Master of Science

This thesis is a study of how information theory can be used to analyse and interpret quantum mechanical systems, specifically, systems that are being used for information processing and communication tasks. Quantum information science is a relatively new field, and it has only been recently that we have started to look at using quantum mechanical systems to accomplish information processing tasks that are not possible using classical systems. As this method of information processing is a relatively new paradigm, there are many possibilities for new algorithms and discoveries in this area of computing. The application of information theory towards the analysis and description of quantum mechanical systems yields new interesting results - it can even be argued that information theory provides a natural and intuitive way to intepret quantum mechanics, as correlations and measurement outcomes in quantum processes can be described using information theory. Chapter 1 is an introduction to the strange new world of quantum mechanics and takes a brief look at one of the applications of quantum information science in quantum cryptography. A mathematical overview of the postulates and framework that quantum mechanics is built on is presented in Chapter 2. Chapter 3 is a brief study of classical information theory and some of the fundamental concepts and ideas that form the crux of the subject. In Chapter 4, we present the main results of the thesis by taking a look at an optimal measurement scheme that improves on previous schemes by allowing us to extract the maximal amount of information from states in a quantum channel of specific geometry. The findings have an impact on various other areas of the field, including security analysis of quantum cryptographic protocols and building better quantum cloning machines.

The overall aim of this thesis is to develop liquid chromatographic techniques for analyzing different classes of phytohormones in tropical plants. Today high performance liquid chromatography (HPLC) and liquid chromatography-mass spectrometry (LC-MS) are the key techniques in the field of biochemical analysis of complex mixtures due to the many advantages offered by these two methods. HPLC and LC-MS methods have been developed for the separation and identification of different classes of phytohormones in tropical plant materials, i.e. coconut water, sugar cane juice, and watermelon juice, etc. Such tropical fruit juices have been widely used in the tissue culture industry due to these media being rich sources of crude phytohormones.

Chapter 1 provides an introduction to HPLC and LC-MS techniques which are the most powerful techniques for the separation and analysis of chemical compounds from small molecules/ions to bio-macromolecules. Many articles and reviews published using HPLC and LC-MS over the last decade illustrating that these two techniques are now the most useful and widely applied tools in biochemical separation and analyses. Sample isolation and preparation, key to biochemical analyses, have been briefly introduced in this chapter.

Chapter 2 covers the theoretical aspects of HPLC and LC-MS. Several important parameters affecting separation, including capacity, efficiency, selectivity, and resolution, have been discussed. A variety of commercially available detectors for HPLC, several operation modes of HPLC, and the column used in HPLC, are also described. At the same time, some introductions of rapid resolution liquid chromatography (RRLC) are given in this chapter. Five distinct parts of MS, including the sample inlet, the ion source, the mass analyzer, the detector, and the data recorder are briefly introduced in this chapter. The introduction and application of a combination of liquid chromatography with mass spectrometry are also presented.

Chapter 3, the general characteristics of different classes of phytohormones are introduced, with a description about their chemical properties and their biological functions in plant growth and development. Brief introductions of different isolation and sample preparation methods, analytical techniques for phytohormones are presented. The discussion is not fully comprehensive but is intended to provide a brief description of those aspects of each technique.

Chapter 4, 5 and 6 are the results and discussion of this dissertation. These chapters describe the applications of HPLC and LC-MS for the analyses of different phytohormones and indole compounds in different tropical plant materials, such as coconut water, sugar cane juice and watermelon juice. The optimum HPLC separation and LC-MS identification conditions have been used for the analyses of different phytohormones and indole compounds in real samples. Some key findings of the phytohormones are as follows: indole-3-acetic acid (IAA, one of the most important auxins) and abscisic acid (ABA) have been detected in coconut water; IAA, 3-acetylindole, Indole-3-pyruvic acid (IPA) and indole-2-carboxylic acid have been found in sugar cane juice; and IAA and indole-3-butyric acid (IBA) have been found in the watermelon juice. Solid-phase extraction (SPE) with C18 columns has been used as an efficient method to pre-concentrate of different phytohormones and indole compounds. In addition, a new SPE method was developed for the isolation and pre-concentration of indole compounds in sugar cane juice successfully.

Chapter 7 presents a general conclusion for the successful application of HPLC and LC-MS for analyzing different classes of phtohormones and indole compounds in tropical plant materials. Further perspectives are also given in this chapter.

This study examined the effects of 3 different types of post-activation potentiation (PAP) protocols [dynamic back squats (DS), maximum voluntary isometric knee extension (IKE) and maximum voluntary isometric back squat (IS)] on subsequent 10-, 20- and 30-metre sprint performance in 12 well-trained track athletes. Before the 30-metre sprint assessment, the subjects performed one of the four randomized PAP protocol; Control (4 minutes of passive rest), DS protocol (3 repetitions of back squats at 90% 1 repetition maximum), IKE protocol (3 repetitions of 3-seconds isometric knee extension) and IS protocol (3 A further 4-minute recovery period separated the PAP protocols and three trials of 30-metre sprint assessment.

A general linear model with repeated measures ANOVA was used to separately detect differences in 10-, 20-, and 30-metre sprint performance among the PAP protocols. Pairwise comparisons were applied post hoc where appropriate. The primary results revealed that there were no significant differences in sprint performance among the control, IKE, IS and DS protocols, regardless of the 1st, 2nd, 3rd, average or best trials. Since the best 30-metre sprint time approached significance (F = 2.57; p = .099), these data were further looked into to examine individual trends.

The follow-up investigation revealed two “Non-responders” and 10 “Responders” to the PAP intervention. Subjects were categorised as Responders” when their best 30-metre sprint time improved in response to at least one of the PAP protocols. For the two “Non-responders”, the sprint performance did not improve after any PAP protocols. The sprint times of the 10 “Responders” showed a significant difference among the PAP protocols in 20-metre sprint time (F = 3.97, p = .043) and 30-metre sprint time (F = 5.30, p=.016). Post hoc analysis reported that only the 30-metre sprint time after the IS protocol (3.98 ± 0.08 seconds) was also significantly faster than the DS protocol (4.03 ± 0.06; 1.26%; p = .004).

