O presente trabalho consiste no desenvolvimento de uma nova técnica de identificação de materiais poliméricos por meio do uso de um nariz eletrônico. Narizes eletrônicos têm sido desenvolvidos para detecção automática e classificação de odores e gases. São instrumentos capazes de medir a concentração ou intensidade odorante de modo similar a um olfatômetro, mas sem as limitações inerentes ao uso de painel humano, o que é altamente desejável. O nariz eletrônico é composto por um sistema de sensores, no nosso caso, utilizamos um arranjo de quatro sensores, que foram confeccionados pela deposição de finos filmes de polímeros condutores dopados sobre a superfície de eletrodos interdigitados. Estes sensores foram conectados a condutivímetros acoplados a um computador de uso pessoal (PC) através de um conversor AD. O PC era dotado de softwares de aquisição e tratamento de dados. Amostras dos materiais a serem analisados foram aquecidas a 257°C e o arranjo de sensores foi exposto aos compostos voláteis produzidos durante esse aquecimento. Realizaram-se 30 ensaios formados por períodos de exposição (5 segundos; compostos voláteis) intercalados por períodos de recuperação (45 segundos; ar puro). Os dados obtidos foram tratados estatisticamente por Análises de Componentes Principais (PCA). Esse arranjo de sensores mostrou-se eficiente, sendo capaz de diferenciar nove tipos de materiais poliméricos testados. Apresentou 100% de acerto em 30 ensaios de classificação realizados. / This research consists in the development of a new technique capable of the identification of polymeric materials using an electronic nose. Electronic noses have been developed for automatic detection and classification of odors, vapors and gases. They are instruments capable of measuring the concentration or intensity of an odorant similarly to an olfactometer, but without the inherent limitations of the human panel, which is highly desired. The electronic nose is composed by a system of chemoresistive sensors, in this case, an array of four sensors was used, which were made through a deposition of thin films of doped conductive polymers, on the surface of interdigitated electrodes. These sensors were connected to conductivity meters coupled to a personal computer (PC) through AD converters. The PC had acquisition and data processing softwares installed on it. Thirty readings were made or each analyzed polymer consisting of alternated 5 seconds exposure periods and 45 seconds recovery periods. The collected data were statistically processed by Principal Component Analysis (PCA). This electronic nose was efficient, being able to identify nine types of polymeric materials through the analysis of the different volatile compounds released when these materials were heated to the heat of 257ºC. A 100% correct classification score was obtained in the 30 sets of analysis.

