An overview of synchrotrons and synchrotron radiation is presented, along with the theory and practical considerations behind several types of X-ray spectroscopy. The theory and practical considerations of density functional theory are also given, with direct reference to some specific software packages.<p> Some synchrotron-excited X-ray spectroscopy measurements and density functional theory calculations of selenium and arsenic-doped selenium films are then outlined. The physical structure of crystalline and amorphous selenium and the electronic structure of amorphous selenium are discussed and comparison is made to the experimental results.<p> A weak feature in the conduction band is identified as a "fingerprint" of the degree of crystallization in amorphous selenium from X-ray absorption measurements. Similarly, a weak feature corresponding to lone-pairs in the valence band is identified as a "fingerprint" of the arsenic concentration from X-ray emission measurements.<p> Finally a detailed model of the structure of amorphous selenium is explained, and compared to experiment. This model is tested both by direct calculations and by a reverse Monte Carlo approach. The implications of this model with respect to the structure of amorphous and arsenic-doped amorphous selenium are discussed. Calculations suggest that simply randomizing the arrangement of "perfect" trigonal selenium is unable to reproduce the measurements of amorphous selenium; a moderate variation in the bond angle of "perfect" trigonal selenium is also necessary.

The electronic properties of amorphous materials are greatly affected by the density of localized states in the mobility gap of these materials. The exact shape of the density of states (DOS) distribution in amorphous selenium (a-Se) is still unresolved despite decades of research. One of the most commonly employed methods to investigate charge transport properties in high resistivity materials is time-of-flight (TOF) transient photoconductivity experiment. The TOF transient photoconductivity technique is used to measure the induced photocurrent in the external circuit when the sample is photoexcited. Information pertaining to carrier mobility and other carrier parameters are deduced from the shape of the photocurrent. The investigation of the charge transport phenomenon is well known to be a complicated task. Monte Carlo (MC) simulation method has become a standard method for carrier transport studies in amorphous materials. The purpose of this research work is to develop a Monte Carlo simulation model for charge transport in typical TOF transient photoconductivity experiment to investigate the DOS distribution in a-Se. The MC simulations were first performed for relatively simpler models for which theoretical and analytical solutions were available. The MC model developed here is based on simulating the drift of carriers resulting from photogeneration, subject to the influence of an applied electric field and multiple trapping events. The free drift time of photocarriers and their dwell time in the traps are stochastic in nature, in accordance with the probabilities of these events. Electron time-of-flight transient photocurrents were calculated in amorphous selenium as a function of the electric field. The distribution of localized states (DOS) in a-Se has been investigated by comparing the experimentally measured and calculated transient photocurrents. The analysis of multiple-trapping transport has been done by the discretization of a continuous DOS. The DOS distribution has been optimized to produce the best agreement between the calculated and measured transient photocurrents. The resulting DOS has distinct features: A first peak at ~0.30 eV below Ec with an amplitude ~1017 eV1 cm3, a second small peak (or shoulder) at 0.450.50 eV below Ec with an amplitude 10141015 eV1 cm3, and deep states with an integral concentration 10111014 cm3 lying below 0.65 eV, whose exact distribution could not be resolved because of the limitations of the available experimental data. The density of states (DOS) distribution in the vicinity of the valence band mobility edge in vacuum coated a-Se films has been investigated by calculating the MC hole transient photocurrents at different temperatures, and also the dependence of the drift mobility on the temperature and field. The calculated TOF transient photocurrents were compared with experimental data published elsewhere. It is shown that, analogous to electron transport in a-Si:H, the DOS near Ev is a featureless, monotonically decreasing distribution in energy up to Ev + 0.4 eV, without the 0.28 eV peak near the valence band which was thought to control the hole drift mobility. Such a DOS was able to account for hole TOF data reported previously by several authors to date.

The "n-like layer" is important in multilayer layer amorphous selenium (a-Se) based Xray detector structures because it blocks the injection of holes from the positive electrode. The dark current in these devices is controlled primarily by hole injection,and the introduction of the n-like layer to block hole injection was a key development in the commercialization of a-Se X-ray detectors. An n-like a-Se layer is defined as a layer in which the electron range is much greater than the hole range, ¦Ìe¦Óe >> ¦Ìh¦Óh, where ¦Ó and ¦Ì are the lifetime and drift mobility of the charge carriers and the subscript e and h represent electrons and holes.<p> This thesis examines the effect of doping a-Se with Group II elements (in particular Mg) towards finding a better n-like layer ¨C that with relatively long electron range (drift mobility ¡Á lifetime) , trap limited hole transport and which is stable against crystallization. Conventional Time of Flight (TOF) and Interrupted Field Time of Flight (IFTOF) transient photoconductivity measurements were used to characterize the electron and hole transport in various Group II doped a-Se layers. The dependence of the electron and hole lifetimes and drift mobilities on the composition of the n-like layer was examined. The addition of Group II materials converts the a-Se starting material from p-like into n-like. It was found that increasing the concentration of Mg increases the electron range while limiting the hole range by modifying the population of deep traps. The addition of As further limits the hole transport but does not alter the electron range. The clear reproducibility of the thermal properties obtained from the Differential Scanning Calorimetry (DSC) implies that small amounts of Mg can be used as a suitable n-type dopant.

