is deemed as a mode variable. In order to avoid the nonlinear transformation
between the target’s state and measurement (defined in the ambiguous range-Doppler domain), the
target states are defined in the ambiguous range-Doppler domain. The evolutions of target state in
successive range-Doppler maps are analyzed and modeled as a hybrid (or multiple-model) system [9,
10

10]: The transitions of target locations (defined by range and apparent Doppler) are governed by the
ambiguity number, i.e., mode variable, while the ambiguity number transitions are state-dependent.
Exploiting the admissible transitions of target locations and ambiguity number, we have proposed a
dynamic programming (DP) based method for joint maximum a posteriori (MAP) estimation of
target’s trajectory in the ambiguous range-Doppler maps and the corresponding ambiguity number
sequence. The proposed method is investigated under both known and unknown nuisance parameters
(target power and noise variance). The detection and tracking performances of the proposed algorithm
are studied with respect to different system settings. Moreover, the effect of the prior uncertainty of
the nuisance parameters on the detection and tracking performance is also investigated.
2. Signal processing model for low PRF surveillance radars
The physical scenario considered here is shown in Fig.1. A low PRF surveillance radar is used for
monitoring a given area with an electronically scanned antenna. The surveillance area is divided into
small angular regions, visited by the antenna at a constant revisit intervalT R
. For many long range
targets, passing from one elevation beam to another during several scans happens infrequently due to
the geometric constraints; however, crossing azimuth beams happens much more frequently.
Therefore, each angular region is composed of a single elevation and N a azimuthal beam-pointing
positions. The azimuthal beamwidth is Δ
a
. The pulse repetition time is T p
and the number of train
pulses coherently processed is Nd
. At each visit, say, the kth
, the radar provides a multi-dimensional
data map (with four dimensions of range, azimuth, elevation and Doppler), each pixel is the reflected
power out of a signal processing unit, e.g., discrete Fourier transform(DFT), of a surveillance radar.
Fig.1. Radar surveillance zone illustration
The conventional signal processing method for surveillance radar first thresholds the data maps to
extract target plots (range, azimuth, elevation and Doppler locations), and then performs tracking
based on these plots. For targets of high SNR, this method works well, however, for targets of relative
low SNR, such as stealthy military aircraft and cruise missiles, thresholding throws away potentially
useful information, causing missing detections. Besides, since we are concerned with low PRF radar,
the Doppler plots are ambiguous; it’s hard to resolve the Doppler ambiguity under low SNR in the
framework of conventional methods. The TBD algorithms use several successive unthresholded data
maps to jointly determine the presence of a target and its trajectory (if present). TBD methods exploit
all the useful information among several scans and are especially preferred when encountering low
SNR targets. Besides, Doppler ambiguity is also implicitly resolved in the framework of TBD methods.
To simplify the exposition, we assume that the azimuth and elevation information are obtained
through antenna processing. The only relevant coordinates are range and Doppler. In an actual
implementation, one can choose to process azimuth-elevation and range-Doppler maps separately.
11

3. Dynamic model for approaching targets
As concerns the choice of target dynamic model, it’s tied to the adapted tracking coordinate system.
The most natural coordinate is the Cartesian coordinate. In a Cartesian system, the target state is
described by ( x, xyy , , ) T x y
x& y&
is the velocity vector.
Φ = & & , where ( , ) k
k
k
is the target’s position; ( , ) k
2 2
Velocity vector can be described by speed and heading pair (, v α , where v = x& + y& ,
and arctan ( x / y)
α = & & .
However, in the situation under our consideration, the measurements are the reflected power on the
ambiguous range-Doppler (i.e., range-apparent Doppler) grid. Thus, the choice of the Cartesian
coordinate requires a nonlinear transformation which results in increased complexity. It’s of interest to
directly model the target dynamics in the ambiguous range-Doppler domain. As illustrated in Fig. 2,
an approaching target travels at a velocity v and headingα . Heading is defined as the angle from the
velocity direction to the north direction. β denotes the difference between the inverse bearing and
target heading,
β = α − ( θ + π)
(1)
0
) k
Fig.2. Cartesian coordinate with radar located at the original and approaching target
The range trajectory for the approaching target is
2 2 2
rt () = r + vt −2rvtcosβ
0 0
2
dr() t 1 dr () t 2
r0 |
t=
0
t |
2 t= 0
t (2)
≈ + ⋅ + ⋅
dt 2 dt
2
1( vsin β ) 2
= r0
− ( vcos β ) t+
t
2 r0
2
Let λ = c/ Ft
be the radar wavelength; fdc
= 2vcos β / λ , the Doppler centroid; fdr
=− 2( vsin β ) /( λr0
) ,
the Doppler rate. (2) can be rewritten as
λ λ 2
rt () = r0
− fdct− fdrt
(3)
2 4
The Doppler trajectory can be written as
2 dr( t)
fd()
t =− = f dc
+ f dr
t (4)
λ dt
The discrete-time target dynamic model can be obtained by sampling the above continuous-time
model. Sampling the above continuous-time models at the interval ofT R
, we get
fdk , + 1
= fd()| t
t= ( k+
1) T
= f ( 1)
R dc+ fdr k+
TR
(5)
= f + f T
dk , dr R
12

λ
λ
2
rk+ 1
= r()| t
t= ( k+
1) T
= r0
− f ( 1) ( )
R
dc
k + TR − fdc
kTR
+ TR
2 4
λ λ 2
= rk − fd,
kTR − fdrTR
2 4
(6)
Expressing the true Doppler fdk
,
in terms of observed apparent Doppler f
da,
k
( 0≤ f da , k
< PRF
), and
ambiguity number m k
,
fdk ,
= fdak ,
+ mkPRF
(7)
The apparent Doppler fda,
k
is a modulo-PRF version of the true Doppler,
fda, k
= mod ( fd , k
, PRF)
(8)
where mod ( A,
B)
denotes the modulo- B version of A . Defining the target state vector
as s ( ,
,
) T
k
= x
k
fdr k
, where x ( ,
,
) T
k
= rk fda k
is the target location in ambiguous range-Doppler maps, the
evolution of target state can be written as a hybrid system
s = k 1
f ( s k, mk) + h( s + k, mk)
w (9)
k
where
⎛ λ
λ 2 ⎞
⎜rk − ( fda, k
+ mkPRF)
TR − fdr,
kTR
2 4
⎟
⎜
⎟
f( sk, mk) = ⎜ mod ( fda, k
+ fdr,
kTR,PRF) ⎟
(10)
⎜
⎟
⎜
fdr,
k
⎟
⎝
⎠
⎛λT
2 /4⎞
⎜ ⎟
h( sk, mk)
= ⎜⎜ TR
⎟
(11)
1 ⎟
⎝ ⎠
wk
is the Doppler rate increment noise, it’s assumed to be zero mean white Gaussian noise with
variance Q = var( w k
) . The ambiguity number, mk
∈{1, 2,..., M } is considered as the mode variable. In
the following, we use the ambiguity number and mode interchangeably.
3.2 Transition probabilities of ambiguity number
The target location transitions among successive ambiguous rang-Doppler maps depends on the
ambiguity number, while the ambiguity number transitions occur when the apparent Doppler falls
within a subset, known as guard condition, of the state space. For example, as illustrated in Fig. 3, a
target decelerates at a constant rate. When the observed Doppler has reached a certain region, say,
below 5 at k = 3 , the ambiguity number starts to change from m
3
= 7 to m
4
= 6 . This implies that the
evolution of the ambiguity number depends on the state, specifically, the apparent Doppler. It makes
sense to define the state-dependent mode transition probabilities [9, 10]
( x ) = pm ( = j| m = i, x ) (12)
π
ij k k + 1
k k
In the following, we consider the mathematic expression of the state dependent ambiguity
number transition probability π
ij
( x
k
). To start with, we assume only transitions to adjacent ambiguity
numbers are feasible. This assumption limits the magnitude of the Doppler rate and its increment noise
variance, achieving more realistic target behavior. Specifically, for a decelerating target shown in
Fig.3, only transition from mode i to mode j = i − 1 is feasible, and the transition occurs only when the
observed Doppler has fallen below a certain value. We express the guard conditions as linear
inequalities. Thus, a mode transition from mode i to mode j = i − 1 occurs when the following guard
condition is met.
Cx
ij k
≤ δij
(13)
C
ij
= (0,1) , andδ ij
is the guard threshold.
We can assume δ ij
as deterministic, then it’s much easier to determine π
ij
( x
k
)
13

⎧π0
if j = i and Cx
ij k
> δij
⎪
π
ij
( xk ) = ⎨1 − π0
if j = i−1 and Cijx k
≤δij
(14)
⎪
⎩0 else
whereπ 0
is the probability of staying in the same mode, the guard thresholdδ ij
is determined by the
level of Doppler rate and its incremental noise rate increment.
Due to the uncertainties in target dynamics, δ
ij
is well modeled as a random variable. For
simplicity, we assume δ ij
is a Gaussian distributed variable with mean δ ij
and variance Σ ij
, denoted
byδ N ( δ ; δ , Σ ), then the conditional mode transition probability from mode i to mode j = i − 1 is
ij ij ij ij
given by
π ( x ) = p δ ≥ C x | m = i,
x
( )
ij k ij ij k k k
(15)
Sinceδ ij
is Gaussian distributed, then (15) is given by
( δ )
π ( x ) = 1 −Φ C x − ; Σ (16)
ij k ij k ij ij
where Φ( δ ij
; Σ ij ) denotes a Gaussian cumulative probability function (CDF) with mean δ ij
and
variance
Σ ij
.The probability that target stays the same ambiguity number is given by
π ( x ) = 1 −π
( x )
(17)
ii k ij k
Fig.3. Illustration of state-dependent ambiguity number transition
4. Measurement Statistical Model
Let x()
t be the signal transmitted by the radar. For simplicity, the transmitted signal is a
rectangular pulse train x() t = a()exp( t j2 π Ft t ), where F t
is the carrier frequency and at () is the pulse
envelope of the form
⎧⎪ 1:0 ≤mod( t,
T p ) ≤τ
(18)
at () = ⎨
⎪⎩ 0:others
whereτ is the pulse width.
The received baseband signal during the kth
visit, k = 1,..., K the can be written as
H : y () t = g () t
0
k
k
Nd
⎛
2r
k ⎞
H1
: yk() t = ∑Aa k ⎜t−nTp−( k−1) TR− exp ( j2π
f t
, ) + g ()
dk k
t
n=
0
c
⎟
⎝
⎠
(19)
14

wherein A k
is the received signal amplitude, and it’s assumed to be unchanged during each pulse train.
r and f d,
k
are the unknown target range, and Doppler at kth visit, respectively. c is the speed of light,
k
gk
() t is the complex white Gaussian noise with variance σ 2 g
. After range sampling, the slow-time
discrete observation on a given range cell, say the ith
, i = 1,..., Nr
, can be written as
H
0
:
y ( n) = g ( nT )
ki , ki , p
( j2π
f nT
d,
k p)
H : y ( n) = A exp + g ( nT )
1 ki , k ki , p
(20)
for n= 1,..., N
d
. The dimension of range cell is given by Δ
r
= cτ /2.
For extracting moving targets from the undesired background clutter, Doppler processing is
desirable. For each range cell i , i = 1,..., Nr
at the kth
scan, a DFT is applied to the slow-time
observation. We first consider the case when only noise present. The N
d
-point DFT of yki
,
( n)
is given
by
N d −1
⎛ 2π
nl ⎞
Yki ,
(| l H0 ) = ∑ gki ,
( nTp)exp⎜−j
⎟
l=
0
⎝ N
, l= 1,..., N (21)
d
d ⎠
Yki
,
( l| H0)
is Gaussian distributed since it’s linear combination of the complex Gaussian
2 2
variable gki
,
( nT
p)
. The mean of Yki
,
( l| H 0
) is zero and the variance isσ
= N d
σ g
.
When a target is present, we ignore leakage effect and assume the apparent Doppler lies in
frequency bin , which means
ki ,
( |
1)
l 0
( )
f = mod f , PRF = l PRF/ N
da, k d , k 0 d
The expression forY
ki ,
( l | H1)
is given by
(22)
⎧⎪ NdAk + Yk, i( l| H0)
l = l0
Yki
,
( l| H1)
= ⎨
(23)
⎪⎩ Yki
,
( l| H0)
l ≠ l0
2
Y l H is also a Gaussian random variable with mean N A and covariance σ = . Through DFT,
the Doppler domain is divided into Nd
cells, the dimension of which is Δ
d
= PRF/ N d
. The reflected
power on the ambiguous range-Doppler grid is given by
z = | Y ( l)| 2 ,( i,
l)
∈S , k = 1,..., K (24)
k
{ k,
i
}
where S = { 0..., N} × { 0..., N }
r
d
denotes the set of all resolution cells in the ambiguous range-Doppler grid.
The total number of resolution cells is N= Nr× Nd. The center of each cell ( il) , is defined to be
(,)
at ( iΔr, lΔd)
. For each pixel of zk
, z il
k
, being the power of a complex Gaussian variable, follows either
a exponential distribution (if only noise exists) or a noncentral chi-square distribution with two
degrees of freedom (if a targets exists). Assume the pixels of z k
are independent, the likelihood
function of z k
is given by
(,) il
1 ⎛ z ⎞
k
p( zk
| bk, H0 ) = ∏ exp
2 ⎜−
2 ⎟ (25)
(,) il∈S
σ ⎝ σ ⎠
if there is no target exists or
( i0, l0)
( i0, l0)
1 ⎛ u
2
k
+ z ⎞ ⎛ uz ⎞
k
k k
p( zk | xk, bk)
== exp I
2 ⎜− 2 ⎟× ⎜ ⎟
0
×
2
σ ⎝ σ ⎠ ⎜ σ ⎟
⎝ ⎠
(26)
(,) il
1 ⎛ z ⎞
2 exp
k
∏ ⎜−
2 ⎟
(,) il∈S
,(,) il ≠( i0, l0)
σ ⎝ σ ⎠
if there is a target located in cell ( i0, l0)
of the ambiguous range-Doppler, where I()
0
⋅ is the zero-order
2
modified Bessel function; uk = ( NdAk)
is the expected target power after Doppler processing.
2
2
bk
= σ under H0
and bk
= ( σ , uk)
under H
1
, denotes the nuisance parameters.
k
The set of all data maps up to the k th visit might be denoted by Z = ( z ,..., 1
z
k ) , the likelihood
k
function of under H takes the form
Z
1
d
k
2
N d
σ g
15

k
k k k
( ) p(
p Ζ | X , B = ∏ zk | x
k,
b
k)
(27)
k = 1
k
k
given the target trajectory X = ( x1, x2,..., xk
) and nuisance parameter B = ( b1
,..., b k
).
5. TBD Procedures
In our situation, given the set of unthresholded ambiguous range-Doppler data maps up to
k
the k th visit, Z = ( z ,..., 1
z
k ) , the TBD procedures aim to compute optimal (in MAP sense) target
k
trajectory X = ( x1, x2,..., xk
) in ambiguous range-Doppler map, and its ambiguity number
k
k k
sequence Μ = ( m1
,..., m k
) . The close-form solution of X , Μ is hardly feasible. Moreover, the
k
existence of nuisance parameters, B in the measurements also increases the complexity of solving
this problem. An exhaustive search over the ambiguous range-Doppler maps, mode parameter and
unknown nuisance parameter space results in unbearable computational burden. Luckily, advantage
k
k
can be taken from the Markovian property of the target trajectory X and mode sequence Μ , the
k
k
search for the MAP estimates of X and Μ can be solved using DP methodology. In the following
k
k
section, we first derive the joint MAP estimates of X and Μ assuming the known nuisance
parameters and then extend the approach to the case of unknown nuisance parameters.
5.1 Known Nuisance Parameters
k
k
When the nuisance parameter is known, the joint MAP estimates of X and Μ is given by
ˆ k
( , ˆ k k k k
X Μ ) = arg max p( X , Μ | Z ) (28)
k X , Μ
k
k k k
Direct maximization of p( X , Μ | Z ) is extremely difficult since a close-form solution may not exist.
Brute-force maximization results in an exhaustive search over the ambiguous range-Doppler maps and
mode parameters, which represents a huge computation burden. Making use of the Markovian
k
k
property of X and Μ , as shown in Section 3, the search for target trajectory and mode sequence that
k k k
maximize the posteriori function p( X , Μ | Z ) can be solved using DP methodology.
The maximization in (28) is split into two parts
ˆ k ˆ k k k k
( X , Μ ) = argmax
⎡
max p( , | )
⎤
, k 1 k 1
k m ⎢
X M Z
x − −
k⎣X , M
⎥
(29)
⎦
The inner maximization is denoted by
( , ) max ( k , k | k
I x m = p X M Z ) (30)
k k k
,
k−1 k−1
X M
Ik( xk, mk)
can be interpreted as the merit of the candidate trajectory ending in any state xk
and mode mk
at the kth visit. The desired iterative relation is obtained using Bayes’ theorem to
express Ik+ 1( x
k+ 1,
mk+
1) in terms of Ik( x
k, mk)
. From Bayes’ theorem it follows that
k+ 1 k+ 1 k+
1
p( X , Μ | Z )
p( zk+ 1| xk+ 1) p( xk+ 1| xk, mk) p( mk+
1| mk, xk)
(31)
k k k
= p( X , Μ | Z )
k
p( zk+
1
| Z )
Using (31), Ik+ 1( x
k+ 1,
mk+
1)
can be written in the form
Ik+ 1( xk+ 1, mk+
1)
k+ 1 k+ 1 k+
1
= max p( X , Μ | Z )
(32)
X
k
, Mk
⎧p( zk+ 1| xk+ 1) p( xk+ 1| xk, mk) p( mk+
1| mk, xk)
⎫
= max ⎨ Ik( xk, mk)
k,
m
k
⎬
x k⎩
p( zk+
1
| Z )
⎭
k
Equation (32) constitutes the desired iterative relation. The function p ( zk
+ 1
| Z ) is used for
normalization purposes only and can be dropped without affecting the maximizer. The probability
density p ( z )
k+ 1
| x
k+
1
is obtained from(26). Working with the natural logarithm of (32), we transform
the merit function into an additive rather than a multiplicative function.
16

I
( x , m )
k+ 1 k+ 1 k+
1
(,) il
( i0, l
( 0)
i0, l0)
1 ⎛ 1 z ⎞ 2 u
1 1 1
k 1z
k+ uk+ + z
⎛ ⎛ ⎞⎞
k+
+ k+
1
= ln − ln + ln I
2 ∑ ⎜ −
2 2 ⎟−
⎜ ⎜ ⎟⎟
2
0 2
σ (,) il∈S
,(,) il ≠( i0, l0)
⎝ σ σ ⎠ σ ⎜ ⎜ σ ⎟ ⎟
⎝ ⎝ ⎠⎠
+ max ln ( | , ) + ln ( | , ) + ( , )
xk,
mk
{ px x m pm m x I x m }
k+ 1 k k k+
1 k k k k k
where ( i0, l0)
is the target location in ambiguous range-Doppler map when in state x k + 1. Note that adding
the same bias to the merit function of all states in any given frame will not affect the maximization
result. Thus, we are justified in dropping the first two terms on the right side of (33), which are
independent of ( x , m ).
k+ 1 k+
1
I
( x , m )
k+ 1 k+ 1 k+
1
( i0, l
( 0)
i0, l0)
u
2 u
1 1
k 1z
k+ + z
⎧
⎫
k+
⎪
⎛ ⎞
+ k+
1 ⎪
=− +ln I
2 ⎨ ⎜ ⎟
0
2 ⎬
σ ⎪ ⎜ σ ⎟
⎩ ⎝ ⎠⎪⎭
+ max ln ( | , ) + ln ( | , ) + ( , )
xk,
mk
{ p x x m p m m x I x m }
k+ 1 k k k+
1 k k k k k
where we are using the same symbol I for the redefined merit function.
Although the maximization in (34) is said to take place over all locations and ambiguity numbers
in the previous visit, in practices, only a small set of locations and ambiguity numbers which are
physically admissible need to be considered. Fig. 4 shows admissible search regions among successive
frames. The search region in range dimension depends on ambiguity number. For a cell (, i j)
at
k + 1st frame, given the ambiguity number mk
, say, m
k
= 2 , the corresponding search region in range
dimension at k th frame is well defined by RT(
mk
= 2) . When conditioned on a different ambiguity
number, say, m = 6 , the search region is redirected to a further region in range, R ( m = 6)
.
k
T
k
(33)
(34)
Fig.4. Illustration of admissible search regions among successive visits
17

As illustrated in Fig.4, the structure information of the ambiguous range-Doppler maps can be well
exploited to limit the search region to a small set of admissible locations and ambiguity numbers.
Limiting the search region can reduce the computational burden.
Once the merit function of last frame has been calculated, the outer maximization of (29) is
performed to yield the last location and ambiguity number estimates
( xˆ
, mˆ
) = argmax I ( x , m )
(35)
k k k k k
xk,
mk
The previous location and ambiguity number estimates are found by tracing backwards from ( xˆ
, ˆ
k
mk)
.
A final remark is that the merit function in (34) consists of two parts, the measurement
information part and the transition probability information parts (including mode conditioned state
transition and state dependent mode transition). When the noise σ
2 is known to be very high, the
measurement information part represented by the first two terms is trivial. As a consequence, the
transition probability information parts are much more relied to determine the target trajectory. When
there is no prior information about the target dynamics, all transition probabilities from feasible
previous state and mode to the current state and mode can be assumed equal. The first two terms in the
maximization can be treated as constants and dropped.
5.2 Unknown Nuisance Parameters
k
When the nuisance parameter B is unknown, we assume it’s an unknown deterministic sequence.
k
k
In fact, the target power sequence U = ( u1,..., u k
) in B could also be considered as a Markov sequence,
but for TBD application, this information is weak comparing to the information on the Markovian
k
k
k
property of X and Μ . For the deterministic parameter B without a priori information, its MAP
k
k
estimate is equivalent to its ML estimate. The MAP estimates of X and Μ , and ML estimates of
k
k k k k
B can be obtained by maximizing the conditional joint PDF p( Z , X , Μ | B ).
ˆ k
( , ˆ k
, ˆ k k k k k
X Μ B ) = arg max p( X , Μ , Z | B ) (36)
,
k k k X Μ , B
The above maximization is split into two parts
ˆ k ˆ k ˆk k k k k
( X , Μ , B ) = argmax
⎡
max p( , , | )
⎤
, k 1
,
k 1
k ,
k
k m ⎢
X Μ Z B
x − −
⎣X Μ B
⎥
(37)
⎦
The inner maximization is denoted by
( , ) max ( k , k , k | k
Γ x m = p X M Z B ) (38)
k k k
,
X
k−1 Mk−
1 , Bk
The desired iterative relation is obtained using Bayes’ theorem to express Γ
k
k
of Γ ( x , m ). Since X is independent of B , from Bayes’ theorem it follows that
k k k
p
k+ 1 k+ 1 k+ 1 k+
1
( X , Μ , Z | B )
k+
1 k k k k+1
=( p zk+ 1, xk+ 1, mk+
1| xk, mk, B ) p( Z , X , Μ | B )
zk +1
xk + 1
b
k + 1
, while k + 1
( x , m
k+ 1 k+ 1 k+
1
) in terms
Since the depends only on and
x depends on xk
and mk
; m
k + 1
depends
on x and m , the first term on the right side of (39) is broken down into
k k
k+
1
( zk+ 1, xk+ 1, k+
1| xk, k, B )
p m m
(40)
= p( zk+ 1| xk+ 1, bk+ 1) p( xk+ 1| xk, mk) p( mk+
1| xk, mk)
Using (39) and (40), we can write Γk+ 1( x
k+ 1,
mk+
1)
in the form
Γk+ 1( xk+ 1, mk+
1)
k+ 1 k+ 1 k+ 1 k+
1
= max p( X , Μ , Z | B )
,
Xk , Μk Bk+
1
= max p( zk+ 1| xk+ 1, bk+ 1) max { p( xk+ 1| xk, mk) p( mk+
1| mk, xk) Γk( xk, mk)
}
bk+
1
xk,
mk
which is the required iterative equation. We work with the natural logarithm of (41) to transform the
merit function into an additive function
Γk+ 1( xk+ 1, mk+
1)
= max ln p( z | x , b ) +
(42)
bk
+ 1
xk,
mk
k+ 1 k+ 1 k+
1
{ p x x m + p m m x +Γ x m }
max ln ( | , ) ln ( | , ) ( , )
k+ 1 k k k+
1 k k k k k
(39)
(41)
18

where we are using the same symbol Γ for the redefined merit function of (42). Note that similar to
(34), the merit function in (42) also consists of two parts, the measurement information part
(represented by the first term on the right side) and the transition information part (represented by the
last three terms on the right side). Equation (42) differs (34) only in the measurement information part,
where the former has to perform maximization over the space of nuisance parameters b
k + 1
. We now
consider the ML estimate of the nuisance parameters. The logarithm of the likelihood function in (42)
can be written as
2 sk+ 1+
uk+
1
0 0
( ) { ( )}
( i
ln , l ) 2
p zk+ 1| xk+ 1, bk+ 1
=−Nlnσ
− + ln I
2
0
2 uk+ 1zk+
1
/ σ
(43)
σ
(,) il
where s = ∑ S +
; ( i0, l0)
denotes the corresponding cell when a target locates at x k +1
.
k+ 1
z
(,) il∈
k 1
We evaluate the partial derivatives of the logarithm likelihood function as
( i0, l0) 2
( )
I ( )
( 0, 0)
1
2
1 1
/
i l
z
uk+ zk+
σ
k+ 1
xk+ 1 k+ 1 1
k+
1
=− +
2
2
∂u ( i0, l0) 2
k+ 1
σ I ( )
1
0
2 u
1 1
/
k
k
zk
σ σ u
+
+ +
( i0, l0) 2
I (
( )
0, 0)
1
2 u
1 1
/ i l
k+ zk+
σ
k+ 1
xk+ 1 k+ 1
2 z
1 1
k
u
k+ k+
k
= − −
2
2
2 2
( i0, l0) 2 2
2
∂σ
( σ ) σ I0( 2 uk+ 1zk+
1
/ σ ) ( σ )
∂ln p | , b 1
z
( z
)
∂ ln p | , b s + u N
+ 1 + 1
where I() = I() ′ is the first-order modified Bessel Function. Equating (44) to zero, we obtain
1 0
( i0, l0) ( uˆk+ zk+
2
σˆ
)
( i0, l0) ( uˆk+ zk+
2
σˆ
)
I1 2
1 1
/
uˆ
k + 1
( i0, l0)
I 1
0
2
1 1
/ zk
+
= (46)
Substituting the result to (45), and equating it to zero, we have
2
uˆk
+ 1= s k + 1− Nˆσ
(47)
By solving (46) and (47) jointly, we can find the MAP estimate of the unknown
2
parameter bˆ 2
ˆ ˆ
k+ 1
( σ , uk+
1)
. Notice that both ˆ σ and u ˆk + 1
enter the function of (46) in a nontrivial way,
therefore, solving the function will be an iterative numerical process, which is computationally
expensive.
Now we consider a simple way of estimating (
2
, ). The target signature only affects the cell it
σ u k + 1
locates, as a result, if a target is present and locates in cell ( i0, l0)
at k + 1st frame, the pixel ( i0 , l0
z )
k + 1
is a
(,) il
noncentral chi-square variable, while the other pixels z
k + 1
, (, il)
∈ S , (, il) ≠ ( i0, l0)
, are exponentially
distributed variables. Both the noncentral chi-square and exponential variables share the same
2
parameterσ , therefore, we can estimateσ 2 (mean of the exponential distribution) by averaging the
exponentially distributed pixels
( i0, l0)
2 1
( , ) 1 1
ˆ σ il sk − zk
= ∑ z
+ +
(,) il ,(,) il ( i 1
0, l0)
k +
=
∈ ≠
N −1 S
(48)
N −1
Substituting (48) into (47) gives
( i0, l0)
Nzk+
1
− sk+1
uˆ
k + 1
=
(49)
N −1
Substituting (48) and (49) into (42) and dropping the terms independent of ( xk+ 1,
mk+
1)
, we end up with
a modified merit function
Γ ( x , m )
k+ 1 k+ 1 k+
1
( i0, l0)
( i0, l0) ( i0, l0)
(2N−1)
z 2 ( N 1)( N z
1
k 1
sk)
z
k+
−s
⎧
⎫
k ⎪
⎛ −
+
− ⎞
k+
1 ⎪
=− +ln I
( i 0
0, l0) ⎨ ⎜
⎟
( i0, l0)
⎬
sk −z ⎜
k+ 1
sk −z
⎟
k+
1
⎩⎪
⎝
⎠⎭⎪
+ max ln ( | , ) + ln ( | , ) +Γ ( , )
xk,
mk
{ p x x m p m m x x m }
k+ 1 k k k+
1 k k k k k
where we are using the same symbol Γ for the redefined merit function of (50).
Once the merit function of last frame has been calculated, the last state and mode estimates are
obtained using
(44)
(45)
(50)
19

( xˆ
, m ˆ ) = argmax Γ ( x , m )
(51)
k k k k k
xk,
mk
The previous states and modes are found by tracing backwards from ( xˆ
, mˆ
).
k
k
6. Simulation and Results
In our simulation, an approaching target, referring back to Fig. 2, travels at initial range r
0
= 300 km,
with velocity v = 2.8 Mach, and headingα = 225 deg. Since the target is not travelling along the radial
direction, the Doppler is inevitably time-varying. We set the Doppler rate increment variance
2 2
Q = (6Hz/s ) .Fig. 5 and Fig.6 show the trajectories of target range and Doppler, respectively. The
target Doppler information is represented by the observed apparent Doppler and corresponding
ambiguity number.
Fig. 5. target range trajectory
Fig. 6. target Doppler trajectory, represented by apparent Doppler trajectory and ambiguity number
trajectory
The ambiguity number transits for i to j = i − 1 when the guard condition is satisfied. To account for
the random variations in Doppler increment, δ ij
is modeled as a Gaussian distributed variable with
20

mean δ
ij
= 50 Hz and variance Σ
ij
= 16 , denoted byδij
N ( δij
;50,16), and then, as discussed in Section
3, the state-dependent ambiguity number transition probability is
⎧Φ( Cx
ij k
−δij ; Σ
ij ) if j = i
⎪
πij ( xk ) = ⎨1 −Φ( Cijx k
−δij; Σ
ij ) if j = i−1
(52)
⎪
0 else
⎪⎩
A maximum target velocity v
max
= 3Mach is assumed in our simulation.
We consider here a low PRF surveillance radar using a carrier frequency of 1.2 GHz and
transmitting trains of N
d
= 64 rectangular pulses with pulse width τ = 1.2 μs
and PRF=860 Hz. Radar
revisit interval is T
R
= 4 s . For the reader’s sake, the main surveillance radar parameters are
summarized in Table I .After discretization and Doppler processing, the received data are transformed
into the ambiguous range-Doppler maps. For low PRF surveillance radars, the measured range is
unambiguous while the measured Doppler is ambiguous. Since we are primarily concerned with early
warning of approaching targets, the range cells under consideration are limited to the
furthest N
r
= 256 cells. Thus, after each visit, we obtained a 256× 64 ambiguous range-Doppler map. In
the simulation, the unknown target’s power u k
at each visit is randomly drawn from an exponential
distribution with average return power u . Note that we ignore the power variation caused by the
changing range since the range variation is small in comparison to the distance between the target and
radar. The SNR is defined in range-Doppler domain. The SNR of the target at kth visit is expressed
2
as (SNR) = 10log( u / σ ) .
k
k
Table I Radar System Parameters
It’s clear that SNR experienced by the target varies at each visit due to the variation of target power. In
the following, in order to characterize the detection performance of the proposed TBD algorithms, we
define an average SNR as
2
SNR = 10log( u / σ ) (53)
Notice that, since we assume the number of coherently processed pulses N
d
= 64 , the average received
SNR per pulse is obtained by subtracting 10log 64 18dB from the reported values in range-Doppler
domain.
Fig.7 shows a mesh plot of the reflected power in ambiguous range-Doppler maps. The target is
located at cell (231, 14). When the SNR is high (13dB) and it’s easy to discern the target as a high
peak (Fig.7(a)). When the SNR is relative low (10dB), it’s not easy to distinguish the target from the
noise (Fig.7(b)). TBD procedures provide a way to improve the detestability of low SNR target by
jointly processing several range-Doppler maps. In the following simulation, we set the number of
jointly processed ambiguous range-Doppler maps to K = 6 .
21

(a)
(b)
Fig. 7. mesh plot of the reflected power in ambiguous range-Doppler maps: (a) SNR=13dB, (b)
SNR=10dB
Fig. 8 shows the final merit function for a fixed mode m
K
= 5 . When no target exists, it’s obvious that
the merit function tends to noisy (Fig.8(a)). When a target with an average SNR=10dB is present (the
nuisance parameters are assumed known), it’s clear that the noisy merit function has a distinct peak at
a particular location in the ambiguous range-Doppler map (Fig.8(b)). The corresponding target final
state can be extracted by thresholding the resultant merit function.
6.1 Selection of Threshold
In order to calculate the threshold, we first have to define the probability of false alarm P fa
. In our
situation, Pfa
is defined as the probability of detecting a false track in the absence of target (i.e., under
the hypothesis H
0
). This is equal to the probability of the maximum noise merit function exceeding
the threshold VT
P
fa
= Pr ( max Γ
K
(
K
, mK ) >
T
xK,
mK
x V ) (54)
where x K
∈{noise states} .
Given the desired P fa
, the threshold is determined as the (1−
Pfa
)-quantile of the test statistic, i.e.,
max
x ,
( , ) given in
K m
Γ
K K
x
K
mK
(54). Unfortunately, the statistical distribution of the test statistic is rather
complicated [11], sometimes even infeasible. This mainly because 1) the max operator in the DP
recursion brings in nontrivial nonlinearities and non-Gaussianity; 2) the samples of the test statistic are
22

actually correlated since the backward search process in the DP recursion may end with a common
previous location xk −1
for two different locations at frame k . The threshold is, hence, evaluated via
computer simulation.
(a)
(b)
Fig. 8. mesh plot of the merit function ( K = 6) for a fixed mode m
K
= 5 : (a) no target exists (b) a target
with average SNR=10dB exists
6.2 Performances Evaluation
We now move to the analysis of the detection and tracking performance of the proposed algorithm.
As in [7], we define the following performance measures for evaluating the proposed TBD algorithms.
Definition 1 The probability of target detection P d
is defined as the probability of the maximum final
merit function exceeds the threshold VT
and meanwhile, the corresponding retrieved cell, say, ( iˆ
, ˆ
K
l
K
),
is within ε r
cells in range and ε d
cells in Doppler from the actual location, ( iK
, l
K
), and the ambiguity
number estimate m ˆ K
is equal to the true one. Hence
Pd = Pr
⎛( max
K( K, mK)
⎞
⎜ Γ x > V
T
,
) ∪E ⎟
⎝ xK
m
d
(55)
K
⎠
where the event E d
is expressed as
abs( iˆ
K
−iK)
≤ ε
r, abs( lˆ
K
−lK)
≤ ε
d
, mˆ K
= m K
. (56)
23

Definition 2 The probability of target detection and reliable tracking P dT ,
is defined as the probability
of making a correct target detection and of recovering a target trajectory such that each retrieved
state, ( iˆ
ˆ
k, lk)
, k = 1,... K , is within ε
r
cells in range and ε d
cells in Doppler from the actual location, and
the estimated ambiguity number sequence ˆ K
Μ = ( mˆ ˆ
1,..., m K
) is equal to the true one. Hence
PdT ,
= Pr
⎛(
max
K( K, mK)
V
⎞
⎜ Γ x >
T
∪
dT , ⎟
⎝ xK,
m
) E (57)
K
⎠
where the event
E dT ,
is expressed as
abs( iˆ
k
−ik)
≤ ε
r, abs( lˆ
k
−lk)
≤ εd, k = 1,... K , Μˆ K = Μ
K
(58)
It’s evident that the overall system performance depends on the choice of the tolerance factors: ε r
and ε d
. The smaller the tolerance factors, the larger the SNR is required in order to maintain a certain
value of detection probability P d
and P under a preset probability of false alarm rate. In the
dT ,
following, the detection and tracking performance with respect to average SNR and tolerance factors
are investigated.
We first consider the case that the nuisance parameters are known in advance; the corresponding
−6
implementation of the DP recursive is given in (33). In Fig.9, we set P fa
= 10 , and plot P
d
and P versus average SNR under different settings of tolerance factor. The tolerance
dT ,
factor ( ε
r, ε
d)
from up to down is (4,2) , (2,1) and (0,0) . ( ε
r, ε
d) = (0,0) means correct detection and
estimation. We note that as the tolerance factor becomes smaller (i.e., the estimation accuracy
becomes higher), both P d
and P dT ,
decreases considerably. We can also notice that compared to P dT ,
,
P is less sensitive to the tolerance factor. Satisfactory detection and tracking performance can be
d
achieved when the tolerance factor is (4,2) .
Fig. 9. P and
d
P dT ,
versus the average SNR in range-Doppler domain, with known nuisance parameters,
( ε , ε ) = (4, 2), (2, 1), (0, 0) from up to down
r
d
We now investigate the situation when the nuisance parameters are unknown; the DP recursive with
−6
unknown nuisance parameters is given in (50). In Fig. 10, we set P fa
= 10 , and report P d
24

and P dT ,
versus the average SNR under several values of tolerance factor. For the sake of comparison,
the tolerance factor ( ε
r, ε
d)
from up to down is also set to (4,2) , (2,1) and (0,0) . Similar result to Fig. 10
is obtained: both P d
and P decreases considerably with the decreasing of tolerance factors. By
dT ,
inspection, we note that P d
and P dT ,
suffer an approximately 1dB loss due to the prior uncertainty as to
the nuisance parameters.
Fig. 10. P and
d
P versus the average SNR in range-Doppler domain, with unknown nuisance
dT ,
parameters,
( ε , ε ) = (4, 2), (2, 1), (0, 0) from up to down
r
d
7. Conclusions
In this paper we have investigated the TBD procedures with reference to a low PRF surveillance
radar framework. We give physical and statistical models for the received power measurements on the
ambiguous range-Doppler maps. In order to avoid the nonlinear transformation between the target’s
state coordinates and measurement coordinates (defined by range and apparent Doppler); we model
the target’s dynamics directly in the measurement coordinate. Exploiting the target dynamics in the
ambiguous range-Doppler maps as a hybrid system, we have proposed DP based methods for joint
MAP estimation of target’s trajectory in the ambiguous range-Doppler maps and the corresponding
ambiguity number sequence. The detection and tracking performance of the proposed procedure is
studied using two measures: 1) detection probability P d
, 2) detection and reliable tracking
probability P dT ,
. Simulation results show satisfactory detection and tracking performance is achieved
when the tolerance factor is set to within four cells in range and two cells in Doppler from the true
location. The impact of the uncertainties in the nuisance parameters (target power and noise variance)
has also been studied. Simulation shows that both P d
and P dT ,
suffer an approximately 1dB loss due to
the prior uncertainty of the nuisance parameters.
25

Acknowledgments
Xiaobo Deng would like to acknowledge his useful discussion with Dr. Xuezhi Wang
(Department of EEE, The University of Melbourne), and thank the China Scholarship Council for
supporting his visit at the Department of EEE, The University of Melbourne.
References:
[1] Reed, I., Gagliardi, R. and Stotts, L.: ‘A recursive moving-target-indication algorithm for optical
image sequences’, IEEE Trans. Aerosp. Electron. Syst., 1990, 26, (1), pp. 434-440
[2] Barniv, Y.: ‘Dynamic programming solution for detecting dim moving targets’. IEEE Trans.
Aerosp. Electron. Syst., 1985, 21, (1), pp. 141-155
[3] Arnold, J., Shaw, S., and Pasternack, H.: ‘Efficient target tracking using dynamic programming.
IEEE Trans. Aerosp. Electron. Syst., 1993, 29, (1), pp. 44-56
[4] Xiong. Y, Peng, J. X., Ding, M. Y., and Xue, D. H.: ‘An extended track-before-detect algorithm
for infrared target detection’, IEEE Trans. Aerosp. Electron. Syst., 1997, 33, (3), pp. 1087-1092
[5] Tonissen, S. M., and Evans, R. J.: ‘Performance of dynamic programming techniques for trackbefore-detect’,
IEEE Trans. Aerosp. Electron. Syst., 1996, 32, (4), pp. 1440-1451
[6] Wallace, W. R.: ‘The use of track-before-detect in pulse-Doppler radar’. Proc. Int. Conf. Radar
2002, pp. 315-319
[7] Buzzi, S., Lops, M., and Venturino, L.: ‘Track-before-detect procedures for early detection of
moving target from airborne radars’, IEEE Trans. Aerosp. Electron. Syst., 2005, 41, (3), pp. 937-
954
[8] Buzzi, S., Lops, M., and Ferri, M.: ‘Track-before-detect procedures in a Multi-target environment’,
IEEE Trans. Aerosp. Electron. Syst., 2008, 44, (3), pp. 1135-1148
[9] Seah, C. and Hwang, I.: ‘State estimation for stochastic linear hybrid systems with continuousstate-dependent
transitions: an IMM approach’, IEEE Trans. Aerosp. Electron. Syst., 2009, 45, (1),
pp. 376-392
[10] Blom, H., and Bloem, E.: ‘Exact Bayesian and particle filtering of stochastic hybrid systems’,
IEEE Trans. Aerosp. Electron. Syst., 2007, 43, (1), pp. 50-70
[11] Johnston, L. A., and Krishnamurthy, V.: ‘Performance analysis of a dynamic programming track
before detect algorithm’, IEEE Trans. Aerosp. Electron. Syst., 2002, 38, (1), pp. 228-242
26