Ouyang, Mengxing. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2009. / Includes bibliographical references (leaves 83-88). / Abstracts in English and Chinese. / Chapter 1. --- Introduction --- p.1 / Chapter 1.1 --- Background and Motivation --- p.1 / Chapter 1.2 --- Objectives --- p.2 / Chapter 1.3 --- Contributions --- p.3 / Chapter 1.4 --- Organization of the Dissertation --- p.3 / Chapter 2. --- CNTs and Their Application as Sensors --- p.5 / Chapter 2.1 --- Introduction to CNTs --- p.5 / Chapter 2.2 --- CNTs Based Sensors --- p.8 / Chapter 3. --- F-CNTs Based Ethanol Sensors --- p.10 / Chapter 3.1 --- Introduction --- p.10 / Chapter 3.1.1 --- Carbon Nanotube Alcohol Sensors --- p.11 / Chapter 3.1.2 --- CNTs Sensor Configuration --- p.12 / Chapter 3.1.3 --- Activation of CNTs Sensor --- p.13 / Chapter 3.1.4 --- Functionalization of CNTs --- p.14 / Chapter 3.2 --- Fabrication of F-CNTs Based Ethanol Sensors --- p.16 / Chapter 3.2.1. --- Fabrication of f-CNTs --- p.16 / Chapter 3.2.2. --- Fabrication of Sensing Elements by DEP Manipulation --- p.17 / Chapter 3.2.3. --- Experimental Setup --- p.19 / Chapter 3.2.4. --- Mechanism of Ethanol Sensor --- p.20 / Chapter 3.3 --- Characterization of F-CNTs Based Ethanol Sensors --- p.21 / Chapter 3.3.1. --- I-V Characteristics --- p.21 / Chapter 3.3.2. --- Thermal Sensitivity --- p.22 / Chapter 3.3.3. --- Stability --- p.23 / Chapter 3.3.4. --- FFT and Spectral Analysis --- p.23 / Chapter 3.4 --- Performance of F-CNTs Based Ethanol Sensors --- p.26 / Chapter 3.4.1. --- Typical Response --- p.26 / Chapter 3.4.2. --- Selectivity --- p.27 / Chapter 3.4.3. --- Towards Low Concentration --- p.28 / Chapter 3.4.4. --- Towards Realistic application --- p.29 / Chapter 3.5 --- Constant Power Configuration --- p.32 / Chapter 3.5.1. --- Constant Power Circuit for Ethanol Detection --- p.32 / Chapter 3.5.2. --- Sensor Response versus Power --- p.35 / Chapter 3.5.3. --- Responsivity --- p.37 / Chapter 3.5.4. --- Noise --- p.38 / Chapter 3.5.5. --- Sensitivity --- p.39 / Chapter 3.5.6. --- Dynamic Response --- p.41 / Chapter 3.6 --- Comparison between F-MWNTs and MWNTs --- p.43 / Chapter 3.6.1. --- I-V Characteristics --- p.43 / Chapter 3.6.2. --- Cycling Response --- p.44 / Chapter 3.6.3. --- Dynamic Response --- p.46 / Chapter 3.6.4. --- Sensor Performance under Different Power --- p.48 / Chapter 3.7 --- Summary --- p.53 / Chapter 4. --- EG-CNTs Based Flow Sensors --- p.55 / Chapter 4.1 --- Introduction to CNTs Flow Sensors --- p.55 / Chapter 4.2 --- EG-CNTs and Their Applications --- p.56 / Chapter 4.2.1 --- Intro to EG-CNTs Sensor --- p.56 / Chapter 4.2.2 --- Fabrication of EG-CNTs Sensor --- p.57 / Chapter 4.2.3 --- Experimental Characterization --- p.59 / Chapter 4.2.3.1. --- I-V Characteristics --- p.59 / Chapter 4.2.3.2. --- Thermal Sensitivity --- p.61 / Chapter 4.2.3.3. --- Humidity responsivity --- p.63 / Chapter 4.2.3.4. --- Stability --- p.65 / Chapter 4.2.3.5. --- Hysteresis --- p.66 / Chapter 4.2.4 --- Summary --- p.68 / Chapter 4.3 --- Fabrication of EG-CNTs Flow Sensor --- p.70 / Chapter 4.3.1. --- Fabrication Procedure --- p.70 / Chapter 4.3.2 --- Experimental Setup --- p.73 / Chapter 4.4 --- Characterization of EG-CNTs Flow Sensor --- p.74 / Chapter 4.4.1. --- Typical Response --- p.74 / Chapter 4.4.2. --- Power Consumption --- p.75 / Chapter 4.4.3. --- Repeatability --- p.77 / Chapter 4.4.4. --- Flow Sensitivity --- p.78 / Chapter 4.5 --- Summary --- p.79 / Chapter 5. --- Conclusion --- p.80 / Chapter 6. --- Bibliography --- p.83