The main finding of this study reported that the PAP protocols did not enhance subsequent sprint performance. However, the 30-metre sprint times after the IS protocol improved after the IS protocol, compared to after the DS protocol. This was only observed in subjects that responded to the IS protocol. Further investigation is required to warrant the utilisation of PAP protocols to augment subsequent acute sprint performance.

Artificial neural networks (ANN), especially the backpropagation neural networks (BPNN) have recently attracted the interests of many researchers in the field of chemistry. The advantages of BPNN over classical methods have been proven, but BPNN methods still have many aspects worth to be studied such as to improve the performance of the algorithm in order to overcome some of its limitation.

The objectives of this work are to improve the neural network methods, to apply the BPNN in some selected chemistry areas, and to compare the neural network method with multivariate linear regression (MLR) method.

In-house developed programs were part of this study. The programs were developed using C language. Moreover, all of the programs were written and run on a Pentium III 450 PC with 64 Mbytes memory under a Window 98 operating system. They were proven to be successful.

BPNN have been used to predict six kinds of physical properties, namely heat capacity, boiling point, density, refractive index, Gibbs free energy and enthalpy of alkanes. The molecular distance-edge vector λ was used as input for the BPNN. It was noted that the BPNN prediction results for 25 alkanes are indeed better than the MLR results.

The BPNN method was also used to predict the 13C NMR chemical shifts for alkanes. From the experimental results obtained by the BPNN and MLR methods, it was noted that the BPNN calculation provides better prediction results than the MLR. The results indicated that there is a nonlinear dependence between the descriptors and the intended property, ¹³C chemical shift. Undoubtedly the use of BPNN is justified.

The performance of the BP algorithm is affected by the presence of outliers in the experimental data. In order to solve this problem a robust BPNN algorithm has also been developed. The developed robust BPNN demonstrated better predicted results even when the training data set contains 25% of outliers.

This thesis is organized into seven chapters :

Chapter 1 outlines what was done in this thesis and how it was implemented.

Chapter 2 introduces the typical architecture, general features and characteristics of multilayer neural nets (MLNNs) studied in this research.

Chapter 3 focuses on the most popular learning algorithm, the standard Backpropagation (BP) method. The standard steepest descent algorithm is studied in detail and a fast adaptive steepest descent (FASD) algorithm, is proposed. The objective of the FASD is to maximize the stepsize every step, for the steepest descent algorithm. Also, the standard conjugate gradient algorithm and its extension to nonquadratic problems are studied while the Scaled Conjugate Gradient Algorithm (SCG) is studied in detail and a Modified Scale Conjugate Gradient Algorithm (MSCG) is proposed and implemented. The MSCG is an effective and feasible method to implement the principle of the standard conjugate gradient to nonquadratic problems and it can make the learning fast and has less calculation complexity and memory usage involved in training the neural networks.

Chapter 4 presents a heterogeneous model of the ANNs, an alternative method to train the multilayer neural nets faster which has different activation functions among different hidden neurons.

Chapter 5 discusses a global minimum algorithms, i.e. the simulated annealing algorithm in detail.

In Chapter 6, a novel application of artificial intelligence, forecasting force-field parameters and lattice energies using neural networks, is discussed and it provides a complementary alternative to traditional methods and demonstrates the innate strength of neural networks and their potential as forecasting techniques in solid state physics and condensed matter physics.

Chapter 7 summarizes the student's research work and recommends a Multiple Conjugate Gradient Algorithm.

The established CAL3QHC model for air pollution modeling of carbon monoxide (CO) and particulate matter (PM) is studied to examine its applicability for use under Singapore envirom~ental settings. To date, the CAL3QHC model is the best one to predict air pollutant CO concentration and is able to predict PM concentration in the atmosphere at road intersections under North American settings. However, the model has not been applied to predict CO and PM concentrations in a metropolitan country that is near the equator such as Singapore, which has predominantly unstable tropical climatic conditions, with high humidity level, high rainfall, warm weather and variable wind directions throughout the year.

In our study, we tested the model's suitability for used in four separate road intersections in Singapore. The first location was a "T-shape" road intersection. The second location was a cross road intersection that was close to an industrial area. The third location was a complex road intersection where an expressway, several main and minor roads converged. The last location was also a complex intersection that included a circus and fly-over road.

In order to improve the predicted values, certain assumptions and modifications of input parameters for the model were made. The air pollutant concentrations at a "Tshape" intersection were calculated by using a standard cross road intersection to model the "T-shape" intersection where the input parameters for the traffic flux of the imaginary fourth road leading to this "T-shape" intersection were set to relatively low values. The mathematical axiom of superposition was a road intersection that included a fly-over road. The application of the mathematical axiom allowed roads at the intersection that were not on the same plane to be treated as if they were on one imaginary plane for the purpose of pollutant concentration calculations.

After the appropriate revisions, the Cal3QHC model was able to predict the CO concentrations as the four locations satisfactorily but the calculated PM concentrations at these locations suffered from gross error. Generally, the predicted concentrations for CO from the revised model deviated on an average of *l 5% from the monitored values.

The time spent and suite of behaviours associated with water uptake and circulation of three sympatric species of fiddler crabs (Austruca annulipes, A. perplexa and Gelasimus vocans) on Pulau Hantu Besar were determined. Since G. vocans individuals live nearer the water edge, they allocated significantly less time to water uptake and circulation. There were also slight, non-significant differences in the suite of behaviours displayed by each species of crabs, revealing the relative efficacies of each behaviour in relation to moisture content of the substrate. Results suggest that ‘dip’, the predominant behaviour, was associated with drier substrates and ‘sit’ was observed more commonly on wet substrates. Increase in feeding duration consistently increased water uptake activity across species and significantly for A. perplexa individuals, validating ‘flotation process’ as a possible feeding mechanism. Elevation in substrate temperature did not significantly increase water uptake activity or time spent underground, suggesting that other factors could also influence water uptake and circulation from the surface of substrate in the field. The presence of major chela appeared to hinder water uptake in A. perplexa individuals, but not significantly so. This is likely a small cost relative to the importance of this enlarged structure in increasing reproductive fitness. The removal of setal tufts in laboratory experiments showed that these setae could be more useful in water uptake only for G. vocans individuals and smaller individuals of A. annulipes. These findings suggest that species dwelling on moist substrates (e.g., G. vocans) depend more on capillary water uptake through the setae while those living on drier substrates (e.g., A. annulipes and A. perplexa) could favour the use of pressure difference between the inside of the branchial chambers and external environment, warranting further work in this aspect Examination of setal morphology revealed substantial differences in the area of setal coverage and types of setae in individuals of G. vocans compared to those of the other two species, indicating that some morphological adaptation to their respective microhabitats. New types of setae were also discovered, which could possibly be unique to the populations at the study site. The behavioural and morphological aspects of this study corroborated to suggest that capillary water uptake could be a derived trait. This is the first study to characterise the water-uptake and circulatory behaviours and uncover the functional morphology and importance of setal tufts in water uptake among fiddler crabs. These results point to the possibility of behavioural and morphological adaptations of each species of fiddler crabs for water uptake in relation to the moisture content in their microhabitats and shed light on their success in occupying the semi-terrestrial intertidal zone, where large fluctuations in microhabitat conditions often pose as a challenge for many other organisms. The scope of further work was also described to expand the study in greater relevance to current environmental issues.