Fast realization of Automatic optical and infrared image registration
Weiping Yang 1 2 * , Xuezhi Wang 2 , Bill Moran 2
* Presenter
1. National University of Defence Technology, Hunan, 410073, China
2. The University of Melbourne, Parkville, VIC 3010, Australia
Because of the requirement of practical applications, it is necessary for image
registration between different source images. However, the image characteristics from
different imaging sensors are different, so the wanted results can not be obtained if the
intensive images are directly used. In order to solve this problem, edge detection and
extraction methods should be applied, then the correlation tactics is used. But because the
standard cross correlation method is time expensive, fast realization method should be
adopted. Here we use the changeable resolution method to do this, and get good results with
only less than one fortieth time consuming of the standard algorithm.
Keywords automatic image registration, edge detection, changeable resolution, normalized
cross correlation
1. Introduction
Water resource is the key factor of human being social development, and paid more attention by
the world. The issue of how to use the limited freshwater effectively is one of the subjects in all over
the world. In order to study automatic water saving irrigation technique, we need to know the water
status of the plants. To say further, if we want to know the plant water status, we need to know the
canopy temperature at first. Therefore, researchers have done many fruitful results [1-5] . Using
nondestructive optical image sensing technique to obtain the canopy temperature is a challenging
project. The first step is to automatically register the infrared image with the optical image, then
extract the temperature of the interested regions, and finally can get the temperature estimation of the
canopy.
As we all know, researchers have presented many algorithms on image registration and image
match, for example, cross correlation [6] , mutual information [7] , correlation ratio [8] , and SIFT based
methods [9] . However, for image registration between different source image pairs, there are some
practical problems if we directly use the intensive images to do the image registration processing, in
reference [10] we have pointed out, here we only discuss its fast realization. It is because of the
existing differences between optical and infrared images that we should edge the images at first, then
using certain image registration method to process the edge images. Because of the time consuming,
standard cross correlation( normalized cross correlation) method is improper, the changeable
resolution method based normalized cross correlation may be the best choice. Experiment tests show
that the changeable resolution normalized cross correlation method has not only the same good
registration results, but also fast running speed with less than one fortieth time consuming of the
standard cross correlation method.
The paper is organized as follows: In Section 2 the algorithm is described and the flow diagram is
given. Section 3 is the implementation of the above algorithm. The experiment results are presented in
Section 4, then followed by the discussion and conclusion.
2. Changeable resolution algorithm based on normalized cross correlation
When we want to register the IR image to the reference optical image, we should obtain its
location and rotation angle. In general, the IR image is wholly overlapped with the reference optical
image. Moreover, we assume that the space resolution of the two images is the same. And then the
changeable resolution algorithm based on normalized cross correlation can be shown as Fig 1.
Because it needs to calculate the correlation coefficient pixel by pixel, normalized cross
correlation (NCC) algorithm is time expensive. In order to solve the problem of the slowly running
27

speed, several methods can be utilized. Here we use the changeable resolution algorithm to accelerate
it. The key step of the changeable resolution algorithm is to reduce the resolution at first, then roughly
do image registration in the lower resolution images, and then do it accurately near the possible
positions in the full resolution images, finally we can get the registration results.
Optical image
IR image
Edge extraction
Edge extraction
Resolution Reducing
Resolution Reducing
NCC
Roughly registration
NCC at certain positions
Selecting optimal position
Rotation angle estimation
Image registration parameters
Fig 1 Flow diagram of changeable resolution algorithm based on NCC
Suppose Im O and Im IR (image size is M×N) stands for the optical and IR image respectively, f is
the image zoom factor, Im Ol and Im IRl (image size is Mh×Nh) stands for the lower resolution one,
accordingly, Im Oe , Im IRe , Im Ole , Im IRle are the edged images of the corresponding images, then the
generalize cross correlation can be expressed as
ρ ( u,
v)
=
l
ρ(
k,
l)
=
Mh
Nh
∑∑
i= 0 j=
0
M
N
∑∑
i= 0 j=
0
(Im
(Im
Oleu, v
( i,
j)
− ImIRle
( i,
j))
σ
lu,
v
k , l
σ
IR
IRl
Oek, l
( i,
j)
− ImI
Re
( i,
j))
σ
σ
Where ρ
l
( u,
v)
is the cross correlation coefficient calculated from the lower resolution image pair
Im Ole and Im IRle , (u,v) is the coordinate index of the optical image Im Ole , Im Oleu,v
is an image located in
(u,v)th of the image Im Ole and its size is the same as image Im IRle , σ lu,v and σ IRl are the standard
deviation of the corresponding images respectively; ρ ( k,
l)
is the cross correlation coefficient
calculated from the raw resolution image pair Im Oe and Im IRe , (k,l) is the coordinate index of the
(1)
(2)
28

optical image Im Oe ,
Im Oek,l
is an image located in (k,l)th of the image Im Oe and its size is the same as
image Im IRe , σ k,l and σ IR are the standard deviation of the corresponding images respectively.
3. Algorithm implementation of the changeable resolution algorithm
(a) Raw optical image
(b) Raw IR image
Fig 2 Raw images
In order to implement the algorithm, the first step is edge extraction. It is important to the results
of image registration. There are many methods on edge detection and extraction, such as Prewitt,
Sobel, Robert operators and Canny algorithm, and so on. As the above mentioned, the infrared image,
different with the reference optical image(Fig 2), has only a lower resolution and less details, we need
to reserve its edge information while edge processing. Here the edge operation is based on the Sobel
operator, and it can be described as follow,
⎡1
0 −1⎤
⎡ 1 2 1 ⎤
S =
⎢ ⎥
⎢
2 0 − 2 , , (3a)
1
⎥ S
⎢
⎥
2
=
⎢
0 0 0
⎥
⎢⎣
1 0 −1⎥⎦
⎢⎣
−1
− 2 −1⎥⎦
⎡2
1 0 ⎤ ⎡ 0 1 2⎤
S =
⎢ ⎥
⎢
1 0 −1
,
3
⎥
S
⎢ ⎥
(3b)
4
=
⎢
−1
0 1
⎥
⎢⎣
0 −1
− 2⎥⎦
⎢⎣
− 2 −1
0⎥⎦
Using these four operators to convolute with the raw images, we can get
2
2
∑∑
E ( i,
j)
= Im( i + m −1,
j + n −1)*
S ( m,
n)
. k=1,2,3,4
k
m= 0 n=
0
E( i,
j)
= max E ( i,
j)
k
k
k
(5)
Where E( i,
j)
is the edge of the point (i,j) of image Im.
After edge detection, we can get the edged image Im Oe , Im IRe , Im Ole , Im IRle , then using NCC
algorithm to do the rough registration. On account of possible disturbing, we choose some points with
the former higher correlation coefficients, marked as {( u
k
, vk
)}.
When we get the candidate points{ ( u k
, v , next step is to do the accurate image registration near
k
)}
these specified points of the full resolution images. Here we adopt a weighted coefficient c to ensure
that the final point with the maximum correlation coefficient is the optimal one.
ρ
k
= c ρl
( uk
, vk
) + (1 − c)
ρ(
u′
k
, v′
k
)
(6)
ρ max ρ
* =
k
k
k
(4)
(7)
29

Where u ′ , v′
) is the point position in the full resolution image Im Oe , ρ u ′ , v′
) is the correlation
(
k k
(
k k
coefficient of accurate image registration, ρl
( uk
, vk
) is the correlation coefficient of rough image
registration. Through the step, we can get the registration position u ′ , ′ ) .
( * v *
k k
The final step is the estimation of rotation angle. At present the permitted rotation angle θ range is
from -10°to 10°. We divide the range into 200 angles. To each angle, we calculate its correlation
coefficient, then choose the angle θ * with the maximum correlation value.
4. Experiment results
The above algorithm was tested using many image pairs. All of the results are successful with
tolerable errors(Fig 3). In order to test the running speed, contrasting experiment was done between
the changeable resolution algorithm and standard NCC in the same running environment and
conditions, the results can be shown in table 1. The experiments show that the changeable resolution
algorithm has the same good registration results with a fast speed, only spending less than one fortieth
running time compared with the standard algorithm. Because we don’t know the true registration
positions, we compare them with the results of manual registration.
Fig 3 Raw image pairs and their registration results
Table 1 experiment results of changeable resolution algorithm based on NCC
Number of Success Running time per pair Registration error
Image pairs number (average)
(max, min, average)
Standard NCC 20 20 610s (6,0,2.6)
Changeable
resolution
algorithm
20 20 15s (6,0,2.6)
All of the tests are done in the condition of f = 2 and using Sobel operator to get the image edges.
The contrasting experiments with the NMI (normalized mutual information) and CR(correlation ratio)
algorithms have also been done( Fig 4, Table 2). From these figures and tables, we can see the
changeable resolution algorithm based on normalized cross correlation is the best choice.
30

Table 2 experiment results of several different algorithms
Number of Image Success* Registration error
pairs
number (max, min, average)
Changeable resolution 10 10 (6,0,2.6)
algorithm
NMI algorithm 10 8 (9,0,2.7)
CR algorithm 10 0 Null
Remark: *here a successful registration is whose position error is less than 10 pixels.
(a) Image pair 1(Optical(left), IR(right))
(299,120,-0.8)
(b) registration result 1 of Changeable Res Algo
(c) registration result 1 of NMI alg (304,125,-0.5) (d) registration result 1 of CR Alg (318,257,0.5)
(e) Image pair 2(Optical(left), IR(right)) (f) registration result 2 of Changeable Res Algo (364,153,1.5)
(g) registration result 2 of NMI alg (363,157,1.0) (h) registration result 2 of CR Alg (283,93,3.9)
Fig 4 Raw image pairs and their registration results using different algorithms
5. Discussion
From the experiment results we can find that NMI and CR algorithms are unsuitable to the
different source image registration because these two methods are dependent on the intensity
distributions of the images, while the changeable resolution algorithm is related to the edge of the
images, that is to say, the latter algorithm is dependent on the image features and independent with the
image intensities. Though they have good results in many applications, the image registration methods
31

ase on SIFT are also unsuitable to our application because common key points can hardly be found.
The important reason is that the key points are wholly dependent on the image intensities.
With regard to the running speed, though the images are only zoomed out at 2 times, a higher
running efficiency is obtained. The changeable resolution algorithm only takes less than one fortieth
running time of the standard pixel by pixel cross correlation method, but they have same registration
results, so the changeable resolution algorithm is a fast and effective realization of the NCC method.
6. Conclusion
The paper presented a fast realization algorithm of automatic optical and infrared image registration.
It reduces the computation considerably by utilizing changeable resolution algorithm and has obtained
inspiring results at a fast running speed. It can also be used in other applications. As we mentioned
above, the image registration is only the first work to do, the aim is to estimate the canopy temperature.
But delightful thing is that it is our ongoing research, and some great progresses have been made.
Moreover, how to get more accurate registration and how to make algorithm suit to more variational
background are also our ongoing research.
7. Acknowledgement
We would like to thank Mr Ashley Wheaton and Dr Nicola Cooley for their providing images and
some useful discussions.
References:
[1] H.G. Jones. Use of thermography for quantitative studies of spatial and temporal variation of
stomatal conductance over leaf surfaces. Journal of Plant, Cell and Environment (1999) 22, 1043–
1055.
[2] H.G. Jones. Use of infrared thermometry for estimation of stomatal conductance as a possible aid
to irrigation scheduling. Journal of Agricultural and Forest Meteorology 95 (1999) 139-149.
[3] L. Guilioni , H.G. Jones, I. Leinonen, J.P. Lhomme. On the relationships between stomatal
resistance and leaf temperatures in thermography. Journal of Agricultural and Forest Meteorology
148 (2008): 1908 – 1912.
[4] O. Grant, L. Tronina1, H. G. Jones and M. M. Chaves (2007) Exploring thermal imaging variables
for the detection of stress responses in grapevine under different irrigation regimes. Journal of
Experimental Botany, Vol. 58, No. 4, pp. 815–825, 2007.
[5] Wheaton, A.D., Cooley, N., Dunn, G., Goodwin, I., Needs, S., (2007) Evaluation of infrared
thermography to determine the crop water status of Cabernet Sauvignon grapevines. Poster paper.
13th Australian Wine Industry Technical Conference, Adelaide, 28 July – 2 August.
[6] Du-Ming Tsai, Chien-Ta Lin. Fast normalized cross correlation for defect detection. Journal of
Pattern Recognition Letters 24 (2003) 2625–2631.
[7] A. Roche, G. Malandain, etc. The correlation ratio as a new similarity measure for multimodal
image registration. Medical Image Computing and Computer Assisted Intervention - MICCAI'98,
1496: 1115-1124,1998.
[8] S. Klein. Evaluation of Optimization Methods for Nonrigid Medical Image Registration Using
Mutual Information and B-Splines. IEEE Transactions on Image Processing, 16(12): 2879-2890,
2007.
[9] Jian Chen, Jie Tian. Real-time multi-modal rigid registration based on a novel symmetric-SIFT
descriptor. Progress in Natural Science 19 (2009): 643–651.
[10] Yang Weiping, Wang Xuezhi, etc, Automatic optical and IR image fusion for plant water stress
analysis, Proc of 12 th International Conference on Information Fusion, 1053-1059, 2009.
32

Impulsive Interference Detection Method Based on Morlet Wavelet and
Maximum Likelihood Estimation
Yuan Y. He 1 2 * , Chang J. Yu 2 , and Tai F. Quan 2
* Presenter
1. Harbin Institute of Technology, Harbin, 150001, China
2. The University of Melbourne, Parkville, VIC 3010, Australia
The performance of the frequency monitor system (FMS) of high frequency (HF) radar is
degraded by signal corruption due to impulsive interferences such as lightning and meteor
echoes. These interferers raise the FMS spectrum noise level, as a result, FMS can’t pick out
the real minimum disturbance frequency for HF radar. It is desirable to extract the impulsive
interference components from the FMS data without removing any other information.
However, Current wavelet detecting impulsive interference techniques, not matching
impulsive interference very well, always lead to non-impulsive interference components
degraded and lost. In this paper, we construct a more suitable impulsive interference detection
method, by using the Morlet wavelet, matching well the shape of the impulsive interference
signal, decomposition algorithm and maximum likelihood estimation thresholding rule,
specifically designed to detect impulsive interference, so as to less disruption to the nonimpulsive
interference information when extracting the impulsive interference. The
effectiveness of the proposed technique has been proved by both simulated and practical
experiments.
33

Demonstration and Performance Analysis of an Uplink based on Digitized
RF-over-Fiber Signal Transport
Yizhuo Yang 1 * , Christina Lim 1 , Prasanna Gamage 1 and Ampalavanapillai Nirmalathas 1
* Presenter
1. The University of Melbourne, Parkville, VIC 3010, Australia
In this paper, we demonstrate and analytically analyze a digitized RF-over-fiber transport
scheme for uplink transmission in a Radio-over-Fiber system. We achieve significant
improvement in both signal-to-noise ratio (SNR) and dynamic range. The experimental and
analytical results show that digitized RF link is less susceptible to nonlinearity and fiber
attenuation which makes this scheme well suited for long-reach optical backhaul for wireless
networks.
34

Fiber Nonlinearity Compensation for CO-OFDM Systems with Periodic
Dispersion Maps
Liang B. Y. Du 1 * and Arthur J. Lowery 1
* Presenter
1. Electrical and Computer Systems Engineering, Monash University, Clayton, VIC 3800,
Australia
We show that the nonlinear limit of CO-OFDM systems at 100+ Gbps in periodic
dispersion maps can be increased by 4 dB using nonlinear precompensation. We also show
compensation is beneficial in WDM systems.
35

Effect of mini-tyrosl-tRNA synthetase / mini-tryptophanyl-tRNA synthetase
on ischemic angiogenesis in rats with acute myocardial infarction
Rui Zeng 1 * , Yu-cheng Chen 1 , Zhi Zeng 1 , Wei-qiang Liu 1 , Xiao-xia Liu 2 , Rui Liu 3 , Ou
Qiang 3 , Xian Li 3 , Smith AI 4
* Presenter
1. Department of Cardiology, West China Hospital, School of Clinic Medicine, Sichuan
University, Chengdu ,China .610041
2. Department of Epidemiology and Health Statistics,School of Public Health, Sichuan
University,Chengdu ,China .610041
3. Laboratory of Peptides Related with Human Diseases,The National Laboratory of
Biomedicine, Sichuan University, Chengdu,China .610041
4. Department of Biochemistry and Molecular biology, Monash university, Melbourne,
Australia.3800
Background: To clarify the mechanism of the angiogenesis effect of mini-TyrRS/mini-
TrpRS in rodent primates with acute myocardial infarction. Methods: Left coronary artery
ligation was used to establish the model of acute myocardial infarction in rats (Sprague-
Dauley male rats,250-300g, 2-3 months old), mini-TyrRS/mini-TrpRS were administered by
coronary artery polyvinyl catheter injection to rats. Rats were randomly divided into four
experimental groups: sham operated group; coronary artery ligation (CAL); CAL+mini-
TyrRS(20μl, twice daily, 600μg. Kg -1 .day -1 );. CAL+mini-TrpRS(20μl, twice daily, 600μg.
Kg -1 .day -1 ). The experiment was carried out at 4 time points as the 3rd, 7th, 14th, and 28th
day after ligation. To determine whether mini-TyrRS/mini-TrpRS affected the angiogenesis
activity of rats with myocardial infarction, we measured the myocardial infarction size by
TTC staining, and the density of capillary using immunohistochemistry staining to investigate
the expression of VIII factor. Results: Compared with CAL group
(9.3%,15.0%,30.9%,39.3%),the myocardial infarction size of mini-TyrRS group at 3rd, 7th,
14th, and 28th day were respectively 8.0%,10.9%,22.4%,28.1%, as for mini-TrpRS group,
were 10.7%,18.8%,37.5%,51.2%, but only in day 14th and 28th has significant difference,
p

A rodent model of acute myocardial infarction(MI) was first developed in the rat[7].More recently
a murine equivalent has been described[8],providing the means to exploit the increasing availability of
many useful transgenic and knockout mouse strains. Complete occlusion of the left anterior
descending(LAD)coronary artery induces an acute MI. The resulting ischaemia in the left ventricular
wall has been visualized with Evans blue/TTC perfusion assays[9-12],which allows quantification of
the infarct area. Further work by Guo et al.[13] has demonstrated ischaemic preconditioning after
short-term LAD occlusion in mouse, thus validating the physiologic relevance of this infarct size.
Aminoacyl-tRNA synthetases catalyze the first step of protein synthesis that consists of the
aminoacylation of tRNAs. But they have a broad repertoire of functions beyond protein synthesis,
including transcriptional and translational regulation as well as cell signaling [14]. Recently, it has
been demonstrated that two of the tRNA synthetases, human tyrosyl-tRNA synthetase (TyrRS) and
human tryptophanyl-tRNA synthetase (TrpRS), have novel cytokine functions[15]. This
demonstration established a link between protein synthesis and signal transduction. At the same
time,mammalian TyrRS and TrpRS have also been shown to regulate angiogenesis [16-21].
Under apoptotic conditions in culture, full lengthTyrRS is secreted, and two distinct cytokines can
then be generated by an extracellular protease such as leukocyte elastase[15].The NH2-terminal
catalytic fragment, mini-TyrRS, binds strongly to the CXC-chemokine receptor CXCR1 and, like IL-8,
functions as a chemoattractant for polymorphonuclear leukocytes (PMNs), to promote
angiogenesis[15], whereas the full-length enzyme lacks cytokine activity.The catalytic core domain of
TrpRS is a close homologue of the catalytic domain of TyrRS [22-24]. In normal cells, human TrpRS
exists as two forms. The major form is the full-length protein, and the other is a truncated TrpRS
(mini-TrpRS) in which most of the extra NH2-terminal domain is deleted because of alternative
splicing of the pre-mRNA[25-26], with Met-48 being deduced as the NH 2 -terminal residue of mini-
TrpRS [25]. PMN elastase digestion of recombinant full-length TrpRS produced two fragments
designated T 1 -TrpRS and T 2 -TrpRS, respectively.These fragments were similar in size to mini-TrpRS.
In addition, mini-TrpRS and T2-TrpRS blocked vascular endothelial growth factor-stimulated
angiogenesis in both chick cell adhesion molecule and mouse matrigel assays in vivo [18-19]. The
full-length enzyme lacks cytokine activities. The construction and relationship of TyrRS, TrpRS, and
their variants is shown in Fig. 1.
Fig 1. Schematic representation of human TyrRS, TrpRS and their variant constructs used in this study.
Shaded regions of full-length TyrRS and TrpRS represent COOH- and NH 2 -terminal appended
domains, respectively. Numbers on the left and right correspond to the NH 2 - and COOH-terminal
residues relative to the human full-length enzymes, respectively.
Despite these intriguing in vitro actions, no studies have examined the mini-TyrRS and mini-
TrpRS in physiological or pathological settings in vivo, especially in myocardial ischemia condition.
Therefore, the purpose of this study was to determine whether exogenous mini-TyrRS augments
angiogenesis while mini-TrpRS inhibited, and to preliminary comprehend their angiogenesis
mechanism. All of this may be helpful for healing ischemia diseases, such as CAHD.
2. Methods
38

2.1 Materials
Sprague-Dauley(SD) male rats,(250-300g, 2-3 month old) were provided from experimental
animal center of Sichuan university.(Sichuan, China). Human recombinant mini-TyrRS and mini-
TrpRS were from aTyr Pharma (La Jolla, CA).Ⅷ factor related antigen and antibody was from
Boster(Wuhan, china). Immunohistochemisty staining kit was purchased from Zhongshan
Goldenbridge (Beijing, China). Revert Aid First Strand cDNA Synthesis Kit was purchased from
MBI Company (Lithuania).
2.2.Ethics
All animal procedures were conducted with prior institutional ethical approval under the
requirements of the Chinese Prevention of Cruelty to Animals Act and the Code of Practice for the
Care and Use of Animals for Scientific Purposes. Prior clearance was obtained from the Animal
Experimentation Ethics Committees of West china Medical Centre and Institutes of Animal Science.
The animals of this study were inspected by members of the West china Medical Centre Animal Ethics
Committee.
2.3.Left anterior descending (LAD) artery ligation
Animals were anaesthetized with 100g/L chloral hydrate (0.2ml per 100g body weight injected
intraperitoneally).The analgesic buprenorphine was given preoperatively(10-20 µg/kg body
weight).Occlusion of the left anterior descending coronary artery(LAD) was performed as previously
reported[11],under sterilec conditions, with minor alterations. Briefly,the anaesthetized animal was
incubated endotracheally in a supine position, and ventilated with a Harvard Mouse Mini-
Vent(Harvard apparatus, Marchhugstetten,Germany),which supplied 0.2-0.25ml room air 120 times
per minute. The animal was moved onto its right side, and a left thoracotomy in the third intercostal
space was provided access to the beating heart. After removing the pericardium,the LAD was
visualized with a stereomicroscope (Leica MZ6,Heerbrugg,Switzerland),and occluded with 8/0
prolene suture. The suture position of the LAD coronary artery was 0.3mm distal to the
atrioventricular junction. Occlusion was confirmed by observation of left ventricular pallor
immediately post ligation and an electrocardiogram was used to observe changes such as widening of
QRS and ST-T segment elevation. The chest was closed, the lungs re-inflate and the animal moved to
a prone position until spontaneous breathing occurred. Animal were monitored closely for signs of
infection at the surgical site; none were observed in any animals.
2.4.Experimental groups
A total of 80 rats(200-250g) were divided into four different groups(n=20 per each group, and
n=5 for each time point).⑴sham group: animals underwent a thoracotomy with removal of the
pericardium, but no coronary artery ligation (CAL). No suture was placed in the sham animals’heart,
in order to avoid unintended vessel damage or occlusion; ⑵ CAL group, but no mini-TyrRS/mini-
TrpRS-siRNA injection;⑶: CAL+mini-TyrRS(20μl, twice daily, 600μg. Kg -1 .day -1 ) ⑷: CAL+mini-
TrpRS(20μl, twice daily, 600μg. Kg -1 .day -1 ). Mini-TyrRS or mini-TrpRS were administrated by
coronary artery polyvinyl catheter injection to rats.
2.5.Histologic expression
All rats were killed after ligation 3rd, 7th, 14th, and 28th day ,the heart was removed and fixed in
fresh 4% paraformaldehyde, pH 7.4.The tissue was processed and embedded in paraffin using routine
histological procedures. Five micrometre transverse step sections were collected every 200µm
through the entire ventricle(approximately 10-12 sections per animal),and stained with Haematoxylin
and Eosin(HE).the cells were observed with an inverted phase-contrast microscope (Olympus, Japan)
and photographed. The total tube area were analyzed in 3 different fields at ×400 magnification
2.6.Measurement of irreversible ischemic injury
After ligation 3rd, 7th, 14th, and 28th day, 1% of Evans blue solution (5 mL),was infused into the
abdominal vena cava to delineate the ischemic area at risk of the left ventricle. The heart was excised
and cross-sectioned from the apex to the atrioventricular groove into four specimens of 0.8mm in
thickness with the use of a stereoscope. The two middle heart slices were incubated in 2,3,5-
39

triphenyltetrazolium chloride (TTC) solution (1%) for 30 min in phosphate buffer at 37℃. Sections
were fixed overnight in 4% paraformaldehyde for contrast enhancement between stained and
unstained tissue. TTC stained the viable tissue with red, while the necrotic tissue remained discolored.
The sections were then placed between two cover slips and digitally photographed using a Nikon
coolpix S10 camera, and quantified with the weight respectively. The area of irreversible injury (TTC
negative) is presented as a percentage of the area(the irreversible injury area/the total weight of
ventricles).
2.7.Measured the density of capillary
The streptavidin peroxidase (SP) immunohistochemical method was used to detect the expression
of factor VIII in myocardial infarction margin areas. The dilution of factor VIII rat monoclonal
antibody (Santa Cruz, CA, USA) was 1:100. The procedure was performed according to the
manufacturer’s instructions.The positive cells were identified, counted and analyzed under the
inverted phase-contrast microscope (Olympus, Japan) in 3 different fields(0.1mm 2 ) at 400
magnification using Image-proplus 6.0 software.Results were normalized by arbitrarily setting the
total tube area of control to 100%.
2.8. Real time fluorescent quantitation PCR (RT-PCR)
After ligation, myocardial tissues were harvested at the times indicated, washed twice with icecold
phosphate-buffered saline (PBS) and collected by centrifugation. Total RNA was isolated using
the Trizol reagent (MRC, USA) according to the manufacturer's instructions. Total RNA (5 µL) was
converted to complementary DNA (cDNA) using Revert Aid First Strand cDNA Synthesis Kit. A
5µL aliquot of the resulting cDNA was used as template for PCR amplification with the following
primers: mini-TrpRS, P1 (forward, 5’- CCC TGC TGC ACT CCA CCT T-3’), P2 (reverse, 5’- ACG
CAT GCT TAT TGA CCT TG-3’); mini- TyrRS, P3 (forward, 5’- CAT CTG ATG AAT CCT ATG
GTT -3’), P4 (reverse, 5’- GGA TCA CAA ACT CGG ACT TA-3’); β-actin, P5 (forward, 5’-GCC
AAC ACA GTG CTG TCT -3’), P6 (reverse, 5’-AGG AGC AAT GAT CTT GAT CTT -3’). The
amplifications were performed by an initial denaturation (94 °C for 2 min), followed by 45 cycles of
denaturation, annealing and extension (94 °C for 20 s, 54 °C for 20 s, 72 °C for 30 s), and a final
extension (72 °C for 5 min). The transcript of β-actin was also amplified by RT-PCR from the same
cDNA template and was used as an internal control. All the primers were designed and synthesized by
Genepharma (Shanghai, China). The identity of each PCR product was confirmed by DNA sequencing.
The pictures were scanned and analyzed by the Gel Doc 1000 gel imaging system.
2.9.Statistical analysis
Results are expressed as mean ± standard deviation. Comparison of means was performed by
means of the analysis of variance procedure (Student–Newman–keuls test, SPSS 13.0 for Windows).
P

Fig. 2. The representative photomicrographs of neoformative blood capillaries in infarcted LV area,
stained with hematoxylin and eosin. Black arrowheads point to newly formed blood vessels and
neoformative blood capillary at different time point. Sham, sham operation; CAL, coronary artery
ligation; CAL + mini-TyrRS, mini-TyrRS (20μl, twice daily, 600μg. Kg -1 .day -1 ); CAL + mini-TrpRS,
mini-TrpRS (20μl, twice daily, 600μg. Kg -1 .day -1 ).
3.2.Measurement of irreversible ischemic injury(TTC staining)
No myocardial infarction was found in the sham group, , but obvious infarction areas were seen in
other groups. Compared with CAL group (9.3%,15.0%,30.9%,39.3%),the myocardial infarction size
of mini-TyrRS group at 3rd, 7th, 14th, and 28th day were respectively 8.0%,10.9%,22.4%,28.1%, as
for mini-TrpRS group, were 10.7%,18.8%,37.5%,51.2%, but only in day 14th and 28th has significant
difference, p

Fig 3. Mini-TyrRS-siRNA / mini-TrpRS-siRNA transfected TTC stain after myocardial infarction.
Black arrowheads point to myocardial infarction zone. Sham, sham operation; CAL, coronary artery
ligation; CAL + mini-TyrRS, mini-TyrRS (20μl, twice daily, 600μg. Kg -1 .day -1 );CAL + mini-TrpRS,
mini-TrpRS (20μl, twice daily, 600μg. Kg -1 .day -1 ). No myocardial infarction was found in the sham
group, but more obvious infarction areas were seen in other groups. Mini-TrpRS group had the most
myocardial infarction areas, while in mini-TyrRS group were present the least .
Fig 4 Measurement of irreversible ischemic injury area of mini-TyrRS /mini-TrpRS injected.
Compared with CAL, the myocardial infarction area was decreased in mini-TyrRS group, while
increased in mini-TrpRS group, but only in day 14 and 28 considered statistically significant
difference .Data are shown as the means±S.D. (n = 5 for each time point, each group). #P < 0.05, vs.
CAL group. No any infarction size were seen in sham.
3.3.Measured the density of capillary(SP method)
The streptavidin peroxidase (SP) immunohistochemical method was used to detect the expression
of factor VIII in myocardial infarction areas. The density of capillary of mini-TyrRS group at 3rd, 7th,
14th, and 28th day were respectively 5.5 mount/0.1mm 2 ,8.6 mount/0.1mm 2 , 11.7mount/0.1mm 2 ,13.4
42

mount/0.1mm 2 , as for mini-TrpRS group, were 1.2 mount/0.1mm 2 ,1.6 mount/0.1mm 2 , 1.9
mount/0.1mm 2 ,1.8 mount/0.1mm 2 , has significant difference compared with CAL group (3.0
mount/0.1mm 2 ,5.6 mount/0.1mm 2 , 6.7mount/0.1mm 2 ,5.5 mount/0.1mm 2 ), p

Fig.6 Angiogenesis (capillary density, measured at day 3, 7, 14 and 28 after ligation) was inhibited by
mini-TrpRS and augmented by mini-TyrRS, while sham had no angiogenesis effect compared with
CAL. Data are shown as the means±S.D. (n = 5 for each time point, each group). #P < 0.05, ##P <
0.01 vs. CAL group.
Fig.7 Mini-TyrRS mRNA expression in infarcted left ventricular tissue in CAL rats at different time
points. (A) mRNA of mini-TyrRS at 3 days;(B) mRNA of mini-TyrRS at 7 days; (C) mRNA of mini-
TyrRS at 14 days; (D) mRNA of mini-TyrRS at 28 days. Sham, sham-operation; CAL, coronary artery
ligation; CAL + mini-TyrRS (20μl, twice daily, 600μg. Kg -1 .day -1 ). Lane 1, Sham; Lane 2, CAL;
Lane 3, CAL + mini-TyrRS. Data are shown as the means±S.D. (n = 5 for each time point, each
group). * P < 0.05,**P < 0.01 vs. sham operation group; #P < 0.05, ##P < 0.01 vs. CAL group.
44

3.4.The mRNA expression of mini-TyrRS / mini-TrpRS at infarction marginal zone
RT-PCR was used to detect the mRNA expression of mini-TyrRS/mini-TrpRS at infarction zone.
The results indicated that: Compared with sham group, the expression level of CAL was increase and
had significantly different in mini-TyrRS group(P < 0.05 or P < 0.01), after administration of mini-
TyrRS, there was a significant increase in expression level compared with the CAL group and sham
group.(P < 0.05 or P < 0.01).But the expression level between sham-operated and CAL groups was
not significantly different in mini-TrpRS group(p>0.05), after administration of mini-TrpRS, there was
a significant increase in expression level compared with the CAL group and sham group.(P < 0.05 or
P < 0.01).see Fig.7-8.
Fig.8 Mini-TrpRS mRNA expression in infarcted left ventricular tissue in CAL rats at different time
points. (A) mRNA of mini- TrpRS at 3 days;(B) mRNA of mini- TrpRS at 7 days; (C) mRNA of mini-
TrpRS at 14 days; (D) mRNA of mini- TrpRS at 28 days. Sham, sham-operation; CAL, coronary
artery ligation; CAL + mini- TrpRS (20μl, twice daily, 600μg. Kg -1 .day -1 ). Lane 1, Sham; Lane 2,
CAL; Lane 3, CAL + mini- TrpRS. Data are shown as the means±S.D. (n = 5 for each time point). * P
< 0.05,**P < 0.01 vs. sham operation group; #P < 0.05, ##P < 0.01 vs. CAL group.
4. Discussion
Preclinical models of myocardial ischaemia have been reported in several large animals species,
including dogs[27-28]and goats [29].The model that most closely resembles the response seen in
humans is the pig ameroid model[30],which has been used in a variety of therapeutic studies[31-
39].However, the expense and practical demands of porcine surgical facilities severely limit the extent
of such studies,precluding the use of this model for large scale screening studies of novel therapeutic
approaches.
In normal cells, human TrpRS exists as a full length form and as a truncated form designated mini-
TrpRS, which is produced by alternative splicing[40]. Expression of mini-TrpRS is highly stimulated
in human cells by the addition of IFN-c [41-43]. Although both human full-length TrpRS and mini-
TrpRS are enzymatically active in aminoacylation, they differ in angiostatic activity[23; 44]. The same
45

phenomenon is also present between full-length TyrRS and mini-TyrRS. Researches have found a
kind of angiogenesis regulatory factor, CXC chemotactic factor, but their function was dependent on
whether or not they have the Glu-Leu-Arg (ELR) motif [45]. The CXC chemotactic factors which
have the ELR motif could promote angiogenesis, while CXC chemotactic factors which do not have
the ELR motif would be the antagonistic factors of angiogenisis. The situation is analogous to human
mini-TyrRS and mini-TrpRS. Mini-TyrRS has an ELR motif and promotes angiogenesis, but mini-
TrpRS does not have the ELR motif. Therefore, their function was totally opposite[46].
We were aware that angiogenic and angiostatic factors may work together to regulate angiogenesis.
To test our hypothesis that mini-TyrRS (containing a natural ELR sequence) is an angiogenic factor,
and that mini-TrpRS(containing a natural ELQ sequence) is an angiostatic factor, we evaluated
whether mini-TyrRS/mini-TrpRS induced myocardial cells to format new capillaries using rats
myocardial infarction model. We found out that irreversible ischemic injury areas was decreased and
new capillaries formation was increased after mini-TyrRS administration by using TTC staining(Fig.3-
4) and the SP immunohistochemical method(Fig.5-6), but in mini-TrpRS group, myocardial infarction
zone was increased and new capillaries formation was decreased after administration. In contrast,
angiogenesis was not observed with full-length TyrRS and TrpRS. Interestingly, angiogenesis is
stimulated by either mini-TyrRS or is inhibited by mini-TrpRS in rats myocardial infarction models,
raising the possibility that mini-TyrRS/mini-TrpRS stimulates a common downstream signaling event.
Thus, naturally occurring fragments of the two proteins involved in translation, TyrRS and TrpRS,
have opposing activity on angiogenesis in the rat model. The opposing activities of the two tRNA
synthetases suggest tight regulation of the balance between pro- and antiangiogenic stimuli, but the
exact mechanism and relationship between them has not been clearly demonstrated. More research
should be done to further explore the mechanism.
References:
1.Losordo DW, Vale PR, Symes JF, Dunnington CH, Esakof DD, Maysky M, Ashare,AB Lathi K,
Isner JM. Gene therapy for myocardial angiogenesis ;initial clinical results with direct myocardial
injection of phVEGF165 as sole therapy for myocardial ischemia. Circulation 1998;98:2800-4.
2.Laham RJ, Sellke FW, Edelman JD, Pearlman JD, Ware A, Brown DL, Gold JP , Simons M. Local
perivascular delivery of basic fibroblast growth factor in patients undergoing coronary bypass
surgery :results of a phase I randomized, double-blind, placebo-controlled trial. Circulation
1999;100:1865-71.
3.Rosengart TK, Lee LY, Patel SR, Kligfield PD, Okin PM, Hackett NR, Isom OW , Crystal RG. Sixmonth
assessment of a phase I trial of angiogenic gene therapy for the treatment of coronary artery
disease using direct intramyocardial administration of an adenovirus vector expressing the
VEGF121 cDNA. Ann Surg 1999;230,466-72.
4.Rosengart TK, Lee LY, Patel SR, Sanborn TA, Parikh M, Bergman GW, Hachamovitch R, Szulc M,
Kligfield PD, Okin PM, Hahn RT, Devereux RB, Post MR, Hackett NR, Foster T, Grasso TM,
Lesser ML, Isom OW, Crystal RG. Phase I assessment of direct intramyocardial administration of
an adenovirus vector expressing the VEGF121 cDNA to individuals with clinically significant
severe coronary artery disease. Circulation 1999;100:468-74.
5.Symes JF, Losordo DW, Vale PR, Lathi KG, Esakof DD, Maysky M, Isner JM. Gene therapy with
vascular endothelial growth factor for inoperable coronary artery disease. Ann Thoracic Surg
1999;68:830-7.
6.Vale PR, Losordo DW, Milliken CE, McDonald MC, Gravelin LM, Curry CM, Esakof DD, Maysky
M, Symes JF ,Isner JM. Randomized, singe-blind, placebo-controlled pilot study of catheter-based
myocardiol gene transfer for therapeutic angiogenesis using left ventricular electromechanical
mapping in patients with chronic myocardial ischemia. Circulation 2001;103:2138-43.
7.Pfeffer MA, Pfeffer JM, Fishbein MC, Fletcher PJ, Spadoro J, Kloner RA, Braunwald E. Myocardial
infarct size and ventricular function in rats. Circulation Res 1979 ;44:503-12.
8.Michael LH, Entman ML, Hartley CJ, Youker KA, Zhu J, Hall SR, Hawkins HK, Berens K,
Ballantyne CM. Myocardial ischemia and reperfusion : a murine model. Am J Physiol
1995;269:2147-54.
46

9.Du XJ, Gao X, Ramsey D. Surgical methods of inducing transverse aortic stenosis and myocardial
infarction in the mouse. Asia Pacific Heart 1998;7:187-92.
10.Lutgens E, Daemen MJAP, de Muinck ED, Debet J, Leenders P, Smits JFM. Chronic myocardial
infarction in the mouse :cardiac structural and functional changes. Cardiovascular Res
1999;41:586-93.
11.Michael LH, Ballantyne CM, Zachariah JP, Gould KE, Pocius JS, Taffet GE, Hartley CJ, Pham TT,
Daniel SL, Funk E, Entman ML. Myocardial infarction and remodeling in mice. Effect of
reperfusion. Am J Physiol 1999;277:660-8.
12.Scherrer-Crosbie M, Steudel W, Ullrich R, Hunziker PR, Liel-Cohen N, Newell J, Zaroff J, Zapol
WM, Picard MH. Echocardiographic determination of risk area size in a murine model of
myocardial ischemia. Am J Physiol 1999;277:986-92
13.Guo Y, Wu WJ, Qin Y ,Tang XL, Yang Z, Bolli R.Demonstration of an early and a late phase of
ischemic preconditioning in mice. Am J Physiol 1998;275:1375-87.
14. Martinis SA, Plateau P, Cavarelli J, Florentz C .AminoacyltRNA synthetases: a family of
expanding functions. EMBO J 1999;18:4591–6.
15.Wakasugi K, Schimmel P.Two distinct cytokines released from a human aminoacyl-tRNA
synthetase. Science 1999;284:147–51.
16.Wakasugi K, Schimmel P.Highly differentiated motifs responsible for two cytokine activities of a
split human tRNA synthetase. J Biol Chem 1999;274:23155–9.
17.Wakasugi K, Slike BM, Hood J, Ewalt KL, Cheresh DA, Schimmel P. Introduction of angiogenesis
by a fragment of human tyrosyl-tRNA synthetase. J Biol Chem 2002;277:20124–6.
18.Wakasugi K, Slike BM, Hood J, Otani A, Ewalt KL, Friedlander M, Cheresh DA, Schimmel P.A
human aminoacyl-tRNA synthetase as a regulator of angiogenesis. Proc Natl Acad Sci
2002;99:173–7.
19.Otani A, Slike BM, Dorrell M, Hood J, Kinder K, Ewalt KL, Cheresh D, Schimmel P, Friedlander
M.A fragment of human TrpRS as a potent antagonist of ocular angiogenesis.Proc Natl Acad Sci
USA 2002;99:178–83.
20.Ewalt KL, Schimmel P.Activation of angiogenic signaling pathways by two human tRNA
synthetases. Biochemistry 2002;41:13344–9.
21.Ibba M.Aminoacyl-tRNA synthesis. Annu Rev Biochem 2000;69:617–50.
22.Ribas de Pouplana L, Frugier M, Quinn CL, Schimmel P.Evidence that two present-day
components needed for the genetic code appeared after nucleated cells separated from eubacteria.
Proc Natl Acad Sci USA 1996;93:166–70.
23.Lee SW, Cho BH, Park SG, Kim S.Aminoacyl-tRNA synthetase complexes: beyond translation. J
Cell Sci 2004;117:3725–34.
24.Fleckner J, Martensen PM, Tolstrup AB, Kjeldgaard NO ,Justesen, J.Differential regulation of the
human interferon inducible tryptophanyl-tRNA synthetase by various cytokines in cell lines.
Cytokine 1995;7:70–7.
25.Kise Y, Lee SW, Park SG, Fukai S, Sengoku T, Ishii R, Yokoyama S, Kim S, Nureki O.A short
peptide insertion crucial for angiostatic activity of human tryptophanyl-tRNA synthetase. Nat
Struct Mol Biol 2004;11:149–56.
26.Jorgensen R, Sogaard TM, Rossing AB, Martensen PM, Justesen J.Identification and
characterization of human mitochondrial tryptophanyl-tRNA synthetase. J Biol Chem
2000;275:16820–6.
27.Schaper W, De Bradbander M, Lewi P. DNA synthesis and mitoses in coronary collateral vessels
of the dog. Circulation Res 1970;28:671-9.
28.Shou M, Thirumurti V,Sharnini Rajanayagam MA, Lazarous DF, Hodge E, Stiber JA, Pettiford M,
Elliot E, Shah SM, Unger EF. Effect of basic fibroblast growth factor on myocardial angiogenesis
in dogs with mature collateral vessels.J Am College Cardiogl 1997;29:1102-6.
29.Mannion JD, Blood V, Bailey W ,Bauer TL, Magno MG, Dimeo F,Epple A, Spinale FG.The effect
of basic fibroblast growth factor on the blood flow and morphologic features of a latissimus dorsi
cardiomyoplasty. J Thoracic Cardiovascular Surg 1996;111:19-28.
30.Ware JA, Simons M. Angiogenesis and cardiovascular disease. Oxford university press, 1999,New
York.
47

31.Giordano FJ, Ping P, McKirnan D,Nozaki S, DeMaria AN, Dillmann WH, Mathieu CO, Hammond
K. Intracoronary gene transfer of fibroblast growth factor-5 increases blood flow and contractile
function in an ischemic region of the heart. Nature Med 1996;2:534-9.
32.Hariawala MD, Horowitz JR, Esakof D, Sheriff DD, Walter DH, Keyt B, Isner J, Symes JF.VEGF
improves myocardial blood flow but produces EDRF-Mediated hypotension in porcine hearts. J
Surg Res 1996;63:77-82.
33.Lopez JJ, Laham RJ, Stamler A, Pearlman JD, Bunting S, Kaplan A, Carrozza JP, Sellke
FW.VEGF administration in chronic myocardial ischemia in pigs. Cardiovascular Res
1998;40:272-81.
34.Mack CA, Patel SR, Schwarz EA, Zanzonico P, Hahn RT, Ilercil A, Devereux RB, Goldsmith SJ,
Christian TF, Sanborn TA, Kovesdi I, Hackett N, Isom OW, Crystal RG, Rosengart TK. Biologic
bypass with the use of adenovirus-mediated gene transfer of the complementary deoxyribonucleic
acid for vascular endothelial growth factor 121 improves myocardial perfusion and function in the
ischemic porcine heart. J Thoracic Cardiovascular Surg 1998;115:168-77.
35.Tio RA, Tkebuchava T, Scheuermann TH, Lebherz C, Magner M, Kearny M, Esakof DD, Isner JM,
Symes JF. Intramyocardial gene therapy with naked DNA encoding vascular endothelial growth
factor improves collateral flow to ischemic myocardium. Huamn Gene Ther 1999;10:2953-60.
36.Patel SR, Lee LY, Mack CA,Polce DR, El-Sawy T, Hackett NR, Ilercil A, Jones EC, Hahn RT,
Isom OW, Rosengart TK, Crystal RG.Safety of direct myocardial administration of an adenovirus
vector encoding vascular endothelial growth factor 121 .Human Gene Ther 1999;10:1331-48.
37.Laham RJ, Rezaee M, Post M, Novicki D, Sellke FW, Pearlman JD, Simons M, Hung
D.Intrapericardial delivery of fibroblast growth factor-2 induces neovascularization in a porcine
model of chronic myocardial ischaemia. J Pharmacol Exp Therapeutics 2000;292:795-802.
38.Sato K, Laham RJ, Pearlman JD, Novicki D, Sellke FW, Simons M, Post MJ. Efficacy of
intracoronary versus intravenous FGF-2 in a pig model of chronic myocardial ischemia. Ann
Thoracic Surg 2000;70:2113-8.
39.Sato K, Wu T, Laham RJ, Johnson RB, Douglas P, Li J, Sellke FW, Bunting S, Simons M, Post MJ.
Efficacy of intracoronary or intravenous VEGF in a pig model of chronic myocardial ischemia. J
Am College Cardiol 2001;37:616-23.
40.Yeh CH, Peng HC, Huang TF.Accutin, a new disintegrin,inhibits angiogenesis in vitro and in vivo
by acting as integrin alphavbeta-antagonist and inducing apoptosisl.J Blood 1998;92:3268–76
41.Tolstrup AB, Bejder A, Fleckner J, Justesen J.Transcriptional regulation of the interferon-c
inducible tryptophanyltRNA synthetase includes alternative splicing. J Biol Chem 1995;270:397–
403.
42.Fleckner J, Rasmussen HH, Justesen J.Human interferon gamma potently induces the synthesis of
a 552 kDa protein highly homologous to rabbit peptide chain release factor and bovine tryptophan.
Proc Natl Acad Sci 1991;88:11520–4.
43.Shaw AC, Rossel LM, Roepstorff P, Justesen J, Christiansen G, Birkelund S. Mapping and
identification of interferon gamma-regulated HeLa cell proteins separated by immobilized pH
gradient two dimensional gel electrophoresis. Electrophoresis 1999;20:984–93.
44.Tzima E, Reader JS, Irani-Tehrani M, Ewalt KL, Schwartz MA, Schimmel P. VE-cadherin links
tRNA synthetase cytokine to anti-angiogenic function. J Biol Chem 2005;280:2405–8
45. Clark LI, Dewald B, Geiser T ,Moser B, Baggiolini M.Platelet factor 4 binds to interleukin 8
receptors and activates neutrophils when its N terminus is modified with Glu-Leu-Arg. Proc Natl
Acad Sci 1993;90:3574–7.
46.Jeong EJ, Hwang GS, Kim KH, Kim MJ, Kim S, Kim KS.Structural analysis of multi-functional
peptide motifs present in human bifunctional tRNA synthetase: identification of RNA binding
residues and functional implications for tandem repeats. Biochemistry 2000;39:15775–82.
48