O presente trabalho consiste no desenvolvimento de uma nova técnica de identificação de materiais poliméricos por meio do uso de um nariz eletrônico. Narizes eletrônicos têm sido desenvolvidos para detecção automática e classificação de odores e gases. São instrumentos capazes de medir a concentração ou intensidade odorante de modo similar a um olfatômetro, mas sem as limitações inerentes ao uso de painel humano, o que é altamente desejável. O nariz eletrônico é composto por um sistema de sensores, no nosso caso, utilizamos um arranjo de quatro sensores, que foram confeccionados pela deposição de finos filmes de polímeros condutores dopados sobre a superfície de eletrodos interdigitados. Estes sensores foram conectados a condutivímetros acoplados a um computador de uso pessoal (PC) através de um conversor AD. O PC era dotado de softwares de aquisição e tratamento de dados. Amostras dos materiais a serem analisados foram aquecidas a 257°C e o arranjo de sensores foi exposto aos compostos voláteis produzidos durante esse aquecimento. Realizaram-se 30 ensaios formados por períodos de exposição (5 segundos; compostos voláteis) intercalados por períodos de recuperação (45 segundos; ar puro). Os dados obtidos foram tratados estatisticamente por Análises de Componentes Principais (PCA). Esse arranjo de sensores mostrou-se eficiente, sendo capaz de diferenciar nove tipos de materiais poliméricos testados. Apresentou 100% de acerto em 30 ensaios de classificação realizados. / This research consists in the development of a new technique capable of the identification of polymeric materials using an electronic nose. Electronic noses have been developed for automatic detection and classification of odors, vapors and gases. They are instruments capable of measuring the concentration or intensity of an odorant similarly to an olfactometer, but without the inherent limitations of the human panel, which is highly desired. The electronic nose is composed by a system of chemoresistive sensors, in this case, an array of four sensors was used, which were made through a deposition of thin films of doped conductive polymers, on the surface of interdigitated electrodes. These sensors were connected to conductivity meters coupled to a personal computer (PC) through AD converters. The PC had acquisition and data processing softwares installed on it. Thirty readings were made or each analyzed polymer consisting of alternated 5 seconds exposure periods and 45 seconds recovery periods. The collected data were statistically processed by Principal Component Analysis (PCA). This electronic nose was efficient, being able to identify nine types of polymeric materials through the analysis of the different volatile compounds released when these materials were heated to the heat of 257ºC. A 100% correct classification score was obtained in the 30 sets of analysis.

Overabundant nitrate in Iowa’s surface water threatens stream health, drinking water safety, and significantly contributes to hypoxic zones in the Gulf of Mexico. Researchers have quantified surface water nitrate loads historically with grab samples and, more recently, in-situ sensors. In-situ sensor networks capture changes in nitrate concentration over small time scales, providing high temporal resolution data to accurately calculate nitrate loading. However, because advanced sensors are expensive, spatial resolution is often compromised when sensors are deployed on large rivers. To collect high spatial resolution nitrate samples that complement the high temporal resolution data from in-situ sensors, we first used traditional grab samples on small, non-navigable streams in the Clear Creek and the English River watersheds. Dense grab samples across watersheds provide higher resolution data, but not at the spatial resolution achievable on navigable streams with newly developed, boat-deployed sensor technology. We constructed a boat-deployed sensor system that automatically measured nitrate concentrations, temperature, dissolved oxygen, conductivity, and pH as we navigated a boat on a given waterbody. We used the system on the Iowa and Cedar Rivers to capture spatial and temporal changes never previously observed in Iowa. Our data suggest nitrate concentrations and yields were highest in low-relief landforms dominated by row crop agriculture. Nitrate concentrations were lower in higher-relief landforms with less row crop production. We also measured water in Storm Lake, IA with the boat-deployed system. We measured little heterogeneity of nitrate concentrations in the lake, but observed significant nitrate reduction in a large wetland just upstream. The system captured fine scale spatial dynamics of nitrate reduction in the wetland and low nitrate concentrations throughout Storm Lake. Our newly developed sensor platform captured high resolution water quality data, complementing the high temporal resolution data collected with in-situ sensors. High spatial resolution data in this and similar studies provide powerful insights for decision makers to target problematic areas, reduce nitrate, and improve water quality.