In this manuscript, we study the Dzyaloshinkii-Moriya(DM) Anisotropic XX- spin chain model in the presence of an external homogeneous magnetic field applied to one site in a bipartite system. We found that the Berry phase of the system vary interestingly with small and large amount of DM interaction and the magnetic field. The results have potential implementation in quantum logic gate control. Next, we calculate the Berry phase of the thermal states and verify that the results are consistent with that of pure states. In addition, we also considered the concurrence (i.e. the amount of entanglement) of the system, and the relationship between concurrence and the Berry phase.

Although previous studies have addressed the issue of the health of overweight adolescents, few researchers have comprehensively examined the body composition and physiological profile of this population. The purpose of this study was to determine the body composition and selected physiological characteristics of overweight adolescent Singapore Chinese males and to examine the relationship of these parameters with that of an age, gender and race matched non-overweight adolescent group. This study addressed the following specific questions : what are the differences in the results of measurement of body composition using skinfolds and bioelectric impedance; and what are the differences (if any) between overweight and non-overweight male adolescent in body composition, resting blood pressure, heart rate, blood chemistry, pulmonary function variables and functional aerobic capacity?

Forty boys of ethnic Chinese background and whose age ranged from 15 to 16 years from 10 different schools throughout Singapore were assessed in this study. One-half (n=20) of the subject met the criteria of being overweight and the other (n=20) as non-overweight, based on the Singapore standard height for weight charts. Body composition was evaluated using height, weight, sum of skinfolds, circumference of girths, bone widths and bioelectrical impedance. Blood chemistry was determined using the dry chemistry desktop analyzer for total cholesterol, low and high density lipoprotein and triglycerides. A spirometer system was used to measure pulmonary function variables. A graded exercise test was administered to obtain maximal oxygen uptake. Resting and exercise heart rate was determined using a heart rate monitor.

The overweight subjects had significantly higher (p < 0.001) Body Mass Index (BMI), waist/hip ratios, and in all 8 skinfolds, 10 circumferential girths and 2 skeletal widths measurements compared to the non-overweight subjects. There was a significant difference (p < 0.001) in percent body fat between the two groups using four formulae for skinfolds (eight sites) and two formulae for bioelectric impedance. Results showed that body fat was higher in all overweight than non-overweight subjects for all techniques with the Slaughter (1988) 2-site skinfold formulae and Houtkooper (1992) bioelectric impedance equation providing the most realistic data (that is, overweight subjects had predicted body fat greater than 25%.) Other skinfold and bioelectric impedance formulae tended to under-estimate percent body fat in the overweight group (that is, body fat was predicted to be less than 21%). The overweight subjects also showed significantly higher (p < 0.001) systolic blood pressure, triglycerides and very low density lipoproteins but significantly lower (p < 0.001) relative VO₂ max compared to the non-overweight subjects. There were no significant differences in the pulmonary function variables of the two groups. These data provide vital information to the practitioner in planning programmes and identifying precisely where overweight subjects differ physiologically from their non-overweight counterparts.

Timetabling is a problem encountered by many educational institutions from the primary school right up to the universities. It is a particular form of scheduling problem.

The work presented in this thesis concerns the problem of timetabling in secondary schools in Singapore, in particular Yusof Ishak Secondary School (YISS). The main focus of this thesis is to look at the timetabling problem based on actual data and actual requirements of a situation. We examine the various techniques that have been applied to solve timetabling problems. We also devise and examine plausible strategies that might be used to solve the YISS timetabling problem.

From the analysis, we discover that the theoretical techniques and plausible strategies could not be applied directly to the YISS timetabling problem in particular, as the problem in study was very complex. We can also conclude that a fully automated timetabling solution that works in real world timetabling problems is unlikely to exist. Hence a practical computational approach was proposed to address and solve the problem. A detailed study was done to investigate the process of solving a unique real life problem using a practical approach. Techniques to enhance and fine-tune the initial feasible solution were adopted in order to obtain the highly desired feasible solution.

Finally, suggestions are made as to how the study should be extended to other secondary school timetabling problems so that the solution technique can work for any secondary school with minor modifications and not just for one school in particular.

Nasopharyngeal carcinoma (NPC) is a malignant tumor that occurs is prevalent among Chinese in southern China, and in places wherever these Chinese migrate to. Studies on factors affecting the prognosis of NPC have revealed that early diagnosis and treatment are very important in disease management. One of the factors associated with the etiology of NPC is the Epstein-Barr virus (EBV), and antibodies towards this viral antigen have commonly been used in the diagnosis of this disease.

Various antigen complexes of EBV which are useful as serological markers in the diagnosis of NPC are broadly categorized as viral capsid antigen (VCA), diffuse early antigen (EA-D), restricted early antigen (EA-R), membrane antigen (MA) and Epstein-Barr nuclear antigen (EBNA). Each of these antigens has several polypeptides. For the diagnosis of NPC using the whole antigen complex extracted from the cell lines, immunofluorescence assay (IFA) has been shown to be the most sensitive and specific assay compared to other techniques. Enzyme-linked immunosorbent assay (ELISA) is less sensitive but has the advantage of being more suitable for screening populations at high risk for NPC. This test is easily automated, quick to perform and does not involve much skill. The sensitivity and specificity of using ELISA for the detection of NPC can be improved by using gene cloning technology to obtain pure polypeptides from the EBV antigen complex. Polypeptides are used to coat microtiter plates for ELISA.