Distribution of three forms alpha-melanocyte stimulating hormone in the
brain of male and female sheep by using HPLC and Mass Spectrometry
Analysis
Rui Zeng 1,2 * , A.Ian.Smith 2 , Iain.J.Clarke 3
* Presenter
1. Department of Cardiovascular diseases, West China Hospital, School of Clinic Medicine,
Sichuan University, Chengdu 610041, China
2. Department of Biochemistry & Molecular Biology, Faculty of Medicine, Nursing and
Health Sciences, Monash University, Vic 3800,Australia
3. Department of Physiology, Faculty of Medicine, Nursing and Health Sciences, Monash
University, Vic 3800, Australia
To compare des-acetyl α-MSH(des-α-MSH), acetyl α-MSH(act-α-MSH) and di-acetyl α-
MSH(di-act-α-MSH) between the male and female sheep in ten special brain tissues. All of
this will let us ascertain whether different gonadal steroids causes any change in acetylated
peptide. Mass spectrometry (MS) was used for molecular weight determination, HPLC/EIA
was also used for quantitation. Before MS, Immunomagnetic protein isolation using
Dynabeads Protein A also provides a fast and reliable method for capturing Ig for small scale
purification. Mass spectrometry analysis indicated that: In female sheep, all three forms of α-
MSH were detected in hypothalamus, neurointermediate lobe, cortex, pre-optic area and
median eminence. In arcuate nucleus, des- and act-α-MSH were present, and only acetylated
forms(including mono-act and di-act) were detected in anterior pituitary. For male sheep, only
in neurointermediate lobe and pre-optic area were present all three forms of α-MSH. In
hypothalamus, median eminence, cortex, anterior pituitary, des- and act-α-MSH were detected.
We can only find des-α-MSH in arcuate nucleus. Cerebellum, spinal cord and brain stem have
no any forms of α-MSH both in male and female sheep. HPLC/EIA results showed that in
hypothalamus, neurointermediate lobe, pre-optic area and median eminence, there is no any
difference in des-α-MSH between male and female sheep, but female sheep has more act-α-
MSH and di-act-α-MSH compared with male sheep.P < 0.05. In arcuate nucleus, there is no
any difference in des-α-MSH between male and female sheep, but female sheep has more actα-MSH
only compared with male sheep.P < 0.05. In anterior pituitary, there is more des-α-
MSH in male sheep, but female sheep has more act-α-MSH and di-act-α-MSH compared with
male sheep. P < 0.05. Finally, in cerebral cortex, female sheep has more des-α-MSH, act-α-
MSH and di-act-α-MSH compared with male sheep.P < 0.05. These data suggest that N-
acetylation of products formed by the processing of α-MSH can markedly adapt to their
biological properties. The results also illustrated that the distribution of α-MSH may be
affected by the sex steroid hormone.
Keywords
α-MSH, POMC, sheep, HPLC, Mass Spectrometry
Introduction
Proopiomelanocortin (POMC) is a precursor protein that contains the sequence for several
bioactive peptides including adrenocorticotropin (ACTH), β-endorphin(β-EP), and melanocytestimulating
hormone (α-, β- and γ- MSH). The functional peptides have been detected in a variety of
tissues.
Post-translational processing of prohormones is an important biochemical mechanism employed
by many secretory cells to generate a diversity of biological peptides from the same gene product, a
process that occurs in a tissue-specific manner (1). Similarly to many prohormones, POMC is
49

subjected to sequence-specific cleavages to generate bioactive products by the action of the
prohormone convertases (PCs) 1 and 2 (2). Thus far, the processing of POMC has been described in
the pituitary and in the arcuate nucleus (ARC) of the hypothalamus. In the ARC, similar to the pars
intermedia of the pituitary (3-6), POMC is initially cleaved by PC1 to generate proadrenocorticotrophin
(pro-ACTH) and β-lipotrophin. Pro-ACTH is further cleaved by PC1 to generate
a 16 kDa N-terminal peptide and ACTH. ACTH is further cleave to generate ACTH (1-17) and
corticotrophin like intermediate lobe peptide (CLIP). Then, carboxypeptidase E enzyme removes C-
terminal basic amino acids from ACTH (1-17) , and the peptidyl α-amidating mono-oxigenase (PAM)
enzyme amidates the peptide to generate des-acetyl-α-melanocyte stimulating hormone (des-α-MSH).
Finally, an N-acetyl-tranferase enzyme converts des-α-MSH to acetyl-α-MSH (act-α-MSH) and
double- acetyl-α-MSH (di-act-α-MSH)(7-9). POMC (10,11) is also generated in the commissural part
of the nucleus of the solitary tract (NTS). The amount of act-α-MSH in the rodent brain is
controversial, because some studies report undetectable levels of the N-acetylated form in the
hypothalamus (12-14). On the other hand, it is well established that synthetic act-α-MSH is far more
potent than des-α-MSH in reducing food intake when administered centrally in rats (15,16). However,
the fate of its processing and mechanism underlying the different potencies of the two forms of α-
MSH is poorly understood.
The amino acid sequence is the foundation of a protein, and any degradation or modification of a
protein’s primary structure may affect its biological properties and functions. Numerous conventional
analytical techniques have been applied to the determination of the primary structure of therapeutic
proteins.To investigate post-translational processing, we needs to know each product including
acetylated products. Usually methods is high-performance liquid chromatography (HPLC) was used
for protein separation and purity analysis first, and then enzyme immunoassay(EIA) for quantitation,
but the disadvantage of this method was that it would cost at least two days to acquire the results.
Another choice was mass spectrometry (MS), using by molecular weight determination, for its fast
(only several seconds you can obtain all the results), accurate and reliable, except that it could not be
used for quantitation. The amino acid sequence information and molecular weight for the three α-MSH
were seen in Table.1.
Table 1 Amino acid sequence information and molecular weight of three α-MSH
Peptide
MW
Amino acids
Des - Ac - α-MSH (sheep) 1622.4 13
Sequence(One-Letter Code): SYSMEHFRWGKPV-NH2
Sequence(Three-Letter Code): H - Ser - Tyr - Ser - Met - Glu - His - Phe - Arg -
Trp - Gly - Lys - Pro - Val - NH2
Ac -α-MSH (sheep) 1664.9 13
Sequence(One-Letter Code): Ac-SYSMEHFRWGKPV-NH2
Sequence(Three-Letter Code): Ac - Ser - Tyr - Ser - Met - Glu - His - Phe - Arg -
Trp - Gly - Lys - Pro - Val - NH2
Di - Ac-α-MSH (sheep) 1706.8 13
Sequence(One-Letter Code): Di-ac-SYSMEHFRWGKPV-NH2
Sequence(Three-Letter Code): Di-ac - Ser - Tyr - Ser - Met - Glu - His - Phe -
Arg - Trp - Gly - Lys - Pro - Val - NH2
The purpose of this study was therefore threefold: (i)to determine levels of three forms α-MSH in
10 different brain tissues extracts;(ii)to compare the different type of α-MSH between the male and
female sheep;(iii) to compare the results determined by HPLC/EIA and mass spectrometry analysis.
All of this will let us ascertain whether different gonadal steroids causes any change in acetylated
peptide.
50

Materials and Methods
Ethics
All animal procedures were conducted with prior institutional ethical approval under the requirements
of the Australian Prevention of Cruelty to Animals Act 1986 and the Code of Practice for the Care and
Use of Animals for Scientific Purposes. Prior clearance was obtained from the Animal
Experimentation Ethics Committees of Monash Medical Centre and Victorian Institutes of Animal
Science. The animals of this study were inspected by members of the Monash Medical Centre Animal
Ethics Committee.
Animals
Adult Corriedale male sheep(55kg,n=3) and female sheep(53kg, n=3) were maintained on pasture or
in feedlots under natural conditions. For intensive experimentation, the animals were housed in
individual pens with natural lighting and temperature and were fed ad libitum lucerne chaff with free
access to water. Before experiment, animals were conditioned to pen-housing and handling for 1 week.
Animals in this experiment were euthanised by overdose of 20 ml sodium pentobarbital (Lethabarb;
Virbac, Peakhurst, NSW, Australia) iv and the brains were removed, then Hypothalamus , Anterior
pituitary gland (A.P), Cerebellum (CEREB), Arcuate nucleus (ARC), Neurointermediate lobe (NIL),
Median eminence (M.E), Cerebral cortex, Spinal cord, Brain stem(B/S), and Pre-optic area(POA)
were dissected. All tissues were frozen on dry ice and stored at -80°C.
Peptide Extraction
Tissues were immediately immersed in 2N acetic acid (each 600 μl for 300mg sample) and boiled for
10 min. The samples were then homogenized by using electric homogenisers. Vortex, stand on ice
while homogenizing the rest of the samples, and once completed leave on ice for at least half an hour.
Finally, centrifuged at 6,000rpm at 4°C for 30 min. Supernatants were collected, and re-constituted in
distilled water.
Experiment 1: Mass spectrometry analysis of α-MSH in ten brain tissues
Co-immunoprecipitation (Peptide purification)
Polyclonal α-MSH-antibody (1:1000; Bachem, Bubendorf, Switzerland) was cross-linked to
Dynabeads protein A (Dynal Biotech, Olso, Norway) according to the manufacturer’s protocol.
Tissues lysates were precleared with IgG Dynabeads-protein A for 10 min and then incubated with α-
MSH-Dynabeads overnight at 4 °C, then α-MSH-immunoprecipitated complexes were washed five
times with immunoprecipitation buffer (10mM Tris/HCl, pH 7.8,1mM EDTA,150mM NaCl,1mM
NaF,0.5%NonidetP-40,0.5% glucopyranoside, 1μg/ml aprotinin, and 0.5 mM phenylmethylsul-fonyl
fluoride).Proteins were eluted by 0.1M citrate buffer (elution buffer, pH2-3 ) and then processed by
Mass Spectrometry(ABI 4700 Proteomics Analyzer ,Applied Biosystems, USA).
Peptide samples for mass spectrometry analysis
Peptide samples were purified and concentrated using a Zip Tip (Millipore) which has C18 resin fixed
at its end. The resin was rinsed according to the manufacturer’s instructions with 10 μl 0.1% TFA and
50% ACN. Peptides were eluted in 10 μl 50% ACN in 0.1 % TFA. A 0.5 μl volume of the
concentrated peptide-containing sample was mixed with a saturated solution of α-cyano-4-
hydroxycinnamic acid (0.5 μl). Each sample (0.5 μl) was spotted on the mass spectrometer sample
plate . The mass spectrometer determines the mass of the peptides and the sequence (by collisioninduced
dissociation). From the masses of the peptide fragments, sequence data were determined by
comparison with known sequences.
Experiment 2: HPLC/Enzyme immunoassay of α-MSH in ten brain tissues
α-MSH Separation
α-MSH peptides were separated by HPLC using a Symmetry C18 (5μm) column (4.6 × 150mm)
(Waters, Milford, USA). The chromatographic run was performed with an aqueous phase containing
0.1% trifluoroacetic acid (TFA) and an organic phase containing 0.085% TFA in acetonitrile(ACN):
methanol (80:20) gradient at a flow rate of 1.0 ml/min. The gradient used for separation was run 52
mins; the total run time was 72 mins .α-MSH peptides eluted between 25-30% acetonitrile: methanol.
51

Fractions (1 ml) were dried and rehydrated with 50μl of EIA buffer (1.42g/l Na 2 HPO 4 , 0.2g/l KH 2 PO 4 ,
5g/l BSA, 1ml/l Tween 20 in water, pH 7.4) and assayed to identify the different forms of α-MSH
using Enzyme immunoassay (EIA).
Enzyme immunoassay
The EIA plates were first coated with Sheep anti-αMSH (1:10000; Bachem, Bubendorf, Switzerland)
in EIA coating buffer (4.3g NaHCO 3 , 5.3g Na 2 CO 3 , in 1L water, pH 9.4) over night at 4°C. The plate
was then washed in EIA wash buffer (1.42g Na 2 HPO 4 ,0.2g KH 2 PO 4 , 1ml Tween 20 in 1L water,
pH7.4) and blocked (8g NaCl, 1.42g KH 2 PO 4 , 0.2g/KCl, 5.0g BSA in 1L water, pH 7.4) for 1h at
room temperature. After 3 washes, Standards (acetylated αMSH , 0.5 – 1024pg/50μl; Bachem,
Torrance, CA)and samples were added at 50μl/well for 90mins at room temperature. After removing
the supernatant, 50μl of biotinylated αMSH (1:100000, Bachem, Torrance, CA) was added to all wells
except blank and non specific binding wells for 90mins. The plate was washed 3 times followed by
50μl of streptavidin-HRP (BD Pharmingen, Franklin Lakes, NJ) for 1h. Finally, after 3 washes, 50μl
tetramethylbenzidine (TMB) (Pierce, Rockford, IL) was added to each well until a strong blue reaction
developed in the total binding wells (30 mins). The reaction was terminated by the addition of 50μl 2N
H 2 SO 4 and the plates were read at 450nm. The sensitivity of this assay was 4pg/100μl.
Statistical analysis
Results are expressed as mean ± standard deviation. Comparison of means was performed by means of
the analysis of variance procedure (Student–Newman–keuls test, SPSS 13.0 for Windows). P

Fig.1 (A) Mass Spectrometry results for male and female sheep in hypothalamus. All three forms of α-
MSH were detected in female sheep, but only des- and act-α-MSH in male sheep. 1622 for des-α-
MSH ; 1664 for act-α-MSH; 1706 for di-act-α-MSH .(B) HPLC/EIA results indicated that there is no
any difference in des-α-MSH between male and female sheep, but female sheep has more act-α-MSH
and di-act-α-MSH compared with male sheep.* P < 0.05, vs. male sheep group.
53

Fig.2 (A) Mass Spectrometry results for male and female sheep in arcuate nucleus. Des- and act-α-
MSH were detected in female sheep, but only des-α-MSH in male sheep. 1622 for des-α-MSH; 1664
for act-α-MSH. (B) HPLC/EIA results indicated that there is no any difference in des-α-MSH between
male and female sheep, but female sheep has more act-α-MSH compared with male sheep.* P < 0.05,
vs. male sheep group.
Fig.3 (A) Mass Spectrometry results for male and female sheep in anterior pituitary. Only acetylated
forms(including mono-act and di-act) were detected in female sheep and des- and act-α-MSH in male
sheep. 1622 for des-α-MSH ; 1664 for act-α-MSH; 1706 for di-act-α-MSH . (B) HPLC/EIA results
54

indicated that there is more des-α-MSH in male sheep, but female sheep has more act-α-MSH and diact-α-MSH
compared with male sheep.* P < 0.05, vs. male sheep group.
Fig.4 (A) Mass Spectrometry results for male and female sheep in neurointermediate lobe. All three
forms of α-MSH were detected in both female and male sheep. 1622 for des-α-MSH ; 1664 for act-α-
MSH; 1706 for di-act-α-MSH. (B) HPLC/EIA results indicated that there is no any difference in desα-MSH
between male and female sheep, but female sheep has more act-α-MSH and di-act-α-MSH
compared with male sheep.* P < 0.05, vs. male sheep group.
55

Fig.5 (A) Mass Spectrometry results for male and female sheep in cerebral cortex. All three forms of
α-MSH were detected in female sheep, but only des- and act-α-MSH in male sheep. 1622 for des-α-
MSH ; 1664 for act-α-MSH; 1706 for di-act-α-MSH . (B) HPLC/EIA results indicated that female
sheep has more des-α-MSH, act-α-MSH and di-act-α-MSH compared with male sheep.* P < 0.05, vs.
male sheep group.
56

Fig.6 (A) Mass Spectrometry results for male and female sheep in pre-optic area. All three forms of α-
MSH were detected in both female and male sheep. 1622 for des-α-MSH ; 1664 for act-α-MSH; 1706
for di-act-α-MSH. (B) HPLC/EIA results indicated that there is no any difference in des-α-MSH
between male and female sheep, but female sheep has more act-α-MSH and di-act-α-MSH compared
with male sheep.* P < 0.05, vs. male sheep group.
Fig.7 (A) Mass Spectrometry results for male and female sheep in median eminence. All three forms
of α-MSH were detected in female sheep, but only des- and act-α-MSH in male sheep. 1622 for des-α-
MSH ; 1664 for act-α-MSH; 1706 for di-act-α-MSH. (B) HPLC/EIA results indicated that there is no
any difference in des-α-MSH between male and female sheep, but female sheep has more act-α-MSH
and di-act-α-MSH compared with male sheep.* P < 0.05, vs. male sheep group.
57

Table 2 The distribution of three forms α-MSH in brain tissues by Mass spectrometry analysis
Male
Female
des- mono-act di-act des- mono-act di-act
Hypothalamus + + — + + +
Anterior pituitary + + — — + +
Arcuate nucleus + — — + + —
Neurointermediate lobe + + + + + +
Median eminence + + — + + +
Cerebral cortex + + — + + +
Pre-optic area + + + + + +
Spinal cord — — — — — —
Brain stem — — — — — —
Cerebellum — — — — — —
Discussion
Some researches indicated that no authentic α-MSH was detected in extracts of rat brain regions
or human pituitary. In these tissues the only α-MSH variant detected was characterized as des-acetyl
α-MSH. Conversely, acetylated α-MSH was found to be the predominant form in rat pituitary(12). It is
the same result of female sheep in our research, Only acetylated forms(including mono-act and di-act)
were detected, but except acetylated α-MSH ,we can still find des-acetyl α-MSH in male sheep.
The role of the N-acetyl group has been extensively studied in conjunction with
melanocortinergic regulation of food intake. Peripheral injection of α-MSH in viable yellow obese
mice resulted in significantly lower food intake and body weight gain than in mice injected with the
same doses of desacetyl-α-MSH (17). Tsujii and Bray (16) injected acetyl-α-MSH and des-acetyl-α-
MSH intracerebroventricularly (i.c.v.) into food-deprived rats and found that acetyl-α-MSH produced
a highly significant reduction of food intake at doses of 100 and 250 pmol whereas des-acetyl-α-MSH
had no effect at any dose between 100 and 2500 pmol. Similarly, after i.c.v. administration to fasted
rats, doses of 1, 3, or 6 nmol acetyl-α-MSH resulted in decreased food intake (18). Des-acetyl-α-MSH
did not produce a significant effect at these doses but reduced food intake at a dose of 25 nmol. The
most recent demonstration of differential effects of MSH acetylation on the regulation of food intake
has been provided by Guo et al.2004(19), who again demonstrated that i.c.v. injection of 3 nmol
acetyl-α-MSH, but not des-acetyl-α-MSH, resulted in inhibition of food intake in rats.
Differential effects of acetylated and non-acetylated α-MSH have also been noted in the
adrenal cortex, adipose tissue, and bone. In addition to ACTH or melanocortin-2 receptors
(MC2R),MC3R and MC5R are expressed in the adrenal cortex (20,21).Both acetyl-α-MSH
and des-acetyl-a-MSH stimulate aldosterone, corticosterone, and cortisol secretion from
human adrenocortical cells in vitro (22). Des-acetyl-α-MSH is much more potent than acetylα-MSH,
but both are far less effective than ACTH (1–24). ACTH (1–24) elicited significant
secretion of all three steroids at doses of 1.0 pmol whereas the threshold dose was 1.0 nmol
for des-acetyl-α-MSH and 100 nmol for acetyl-α-MSH. The relative potencies of ACTH and
α-MSH agree with the 104-fold difference in potency between the two peptides in rat adrenal
(23). However, others have failed to find any evidence for steroidogenic activity or MC2R
binding of α-MSH in rat or mouse adrenal (24,25). Des-acetyl-, di-acetyl- and mono-acetyl-α-
MSH have been demonstrated to have lipolytic activity in rabbits both in vivo and in vitro,
but none of the forms is effective in rats (24). The threshold dose of des-acetyl-α-MSH to
increase plasma free fatty acid (FFA) concentration in rabbits was 1 mg/kg, but for the two
acetylated forms was 0.33 mg/kg. As determined by peak levels of FFA, the relative potencies
of diacetyl-/monoacetyl-/desacetyl-α-MSH were 8.3/5.3/1. The effects of the three peptides
on lipolytic activity in rabbit adipose tissue slices followed similar patterns except that the
threshold doses and potencies of acetyl-α-MSH and di-acetyl-α-MSH were equivalent. MC2R,
MC4R, and MC5R have now all been identified in adipose tissue by quantitative reverse
58

transcription-polymerase chain reaction analysis (RT-PCR) (26,27). . These data suggest that
N-acetylation of products formed by the processing of α-MSH can markedly alter their
biological properties.
In our research, we detected the different N-acetylation forms of α-MSH in ten brain
tissues of female and male sheep. The results illustrated that the distribution of α-MSH may
be affected by the sex steroid hormone, and further researches should be done in the future.
Acknowledgments
This work were finished at Department of Biochemistry & Molecular Biology and Department of
Physiology, Faculty of Medicine, Nursing and Health Sciences, Monash University .The authors are
grateful to Dr. David Steer(Bio) and Josie Lawrence(Bio) for their valuable discussion on MS analysis;
Iresha Hanchapola (Bio) ,Melanie Clarke(Physi) and Alexandra Rao (Physi) for their help of
HPLC/EIA.
References:
1. Nillni EA ,Sevarino KA. The biology of pro-thyrotropin-releasing hormonederived peptides.
Endocr Rev 1999; 20: 599-648.
2. Benjannet S, Rondeau N, Day R, Chretien M, Seidah NG. PC1 and PC2 are proprotein convertases
capable of cleaving proopiomelanocortin at distinct pairs of basic residues. Proc Natl Acad Sci
USA 1991; 88: 3564-3568
3. Barnea A, Cho G, Porter JC. A reduction in the concentration of immunoreactive corticotropin,
melanotropin and lipotropin in the brain of the aging rat. Brain Res 1982;232: 345-353.
4. Emeson RB , Eipper BA. Characterization of pro-ACTH/endorphin-derived peptides in rat
hypothalamus. J Neurosci 1986; 6: 837-849.
5. Gramsch C, Kleber G, Hollt V, Pasi A, Mehraein P, Herz A. Proopiocortin fragments in human and
rat brain: beta-endorphin and alpha-MSH are the predominant peptides. Brain Res 1980;192: 109-
119.
6. Orwoll E, Kendall JW, Lamorena L, McGilvra R. Adrenocorticotropin and melanocyte-stimulating
hormone in the brain. Endocrinology 1979;104: 1845-1852.
7. Cone RD, Lu D, Koppula S, Vage DI, Klungland H, Boston B, Chen W, Orth DN, Pouton C,
Kesterson RA. The melanocortin receptors: agonists, antagonists, and the hormonal control of
pigmentation. Recent Prog Horm Res 1996;51: 287-317.
8. Pritchard LE, Turnbull AV, White A. Pro-opiomelanocortin processing in the hypothalamus: impact
on melanocortin signalling and obesity. J Endocrinol 2002; 172: 411-421.
9. Wilkinson CW. Roles of acetylation and other post-translational modifications in melanocortin
function and interactions with endorphins. Peptides 2006; 27: 453-471.
10. Bronstein DM, Schafer MK, Watson SJ, Akil H. Evidence that beta endorphin is synthesized in
cells in the nucleus tractus solitarius: detection of POMC mRNA. Brain Res 1992; 587: 269-275.
11. Palkovits M, Mezey E, Eskay RL. Pro-opiomelanocortin-derived peptides (ACTH/betaendorphin/alpha-MSH)
in brainstem baroreceptor areas of the rat. Brain Res 1987; 436: 323-338.
12. Evans CJ, Lorenz R, Weber E, Barchas J D. Variants of alpha-melanocyte stimulating hormone in
rat brain and pituitary evidence that acetylated alpha-MSH exists only in the intermediate lobe of
pituitary. Biochem Biophys Res Commun 1982; 106: 910–919.
13. Emeson RB, Eipper BA. Characterization of pro-ACTH/endorphin-derived peptides in rat
hypothalamus. J Neurosci 1986; 6: 837–849.
14. Jegou S, Tranchand-Bunel D, Delbende C, Blasquez C. Vaudry H. Characterization of alpha-
MSH-related peptides released from rat hypothalamic neurons in vitro. Brain Res Mol Brain Res
1989; 5, 219–226.
15. Abbott CR., Rossi M, Kim M, AlAhmed SH, Taylor GM, Ghatei MA, Smith DM, Bloom SR.
Investigation of the melanocyte stimulating hormones on food intake. Lack Of evidence to support
a role for the melanocortin-3-receptor. Brain Res 2000; 869: 203–210.
16. Tsujii S, Bray GA. Acetylation alters the feeding response to MSH and b-endorphin. Brain Res
Bull 1989; 23:165–169.
59

17. Shimizu H, Shargill NS, Bray GA, Yen TT, Gesellchen PD. Effects of MSH on food intake, body
weight and coat color of the yellow obese mouse. Life Sci 1989;45:543–52.
18. Abbott CR, Rossi M, Kim M-S, AlAhmed SH, Taylor GM, Ghatei MA, Smith DM, Bloom SR.
Investigation of the melanocyte stimulating hormones on food intake. Lack of evidence to support
a role for the melanocortin-3-receptor. Brain Res 2000;869:203–210.
19. Guo L, Mu¨ nzberg H, Stuart RC, Nillni EA, Bjørbæk C. N-acetylation of hypothalamic α-
melanocyte-stimulating hormone and regulation by leptin. Proc Natl Acad Sci USA 2004;
101:11797-11802.
20. Dhillo WS, Small CJ, Gardiner JV, Bewick GA, Whitworth EJ, Jethwa PH, Seal LJ, Ghatei MA,
Hinson JP, Bloom SR. Agouti-related protein has an inhibitory paracrine role in the rat adrenal
gland. Biochem Biophys Res Commun 2003;301:102–107.
21. Liakos P, Chambaz EM, Feige JJ, Defaye G. Expression of ACTH receptors (MC2-R and MC5-R)
in the glomerulosa and the fasciculata-reticularis zones of bovine adrenal cortex. Endocr Res
1998;24:427–432.
22.Henville KL, Hinson JP, Vinson GP, Laird SM. Actions of desacetyl-α-melanocyte-stimulating
hormone on human adrenocortical cells. J Endocrinol 1989;121:579–583.
23. Baumann JB, Eberle AN, Christen E, Ruch W, Girard J. Steroidogenic activity of highly potent
melanotropic peptides in the adrenal cortex of the rat. Acta Endocrinol 1986;113:396–402.
24. Rudman D, Hollins BM, Kutner MH, Moffitt SD, Lynn MJ. Three types of α-melanocytestimulating
hormone: bioactivities and half-lives. Am J Physiol 1983; 245:47–54.
25. Schio¨th HB, Chhajlani V, Muceniece R, Klusa V, Wikberg JES. Major pharmacological
distinction of the ACTH receptor from other melanocortin receptors. Life Sci 1996; 59:797–801.
26. Boston BA. The role of melanocortins in adipocyte function. Ann NY Acad Sci 1999; 885:75–84.
27. Hoggard N, Hunter L, Duncan JS, Rayner DV. Regulation of adipose tissue leptin secretion by α-
melanocyte stimulating hormone and agouti-related protein: further evidence of an interaction
between leptin and the melanocortin signalling system. J Mol Endocrinol 2004;32:145–153.
60

Investigation of the mechanism of the cardioprotective effect of flavonols
Chengxue Qin, 1,2 * Richard A. Hughes, 1 Spencer J. Williams, 2 Owen L. Woodman 3
* Presenter
1. The Department of Pharmacology, University of Melbourne, Parkville, 3010
2. School of Chemistry and Bio21 Molecular Science and Biotechnology Institute, University
of Melbourne, Parkville, 3010
3. School of Medical Sciences, RMIT University, Bundoora, 3083
We have reported that 3’, 4’-dihydroxyflavonol (DiOHF) reduces infarct size after
myocardial ischaemia/reperfusion in sheep [1] and ischaemic stroke in rats [2]. It is however
not known whether the cardioprotective action of this flavonol relies on its anti-calcium or
antioxidant activities.
In this study, we investigated the mechanism of the beneficial action of flavonols using
our recently discovered single-action anti-calcium compound (4’-OH-3’-OCH 3 flavonol) [3]
and the complementary antioxidant compound (DiOHF-6-succinamic acid) [4] in rat isolated
hearts. Hearts were perfused with physiological solution using a Langendorff apparatus and
subjected to global, no-flow 20-minute ischaemia followed by 30-minute reperfusion. Hearts
(n = 6~8 each group) were randomly treated with vehicle (0.05% DMSO), DiOHF, 4’-OH-3’-
OCH 3 flavonol or DiOHF-6-succinamic acid. Treatments (10 -5 M) were included in the
perfusate for 10 min before ischaemia and during reperfusion. In vehicle treated hearts,
LV+dP/dt was significantly reduced at the end of reperfusion (60±8%) compared to the preischaemic
value. Furthermore, total lactate dehydrogenase (LDH) release was significantly
elevated during reperfusion in vehicle treated hearts (895±121 U/L) compared to shams
(70±10 U/L), indicating significant myocardial damage (p

Efficient Ethanol production in glucose/xylose co-fermentations by a novel
co-culture scheme of Zymomonas mobilis and Pichia stipitis
Nan Fu 1 * , Paul Peiris 1 , Julie Markham 1 , John Bavor 1
* Presenter
1. School of Natural Sciences, University of Western Sydney, Locked Bag 1797, Penrith
South DC, NSW 1797, Australia
The efficiency of bioethanol production from lignocellulose would be significantly
improved if the glucose and xylose in the hydrolysate could be simultaneously converted to
ethanol. This study designed a novel fermentation scheme, using a glucose-fermenting strain,
Zymomonas mobilis and a pentose utilizing strain, Pichia stipitis to convert the sugar mixture
substrates including both synthetic media and hydrolysate media.
Fermentation characteristics of the two organisms were optimized in media containing
50 g/l of the corresponding sugars. The inoculum was a key factor in the optimization for the
production of ethanol. A ten fold increase in the cell number in the inoculum of Z. mobilis
reduced the fermentation time from 8 to 2.5 h, resulting in an improvement in the overall
volumetric ethanol productivity from 2.93 to 10.54 g/l/h. Similarly, a five fold increase in the
cell number in the inoculum of P. stipitis resulted in a decrease of the fermentation time from
130 to 28 h, leading to an increase of the volumetric productivity from 0.14 to 0.73 g/l/h. This
is the highest value of volumetric ethanol productivity for Pichia stipitis ever reported.
Co-fermentation of sugar mixture simulating the bagasse hydrolysate (30 g/l glucose and
20 g/l xylose) was carried out with different fermentation schemes which included free cells,
immobilized cells, co-culture and sequential culture. Co-culture with immobilized Z. mobilis
and free cells of P. stipitis proved to be the best scheme. Complete utilization of the sugars
within 19 h with an ethanol yield of 0.49 g/g was achieved, giving a volumetric productivity
of 1.277 g/l/h. Application of this fermentation scheme on sugarcane bagasse hydrolysate
resulted in a complete sugar utilization within 26 h; ethanol yield peaked at 40 h with a yield
of 0.49 g/g. These results are the highest reported values for co-culture and compare favorably
with best results reported with recombinant strains.
62

Dual mode roll-up effect in a multicomponent near adiabatic adsorption
process
Gang Li, 1 * Dong Xu, 1,2 Penny Xiao, 1 Paul A Webley 1
* Presenter
1. Cooperative Research Centre for Greenhouse Gas Technologies, Department of Chemical
Engineering, Monash University, Wellington Road, Clayton, Victoria 3800, Australia
2. School of Materials & Metallurgy, Northeastern University, Shenyang 110004, P.R. China
Adsorption technologies have been widely used in laboratorial analysis/separations, e.g.
chromatography and portable life support systems, and also in industrial process, e.g. air
separation and carbon capture. A common phenomenon in multi-component adsorption in a
fixed bed is that the effluent concentration of some components may exceed their steady inlet
values. This effect, often referred to as roll-up, is attributed to the displacement of the more
weakly adsorbed component by the slower moving more strongly adsorbed component,
according to the conventional explanation in the literature. This has also been applied to some
practical operations in liquid chromatography and some large-scale chromatographic
separations to desorb the pre-loaded components from a column by introducing a desorbent,
which is more strongly adsorbed than any other component in the system. Many published
works reported that in a system containing two adsorbed species, one could easily see a single
roll-up of the weak component during the breakthrough process. However, our recent study
showed that more than one roll-up could occur, in such a system using CO 2 and H 2 O vapour
as the weakly and strongly adsorbed components respectively in double zeolite 3A/13X and
activated alumina/13X layered columns. It has also been found that a pure thermal wave was
generated by H 2 O adsorption and it moved ahead of the H 2 O mass transfer zone, and the first
roll-up coincided with the propagation of the thermal front to the 3A/13X layer interface
while the second roll-up happened when H 2 O concentration front reached the 13X layer. In
theory, enthalpy can be regarded as a direct equivalence to an extra adsorbed component so
that temperature front will sweep off the column before the propagation of the strongly
adsorbed component (H 2 O in this case), but it only signifies under adiabatic or near adiabatic
conditions. The interval of the two roll-ups reduces with the decrease of the length of 3A layer,
and eventually they merge into a single pronounced roll-up. The results were further verified
by computer simulations. Therefore, to the best of our knowledge, this is the first affirmative
report about this unique dual mode roll-up effect, which will be essential in the design of near
adiabatic adsorption/chromatography units.
63

Effect of glucose and sodium chloride on the stability of aged glucose
oxidase from Aspergillus niger
Chun Ming (Eric) Wong 1* , Xiao Dong Chen 1 , Kwun Hei Wong 2
* Presenter
1. Biotechnology and Food Engineering Group, Department of Chemical Engineering,
Monash University, Clayton Campus, Victoria 3800, Australia
2. Auximedic Ltd, PO box 109-175, Newmarket, Auckland 1149, New Zealand
Glucose-1-oxidase (GOX) from Aspergillus niger is a well characterized enzyme used in
many commercial applications and is found naturally in insects and fungi. Various properties
of GOX had been well documented but information on the effects of GOX stabilizer, glucose
and sodium chloride are sparse. This article presents the effectiveness of glucose and sodium
chloride as GOX stabilizer against drying and heat denaturation in solution with a three year
old food grade GOX preparation over time. The primary method used to measure GOX
activity was to log the decrease in dissolved oxygen using clark type electrode. Drying
experiments showed that the additions of thermo-stabilizers do not increase the stability of
dried GOX but actually accelerates its rate of degradation over time. After about a month,
only 1/3 of the initial activity was left in the dried GOX containing glucose and/or sodium
chloride. Heat denaturantion experiments showed that thermostability of GOX increases in
the presence of sodium chloride and/or glucose. Stabilization due to dissolved sodium
chloride and glucose are cumulative, postulated to work via different mechanism. The
contradicting finding presented leave no doubt that there is implications for GOX
manufacturers and users whether or not they are working towards more stable GOX
preparations.
64

Paper as a low-cost base material for diagnostic and environmental sensing
applications
Xu Li 1 * , Junfei Tian 1 , Wei Shen 1
* Presenter
1. Australian Pulp and Paper Institute, Department of Chemical Engineering, Monash
University, Victoria 3800, Australia
Paper has been used as a chromatographic substrate and for indicator strips for a long
time. Recently, a novel application using patterned paper to make microfluidic systems for
diagnostic and environmental sensing has been developed. The patterned paper which can
control liquid penetration within its hydrophilic channels has great potential to make low-cost
and simple analytical devices for bioassays and environmental analysis. This article shows
that basic papermaking and printing techniques can be applied to create paper-based
microfluidic analytical devices. High volume and high speed printing processes hold the key
for the transformation of the novel concept of paper-based microfluidic system into practical
analytical devices that will find applications in many fields.
65

Data Management in Cloud Scientific Workflow Systems
Dong Yuan 1 *
* Presenter
1. Faculty of Information and Communication Technologies, Swinburne University of
Technology, Hawthorn, Melbourne, Australia 3122
Data-intensive scientific applications are posing many challenges in distributed
computing systems. In the scientific field, the application data are expected to double every
year over the next decade and further. With this continuing data explosion, high performance
computing systems are needed to store and process data efficiently, and workflow
technologies are facilitated to automate these scientific applications. Scientific workflows are
typically very complex. They usually have a large number of tasks and need a long time for
execution. Running scientific workflow applications usually need not only high performance
computing resources but also massive storage. The emergence of cloud computing
technologies offers a new way to develop scientific workflow systems. Scientists can upload
their data and launch their applications on the scientific cloud workflow systems from
everywhere in the world via the Internet, and they only need to pay for the resources that they
use for their applications. As all the data are managed in the cloud, it is easy to share data
among scientists. This kind of model is very convenient for users, but remains a big challenge
to the system. This seminar will discuss several research topics of data management in
scientific cloud workflow systems.
Keywords
component, formatting, style, styling, insert (key words)
1. Introduction
Data-intensive scientific applications are posing many challenges in distributed computing systems.
In many scientific research fields, like astronomy, high-energy physics and bio-informatics, scientists
need to analyse terabytes of data either from existing data resources or collected from physical
devices. During these processes, similar amounts of new data might also be generated as intermediate
or final products [9]. According to [20], in the scientific field, the application data are expected to
double every year over the next decade and further. With this continuing data explosion, high
performance computing systems are needed to store and process data efficiently, and workflow
technologies are facilitated to automate these scientific applications. Scientific workflows are typically
very complex. They usually have a large number of tasks and need a long time for execution. Running
scientific workflow applications usually need not only high performance computing resources but also
massive storage [9].
Nowadays, popular scientific workflows are often deployed in grid systems [16] because they have
high performance and massive storage. However, building a grid system is extremely expensive and it
is normally not open for scientists all over the world. The emergence of cloud computing technologies
offers a new way to develop scientific workflow systems.
Since late 2007 the concept of cloud computing was proposed [22], it has been utilised in many
areas with some success. Cloud computing is deemed as the next generation of IT platforms that can
deliver computing as a kind of utility [8]. Foster et al. made a comprehensive comparison of grid
computing and cloud computing [11]. Some features of cloud computing also meet the requirements
of scientific workflow systems. Cloud computing systems provide the high performance and massive
storage required for scientific applications in the same way as grid systems, but with a lower
infrastructure construction cost among many other features, because cloud computing systems are
composed of data centres which can be clusters of commodity hardware [22]. Research into doing
science and data-intensive applications on the cloud has already commenced [18], such as early
experiences like Nimbus [14] and Cumulus [21] projects. The work by Deelman et al. [10] shows that
67

cloud computing offers a cost-effective solution for data-intensive applications, such as scientific
workflows [13]. Furthermore, cloud computing systems offer a new model that scientists from all over
the world can collaborate and conduct their research together. Cloud computing systems are based on
the Internet, and so are the scientific workflow systems deployed in the cloud. Scientists can upload
their data and launch their applications on the scientific cloud workflow systems from everywhere in
the world via the Internet, and they only need to pay for the resources that they use for their
applications. As all the data are managed in the cloud, it is easy to share data among scientists. This
kind of model is very convenient for users, but remains a big challenge for data management to
scientific cloud workflow systems.
Firstly, new data storage strategy is required in cloud scientific workflow systems. In a cloud
computing, theoretically, the system can offer unlimited storage resources. All the application data of
the scientific workflows can be stored, including the generated intermediate data, if we are willing to
pay for the required resources. Storing all the application data in the cloud is obviously not costeffective,
since some data are seldom used and huge in size. However, in scientific cloud workflow
systems, a scientist can not decide whether a piece of application data should be stored or not, since
the data are shared and he is not the only user.
Secondly, new data placement strategy is also required, which means the cloud workflow systems
must have the ability to decide where to store the application data. Cloud computing platform contains
different cloud service providers with different pricing models, where data transfers between service
providers also carry a cost. The cloud scientific workflows are usually distributed, and the data
placement strategy will decide where to store the application data, in order to reduce the total system
cost.
Last but not least, new data replication strategy should also be designed for cloud scientific
workflow systems. A good replication strategy can not only guarantee the security of application data,
but also further reduce the system cost by replicating frequently used data in different locations.
Replication strategy in the Cloud should be dynamic based on the application data’s usage rate.
2. Related Works
As cloud computing has become more and more popular, new data management systems have also
appeared, such as Google File System [12] and Hadoop [1]. They all have hidden infrastructures that
can store the application data independent of users’ control. Google File System is designed mainly for
Web search applications, which are different from workflow applications. Hadoop is a more general
distributed file system, which has been used by many companies, such as Amazon and Facebook.
When you push a file to a Hadoop File System, it will automatically split this file into chunks and
randomly distribute these chunks in a cluster. Furthermore, the Cumulus project [21] introduced a
scientific cloud architecture for a data centre. And the Nimbus [14] toolkit can directly turn a cluster
into a cloud and it has already been used to build a cloud for scientific applications.
Comparing to the distributed computing systems like cluster and grid, a cloud computing system
has a cost benefit [4]. Assunção et al. [5] demonstrate that cloud computing can extend the capacity of
clusters with a cost benefit. Using Amazon clouds’ cost model and BOINC volunteer computing
middleware, the work in [15] analyses the cost benefit of cloud computing versus grid computing. The
idea of doing science on the cloud is not new. Scientific applications have already been introduced to
cloud computing systems. In terms of the cost benefit, the work by Deelman et al. [10] also applies
Amazon clouds’ cost model and demonstrates that cloud computing offers a cost-effective way to
deploy scientific applications. In [13], Hoffa conducted simulations of running an astronomy scientific
workflow in cloud and clusters, which shows cloud scientific workflows are cost-effective. The above
works mainly focus on the comparison of cloud computing systems and the traditional distributed
computing paradigms, which shows that applications running on cloud have cost benefits. When it
comes to how to reduce cost of running applications in clouds, Deelman et al. present in [10] that
storing some popular intermediate data can save the cost in comparison to always regenerating them
from the input data. Furthermore, in [2], Adams proposes a model to represent the trade-off of
computation cost and storage cost in storing application data, but the authors have not given the
specific method of managing the data.
68