Electronic noses combine an array of cross-selective gas sensors with a pattern recognition engine to identify odors. Pattern recognition of multivariate gas sensor response is usually performed using existing statistical and chemometric techniques. An alternative solution involves developing novel algorithms inspired by information processing in the biological olfactory system. The objective of this dissertation is to develop a neuromorphic architecture for pattern recognition for a chemosensor array inspired by key signal processing mechanisms in the olfactory system. Our approach can be summarized as follows. First, a high-dimensional odor signal is generated from a chemical sensor array. Three approaches have been proposed to generate this combinatorial and high dimensional odor signal: temperature-modulation of a metal-oxide chemoresistor, a large population of optical microbead sensors, and infrared spectroscopy. The resulting high-dimensional odor signals are subject to dimensionality reduction using a self-organizing model of chemotopic convergence. This convergence transforms the initial combinatorial high-dimensional code into an organized spatial pattern (i.e., an odor image), which decouples odor identity from intensity. Two lateral inhibitory circuits subsequently process the highly overlapping odor images obtained after convergence. The first shunting lateral inhibition circuits perform gain control enabling identification of the odorant across a wide range of concentration. This shunting lateral inhibition is followed by an additive lateral inhibition circuit with center-surround connections. These circuits improve contrast between odor images leading to more sparse and orthogonal patterns than the one available at the input. The sharpened odor image is stored in a neurodynamic model of a cortex. Finally, anti-Hebbian/ Hebbian inhibitory feedback from the cortical circuits to the contrast enhancement circuits performs mixture segmentation and weaker odor/background suppression, respectively. We validate the models using experimental datasets and show our results are consistent with recent neurobiological findings.

The boronic acid functional group is known to bind compounds with the diol group tightly and reversibly in aqueous environment and has been used as a recognition moiety for the design of carbohydrate sensors. The first chapter of the dissertation studies the synthesis and substitution effect on the affinity and selectivity of a known boronic acid-based glucose sensor. In such a sensor design effort, the availability of a signaling event, whether it is fluorescence or UV, is crucial. The second chapter studies the detailed mechanism on how a well-known fluorescent boronic acid compound changes fluorescent properties upon binding. A new mechanism has been established which corrected a decade old mistake. In the third chapter, a series of boronic acid-based sensors were designed and synthesized for sialic acid, which is part of tetrasaccharide found on many cell surface carbohydrates. Such sialic acid sensors could be very useful for the development of new type of anti-influenza therapy. The fourth is on the design and synthesis novel and selective inhibitors for phosphodiesterase 4 (PDE4), which are potential anti-asthma agents.

It is my thesis that a cost-effective motion capture system for wireless 3D controllers can be developed through the use of low-cost inertial measurement devices and camera systems. Current optical motion capture systems require a number of expensive high-speed cameras. The use of such systems is impractical for many applications due to its high cost. This is particularly true for consumer-level wireless 3D controllers. More importantly, optical systems are capable of directly tracking an object with only three degrees of freedom. The proposed system attempts to solve these issues by combining a low-cost camera system with low-cost micro-machined inertial measurement devices such as accelerometers and gyro sensors to provide accurate motion tracking with a full six degrees of freedom. The proposed system combines the data collected from the various sensors in the system to obtain position information about the wireless 3D controller with 6 degrees of freedom. The system utilizes a number of calibration, error correction, and sensor fusion techniques to accomplish this task. The key advantage of the proposed system is that it combines the high long-term accuracy and low frequency nature of the camera system and complements it with the low long-term accuracy and high frequency nature of the inertial measurement devices to produce a system with a high level of long-term accuracy with detailed high frequency information about the motion of the wireless 3D controller.