The cDNA inserts, which were confirmed to have high potential for NPC diagnosis by immunoblotting, were recloned in the plasmid expression vector, pMAL-C2, and over-expressed in E. coli. The fusion protein was easily purified by affinity chromatography on an amylose column. However, out of the eight cDNA inserts studied, only the cDNA inserts from clones R29-3, R29-2, B19-4 and P24-1 were successfully expressed in large quantities. Large-scale purification of the MBP-fusion proteins was carried out and ELISA tests were done using the polypeptides as target antigens.

DNA sequencing indicated that the EBV antigens expressed from clone R29-3 and R29-2 belonged to the large subunit of the EBV ribonucleotide reductase, which was encoded by the carboxyl-terminal end of the EBV BORF2 reading frame. The recombinant protein p28 expressed from clone R29-3 was 28.1 kDa in length, which was encoded by 759 bp of DNA. The recombinant protein p15 from clone R29-2 was 14.9 kDa in length, which was encoded by 405 bp of DNA. In addition, a 13.1 kDa EA-D antigen p13, which was expressed from clone B19-4, was encoded by the 352 bp carboxyl-terminal end of the BMRF1 reading frame of the EBV genome. Recombinant antigen, p18, expressed from clone P24-1, was a thymidine kinase (TK) coded by the carboxyl-terminal end of the EBV BXLF1 open reading frame. This recombinant antigen was 17.6 kDa in length and was encoded by 480 bp of DNA.

A total of 156 sera from patients with NPC and 100 sera from healthy donors were tested for IgA and IgG antibodies to native VCA and EA, as well as recombinant polypeptides p28, p15, p13 and p18 by ELISA. IgA-VCA antibodies were detected in 80% (125/156) of the NPC samples, whilst only 4/100 (4%) of the sera from healthy donors were positive for IgA to VCA. In the IgA-EA assay, however, sensitivity was calculated as 64% (100/156), but this test was more specific with a specificity of 99% (1/100 positive).

In the diagnosis of NPC, detection of anti-EBV RR IgG antibodies had a higher sensitivity of 80% by p28-ELISA and 69% by p15-ELISA. The detection of anti-EBV RR IgA antibodies had a sensitivity of 60% by p28-ELISA and 50% by p15-ELISA. However, it was more specific for the detection of IgA anti-EBV RR antibodies with a specificity of 100%. The detection of IgG anti-EBV RR antibodies had a specificity of 98%.

Results showed that IgG-ELISA test with recombinant p13 (EA-D) could detect 71% of NPC patients' sera, whereas only 1% of normal individuals gave positive results. IgA-ELISA was less sensitive with a sensitivity of 67% but more specific with a specificity of 100% compared with IgG-ELISA.

The TK-IgG ELISA showed a sensitivity of 57% and a specificity of 91%. The Tk-IgA Elisa, however, was not as good and gave a low sensitivity of 43%, but showed a higher specificity of 100%.

We have also tried to test the suitability of combining different recombinant antigens as target antigens in ELISA for the diagnosis of NPC. Recombinant antigen p28 belonging to EA-R and the recombinant antigen p13 belonging to EA-D were combined at a ratio of 2:1 and coated onto ELISA plates. It was found that 133 out of 156 NPC patients (85%) had IgG antibodies to the combined antigens, whilst only 2 out of 100 normal individuals (2%) had such antibodies. 111 samples from the 156 NPC patients (71%) had IgA antibodies against the combined antigens whereas only 1 out of 100 normal samples (1%) had such antibodies. The ELISA using combined antigens was more sensitive than the ELISA in which individual recombinant antigens as well as the native antigens were used separately.

The IgG ELISA with combined antigens also detected 9 out of 49 NPC sera which were IFA-VCA and IFA-EA negative for IgA antibodies. These results implied that combined antigens might be a good complementary diagnostic test to IFA. Based on the results of 72 patients with other diseases, some of them were having early primary symptoms of NPC, 18 cases were tested positive indicating EBV infection. A follow up study of these samples to see whether the EBV positive cases would finally lead to any EBV related carcinomas would be very useful in ensuring early detection and early treatment.

Conventionally, IgA was the antibody used in both IFA and ELISA for the diagnosis of NPC and this is supported by many research publications. In our study, when pure recombinant protein of different antigen complexes (recombinant ribonucleotide reductase (p28 and p15), EA-D (p13) and thymidine kinase (p18)) were analysed systematically by ELISA, we found that in fact IgG against the different recombinant antigens is much better than IgA in the diagnosis of NPC. In comparison, use of IgG as serological markers was not recommended when the conventional method of native VCA and EA, extracted from the whole cell lines, were used.

The IgG antibody titers against the combined antigens determined by ELISA were found to correlate well with the IgA anti-EA titers determined by IFA. Hence this assay might also be suitable for use in monitoring the recovery of NPC patients after radiation or other therapies. In addition, there was also a positive and significant correlation between the values of the OD readings of IgG anti-p28/p13 combined antigens determined by ELISA and the IgA titers against EA (r = 0.37) and VCA (r = 0.33) determined by IFA. Therefore the ELISA IgG test using combined antigens could be considered as a complementary screening assay to the IgA-IFA for the detection of NPC. Using the p28/p13 combined polypeptides as target antigen, there was also a significant correlation (r = 0.34) between the ELISA IgG antibody titers and the dOD readings determined by ELISA.

In summary, for the diagnosis of NPC, it is strongly recommended to use a mixture of several recombinant antigens in detecting antibodies to EBV by ELISA. Sensitivity is greatly improved. The ELISA with combined antigens p28/p13 is recommended for rapid and inexpensive serodiagnosis of populations with high risk of NPC. The combined recombinant antigens anti-EBV ELISA described here represent an innovative assay system for the diagnosis of NPC.