To the best of our knowledge, research in cost-effectively managing application data of cloud
scientific workflow systems is still in blank. My research will focus on this spot and develop novel
strategies of data storage, placement and replication in cloud scientific workflow systems.
3. Research problems and methodologies
A. Data storage strategy
1) Problem analysis
Traditionally, scientific workflows are deployed on the high performance computing facilities, such
as clusters and grids. Scientific workflows are often complex with huge intermediate data generated
during their execution. How to store these intermediate data is normally decided by the scientists who
use the scientific workflows. This is because the clusters and grids only serve for certain institutions.
The scientists may store the intermediate data that are most valuable to them, based on the storage
capacity of the system. However, in many scientific workflow systems, the storage capacities are
limited, such as the pulsar searching workflow we introduced. The scientists have to delete all the
intermediate data because of the storage limitation. This bottleneck of storage can be avoided if we run
scientific workflows in the cloud.
In a cloud computing environment, theoretically, the system can offer unlimited storage resources.
All the intermediate data generated by scientific cloud workflows can be stored, if we are willing to
pay for the required resources. However, in scientific cloud workflow systems, whether to store
intermediate data or not is not an easy decision anymore.
a) All the resources in the cloud carry certain costs, so either storing or generating an intermediate
dataset, we have to pay for the resources used. The intermediate datasets vary in size, and have
different generation cost and usage rate. Some of them may often be used whilst some others may be
not. On one hand, it is most likely not cost effective to store all the intermediate data in the cloud. On
the other hand, if we delete them all, regeneration of frequently used intermediate datasets imposes a
high computation cost. We need a strategy to balance the generation cost and the storage cost of the
intermediate data, in order to reduce the total cost of the scientific cloud workflow system.
b) The scientists can not predict the usage rate of the intermediate data anymore. For a single
research group, if the data resources of the applications are only used by its own scientists, the
scientists may predict the usage rate of the intermediate data and decide whether to store or delete
them. However, the scientific cloud workflow system is not developed for a single scientist or
institution, rather, developed for scientists from different institutions to collaborate and share data
resources. The users of the system could be anonymous from the Internet. We must have a strategy
storing the intermediate data based on the needs of all the users that can reduce the cost of the whole
system.
Hence, for scientific cloud workflow systems, we need a strategy that can automatically select and
store the most appropriate intermediate datasets. Furthermore, this strategy should be cost effective
that can reduce the total cost of the whole system.
2) Methodology
Scientific workflows have many computation and data intensive tasks that will generate many
intermediate datasets of considerable size. There are dependencies exist among the intermediate
datasets. Data provenance in workflows is a kind of important metadata, in which the dependencies
between datasets are recorded. The dependency depicts the derivation relationship between workflow
intermediate datasets. For scientific workflows, data provenance is especially important, because after
the execution, some intermediate datasets may be deleted, but sometimes the scientists have to
regenerate them for either reuse or reanalysis. Data provenance records the information of how the
intermediate datasets were generated, which is very important for the scientists. Furthermore,
regeneration of the intermediate datasets from the input data may be very time consuming, and
therefore carry a high cost. With data provenance information, the regeneration of the demanding
dataset may start from some stored intermediated datasets instead. In the scientific cloud workflow
system, data provenance is recorded while the workflow execution. Taking advantage of data
provenance, we can build an IDG based on data provenance. All the intermediate datasets once
generated in the system, whether stored or deleted, their references are recorded in the IDG. Based on
the IDG, we can calculate the generation cost of every intermediate dataset in the scientific cloud
69

workflows. By comparing the generation cost and storage cost, our strategy can automatically decide
whether an intermediate dataset should be stored or deleted in the cloud system, no matter this
intermediate dataset is a new dataset, regenerated dataset or stored dataset in the system.
Data provenance is important to this strategy in building the IDG. Fortunately, due to the
importance of data provenance in scientific applications, much research about recording data
provenance of the system has been done. Some of them are especially for scientific workflow systems
[6]. Some popular scientific workflow systems, such as Kepler [16], have their own system to record
provenance during the workflow execution [3]. In [19], Osterweil et al. present how to generate a Data
Derivation Graph (DDG) for the execution of a scientific workflow, where one DDG records the data
provenance of one execution. Similar to the DDG, our IDG is also based on the scientific workflow
data provenance, but it depicts the dependency relationships of all the intermediate data in the system.
Hence we can build our IDG by taking advantage of these related works.
B. Data placement strategy
1) Problem Analysis
Scientific applications are data intensive and usually need collaborations of scientists from different institutions [7], hence application
data in scientific workflows are usually distributed and very large. When one task needs to process data from different data centres, moving
data becomes a challenge. Some application data are too large to be moved efficiently, some may have fixed locations that are not feasible to
be moved and some may have to be located at fixed data centres for processing, but these are only one aspect of this challenge. For the
application data that are flexible to be moved, we also cannot move them whenever and wherever we want, since in the cloud computing
platform, data centres may belong to different cloud service providers that data movement would result in costs. Furthermore, the
infrastructure of cloud computing systems is hidden from their users. They just offer the computation and storage resources required by users
for their applications. The users do not know the exact physical locations where their data are stored. This kind of model is very convenient
for users, but remains a big challenge for data management to scientific cloud workflow systems.
2) Methodology
In cloud computing systems, the infrastructure is hidden from users. Hence, for most of the
application data, the system will decide where to store them. Dependencies exist among these data. In
this paper, we initially adapt the clustering algorithms for data movement based on data dependency.
Clustering algorithms have been used in pattern recognition since 1980s, which can classify patterns
into groups without supervision. Today they are widely used to process data streams. In many
scientific workflow applications, the intermediate data movement is in data stream format and the
newly generated data must be moved to the destination in real-time. We adapt the k-means clustering
algorithm for data placement in scientific cloud workflow systems. Scientific workflows can be very
complex, one task might require many datasets for execution; furthermore, one dataset might also be
required by many tasks. If some datasets are always used together by many tasks, we say that these
datasets are dependant on each other. In our strategy, we try to keep these datasets in one data centre,
so that when tasks were scheduled to this data centre, most, if not all, of the data they need are stored
locally.
Our data placement strategy has two algorithms, one for the build-time stage and one for the
runtime stage of scientific workflows. In the build-time stage algorithm, we construct a dependency
matrix for all the application data, which represents the dependencies between all the datasets
including the datasets that may have fixed locations. Then we use the BEA algorithm [17] to cluster
the matrix and partition it that datasets in every partition are highly dependent upon each other. We
distribute the partitions into k data centres, where the partitions have fixed location datasets are also
placed in the appropriate data centres. These k data centres are initially as the partitions of the k-means
algorithm at runtime stage. At runtime, our clustering algorithm deals with the newly generated data
that will be needed by other tasks. For every newly generated dataset, we calculate its dependencies
with all k data centres, and move the data to the data centre that has the highest dependency with it.
By placing data with their dependencies, our strategy attempts to minimise the total data movement
during the execution of workflows. Furthermore, with the pre-allocate of data to other data centres, our
strategy can prevent data gathering to one data centre and reduces the time spent waiting for data by
ensuring that relevant data are stored locally.
C. Data replication strategy
1) Problem Analysis
Data replication is also an important issue for cloud workflow systems as presented by the two
points below:
70

First, a good replication strategy can guarantee fast data access for the cloud workflow system. In
the scientific workflows with many parallel tasks will simultaneously access the same dataset on one
data centre. The limitation of computing capacity and bandwidth in that data centre would be a
bottleneck for the whole cloud workflow system. If we have several replicas in different data centres,
this bottleneck will be eliminated.
Second, a good replication strategy can reduce data movement between data centres. For example,
if tasks in one data centre always need to retrieve data from the same data set in a remote data centre,
it is better to replicate that data set in the local data centre to reduce the data movement.
Third, a good replication strategy can guarantee data reliability for the cloud workflow system.
Because data centres in cloud workflow systems are built up with massive cheap commodity
hardware, the breakdown of some hardware could happen any time. It is essential to keep several
copies of each data in different data centres for reliability.
However, at present, data replication strategies that utilised in cloud data management systems are
usually static. For example, in Hadoop, users can manually set the number of replicas, and the system
will automatically replicate the application data in different places (racks or clusters, depends on the
scale of the system). Static replication can guarantee the data reliability, but in cloud environment,
different application data have different usage rate, where we should have the dynamic strategy to
replicate the application data based on their usage rate.
2) Methodology
The basic strategy for the replication could be as follow:
a) Always keep fix number copies of each dataset in different data centres to guarantee reliability
and dynamically add new replicas for each dataset to to guarantee data availability.
b) Where to place the replicas is based on data dependency.
c) How many replicas should a dataset have is based on usage rate of this dataset.
Reference:
[1] "Hadoop, http://hadoop.apache.org/", accessed on 25 November 2009.
[2] I. Adams, D. D. E. Long, E. L. Miller, S. Pasupathy, and M. W. Storer, "Maximizing Efficiency
By Trading Storage for Computation," in Workshop on Hot Topics in Cloud Computing
(HotCloud'09), pp. 1-5, 2009.
[3] I. Altintas, O. Barney, and E. Jaeger-Frank, "Provenance Collection Support in the Kepler
Scientific Workflow System," in International Provenance and Annotation Workshop, pp. 118-
132, 2006.
[4] M. Armbrust, A. Fox, R. Griffith, A. D. Joseph, R. H. Katz, A. Konwinski, G. Lee, D. A.
Patterson, A. Rabkin, I. Stoica, and M. Zaharia, "Above the Clouds: A Berkeley View of Cloud
Computing," University of California at Berkeley,
http://www.eecs.berkeley.edu/Pubs/TechRpts/2009/EECS-2009-28.pdf, Technical Report
UCB/EECS-2009-28, accessed on 25 November 2009.
[5] M. D. d. Assuncao, A. d. Costanzo, and R. Buyya, "Evaluating the cost-benefit of using cloud
computing to extend the capacity of clusters," in 18th ACM International Symposium on High
Performance Distributed Computing, Garching, Germany, pp. 1-10, 2009.
[6] Z. Bao, S. Cohen-Boulakia, S. B. Davidson, A. Eyal, and S. Khanna, "Differencing Provenance
in Scientific Workflows," in 25th IEEE International Conference on Data Engineering, ICDE
'09., pp. 808-819, 2009.
[7] R. Barga and D. Gannon, "Scientific versus Business Workflows," in Workflows for e-Science,
pp. 9-16, 2007.
[8] R. Buyya, C. S. Yeo, S. Venugopal, J. Broberg, and I. Brandic, "Cloud computing and emerging
IT platforms: Vision, hype, and reality for delivering computing as the 5th utility," Future
Generation Computer Systems, vol. in press, pp. 1-18, 2009.
[9] E. Deelman and A. Chervenak, "Data Management Challenges of Data-Intensive Scientific
Workflows," in IEEE International Symposium on Cluster Computing and the Grid, pp. 687-692,
2008.
71

[10] E. Deelman, G. Singh, M. Livny, B. Berriman, and J. Good, "The Cost of Doing Science on the
Cloud: the Montage example," in ACM/IEEE Conference on Supercomputing, Austin, Texas, pp.
1-12, 2008.
[11] I. Foster, Z. Yong, I. Raicu, and S. Lu, "Cloud Computing and Grid Computing 360-Degree
Compared," in Grid Computing Environments Workshop, GCE '08, pp. 1-10, 2008.
[12] S. Ghemawat, H. Gobioff, and S.-T. Leung, "The Google file system," SIGOPS Oper. Syst. Rev.,
vol. 37, pp. 29-43, 2003.
[13] C. Hoffa, G. Mehta, T. Freeman, E. Deelman, K. Keahey, B. Berriman, and J. Good, "On the Use
of Cloud Computing for Scientific Workflows," in 4th IEEE International Conference on e-
Science, pp. 640-645, 2008.
[14] K. Keahey, R. Figueiredo, J. Fortes, T. Freeman, and M. Tsugawa, "Science Clouds: Early
Experiences in Cloud Computing for Scientific Applications," in First Workshop on Cloud
Computing and its Applications (CCA'08), pp. 1-6, 2008.
[15] D. Kondo, B. Javadi, P. Malecot, F. Cappello, and D. P. Anderson, "Cost-benefit analysis of
Cloud Computing versus desktop grids," in IEEE International Symposium on Parallel &
Distributed Processing, IPDPS'09, pp. 1-12, 2009.
[16] B. Ludascher, I. Altintas, C. Berkley, D. Higgins, E. Jaeger, M. Jones, and E. A. Lee, "Scientific
workflow management and the Kepler system," Concurrency and Computation: Practice and
Experience, pp. 1039–1065, 2005.
[17] W. T. McCormick, P. J. Sehweitzer, and T. W. White, "Problem Decomposition and Data
Reorganization by a Clustering Technique," Operations Research, vol. 20, pp. 993-1009, 1972.
[18] C. Moretti, J. Bulosan, D. Thain, and P. J. Flynn, "All-Pairs: An Abstraction for Data-Intensive
Cloud Computing," in IEEE International Parallel & Distributed Processing Symposium,
IPDPS'08, pp. 1-11, 2008.
[19] L. J. Osterweil, L. A. Clarke, A. M. Ellison, R. Podorozhny, A. Wise, E. Boose, and J. Hadley,
"Experience in Using A Process Language to Define Scientific Workflow and Generate Dataset
Provenance," in 16th ACM SIGSOFT International Symposium on Foundations of Software
Engineering Atlanta, Georgia: ACM, 2008.
[20] A. S. Szalay and J. Gray, "Science in an Exponential World," Nature, vol. 440, pp. 23-24, 2006.
[21] L. Wang, J. Tao, M. Kunze, A. C. Castellanos, D. Kramer, and W. Karl, "Scientific Cloud
Computing: Early Definition and Experience," in 10th IEEE International Conference on High
Performance Computing and Communications, HPCC '08. , pp. 825-830, 2008.
[22] A. Weiss, "Computing in the Cloud," ACM Networker, vol. 11, pp. 18-25, 2007.
72

HC_AB: A New Heuristic Clustering Algorithm based on Approximate
Backbone
* Presenter
1. Victoria University
Yu Zong 1 * , Guandong Xu 1 , Yanchun Zhang 1 , Mingchu Li 1
Clustering is an important research area with numerous applications in pattern
recognition, machine learning, and data mining. Since the clustering problem on numeric data
sets can be formulated as a typical combinatorial optimization problem, many research studies
have addressed the design of heuristic algorithms for finding local suboptimal solutions in a
reasonable period of time. However, a majority of heuristic clustering algorithms suffer from
the problem of being sensitive to the initialization and not guaranteeing the high quality
clusters. Recently, Approximate Backbone (AB), i.e., the commonly shared intersections of
several local suboptimal solutions, has been proposed in heuristic algorithm design to
improve the performance of heuristic algorithms. In this paper, we aim to introduce the AB
into heuristic clustering to overcome the drawbacks of conventional heuristic clustering
algorithms. The main strength of the proposed method is the capability of restricting the
initial search space around the global optimal results by defining the AB, and in turn,
reducing the impact of initialization on clustering, eventually improving the performance of
heuristic clustering. Experiments on synthesis and real world data sets are performed to
validate the effectiveness of the proposed approach.
73

Web Page Prediction Based on Conditional Random Fields
* Presenter
1. The University of Melbourne
Yong Zhen Guo 1 * , Yuan Miao 1
Web page prefetching is used to reduce the access latency on the Internet. However, if
most prefetched Web pages are not visited by the users in their subsequent accesses, the
limited network bandwidth and server resources will not be used efficiently and may even
worsen the access delay problem. Therefore, enhancing the Web page prediction accuracy is a
key issue of Web page prefetching. In this talk, a Web page prediction method based on the
powerful sequential learning model, Conditional Random Fields (CRFs), is proposed to
improve the Web prediction accuracy. We also show how to scale the CRF-based Web
prediction method to large-size websites by using the ECOC (Error Correcting Output
Coding) technique. Moreover, because the limited class information provided to the binarylabel
sub-classifiers in ECOCCRFs will also lead to inferior accuracy when compared to the
multi-label CRFs, in this talk, we introduce the grouping method which allows us to obtain a
prediction accuracy closer to that of multi-label CRFs while maintaining the advantage of
ECOC-CRFs. The experimental results show that the Web prediction method based on the
grouped ECOC-CRFs is highly accurate and scalable, and is ready for use in large-scale
websites to perform predictions.
74

Provenancing Qualifications Using MEASUR within Higher Education
Institutions: An Australian Case
Xiaochen Li 1 * , Bruce Calway 1 , and Suku Sinnappan 1
* Presenter
1. Swinburne University of Technology
Individuals move across educational institutions, learning modes and borders for further
study and employment purpose. Properly provenanced learning credentials can be carried
across institutes and borders and be recognised, thus facilitate individuals’ mobility.
Provenance is the origin or source from which an item comes, and the history of subsequent
owners or proof of authenticity. Provenance data has been collected and used in areas such as
online learning, social networks, and e-science research for a range of uses. This workinprogress
paper models the existing workflows adopted by higher education institutions on
assessing qualifications using MEASUR method. The workflow model has implications on
the collection of provenance data of learning credentials for authentication and verification
purpose.
75

A Multi-modal Gesture Recognition System in a Human-Robot Interaction
Scenario
* Presenter
1. Monash University
Zhi Li 1 * and Ray Jarvis 1
Recognition of non-verbal gestures is essential for robots to understand a user’s state and
intention in a Human-Robot Interaction (HRI) scenario. In this paper a multi-modal system is
proposed to recognize a user’s hand gestures and estimate body poses from a single viewpoint.
A newly emerged active sensing technology is employed to derive depth data at a high frame
rate. Depth data is useful for objects detection and localization in 3D spaces and for image
segmentation. A pair of stereo cameras is used to sense the head gestures of the user, which
provides useful information about the user’s attention. Both hand shapes and hand trajectories
are recognized. Full configurations of body poses are estimated using a model-based
algorithm. Poses are tracked by a Particle Filter, and refined by a gradient-based searching
method in the neighborhood of the particles.
76

Distributed Agent based Interoperable Virtual EMR System for Healthcare
System Integration
* Presenter
1. Vicotoria University
Xuebing Yang 1 * , Yuan Miao
One of the major problems in health care system integration is the formidable cost of
mediating between myriad vendors and policy makers for updating existing heterogeneous
systems to support a great variety of standards or interfaces. To provide cost-effective
healthcare system integration solution, this paper presents a Graphical User Interface state
model (GUISM) for automatically exchanging information with existing healthcare software
through their GUIs with no modifications needed to them. This can save the huge cost of
upgrading, testing and redeploying the existing systems. By using the GUISM model,
distributed agents are deployed to the client computers interacting with the local electronic
medical system (EMR) for communicating with other EMR systems. The whole system is
called virtual EMR system and each client in this system can request needed patient
healthcare information without knowing the actual location of the data.
77

Searching for Fair Joint Gains in Agent-based Negotiation
* Presenter
1. Swinburne University
Minyi Li 1 *
In multi-issue negotiations, autonomous agents can act cooperatively to benefit from
mutually preferred agreements. However, empirical evidence suggests that they often fail to
search for joint gains and end up with inefficient results. To address this problem, we propose
a novel mediated negotiation procedure to support the negotiation agents in reaching an
efficient and fair agreement in bilateral multi-issue negotiation. At each stage of negotiation,
the mediator searches for the compromise direction based on a new E-DD (Equal Directional
Derivative) approach and computes the new tentative agreement. We experimentally evaluate
the proposed approach with different kinds of utility functions. The experimental result
demonstrates that the proposed approach not only guarantees Pareto efficiency, but also
produces fairer improvements for two negotiating agents compared with other existing
methods.
78

Mesoporous Silica-Templated Assembly of Luminescent Polyester Particles
[1]
Jiwei Cui, 1 2 * Yajun Wang, 1 2 Jingcheng Hao, 2 and Frank Caruso 1
* Presenter
1. Centre for Nanoscience and Nanotechnology, Department of Chemical and Biomolecular
Engineering, The University of Melbourne, Parkville, Victoria 3010, Australia.
2. Key Laboratory of Colloid and Interface Chemistry of Ministry of Education, Shandong
University, Jinan 250100, P. R. China.
We report the template assembly of luminescent poly-3-hydroxybutyrate particles doped
with rare-earth complexes. The hydrophobic polymer, poly-3-hydroxybutyrate, has been
infiltrated into the nanopores of mesoporous silica particles in organic solvent. Owing to the
van der Waals interaction of the polymer chains, poly-3-hydroxybutyrate loaded in the
nanopores yields replicated particles following removal of the mesoporous silica template. To
prevent aggregation of the hydrophobic poly-3-hydroxybutyrate particles in aqueous media,
the poly-3-hydroxybutyrate-loaded mesoporous silica particles were coated with a
polyelectrolyte multilayer shell through the layer-by-layer assembly of poly(allylamine
hydrochloride) and poly(sodium 4-styrenesulfonate). Following removal of the silica core, the
polyelectrolyte multilayer-coated poly-3-hydroxybutyrate replicas were used to effectively
coordinate rare-earth complexes (europium β-diketone complexes). The rare-earth-loaded
poly-3-hydroxybutyrate replicas coated with polyelectrolyte multilayer emit intense
luminescence over a wide range of pH (3~11) and for at least several months in aqueous
solution, which is due to the intramolecular energy transfer from the ligand to the luminescent
center in the rare-earth complexes. The poly-3-hydroxybutyrate replicas with stable and
intense luminescence may find application in diagnostics and drug delivery.
Introduction
Mesoporous silica (MS) materials have attracted considerable attention in the areas of chemistry,
physics, biology, and materials science since their introduction in the early 1990s. 1 Owing to their high
surface areas (up to ~1200 m 2 g -1 ), unique pore structures (in the range of 2-50 nm), and tunable
particle morphology, MSs have been widely used to load and encapsulate various species, and for the
template synthesis of diverse nanostructured materials. 2-4 Replication of MSs with metals, 5 metal
oxides, 6 carbon, 7 and polymers 8 have been achieved by filling the nanopores with small molecule
precursors such as metal alkoxides, sucrose, and organic monomers, followed by hydrolysis,
carbonization, or polymerization of the precursors and removal of the template.
Recently, we introduced a method to prepare nanoporous, polymer-based spheres (NPS) by
templating MS spheres. This approach involves infiltration of preformed macromolecules (i.e.,
polymers or proteins) into the nanopores of the MS particles, followed by crosslinking of the polymer
chains, and removal of the MS particle templates. 9 The general applicability of this technique
facilitates the design of polymeric particles with tunable properties by varying the macromolecules
used, which can include synthetic polyelectrolytes, proteins, polypeptides, and polymer-drug
conjugates. 10 Furthermore, this method permits tuning of the morphology and size of the nanoporous
particles. Our previous studies have focused on infiltrating water-soluble macromolecules from
aqueous solution to prepare particles that are dispersed in water. The preparation of similar but
hydrophobic polymeric particles is also fundamentally interesting, as such particles can provide
significantly different physicochemical properties to their hydrophilic counterparts. This may provide
new opportunities for the loading and release of hydrophobic drugs, and for the protection and
stabilization of molecules and macromolecules that are sensitive to aqueous
[1]. Published at Chemistry of Materials (2009) 21:4310-4315

environments. However, challenges associated with preparing hydrophobic particles from
hydrophobic constituents include a lack of control over the size of the particles and difficulties in
dispersing them in aqueous media for water-based applications.
In this work, we report a facile method to prepare hydrophobic poly-3-hydroxybutyrate (PHB)
particles through templating MS spheres. PHB is hydrophobic polyester that has similar physical
properties to those of polypropylene. Compared with previous reports, a significant difference of the
current work is that no crosslinking of the infiltrated polymer chains is required to obtain intact replica
particles, owing to the hydrophobic PHB chains, which can strongly associate and maintain the
template particle morphology after MS template removal in aqueous solution. Furthermore, PHB is
biocompatible and biodegradable, and has been evaluated as a material for use in tissue engineering
scaffolds and controlled drug-release carriers. 11 Recently, it was reported that PHB can also be
coordinated with rare-earth complexes to construct luminescent films. 12 We demonstrate that the
obtained PHB replicas can be effectively used to coordinate rare-earth complexes. The rare-earth
compounds (e.g., rare-earth β-diketone complexes) have unique spectral characteristics, including
sharp emission peaks with a small peak width at half-height, long lifetimes, high fluorescence
quantum yields, and a large Stokes shifts, 13 which makes them of interest in clinical diagnostic assays,
genomic screening, and fluorescence immunoassays. 14 However, coordination of −OH groups
significantly reduces the luminescence emission intensity and decay time of the rare-earth complexes
because of nonradiative dissipation of energy of the high-energy −OH vibrations. 15 This has largely
limited the application of rare-earth complexes in biology and related fields, in which aqueous media
are typically required. Hence, the design of rare-earth materials with both strong luminescence and
good stability in aqueous media is fundamentally important. 16 The rare-earth immobilized in the PHB
replicas reported here can emit intense luminescence over a wide range of pH (3~11) and for at least
several months in aqueous solution. The polyelectrolyte multilayer (PEM) coating (poly(allylamine
hydrochloride) (PAH) and poly(sodium 4-styrenesulfonate) (PSS)) on the PHB replicas provides
stable and dispersed particles in aqueous solution.
Experimental Section
Materials. The MS particles with a bimodal pore structure (particle diameter, 2-4 μm; pore
diameters, 2-3 nm and 10-40 nm) were synthesized via the protocol reported by Schulz-Ekloff et al. 17
Poly(allylamine hydrochloride) (PAH, M w 70 000), poly(sodium 4-styrenesulfonate) (PSS, M w 70
000), europium(III) chloride hexahydrate (EuCl 3·6H 2 O), 2-thenoyltrifluoroacetone (TTA), chloroform,
and hydrofluoric acid (HF) were obtained from Sigma-Aldrich. Poly-3-hydroxybutyrate (M w 10 000)
was purchased from Polysciences, Inc. The water used in all experiments was prepared in a three-stage
Millipore Milli-Q Plus 185 purification system and had a resistivity greater than 18 MΩ·cm.
Synthesis of the Rare-Earth Complex. The europium β-diketone complex, Eu(TTA) 3·2H 2 O
(denoted as EuC), was synthesized according to published methods. 18 Briefly, 266 mg of TTA was
dissolved in a solution of 6 mL ethanol and 1.2 mL of 1 M NaOH. Then, EuCl 3·6H 2 O (146 mg) in 42
mL of water was added to the above solution, and the mixture was heated at 60 ºC for 30 min. The
yellow complex of EuC precipitated during cooling the solution to room temperature. The precipitate
was dried in vacuum after removal of the solution through filtration and water washing.
PHB Replicas from MS Particles, Surface Modification and EuC Loading. PHB was firstly
loaded in the MS particles. Approximately 5 mg of MS particles were dispersed in 0.5 mL of the PHB
solution (concentration of 35 mg mL -1 ) in chloroform at room temperature for 16 h. After removing
the supernatant with centrifugation (1500 g for 2 min), the PHB-loaded MS particles were dried in a
vacuum desiccator to remove the chloroform. The PHB-loaded MS particles were dispersed in water
with brief sonication. After removing the supernatant with centrifugation (1000 g for 2 min), 500 μL
of 2.5 M HF was added into the PHB-loaded MS particles. PHB replicas were obtained after three
washing with water. PHB replicas coated with PAH/PSS were also prepared. Before removal of the
MS templates, one PAH/PSS bilayer was coated on the PHB-loaded MS particles using the layer-bylayer
(LbL) technique. 19 The PHB replicas coated with PAH/PSS multilayers were obtained after
dissolving the silica particles. Details of this process are given in Figure 1. For the EuC loading, the
PAH/PSS-coated PHB replicas were then dispersed in an ethanol solution of EuC for 16 h. After
removal of the supernatant with centrifugation (2000 g for 2 min), three washings with water were
carried out to remove the unadsorbed EuC.
81

Figure 1. Schematic representation of the formation of EuC-loaded PHB particles obtained by
templating MS spheres. A polyelectrolyte multilayer (PEM) coating is used to stabilize the PHB
particles, thus avoiding aggregation in water and to impart functional groups to the surface of the
particles.
Instruments. Thermogravimetric analysis (TGA) experiments were conducted on a Mettler
Toledo/TGA/SDTA851e Module analyzer. The samples were heated from 25 to 120 ºC with a heating
rate of 5 ºC per min and kept at 120 ºC for 20 min under nitrogen (30 mL min -1 ). Then, the samples
were heated from 120 to 550 ºC with a heating rate of 10 ºC per min under oxygen (30 mL min -1 ). ζ-
potential measurements were performed on a Malvern 2000 Zetasizer. Fourier transform infrared
(FTIR) experiments were conducted on a Varian 7000 FT-IR spectrometer. Transmission electron
microscopy (TEM, Philips CM120 BioTWIN, operated at 120 kV) and scanning electron microscopy
(SEM, FEI Quanta 200 FEG, operated at 5 kV) were used to examine the morphology of the PHB
replicas. The TEM samples (2 μL) were placed onto Formvar-coated copper grids and allowed to airdry.
The SEM samples (2 μL) were placed onto silicon wafers and allowed to air-dry prior to gold
sputter-coating. Luminescence emission spectra were recorded on a Horiba Yvon-Jacob FL3-22
fluorimeter using a quartz cuvette (excitation wavelength, 350 nm; emission wavelength range, 500-
670 nm; increment size, 1 nm; excitation and emission slit widths, 5 nm). 40 μL of EuC-doped PHB
replicas coated with PAH/PSS (ca. 1.5 × 10 9 particles mL -1 ) was dispersed in 160 μL of 5 mM
phosphate buffer solution of different pH for the luminescence measurements. Fluorescence
microscopy images were taken with an Olympus IX71 inverted fluorescence microscope. The samples
(1.5 μL) were placed onto glass slides and viewed using a 60× oil immersion objective.
Results and Discussion
As illustrated in Figure 1, PHB was first infiltrated into the nanopores of the MS particles in
chloroform solution. After removal of the chloroform solvent, the PHB-loaded MS particles were
dispersed in water by brief sonication. TEM images reveal that the PHB-loaded MS particles (Figure
2a) have the same morphology and diameter (2-4 μm) as the pristine MS template particles (data not
shown). No signs of PHB clustering were observed either in the solution or on the MS particle surface.
TGA was employed to estimate the PHB loading in the MS particles. The thermogravimetric analysis
of the PHB-loaded particles showed a mass loss of 31.5% at 215 ºC (Figure S1). Hence, it can be
estimated that 1 g of MS particles can load ca. 0.46 g of PHB. Based on the TEM and TGA data, we
82

conclude that the hydrophobic PHB can be effectively loaded in the nanopores of the MS particles
from organic solvent (chloroform).
Figure 2. (a) TEM image of PHB-loaded MS particles. (b) Optical micrograph of PHB replicas coated
with PAH/PSS dispersed in water. (c) TEM images of PHB replicas coated with one PAH/PSS bilayer.
The inset is a higher magnification image. (d) SEM images of the PHB replica coated with one
PAH/PSS bilayer. A higher magnification image of the surface of the replica is shown in the inset.
PHB replicas were obtained after removal of the MS template particles, indicating that the van der
Waals interaction among the hydrophobic polymer chains plays a significant role in maintaining the
morphology of the particles in aqueous media. This is in contrast to our previous reports that use water
soluble macromolecules, where cross linking of the polymer chains in the porous particles is required
to obtain a stable polymer network in solution. 9,10 However, the replica PHB particles easily
aggregated in water due to the hydrophobic nature of PHB, hampering their application as colloidal
particles in aqueous media. To avoid such aggregation, a PEM (PAH/PSS) coating was assembled on
the surface of the PHB-loaded MS particles before MS template removal. The PEM shell provides the
dual role of stabilizing the PHB replicas from aggregation and functionalizing the particle surface.
The polyelectrolyte coating was assembled through alternately depositing PAH and PSS. ζ-potential
measurements were used to monitor the variation in surface charge of the MS particles after PHB
infiltration and polyelectrolyte coating (Figure S2). The PHB-loaded MS particles had a ζ−potential of
−25mV in water. After deposition of each layer of PAH and PSS, the ζ−potential of the particles
alternated between ca. 30 mV (corresponding to PAH deposition) and ca. −40 mV (corresponding to
PSS deposition), indicating the consecutive deposition of PAH and PSS, respectively.
83

Figure 3. (a) Transmission and (b) fluorescence microscopy images of EuC-doped PHB replicas
coated with one PAH/PSS bilayer. The insets in (a) and (b) are photographs of the particles after
centrifugation in water and exposed to sunlight and ultraviolet light, respectively. The EuC incubating
concentration was 0.5 mg mL -1 .
After removal of the silica template, PAH/PSS-coated PHB replicas readily dispersed in water, as
observed by optical microscopy (Figure 2b). The particles have a diameter of 2-4 μm, similar to the
size of the MS template spheres. Figure 2c shows a TEM image of the PHB replicas. The diameter of
the PHB replicas decreased to 1-2 μm and a rough particle surface was observed under high
magnification conditions (insert in Figure 2c). Figure 2d shows the SEM images of the PHB replicas
coated with PAH/PSS, which is a free-standing, noncollapsed sphere. The porosity on the surface is
apparent at higher magnification (Figure 2d inset). This noncollapsed structure is in contrast to what is
typically observed for polyelectrolyte capsules prepared via the sequential assembly of polymers on
the surface of MS particles without PHB loading, where collapsed, hollow structures are observed. 20
FTIR was also used to monitor the composition of the particles. For the PHB-loaded MS particles, a
strong and broad absorption band at 1065 cm -1 was observed (Figure S3). The band can be assigned to
the Si-O stretching vibration. The peak at 1723 cm -1 is attributed to the C=O stretching vibration of
PHB. 21 No absorption band at 1065 cm -1 was observed for the replicas, suggesting complete removal
of the silica. The spectrum of the PAH/PSS-coated PHB replicas is similar to the spectrum of pure
PHB. The absorption peak at 1007 cm -1 originates from the SO 3 - symmetric vibration of PSS. 22
Luminescent materials are widely employed in fields as diverse as biology and information
technology. Commonly sought properties for luminescent materials are brightness and stability. 23 EuC
has emerged as a novel luminescent material because of its unique spectral characteristics, including
sharp emission peaks, long lifetimes, high fluorescence quantum yields, and a large Stokes shift.
However, the luminescence of EuC is not stable in aqueous solution because coordination of −OH
groups extensively reduces its luminescence emission intensity and decay time. Our aim was to
immobilize the EuC into the PHB replicas to construct particles with stable and intense luminescence.
84

Figure 4. (a) Energy transfer model from the ligand to the Eu III luminescent center in a PHB replica
doped with EuC. PHB serves as an efficient cosensitizer for the Eu III ion. (b) Luminescence emission
spectra of EuC-doped PEM capsules composed of three bilayers of PAH/PSS (black line) and PHB
replicas coated with one PAH/PSS bilayer (red line) dispersed in water (λ excitation = 350 nm). The EuC
incubating concentration was 0.05 mg mL -1 .
The EuC-doped PHB replicas coated with PAH/PSS were visualized in situ by fluorescence
microcopy (Figure 3). From the images, the replicas are observed to be spherical with a diameter of 2-
4 μm, and no aggregation was observed in both transmission and fluorescence modes. The intact
replicas show no shrinkage or swelling after loading EuC in ethanol, suggesting the PHB replicas have
good stability in both aqueous and alcohol solutions. The bright luminescent spheres confirm the
effective loading of EuC in the replicas (Figure 3b). The insets in Figure 3a and 3b show the images of
the EuC-doped PHB replicas coated with PAH/PSS in water after centrifugation and exposed to
sunlight and ultraviolet light, respectively. The intense monochromatic red luminescence of the EuCdoped
replicas is easy to detect by a laboratory ultraviolet lamp at 365 nm (Figure 3b inset).
The intense luminescence observed from the EuC-doped PHB replicas originates from the loaded
EuC. Figure 4a displays an energy transfer model for the ligands, PHB, and the central rare-earth ions
in the replica. Firstly, the β-diketone ligand TTA is excited by ultraviolet light. Efficient
intermolecular energy transfer from the ligand to the luminescent center and subsequent highly
efficient emission from the excited state of Eu III to the low level state endow the EuC-doped replicas
with luminescent properties. This is the so-called “antenna effect”. 24 A detailed energy level diagram
for Eu III along with the singlet and triplet levels of TTA is given in Figure S4. Here, the PHB also
serves as an efficient cosensitizer for the Eu III ions. 12
85

Figure 4b shows the luminescence emission spectrum of EuC-doped PHB replicas coated with
PAH/PSS There are four characteristic peaks (580 nm, 595 nm, 613 nm and 650 nm), corresponding
to the electric dipole transitions of 5 D 0 → 7 F J (J = 0, 1, 2 and 3), which is similar to that of EuC in
ethanol solution. The red luminescence observed is mainly attributed to the 5 D 0 → 7 F 0 transition
around 613 nm.
25
Intensity (/10 6 ) / a.u.
20
15
10
5
0
0.00 0.05 0.10 0.15 0.20 0.25 0.30
C EuC
/ mg mL -1
Figure 5. Luminescence intensity of EuC-doped PHB replicas coated with one PAH/PSS bilayer
dispersed in water as a function of the EuC incubating concentration. The luminescence intensity was
recorded at 613 nm (λ excitation = 350 nm).
The EuC content in the PHB particles was studied at different EuC incubating concentrations. PHB
replicas coated with PAH/PSS were dispersed in ethanol solutions of EuC of different concentration
for 16 h. After removal of the ethanol supernatant, and three water washing cycles, the luminescence
intensity of the particles was measured in water. Figure 5 plots the relationship between the
luminescence intensity and the EuC incubating concentration. The luminescence intensity linearly
increases at EuC incubating concentrations in the range 0.01−0.05 mg mL -1 . The luminescence
intensity of the PHB replicas obtained using 0.05 mg mL -1 EuC incubating solutions is equivalent to
that of neat EuC in bulk ethanol with a concentration of 6.05 × 10 -3 mg mL -1 . According to the TGA
results, 1 g of MS can adsorb 0.46 g of PHB, and hence the EuC doping is ca. 0.5% (EuC/PHB, w/w).
At this incubating concentration, the luminescence intensity of the EuC-doped PHB replicas is more
than 20 times higher than that of the EuC-doped PAH/PSS multilayer capsules (3 bilayers) templated
by the same MS particles (Figure 4b). With further increasing the EuC concentration, the
luminescence intensity increases only marginally. This is likely caused by concentration quenching in
the EuC-doped particles.
86

20
Intensity (/10 6 ) / a.u.
15
10
5
0
0 2 4 6 8 10 12 14
pH
Figure 6. Luminescence intensity of EuC-doped PHB replicas coated with one PAH/PSS bilayer and
dispersed in a 5 mM phosphate buffer solution of different pH. The EuC incubating concentration was
0.05 mg mL -1 . The luminescence intensity was recorded at 613 nm (λ excitation = 350 nm).
The stability of EuC in the PHB particles was also investigated. The saturated complexes can
effectively preserve the rare-earth ions from being attacked by −OH groups in solution. 25 Here, PHB is
used to coordinate the EuC to form saturated EuC complexes in the PHB particles. Figure 4a
illustrates the EuC coordinated to the PHB backbone and the ligands. The formation of saturated
complexes in the hydrophobic PHB particles protects and stabilizes the EuC in aqueous solution.
Figure 6 shows the luminescence intensity of EuC-doped PHB replicas coated with PAH/PSS in
phosphate buffer solution of different pH. EuC-doped PHB replicas can emit intense luminescence
under ultraviolet radiation over a wide pH of 3~11. No measurable decrease of the luminescence of the
EuC-doped PHB particles was observed after several months. However, the luminescence of EuCdoped
PEM (e.g., PAH/PSS) capsules dispersed in water becomes weaker and weaker, suggesting that
the PHB stabilizes the EuC through coordination. As the EuC is water insoluble, uncharged and has a
low molecular weight, it has similar physical properties to some hydrophobic drugs. Therefore, the
PHB replicas are interesting systems for the loading of hydrophobic (e.g., anticancer) drugs. This is an
important property for the EuC-doped PHB replicas to find application in various applications.
Conclusions
We have demonstrated the preparation of PHB replicas through templating MS particles, and their
subsequent surface coating with PEMs and doping with rare-earth complexes to obtain bright and
stable luminescent PHB particles dispersed in aqueous media. Hydrophobic PHB was effectively
assembled in MS particles and intact PHB replicas are obtained after removal of the silica template.
The broad range of MS materials available with tunable size, morphology and porosity will enable the
preparation of PHB materials with tailored dimensions and morphologies. The surface properties of
the PHB replicas can be easily tuned by assembly of a PEM coating. The PEM provides the dual role
of preventing aggregation of the PHB replicas in aqueous solution and also imparting functionality
through the polymer coating. The biodegradable PHB replicas can be effectively used to coordinate
rare-earth complexes. The EuC-doped PHB replicas emit intense luminescence over a wide range of
pH (3~11) and for at least several months in aqueous solution, making them potentially useful for a
range of bioapplications.
Acknowledgements: This work was supported by the Australian Research Council under the
Federation Fellowship and Discovery Project schemes. Jiwei Cui acknowledges the China Scholarship
Council for a Joint-Education PhD Program Scholarship. Philipp Senn and Cameron R. Kinnane are
thanked for the SEM measurements.
87