It is my thesis that a cost-effective motion capture system for wireless 3D controllers can be developed through the use of low-cost inertial measurement devices and camera systems. Current optical motion capture systems require a number of expensive high-speed cameras. The use of such systems is impractical for many applications due to its high cost. This is particularly true for consumer-level wireless 3D controllers. More importantly, optical systems are capable of directly tracking an object with only three degrees of freedom. The proposed system attempts to solve these issues by combining a low-cost camera system with low-cost micro-machined inertial measurement devices such as accelerometers and gyro sensors to provide accurate motion tracking with a full six degrees of freedom. The proposed system combines the data collected from the various sensors in the system to obtain position information about the wireless 3D controller with 6 degrees of freedom. The system utilizes a number of calibration, error correction, and sensor fusion techniques to accomplish this task. The key advantage of the proposed system is that it combines the high long-term accuracy and low frequency nature of the camera system and complements it with the low long-term accuracy and high frequency nature of the inertial measurement devices to produce a system with a high level of long-term accuracy with detailed high frequency information about the motion of the wireless 3D controller.

Electronic noses combine an array of cross-selective gas sensors with a pattern recognition engine to identify odors. Pattern recognition of multivariate gas sensor response is usually performed using existing statistical and chemometric techniques. An alternative solution involves developing novel algorithms inspired by information processing in the biological olfactory system. The objective of this dissertation is to develop a neuromorphic architecture for pattern recognition for a chemosensor array inspired by key signal processing mechanisms in the olfactory system. Our approach can be summarized as follows. First, a high-dimensional odor signal is generated from a chemical sensor array. Three approaches have been proposed to generate this combinatorial and high dimensional odor signal: temperature-modulation of a metal-oxide chemoresistor, a large population of optical microbead sensors, and infrared spectroscopy. The resulting high-dimensional odor signals are subject to dimensionality reduction using a self-organizing model of chemotopic convergence. This convergence transforms the initial combinatorial high-dimensional code into an organized spatial pattern (i.e., an odor image), which decouples odor identity from intensity. Two lateral inhibitory circuits subsequently process the highly overlapping odor images obtained after convergence. The first shunting lateral inhibition circuits perform gain control enabling identification of the odorant across a wide range of concentration. This shunting lateral inhibition is followed by an additive lateral inhibition circuit with center-surround connections. These circuits improve contrast between odor images leading to more sparse and orthogonal patterns than the one available at the input. The sharpened odor image is stored in a neurodynamic model of a cortex. Finally, anti-Hebbian/ Hebbian inhibitory feedback from the cortical circuits to the contrast enhancement circuits performs mixture segmentation and weaker odor/background suppression, respectively. We validate the models using experimental datasets and show our results are consistent with recent neurobiological findings.

The purpose of this work is the development of MEMS-based resonant sensors for liquid-phase biochemical sensing applications. Specifically, the sensors developed here are aimed at embedded or point-of-sampling applications: (1) when there is not enough time to send a sample to a lab for analysis, (2) in resource-poor settings, (3) when collecting analyte and shipping it to a lab would damage the sample, or (4) for in-situ monitoring. To this end, a bulk micromachined resonant cantilever sensor and a surface micromachined sensor based on the spring-softening effect are investigated as transducer elements. The developed cantilever resonators are operated in an in-plane vibration mode to reduce fluid damping and mass loading by the surrounding fluid. The surface of the resonator is either coated with a chemically sensitive polymer film for chemical sensing or with a layer of protein or antibody for biosensor testing. Chemical tests for sensing volatile organic compounds using polymer-coated in-plane resonators in the liquid-phase give estimated limits of detection below 100 ppb. In addition, biosensor tests for the detection of anti-IgG yield estimated limits of detection around 100 ng/ml. In an attempt to further improve sensor reliability and to further lower the limits of detection, a second sensing concept has been investigated. The presented sensing scheme is capacitive with a resonator acting as an analog-to-digital converter. The resonator and the sensing capacitors are coupled via the spring softening effect. Through this mechanism a change in capacitance causes a shift in resonant frequency. Extensive device modeling has been performed and a process has been developed allowing for fabrication and on-chip packaging of these sensor structures. Initial mechanical characterization data show that the resonators do in fact vibrate.