The electronic properties of amorphous materials are greatly affected by the density of localized states in the mobility gap of these materials. The exact shape of the density of states (DOS) distribution in amorphous selenium (a-Se) is still unresolved despite decades of research. One of the most commonly employed methods to investigate charge transport properties in high resistivity materials is time-of-flight (TOF) transient photoconductivity experiment. The TOF transient photoconductivity technique is used to measure the induced photocurrent in the external circuit when the sample is photoexcited. Information pertaining to carrier mobility and other carrier parameters are deduced from the shape of the photocurrent. The investigation of the charge transport phenomenon is well known to be a complicated task. Monte Carlo (MC) simulation method has become a standard method for carrier transport studies in amorphous materials. The purpose of this research work is to develop a Monte Carlo simulation model for charge transport in typical TOF transient photoconductivity experiment to investigate the DOS distribution in a-Se. The MC simulations were first performed for relatively simpler models for which theoretical and analytical solutions were available. The MC model developed here is based on simulating the drift of carriers resulting from photogeneration, subject to the influence of an applied electric field and multiple trapping events. The free drift time of photocarriers and their dwell time in the traps are stochastic in nature, in accordance with the probabilities of these events. Electron time-of-flight transient photocurrents were calculated in amorphous selenium as a function of the electric field. The distribution of localized states (DOS) in a-Se has been investigated by comparing the experimentally measured and calculated transient photocurrents. The analysis of multiple-trapping transport has been done by the discretization of a continuous DOS. The DOS distribution has been optimized to produce the best agreement between the calculated and measured transient photocurrents. The resulting DOS has distinct features: A first peak at ~0.30 eV below Ec with an amplitude ~1017 eV1 cm3, a second small peak (or shoulder) at 0.450.50 eV below Ec with an amplitude 10141015 eV1 cm3, and deep states with an integral concentration 10111014 cm3 lying below 0.65 eV, whose exact distribution could not be resolved because of the limitations of the available experimental data. The density of states (DOS) distribution in the vicinity of the valence band mobility edge in vacuum coated a-Se films has been investigated by calculating the MC hole transient photocurrents at different temperatures, and also the dependence of the drift mobility on the temperature and field. The calculated TOF transient photocurrents were compared with experimental data published elsewhere. It is shown that, analogous to electron transport in a-Si:H, the DOS near Ev is a featureless, monotonically decreasing distribution in energy up to Ev + 0.4 eV, without the 0.28 eV peak near the valence band which was thought to control the hole drift mobility. Such a DOS was able to account for hole TOF data reported previously by several authors to date.

The "n-like layer" is important in multilayer layer amorphous selenium (a-Se) based Xray detector structures because it blocks the injection of holes from the positive electrode. The dark current in these devices is controlled primarily by hole injection,and the introduction of the n-like layer to block hole injection was a key development in the commercialization of a-Se X-ray detectors. An n-like a-Se layer is defined as a layer in which the electron range is much greater than the hole range, ¦Ìe¦Óe >> ¦Ìh¦Óh, where ¦Ó and ¦Ì are the lifetime and drift mobility of the charge carriers and the subscript e and h represent electrons and holes.<p> This thesis examines the effect of doping a-Se with Group II elements (in particular Mg) towards finding a better n-like layer ¨C that with relatively long electron range (drift mobility ¡Á lifetime) , trap limited hole transport and which is stable against crystallization. Conventional Time of Flight (TOF) and Interrupted Field Time of Flight (IFTOF) transient photoconductivity measurements were used to characterize the electron and hole transport in various Group II doped a-Se layers. The dependence of the electron and hole lifetimes and drift mobilities on the composition of the n-like layer was examined. The addition of Group II materials converts the a-Se starting material from p-like into n-like. It was found that increasing the concentration of Mg increases the electron range while limiting the hole range by modifying the population of deep traps. The addition of As further limits the hole transport but does not alter the electron range. The clear reproducibility of the thermal properties obtained from the Differential Scanning Calorimetry (DSC) implies that small amounts of Mg can be used as a suitable n-type dopant.