Supporting Information Available: Thermogravimetric analysis of PHB-loaded MS particles; ζ-
potential of the PAH/PSS-coated PHB particles as a function of the number of polyelectrolyte layers
deposited; FTIR spectra of PHB-loaded MS, PHB replicas coated with PAH/PSS, and neat PHB; and
energy level diagram for Eu III and the singlet and triplet levels of TTA.
References:
(1) (a) Kresge, T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C.; Beck, J. S. Nature 1992, 359, 710-
712.
(2) For reviews see: (a) Stein, A. Adv. Mater. 2003, 15, 763-775. (b) Yang, H.; Zhao, D. J. Mater.
Chem. 2005, 15, 1217-1231. (c) Hartmann, M. Chem. Mater. 2005, 17, 4577-4593. (d) Trewyn,
B. G.; Slowing, I. I.; Giri, S.; Chen, H.-T.; Lin, V. S.-Y. Accounts Chem. Res. 2007, 40, 846-853.
(e) Balas, Vallet-Regi, F.; Arcos, D. Angew. Chem. Int. Ed. 2007, 46, 7548-7558. (f) Wang, Y.;
Angelatos, A. S.; Caruso, F. Chem. Mater. 2008, 20, 848-858. (g) Wang, Y.; Price, A. D.; Caruso,
F. J. Mater. Chem. 2009, DOI: 10.1039/b901742a.
(3) (a) Wang, Y.; Caruso, F. Chem. Commun. 2004, 1528-1529. (b) Wang, Y.; Caruso, F. Chem.
Mater. 2005, 17, 953-961. (c) Deng, Y.; Qi, D.; Deng, C.; Zhang, X.; Zhao, D. J. Am. Chem. Soc.
2008, 130, 28-29. (d) Price, A. D.; Zelikin, A. N.; Wang, Y.; Caruso, F. Angew. Chem. Int. Ed.
2009, 48, 329-332. (e) Johnston, A. P. R.; Lee, L.; Wang, Y.; Caruso, F. Small 2009, 5, 1418-
1421.
(4) (a) Yu, C.; Fan, J.; Tian, B.; Zhao, D.; Stucky, G. D.; Adv. Mater. 2002, 14, 1742-1745. (b) Dong,
A.; Ren, N.; Tang, Y.; Wang, Y.; Zhang, Y.; Hua, W.; Gao, Z. J. Am. Chem. Soc. 2003, 125,
4976-4977. (c) Wang, Y.; Caruso, F. Adv. Funct. Mater. 2004, 14, 1012-1018.
(5) Crowley, T. A.; Ziegler, K. J.; Lyons, D. M.; Erts, D.; Olin, H.; Morris, M. A.; Holmes, J. D.
Chem. Mater. 2003, 15, 3518-3522.
(6) Tian, B.; Liu, X.; Yang, H.; Xie, S.; Yu, C.; Tu, B.; Zhao, D. Adv. Mater. 2003, 15, 1370-1374.
(7) Ryoo, R.; Joo, S. H.; Kruk, M.; Jaroniec, M. Adv. Mater. 2001, 13, 677-681.
(8) (a) Kageyama, K.; Tamazawa, J.-I.; Aida, A. Science 1999, 285, 2113-2115. (b) Yilmaz, E.;
Ramsrtöm, O.; Möller, P.; Sanchez, D.; Mosbach, K. J. Mater. Chem. 2002, 12, 1577-1581.
(9) Wang, Y.; Yu, A.; Caruso, F. Angew. Chem. Int. Ed. 2005, 44, 2888-2892.
(10) (a) Wang, Y.; Caruso, F. Adv. Mater. 2006, 18, 795-800. (b) Wang, Y.; Caruso, F. Chem. Mater.
2006, 18, 4089-4100. (c) Wang, Y.; Angelatos, A. S.; Dunstan, D. E.; Caruso, F.
Macromolecules 2007, 40, 7594-7600. (d) Wang, Y.; Bansal, V.; Zelikin, A. Z.; Caruso, F. Nano
Lett. 2008, 8, 1741-1745.
(11) Köse, G. T.; Ber, S.; Korkusuz, F.; Hasirci, V. J. Mater. Sci. - Mater. Med. 2003, 14, 121-126.
(12) Kai, J.; Parra, D. F.; Brito, H. F. J. Mater. Chem. 2008, 18, 4549-4554.
(13) (a) Bekiari, V.; Lianos, P. Adv. Mater. 1998, 10, 1455-1458. (b) Cui, J.-W.; Zhang, R.-J.; Lin, Z.-
G.; Li, L.; Jin, W.-R. Dalton Trans. 2008, 895-899.
(14) (a) Mathis, G. J. Biomol. Screen 1999, 4, 309-313. (b) Mathis, G. Clin. Chem. 1993, 39, 1953-
1959. (c) Saha, A. K.; Kross, K.; Kloszewski, E. D.; Upson, D. A.; Toner, J. L.; Snow, R. A.;
Black, C. D. V.; Desai, V. C. J. Am. Chem. Soc. 1993, 115, 11032-11033.
(15) (a) Maas, H.; Currao, A.; Calzaferri, G. Angew. Chem. Int. Ed. 2002, 41, 2495-2497. (b) Petoud,
S.; Cohen, S. M.; Bünzli, J.-C. G.; Raymond, K. N. J. Am. Chem. Soc. 2003, 125, 13324-13325.
(16) Selvin, P. R. Nat. Struct. Biol. 2000, 7, 730-734.
(17) Schulz-Ekloff, G.; Rathouský, J.; Zukal, A. Int. J. Inorg. Mater. 1999, 1, 97-102.
(18) (a) Melby, L. R.; Rose, N. J.; Abramson, E.; Caris, J. C. J. Am. Chem. Soc. 1964, 86, 5117-5125.
(b) Charles, R. G.; Ohlmann, R. C. J. Inorg. Nucl. Chem. 1965, 27, 255-259.
(19) (a) Decher, G.; Hong, J. D. Ber. Bunsen-Ges. Phys. Chem. 1991, 95, 1430-1433. (b) Decher, G.
Science 1997, 277, 1232-1237. (c) Donath, E.; Sukhorukov, G. B.; Caruso, F.; Davis, S. A.;
Möhwald, H. Angew. Chem. Int. Ed. 1998, 37, 2201-2205. (d) Cui, J.; Liu, Y.; Hao, J. J. Phys.
Chem. C 2009, 113, 3967-3972.
(20) Yu, A.; Wang, Y. Barlow, E.; Caruso, F. Adv. Mater. 2005, 17, 1737-1741.
(21) (a) Li, X.; Loh, X. J.; Wang, K.; He, C.; Li, J. Biomacromolecules 2005, 6, 2740-2747. (b) Zhang,
J.; Sato, H.; Furukawa, T.; Tsuji, H.; Noda, I.; Ozaki, Y. J. Phys. Chem. B 2006, 110, 24463-
24471.
88

(22) (a) Cho, J.; Quinn, J. F.; Caruso, F. J. Am. Chem. Soc. 2004, 126, 2270-2271. (b) Caruso, F.;
Furlong, D. N.; Ariga, K.; Ichinose, I.; Kunitake, T. Langmuir 1998, 14, 4559-4565.
(23) Burns, A.; Ow, H.; Wiesner, U. Chem. Soc. Rev. 2006, 35, 1028-1042.
(24) (a) Sabbatini, N.; Guardigli, M.; Lehn, J.-M. Coord. Chem. Rev.1993, 123, 201-228. (b) Bredol,
M.; Kynast, U.; Ronda, C. Adv. Mater. 1991, 3, 361-367.
(25) Gago, S.; Fernandes, J. A.; Rainho, J. P.; Sá Ferreira, R. A.; Pillinger, M.; Valente, A. A.; Santos,
T. M.; Carlos, L. D.; Ribeiro-Claro, P. J. A.; Gonçalves, I. S. Chem. Mater. 2005, 17, 5077-5084.
Supporting Information:
100
95
Mass loss / %
90
85
80
75
70
31.5 %
65
0 100 200 300 400 500 600
Temperature / o C
Figure S1. Thermogravimetric analysis of PHB-loaded MS particles.
40
PAH
PAH
ζ-potential / mV
20
0
-20
-40
PHB
PAH
PSS
PSS
PSS
1 2 3 4 5 6 7
Layer number
Figure S2. ζ-potential of the PAH/PSS-coated PHB particles as a function of the number of
polyelectrolyte layers deposited.
89

a
b
Absorbance
c
1800
1650
1500
1350
1200
1050
900
Wavenumber / cm -1
Figure S3. FTIR spectra of (a) PHB-loaded MS, (b) PHB replicas coated with PAH/PSS, and (c) neat
PHB.
S 1 5D 1
S 1
T 1
5D 1
5D 0
7F 2
S 0
TTA
Eu Ⅲ
7F 0
Figure S4. Energy level diagram for Eu III and the singlet and triplet levels of TTA.
90

Grain refinement of pure magnesium by back pressure equal channel
angular pressing at room temperature
Ji-zhong Li 1 *
* Presenter
1. School of Materials and Metallurgy, Northeastern University, Shanyang 110004, China
Systematic investigation on feasibility of refining grain structure in commercial pure
magnesium was conducted by means of back pressure equal channel angular processing (BP-
ECAP) at temperatures from 200 °C to room temperature. It was clearly shown that the grain
refinement of commercial pure Mg casting ingot can be successfully achieved from greater
than 800 µm down to submicron scale with 90° channel die, 100 MPa back pressure and
2mm/min pressing speed at room temperature. The temperature effect on the microstructure
and mechanical behaviour in compression was discussed.
Keywords Back Pressure, Equal Channel Angular Pressing, Pure Magnesium, Mechanical
properties, Twinning
Introduction
Materials with ultra-fine and nano-scaled grain structures have attracted growing interest during
the past decade because of their unique mechanical properties in comparison of their coarse grained
ones [1-4]. A popular method to produce such a material is severe plastic deformation (SPD) as grain
refinement of a metallic material can be achieved through large amount of plastic deformation.
Although there are several SPD techniques, such as equal channel angular processing (ECAP), high
pressure torsion (HPT) and accumulated roll bonding, ECAP is an effective practice to refine large
grains of a bulk metallic material into ultra-fine or nano-scaled level though one or multi-pass in an
angular die [5, 6]. To date there are significant ECAP works on materials with the fcc and bcc
structures [7-10], but limited researches on hcp structural materials that have been claimed as hard-todeformed
metals, such as Mg and Ti [11,12].
Mg and its alloys, having the low density and workability, are very attractive for automotive,
aviation and electronic industries [13, 14]. However, their applications have been restricted because of
their low ductility determined by the hcp structure, in which only a limited number of slip systems
exists. To improve their ductility while retaining or enhancing their strength, refining their gain
structures has been an objective in recent studies. ECAP has recently been effectively used in refining
grains of Mg and its alloys with enhanced ductility, strength and super plasticity [5-6, 15-18].
However, in the most works, the processing temperature used was at or higher than 200 ˚C. Otherwise,
during ECAP processing the material cracked into many segments along shearing plan which was
about 45˚ to the longitudinal direction [11, 12]. At such a temperature, grain size could not be refined
smaller than 5 µm as recrystallisation and grain growth occurred. By means of BP-ECAP, submicron
grain size in AZ31 was obtained after 4 passes at 150 ˚C plus another 4 passes at 100 ˚C following the
Bc route [6]. In the current work, refinement of gain structures in commercial purity Mg was initiated
using BP-ECAP at a range of temperatures from 200 ˚C to room temperature. For comparison purpose
a specimen was also ECAP pressed at 250 °C without back pressure. The microstructures and the
mechanical properties at room temperature were also investigated.
Experimental procedure
Specimens of 9 mm diameter with a length of 70 mm were machined from a commercial-purity
Mg ingot (purity >99.5%). ECAP die was consisting of two round channels with the intersect-angle of
90˚. The die was heated to a pressing temperature after a specimen was inserted into the entrance
channel. Graphite lubricant was used between the specimen and the channel surface. During ECAP
processing, a back pressure of 100 MPa was applied and low speed of 2 mm/min was used. Each
specimen was pressed for 4 passes with route C at a given temperature.
91

Microstructures in the as-cast and BP-ECAPed materials were observed using optical
microscopy (OM) and scanning electron microscopy (SEM). The samples for OM and SEM were cut
perpendicular to the longitudinal direction and prepared by standard metallographic procedures
followed by chemical etching using a solution of 10 ml nitric acid, 30 ml acetic acid, 40 ml distilled
water and 120 ml ethanol. Average grain sizes of the as-cast and BP-ECAPed materials were obtained
following the linear method.
Compression specimens of 3 mm x 3 mm x 6 mm were cut along the longitudinal direction of
the BP-ECAPed samples. Compression tests were performed at room temperature with an initial strain
rate of 1×10 -3 s -1 . For each material at least three compression tests were performed. Microstructures on
the samples after the compression tests were also observed along cross section.
Results and discussion
Fig. 1 showed that the majority grains in the as-cast Mg were equiaxed and very coarse with
irregular morphology. The average grain size was larger than 800 µm. Fig. 2 (a) displayed the
appearances of a sample after 1 pass through the die at 250 °C without back pressure. Fracture
segments distributed uniformly along the sample and the cracks went though the bottom to top along
shearing direction which was about 45° to the extruded direction. In contrast, all the samples after BP-
ECAP at temperatures ranging from 200 °C down to room temperature with 23 °C interval showed no
cracks at all. Fig. 2 (b) exhibited the sample after 4 passes at room temperature. It demonstrated that
applying back pressure was significantly helpful in ECAP processing of hard-to-deform metals at a
relatively low temperature or even at room temperature as an appropriate back pressure can effectively
prevent the material from cracking during ECAP processing.
Fig.1 Optical microstructure in pure cast ingot
Fig. 2 Appearance of pure Mg billets: (a) 250 ºC
without back pressure; (b) 23 ºC withbackpressure
Fig. 3 (a) through (c) presented the typical microstructures under OM for the materials after 4
passes at temperatures of 200, 150 and 100 °C, respectively, and Fig. 3 (d) revealed the representative
microstructure under TEM for the material after 4 passes at room temperature. The grain refinement
was significant. Especially, the grain size was refined into submicron level when the BP-ECAP
process was carried out at room temperature. The average grain sizes of the materials were listed in
Table 1.
92

Fig. 3 The optical and SEM microstructure of pure Mg: (a) 200 ºC; (b) 150 ºC; (c) 100 ºC; (d) room
temperature (TEM).
Table 1 average grain size of different temperature
Temperature (°C) 200 150 100 23
Average grain size (µm) ~8 ~6 ~2 ~0.8
Another clear feature was that twining existed in both as-cast and the materials BP-ECAPed at
200 °C with the grain size of 10 µm. But no twining was detected when the materials BP-ECAPed at
the temperature lower than 200 °C and having the grain size of 5 µm or less. Mechanism behind this
phenomenon was not clear yet and further investigation was necessary.
Fig. 4 showed the true stress-strain curves of the materials. It can be seen that the as-cast
material showed almost no ductility and the low yield strength (0.2% proof stress) of about 40 MPa.
The yield strength was increased as the grain size refined into 5 to 1 µm and then decreased when
grain size was within submicron level. The material after 4 passes at 200 °C displayed good ductility
and highest ultimate strength among those materials. Both the as-cast and the 200 °C processed
materials had high strain hardening rate. On the other hand, the other BP-ECAPed materials showed
almost no strain hardening. The compression behaviours of those materials also displayed obviously
that the plastic deformation after yielding increased to about 15% as the grain size reduced to
submicron scale. This indicated that grain boundary sliding and texture softening played dominant role
and contributed to plastic deformation extensively.
Another interesting point is that the material processed at 200 °C showed double yield strengths,
as indicated by arrows. It can be explained that the first yielding was attributed to twins as the twins
occurred only in the grains at low strain, and the second one was due to the combination effect of
twins and dislocation movement which had contribution to further strain hardening and some ductility.
93

Fig. 4 The true stress-strain curves for cast state and ECAPed with 4 passes at different temperatures.
Fig. 5 (a) though (c) showed the OM microstructures of the materials processed at 200, 150 and
100 °C after compression tests and Fig. 5 (d) exhibited the microstructure of the material processed at
room temperature. Fig. 5 demonstrated that there were no twins when the grain size was less than 5
µm, but some twins existed in the grains about 5 µm or larger, as in Fig. 5 (b). Combining this feature
with the compression behaviours in Fig. 4, it might conclude that strain hardening was, at least
partially, attributed to mechanical twining. In addition, the high yield stress and low ductility for the
materials ECAPed at 150 and 100 ˚C may due to formation of texture during the BP-ECAP. Cisar [10]
and Kim [19] etc have reported that ECAP processing makes basal plan parallel to extrusion direction
results in Schmid factor closed to 0, which is also makes yield strength increase and ductility decrease.
Fig. 5 The microstructure of compressive deformed samples: (a) 200 ºC; (b) 150 ºC; (c) 100 ºC; (e)
room temperature (SEM).
94

Conclusions
(1) The large grains in the as-cast pure Mg were successfully refined into submicron sized grains
using BP-ECAP at room temperature without cracks.
(2) The compressive tests at room temperature showed that the yield strength might decrease as the
grain size was reduced into submicron level. However, the ductility was significantly enhanced
due to submicron grain size.
(3) Strong strain hardening was attributed to large grain sizes, which was larger than 10 µm.
Deformation twining, might contribute to the yield strength for the pure Mg materials having
grains larger than 5 µm.
References:
[1] Y.T. Zhu, J.Y. Huang, J. Gubicza, T. Ungár, Y.M. Wang, E. Ma, R.Z. Valiev, J. Mater. Res. 18
(2003) 1908.
[2] D. Jia, Y.M. Wang, K.T. Ramesh, E. Ma, Y.T. Zhu, R.Z. Valiev, Appl. Phys. Lett. 79 (2001) 611.
[3] R. Valiev, Nature Mater. 3 (2004) 511.
[4] Z. Horita, K. Ohashi, T. Fujita, K. Kaneko, T.G. Langdon, Adv. Mater. 17 (2005) 1599.
[5] K. Matsubara, Y. Miyahara, Z. Horita, T. G. Langdon. Acta Mater. 51 (2003) 3073.
[6] K. Xia , J. T. Wang, X. Wu, G. Chen, M. Gurvan. Mater. Sci. Eng. A 410-411 (2005)324.
[7] C. Mallikarjuna, S. M. Shashidhara, U. S. Mallik. Mater. Design. 30 (2009) 1638.
[8] P. L. Sun, P. W. Kao, C. P. Chang. Mater. Sci. Eng. A. 283 (2000) 82.
[9] M. Furukawa, Y. Ma, Z. Horita, M. Nemoto, R. Z. Valiev, T. G. Langdon. Mater. Sci. Eng. A. 241
(1998) 122.
[10] L. Cisar, Y. Yoshida, S. Kamado, Y. Kojima, F. Watanabe. Mater. Sci. Forum. 419-4 (2003) 249.
[11] E. J. Kwak, C. H. Bok, M. H. Seo, H. S. Kim. Mater. Trans. 49 (2008) 1006.
[12] X. Zhao, W. Fu, X. Yang, T. G. Langdon. Scripta Mater. 59 (2008) 542.
[13] J. Koike, R. Ohyama. Acta Mater. 53 (2005) l963.
[14] N. Ogawa, M. Shiomi, K. Osakada. Int J Mach Tool Manu. 42 (2002) 607.
[15] W. M. Gan, M. Y. Zheng, H. Chang, B. Schwebke. J. Alloys. Comp. 470 (2009) 256.
[16] A. Yamashita, Z. Horita, T. G. Langdon. Mater. Sci. Eng. A 300 (2001) 142.
[17] K. Matsubara, Y. Miyahara, Z. Horita, T. G. Langdon. Metall. Mater. Trans. A 35 (2004) 1735.
[18] S. Y. Chang, S. W. Lee, K. M. Kang, S. Kamado, Y. Kojima. Mater. Trans. 45 (2004) 488.
[19] W. J. Kim, S. I. Hong, Y. S. Kim, S. H. Mind, H. T. Jeong, J. D. Lee. Acta Mater. 51 (2003) 3293.
95

Dielectrophoretic Separation of Carbon Nanotubes and Polystyrene
Microparticles [1]
C. Zhang 1 * , K. Khoshmanesh 2 , F. J. Tovar-Lopez 1 , A. Mitchell 1 , W. Wlodarski 1 , K. Klantarzadeh
1
* Presenter
1. School of Electrical and Computer Engineering, RMIT University, Australia
2. School of Engineering and IT, Deakin University, Australia
The separation of multi-walled carbon nanotubes (MWCNTs) and polystyrene
microparticles using a dielectrophoresis (DEP) system is presented. The DEP system consists
of arrays of parallel microelectrodes patterned on a glass substrate. The performance of the
system is evaluated by means of numerical simulations. The MWCNTs demonstrate a
positive DEP behaviour and can be trapped at the regions of high electric field. However, the
polystyrene microparticles demonstrate a negative DEP behaviour at a certain range of
frequencies and migrate to the regions of low electric field. Experiments are performed on the
microparticles at the frequencies between 100 Hz and 1 MHz to estimate their crossover
frequency and select the range of separation frequencies. Further, experiments are conducted
at the obtained range of separation frequencies to separate the MWCNTs and polystyrene
microparticles.
Keywords Diekectrophoresis, Particle separation, Carbon nanotube, Polystyrene
microparticle
1. Introduction
When a neutral particle is placed in a non-uniform electric field, it experiences a translational
force due to the polarization effect of the particle [1]. Pohl [2] was one of first who recognized and
explored the applications of this force, known as dielectrophoretic (DEP) force, for manipulating and
alignment of particles.
DEP force could be positive or negative. Positive DEP force traps particles into the regions with
high electric field gradient (e.g. the corners of electrodes), and negative DEP force pushes particles
into weak field regions (e.g. the spaces between electrodes) [3]. Due to the different trapping effects of
positive and negative DEP forces, particles with different properties (i.e. dielectric constant, size, and
conductivity) could respond differently to the applied electric fields with different oscillation
frequencies. DEP can be employed in a number of microfluidic technology applications including
particle separation and sorting [1, 4-6], particle trapping and assembling [7-11], as well as patterning
[12-14].
Due to their unique characteristics, one dimensional nanostructures have been extensively studied
and applied for a number of novel nanotechnological applications [15]. They have been employed in
developing electronic devices such as sensors [16, 17], oscillators [18] and transistors [19].
As the manipulation of nanostructures is most effectively conducted within liquid suspensions, it
is important to realize the methods for separating the desired nanostructures from the liquid mixture.
The application of DEP to particle discrimination, separation, and DEP field-flow fractionation was
reviewed in [1]. Numerical scheme based on the distributed Lagrange multiplier method was used for
studying the motion of the nano-sized particles of dielectric suspensions subjected to uniform and
nonuniform electric fields [3]. Cells with different size were separated using direct current DEP in [20].
The separation of polystyrene beads and cells by different dielectric fields was reported in [4], where
opposite DEP forces were used for accurately focusing a stream of beads and yeast cells at different
locations across the channel by adjusting potentials and frequencies. A microparticle
filter (i.e. polystyrene beads, yeast cells, spores and bacteria) was reported to be successfully
developed and characterized by Li et al [21] where DEP force was employed to capture desired
particles. Live and nonviable cells were separated due to their significant differences in DEP mobility
[1]. Published at Microfluidics and Nanofluidics (2009) 7:633-645

[22-25]. Sorting of normal and Babesia bovis infected erythrocytes was reported using DEP by
exploiting the higher ionic membrane permeability of Babesia bovis infected cells [5]. Separating and
sorting of other bioparticles including DNA [26, 27], proteins [28] and viruses [29-31] were also
reported. Additionally, the possibility of separating metallic and semi-conducting single-walled carbon
nano-tubes (CNTs) using DEP was theoretically investigated in [32] by applying Brownian dynamics
method, where conductivity difference produces different DEP behaviours for semiconducting and
conducting CNTs.
The DEP separations have been reported extensively. However, there is rarely any report
concerning the separation of nanostructures from microstructures. This technique can be implemented
in novel applications including: purifying cell suspensions with nanoscale impurities, separation of
organelles, and nanostructure assembling and separation for microelectronic fabrications. For instance,
nanotechnology approaches are recently applied for drug delivery to target cells. In these approaches,
drug materials are encapsulated within or coated around different nanoscale containers and vehicles
such as nanospheres [33], lipid coated nanoparticles [34], magnetic nanoparticles [35] and carbon
nano-tubes [36] to be delivered to the target cells. For assessing whether the delivery of drugs to cells
is achieved successfully or their therapeutic functionalities are effective, the cells are lysed and their
ingredients are released for analysis. Consequently, DEP techniques can play a major role in
separating the nanoscale particles and microscale organelles (such as nucleus and mitochondria) from
the lysed cells.
Both micro and nano structures can be employed as sensing elements in sensor development and
fabrication. Due to the large surface area to volume ratio of such materials, the developed devices are
highly sensitive; however, their selectivity should be improved. DEP trapping and separation
techniques enable the deposition of multiple and dissimilar materials on a single sensing device at
different locations, which can be used for the development of an array of sensors with multiple
sensitive layers.
In this work, MWCNTs were separated from polystyrene microparticles in stationary liquid by
using DEP. The polystyrene microparticles were mixed with MWCNTs in suspension. The mixture
was then placed on a micro-gapped interdigital transducer (IDT) and sealed with a
polydimethylsiloxane (PDMS) block. An alternating voltage with amplitude 10 V peak-to-peak and
frequencies from 100 Hz to 1 MHz was applied across the electrodes. An inverted microscope was
used to observe the behaviour of the particles. The observed DEP separation performance was
compared with computational fluid dynamic predictions.
2. DEP Theory
2.1 DEP Force
If damping and viscous forces are negated, the DEP force equation can be expressed as [37]:
* * 2
FDEP
= Γ ⋅ εm
⋅ Re[fCM
(ε
p,
εm
)] ⋅ ∇E
(1)
,
where, Γ is the geometric factor of the particles, ε m is the dielectric constant of suspending medium,
ε * m and ε * p are the complex permittivities of suspending medium and particles, respectively, f CM is the
Claussius-Mossotti (CM) factor, (also known as the complex polarisation factor), and E refers to the
alternating electric field. The complex permittivity is defined as:
σ
(2)
ε * = ε − i
ω ,
where, ω is angular frequency of the electric field, σ is electric conductivity and i=√-1.
Γ depends on the geometry of the particles that is expressed distinctly for cylindrical [20, 21] and
spherical structures [3] as below:
2
πr l
(3)
Γ = cylindrica l
6 , (4)
3
Γ
spherical
= 2 πr
,
where, r indicates the radius of the sphere or the cross section of the cylinder and l represents the
length of the cylindrical structures.
97

The CM factor also depends on the geometry of structures and is expressed differently for
cylindrical [38] in (5) and spherical [3] structures in (6) as below:
* *
ε
(5)
p
− ε
* *
m
f
CM
(ε
p
,ε
m
) =
*
ε
m ,
* *
ε
p
− ε
* *
m
f
CM
(ε
p
,ε
m
) =
* *
(6)
ε
p
+ 2ε
m
,
2.2 Characterisation of the CM factor for MWCNT and polystyrene microparticles
The DEP force takes the same sign of the real part of the CM factor, which can be either positive
or negative. To explore the ability of DEP force to sort polystyrene microparticles and WMCNTs,
calculations of the CM factor for various electric field frequencies were conducted using Equations
(1)-(6) for each of these particles suspended in DI water. Because of the significant size difference
between MWCNTs and polystyrene microparticles, the geometric factor Γ was also calculated and
taken into the consideration.
Firstly, the value of Г×Re[f CM ] of MWCNTs was calculated. The relative permittivity of
MWCNTs was taken as 2.5 while the conductivity was assumed between 1000 S/m and 150 S/m
according to [38, 39]. The diameter and length of MWCNTs were assumed to be 50 nm and 2 μm
respectively. The relative permittivity of the DI water is taken 78 and the conductivity is measured as
0.2 mS/m.
Figure 1 presents the value of Г×Re[f CM ] as a function of applied electric field frequency,
calculated for MWCNTs suspended in deionised (DI) water. Since the value stays above zero for all
frequencies considered in this investigation (1 kHz-100 MHz), MWCNTs always experience a
positive DEP force.
Next, Г×Re[f CM ] of polystyrene microparticles was calculated. This calculation is more
complicated than that of MWCNT as the conductivity of polystyrene must be carefully defined.
Polystyrene has a very low conductivity, and thus the inherent conductivity of material is zero.
However, placing the microparticles in DI water results in the inducing of surface conductance [40].
The total surface conductance can be modelled as two separate components: a conductance due to the
movement of charges in the diffuse double layer and another due to charge movements in the Stern
layer. As a result, the microparticle has an non-zero effective conductivity σ p , which could be
expressed as [41]:
2K
(7)
S
σ
p
=
r
where K s represents the surface conductance and r is the radius of the microparticle. K s can be
calculated as [41]:
r
(8)
⎜⎛
2
2
K = − + − ( − )( + )( ⋅ ) ⎟⎞
S
σ
m
9σ
m
4 ε
p
ε
m
ε
p
2ε
m
f
0
2π
4 ⎝
⎠
Here, f 0 is known as the crossover frequency where DEP force is zero. Therefore, at the crossover
frequency the value of K s has a linear relationship with the particle diameter if permittivity of the
medium and particles are constant.
The surface conductance has been calculated for different dimensions of latex beads [41].
According to the results, the value of K s is found to be 0.25 nS. In this case, the microparticle
conductivity is calculated to be 1 mS/m. The relative permittivity of microparticle is assumed to be 2.5.
Regarding these values, Г×Re[f CM ] was calculated as a function of frequency and presented in Figure 2.
It is evident that in this case, the microparticles experience positive DEP forces below 200 kHz, and
negative DEP force above 200 kHz.
98

Figure 1. Г×Re[fCM] vs. frequency for the MWCNTs in DI water; (up): log to log scale; (down):
normal scale.
By comparing Figures 1 and 2, it is evident that MWCNTs experience strong positive DEP force
for frequencies smaller than 1 MHz while polystyrene microparticles experience neutral or moderate
negative DEP force for frequencies larger than 200 kHz. As a result, it should be possible to use DEP
to separate the MWCNT and polystyrene microparticles, particularly by applying an electric field with
frequency in the range of 200 kHz to 1 MHz.
In this paper, the used MWCNTs were metallic and their conductivity was assumed to lie in a
range between 150 to 1000 S/m and the calculations and simulations were based on this assumption.
Semi-conductive carbon nanotubes (CNTs) have also been used in DEP experiments by other groups
[42]. However, due to their dielectric properties, they experience negative DEP force at high
frequencies. As, the crossover frequencies of polystyrene microparticles and the semi-conductive
MWCNTs are different, it is possible to perform DEP separation of these two materials for ranges
between the two crossovers frequencies.
99

Figure 2. Г×Re[fCM] vs. frequency for polystyrene microparticles in DI water, as the existence of
negative value, plot in log to log scale is not applied.
2.3 Particle Geometry and Conductivity
According to Equation (3) to (6), the DEP force is a strong function of particle geometry, as well
as particle conductivity. In order to identify the influence of these two parameter on the DEP force, the
Г×Re[f CM ] value of cylindrical and spherical particles with similar volumes, for both similar and
starkly different conductivities were calculated and compared as shown in Figure 3. DI water (relative
permittivity of 78, conductivity of 0.2 mS/m) was selected to be the particle suspending medium. In
Figure 3, it is assumed that cylindrical particles have the same dimensions as the tested MWCNTs
(diameter of 50 nm, length of 2 μm) and the spherical particle have a diameter of 0.2 μm. The relative
permittivity of both particles was assumed to be 2.5. Calculations were conducted for two different
sets of conductivities: 1 S/m and 1000 S/m.
Figure 3 shows that the DEP spectra of cylindrical particles with different conductivities follow
similar trends. However, the DEP force magnitudes are proportional to conductivity. The crossover
frequency is the same for both conductivities in the cylindrical structure. On the contrary, the spherical
particles with conductivity of 1000 S/m experience stable positive DEP force for the whole range of
investigated frequencies. This is because in equation (6), the CM factor converges to one for large
particle conductivities. However, the DEP force, applied on spherical particles with conductivity of 1
S/m, decreases when frequency is higher than 10 MHz. As a conclusion, the conductivity of spherical
particle plays a more important role in DEP separation, as it alters the crossover frequency.
3. Design and Simulation
The DEP platform consists of 8 arrays of microelectrodes, each containing 20 interdigitated
microgaps, as shown in Figure 4. The gap separation is 10 μm, and gaps are placed at a distance of
100 μm from each other.
100

Figure 3. The DEP spectrum of cylindrical and spherical particles with conductivities of 1000 S/m
and 1 S/m. The calculation was based on comparable particle size and relative permittvities.
Figure 4. Design of the DEP platform. (A): top overview of the DEP platform; (B): the arrays of
micro electrode gap; and (C): the close view of a gap section.
The 10 μm gap was chosen because it was less likely to be damaged when high potential AC
voltages were applied. The purpose to separate each gap by 100 μm is to maximize the difference
between strong field and weak field regions, in order to clearly distinguish between positive and
negative DEP behaviours. It also provides enough space for polystyrene microparticles to have
constant motions without colliding with the MWCNTs trapped in the gap. The distance between arrays
was designed to be 260 μm to minimize the field interference between arrays. Similar designs have
been extensively used in aligning and assembling of nanowires [37, 43, 44] for micro and nano
electronics applications.
The Computational Fluid Dynamics (CFD) method is applied to estimate the performance of the
DEP system. The CFD method enables us to analyse the distributions of the electric potential as well
as the electric field that are generated by the applied signals around electrodes.
101

The electric potential field is governed by the Poisson’s equation as follows:
2
∇ ϕ = 0 ,
(9)
where, φ is the electric potential and ∇ is the gradient operator. The electric field is obtained by
calculating the derivative of the electric potential field as follows:
E r = −∇ϕ , (10)
where, E r 2
is the electric field that is a vector variable. Having this variable, the value of ∇E was
obtained throughout the field and substituted into Equation (1) to calculate the DEP force in different
locations of the device.
The simulations were conducted in a two-steps process. First, the geometry of the device was
created and then divided into small elements using the Gambit-2.3 software package (Fluent USA,
Lebanon, NH). Along the electrode edges, fine structured quadrilateral elements were applied,
enabling the control of the thickness of these elements. Selecting the proper thickness of these
elements was crucial to predict the sharp gradients of the electric field along the electrodes and to
assure the convergence of the simulation. In the interior regions of the system, unstructured
quadrilateral elements were applied due to their flexibility.
Next, the Fluent-6.3 software package (Fluent USA, Lebanon, NH) was applied to solve the
governing Equations (9)-(10). The software applied the finite volume method to discretise the
governing equations across each element. The software was developed to solve the differential Navier-
Stokes equations in fluidic systems and predict the velocity, pressure and other flow variables.
However, other differential equations can be solved using the UDS (user-defined scalars) module of
the software.
The boundary conditions applied were as follows: the electrodes on one side of the device were
set at 5 V while the electrodes on the opposite side are grounded. The bottom, top and side walls of the
device were regarded as insulators, where the gradient of the electric potential was zero.
Figures 5-A, B and C show the gradient of the electric field square along x, y and z directions.
Figures 5-A and B show that the field has a much higher gradient at the corners of electrode tips.
Figure 5-C shows that the field gradient is positive in region between the electrode tips and is negative
at electrode edges. Therefore, particles with a positive CM factor are trapped when they are close to
the electrode tips, and are strongly levitated when are located within the intermediate region.
The simulation results assert that the design of this DEP platform is appropriate for the separation
experiments. The region between electrodes has a weak and uniformly distributed electric field, which
is suitable for producing negative DEP while the tips of electrodes have a sharp electric field, which is
suitable for producing positive DEP. Hence spatial separation of the particles will occur if some of
them experience negative DEP and the others experience the positive DEP forces.
4. Experimental
The DEP platform was microfabricated using photolithography on glass substrates. The glass
substrate was first metallised (Electron-beam evaporation technique) using Cr and Au with the
thickness of 50 nm and 150 nm, respectively. The AZ 1512 photoresist was then spin coated at 3000
rpm (acceleration of 1000 rpm/sec) for 25 seconds, and this procedure was followed by soft baking for
20 minutes at 90 °C. A mask aligner (Suss MJB3) was used to expose the mask pattern on to the
photoresist. The photoresist was developed and then the metal film was etched using reactive ion
etching. After patterning the sample was cleaned using isopropyl alcohol and acetone and then dried
with nitrogen.
102

A
B
C
Figure 5. The gradient of the electric field square along x, y and z directions obtained by the CFD
∂ ∂
simulations, which depicts the intensity of the DEP forces. (A): , (B): , (C):
∂
∂x
∂y
∂z
A PDMS block (30 mm×30 mm×10 mm) with a microfluidic channel (20 mm ×1 mm×80 μm)
was used to immobilise the particle suspension in the vicinity of the electrodes. To fabricate the
PDMS block, an 80 μm (channel height) KMPR1025 photoresist layer was coated onto a polished
silicon wafer, and photolithography was used for transferring the channel pattern onto the photoresist
layer. After developing and hard baking, liquid PDMS was applied on the surface, and cured on
hotplate at 100 ºC for 2 minutes. After PDMS was cured, 2 holes were punched in for liquid inlet and
E 2
E 2
E 2
.
103

outlet. The channel was assembled on to the DEP platform such that the interdigital electrodes were
sealed within the channel but with contacting pads exposed, as shown in Figure 6.
Figure 6. PDMS channel (80 μm) block for liquid sealing
MWCNTs (Sigma Aldrich, 636649), 20-50 nm in diameter, 0.5-2 μm in length, with single-wall
thickness of 1-2 nm were dispersed into DI water with an initial concentration of 0.1 mg/ml. The
suspension was stabilized with the aid of Triton X-305 surfactant (Sigma Aldrich, approximately 40 μl
surfactant per 10 ml suspension). The liquid was then mixed with polystyrene microparticles (Duke
Scientific, 1μm in diameter, and concentration of 1.8×10 10 ) in the ratio 1:1. The mixture was then
placed in a 125-W ultra-sonicator (Branson) bath for 45 minutes and left intact for 6 hours to allow
visible agglomerates to precipitate from the suspension. The suspension was then ready for DEP
experiments.
After the liquid suspension was injected into the microfluidic channel, alternating
voltages with amplitudes of 10 V peak-to-peak and frequencies from 100 Hz to 1 MHz was
applied across the electrodes. An inverted microscope was used to observe the behaviour of
the particles.
5. Result and Discussion
5.1 Selection of Voltage Amplitude
During DEP separation experiments, an AC voltage of 10 V peak to peak was employed as the
optimum amplitude. According to our preliminary tests, only AC voltages with the amplitudes higher
than 1 V can produce visible DEP behaviour for 1 μm polystyrene microparticles, and higher
amplitudes resulted in higher separation rates. However, applying AC voltages of more than 10 V
damaged and delaminated the electrodes, as shown in Figure 7. It is seen that the Au electrodes were
etched and the gap between them was extended by an extra 10 μm. According to this limitation, the
amplitude of 10 V peak to peak appeared to be the optimum voltage.
104

Figure 7. SEM image of damaged electrodes which used for performing DEP trapping with AC
voltage of over 10 V peak to peak. The tip of electrode was etched (circled part) during the experiment.
5.2 Verification of the Practical Crossover Frequencies
To verify the value of K s and hence the crossover frequencies obtained in Section 2.2, the DEP
experiments of polystyrene microparticles in DI water was carried out for frequencies in the range of
100 to 150 kHz. An inverted microscope was used for observing the particle behaviour.
Figure 8 present the DEP behaviour for different frequencies of applied electric field, the subfigures
show 100 Hz (A), 500 Hz (B), 1 kHz (C), 5 kHz(D), 10 kHz (E), 50 kHz (F), 150 kHz (G), 500
kHz (H) and 1 MHz (I). Positive DEP behaviour can be observed in Figures 8-A to 8-D, as the
microparticles are trapped at the edge of electrode. The decrease in the number of microparticles
trapped while frequency of applied field was increased from 100 Hz to 5 kHz is due to the decrease of
the DEP force magnitude. Figures 8-E and 8-F show the behaviour of microparticles at the frequencies
close to the crossover frequency, as the microparticles lost the tendency to get trapped at the edges of
the microelectrodes. At 150 kHz (as shown in Figure 8-G), a typical negative DEP behaviour of
microparticles is evident, as the microparticles are repelled from the microelectrodes and are clustered
in the spaces between them; chaining of microparticles is due to the polarization effect in the electric
field. There is not much difference in microparticle behaviours in Figures 8-H and I (500 kHz and 1
MHz) compared to that of Figures 8-G (150 kHz).
According to the theoretical calculations of Section 2.2, the crossover frequency should be
approximately 200 kHz, however, the results of Figure 8 suggest the crossover frequency was between
10 kHz to 50 kHz. This mismatching might be due to the fact that the assumed values for surface
conductance K s were not accurate. In [41], the author only calculated K s for the particles with diameter
in the range of 2 μm to 10 μm, K s was found to be equal to 1 nS from extrapolation . Additionally, the
measurement environment is different, the DI water was obtained from a different purifier,
microparticles are purchased from different supplier, and experiment temperatures could be varied
(both nominally room temperature). As small changes in surface conductance can give rise to large
changes in the DEP spectrum, the crossover frequency could be tuned by change of K s [40]. And when
microparticle conductance σ p was adjusted to 0.3 mS/m in the model of Section 2.2, theoretical and
practical results were matched.
105

Figure 8. Particle behaviours at frequencies from 100 Hz to 1 MHz. (A-D): positive DEP observed; (E
& F): close to crossover frequency. (G): typical negative DEP behaviour. (H & I): DEP behaviour of
particles at 500 kHz and 1 MHz is similar to that at 150 kHz.
5.3 Particle Separation
The ability to use DEP to sort particles was investigated by applying electric fields with
frequencies of 100 Hz, 1 kHz, 100 kHz and 150 kHz and examining the system for spatial separation
of the particles. The separation progress in the stationary liquid is shown in Figures 9 and 10.
At 100 Hz (Figure 9-A), strong positive DEP was observed. Large numbers of polystyrene
microparticles and MWCNTs were in constant motion in the region near the electrode tips. This was
due to the existence of large levitating and trapping forces described in Section 3. When the frequency
of applied field was increased to 1 kHz (Figure 9-B), microparticles in motion could no longer be
observed. As this frequency increased to values closer to the crossover frequency (approximately 30
kHz), the DEP force became weaker and the number of assembled microparticles decreased. Figure 9-
C illustrates initial separation behaviour of polystyrene microparticles and MWCNTs at 100 kHz, as
all polystyrene microparticles were repelled from strong electric field region, while MWCNTs stayed
on the edge of the electrodes. At 150 kHz (Figure 10-A), the separation of MWCNTs and polystyrene
microparticles was complete. Since the magnitude of DEP forces affecting MWCNTs and polystyrene
microparticles is not changing for frequencies larger than 100 kHz, applied signals in the range of 100
kHz to 1 MHz resulted in similar DEP behaviour (Figure 9-D illustrates the particle behaviour at 1
MHz).
After the polystyrene microparticles were removed and the liquid was evaporated, the substrate
was characterised by a scanning electron microscope (SEM) to observe the trapping effect of
MWCNTs; and as can be in Figure 10-B, MWCNTs were trapped on the tip and edges of electrodes.
106

Figure 9. Particle separation progress observed from the frequency of 100 Hz to 100 kHz. (A): at 100
Hz, both microparticles and MWCNTs were experiencing strong positive DEP forces; (B): at 1 kHz,
positive DEP forces were weaker, (C): at 100 kHz, MWCNTs were still experiencing positive DEP
forces, while microparticles were influenced by negative ones. (D): at 1 MHz, particle behaviours are
similar to that at 150 kHz.
A
B
Assembled
MWCNTs
Assembled
MWCNTs
Assembled
MWCNTs
20 µm
Slide 1
Figure 10. A: At 150 kHz, complete particle separation achieved, polystyrene microparticles were
repelled from electrode tips, while MWCNTs trapped at tips. B: SEM image of assembled MWCNTs
on the tip and edge of electrodes.
5.4 Collection of Separated Particles
According to the calculations presented in Section 2.2, the magnitude of the DEP force affecting
MWCNTs is much larger than polystyrene microparticles. Therefore, maintaining a proper flow in the
microfluidic channel, the polystyrene microparticles can be washed away and removed. MWCNTs can
be released when the alternating voltage is switched off. In this case, the collection of polystyrene
microparticle can be achieved by maintaining the flow in positive direction; and after that, the
107

MWCNTs can be collected by reversing the flow direction. The collecting process is shown in Figure
11.
Figure 11. The collecting process for polystyrene microparticles and MWCNTs. (A) Particles injected
in microfluidic channel. (B): DEP trapping when the electric field is applied on the electrodes. (C):
Collection of polystyrene microparticles with the fluidic flow. (D): Releasing captured MWCNTs by
switch off the applied electric field. (E): Collection of MWCNTs by reversing the flow direction.
5.5 More Discussions
As in this work, all particles used for experiment were non-bioparticles, there are no limitations.
However, causation should be experienced when the electric field stresses the live cells membrane
[45]. Experimentally, strong electric field (large amplitude of alternating voltage) can generate large
DEP force, which improves the separation efficiency. However, cells are damaged just in a short time
after they are exposed to a very high electric field level [20] specially near the electrode tips.
108