Purpose: In this study, a comparison of ratings of perceived exertion (RPE) [undifferentiated (RPE-overall) and differentiated (RPE-Legs, and RPE-Chest)] that correspond to the ventilatory breakpoint (Vpt) between different exercise modes (cycling and walking/running) using the Children’s OMNI Scale of Perceived Exertion was investigated. A response normalized rating of perceived exertion (RPE) that corresponds to the Vpt in 13- to 17-year-old male adolescents was identified. An estimation-production paradigm was used to determine whether the adolescents could self-regulate intensity on a cycle ergometer and treadmill exercise using a prescribed target RPE, discriminate between target RPEs and produce intermittent RPEs in both ascending and descending sequence.

Methods: Participants were randomly assigned into cycling and walking/running groups and underwent one orientation trial (O), one estimation trial (E) and two production trials (P). During O, each participant’s peak oxygen uptake (VO_2peak) was determined using a progressive multi-stage cycle ergometer protocol for cycling participants and a modified Astrand treadmill protocol for walking/running participants. Oxygen uptake (VO₂) and heart rate (HR) were recorded at the end of each stage of exercise. During E, undifferentiated (RPE-Overall) and differentiated RPE (RPE-Legs, RPE-Chest) were estimated and VO2 and HR were taken at every stage of a progressive cycle ergometer or treadmill test. Participants performed the exercise trials in O and E to volitional exhaustion. Significance for all statistical analysis was set at P < 0.05, unless otherwise stated. During P, participants adjusted the brake force on the cycle ergometer or speed on the treadmill to produce either an RPE sequence of 4 and 6 (ascending) or 6 and 4 (descending) in the 8-minute intermittent production trials.

Results: Exercise mode effect was not observed for any descriptive or dependent variable. Vpt corresponded to 67.7% VO_2peak for the cycling group and 70.4% VO_2peak for the walking/running group. RPE-Overall-Vpt (mean cycling, 5.1; mean walking/running, 4.8), RPE-Legs-Vpt (mean cycling, 5.6; mean walking/running, 5.2) and RPE-Chest-Vpt (mean cycling, 4.4; mean walking/running, 5.2) did not differ between the two groups. Response normalized RPE-Overall-Vpt for male adolescents aged 13-17 years was identified at 5. VO2 did not differ between E and P at target RPE of 4 (1.59 versus 1.57 L.min-1) and 6 (1.87 versus 1.79 L.min-1). HR did not differ between E and P at a target RPE of 4 (152.4 versus 151.1 beats.min-1) and 6 (167.1 versus 162.4 beats.min-1). Both VO2 and HR were higher (P < 0.01) at target RPE-6 than -4. Reponses were not affected by exercise mode or production sequence.

Conclusion: Undifferentiated and differentiated RPE-Vpt were similar between cycling and walking/running participants. A comparatively stable RPE-Vpt for 13- to 17-year old male adolescents indicates a group normalized perceptual response at RPE-5. Male adolescents were able to use the OMNI Scale to self-regulate short-duration intermittent cycle and treadmill exercise intensity. The responses support prescription congruence and intensity discrimination of exercise intensity in 13- to 17-year old male adolescents.

Background: Rehydration plays an important role in the recovery process after exercise especially in hot, humid conditions so that performance in subsequent training sessions or competition is not affected. Inclusion of sodium in drinks will assist in fluid retention and rehydration but most commercially available sports drinks only partially replace the sodium loss through sweating in humans during exercise. A recently developed rehydration solution called Drip Drop compensates for sodium sweat loss (20 – 80 mmol/L) without compromising the drink’s palatability but its efficacy in post-exercise rehydration or performance has not been evaluated.

Objective: This study aimed to investigate the efficacy of ingesting the Drip Drop beverage, with a high electrolyte concentration, on post-exercise rehydration and subsequent pre- load time trial performance in comparison with a sports drink and water.

Methods: Nine healthy males aged 24 ± 2 years (mean ± SD), with peak oxygen uptake (V0₂ peak) 55.2 ± 5.9 mL/kg/min completed three experimental trials in a randomised manner ingesting water (WA), a carbohydrate-electrolyte drink (SD; carbohydrate: 62 g/L, sodium: 31 ± 3 mmol/L, potassium: 5.3 ± 0.3 mmol/L) or the Drip Drop solution (DD; carbohydrate: 33 g/L, sodium: 60 ± 3 mmol/L, potassium: 18.2 ± 0.4 mmol/L). A total fluid volume equivalent to 150% of sweat loss was ingested during the preceding 75 min cycle at 65% V0₂ peak (temperature: 30.4 ± 0.3°C, relative humidity: 76 ± 1%, simulated wind speed: 8.0 ± 0.6 m/s) and within the first 2 h of recovery (temperature: 23.0 ± 1.0°C, relative humidity: 67 ± 2%). A pre-load time trial consisting of a 45 min cycle at 65% V0₂ peak followed by a 20 km time trial was conducted after a further 3 h of recovery.

Results: Fluid retention was higher with Drip Drop (30 ± 15%) than with water (-4 ± 19%; p < 0.001) or sports drink (10 ± 15%; p = 0.002). Drip Drop (6.03, CI: 0.10 – 7.46; p = 0.008) and sports drink (4.71, CI: 0.04 – 6.32; p = 0.015) were rated saltier than water (0.10, CI: 0.02 – 2.06) but similar between Drip Drop and sports drink (p = 0.02). Mean ratings of palatability were similar among drinks (WA: 4.25 ± 2.60; SD: 5.61 ± 1.79; DD: 5.40 ± 1.58; p = 0.33). Although time trial performance was similar across all three trials (WA: 2365 ± 321 s; SD: 2252 ± 174 s; DD: 2268 ± 184 s; p = 0.65), completion time was faster in 8 participants with sports drink and 7 participants with Drip Drop than with water. Comparing sports drink with Drip Drop, completion time was reduced in 5 participants and increased in 4 participants.

Conclusion: The Drip Drop solution is more effective at restoring fluid deficit during recovery from exercise than a commercially available sports drink or water without compromising the drink’s palatability with increased sodium concentration. There is a trend towards better endurance exercise time trial performance with Drip Drop and sports drink than with water.

This thesis described the complexation reaction of selected ligands with organotin chlorides.