An overview of synchrotrons and synchrotron radiation is presented, along with the theory and practical considerations behind several types of X-ray spectroscopy. The theory and practical considerations of density functional theory are also given, with direct reference to some specific software packages.<p> Some synchrotron-excited X-ray spectroscopy measurements and density functional theory calculations of selenium and arsenic-doped selenium films are then outlined. The physical structure of crystalline and amorphous selenium and the electronic structure of amorphous selenium are discussed and comparison is made to the experimental results.<p> A weak feature in the conduction band is identified as a "fingerprint" of the degree of crystallization in amorphous selenium from X-ray absorption measurements. Similarly, a weak feature corresponding to lone-pairs in the valence band is identified as a "fingerprint" of the arsenic concentration from X-ray emission measurements.<p> Finally a detailed model of the structure of amorphous selenium is explained, and compared to experiment. This model is tested both by direct calculations and by a reverse Monte Carlo approach. The implications of this model with respect to the structure of amorphous and arsenic-doped amorphous selenium are discussed. Calculations suggest that simply randomizing the arrangement of "perfect" trigonal selenium is unable to reproduce the measurements of amorphous selenium; a moderate variation in the bond angle of "perfect" trigonal selenium is also necessary.

Electron transport in vacuum deposited a-Se films has been investigated by Interrupted-Field Time-of-Flight (IFTOF) transient photoconductivity experiments to examine the effect of sample temperature (T) and applied electric field (F) on X-ray induced changes in the electron lifetime. Upon exposure to x-rays, the electron lifetime decreases. The decrease in normalized lifetime is almost linearly proportional to the absorbed dose, and is more significant at higher temperatures. Upon the cessation of x-ray irradiation, the lifetime recovers towards its equilibrium value through a structural relaxation process, and is characterized by a structural relaxation time. The structural relaxation time decreases with temperature in an Arrhenius fashion, and exhibits an activation energy that is roughly 1.4 eV. The structural relaxation time at room temperature (21 C) is 2 – 4 hrs whereas at 35 C, 6 – 10 mins. These measurements are important in characterizing the charge collection efficiency of a-Se based x-ray detectors, and its dependence on x-ray exposure and temperature. The results indicate that the rate of change of electron lifetime per unit exposure is less than 2%/Gy.

Amorphous selenium based digital radiography has attracted much attention because of selenium's high X-ray absorption and excellent charge transport properties, and the ability to be created thick (typically 100 to 1000 micron) uniform layers over a large area (typically 30 cm X 30 cm) at low processing temperatures (typically at around 50 degree C substrate temperature). In this work, the excess noise in amorphous selenium has been studied. A number of device parameters were altered to study the noise characteristics, such as the metal of the electrodes, bulk material composition, device volume, surface conditions and substrate temperature. All the samples had a transverse geometry with 20 to 200- micron thick layers of amorphous selenium electroded with metal at the top and at the bottom. Sample devices were fabricated by conventional vacuum deposition.<p> Noise power was measured over a limited bandwidth of 1 kHz. The fluctuations for one sample amounted to 1% of the bias current. The excess noise was mainly 1/<i>f</i> noise with the slope ranging from -0.77 to -1.4. Interpretation of the noise spectra was complicated due to the samples' highly non-linear I-V relation and long time transients.<p> The metals of the electrode clearly showed a large effect on both the magnitude and shape of the noise spectrum. Of the metals studied, aluminum produced the least normalized noise and platinum the most. The addition of arsenic caused a decrease in the normalized noise. An additional 0.2% (% wt.) arsenic decreased the 1/<i>f</i> noise magnitude by more than a decade, but did not change the slope. The addition of chlorine did not affect the noise magnitude. Amorphous selenium is quite vulnerable to stress and in particular, external mechanical stress causes crystallization. The surface of the sample was gently abraded, applying the least possible amount of stress to the selenium layer. A change in the surface condition before the top electrode was deposited showed that a roughened surface decreased the noise magnitude substantially. These results strongly indicate that the noise is controlled by the metal-semiconductor interface.<p> Noise characteristics in multilayered samples were examined. The p-i-n and n-i-p structures consisted of 200 micron i-layer with 2 to 6 micron p- and n-like layers. The noise fluctuation in the current are typical of 1/<i>f</i> noise showing a power-law spectrum with slopes between -0.9 to -1.1. These samples showed a substantial decrease in the noise power compared to single layer samples; the additional n-like and p-like layers acted as carrier sources so that the current was not controlled by the metal interface. Hence, the measurements are closer to the intrinsic noise of a-Se. After exposure to 14 R (Roentgen) of X-rays, the normalized noise decreased by a factor of 1.6 for the n-i-p structure.