The separation efficiency can be improved by optimising the system design. A multi-stage
separation system could be one of the solutions and much work need to be conducted in order to
minimise the detrimental effects of the applied electric field on live tissues.
6. Conclusion
A platform for separating particles according to their dielectric response to alternating electric
fields at specific frequencies has been presented. The performance of the device was simulated using
CFD method. The dielectrophoretic spectrums for both polystyrene microparticles and MWCNTs
were studied and compared to obtain the optimal frequency for particle separation. Experiments were
carried out and MWCNTs were successfully separated from polystyrene microparticles for frequencies
over 100 kHz. Particle behaviours at different frequencies were characterized and analysed. The
developed device has potential novel applications such as: purifying cell suspensions with nanoscale
impurities, tissues and organelles separation, as well as microelectronic and nanoelectronic device
fabrications.
Reference:
[1] P. R. C. Gascoyne, and J. Vykoukal, “Particle separation by dielectrophoresis,”
Electrophoresis, vol. 23, no. 13, pp. 1973-1983, Jul, 2002.
[2] H. A. Pohl, Dielectrophoresis: The Behaviour of Neutral Matter in Nonuniform Elctric Field,
Cambridge: Cambridge University Press, 1978.
[3] A. T. J. Kadaksham, P. Singh, and N. Aubry, “Dielectrophoresis of nanoparticles,”
Electrophoresis, vol. 25, no. 21-22, pp. 3625-3632, Nov, 2004.
[4] N. Demierre, T. Braschler, R. Muller et al., “Focusing and continuous separation of cells in a
microfluidic device using lateral dielectrophoresis,” Sensors and Actuators B-Chemical, vol.
132, no. 2, pp. 388-396, 2008.
[5] E. M. Nascimento, N. Nogueira, T. Silva et al., “Dielectrophoretic sorting on a
microfabricated flow cytometer: Label free separation of Babesia bovis infected erythrocytes,”
Bioelectrochemistry, vol. 73, no. 2, pp. 123-128, 2008.
[6] T. Yasukawa, M. Suzuki, T. Sekiya et al., “Flow sandwich-type immunoassay in microfluidic
devices based on negative dielectrophoresis,” Biosensors & Bioelectronics, vol. 22, no. 11, pp.
2730-2736, May 15, 2007.
[7] S. Grilli, and P. Ferraro, “Dielectrophoretic trapping of suspended particles by selective
pyroelectric effect in lithium niobate crystals,” Applied Physics Letters, vol. 92, no. 23, 2008.
[8] X. Xiong, A. Busnaina, S. Selvarasah et al., “Directed assembly of gold nanoparticle
nanowires and networks for nanodevices,” Applied Physics Letters, vol. 91, no. 6, 2007.
[9] A. Kuzyk, B. Yurke, J. J. Toppari et al., “Dielectrophoretic trapping of DNA origami,” Small,
vol. 4, no. 4, pp. 447-450, 2008.
[10] B. H. Lapizco-Encinas, S. Ozuna-Chacon, and M. Rito-Palomares, “Protein manipulation with
insulator-based dielectrophoresis and direct current electric fields,” Journal of
Chromatography A, vol. 1206, no. 1, pp. 45-51, Oct 3, 2008.
[11] G. B. Salieb-Beugelaar, J. Teapal, J. van Nieuwkasteele et al., “Field-dependent DNA
mobility in 20 nm high nanoslits,” Nano Letters, vol. 8, no. 7, pp. 1785-1790, Jul, 2008.
[12] M. Yang, and X. Zhang, “Electrical assisted patterning of cardiac myocytes with controlled
macroscopic anisotropy using a microfluidic dielectrophoresis chip,” Sensors and Actuators a-
Physical, vol. 135, no. 1, pp. 73-79, 2007.
[13] M. Suzuki, T. Yasukawa, H. Shiku et al., “Negative dielectrophoretic patterning with colloidal
particles and encapsulation into a hydrogel,” Langmuir, vol. 23, no. 7, pp. 4088-4094, Mar 27,
2007.
[14] A. Rosenthal, and J. Voldman, “Dielectrophoretic traps for single-particle patterning,”
Biophysical Journal, vol. 88, no. 3, pp. 2193-2205, 2005.
[15] E. T. Thostenson, Z. F. Ren, and T. W. Chou, “Advances in the science and technology of
carbon nanotubes and their composites: a review,” Composites Science and Technology, vol.
61, no. 13, pp. 1899-1912, 2001.
109

[16] T. Ueda, M. M. H. Bhulyan, H. Norimatsu et al., “Development of carbon nanotube-based gas
sensors for NOx gas detection working at low temperature,” Physica E-Low-Dimensional
Systems & Nanostructures, vol. 40, no. 7, pp. 2272-2277, 2008.
[17] J. H. Lee, J. Kim, H. W. Seo et al., “Bias modulated highly sensitive NO2 gas detection using
carbon nanotubes,” Sensors and Actuators B-Chemical, vol. 129, no. 2, pp. 628-631, 2008.
[18] A. A. Pesetski, J. E. Baumgardner, S. V. Krishnaswamy et al., “A 500 MHz carbon nanotube
transistor oscillator,” Applied Physics Letters, vol. 93, no. 12, 2008.
[19] P. Stokes, and S. I. Khondaker, “Local-gated single-walled carbon nanotube field effect
transistors assembled by AC dielectrophoresis,” Nanotechnology, vol. 19, no. 17, 2008.
[20] Y. J. Kang, D. Q. Li, S. A. Kalams et al., “DC-Dielectrophoretic separation of biological cells
by size,” Biomedical Microdevices, vol. 10, no. 2, pp. 243-249, Apr, 2008.
[21] H. B. Li, Y. N. Zheng, D. Akin et al., “Characterization and modeling of a microfluidic
dielectrophoresis filter for biological species,” Journal of Microelectromechanical Systems,
vol. 14, no. 1, pp. 103-112, Feb, 2005.
[22] B. H. Lapizco-Encinas, B. A. Simmons, E. B. Cummings et al., “Dielectrophoretic
concentration and separation of live and dead bacteria in an array of insulators,” Analytical
Chemistry, vol. 76, no. 6, pp. 1571-1579, 2004.
[23] M. D. Vahey, and J. Voldman, “An equilibrium method for continuous-flow cell sorting using
dielectrophoresis,” Analytical Chemistry, vol. 80, no. 9, pp. 3135-3143, May 1, 2008.
[24] Z. Y. Wang, O. Hansen, P. K. Petersen et al., “Dielectrophoresis microsystem with integrated
flow cytometers for on-line monitoring of sorting efficiency,” Electrophoresis, vol. 27, no. 24,
pp. 5081-5092, Dec, 2006.
[25] J. T. Huang, G. C. Wang, K. M. Tseng et al., “A chip for catching, separating, and
transporting bio-particles with dielectrophoresis,” Journal of Industrial Microbiology &
Biotechnology, vol. 35, no. 11, pp. 1551-1557, Nov, 2008.
[26] R. Krishnan, B. D. Sullivan, R. L. Mifflin et al., “Alternating current electrokinetic separation
and detection of DNA nanoparticles in high-conductance solutions,” Electrophoresis, vol. 29,
no. 9, pp. 1765-1774, May, 2008.
[27] G. O. F. Parikesit, A. P. Markesteijn, O. M. Piciu et al., “Size-dependent trajectories of DNA
macromolecules due to insulative dielectrophoresis in submicrometer-deep fluidic channels,”
Biomicrofluidics, vol. 2, no. 2, Apr-Jun, 2008.
[28] L. L. Jia, S. G. Moorjani, T. N. Jackson et al., “Microscale transport and sorting by kinesin
molecular motors,” Biomedical Microdevices, vol. 6, no. 1, pp. 67-74, Mar, 2004.
[29] I. Ermolina, J. Milner, and H. Morgan, “Dielectrophoretic investigation of plant virus
particles: Cow Pea Mosaic Virus and Tobacco Mosaic Virus,” Electrophoresis, vol. 27, no.
20, pp. 3939-3948, Oct, 2006.
[30] H. Morgan, M. P. Hughes, and N. G. Green, “Separation of submicron bioparticles by
dielectrophoresis,” Biophysical Journal, vol. 77, no. 1, pp. 516-525, Jul, 1999.
[31] N. G. Green, H. Morgan, and J. J. Milner, “Manipulation and trapping of sub-micron
bioparticles using dielectrophoresis,” Journal of Biochemical and Biophysical Methods, vol.
35, no. 2, pp. 89-102, Sep 25, 1997.
[32] M. J. Mendes, H. K. Schmidt, and M. Pasquali, “Brownian dynamics simulations of singlewall
carbon nanotube separation by type using dielectrophoresis,” Journal of Physical
Chemistry B, vol. 112, no. 25, pp. 7467-7477, 2008.
[33] P. Blasi, S. Glovagnoli, A. Schoubben et al., “Solid lipid nanoparticles for targeted brain drug
delivery,” Advanced Drug Delivery Reviews, vol. 59, pp. 454-477, 2007.
[34] A. J. Almeida, and E. Souto, Solid lipid nanoparticles as a drug delivery system for peptides
and proteins, Advanced Drug Delivery Reviews,Volume 59, Issue 6, 10 July 2007, Pages 478-
490
[35] M. Arruebo, R. Fernández-Pacheco, M. Ricardo Ibarra and J. Santamaría, Magnetic
nanoparticles for drug delivery, Nanotoday, Volume 2, Issue 3, June 2007, Pages 22-32
[36] C. Klumpp, K. Kostarelos, M. Prato et al., “Functionalized carbon nanotubes as emerging
nanovectors for the delivery of therapeutics,” Biochimica Et Biophysica Acta-Biomembranes,
vol. 1758, no. 3, pp. 404-412, 2006.
110

[37] A. Motayed, M. Q. He, A. V. Davydov et al., “Realization of reliable GaN nanowire
transistors utilizing dielectrophoretic alignment technique,” Journal of Applied Physics, vol.
100, no. 11, Dec 1, 2006.
[38] M. Dimaki, and P. Boggild, “Dielectrophoresis of carbon nanotubes using microelectrodes: a
numerical study,” Nanotechnology, vol. 15, no. 8, pp. 1095-1102, Aug, 2004.
[39] D. J. Yang, S. G. Wang, Q. Zhang et al., “Thermal and electrical transport in multi-walled
carbon nanotubes,” Physics Letters A, vol. 329, no. 3, pp. 207-213, Aug 23, 2004.
[40] M. P. Hughes, H. Morgan, and M. F. Flynn, “The dielectrophoretic behavior of submicron
latex spheres: Influence of surface conductance,” Journal of Colloid and Interface Science,
vol. 220, no. 2, pp. 454-457, 1999.
[41] L. Cui, D. Holmes, and H. Morgan, “The dielectrophoretic levitation and separation of latex
beads in microchips,” Electrophoresis, vol. 22, no. 18, pp. 3893-3901, Oct, 2001.
[42] R. Krupke, Hennrich, F., Kappes, M. M., Lohneysen, H., “Surface Conductance Induced
Dielectrophoresis of Semiconducting Single-Walled Carbon Nanotubes,” NANO LETTERS,
vol. 4 no. 8, pp. 1395–1399, 2004.
[43] J. Kim, Y. H. Shin, J. H. Yun et al., “A nickel silicide nanowire microscopy tip obtains
nanoscale information,” Nanotechnology, vol. 19, no. 48, Dec 3, 2008.
[44] J. W. Lee, K. J. Moon, M. H. Ham et al., “Dielectrophoretic assembly of GaN nanowires for
UV sensor applications,” Solid State Communications, vol. 148, no. 5-6, pp. 194-198, Nov,
2008.
[45] J. Voldman, “Electrical forces for microscale cell manipulation,” Annual Review of
Biomedical Engineering, vol. 8, pp. 425-454, 2006.
111

Use of a Radially Polarized Beam for Ultra-low Energy Threshold
for Cancer Photothermal Therapy with Gold Nanorods [1]
Hong Kang 1 * , Jingliang Li 1 , Baohua Jia 1 , Dru Morrish 1 and Min Gu 1
* Presenter
1. Centre for Micro-Photonics , Swinburne University of Technology, VIC, Austraila,
3122
We demonstrate that the use of a radially polarized beam is more efficient for
both imaging and therapy for cancer cells labelled with gold nanorods, compared with
that of a linearly polarized beam.
OCIS codes
(180.0180) Microscopy, (170.0170) Medical optics and biotechnology
1. Introduction
Advances in the development of nanosized materials have created new opportunities and
applications in biomedical and biological fields. Amongst these nanomaterials, metallic nanoparticles
have attracted more interest owing to their flexible synthesis, enhanced collective dipole oscillation
(surface plasmon resonance) and their strong luminescence emission when they are exposed to laser
irradiation [1-3]. The luminescence property makes them a novel contrast agent for diagnosing and
detecting specific cells. Furthermore, it enables plasmonic photothermal therapy (PPTT) in which the
photon energy is converted into heat to induce hyperthermia in adjoining cells. Due to its localised and
minimum invasive feature, PPTT has become a promising method for cancer treatment.
Distinguished from spherically shaped particles, nanorods exhibit two absorption bands due to the
surface plasmon resonance as shown in Fig. 1. The absorption peak located in the shorter wavelength
region can be attributed to the oscillation along the transverse axis, which experiences relatively
insignificant change with the aspect ratio (length divided by width) of nanorods. The other much
stronger absorption peak resulting from the longitudinal oscillation is highly sensitive to the size
variation and shifts to a longer wavelength as the aspect ratio increases. The absorption spectra for the
longitudinal mode can be tuned to the near infrared (NIR) region, which is highly desirable for in vivo
applications as NIR excitation experiences less absorption, scattering and a larger penetration depth
[4-6]. Fig. 1 illustrates the absorption bands of the gold nanorods used in this paper. The transverse
and longitudinal absorption peaks at wavelengths 517 nm and 780 nm, respectively, can be clearly
identified from Fig. 1.
Due to the surface plasmon property, gold nanorods are most efficiently excited by light polarized
parallel to their longitudinal axis. However, when cells are labelled with gold nanorods, the nanorods
are randomly dispersed and orientated. This makes the excitation of the nanorods by a linearly
polarized beam less efficiency because only a small portion of the randomly oriented nanorods
aligning with the incident polarization direction can be excited. Higher incident laser power is required
to compensate for the less efficient nanorod excitation. Too high an incident power can kill healthy
cells outright. To overcome this challenge, a highly localized focal spot polarized in all three
directions is highly desired. By utilizing such a radially polarized beam, nanorods in every orientation
can be more efficiently excited, which subsequently reduces the required laser power and leads to an
ultra-low energy threshold for cancer photothermal therapy. In this paper, a radially polarized beam is
employed for this purpose.
[1]. Published at the Conference of Novel Techniques in Microscopy, Vancouver Canada,
April 30, 2009, @2009 Optical Society of America

Fig. 1. The transverse and longitudinal surface plasmon absorption spectra of gold nanorods. The solid
arrows indicate the direction of the surface plasmon resonance modes within the nanorod. The dashed
arrows indicate the transverse and longitudinal absorption peaks at wavelength 517 nm and 780 nm,
respectively.
As shown in Fig. 2, a radially polarized beam possesses even electric field aligning in a radial
direction in free space. Once focused by an objective with high numerical aperture (NA), a strong
longitudinal component is produced due to the depolarization effect [7,8]. As a result, a light field
polarized in all directions is generated composing these three orthogonal electric fields. Thus by using
such a tightly focused radially polarized beam in PPTT nanorods of arbitrary orientation can be
efficiently excited. Therefore, the required energy fluence for imaging and treatment can be greatly
reduced to a level below the medical standard for clinical applications.
Fig. 2. Experimental intensity distribution of a radially polarized beam. The arrows indicate the
polarization direction of the light.
2. Experiments
The radially polarized beam was generated by a radial polarization converter (Arcoptix S.A.). Fig.
2 shows the experimental intensity distribution of a typical hollow radial beam in free space. To
examine the polarization profile, an analyser is placed rotated to three orientations within the beam.
Fig. 3 shows the obtained intensity patterns at the three different directions. The two high-intensity
113

lobes align with the analyser direction with approximately equal intensities, demonstrating the good
polarization uniformity and the good beam quality of the converted radially polarized beam.
Fig. 3. Intensity distributions at the output of the analyser at three polarization angles (a) 0° (b) 45° (c)
90° (polarization angle is defined as the angle between the incident polarization direction and the
polarization direction of the analyser). The arrows indicate the polarization direction of the analyser.
The in vitro two-photon imaging and treatment were both carried out on a confocal scanning
microscopic system (Olympus: IX70). HeLa (cervical cancer) cells mediated with poly(4-
styrenesulfonic acid) (PSS)-coated gold nanorods were used for all experiments [4]. The samples were
imaged and treated with femtosecond laser pulses at the wavelength of 780 nm generated with an
ultrafast laser with a repetition rate of 80 MHz and a pulse width of approximately 80 fs (Spectra
Physics: Mai Tai). A water immersion objective of NA 1.2 was used both for imaging and treatment.
Figs. 4(a) and 4(b) show the fluorescent images of gold nanorods directly on a cover slip excited with
the linearly polarized beam (Fig. 4(a)) and the radially polarized beam (Fig. 4(b)), respectively, under
the same power level.
The luminescence intensity under linear polarization excitation is much weaker than that under
radial polarization illumination. In Fig. 4(c), the combined luminescence of the nanorods and
transmission images of HeLa cells incubated with PSS coated gold nanorods is presented. It has been
found that by using the radially polarized beam, the energy fluence needed for PPTT was just one fifth
of the clinical safety level of 100 mJ/cm 2 [9].
Fig. 4. Luminescence of gold nanorods excited with (a) a linearly polarized beam, (b) a radially
polarized beam and (c) combined luminescence of nanorods and transmission images of HeLa cells
incubated with PSS coated gold nanorods. The scale bars are 10 μm.
3. Summary
In this paper, we have demonstrated that a radially polarized beam is an effective source for PPTT
of cancer cells labelled with gold nanorods. By using a tightly focused radially polarized beam,
nanorods randomly oriented on HeLa cells can be efficiently excited leading to a significantly reduced
energy threshold for both imaging and treatment.
114

References:
[1] Xiong Liu, Mark C. Lloyd, Inna V. Fedorenko, Priya Bapat, Tatyana Zhukov and Qun
Huo, ”Enhanced imaging and accelerated photothermalysis of A549 human lung cancer cells by
gold nanospheres,” Nanomedicine 3, 617-626 (2008).
[2] Xiaohua Huang, Prashant K. Jain, Ivan H. El-Sayed and Mostafa A. El-Sayed,, “Plasmonic
photothermal therapy (PPTT) using gold nanoparticles,” Lasers Med Sci 23, 217-228 (2008).
[3] Jingyi Chen, Danling Wang, Jiefeng Xi, Leslie Au, Andy Siekkinen, Addie Warsen, Zhi-Yuan Li,
Hui Zhang, Younan Xia and Xingde Li, “Immuno gold nanocages with tailored optical properties
for targeted photothermal destruction of cancer cells,” Nano Lett 7, 1318-1322 (2007).
[4] Jingliang Li, Daniel Day and Min Gu, “Ultra-low energy threshold for cancer photothermal
therapy using transferrin-conjugated gold nanorods,” Adv. Mater, 20, 3866-3871 (2008).
[5] Ling Tong, Yan Zhao, Terry B. Huff, Matthew N. Hansen, Alexander Wei and Ji-Xin Cheng,
“Gold nanorods mediate tumor cell death by compromising membrane integrity,” Adv Mater 19,
3136-3141 (2007).
[6] Takuro Nidome, Masato Yamagata, Yuri Okamoto, Yasuyuki Akiyama, Hironobu Takahashi,
Takahito Kawano, Yoshiki Katayama and Yasuro Niidome, “PEG-modified gold nanorods with a
stealth character for in vivo applications,” Journal of Controlled Release 114, 343-347 (2006).
[7] Min Gu, Advanced Optical Imaging Theory (Springer, 2000), Chap. 6.
[8] Baohua Jia, Xiaosong Gan, and Min Gu, “Direct measurement of a radially polarized focused
evanescent field facilitated by a single LCD,” Optics Express 13, 6821–6827 (2005).
[9] American National Standard for Safe Use of the Lasers ANSI Z136.1 (American Laser Institute,
2000).
115

Electrospinning Thermoplastic Polyurethane-contained Collagen
Nanofibers for Tissue Engineering Applications [1]
Rui Chen 1 * , Qinfei Ke 1 , Pujang Shi 3 , Yos Morsi 3 and Xiumei Mo 2
* Presenter
1. College of textiles, Donghua University, Shanghai, 201620, China
2. College of chemistry and chemical engineering and biological engineering, Donghua
University, Shanghai, 201620, China
3. Biomechanics and Tissue Engineering Group, Faculty of Engineering and Industrial
Science, Swinburne University of Technology, Hawthorn, VIC3122, Australia
Electrospinning is a new method used in Tissue Engineering field. It can spin fibers in
nanoscale by electrostatic force. A serious of thermoplastic polyurethane (TPU)/collagen
blend nanofibrous membranes were prepared with different weight ratios and concentrations
via electrospinning. The two biopolymer were all used 1,1,1,3,3,3,-hexafluoro-2-
propanol(HFP) as solvent. The electrospun thermoplastic polyurethane-contained collagen
nanofibers was characterized using scanning electron microscopy (SEM), XPS spectroscopy,
atomic-force microscopy, apparent density and porosity measurement, contact angle
measurement, mechanical tensile testing and viability of Pig iliac endothelial cells (PIECs) on
blended nanofiber mats. Our date indicate that fiber diameter was influenced by both polymer
concentration and blend weight ratio of collagen to TPU. The average diameter of nanofibers
gradually decreases with increasing collagen content in the blend. XPS analysis indicates that
collagen is found to be present at the surface of blended nanofiber. The results of porosity and
contact angle measurement suggest that with the collagen content in the blend system, the
porosity and hydrophicity of the nanofiber mats is greatly improved.We have also
characterized the molecular interactions in TPU/collagen complex by fourier transform
(FTIR). The result could demonstrate that there were intermolecular bonds between the
molecules of TPU and collagen. The ultimate tensile stress and strain were carried out and the
dates could also prove the analysis of FTIR. The TPU/collagen blend nanofibrous mats were
further investigated as promising scaffold for PIECs culture. The cell proliferation and SEM
morphology observations showed that the cells could not only favorably grow well on the
surface of blend nanofiberous mats, but also able to migrate inside the scaffold with 24 hours
of culture. These results suggest tha the blend nanofibers of TPU/collagen are designed to
mimic the native extracellular matrix for tissue engineering and develop functional
biomaterials.
Keywords
Engineering
Electrospinning, Thermoplastic polyurethane /collagen blend, Nanofiber, Tissue
1 Introduction
The technology of tissue engineering (TE) aims to generate new or substitute or malfunctioning
and could well become an alternative method to whole organ transplantation [1, 2]. So far as we
know, the TE methods have been applied to different types of tissue and organ such as skin[3],
bone[4], liver[5], intestine[6], heart valve[7], muscle[8] and tongue[9]. One of the most important
factor in TE is the three-dimensionally porous scaffold, which allows biological activities such as
cell adhering, migrating, growing and differentiating to attain a proper integration between cells and
scaffold for synthesizing a
[1]. Published at Journal of Biomaterials Science (2009) 20:1513-1536

new tissue[10]. Most of these human organs deposited on fibrous structures with the fibril/fiber size
realigning from nanometer realigning from nanometer to millimeter scale. So nanofibers have now
been extensively used to mimic these natural tissue matrixes. At present, electrospinning is the most
prevalent process that can creates nanofibers through an electrically charged jet of polymer solution
or polymer melt. Different processing parameters such as kind of polymer, viscosity, surface tension,
jet charge density, temperature and humidity control the electrospinning process, especially the
diameter and morphology of the resulting fibers[11]. Recently, researchers have found that the
nanofibrous structure formed by electrospinning method will improve the function in vitro tissue
regeneration and decrease the formation of scar tissue[12]. So the scaffolds prepared from
electrospinning method can be considered to mimic the native extracellular matrix (ECM).
Representative polymers including synthetic ones such as poly(lactic-acid)(PLA)[13,
14],poly(glycolic-acid)(PGA)[15], poly(lactic-co-glycolic acid)(PLGA)[16], poly(ε-caprolactone)
(PCL)[17, 18]and natural ones such as collagen[19], chitosan[20], gelatin[21] and silk[22]have been
electrospun into nanofibers.
Native ECM is the complex of polyprotein and polysaccharide with nanofibrous structure.
Among the natual biopolymers, collagen as polyprotein has been widely used in TE for its excellent
biocompatibility and nonimmunogenicity, but the mechanical property can’t achieve the request of
native ECM. Thermoplastic polyurethanes (TPUs) are a widely used class of polymer with excellent
mechanical properties and good biocompatibility, and have been evaluated for a variety of
biomedical applications such as coating materials for brest implants, catheters, and prosthetic heart
valve leaflets[23]. Conventional TPU are among biomaterials not intended to degrade but are
susceptible to hydrolytic, oxidative and enzymatic degradation in vivo. While the susceptibility of
TPU to such degradation is a problem for long lasting biomedical implants, it can be deliberately
exploited to design biodegradable polyurethane[24]. The TPU used in this research one kind of
medical-grade, aliphatic, polyether-based TPUs. It can biodegrade and the biostability is better than
poly(ester urethane).
As candidate materials, pure TPU and collagen have already been electrospun into nanofibers
as biomaterials respectively[19, 25]. And they all have biological benefits used as TE scaffolds. The
promising study of electrospun TPU and collagen complex has not been well understood. Especially,
the relationship between processing parameters and microstructures in the electrospun nanofbers has
not been reported.
In this study, collagen and and TPU were co-electrospun to produce collagen/TPU composite
nanofibrous scaffolds. We created these electrospun hybrid scaffolds that combined synthetic
material with natural proteins to overcome limitations seen with scaffolds constructed with either
one alone. We characterized the properties of these novel scaffolds and optimized for porosity,
strength, cell seeding. Collagen/TPU hybrid fibrous scaffolds were determined to provide optimal
fiber diameter, pore size and mechanical strength, leading to enhance the seeding of the electrospun
scaffolds with cells. The object of this project is to study the electrospinning of TPU/collagen
complex and to develop the biomimetic extracellular matrix for TE application.
2 Materials and methods
2.1 Materials
The polymer of TPU (Tecoflex EG-80A) was purchased from Noveon, Inc.(USA) and collagen
(mol.wt, 0.8-1×105Da) was purchased from Sichuan Ming-rang Bio-Tech Co. Ltd (China). The two
materials all used 1,1,1,3,3,3,-hexafluoro-2-propanol (HFP) as solvent, which was brought from
Daikin Industries Ltd( Japan). Pig iliac endothelial cells (PIECs) were obtained from institute of
biochemistry and cell biology (Chinese Academy of Sciences, China). Except specially explained, all
culture media and reagents were purchased from Gibco Life Technologies CO, USA.
2.2 Scaffold fabrication
The polymers of TPU and Collagen solutions were dissolved in HFP separately. When they were
prepared already, the two solutions were blended at different weight ratios and volume ratios with
sufficient stirring at room temperature before electrospinning.
The pure and blend polymer solutions were placed into a 2.5 ml plastic syringe with a bluntended
needle with an inner diameter of 0.21mm. The needle was located at a distance of 130~150mm
from the grounded collector. A syringe pump (789100C, cole-pamer, America) was employed to feed
117

solutions to the needle tip at a feed rate of 1.6~2.0 ml/h. A high electrospinning voltage was applied
between the needle and ground collector using a high voltage power supply (BGG6-358,
BMEICO.LTD, China). The applied voltage was between 16 and 20KV. The electric field generated
by the surface charge caused the solution drop at the tip of the needle to distort into a Taylor cone.
2.4 Characterization of collagen-blended TPU nanofibers
The morphologies and diameters of the nanofibers electrospunned by pure and different weight
ratio concentration of TPU to collagen were determined with SEM (JEOL, JSM-5600, Japan) at a
accelerated voltage of 15KV. The diameter range of the fabricated nanofibers was measured based on
the SEM images using an image visualization software ImageJ 1.34s (National Institutes of Health,
USA).
Surface chemistry analysis of the electrospun scaffolds were also analyzed using X-ray
photoelectron spectroscopy (XPS) (Escalab 250; Thermo Scientific Electron, East Grinstead, UK)
equipped with Mg Kα at 1,486.6 eV and 150W power at the anode. A survy scan spectrum was taken
and the surface elemental compositions relative to carbon were calculated from the peak height with a
correction for atomic sensitivity.
Surface propeties of the nanofibers were examined using a a Nanoscope Ⅳ atomic-force
microscope (Digital Instruments), in the tapping mode and expressed as height and phase images.
2.5 Contact angle measurement
The electrospun mats made from pure TPU and TPU/collagen were characterized by the water
contact angle measurement to find the effect of addition of collagen on the hydrophicity of the
materials. The images of the droplet on the membrane were visualized through the image
analyzer(OCA40, Datephysics Co, German) and the angles between the water droplet and the surface
were measured. The measurement used distilled water as the reference liquid and was automatically
dropped onto the electrospun mats. To confirm the uniform distribution of blend nanofiber mats, the
contact angle was measured 3 times from different positions on each mat and an averge value was
calculated by statistical method.
2.6 The porosity of electrospun mats
The electrospun mats of TPU/collagen with different blend ratio at a concentration of 8wt%. The
thickness of the nanofiber mesh was measured with a micrometer (Shanghai, China). The apprent
density and porosity were calculated accroding to the following equations [26]:
Apparent density of nanofiber mats(g/cm 3 )=
Mass of nanofibrous mast (g)
2
Thickness of nanofibrous mats (cm) × Nanofiber mats area (cm )
Porosity of nanofiber mats (%)=
3
Naonfiber mats apparent density (g/cm )
) × 100%
3
bulk density fo raw TPU/collagen (g/cm )
2.6 FTIR spectra
Electrospun TPU nanofibers, collagen nanofiber and TPU/collagen with different weight ratios
were prepared for the FTIR test on AVATAR 380 FTIR instrument (Thermo Electron, Waltham, MA).
All spectra were recorded by an absorption mode in the wave length range of 4000-500 cm -1 .
2.7 Mechanical measurement
Mechanical measurements were carried out by applying tensile test loads to samples which were
prepared from electrospun ultra fine non-woven fiber mats. The specimens were electrospun with
different blend ratios of collagen to TPU (0:1, 1:3, 1:1, 3:1). In this study, the test was performed in
ambient temperature at 20 ℃ and relative humidity of 65%. Four specimens of each sample were
prepared according to the method described by Huang et al [21]. First, a white paper was cut into
templates with a planner dimension of width × gauge length=10mm×30mm and then double side tapes
were glued onto the top and bottom areas of one side. Secondly, the aluminum foil was carefully
scraped off and single side tapes were applied onto the griping areas as end-tabs. Finally, the resulting
specims were conducted on a commercial materials testing machine(H5K, Hounsfield, England) with
a load cell of 10N, and the elongation speed is 10mm/min.
118

2.8 Viability and morphology study of PIEC on nanofiber mats
Pig iliac endothelial cells(PIECs) were cultured in DMEN medium with 10% fetal serum, and
100 units/ml and 100units/ml streptomycin in humidified incubator under standed culture conditions
(5% CO 2 , content at 37℃), and the medium was replaced every three days. Mats of blend nanofibers
with different ratios of TPU and collgen were electrospun on circular 14mm glass coverslips. After
the coverslips with nanofibers were prepared already, they were dried in vacuum for over one week to
release the residua solvents. Then the mats were placed in desiccator to crosslink using glutaraldehyde
(25%water solution) steam for two days. Whereafter the blend nanofibers mats were dried in vacuum
for over two weeks to release the residual glutaradehyde. And the next step is to fixed the coverslips
into 24-well plates with stainless ring. Before seeding cells, fiber scaffolds were sterilized with 75%
alcohol solution, which were placed with phosphate-buffered saline solution (PBS) after 2 hours.
Cells viability on the nanofibers were determined by MTT method. Briefly, the cell and
nanofiber complex was incubated with 5mg/ml 3-(4,5)-dimethylthiahiazo (-z-y1)-3,5-diphenytetrazoliumromide
(MTT) for 4h. Thereafter the culture media were extracted and added 150ul
Dimethyl Sulfoxide (DMSO) to have low speed surge for about 10 minutes. When the crystal was
sufficient resolved, aliquots were pipetted into the wells of a 96-well plate and placed into a Enzymelabeled
Instrument (MK3, Thermo, USA), and the absorbance at 490nm for each well was measured.
For the cell viability test, endothelium cells were seeded onto nanofiber mats (n=3) with
different weight ratio of collagen to TPU and control glass coverslips at a density of 5×10 4 cells/cm 2 .
On days 1, 3, 5 and 7 after cell seeding, untached cells were quantified by MTS kit ((C0009, Beyotime
Institute if Biotechnology, China) and Enzyme-labeled Instrument (MK3, Thermo, USA).
After 24 hours of culturing, the electrospun fibrous scaffolds with cells (density is 1.5×10 5
cells/cm 2 ) were examined by SEM. To prepare the cell-cultured samples for SEM observation, the
scaffolds were rinsed twice with PBS in the first place, followed by fixation with 4% glutaraldehyde
water solution for 2h and then the samples were again rinsed twice with PBS. Thereafter dehydrated in
graded concentrations of ethanol (30, 50, 70, 80, 90, 95 and 100). Finally, they were dried in vacuum
overnight. The dry cellular constructs were coated with gold sputter and observed under the SEM at a
voltage of 15KV.
2.9 Statistics analysis
Statistics analysis was performed using origin 7.5 (Origin lab Inc, USA). Values (at least
triplicate) were averaged and expressed as means ± standard deviation (SD). Statistical differences
were determined by the anaylsis of One-Way ANOVA and differences were considered statistically
significant at p

Blend nanofiber diameters distribution of different blend ratios were further determined and the
results were shown in Fig.2. Nanofiber diameters of every weight ratios were calculated from the
diameter of 100 nanofibers each sample which were directly measured from SEM photographs. From
numerical statement of average diameters and standard deviation we can see that the fiber diameter
gradually deceased with increasing collagen content in the blend. This phenomenon could be
explained as the conductivity increase of the blend solution. It has been known that in electrospinning,
the force that causes the stretching of the solution is due to the repulsive forces between the charges on
the electrospinning jet[29]. Collagen is typical amphiprotic macromolecule electrolyte. When collagen
was affiliated, more ions were formed in the blend solution. So a higher discharge density could be
carried by the electrospinning jet. The conductivity of the solution could be increased by the addition
of ions. On the other hand, the increased charge carried by the solution will increase the stretching of
the solution. And the increased in the stretching of the solution also could tend to yield fibers of
smaller diameter.
In order to investigate the factor of solution concentration, we selected one weight ratio (80/20,
TPU/collagen) to prepare this blend solution with six different concentrations.
Fig 1 and 2 SEM images of nanofiber mats at concentration of 6wt% with different blend ratios of
TPU to Collagen(T/C) and their fiber diameter distribution
Then the mixed solutions were subjected to the electrospinning process under the same
processing condition. Fig.3 shows SEM images of TPU/collagen fibers with the concentration which
varied from 1.5% to 9% at equal interval. Their fiber diameter distributions were calculated in the
same method and the results were shown in Fig.4. From the SEM images we could observe that the
smooth fibers could be obtained when the concentrations of the solutions were located between 4.5%
and 7.5%. At the same time, we could find that either the concentrations were too low or too high,
more beads structure could appear. It is established that chain entanglements in the solution plays an
important role in the fiber diameter and morphology. The increase of the solution concentration will
result in great polymer chain entanglements. Thus, if the concentration is to low, bead-only structure
will be produced due to a lack of chain entanglements in the solution[30]. On the contrary, if the
concentration is too high, that will make it very difficult to pump the solution through the syringe
needle. Morevoer, when the concentration is too high, the solutions may dries at the tip of the needle
before the electrospinning can be initiated.
120

Fig 3 and 4 SEM images of electrospun TPU/collagen (80/20) blend nanofibers as a functions of
different concentrations.(A: 1.5wt%; B: 3wt%; C: 4.5%; D: 6%; E: 7.5%; F: 9% ) and their fiber
diameter distribution
The surface chemistry changes of blended nanofibers was verified by XPS spectroscopy. Fig.5
shows the XPS survey scans of nanofibrous scaffold surfaces. Table 1 shows the atomic ratios of
carbon, nitrogen and oxygen on pure and blended nanofibers with weight ratio of 50:50. As expected,
the blend-electrospun scaffold collagen/TPU shows three peaks corresponding to C1s (binding energy,
286 eV), N1s (binding energy, 400 eV) and O1s (binding energy, 532 eV). The atomic ratios of carbon,
nitrogen and oxygen on nanofibrous scaffolds calculated from XPS survey scan spectra are shown in
Table 1. The oxygen content (20.29%) of TPU nanofiber surface was not changed by the
incorporation of collagen (collagen/TPU, 19.46). On the other hand, nitrogen (7.4%) found on the
blend nanofiber surface was much higher than TPU nanofiber (1.77%). This result illustrate that
collagen was found to be present on the surface of blended nanofibers. However, it is hard to say that
collagen was located on the surface of blended nanofibers, because both TPU and collagen have
Nitrogen content in their chemical structures. The existence of collagen molecules of collagen on the
surface and inside the nanofibers provides sustained cell recognition sigals with polymer degradation,
which is important for cell function development .
Fig 5 X-ray photoelectron Spectroscopy survy scan spectra of (a) collagen nanofiber, (b) TPU
nanofiber, (c)collagen/TPU nanofiber(50:50,w/w)
To study the surface morphologies of TPU, collagen and TPU/collagen nanofiber surface, AFM
image was examined using a height mode (Fig. 6). From the analysis of AFM, we found that the
121

surface of TPU nanofibers is much smooth than collagen nanofibers. The blend nanofbiers wers also
even, but there were a lot of lines on the surface at the surface of blend nanofibers.
Fig 6 AFM images represented by height mode:(a) TPU fiber (b) collagen fiber (c) TPU/collagen
fiber(50:50,w/w)
3.2 Water contact measurements
The hydrophobic (or hydrophilic) nature of a substrate has a direct impact on the avenue of its
usage. Fig.7 gives the contact-angle dates variation and the water convex shape on different
electrospun fibrous mats measured with a surface analysis instrument. As expected, collagen showed
much better wettability than the TPU because of its hydrophilic group. The electrospun pure TPU mats
showed an angle around 134.1° which could seemed to be as lyophobic. With the increasing collagen
component in the blend nanofiber, the various electrospun mats exhibited the contact angle around 60-
80°. The result indicated that the hydrophilicity of the membranes based on TPU was significantly
improved by adding the collagen ingredient. Such a property will be extremely useful for creation of
tissue with cellularity throughout the scaffold by cell culture.
Fig 7 Water contact angle of electrospun TPU/collagen nanofibrous mats with different blend ratios.
(Inset this figure shows the variation shapes of contact angle on different mats. (A) T/C =1:0; (B) T/C
= 3:1; (C) T/C =1:1; (D) T/C =1:3; (E) T/C =0:1; (F) T/C =0:1.Date are means±SD (n=3)
The hydrophilicity of tissue engineering scaffolds is very important for their application. Many
works have reported that cells attached and spread more easily and effectively on surfaces with proper
hydrophilicity than on hydrophobic surfaces[31]. From the cell culture result, we could find that there
the improved surface hydrophilicity was expected to facilitate cell attachment and proliferation. With
collagen ingredient increasing in the blend nanofibers, the hydrophilicity was greatly improved. The
blend ratio of TPU/collagen from 100:0 to 75:25, the contact angle had a great change from 134 to 82°.
Although there were no sigificant difference among these blend nanofibrous mats in the first 24 hours,
they all showed better proliferation rate at 72, 120 and 144 hours. Furthermore, Cell attachment to the
substratum is almost always mediated by extracellular matrix (ECM) proteins absorbed on the culture
surface. ECM proteins are present in the serum used in most cell culture applications. Numerous
122

studies have indicated that surface properties such as surface hydrophilicity and charge could affect
cell attachment by influencing the ability of the substratum to absorb protein and/or by altering the
conformation of the absorbed protein[32, 33].
3.3 Porosity of nanofibrous mats
During the process of electrospinning, the thickness of the blended nanofibers was controlled by
the deposition time if the electrospinning parameters were all fixed. We can see that all the apparent
density of blended nanofibers was in the range of 0.30-0.60g/cm 3 and it slowly decreased with the
collagen increasing in the blend.
As the electrospun TPU blended collagen nanofibous mats are having highly pore size, using the
apprent density of all different mats and the bulk density of TPU/collagen blends, we could calculate
the porosity of the blended nanofibers. However, it was not easy to accurately measure the bulk
density of the TPU/collagen blend for it was a high porous material and contained a mixture of
different ingredient, so the bulk density of the blends were estimated to be 1.04g/cm 3 on the basis of
the bulk density of TPU. From Fig.8 we can see that the porosity of TPU/collagen nanofibers had a big
increase with the increasing weight ratio of collagen to TPU and we can also got the conclusion from
Fig.1 that the pore size of D, E, F were larger than A, B, C. The porosity increase of nanofiber mats by
adding collagen in the blend can also be explained as the increase of solution conductivity[34]. From
the fiber diameter analysis we could know that with the increase of the conductivity of the solution,
the finer of the fiber would be. On the other hand, with more electronic charges carried by the
nanofiber, there would be stronger repulsive forces among fibers during depositing to the collector.
Fig 8 The porosity of TPU/collagen blend nanofibrous mats with various blend ratios.
Pores play an important role in determining the physical and chemical properties of porous
substrates and have a deterministic effect on the performance of membranes, catalysts, adsorbents etc.
On the other side, the mats must have a large pore volume fraction as well as an interconnected pore
work to permit the transport, multiplication and metabolites. High porosity is propitious to cell
adhesion on the mat, promotes extracellular matrix regeneration. Thus, porosity is an essential factor
for materials to minic extracellular matrix in tissue engineering. From the analysis of the blended
nanofibrous mats, they all showed high porosity and adequate mechanical property for cell ingrowth.
Date are representatives of three independent experiments and all date points are plotted as
means±SD (n=3)
3.4 FTIR spectra
Fig 9 depicts the spectra of TPU/ collagen blend with different weight ratio. It can be seen that
although the blends prepared in different ratios, these spectras are quite simlar to each other and the
peaks appear at the same bands. The difference among the three spectra is the intensity of some peaks.
The characteristic peaks of thermoplastic polyurethane at 3290 cm-1, 2930 cm-1 and 2860 cm-1 can
be found in the range of 3500-2500 cm-1,while the peaks between 2000-1000 cm-1, such as those at
123

1640 cm-1, 1540 cm-1 and 1240 cm-1, can be assigned to the characteristic absorption of collagen. It
is noticed that the characteristics absorption peak of TPU at 1530 cm-1 and 1220 cm-1 were overlaped
with characeristic band of collagen, and the peak at 1710 cm-1 is belong to the stretching vibration of
carbonyl group. The spectras show both characteristics peaks of electrospun collagen and TPU. So this
is the evidence to illuminate that there may be no reaction between collagen and TPU.
Fig 9 FTIR spectra of TPU and collagen blend nanofibrous membranes (a: TPU/collagen=75:25; b:
TPU/collagen=50:50; c: TPU/collagen=25:75).
3.5 Mechanical test
The mechanical properties of blend nanofibers are important for their successful applications in
tissue engineering. Collagen and TPU blend nanofibers were electrospun into 0.5mm thick fiber mats
to measure their mechanical properties. Fig.11 shows the tensile stress-strain curve of collagen/TPU
nanofiber mats with various blend ratios. The electrospun TPU material gives a characteristic response
for elastomeric materials-sigmoidal in shape. It showed a very soft and flexible characteristic with low
Young’s modulus and the high elongation at break of 365%. With the increase of collagen content in
the blend, the initial modulus of the mats became large. This phenomenon indirect that electrospun
collagen has a plastic property which is different from TPU. From the stress-strain curves, we can see
that with collagen content increasing, the materials mechanical properties have changed from
elastomeric to plasic. Therefore, we can adjust the mechanical property to meet the requirement in
practice through changing the blend ratio of collagen to TPU.
To design an ideal scaffold, various factors should be considered such as pore size and
morphology, mechanical properties versus porosity, surface properties and appropriate
biodegradability. Of these factors, the importance of mechanical properties on cell growth is
particularly obvious in tissues such as bone, cartilage, blood vessels, tendors, heart valve and muscles.
Different natual tissues have different mechanical propertiies. For example, heart valve of human and
porcine have different mechanical tensile stress and ultimate strain. TPU has high tensile strength,
good tear and abrasion resistance. But its Young’s modulus is very low. Collagen has high Young’s
modulus and excellent biocompatiblity, but its tensile strenght and strain are poor. In order to mimic
the mechanical properties in the radial and circuit direction, we need to adjust the blend ratios of TPU
and collagen. From the analysis, we can find that with the collagen ingredient increasing, the Young’s
modulus, tensile strength and tensile strain changed. Blended nanofibers may have improved
mechanical strength compared with pure non-crosslinked collagen nanofibers, which could combine
the advantages of both synthetic and natual materials.
124