In the first chapter, the basic concepts of organotin compounds, ligands and multi-dentate ligands were described. Organotin complexes, with examples, were classified into several categories according to their structures. The second part of this chapter is a review on the factors that affect the formation of adducts.

In the second chapter, the reactions of 8-aminoquinoline (Aq) with diorganotin dichlorides (R₂SnCl₂, R = Ph, Me, Bu) were described in the first part. With Ph₂SnCl₂, a stable adduct, Aq-Ph₂SnCl₂, was obtained. However, the reaction of BU₂SnCl₂ with Aq, irrespective of the solvent used, gave hydrolytic products, [(BU₂SnCl)₂O]₂ and HCI, which subsequently protonated the ligand to form Aq-HCI. Depending on the solvent used, Me₂SnCl₂ reacted with Aq and yielded either an adduct, Aq-Me₂SnCl₂, or hydrolytic and protonated products, [(Me₂SnCl)₂O]₂ and Aq-HCI. Additionally, the adduct, Aq-Me₂SnCl₂ was found to decompose into Aq-HCI and [(Me₂SnCl)₂O]₂ upon recrystallisation in a commercial solvent. The adducts, Aq-Ph₂SnC1₂ and Aq-Me₂SnCl₂, were characterised with various physical methods including X-ray crystallography. These results confirmed that the decreasing sequence of acceptor strength for diorganotin dichlorides with respect to Aq is Ph₂SnC1₂ > Me₂SnC1₂ > Bu₂SnC1₂.
,br>The second part of Chapter 2 was concerned about the reactions of 8-methyl-aminoquinoline (MeAq) with R₂SnC1₂ (R = Ph, Me, Bu) in commercial chloroform. The reaction of Me₂SnC1₂ with MeAq gave (HMeAq)₂Me₂SnC1₄, together with [(Me₂SnC1)₂O]₂ while MeAq-HCI and [(Bu₂SnC1)₂O]₂ were isolated from the reaction with Bu₂SnC1₂. Diphenyltin dichloride reacted with MeAq and afforded an unstable adduct, MeAq-Ph₂SnC1₂, which was easily decomposed into (HMeAq)₂Ph₂SnC1₄ and [(Ph₂SnC1)₂O]₂ upon recrystallisation in hot CHC1₃. The adduct, MeAq-Ph₂SnC1₂, and the complex salts, (HMeAq)₂R₂SnC1₄ (R = Ph, Me), were characterised. All the products isolated revealed that the hydrolytic rate of R₂SnC1₂ followed the order of Ph₂SnC1₂ < Me₂SnC1₂ < Bu₂SnC1₂ . In addition, results derived from product studies showed that the electron-donating ability of MeAq was reduced compared to that of Aq. It indicated that the methylation on donor atom (N) of a ligand reduced the coordination ability of this ligand.

The reactions of 8-methoxyquinoline (Mq) with R₂SnC1₂ (R = Ph, Me, Bu) and Ph₃SnC1 were discussed in the third chapter. All these RSnC1₂ reactions yielded not adduct but complex salts and protonated compounds, together with the corresponding distannoxanes. The reactions of Bu₂SnC1₂, irrespective of the solvent used, gave Mq-HC1 and [(Bu₂SnC1)₂O]₂ while (HMq)₂Me₂SnC1₄ and [(Me₂SnC1)₂O]₂ were isolated from the Mq reaction with Me₂SnC1₂. When Mq reacted with Ph₂SnC1₂ in chloroform and it yielded (HMq)₂Ph₂SnC1₄ and [(Ph₂SnC1)_2O]2 . However, when the same reaction was carried out in cyclohexane, an additional complex salt (Mq-HMq)Ph₂SnC1₃, besides (HMq)₂Ph2SnC1₄ and [(Ph₂SnC1)₂O]₂, was isolated. Furthermore, (Mq-HMq)Ph₂SnC1₃ was found to decompose into (HMq)₂Ph₂SnC1₄, [(Ph_2MSnC1)₂O]₂ and Mq-HCI. All the protonated products wre characterised, including X-ray structural analyses of (Mq-HMq)Ph₂SnC1₃ and (HMq)₂Ph₂SnC1₄. Isolation of these products reaffirmed that the hydrolytic rate of R₂SnC1₂ followed the sequence of Ph₂SnC1₂ < Me₂SnC1₂ < Bu₂SnC1₂. Additionally, the composition of (Mq-HMq)Ph₂SnC1₃ displayed that the [Ph₂SnC1₃] anion disproportionated into [Ph₂SnC1₄]^2-, a dianion, and Ph₂SnC1₂.

An outer-sphere coordination adduct, Mq-H₂O-Ph₃SnC1, with a different structure as usual, was prepared and characterised with various physical methods through the reaction of Mq and Ph₃SnC1.

The reactions of 3-methyladenine (3-MeAd) and N,N,9-trimethyladenine (Me3Ad) with R₂SnC1₂ (R = Ph, Me, Bu) were depicted in the fourth chapter. Dimethyltin dichloride reacted with 3-MeAd and produced a 2:1 adduct, (3-MeAd)₂Me₂SnC1₂, irrespective of the ratio of the donor and acceptor used. The reaction of Bu₂SnC1₂ with 3-MeAd gave [(Bu₂SnC1)₂O]₂ and 3-MeAd-HCI. However, the reaction of Ph₂SnCI₂ with 3-MeAd only afforded, besides Ph₃SnC1 and 3-MeAd-HCI, the disproportionated product, (3-MeAd)₂PhSnC1₂, which cannot be characterised probably due to its hydrolysis. From the structural characterisation of (3-MeAd)₂Me₂SnC1₂, the coordination site of 3-MeAd was confirmed to be at the N7 atom.

Me3Ad reacted with R₂SnC1₂ (R = Ph, Me, Bu) and yielded Me₃Ad-HCI, together with the corresponding distannoxanes, [(R₂SnC1)₂O]₂. These results reaffirmed that the methylation on donor atom (N) of a ligand reduced the coordination ability of this ligand.

This thesis is concerned with the development of amperometric enzyme electrodes based on chitosan and silica sol-gel immobilization techniques. Immobilization of enzymes onto electrode surfaces as applied to electrochemical biosensors has been a challenging and active research field among sensor technologists. Chitosan and novel silica sol-gel/chitosan derived carbon paste amperometric biosensors were developed.