Fig 10 Tensile Stress curve of electrospun TPU and TPU/collagen blend nanofibrous mats (A)
electrospun TPU, (B) electrospun T/C=75:25, (C) electrospun T/C=50:50, (D) electrospun T/C=25:75.
3.6 Viablity of cells on collagen-blended TPU nanofibers
The viability of PIECs on days 1, 3, 5 and 7 after seeding on blend nanofibers with various
weight ratios of TPU to collagen is shown in Fig.11. The viability of cells cultured on blend
nanofibers was compared with that of cells cultured on coverslips (control), pure collagen nanofibers
and TPU nanofibers. It was revealed that all the nanofiber mats had good cell viability than coverslips
and cell viability had no obvious difference among the blend nanofiber mats at 24 hours, but cell
proliferation was very fast and the highest MTT absorbance index could reach 0.7. On days 3, among
the different blend nanofibers, the one with the TPU to collagen blend weight ratios of 3:1 had showed
better cell viability than the others, at the same time blend ratios of 3:1 had most excellent cell
viability on days 5 and 7. Meanwhile, the MTT absorbance index on day 7 descended a little could be
ascribed that there was no more available area for cells to proliferate. The fibers diameters, porosity
and mehchanical property are very important for cell growing and migration. Among the blending
system, the diameters, porosity and mechanical stress changed along with the addition of collagen.
Thus, compared with pure collagen and TPU, blend nanofibers could provide better growth condition
for cell proliferation. In our studies, the one with TPU to collagen weight ratios of 3:1 might offer the
most suitable qualification for cell culture.
Fig12 shows SEM micrographs of PIECs on TPU(A), TPU/collagen=75:25(B),
TPU/collagen=50:50(C), TPU/collagen=25:75(D) mats after seeding for 24hours. As SEM images
shown, cells could spread well both on on TPU nanofibers and TPU/collagen blend nanofibers. Fig.9
(B) and (C) behave better cell proliferation than others with TPU/collagen blend ratios of 3:1 and 1:1.
125

Fig 11 Viability of PIECs cultured on TPU, collagen/TPU and collagen nanofibers. PIECs cultured on
the coverslips acted as a negative control. Date are representative of three independent experiments
and all date points are ploted as means±SD,*P

well as on a usual nanofibrous scaffold, especially collagen/TPU blend ratio of 1:3 is the most
appropriate for endothelial cell proliferation. Blended nanofibers can improve bioactivity relative to
pure polymer nanofibers and possess the potential to refine the composition of nanofibers readily by
adjusting ingredients according to the cell type. All these dates strongly suggested the potential
application of blended nanofibers as vascular engineering scaffolds with good endothelialization.
Acknowledgement
This research was supported by the National High-Tech Research and Development Programme of
China (2008AA03Z305), Open foundation of State Key Laboratory for Modification of Chemical
Fibers and Polymer Materials, National Nature Science Foundation of China (30570503), Natural
Science Foundation of Shanghai (07ZR14001) and the 111 project (B07024)"
References:
1. Langer R, Vacanti JP. TISSUE ENGINEERING. Science 1993;260(5110):920-926.
2. Weinand C, Pomerantseva I, Neville CM, Gupta R, Weinberg E, Madisch I, et al. Hydrogel-beta-
TCP scaffolds and stem cells for tissue engineering bone. Bone 2006;38(4):555-563.
3. Bannasch H, Fohn M, Unterberg T, Bach AD, Weyand B, Stark GB. Skin tissue engineering.
Clinics in Plastic Surgery 2003;30(4):573-+.
4. Li CM, Vepari C, Jin HJ, Kim HJ, Kaplan DL. Electrospun silk-BMP-2 scaffolds for bone tissue
engineering. Biomaterials 2006;27(16):3115-3124.
5. Feng QL. Materials selection and scaffold construction for liver tissue engineering. Pricm 5: The
Fifth Pacific Rim International Conference on Advanced Materials and Processing, Pts 1-5
2005;475-479:2391-2394.
6. Chen MK, Beierle EA. Animal models for intestinal tissue engineering. Biomaterials
2004;25(9):1675-1681.
7. Neuenschwander S, Hoerstrup SP. Heart valve tissue engineering. Transplant Immunology
2004;12(3-4):359-365.
8. Levenberg S, Rouwkema J, Macdonald M, Garfein ES, Kohane DS, Darland DC, et al.
Engineering vascularized skeletal muscle tissue. Nature Biotechnology 2005;23(7):879-884.
9. Bunaprasert T, Hadlock T, Marler J, Kobler J, Cowan D, Faquin W, et al. Tissue engineered
muscle implantation for tongue reconstruction: A preliminary report. Laryngoscope
2003;113(10):1792-1797.
10. Zhang YZ, Venugopal J, Huang ZM, Lim CT, Ramakrishna S. Characterization of the surface
biocompatibility of the electrospun PCL-collagen nanofibers using fibroblasts.
Biomacromolecules 2005;6(5):2583-2589.
11. Nair LS, Bhattacharyya S, Laurencin CT. Development of novel tissue engineering scaffolds via
electrospinning. Expert Opinion on Biological Therapy 2004;4(5):659-668.
12. Webster TJ, Waid MC, McKenzie JL, Price RL, Ejiofor JU. Nano-biotechnology: carbon
nanofibres as improved neural and orthopaedic implants. Nanotechnology 2004;15(1):48-54.
13. Kim K, Yu M, Zong XH, Chiu J, Fang DF, Seo YS, et al. Control of degradation rate and
hydrophilicity in electrospun non-woven poly(D,L-lactide) nanofiber scaffolds for biomedical
applications. Biomaterials 2003;24(27):4977-4985.
14. Yang F, Murugan R, Wang S, Ramakrishna S. Electrospinning of nano/micro scale poly(L-lactic
acid) aligned fibers and their potential in neural tissue engineering. Biomaterials
2005;26(15):2603-2610.
15. Boland ED, Wnek GE, Simpson DG, Pawlowski KJ, Bowlin GL. Tailoring tissue engineering
scaffolds using electrostatic processing techniques: A study of poly(glycolic acid)
electrospinning. Journal of Macromolecular Science-Pure and Applied Chemistry
2001;38(12):1231-1243.
16. Ayutsede J, Gandhi M, Sukigara S, Ko F. Carbon nanotube reinforced Bombyx mori nanofiber
composites by the electrospinning process. Mechanical Properties of Bioinspired and Biological
Materials 2005;844:281-286.
127

17. Li WJ, Tuli R, Okafor C, Derfoul A, Danielson KG, Hall DJ, et al. A three-dimensional
nanofibrous scaffold for cartilage tissue engineering using human mesenchymal stem cells.
Biomaterials 2005;26(6):599-609.
18. Yamane S, Iwasaki N, Majima T, Funakoshi T, Masuko T, Harada K, et al. Feasibility of
chitosan-based hyaluronic acid hybrid biomaterial for a novel scaffold in cartilage tissue
engineering. Biomaterials 2005;26(6):611-619.
19. Matthews JA, Wnek GE, Simpson DG, Bowlin GL. Electrospinning of collagen nanofibers.
Biomacromolecules 2002;3(2):232-238.
20. Desai K, Kit K, Li J, Zivanovic S. Morphological and surface properties of electrospun chitosan
nanofibers. Biomacromolecules 2008 Mar;9(3):1000-1006.
21. Huang ZM, Zhang YZ, Ramakrishna S, Lim CT. Electrospinning and mechanical
characterization of gelatin nanofibers. Polymer 2004;45(15):5361-5368.
22. Min BM, Lee G, Kim SH, Nam YS, Lee TS, Park WH. Electrospinning of silk fibroin nanofibers
and its effect on the adhesion and spreading of normal human keratinocytes and fibroblasts in
vitro. Biomaterials 2004;25(7-8):1289-1297.
23. Pedicini A, Farris RJ. Mechanical behavior of electrospun polyurethane. Polymer
2003;44(22):6857-6862.
24. Tatai L, Moore TG, Adhikari R, Malherbe F, Jayasekara R, Griffiths I, et al. Thermoplastic
biodegradable polyurethanes: The effect of chain extender structure on properties and in-vitro
degradation. Biomaterials 2007;28:5407-5417.
25. Demir MM, Yilgor I, Yilgor E, Erman B. Electrospinning of polyurethane fibers. Polymer 2002
May;43(11):3303-3309.
26. He W, Yong T, Teo WE, Ma ZW, Ramakrishna S. Fabrication and endothelialization of
collagen-blended biodegradable polymer nanofibers: Potential vascular graft for blood vessel
tissue engineering. Tissue Engineering 2005;11(9-10):1574-1588.
27. Pham QP, Sharma U, Mikos AG. Electrospinning of polymeric nanofibers for tissue engineering
applications: A review. Tissue Engineering 2006;12(5):1197-1211.
28. Heydarkhan-Hagvall S, Schenke-Layland K, Dhanasopon AP, Rofail F, Smith H, Wu BM, et al.
Three-dimensional electrospun ECM-based hybrid scaffolds for cardiovascular tissue
engineering. Biomaterials 2008;29(19):2907-2914.
29. Yagi K, Michibayashi N, Kurikawa N, Nakashima Y, Mizoguchi T, Harada A, et al.
Effectiveness of fructose-modified chitosan as a scaffold for hepatocyte attachment. Biological &
Pharmaceutical Bulletin 1997;20(12):1290-1294.
30. Eda G, Liu J, Shivkumar S. Flight path of electrospun polystyrene solutions: Effects of molecular
weight and concentration. Materials Letters 2007;61(7):1451-1455.
31. Vanwachem PB, Beugeling T, Feijen J, Bantjes A, Detmers JP, Vanaken WG. INTERACTION
OF CULTURED HUMAN-ENDOTHELIAL CELLS WITH POLYMERIC SURFACES OF
DIFFERENT WETTABILITIES. Biomaterials 1985;6(6):403-408.
32. Dhiman HK, Ray AR, Panda AK. Characterization and evaluation of chitosan matrix for in vitro
growth of MCF-7 breast cancer cell lines. Biomaterials 2004;25(21):5147-5154.
33. Chatelet C, Damour O, Domard A. Influence of the degree of acetylation on some biological
properties of chitosan films. Biomaterials 2001;22(3):261-268.
34. Zong XH, Kim K, Fang DF, Ran SF, Hsiao BS, Chu B. Structure and process relationship of
electrospun bioabsorbable nanofiber membranes. Polymer 2002;43(16):4403-4412.
35. Jiang Zhi YK-j, Li Shu-fen, CHOW W.K. Study of FTIR spectra and thermal analysis of
polyurethane. Spectroscopy and Spectral Analysis 2006;26:624-628.
36. Jackson M, Choo LP, Watson PH, Halliday WC, Mantsch HH. BEWARE OF CONNECTIVE-
TISSUE PROTEINS - ASSIGNMENT AND IMPLICATIONS OF COLLAGEN
ABSORPTIONS IN INFRARED-SPECTRA OF HUMAN TISSUES. Biochimica Et Biophysica
Acta-Molecular Basis of Disease 1995;1270(1):1-6.
37. Chen ZG, Mo XM, Qing FL. Electrospinning of collagen-chitosan complex. Materials Letters
2007;61(16):3490-3494.
128

Uniform Coating of WO x on TiO 2 Nanotubes for Enhanced Electrochromic
Performance
Jian Zhen Ou 1 * , Hai Dong Zheng 1 , Kourosh Kalantar-zadeh 1
* Presenter
1. School of Electrical and Computer Engineering, RMIT University, Australia
Both WOx and TiO 2 are very good electrochromic materials due to their strong electronphonon
interaction capability. Thus, the combination of these two materials could possibly
enhance the electrochromic performance. In our work, WOx was uniformly coated on TiO 2
nanotubes using electrodeposition method from the solution of 0.5M Na 2 WO 4. . SEM and
EDX characterisation techniques were used to determine the surface morphology and film
composition respectively. Electrochromic measurement was also conducted and the device
performance was evaluated based on optical power reflective density. It was also compared
with the performance of conventional RF sputter WO3 thin film coated on FTO glass.
129

Nanoparticle trapping by surface plasmon resonance assisted thermal
forces
Jingzhi Wu 1 * , Xiaosong Gan 1
* Presenter
1. Centre for Micro-Photonics, Swinburne University of Technology, VIC, Austraila, 3122
Surface plasmon resonance of metal nanostructures allows strong field intensity
enhancement and field gradients, which is promising for trapping and manipulation of
nanoparticles. However, metallic nanostructures generate heat in the presence of
electromagnetic radiation especially under plasmon resonance. Here, we report an approach
for nanoparticles trapping explores thermal forces induced by temperature gradients in the
medium surrounding the nanostructure. Simulation results show that light absorptions of the
nanostructure generate great temperature gradients corresponding to strong thermal forces.
This suggests stable nanometric trapping can be realized with thermal forces in nanostructures.
130

Union as a Social Regulator of Market Risk?
--Empirical Evidence from Professional Union Leader Program in B City
Youqing Fan 1 * , Peter Gahan 1
* Presenter
1. Work and Employment Rights Research Centre, Department of Management, Monash
University, Caulfield East Victoria 3145, Australia
Polanyi and his followers summarized a ‘double movement’ theory to understand the
interaction of marketization and social protectionism during England’s transformation in 18th
century. In the recent 30 years, China is also experiencing a similar marketization expansion,
according to Polanyi; it is supposed to be a social protectionism developed as a
countermovement.
Despite Chinese unions’ influence and membership has been seriously undermined
during the SOE reform, there has been an attempt by some regional offices of the All-China
Federation of Trade Unions (ACFTU) to recruit professional leaders, and establish unions
independently from management. This attempt does have its potential to make Chinese union
a social regulator in the rising private sector.
This paper draws empirical evidence from professional union leader program in B city to
examine the question that ‘to what extent does this professional union leader program has
made union as a social regulator to protect workers from market risks?’
1. Introduction
Writing near the mid-point of the last century, Karl Polanyi wrote his seminal work, The Great
Transformation (1944[1957]), which sought to provide an understanding of England’s transformation.
Central to his analysis was the Polanyi’s analysis of these processes have proved enduring,
particularly over the last decade, as a useful approach to examining processes of globalisation (e.g.,
Munck, 2002; Standing, 2007) and rise of new transnational social movements seeking to reverse or
ameliorate the new risks associated with globalisation (Edelman, 2005; Evans, 2008).
To what extent have Chinese unions developed an effective countermovement to the expansion of
market? There are, of course, stark differences between China’s economic transformation and
England’s industrial revolution. But we would suggest that Polanyi’s conceptual framework for
understanding the economic processes associated with industrialisation and subsequent responses to it
– notably the idea of the double movement and social embeddedness of economic institutions –
provide a useful approach for thinking about China’s economic transformation and the social
responses to this transformation. Here, we do not provide any attempt to analyse the transition from a
Polanyian perspective (for an early attempt to incorporate Polanyi ideas into such an analysis, see Nee,
1992; 1996; and more recently Deyo and Ağarton, 2007). Our aim is to sketch briefly Polanyi’s key
ideas and examine the role of professional union leader program in protecting workers against market
risks in China’s new economy.
Thus, the paper will: firstly, introduce Polanyi’s ‘Double Movement’ theory; secondly, briefly
summarize China’s economic transition and changing role of Chinese trade union; thirdly, discuss the
role of professional union leader program in protecting workers against market risks in China’s new
economy; and finally make some concluding comments.
2. Polanyi and the Concept of “Double Movement”
In his book, The Great Transformation (1944), Karl Polanyi described the impact of
industrialisation on established institutions that had regulated production and labour markets.
Industrialisation brought with it an extension of market-based regulation to new areas of economic and
social relations, including the labour market. Consequently, the regulation of labour through the
132

system of guilds and government regulation gave way to a growing reliance on markets and price
signals, permitting employers to hire and fire when required, without any broader obligation for
worker welfare. While industrialisation and market liberalisation may have improved overall
economic welfare, it subordinated all other relationships to economic concerns, increasing worker
exposure to risk. The unleashing of market capitalism, Polanyi posited, gave rise a ‘countermovement’,
a movement for greater social protection against market forces:
“It can be personified as the action of two organising principles of society... The one was the
principle of economic liberalism, aiming at the establishment of a self-regulating market... using
largely laisez-faire and free trade as its methods; the other was the principle social protection
aiming at the conservation of man and nature...using protective legislation, restrictive
associations, and other instruments of intervention as its methods (Polanyi, 1944[1957]: p. 132).
Labour unions were one form that the second of these organising principles manifested itself as a
‘countermovement’ to the expansion of markets and free trade as institutions regulating economic life.
Unions, along with other groups, sought to regulate the terms on which employers could hire and fire
workers, as well as to agitate for legislation that afforded greater social protection. This
countermovement reflected the inevitably process by which the extension of market relations became
“dis-embedded” socially, disconnected from its social and political institutions, thereby generating
insecurity and social anxiety (Munck, 2002: p. 18). Counter movements formed a reaction to these
effects; and were attempts by social actors to re-embed markets into a more sustainable framework
(Dale 2008). These opposing forces, Polanyi argued, co-exist in an uneasy tension, both having the
potential to undermine the logic of the other. Just as ‘blind market relations’ resulted in dis-embedded
social institutions, social protectionism undermined the efficiency and effectiveness of markets.
Scholars seeking to translate Polanyi’s logic of the double movement to the twenty first century
have typically begun with the notion of globalisation (Edelman, 2005; Standing, 2007; Evans, 2008),
although as Evans (2008, p.273) in particular has pointed out Polanyi’s critique of industrialism is
squarely centred on the role of countermovements within relatively autonomous nation states.
Consequently, the scope for countermovement and new forms of protectionism under globalization
necessitate social actors mobilising at a transnational level. As such, countermovements against
globalisation are more problematic and are difficult to sustain.
3. China’s Economic Transition and the Changing Role of Unions
Transition to a market economy in China has now extended over a thirty year period. This process
has been widely documented by a number of scholars in the research on China’s economic transition
(e.g., Naughton, 1995; 2007; Wu, 2005; Qian, 2003). It has consisted of successive waves of reforms
and, in contrast to the ‘shock therapy’ administered to the command economies of Eastern Europe
(McMillan 2004), the Chinese approach to transition has been a gradualist one (Child and Tse, 2001).
The transition process has typically been characterized as consisting of two distinct phases
(Naughton, 2007). In the first phase, reformers established a ‘dual track’ economic system in which
reforms allowed for an increasing level of market activity alongside the planned economy (Lin et al.,
1998; Lau et al., 2000). While this phase allowed for growing competition in product markets through
the formation of town and village enterprises (TVEs), domestic private enterprises (POEs), joint
venture enterprises (JVEs) with foreign multinationals, and wholly foreign owned enterprises (FOEs),
the activities of SOEs remained virtually untouched and the development of competitive labor markets
was weak. Then, from 1992, a series of reforms were introduced intended to integrate these dual tracks
through the strengthening of market institutions, particularly in the area of labor and capital markets,
and exposing SOEs to greater market competition through restructuring, privatization and corporate
governance reforms (Qian, 2003).
As a consequence of these reforms, China has experienced a transformation in the institutional,
technological and organizational foundations of industry (Naughton, 2007). New industries have been
created, and existing ones have undergone major structural reform. These reforms included policies
and institutional changes intended to promote great labour mobility across sectors and regions,
incentives for greater direct foreign investment (DFI), the privatization of state-owned enterprises
(SOEs) and the promotion of domestic private firms. These developments have, in turn, been
133

associated with a reinvention of institutions governing industrial relations and labor management,
which have inevitably involved the creation of more conventional labour market (Shen, 2008; Frenkel
and Kuruvilla, 2002). Consistent with the more general policy orientation toward market
liberalization, these policy shifts were aimed at creating a free labour market and enabling individual
enterprises to determine a broad range of issues at the workplace level, including recruitment and
dismissal, wages and other forms of remuneration, and investments in training –see Table 1 for a
summary of major labour law reforms. These reforms have not only altered the structure and dynamics
of China’s economy, but have also profoundly altered the exposure of workers to market risk. In this
context our aim is to consider the emergence of new forms of union organisation as a
‘countermovement’ against the extension of the market, providing workers with new forms of social
protection.
Table 1. Summary of Main Labour Law Reforms in China, 1992-2008
Law/Regulations
Major Contribution
Labour Union Law (1992)
Stipulating the status, obligations and rights of trade union in
new economic system
Regulations On Enterprises Labour
Disputes Settlement(1993)
Labour Law (1994)
Labour Union Law (2001)
Labour Contract Law (2008)
Labour Dispute, Mediation and
Arbitration Law (2008)
Setting up framework for labour dispute resolution in new
economic system
Basic legal rules for labour market and of labour relations
mediation
More emphasis on unions’ role of worker representation and
participation in labour process
More pro-labour provisions:
-Workplace and occupational health provisions;
-Work time and holiday provisions;
-Compensative payment for pre-maturity contract
termination.
Provisions to make arbitration procedures more accessible:
-Arbitration terminated in the first instance;
-Free of charge for arbitration;
-Employers have more ‘burden of proof’ responsibilities in
arbitration.
The Role of Unions in Socialist China
The peak union organization, the All-China Federation of Trade Union (ACFTU) was established
as the formal union movements under the leadership of the Communist Party of Chinese (CPC
hereafter), and incorporated other union movements. Officially, “the trade unions of China are a
voluntarily organized mass organizations of Chinese working class led by the CPC and formed by the
staff and members voluntarily, are the bridges and transmission belt linking the Party and the masses
of workers and staff members, are the important social pillar of the state power, and are the
representatives of the interests of the trade union members workers and staff” (Chinese Union Charter,
2003).
The ACFTU formally functions as both a representative institution of workers and a state
institution. In private workplaces, grassroots unions are largely affiliated to the management (Shen,
2007, p.236-237). Chinese unions assumed three objectives:
1) to assist state administrative function;
2) to collaborate with enterprise management to improve production efficiency; and
3) to represent and protect the interests of employees (Ge, 2007, p.2).
All trade union organisations at the provincial, local and workplace levels are required by
law to be affiliated to the ACFTU.
134

Chinese Unionism in Reform Period
As noted above, the economic reforms over the last 30 years have been associated with significant
changes in the industrial relations framework in China. The major legal reforms have formally
provided workplace unions with an expanded role in collective bargaining and greater capacity to
represent members in the workplace. What is not clear, however, is whether these developments have
been associated with any substantial change in the role of unions at the workplace, or the nature of the
relationship between unions and their members.
The available evidence indicated that the restructure of Chinese industry has created a number of
adverse effects for unions, most notably in relation to union membership. Figure 1 graphs union
membership for the period 1980 to 2005. This figure shows that from the mid 1990s, membership fell
in absolute terms before increasing again after 1999. A second major decline in membership is then
evident in 2003, after which membership increased again. This decline in membership during the
1990s has been attributed to the effects of the economic reforms, particularly the restructuring of state
owned enterprises (SOEs). Taylor and Li (2005) estimate that between 1998 and 2005, SOEs
dismissed approximately 60 percent of their workforce, or around 30 million employees. Although
there are no reliable statistics, the evidence also indicates that unions have found it difficult to recruit
and retain members in foreign invested forms (Chan, 2006).
Following this period of restructuring, the ACFTU sought to reverse the decline member by
organise the emerging private sector (Chan, 2006). This is evident in both the union membership
figures and the decline and subsequent growth in workplace union organisations. Both of the two
indicators are summarised in Figure 1. This data likewise reveals the decline in union organisation
during the mid 1990s and subsequent recovery after the turn of the century.
Figure 1. Chinese Trade Union Organization and Membership Change Trend (1980-2000)
Given the limited capacity of organizing foreign enterprises rapidly, Chan (2006) reports that the
ACFTU has no choice but accept management compliant unionism. The main pattern of workplace
union branches organizing strategy in foreign-funded enterprises is “top-down”, of which process and
basic features has been summarized by Chan (2006) as follows:
“The district-level union would have sought management approval and cooperation to set up
a union branch. Once an agreement was struck, management and the local union would have
decided together on a mid-level PRC Chinese manager to serve as the union chair, without a
union election. After the fact, an announcement would have been made to the employees about
the formation of a new union branch, or in some cases, no announcement would have been made
at all… [Such union chairs] have no power or independence from management. More often than
not, this “union” does not even perform the traditional welfare functions it does in state-owned
135

enterprises…There obviously is no collective bargaining or other actions that we associate with
unions.” (Chan, 2006)
This subordinate role has been reinforced by unions’ financial and personnel dependence on the
enterprise (Chan, 1998, pp.134-140; Shen, 2007, p.236). As a consequence, unions have not been
capable or willing to represent workers, but have generally taken a pro-management or at least not
pro-labour when labour disputes happen (Taylor and Li, 2005, pp.3-4). As a consequence, the fast
growing in private sector unionism has not been associated with effective representation of workers
interests in the workplace.
4. Empirical Evidence--Professional Union Organizer/Leader Program in B City
Faced with a new type of capitalism -- neo-liberal globalization -- the surge of unionism in China
shows signs of taking different forms from the past to counter the new risks. There are some reports
of a recent attempt by regional offices of the ACFTU to recruit professional union leaders and form
region-based unions, which is independent from management. In particular, this paper examines the
role of these unions in protecting labour from market risks.
Fieldwork Method
Empirical evidence of this case is mainly originated from a major fieldwork. The fieldtrip involves:
1) a major interview with district level, sub-district level union leaders and professional union leaders;
2) a special interview with 2 professional leaders; and 3) a half-day long participative observation in
professional union leaders’ daily work. Government policy documents are also used as secondary data
source.
Case Description
Since 2002, a number of regional branches of the ACFTU started to try out two top-down trials
aiming at organizing workers and form independent leadership from management in workplace/local
unions. Among these attempts, the practice in P sub-district, S district, B city is a unique combination
of recruiting professional union organizers and leaders, which represents characteristics of both the
professional leader and organizer program. The critical innovations are:
1. Recruit organizers outside the firms to organize rank-and-file workers and establish
local united unions;
2. Get the organizers elected as union leaders, and get them paid by various levels of
unions rather than by management.
P sub-district is a suburb right lies on the border of rural and urban area of B city with 110
thousand populations. Among all the 900 employers in this sub-district, around 600 are small business.
In recent years, urban expansion transfers all the former farmers in the sub-district into urban residents.
Some of them set up their own business in various occupations with 2-8 employees.
Faced with the complex situation, P sub-district union leader initiated an innovative form of
grassroots union, which is called “Small Firms United Grassroots Union”. The sub-district level union
leader explains why they adopt this new form of grassroots union:
We have lots of petty shops in our area. These shops are too tiny to be organized as an
individual grassroots union, since at least 25 employees are required by the Trade Union Law
for a workplace to be eligible for setting up an individual grassroots union. If we set up unions
in these petty firms, it could leads to two different situations. One possibility is that the boss
holds a concurrent post as union president, which will undermine union’s role of maintaining
workers’ interests. Especially, these petty unions are mostly run by former farmers, who lack
of sense of conforming to labour legislations. The other possibility is rank-and-file workers
take the position of workplace union president, which will put themselves in danger of being
fired by their boss when they try to bargain with boss over labour relations issues, which is
simply because they are hired by, and get remuneration from, their boss. Both of the
possibilities will not be favourable for union’s worker representation.
136

Obviously, the trade union leaders realize that if they want P sub-district union to represent
workers, firstly, they should organize those small firms to cover the workers; and secondly, make
grassroots union leader independent from management both in personnel and in finance. Consequently,
after consulting with district level union leaders, they worked out a solution: first, recruit union
organizers outside the firms and get them paid by upper level union; second, dispatch them to organize
the tiny firms, and then sign collective contract with the firm employers; third, during the organizing
campaign, union member representatives are selected from the firms and convened to attend a ‘small
firm grassroots union establishment congress’ and directly elect the organizers as united grassroots
union leader; finally, use revisit as main method to detect labour complaint and maintaining workers’
rights.
The above blueprint forms a framework for the professional union leader practice. In the following
paragraphs, we will describe the implementation of the practice in detail.
Personnel and Financial Sources of Organizing Campaign
To secure personnel independence, initially, P sub-district level union leader intended to recruit
‘union assistants’ from society. However, there is no special funding for recruiting such union
assistants within the union system. As a result, the city and district level union leaders requested
Municipal Labour Bureau for using the Social Welfare Fund, which is administrated by the Bureau, to
fund recruitment of union leaders. Finally, the Bureau approved the request on the condition that the
recruited assistant must be the “4050” unemployed. 1
Since 2007, 27 “4050” unemployed party member have been selected and recruited by sub-district
level unions as union assistants. Municipal Labour Bureau pays salary (at the minimum wage level in
B city) and social insurance contributions for them. P sub-district also immediately recruited 2 union
assistants after securing the financial source. The recruited leaders are both SOE laid-off workers.
Organizing Process
P sub-district is divided by an ancient river into two sub-areas. The numbers of small businesses in
each sub-area are nearly the same. Therefore, before the proposed organizing campaign, sub-district
level union leader divided the whole sub-district into two zones, ‘river east’ and ‘river west’. The
assistants are separately responsible for organizing and setting up united grassroots union in 1 zone.
The main organizing targets are small businesses with less than 8 employees.
Formal organizing campaign was initiated in May 2007 with a focus on organizing the river east
zone. The campaign adopted a door to door organizing method, which involves ask each employer in
the zone to participate in the proposed united grassroots union. In the process, normally, union
assistants firstly try to talk to bosses and explain the function of Chinese unions and benefits of
participating in union. After gaining the permission from bosses, assistants would require bosses to
sign an application form for participating union and a prepared collective agreement, and then acquire
a copy of their business licence. After 3 months organizing, about 75 firms have been organized in this
manner.
Union Election
In order to convene union establishment congress and elect union leaders, union assistants also
require each unionized firm to nominate a rank-and-file employee as union member representative
during organizing campaign. Finally, in the already organized 75 firms, 50 representatives were
nominated. Nearly 4/5 of them are rank-and-file employees. The rest 8-9 representatives are bosses,
who are all party members selected by the union assistants. This measure secures quality of member
representative team.
1 The “4050” unemployed mainly refers to those middle-aged former SOE workers who got unemployed in SOE
reform. Normally, the female workers are in their 40s, and male workers are in their 50s. Basically, they are
associated with low job skills except their original job skills, while still have 1-10 years to be eligible for pension
benefits. Thus, it is quite critical while hard for them to find a job in labour market. Thus, the government
prioritizes the 4050 unemployed to take casual positions in public sector, such as union assistant, traffic assistant,
community assistant etc.
137

Union assistants participate in organizing campaign, thus they are quite familiar with union
members. Once they become union leaders, they are more likely to win trust of members.
Union Funding
Lacking of enough funding becomes the most serious problem for the new united grassroots unions.
The main financial source of the united grassroots unions comes from employers’ contribution of
union expenditure, which is 2% of the total payroll. However, nearly all the participant firms don’t
transfer union dues to their union accounts. S district level union leader attributes this to the weak
Trade Union Law, which doesn’t stipulate any penalty for union expenditure arrears. Other sources of
funding are almost unavailable: unions are not allowed to receive any donation; financial support from
the government is quite limited, and only focuses on support city and district level unions.
Union Activities
The primary activities of the united grassroots union mainly include signing collective agreements
and revisiting small firm workplaces.
Firstly, during the organizing campaign, collective agreements were signed between the proposed
union and employers, which includes collective contract, wage agreement and female worker
agreement. The content is all the same for different firms. Terms and conditions in the agreement is
almost a repeat of Chinese Labour Law and Municipal Wage Bylaw of B City over minimum wage and
working condition issues. Sub-district level union leader explained:
The union is just a mass organization, it doesn’t have any power to force firms to sign
collective contract. So, in private firms, a feasible way is to lower the collective contract terms
in exchange for bosses’ willingness of signing collective contracts. After signing the contract,
at least, we can monitor the implementation of collective contract and guarantee the minimum
wage and basic working conditions for our members.
Secondly, the union presidents/union assistants pay revisits to workplace, and have informal
discussion with workers. The revisit bears multiple roles: first, to get to know their members’ needs
and concerns at their workplace; second, to monitor the implementation of collective agreements; third,
do some labour legislation propaganda.
Normally, if there is a labour dispute detected during revisit, united grassroots union president
would call on employer or manager for a mediation meeting. If dispute cannot be mediated, then
united grassroots union president will report this to the district level legal aid centre, and lawyers there
will help the member out by representing him/her at labour tribunal or court. Some labour disputes
have been successfully mediated and solved by the union. For instance, one union member wanted to
get medical reimbursement of his kid, the union coordinated this issue with local residence community,
and finally got the reimbursement for the member.
At present, few members go to grassroots union to address their problems on their own initiative.
However, the grassroots leaders believe that if they continue the revisit practice, members will get to
know them, trust them and are willing to address their problems to them.
The next two tasks of the united grassroots union are setting up a union service centre and ten
service hubs based on the centre, with an aim to provide special service to all current members
according to members’ interests. For example, sports hub could organize training and matches for
members who like doing sports. The service centre is located in the sub-district government building,
equipped with reading room and meeting spaces.
5. Discussion: Professional Leader as a Regulator of Market Risks
After reviewing double movement theory and professional union leader practice, we back to our
question: ‘to what extent has the practice developed an effective countermovement to the expansion of
market risks in China’s new economy? In other words, what is the effectiveness of the new form in
protecting their members from labour market risks?’ In this section, we will identify the evidence of
worker protection in the practice.
138

In terms of union internal structure, firstly, this practice organizes membership in small firms,
which are almost not organized at all before; secondly, the initiative innovatively creates a form of
‘united grassroots union’ to cover as many petty shop employees as possible in the area.
In terms of union-management relations, through the professional leader programs, grassroots
leaders gain the personnel and financial independency from the management, which enables them to
genuinely represent worker to balance management extend union’s influence into workplace.
In reference to union-member relations, firstly, the organizing campaign led by professional union
leaders extends unions’ influence to both small private businesses and precarious employees within
those small shops; secondly, after the establishment of united grassroots union, rural migrant workers
working in P sub-district can find a place to address their labour relations concerns; thirdly, the union
plays the role of workplace labour dispute mediator and contact person for district level legal aid
centre; Finally, the grassroots union leaders’ “revisit” has formed a kind of surveillance to firm bosses
on their conformity to labour legislations and collective contract.
In terms of union democracy, mutual-trust has been established to some extent between
professional leaders/organizers and rank-and-file members during the union organizing campaign and
following ‘revisit’ process. Based on the mutual-trust, more member participation on union
governance and union activities could be expected.
Besides, the practice, together with union professional leader trials in other cities, influences the
revision of the ACFTU regulations, which will get the practice implemented throughout all the
branches of the ACFTU. In October 2008, an amendment of the Constitution of the All-China
Federation of Trade Unions, Chapter 6, provision 34, two more clauses were added to further stipulate
the personnel and financial independence from the management, which are, “county level unions and
union branches above the county level should set up a fund for union leadership rights and interests
protection” and “county level unions and union branches above the county level may select and
employ union officials for grassroots unions”.
6. Concluding Comments
The union leader/organizer professionalization program emerged within Chinese official unions
suggests a clear response to the market liberalization process in the last 30 years reform in China. This
response manifests itself by involving organizing more membership, more independency of grassroots
unions from management and more formal way of worker presentation, and expected closer and more
interactive union-member relationships.
All these changes have not only paved a way for unions to represent their members, but also
formed a relatively formal institution to protect their union members from market risks. The increase
of union membership and influence helps unions get involved in the new industrial relations
framework; independency from management enables union leaders to balance interests of
management and workers; the promotion of union-member relationships makes union rank-and-file
members participate more in union activities.
To sum up, despite its limitations, the professional union leader practice does have its subtle
implications that Chinese unions do have their potential to act as social actor being able to set up a
countermovement against market risks exerted by globalized capitalism, like other social actors
mentioned in the Polanyi’s theory, e.g. farmers and consumer groups.
References:
[1] ACFTU (2003). Chinese Union Charter (Vol. General Principle, Paragraph 1).
[2] Chan, A. (1998). Labour Relations in Foreign-funded Ventures. In G. O'Leary (Ed.), Adjusting to
Capitalism: Chinese Workers and the State (pp. xviii, 181 p.). Armonk, N.Y. London: M.E.
Sharpe.
[3] Chan, Anita (2006) "Organizing Wal-Mart: The Chinese Trade Union at a Crossroads," Japan
Focus, September 8, 2006, http://japanfocus.org/products/topdf/2217
[4] Dale, Gareth (2008) “Karl Polanyi’s The Great Transformation: Perverse Effects, Protectionism
and Gemeinschaft.” Economy and Society, 37(4): 495-524.
139

[5] Deyo, Frederic C and Ağartan, Kaan (2007) “Reforming East Asian Labor Systems: China, Korea
and Thailand.” In Ayşe Buğra and Kaan Ağartan, eds., Reading Karl Polanyi for the Twenty-First
Century: Market Economy as a Political Project, Palgrave Macmillan, New York, pp. 191-219.
[6] Edelman, Marc (2005) “Bringing the Moral Economy Back in…to the Study of 21st-Century
Transnational Peasant Movements.” American Anthropologist, 107(3): 331-345.
[7] Evans, Peter (2008) “Is an Alternative to Globalization Possible?” Politics and Society, 36(2):
271-305.
[8] Ge, Y. (2007). What Do Unions Do in China? : SSRN Working Paper.
[9] Huang, X., and Ji, L. (2006, May 11st) Report on Trade Unions in Foreign-invested Companies
(Waiqi Gonghui Shengcun Diaocha), Southern Weekly (Nanfang Zhoumo), May 11st, 2006.
[10] Ma, H., Wu, D. (2008) Guangzhou Trade Union Organizers Officially Appear Today (Guangzhou
Gonghui Zuzhiyuan Jinzao Liangxiang) Yangcheng Evening News (Yangcheng Wanbao), August
6th, 2008.
[11] Munch, Ronaldo (2002) “Globalization and Democracy: A New “Great Transformation”?”
Annals of the American Academy of Political and Social Science, 581(May): 10-21.
[12] Nee, Victor (1992) “Organizational Dynamics of Market Transition: Hybrid Forms, Property
Rights and Mixed Economy in China.” Administrative Science Quarterly, 37: 1-27.
[13] Nee, Victor (1996) “The Emergence of a Market Society: Changing Mechanisms of Stratification
in China.” American Journal of Sociology, 101(4): 908-949.
[14] Ngok, K.L. (2007) Chinese Labor Policy: A Market and Globalization Perspective, Beijing:
Social Sciences Academic Press.
[15] Polanyi, K (1944[1957]) The Great Transformation: The Political and Economic Origins of Our
Time, Beacon Press: Boston.
[16] Sang, W. (2007) Guangzhou: 100 Union Organizer Entering the Firms (Guangzhou: Baiming
Gonghui Zuzhiyuan Jin Qiye) Retrieved 18th Oct, 2008, from
http://www.acftu.net/template/10004/file.jsp?cid=731&aid=78071
[17] Shen, J. (2007). Labour disputes and their resolution in China. Oxford: Chandos.
[18] Standing, Guy (2007) “Labor Recommodification in the Global Transformation.” In Ayşe Buğra
and Kaan Ağartan, eds., Reading Karl Polanyi for the Twenty-First century: Market Economy as
a Political Project, Palgrave Macmillan, New York, pp. 67-94.
[19] Taylor, B., and Li, Q. (2005). Review on ACFTU’s Union Organizing Practice. Paper presented
at the Globalization and Labour Turnover in India and China, Monash University, Australia.
[20] Taylor, B. (2008). Does China Have a Labour Movement? Prospects for Industrial Relations
Reform in China. Montreal: CRIMT.
[21] Taylor, B., and Li, Q. (2007). Is the ACFTU a Union and Does it matter? Journal of Industrial
Relations, 49(5), 701-715.
[22] Unger, J., and Chan, A. (1995). China, Corporatism and the East Asian Model. Australian Journal
of Chinese Affairs, 33, 29-53.
[23] Wang, W. (2007) Union Organizer System Initiated in Our Province (Wosheng Zai Quanguo
Shouchuang Gonghui Zuzhiyuan Zhidu), Hebei Daily (Hebei Ribao), June 20th, 2007.
140

Developing Dynamic Capability through Partnership: The Role of
Capabilities
Wei Jiang 1 * , Felix Mavondo 1
* Presenter
1. Department of Marketing, Monash University, Clayton Compus, Wellington Rd, Clayton
3800, Australia
Partnerships have been adopted by most organizations as a major source of competitive
advantage and to manage technological turbulence and dynamic market environments. The
paper aims to investigate some key organizational capabilities in partnerships that were
proposed to contribute to the improvement of the organization’s agility—a dynamic capability
for managing capricious environments. The role of trust between partnering organizations as a
moderating variable is examined. A samples (n=300) from Chinese managers in
manufacturing industries was used for this study. The results indicate that trust is an important
moderator of the relationship between market orientation and learning orientation and
organisational agility.
1. Introduction
The business environment has been described in the literature as hyper turbulent, unpredictable,
hypercompetitive [39] [42] [45]. Environmental changes force acquired skills and capabilities to
become obsolete or create new opportunities, both of which may require firms to build new
capabilities [3]. Failure or slowness of building new capabilities endangers firm’s prosperity or even
survival. Such environments, thus, have been recognized as the cause of most organizations’ failures
[38]. Dynamic capability theory contends that dynamic capabilities underlie the source of sustainable
competitive advantage in such market environments; they are the important mechanisms for
reconfiguring, recombining and deleting ordinary organizational resources to achieve a fit with the
environment and strategic imperatives [43] [50] [28]. Despite its importance to firm development and
the scholarly attention devoted to it, dynamic capability remains underspecified [3], and empirical
work is still in its infancy [36] [55]. It has been generally acknowledged in the literature that firms can
develop dynamic capabilities through long-term partnerships [53], but ‘the how question’ has not been
systematically conceptualized or empirically grounded. To address this gap, this research seeks to
examine the implications of organizational capabilities for the development of dynamic capabilities
and to identify any boundary limiting conditions. Since trust is an important relational variable this is
conceptualized as a possible boundary condition—a moderator of the various relationships.
The paper is organized as follows: first, the conceptual framework is presented. Second, relevant
literatures and propositions are presented, followed by a discussion of methodological issues. Finally,
the results and discussion are presented with limitations of the study and its potential implications for
managers and academics.
2. Literature review
2.1. Dynamic capability
Dynamic capability theory provides a compelling explanation for the ability of some companies to
continuously create, define, discover and exploit entrepreneurial opportunities [55]. It proposes that
possessing strategic resources and capabilities is not a sufficient condition to generate rents, but that
the dynamic capabilities which grow and evolve in response to the external environment are the real
source of rent generation [44]. Capabilities are the skills and routines that combine both tangible and
intangible resources together coherently in a synergistic manner to enable firms to produce efficiently
and/or effectively valued market offerings [5]. While dynamic capabilities can be any capabilities
which enable firm to modify its resources and routines to changing environment, so it can continue
producing market offerings efficiently and/or effectively [28] [55].
141

Organizational Capabilities
• Learning Orientation
• Market Orientation
• Manufacturing Capability
• Managerial Capability
Trust
Dynamic Capability
• Agility Capability
Agility, which emerged since 1990s in the literature is considered a major capability that
emphasizing the ability to adjust and change [39]. Agility is the ability to survive and prosper in a
competitive environment with continuous and unpredictable changes by reacting quickly and
effectively [12]. Some important attributes of agility emphasized in all definitions include: speed,
effective responsiveness; proactiveness; and availability of slack resources [57]. Speed is the firm’s
ability to accomplish tasks in the shortest possible time, such tasks includes quick new product
development, fast operation, quick learning of new technology, and fast adaptation to change, etc. [39].
Responsiveness is the ability to identify changes and opportunities, respond reactively or proactively
to them, and recover from them [38]. Proactiveness is the capability to act proactively, taking initiative
in improving current circumstances or creating new favorable ones [2]. Organizational slack is defined
by Bourgeois (1981, p. 29) as a cushion of actual or potential resources, which firms can use to initiate
changes in strategies for environment adapting; they are accumulated for pursuing market and
competitive opportunities in the future [43].
2.2. Organizational capabilities
Four key organizational capabilities are identified from the literature as significant for firm’s
competitive advantages. They are market orientation, learning orientation, managerial capability and
manufacturing capability. Market orientation has been defined as an important organizational culture
that creates the necessary behaviors for understanding and fulfilling customers’ expressed needs [32],
and also latent needs [35]. The concept of market orientation is focusing continuously on discovering
new opportunities for target-customers, especially under dynamic market or competitive environments
[34]. Managerial capability is innate and learned abilities, and expertise of managers in a firm [1] [14]
[24]. It directly affects the firm’s performance by integrating, reintegrating firms’ resources and
capabilities [29], and the partnership performance by the managers’ coordination and control of
operation [8]. Learning orientation “is comprised by the continually evolving knowledge stocks
existing in individuals, groups and the organization, which flow to continuously exploit and explore
knowledge in accordance with the environmental conditions” [37] (p.501). It allows the transfer of
tacit knowledge, resources and capabilities in the partnership, thus providing opportunities to develop
new capabilities [7]. Manufacturing capability directly influences a firm’s success because such a
capability has the potential to support and shape corporate strategy [11]. Various studies have
indicated the direct impact of manufacturing capability on the manufacturing and organizational
performance [25]. A core manufacturing capability is a multifaceted complex concept and is multidimensional
and implies cost, quality, delivery, and flexibility [11] [25]. Based on the above, we
develop the following hypotheses:
H1: Market Orientation is positively associated with Agility.
H2: Learning Orientation is positively associated with Agility.
H3: Managerial Capability is positively associated with Agility.
142

H4: Manufacturing Capability is positively associated with Agility.
2.3. Trust
In economic exchange, trust implies a general expectation of good faith efforts by parties to honor
commitments, to be honest in negotiations, and to decry opportunistic behavior. Trust has been argued
as playing a critical role in creating synergistic outcomes in partnership [2] [32]. First, trust is regarded
as one of the most widely acknowledged means for governing and coordinating inter-organizational
exchange [18] [31]. Second, partnerships of two or more companies produce a strong potential for
dysfunctional conflicts, where trust is a key factor to make cooperation smooth by alleviating the
functional conflicts, facilitating mutual understanding and bilateral communication [2] [48]. Third,
trust facilitates knowledge exchange, both in quantity and quality, and learning in alliances [48]. Thus,
the hypotheses are stated as:
H5: Trust is positively associated with Agility.
H6: Trust moderates the relationship between Market Orientation and Agility.
H7: Trust moderates the relationship between Learning Orientation and Agility.
H8: Trust moderates the relationship between Managerial Orientation and Agility.
H9: Trust moderates the relationship between Manufacturing Orientation and Agility.
3. Research methodology
Inspired from the literature in typologies of strategic alliances [4], the long-term partnership defined
in this study includes: (1) long-term sourcing agreements (major customer or major supplier), (2) joint
R&D, (3) joint marketing and promotion, and (4) joint manufacturing. In this research, the unit of
analysis is individual manufacturing firm in long-term partnership where at least one partner is a
Chinese firm. In the partnership which involves more than two companies, the main partner will be
only one considered. Senior managers in various manufacturing industries were selected as
respondents. Mail-out questionnaires were employed in seven manufacturing provinces in China,
including major cities representing each part of China. A usable sample of 300 was received
representing a 35% response rate.
4. Results and discussions
Table 1 shows the correlations and reliabilities of the constructs in the model. In Table 2 the results
of hypothesis testing are presented. In Model 2 Market orientation is significantly related to marketing
agility (p

Table 1. Correlations, Reliabilities of the variables in the Model
n=300 1 2 3 4 5 6
1 Market orientation 0.932
2 Learning orientation 0.488 0.941
3 Managerial capability 0.540 0.668 0.885
4 Manufacturing capability 0.477 0.540 0.618 0.891
5 Trust 0.532 0.402 0.418 0.351 0.881
6 Agility 0.776 0.524 0.547 0.481 0.486 0.936
Mean 5.335 5.274 5.609 5.449 5.845 5.331
Standard deviation 0.895 0.986 0.816 0.900 0.870 0.857
Note: The diagonal (in italics and bold) shows the Cronbach’s α for each construct
Table 2: Regression models for the antecedents of Marketing Agility
Variables
Model 1 Model 2
β t-value β t-value
H1: Market Orientation 0.622*** 13.379 -0.644 -1.103
H2: Learning Capability 0.118** 2.402 1.072*** 2.557
H3: Managerial Capability 0.074 1.396 -0.924 -1.471
H4: Manufacturing Capability 0.055 1.186 0.507 1.322
H5: Trust 0.058 1.361 -0.963* -1.763
H6: Trust x Market Orientation 2.014* 2.134
H7: Trust x Learning
-1.405* -2.284
orientation
H8: Trust x Managerial
1.768* 1.682
Capability
H9: Trust x Manufacturing
-0.713 -1.223
Capability
R 2 0.640 0.656
Adj R 2 0.634 0.646
F-Ratio 104.454 61.565
Table 3: Investigation of Slopes of Market and Learning Orientation at different levels of Trust
Low (t-value) Medium (t-value) High (t-value)
Market Orientation 0.6530 (11.1114) 0.6870 (16.5933) 0.7211 (14.0508)
Learning Orientation 0.3103 (5.4074) 0.3413 (7.7685) 0.3724 (6.4487)
5. Conclusions
In this paper, we identified the four key organizational capabilities, decomposed their effects on the
development of agility respectively, and examined the moderating role of trust in such relationships.
The proposed model showed that all the four capabilities have positive and significant impacts on
agility. Trust moderates the impacts of capabilities on agility, but only for market orientation, learning
orientation. That means in a partnership with high level of trust, the contribution of market orientation
and learning orientation to the development of agility capability is higher than in a partnership with
lower trust. Market orientation and learning orientation are the most critical input to marketing agilitydynamic
capability in a partnership. Very surprisingly, data from Chinese manufacturing industries
showed that manufacturing capability does not influence the development of agility in the partnership
no matter whether the level of trust that exists.
This study has several academic contributions in terms of empirically examining dynamic
capability, its operationalisation, and the relationship between ordinary capabilities and dynamic
capabilities. It extends studies in dynamic capability theory into inter-organizational partnerships.
Moreover, the study contributes to the literature related to business issues in China.
144

6. Managerial implication and limitations
This research also offers a number of implications that could add to our understanding of
competence-centered practices and their contribution to firm agility. China has been considered as one
of the biggest emerging markets in the world for its rapid and sustainable development, after the
transforming from centralized economy to a more market-orientated economy [51]. Firm agility has
become extremely important to keep firms energetic enough to survive in the dynamic global
environment.
This study has tested one moderator—trust, which is only one aspect of the partnership connection,
in the cooperation process. Other factors, such as type of partnership, length of the partnership,
conflict management, commitment, et al., may indicates various degree of influences on the
capabilities’ contribution to dynamic capability development.
References:
[1]. Castanias, R. P., & Helfat, C. E. (2001). The managerial rents model: theory and empirical
analysis. Journal of Management, 27(6), 661-678.
[2]. Crant, J. M. (2000). Proactive behaviour in organizations. Journal of Management, 26(3), 435-462.
[3]. Cullen, J. B., Johnson, J. L., & Sakano, T. (2000). Success through commitment and trust: the soft
side of strategic alliance management. Journal of World Business, 35(3), 223-241.
[4]. Danneels, E. (2008). Organizational antecedents of second-order competences. Strategic
Management Journal, 29(5), 519-543.
[5]. Das, T. K., & Teng, B.-S. (2000). A resource-based theory of strategic alliances. Journal of
Management, 26(1), 31-61.
[6]. Day, G. S. (1994). The capabilities of market-driven organizations. Journal of Marketing, 58(1),
37-52.
[7]. Eisenhardt, K. M., & Martin, J. A. (2000). Dynamic capabilities: what are they? Strategic
Management Journal, 21(10/11), 1105-1121.
[8]. Eisenhardt, K. M., & Schoonhoven, C. B. (1996). Resource-based view of strategic alliance
formation: strategic and social effects in entrepreneurial firms. Organization Science, 7(2), 136-
150.
[9]. Fahy, J., Hooley, G., Cox, T., Beracs, J., Fonfara, K., & Snoj, B. (2000). The development and
impact of marketing capabilities in central Europe. Journal of International Business Studies,
31(1), 63-81.
[10]. Griffith, D. A., & Harvey, M. G. (2001). A resource perspective of global dynamic
capabilities. Journal of International Business Studies, 32(3), 597-607.
[11]. Grobler, A., & Grubner, A. (2006). An empirical model of the relationships between
manufacturing capabilities. International Journal of Operations & Production Management, 26(5),
458-485.
[12]. Gunasekaran, A. (1999). Agile manufacturing: A framework for research and development.
International Journal of Production Economics, 62(1/2), 87-105.
[13]. Helfat, C. E., & Peteraf, M., A. (2003). The dynamic resource-based view: capability
lifecycles. Strategic Management Journal, 24(10), 997-1010.
[14]. Harris, D., & Helfat, C. E. (1997). Specificity of CEO human capital and compensation.
Strategic Management Journal, 18(11), 895-920.
[15]. Hitt, M. A., Ahlstrom, D., Dacin, M. T., Levitas, E., & Svobodina, L. (2004). The institutional
effects on strategic alliance partner selection in transition economies: China vs. Russia.
Organization Science, 15(2), 173-185.
[16]. Hitt, M. A., Bierman, L., Shimizu, K., & Kochhar, R. (2001). Direct and moderating effects of
human capital on strategy and performance in professional service firms: a resource-based
perspective. Academy of Management Journal, 44(1), 13-28.
[17]. Hitt, M. A., Dacin, M. T., Levitas, E., Arregle, J.-L., & Borza, A. (2000). Partner selection in
emerging and developed market contexts: resource-based and organizational learning perspectives.
Academy of Management Journal, 43(3), 449-467.
145

[18]. Jap, S. D. (1999). Pie-expansion efforts: collaboration processes in buyer-supplier
relationships. Journal of Marketing Research, 36, 461-475.
[19]. Kale, P., Singh, H., & Perlmutter, H. (2000). Learning and protection of proprietary assets in
strategic alliances: building relational capital. Strategic Management Journal, 21, 217-237.
[20]. Kale, P., Dyer, J., & Singh, H. (2001). Value creation and success in strategic alliances:
alliancing skills and the role of alliance structure and systems. European Management Journal,
19(5), 463-471.
[21]. Kogut, B., & Zander, U. (1992). Knowledge of the firm, combinative capabilities, and the
relationship of technology. Organization Science, 3(3), 383-397.
[22]. Lambe, C. J., Spekman, R. E., & Hunt, S. D. (Spring2002). Alliance competence, resources,
and alliance success: conceptualization, measurement, and initial test. Academy of Marketing
Science, 30(2), 141-158.
[23]. Lee, D. Y., & Dawes, P. L. (2005). Guanxi, trust, and long-term orientation in Chinese
business markets. Journal of International Marketing, 13(2), 28-56.
[24]. Li, H., & Zhang, Y. (2007). The role of managers' political networking and functional
experience in new venture performance: evidence from china's transition economy. Strategic
Management Journal, 28(8), 791-804.
[25]. Li, L. L. X. (2000). Manufacturing capability development in a changing business
environment. Industrial Management & Data Systems, 100(6), 261-270.
[26]. Lin, H. (2006). Inter-organizational collaboration, social embeddedness, and value creation: a
theoretical analysis. International Journal of Management, 23(3), 548-558.
[27]. Luo, Y. (2000). Partnering with Chinese firms: lessons for international managers: Ashgate.
[28]. Madhavaram, S., & Hunt, S. D. (2008). The service-dominant logic and a hierarchy of operant
resources: developing masterful operant resources and implications for marketing strategy.
Journal of the Academic Marketing Science, 36(1), 67-82.
[29]. Mahoney, J. T. (1995). The management of resources and the resource and management.
Journal of Business Research, 33(1), 91-101.
[30]. Makadok, R. (2001). Toward a synthesis of the resource-based and dynamic-capability views
of rent creation. Strategic Management Journal, 22(5), 387-401.
[31]. Morgan, R. M., & Hunt, S. D. (1994). The commitment-trust theory of relationship marketing.
Journal of Marketing, 58(3), 20-38.
[32]. Muthusamy, S. K., White, M. A., & Carr, A. (2007). An empirical examination of the role of
social exchanges in alliance performance. Journal of Managerial Issues, 11(1), 53-75.
[33]. Narver, J. C., & Slater, S. F. (1990). The effect of a market orientation on business
profitability. Journal of Marketing, 54, 20-35.
[34]. Narver, John C.; Slater, S. F., & MacLachlan, D. L., (2004) Responsive and Proactive Market
Orientation and New-Product Success. Journal of Product Innovation Management, 21 (5) 334-
347.
[35]. Nasution, H., N.,, & Mavondo, F., T. (2008). Organisational capabilities: antecedents and
implications for customer value. European Journal of Marketing, 42(3/4), 477-501.
[36]. Newbert, S. L. (2007). Empirical research on the resource-based view of the firm: an
assessment and suggestions for future research. Strategic Management Journal, 28(1), 121-146.
[37]. Prieto, I. M., & Revilla, E. (2006). Assessing the impact of learning capability on business
performance: empirical evidence from Spain. Management Learning, 37(4), 499-522.
[38]. Sharifi, H., & Zhang, Z. (1999). A methodology for achieving agility in manufacturing
organizations: An introduction. International Journal of Production Economics, 62(1/2), 7-22.
[39]. Sherehiy, B., Karwowski, W., & Layer, J. K. (2007). A review of enterprise agility: concepts,
frameworks, and attributes. International Journal of Industrial Ergonomics, 37(2), 445-460.
[40]. Slater, S. F., & Narver, J. C. (1994). Does the competitive environment moderate the marketorientation
performance relationship. Journal of marketing, 58, 46-55.
[41]. Sirman, D., G.;, & Hitt, M. A. (2003). Managing resources: linking unique resources,
management, and wealth creation in family firms. Entrepreneurship: Theory & Practice, 27(4),
339-358.
[42]. Tan, J. (2003). Curvilinear relationship between organizational slack and firm performance:
evidence from Chinese state enterprises. European Management Journal, 21(6), 740-749.
146

[43]. Tan, J. & Peng, M.W., (2003) Organizational Slack and Firm Performance during Economic
Transition: Two Studies from an Emerging Economy. Strategic Management Journal, 24 (4)
1249-1263.
[44]. Teece, D. J., Pisano, G., & Shuen, A. (1997). Dynamic capabilities and strategic management.
Strategic Management Journal, 18(7), 509-533.
[45]. Thoumrungroje, A., & Tansuhaj, P. (2004). Globalization effects, co-marketing alliances, and
performance. Journal of American Academy of Business, 5(1/2), 495-502.
[46]. Van Den Bosch, F. A. J., Volberda, H. W., & Boer, M. d. (1999). Coevolution of firm
absorptive capacity and knowledge environment: organizational forms and combinative
capabilities. Organization Science, 10(5), 551-568.
[47]. Vorhies, D. W., & Morgan, N. A. (2005). Benchmarking marketing capabilities for
sustainable competitive advantage. Journal of Marketing, 68, 80-94.
[48]. Voss, K. E., Johnson, J. L., Cullen, J. B., Sakano, T., & Takenouchi, H. (2006). Relational
exchange in US-Japanese marketing strategic alliances. International Marketing Review, 23(6),
610-635.
[49]. Wang, C. L. (2007). guanxi vs. relationship marketing: exploring underlying differences.
Industrial Marketing Management, 36(1), 81-86.
[50]. Wang, C. L., & Ahmed, P. K. (2007). Dynamic capabilities: a review and research agenda.
International Journal of Management Reviews, 9(1), 31-51.
[51]. Wang, Y., Lo, H., & Yang, Y. (2004). The constituents of core competencies and firm
performance: evidence from high-technology firms in China. Journal of Engineering and
Technology Management, 21, 249 - 280.
[52]. Xin, K. R., & Pearce, J. L. (1996). Guanxi: connections as substitutes for formal institutional
support. Academy of Management Journal, 39(6), 1641-1658.
[53]. Yusuf, Y. Y., Sarhadi, M., & Gunasekaran, A. (1999). Agile manufacturing: The drivers,
concepts and attributes. International Journal of Production Economics, 62(1/2), 33-43.
[54]. Zahra, S. A., & George, G. (2002). Absorptive capacity: a review, reconceptualization, and
extension. Academy of Management Review, 27(2), 185-203.
[55]. Zahra, S. A., Sapienza, H. J., & Davidsson, P. (2006). Entrepreneurship and dynamic
capabilities: a review, model and research agenda. Journal of Management Studies, 43(4), 917-
955.
[56]. Zhang, Y., & Zhang, Z. (2006). Guanxi and organizational dynamics in China: a link between
individual and organizational levels. Journal of Business Ethics, 67(5), 375-392.
[57]. Zhang, Z., & Sharifi, H. (2007). Towards theory building in agile manufacturing strategy-a
taxonomical approach. IEEE Transactions on Engineering Management, 54(2), 351-370.
[58]. Zollo, M., Reuer, J. J., & Singh, H. (2002). Inter-organizational routines and performance in
strategic alliances. Organization Science, 13, 701-713.
[59]. Zollo, M., & Winter, S., G. (2002). Deliberate learning and the evolution of dynamic
capabilities. Organization Science, 13(3), 339-351.
147

Early Childhood Education Matters: Functions, Situations, and Perceptions
Beibei Pan 1 *
* Presenter
1. Department of Management, Monash University, Clayton Compus, Wellington Rd,
Clayton 3800, Australia
This paper is aiming at providing an overall picture of former and current views on early
childhood education. The general developmental tendency of this field is that early education
for young children has increasingly received attention of parents, educationists, and scholars
from diverse research backgrounds. After clarifying the educational and socioeconomic
functions, the focus of this paper shifts to digging out the underpinning determinants which
make quality services of preschool programs.
Keywords
early childhood education; development; essential factors
Introduction
During the first decades of the nineteenth century, industrialization brought changes to family
structure as well as to the initial concept of early childhood care and education. Since then, early
childhood education (ECE) has become an international interest (Tietze, Cryer, Bairrão, Palacios, &
Wetzel, 1996). ‘In the international arena, the collocation of a variety of integrated, flexible services
for children has gained momentum’ (Noble, 2007, p. 51, as cited in OECD, 2001). Literally, ECE has
developed as an association which concerns individuals, families, and societies, and also a conjunction
which involves education, economy, and policy. For its impacts on a rich variety of fields, ECE
attracts attentions of scholars from differing research backgrounds.
Significant Influence of Early Childhood Education
The importance of ECE in both individual development and social progress has been clarified by
recent studies. Traditionally, child care centres were established to propel women’s participation in
labour market; later on, the educational functions for the sake of human development has been
attached to ECE as well (Nyland, Nyland, & Maharaj, 2009).
Contemporary scholars have well acknowledged the favourable outcomes of early years’
education. Researchers hold the view that investment in early education should be encouraged for
reasons summarized as follows (Heckman, 2006):
• The formation of both cognitive and socio-emotional abilities begins at early age, which
greatly influences a person’s whole life.
• This is an affair of social justice. Children from less advantaged families need more financial
support to get access to ECE resources.
• Early intervention can effectively guarantee better school outcomes. Correspondingly,
promoted quality of workforce will lead to increased productivity in economy and society at
large.
• Quality child care can reduce crime rate, teenage pregnancy and welfare dependency.
From an educational perspective, early years are a sensitive period which is essentially influential
to overall competency development during a life time according to behavioural research (Knudsen,
Heckman, Cameron, & Shonkoff, 2006, as cited in Shonkoff & Phillips, 2000; Heim & Nemeroff,
2001). An array of studies have demonstrated that child care quality is positively related to the
cognitive development and social competence, with other factors generating variations in early
education effects, such as socioeconomic status or family structure (Peisner-Feinberg & Burchinal,
1997). Peisner-Feinberg and Burchinal’s (1997) research also provides evidence to the positive
relation for children from all backgrounds. Furthermore, they observed that the ECE influence is
stronger on children from less advanced families. And it has been reiterated by Knudsen et al. (2006, p.
10155)as ‘that the quality of the early childhood environment is a strong predictor of adult
148

productivity (As cited in Carneiro & Heckman, 2003) and that early enrichment for disadvantaged
children increases the probability of later economic success (As cited in Currie & Blau, 2005)’. In the
light of those positive research results, now it is widely agreeable that ECE matters in a large range of
aspects of human competency development.
From a socioeconomic perspective, Cunha et al. (2005) referred to relevant literatures and
integrated main findings by presenting the rates of return to educational funding according to critical
stages of life circle in figure 1. It is noticeable that a remarkable return rate appears in the stage of
preschool programs. Along with the growth of age, human capital investment sees a dramatic decline
in the rate of return. This finding turns out to be a persuasive evidence to convince the essential
significance of ECE.
Figure 1. Rates of return to human capital investment initially setting investment to be equal across all
ages
Source: Cunha et al. (2005) Interpreting the evidence on life cycle skill formation.
As observed in those studies above, it is the high economic return of investing in preschool
programs that make early education a crucial issue to the development of human capacities and the
progress of contemporary societies. Responsively, educational requirements extend to early years
spontaneously. The mode of contemporary ECE is seeking to better meet the criteria for quality human
resources.
Taking the points above into account, we can safely draw a conclusion that ECE features a series
of critical characteristics in relation to a complexity of education and socio-economy. This explains
the particular attentions paid by parents, educationists, economists, and policy-makers on ECE.
Overview of Early Childhood Education Development
With all the endeavour of educators and researchers, ECE has been theoretically and practically
well developed during recent decades. ECE systems have undergone regulations in every aspect that
could possibly be improved across the world. The UK launched the Sure Start and other programmes;
the USA initiated the Head Start and family support approaches; great subsidies are commonly offered
by governments to support child care and education among OECD countries (Kamerman, 2000).
Whereas, with a diversity of historical and socio-economic circumstances, child care in different areas
share similarities as well as disparities.
Due to the active interactions between countries in Europe for centuries, early education
approaches share ‘a remarkable degree of similarity among ECE systems in western industrialized
countries’ (Tietze, et al., 1996, p. 449). Take Germany for example. According to the point of Cryer et
al. (2002), Germany takes measures to make ECE services affordable to most families based on a
subsidized sliding scale. Thus cost is not a strong consideration when German parents make their
149

decisions on selection of child care centre. And the educational functions and availability of ECE
institutions are taken into serious consideration (Cryer, et al., 2002). Cyrer et al. (2002) also claimed
that the range of ECE programme quality in Germany is relatively restricted with little variation which
leaves parents not many options as well.
While in the USA, subsidies are offered to disadvantaged families, but considering the high fees
for child care parents normally have very limited choices of both affordable and accessible centrebased
ECE services (Cryer, et al., 2002). The availability of certain types of ECE provisions becomes
the limitation of parents’ choices (Peyton, Jacobs, O'Brien, & Roy, 2001). Another outstanding feature
is the considerable variation existing in both sponsorship and quality of American ECE system. The
arrangements of child care are of great variety to meet differing American families. However, the
provisions of high quality child care, in particular the affordable ones, are scarce resource (Cryer, et al.,
2002).
China’s case is more like the situation in the USA. The types of ECE services range widely with
certain constraints in fact confining parents’ options to few choices. ‘Parental choice is limited in each
country, but in different ways‘ (Cryer, et al., 2002, p. 263).
Overall, progressive development has been made in many aspects across various regions, but
new progresses always deserve continuous efforts to make it better and better.
Access to Early Childhood Education
The aim of investigations relating to this topic is to figure out what factors really matter in ECE
systems to make high quality services for child care and education. The concept of ECE programme
quality varies in accordance with values, beliefs, needs etc. (Cryer, et al., 2002, as cited in Bernhard &
Gonzalez- Mena, 2000; Dhalberg, Moss, & Pence, 1999; Moss, 1994), whereas ‘when quality
definition are closely inspected, the themes of these core elements appear repeatedly, with only the
details differing’ (Cryer, et al., 2002, p. 261).
Various research methods of social science have been adopted to examine each and every aspect
in this field. A range of elements related to quality characteristics have influenced developmental
outcomes to various extent (Burchinal & Cryer, 2003). Features of quality programming are to be
examined ‘including standards for curriculum and assessment, health and safety, class size and
teacher/child ratios, parent involvement, professional development and staff qualifications' (Warash,
Ward, & Rotilie, 2008, p. 646). In brief, four categories consisting of quality, safety, convenience and
availability are main reasons decide the result (Davis & Connelly, 2005). Normally, the demographic
characteristics of family also affect the process of parents decision-making (Peyton, et al., 2001).
Some scholars (Taylor, Dearing, & McCartney, 2004) find that family economic status matters in
children’s development, especially for children from low income families. Through their own survey,
Tietze and Cryer (1999) view availability and affordability as major criteria for quality ECE
programmes. Availability represents provision of services, acceptable distance, convenient schedule,
suitable programmes etc. Affordability is the principle in the light of European original idea to make
ECE services available to parents. (Tietze & Cryer, 1999)
However, the needs of children and parents have received a comprehensive awareness of both
professionals and the public nowadays in the European countries (Tietze & Cryer, 1999). In most
cases, parents take full responsibility to select ECE services for their children and family based on
their own criteria as child care consumers (Cryer & Burchinal, 1997). Consequently, the opinion of
parents, the decision-makers of early education selection, will be valued in the process of assessing the
quality ECE services. Cryer et al. (2002) claimed that factors influencing parents perceptions of ECE
quality, such as affordability, accessibility, traditions related to ECE etc., is not fixed in different
countries. By rating, ranking and conjoint analysis, Rose and Elicker (2008) find that warmth and
educational level of caregivers score very high in mothers’ considerations. Grace and O’cass (2003)
mention that ethics, service delivery, and service failure response are the main reasons when parents
choose and switch child care centres.
Another point made by Cryer and Burchinal (1997) is that child care area has been viewed as a
market obeying the basic principle of supply and demand. Supply is provided by all types of ECE
programmes, as demand, in most cases, is the needs of parents having their children receive quality
care and education. According to market rules, the magnitude of child care supply is inevitably
affected by the preference of parents. They pointed out that parents are not well informed with
150

adequate and accurate information of the services their children have received. They cited economist
James Walker (1991) as ‘The lack of perfect information is the most striking difference between the
child care market and the idealized perfect market…. Consumers do not know the quality of care
offered by providers once identified. Even after a long period of use, consumers will not be fully
informed about the behaviour of the provider’ (p. 65).
Since parent attitudes matter, what are the determinants in their decisions? Noble (2007) set four
sub categories to cover the major influencing elements in parents’ decisions (p. 53):
1. Parent relationship with child
2. Influence of significant others
3. Parent understandings of childhood
4. Maximising the child’s potential
She even emphasises the essential relationship between the needs of family and children (Noble,
2007). Moreover, ‘ As evidenced from the pilot study, adults’ needs (such as hours of operation, cost,
and location) were rated as influential parents’ child care choices’ (Seo, 2003, p. 648). Interestingly,
survey result indicates that parents consider convenience factors (cost, location etc.) more than quality
indicators (safety, pedagogy, interaction etc.) when weighing determinants of ECE services (Cryer &
Burchinal, 1997, as cited in Kisker & Maynard, 1991). It shows that parents place their own needs
prior to children’s which may lead to a hasty decision (Gable & Cole, 2000). Nevertheless, Gable and
Cole (2000) give reasonable explanations to those phenomena: if the service is too expensive to afford,
too far to deliver their child, or difficult to cooperate in time schedule, the decision is actually
unrealistic. ‘In some ways, the priorities of parents, the state and experts in the field are not the same’
("Parents pleased with child care options and quality," 2008, p. 1). It is uncertain that to what extent
the quality factors in parental values match those of professionals (Peyton, et al., 2001).
In summary, up until now, the importance of ECE has been acknowledged all over the world.
Supportive forces come from all fields of society to attribute to this career. ECE has become an
integrated concept which involves complex factors concerning the prospect of entire human
community. Accompanying with accelerated development of socio-economy and human capacity,
ECE is looking forward to contribution of everybody concerned to meet the demands of both societies
and individuals.
References:
Burchinal, M. R., & Cryer, D. (2003). Diversity, child care quality, and developmental outcomes.
Early Childhood Research Quarterly, 18(4), 401-426.
Cryer, D., & Burchinal, M. (1997). Parents as child care consumers. Early Childhood Research
Quarterly, 12(1), 35-58.
Cryer, D., Tietze, W., & Wessels, H. (2002). Parents' perceptions of their children's child care: a crossnational
comparison. Early Childhood Research Quarterly, 17(2), 259-277.
Cunha, F., Heckman, J. J., Lochner, L., & Masterov, D. V. (2005). Interpreting the evidence on life
cycle skill formation. In E. Hanushek & F. Welch (Eds.), Handbook of the Economics of
Education. North Holland.
Davis, E. E., & Connelly, R. (2005). The influence of local price and availability on parents' choice of
child care. Population Research and Policy Review, 24, 301-334.
Gable, S., & Cole, K. (2000). Parents' child care arrangements and their ecological correlates. Early
Education & Development, 11(5), 549-572.
Grace, D., & O'cass, A. (2003). Child care services: an Exploratory study of choice, switching and
search behaviour. European Journal of Marketing, 37(1/2), 107-132.
Heckman, J. J. (2006). The economics of investing in early childhood. Paper presented at the 2006
NIFTeY Conference - Prevention: Invest Now or Pay Later, Sydney, Australia.
Kamerman, S. B. (2000). Early childhood education and care: an overview of developments in the
OECD countries. International Journal of Educational Research, 33(1), 7-29.
Knudsen, E. I., Heckman, J. J., Cameron, J. L., & Shonkoff, J. P. (2006). Economic, neurobiological,
and behavioral perspectives on building America's future workforce. Paper presented at the
The National Academy of Science of the United States of America, Washington.
151

Noble, K. (2007). Parent choice of early childhood education and care services. Australian Journal of
Early Childhood, 32(2), 51-57.
Nyland, B., Nyland, C., & Maharaj, E. A. (2009). Early childhood education and care in urban China:
the importance of parental choice. Early Child Development and Care, 179(4), 517-528.
Parents pleased with child care options and quality. (2008) & A. Dickman (Vol. Ed.): Vol. 96. Survey
results suggest contrasting child care priorities (pp. 1-11). Wisconsin Public Policy Forum.
Peisner-Feinberg, E. S., & Burchinal, M. R. (1997). Relations between preschool children's child-care
experiences and concurrent development: The cost, quality, and outcomes study. Merrill -
Palmer Quarterly, 43(3), 451-477.
Peyton, V., Jacobs, A., O'Brien, M., & Roy, C. (2001). Reasons for chosing child care: associations
with family factors, quality, and satisfaction. Early Childhood Research Quarterly, 16, 191-
208.
Rose, K. K., & Elicker, J. (2008). Parental decision making about child care. Journal of Family Issues,
29(9), 1161-1184.
Seo, S. (2003). Early child care choices: a theoretical model and research implications. Early Child
Development and Care, 173(6), 637-650.
Taylor, B. A., Dearing, E., & McCartney, K. (2004). Incomes and outcomes in early childhood. The
Journal of Human Resources, 39(4), 980-1007.
Tietze, W., & Cryer, D. (1999). Current trends in European early child care and education. The Annals
Of The American Academy Of Political And Social Science, 563(1), 175-193.
Tietze, W., Cryer, D., Bairrão, J., Palacios, J., & Wetzel, G. (1996). Comparisons of observed process
quality in early child care and education programs in five countries. Early Childhood
Research Quarterly, 11(4), 447-475.
Walker, J. (1991). Public policy and the supply of child care services. New York: Russel Sage
Foundation.
Warash, B. G., Ward, C., & Rotilie, S. (2008). An exploratory study of the application of early
childhood environment rating scale critera. Education, 128(4), 645-658.
152

Five Dimensions of Entrepreneurship: A Study of First and Second
Generations Chinese Entrepreneurs in Melbourne
Henri Lee 1 *
* Presenter
1. Faculty of Business and Enterprise, Swinburne University of Technology
The primary objective is to investigate the success achieved by first and second
generation Chinese entrepreneurs so as to improve understanding of their impact and
contribution in the development of Melbourne. Ultimately, this study seeks to provide an
insight into their entrepreneurial activities and their social and economic contributions to
themselves and the nation, leading to Chinese being recognized as an important source of
economic growth. Five dimensions of Chinese entrepreneurs from the first and second
generations were identified for investigation and exploration. These are their background,
their motivations for business ownership, traditional Confucian values, difficulties and
hurdles, and their overseas expansionary vision. Ultimately, the study concentrates on their
successes.
153

An Exploration of Country of Origin Effect on Union Policies in Chinese
Multinational Enterprises
Shuqin Zhu 1 *
* Presenter
1. Department of Management, Monash University, Clayton Compus, Wellington Rd,
Clayton 3800, Australia
There are an increasing number of studies investigating country of origin effect on the
management practices of multinational enterprises (MNEs). On the basis of the research on
multinational corporations from developed countries, these studies argue that country of
origin effect has a significant impact on MNEs’ operation in other countries. However, few of
them have investigated country of origin effect on the management practices of MNEs from
developing country. Using qualitative methodology, this research aims to explore whether
and how country of origin effect exemplifies itself in the union policies of Chinese
multinational enterprises. This paper first reviews the relevant literature on country of origin
effect. Then, it proceeds to examine the studies on the union policies of multinational
corporations. Following this, the paper discusses the industrial relations system in China.
Finally, the paper proposed some questions for future research.
154

Authors Title Affiliation
Wei Tao, Bao
Hongchun, Dru
Morrish, & Min Gu
Gang Li, Ranjeet
Singh, Dan Li, Chunxia
Zhao, Liying Liu, &
Paul A. Webley
Shan Liu, Charles Ma,
Lina Wng, Ross
Coppel, & Gareth M.
Forde
Xin Gao & Philip J.
Marriott
Junfei Tian, Xu Li, &
Wei Shen
Supercontinuum generation for fiber-optic nonlinear
microscopy
Synthesis of biomorphic zeolite honeycomb
monoliths with 16000 cells per square inch
Synthesis and delivery of Malaria DNA vaccines via
biodegradable polymer carrier systems
Gas chromatographic retention indices and mass
spectra of trimethylsilyl derivatives of flavonoids on
capillary columns with non-polar stationary phase
Microfluidic systems on paper
Swinburne
University of
Technology
Monash
University
Monash
University
RMIT
University
Monash
University
Yan-e Gao, TG Johns,
E Dimitriadis, E
Menkhorst, Joanne
Mockler, BRG
Williams, & S Tong
Ying Lu, Ramanie
Samarathunge, &
Charmine Hartel
Dan Wang, Susan
Freeman & Kate
Hutchings
Mingqiong Zhang,
Chris Nyland, &
Cherrie J. Zhu
Liang B. Y. Du &
Arthur J. Lowery
Liang B. Y. Du &
Arthur J. Lowery
Liang B. Y. Du &
Arthur J. Lowery
Haidong Zheng, Abu
Sadek, David Yao,
Michael Breedon, &
Kourosh Kalantarzadeh
MicroRNA expression changes detected by
microarray with forskolin syncytialisation of BeWo
cells
The Impact of Acculturation on Professional Chinese
Immigrants in the Australian Workplace
Outward Foreign Direct Investment from China
The Institution of Mingongzhi in Contemporary
China and the Strategies of MNEs: An Institutional
Analysis
Fiber Nonlinearity Precompensation for Long-hual
links using Direct-Detection Optical OFDM
Improving Nonlinear Precompensation in Direct-
Detection Optical OFDM Communications Systems
Improved Nonlinearity precompensation for Longhual
High-date-rate Transmission using Coherent
Optical OFDM
Rapid Formation of Thick and Transparent Anodic
TiO 2 Nanotubular Films from Sputtered Ti
Monash
Institute of
Medical
Research
Monash
University
Monash
University
Monash
University
Monash
University
Monash
University
Monash
University
RMIT
University
156

Supercontinuum generation for fiber-optic nonlinear microscopy
Wei Tao 1 , Bao Hongchun 1 , Dru Morrish 1 , & Min Gu 1
1. Centre for Micro-Photonics, Swinburne University of Technology
We report the experimental investigation of the transmission characteristics of a
supercontinuum through double-clad photonic crystal fiber for fiber-optic nonlinear
microscopy.
157

Synthesis of biomorphic zeolite honeycomb monoliths with 16000 cells per
square inch
Gang Li, 1,2 Ranjeet Singh, 1,2 Dan Li, 1 Chunxia Zhao, 1,3 Liying Liu 1,2,4 and Paul A. Webley 1,2
1. Department of Chemical Engineering, Monash University, Wellington Road, Clayton, VIC
3800, Australia
2. Cooperative Research Centre for Greenhouse Gas Technologies, Grnd Flr NFF House, 14-
16 Brisbane Ave, Barton, ACT 2600, Australia
3. State Key Laboratory of Advanced Technology for Materials Synthesis and Processing,
School of Materials Science and Engineering, Wuhan University of Technology, Wuhan
430070, PR China
4. School of Materials & Metallurgy, Northeastern University, Shenyang 110004, PR China
Zeolite NaX and silicalite honeycomb monoliths were synthesized via hydrothermal
growth of zeolites on the interior surface of cuttlefish bone by a novel flow coating technique.
The resultant samples were characterized by SEM, XRD, EDX, N2 adsorption and TGA. The
loading of nano-crystalline silicalite (36.4 wt%) was higher than for NaX type zeolite. The
biomorphic honeycomb monoliths were found to have an exceptionally high cell density of
16000 cells per square inch which is ten times higher than the best synthetic ones. The
honeycombs were evaluated for their hydrodynamic and kinetic properties, showing
substantial improvement of mass transfer rate with lower pressure drops than conventional
zeolite packing.
158

Gas chromatographic retention indices and mass spectra of trimethylsilyl
derivatives of flavonoids on capillary columns with non-polar stationary
phases
Xin Gao 1 2 & Philip J. Marriott 2
1. Health Science Centre, Peking University
2. School of Applied Sciences, RMIT University
Flavonoids are one of the most important groups of compounds occurring in plants,
which exhibit a wide range of biological effects and act as natural antioxidants. They
comprise the main class of polyphenols in diet and Chinese herbs, and usually exist in the
presence of complex matrix, which considerably hinders the identification of these analytes.
Chemical analysis of flavonoids in samples allows for their precise compositional assignment,
and a number of advanced methods exist for this task. Analysis by using gas chromatography
(GC) and comprehensive two-dimensional gas chromatography (GC×GC) may be regarded as
very promising methods for studying flavonoids in diet and herbs due to the ability to speciate
different types of flavonoids. In the present work, it was of interest to analyse a large database
of flavonoids by GC and GC×GC in order to present the GC and GC×GC characteristics of 29
compounds. These data include basic retention index information on high-performance
capillary columns with non-polar stationary phases. A number of previously unreported
flavonoids have been included in this study, and so this report presents for the first time mass
spectral and retention data for a series of trimethylsilyl derivatives of those flavonoids.
159

The Impact of Acculturation on Professional Chinese Immigrants in the
Australian Workplace
Ying Lu 1 , Ramanie Samarathunge 1 , & Charmine Hartel 1
1. Department of Management, Building 11, Monash University, Clayton Campus,
Wellington Road Clayton Victoria 3800 AUSTRALIA
Professional Chinese immigrants (PCIs) encounter many adjustment difficulties arising
from cultural and social differences after entering into the Australian workplace (Birrell &
Healy, 2008; Ho, 2006). Evidence shows that Australian employers do not pay adequate
attention to the integration of PCIs, which has already caused adjustment problems in the new
homeland and has negatively impacted on their career development and success (Birrell &
Healy, 2008). Given the important role of PCIs in Australia with regards to creating a diverse
workforce, boosting economic growth and narrowing the gap in the skills shortage of the
country, their integration process into the Australian workplace deserves due consideration.
The research looks at this issue from acculturation, which is an inevitable process that every
immigrant has to experience. To sum up, the aim of this research is to examine the individual
and workgroup contextual factors which may affect PCIs’ acculturation strategy as well as the
effect of various acculturation strategies on selected individual work outcomes in the
workgroup. In particular, it explores the possible mediating function of acculturation strategy
in a workplace context.
160

Outward Foreign Direct Investment from China
Dan Wang 1 , Susan Freeman 1 & Kate Hutchings 1
1. Department of Management, Building N , 27 Sir John Monash Drive, Caulfield East
Victoria 3145, AUSTRALIA
This poster examines the unique motivations and strategies for outward foreign direct
investment (OFDI) undertaken by Chinese multinationals (MNEs), and addresses the
challenges associated with these unique motivations and strategies. We find that the activities
of Chinese MNEs are strongly influenced by the institutional environment of China, and that
despite the unique resources Chinese MNEs possess, institutional theory rather than the
resource based view is best placed as a theoretical framework when examining Chinese OFDI.
161

The Institution of Mingongzhi in Contemporary China and the Strategies of
MNEs: An Institutional Analysis
Mingqiong Zhang 1 , Chris Nyland 1 , & Cherrie J. Zhu 1
1. Department of Management, Building 11, Monash University, Clayton Campus,
Wellington Road Clayton Victoria 3800 AUSTRALIA
From the early 1980s, China has entered into an age of rural-urban migration. Millions of
farmers have flooded into cities and towns for work and significantly reshaped the character
of Chinese society. Confronted with the flood of rural-urban migrants, the central government,
city governments, private and public companies and urban residents have responded on the
various bases. Consequently, a combination of linked rules, values, norms, and patterned
practices that structure the way that rural migrants should be treated in urban areas has
emerged and rural migrants are being socially excluded. From the perspective of
neoinstitutionalism, these linked hukou-based regulative, normative and cognitive elements
demonstrate that a new institution has emerged, which is termed in this study the institution of
mingongzhi. Like the terms slavery and apartheid, the institution of mingongzhi refers to a
social-economic system under which rural migrant workers are socially excluded in urban
areas of contemporary China based on their household registration (hukou) status. As a major
institution of the labour market in contemporary China, mingongzhi has had a profound
influence on China’s urban employment environment, industrial relations (IR) and human
resource management (HRM) practices. It is an ideal variable to conduct analysis of the
strategic response of MNEs to the institutional characteristics of a host country. This is the
more so as international business scholars have accorded little systematic attention to the
relationship between institutions and multinational enterprises (MNEs).
162

Fiber Nonlinearity Precompensation for Long-hual links using Direct-
Detection Optical OFDM
Liang B. Y. Du 1 * and Arthur J. Lowery 1
1. Electrical and Computer Systems Engineering, Monash University, Clayton, VIC 3800,
Australia
The use of nonlinearity precompensation in direct-detection optical orthogonal frequency
division multiplexed links is investigated by simulation. Because of the presence of a strong
optical carrier its performance is poorer than for coherent systems: with compensation the
signal quality is found to vary almost periodically across the signal band. We propose and
explain the operation of two optical, one electrical and one computational method of
removing this periodic variation. Optical filtering of one sideband at the receiver is most
effective, but a substantial improvement can be obtained by a simple modification to the
precompensation algorithm.
163

Improving Nonlinear Precompensation in Direct-Detection Optical OFDM
Communications Systems
Liang B. Y. Du 1 * and Arthur J. Lowery 1
1. Electrical and Computer Systems Engineering, Monash University, Clayton, VIC 3800,
Australia
Carrier boosting at the receiver enables direct detection optical OFDM (DDO-OFDM) to
outperform coherent OOFDM in the nonlinear limit. Boosting also improves the effectiveness
of nonlinearity precompensation substantially.
164

Improved Nonlinearity precompensation for Long-hual High-date-rate
Transmission using Coherent Optical OFDM
Liang B. Y. Du 1 * and Arthur J. Lowery 1
1. Electrical and Computer Systems Engineering, Monash University, Clayton, VIC 3800,
Australia
We show that the performance of precompensation of fiber nonlinearity in coherent
optical OFDM systems operating at up to 60 Gbps/ polarization can be improved by electrical
filtering the precompensation signal. The optimal filter bandwidth is related to the FWM
efficiency spectrum when dispersion is considered.
165

Rapid Formation of Thick and Transparent Anodic TiO2 Nanotubular
Films from Sputtered Ti
Haidong Zheng 1 , Abu Sadek 1 , David Yao 1 , Michael Breedon 1 , & Kourosh Kalantarzadeh
1
1. School of electrical and computer engineering, RMIT University.
This poster reports the new insights found in fast anodization of sputtered Ti .
166