Chapter 1 is an introduction to the biosensors, which are emerging as an important analytical tool widely applied in biochemical, pharmaceutical, and clinical fields. Amperometric enzyme electrode, which is a branch of biosensors, is discussed in detail with respect to the enzyme reaction kinetics, immobilization techniques and the performance of enzyme electrodes. It also covers the theoretical aspects involved in this thesis. Chitosan and silica sol-gel as immobilization support matrix, and electrochemical techniques including cyclic voltammetry and amperometry are discussed. The aims of the study are given at the end of the chapter.

In Chapter 2, a new enzymatic amperometric biosensor for the detection of hydrogen peroxide was developed via an easy and effective enzyme immobilization method using chitosan film cross-linked with glutaraldehyde. Horseradish peroxidase (HRP) was immobilized on the surface of a carbon paste electrode (CPE). Hexacyanoferrate (II) was present in the solution as a mediator. The fabrication procedure of the biosensor was optimized and the resulting biosensor exhibited a relatively fast response, wide linear range and good stability.

Chapter 3 describes the construction of a glucose biosensor with the following "sandwich" configuration: chitosan-ferrocene:GOX:chitosan using CPE as the basic electrode. The influence of operational parameters were explored in detail. Applicability of the biosensor was demonstrated in real samples.

In Chapter 4, a new type of silica sol-gel/chitosan hybrid material was used for the fabrication of an amperometric H₂O₂ biosensor. The composite consisting of sol-gel and natural polymer chitosan was used to immobilize HRP on a CPE electrode. The fabrication procedure was systematically optimized to improve the biosensor performance.

The significance and motivation of research on synthesis of a novel material, PC₃N₄ were reviewed. It is seen that this subject is both fundamentally and technologically significant and presents a great challenge to condensed matter physicists, material scientists and chemists as well. Liu and Cohen's prediction that a type of carbon nitride compound, β-C₃N₄. is comparable in hardness with or even harder than diamond, was also conceptionally described.

The progress and current status of research on this novel material were selectively described, showing that much progress had been achieved, particularly in four years before.

Based on physical and technical considerations, a reactive d magnetron sputtering was chosen as a technique to synthesise carbon nitride films in this experiment. The general principle of magnetron sputtering, and the structure and main features of the present purpose-designed rf magnetron sputtering equipment were briefly described. Using this equipment, the carbon nitride films were successfully deposited on various substrates, such as single- crystal KC1 (100), KBr (100) and Si (100) substrates, and stainless steel and glass wafers. The morphology, chemical bond states and compositions, structures and microstructures of these films were characterised by means of various analyses, such as infrared spectrum (IR), x-ray photoelectron spectroscopy (XPS), scanning electron microscopy (SEM), energy dispersive x-ray analysis (EDX), transmission electron microscopy (TEM), high resolution electron microscopy (HREM), x-ray diffraction (XRD) and atomic force microscopy (AFM).

The main findings in our experiment are summarised as follows.

1. The films deposited on single-crystal KBr (100), KC1 (100) and Si (100) substrates were of polycrystalline structure whereas the films on stainless steel wafers were amorphous.

2. CN grains with various sizes were distributed in different regions of the polycrystalline films. And the biggest grain was estimated to be over 10 μm in size.

3. Nitrogen atoms were bonded to carbon atoms and the composition ratio N/C for C₃N₄ component reached 1.23 in bulk and 1.27 at surface. These numbers were close to the expected stoichiometric value of 1.33, indicating a high quality of the film both in bulk and at surface.

4. The observed Laue patterns, Kikuchi lines, and HREM images exhibited that the observed CN grains were of single-crystalline and perfect structure.

5. The composition ratio both for overall nitrogen and carbon contents and for C₃N₄ component increased respectively up to saturation values, 0.87 and 1.24 in the bulk of the film deposited on Si (100) substrate.

6. The time behaviours of the composition ratio NIC measured by XPS, the C-N bonding states revealed in IR spectra and the measured growth rate of the C-N film were well consistent with each other. The deposition condition for improving the C-N film quality was recommended.

Notably, this is the first time that polycrystalline C₃N₄ films were successfully prepared by means of reactive rf magnetron sputtering. And some results about the film structure were acquired from those films also for the first time.

Finally, some important issues in the research on carbon nitride film were reviewed and some subjects were also suggested for further study.

The reported benefits underlying the importance of regular resistance exercise (RE) drive numerous national health organizations to recommend at least 2-3 days per week of RE, preferably spread 48-72 h apart to improve muscle strength and size. However, given the current recommendation for 3-5 days per week of aerobic exercise, the busy lifestyle of working individuals, and/or training schedule of elite athletes, it is inevitable that a portion of these individuals perform RE on consecutive days. Recommendations for the recovery period in between RE sessions from current guidelines stem from studies that examined molecular and cellular responses on muscle hypertrophy and strength, yet the influence of recovery period on cytokines and red blood cell (RBC) parameters remains unclear. Therefore, this study seeks to explore the acute cytokines and RBC responses to RE when performed on 3 non-consecutive (NC) or consecutive (C) days in recreationally active men.

Thirty healthy, recreationally active young males were randomized to 3 NC (~48-72 h between sessions) or 3 C (~24 h between sessions) days of RE in a week. For each RE session, both groups performed 3 sets x 10 repetitions x 10- repetition maximum (RM) of leg press, latissimus pulldown, leg curl, shoulder press, and leg extension. Cytokines [Chemokine ligand 2 (CCL2), Fractalkine (CX3CL1), Interleukin-8 (CXCL8), Interferon (IFN)-γ, and Tumor Necrosis Factor-like weak inducer of apoptosis (TWEAK)] were measured at Pre-RE, 0 h, 1 h and 24 h post-3rd RE, while RBC parameters [RBC count, hemoglobin (Hb), hematocrit (Hct), mean cell volume (MCV), and mean corpuscular hemoglobin concentration (MCHC)] were measured at Pre, 0 h, and 24 h post-3rd RE. Analysis was done using (time x group) mixed repeated measures ANOVA or Aligned Rank Transformation ANOVA.

Both groups were similar in age [mean 25 (SD 2) years], height [1.72 (0.06) m], body mass [65 (10) kg], BMI [22.2 (2.7) kg·m² ], systolic and diastolic blood pressure [114 (5)/ 69 (8) mmHg], fasting glucose [4.5 (0.3) mmol·L^-1 ], and physical activity level [2144 (1428) MET-min·week-1 ] at baseline (p = 0.161-0.999). There were no time x group interaction found for all cytokines (p = 0.476 – 0.948) and RBC parameters (p = 0.359-0.680). Time effect was significant for all cytokines (p = 0.002 – 0.028) except for CXCL8 (p = 0.057). Additionally, no group differences were found (p = 0.277 – 0.941). Time effect was significant for all RBC parameters (p = < 0.001 – 0.013). Although group differences were noted in both Hb and MCHC, it was because the mean (SD) for C group (Hb: 15.2 (0.9) g•dL^-1 , MCHC: 35.3 (0.9) g•dL^-1) were higher than NC group (Hb: 14.4 (1.2) g•dL^-1, MCHC: 34.1 (0.9) g•dL^-1 ) at the start of the program.

Acute C days of RE does not have any detrimental effect on cytokines and RBC parameters compared to NC days of RE. Respectively, the parameters reviewed are a reflection of the immune system and health indices. This study is significant as the findings would provide further evidence in the consideration of recovery period for exercise guidelines.

This thesis describes in detail two novel applications of the back-propagation neural network. In the first application, the neural network is viewed as a component extractor. Here, the network attempts to dynamically find the best indicators (through non-linear weighted averaging) that give a trading signal that matches as close as possible the perfect foresight. The results obtained after training the network on the Kuala Lumpur Stock Exchange Composite Index are presented and discussed. In the second application, the network is used to forecast the modified regressed slopes of price returns. Thus, the ideas of both regression and neural networks are fruitfully combined. The forecasted value is used in a trading strategy that is reasonable and intuitive. The assumptions involved in using the regressed slope are inspected critically. Attention is given to performance. This is accomplished by means of the Sharpe Ratio by which ambiguity that may result from benchmarking a strategy against other indicators is avoided.

In the early chapters, we start off by considering the basic neural network unit, the linear threshold gate, and investigate its computational capability. Then, the ideas are generalized to polynomial threshold gates. Having seen how Boolean functions can be realized, some fundamental results are summarized about the capability of neural networks to approximate continuous functions. Back-propagation is introduced as an intuitive but powerful rule that is used in neural network learning. Some issues pertaining to the convergence and speed of convergence of back-propagation networks are discussed in detail. These topics include weight initialisation, tuning the learning rate and momentum parameters, and the effects of the nature of the activation function. The generalization capability of back-propagation networks is also investigated. A widely used generalization technique called cross-validation is presented.

We then proceed with the background for the applications, by first giving an overview of fundamental and technical analysis and then illustrate why technical analysis is useful in gauging the market. The indicators used for the first application, the portfolio selection problem, are discussed. Some aspects of the theory of least squares regression are seen. The motivation for using the regressed slope of returns in our second application is presented. Then, the properties of regression parameter estimates are explored. These properties are found to be attractive in view of our applications. The Guess-Markov conditions which must be satisfied for the desirable properties of unbiasedness and "bestness" are stated.

The applications are then discussed, the essentials of which have been described in the first paragraph. In conclusion, we summarize all the work done and explore some related areas where further research can be conducted. The appendix includes the complete source-code of the back-propagation neural network with the additional enhancements of adaptive learning rates and momentum. The program, which was used for all the simulations, is designed to work in all C-compatible platforms as minimum reliance is placed on special functions (like the random number generator) that may be used in a particular platform.

Dinitrobenzenes are one of the main contaminants appeared in the discharged wastewater around army ammunition production plants and certain chemical industries. These organic compounds are very stable in nature and are difficult to be treated using conventional technologies. In this work, the advanced oxidation processes (AOPs), in the field of which much attention has been attracted recently for wastewater treatment, were applied to degrade 1,3-dinitrobenzene in aqueous solution.

Direct photolysis of 1,3-dinitrobenzene by polychromatic light was studied. The main detected intermediates were 2,4-dinitrophenol, 3-nitrophenol and resorcinol. The direct photolysis of 1,3-dinitrobenzene followed zero-order kinetics. The degradation rate was around 10^-9 M.s^-1 under the experimental conditions.

In UV/H₂O₂ process, the degradation of 1,3-dinitrobenzene followed first-order kinetics and the range of degradation rate was from 10^-8 to 10^-7 M.s^-1. 2,4-dinitrophenol, 3-nitrophenol, nitrohydroquinone, 4-nitrochatechol, resorcinol, catechol, hydroquinone were identified as the major aromatic intermediates. Reaction mechanisms and kinetics were discussed. The effect of initial concentration of H₂O₂, initial pH value and mixing rate was studied for the degradation of 1,3- dinitrobenzene.

Photocatalytic degradation was carried out for 1,3-dinitrobenzene in aqueous solution with UV/TiO₂ and UV/ZnO processes. 2,4-dinitrophenol and 3-nitrophenol were identified as the main organic intermediates. Both processes were proved to be inefficient for 1,3-dinitrobenzene under the experimental conditions. The initial reaction rates were around 10^-9 M.s^-1.

UV/Fenton and UV/Fenton-like processes were used to degrade 1,3-dinitrobenzene in aqueous solution. It was found that the addition of ferric and ferrous ions enhanced the reaction rate greatly. The initial reaction rates in UV/Fe²⁺/H₂O₂, UV/Fe³⁺/H₂O₂ processes were around 10^-7 M.s^-1. The UV/ferrioxalate/H₂O₂ process with the initial reaction rate around 10^-6 M.s^-1. A conclusion can be made from the above results that UV/ferrioxalate/H₂O₂ process is the most efficient technology among all the AOPs used in the experiment.

Photolysis of H₂O₂ in aqueous solution irradiated by monochromatic light was also studied. A possible reaction mechanism and corresponding reaction rate expressions were proposed. The experimental results also showed that the decomposition of H₂O₂ in aqueous solution irradiated by both polychromatic and monochromatic light followed zero-order kinetics.

In addition, a detailed literature review related to the degradation of nitroaromatic compounds was also given in the thesis.