Proteins are commonly found molecules in biological systems: our fingernails, hair, skin, blood, muscle, and eyes are all made of protein. Many diseases simply arise because a protein is not folded properly. Therefore, knowledge of protein structure is considered a prerequisite to understanding protein function and, by extension, a cornerstone for drug design and for the development of therapeutic agents. Protein crystallography is a tool that allows structural biologists to discern protein structures to the highest degree of detail possible in three dimensions. The recording of x-ray diffraction data from the protein crystal is a central part of protein crystallography. As such, an important challenge in protein crystallography research is to design x-ray detectors to accurately determine the structures of proteins. This research presents the design and evaluation of a solid-state large area at panel detector for protein crystallography based on an amorphous selenium (a-Se) x-ray sensitive photoconductor operating in avalanche mode integrated with an amorphous silicon (a-Si:H) charge storage and readout pixel. The advantages of the proposed detector over the existing imaging plate (IP) and charge coupled device (CCD) detectors are large area, high dynamic range coupled to single x-ray detection capability, fast readout, high spatial resolution, and inexpensive manufacturing process. The requirement of high dynamic range is crucial for protein crystallography since both weak and strong diffraction spots need to be imaged. The main disadvantage of a-Si:H thin film transistor (TFT) array is its high electronic noise which prohibits quantum noise limited operation for the weak diffraction spots. To overcome the problem, the x-ray to charge conversion gain of a-Se is increased by using its internal avalanche multiplication gain. Since the detector can be made approximately the same size as the diffraction pattern, it eliminates the need for image demagnification. The readout time of the detector is usually within the ms range, so it is appropriate for crystallographic application. The optimal detector parameters (such as, detector size, pixel size, thickness of a-Se layer), and operating parameters (such as, electric field across the a-Se layer) are determined based on the requirements for protein crystallography. A complete model of detective quantum efficiency (DQE) of the detector is developed to predict and optimize the performance of the detector. The performance of the detector is evaluated in terms of readout time (< 1 s), dynamic range (~10^5), and sensitivity (~ 1 x-ray photon), thus validating the detector's efficacy for protein crystallography. The design of an in-house a-Si:H TFT pixel array for integration with an avalanche a-Se layer is detailed. Results obtained using single pixel are promising and highlight the feasibility of a-Si:H pixels coupled with avalanche a-Se layer for protein crystallography application.

Proteins are commonly found molecules in biological systems: our fingernails, hair, skin, blood, muscle, and eyes are all made of protein. Many diseases simply arise because a protein is not folded properly. Therefore, knowledge of protein structure is considered a prerequisite to understanding protein function and, by extension, a cornerstone for drug design and for the development of therapeutic agents. Protein crystallography is a tool that allows structural biologists to discern protein structures to the highest degree of detail possible in three dimensions. The recording of x-ray diffraction data from the protein crystal is a central part of protein crystallography. As such, an important challenge in protein crystallography research is to design x-ray detectors to accurately determine the structures of proteins. This research presents the design and evaluation of a solid-state large area at panel detector for protein crystallography based on an amorphous selenium (a-Se) x-ray sensitive photoconductor operating in avalanche mode integrated with an amorphous silicon (a-Si:H) charge storage and readout pixel. The advantages of the proposed detector over the existing imaging plate (IP) and charge coupled device (CCD) detectors are large area, high dynamic range coupled to single x-ray detection capability, fast readout, high spatial resolution, and inexpensive manufacturing process. The requirement of high dynamic range is crucial for protein crystallography since both weak and strong diffraction spots need to be imaged. The main disadvantage of a-Si:H thin film transistor (TFT) array is its high electronic noise which prohibits quantum noise limited operation for the weak diffraction spots. To overcome the problem, the x-ray to charge conversion gain of a-Se is increased by using its internal avalanche multiplication gain. Since the detector can be made approximately the same size as the diffraction pattern, it eliminates the need for image demagnification. The readout time of the detector is usually within the ms range, so it is appropriate for crystallographic application. The optimal detector parameters (such as, detector size, pixel size, thickness of a-Se layer), and operating parameters (such as, electric field across the a-Se layer) are determined based on the requirements for protein crystallography. A complete model of detective quantum efficiency (DQE) of the detector is developed to predict and optimize the performance of the detector. The performance of the detector is evaluated in terms of readout time (< 1 s), dynamic range (~10^5), and sensitivity (~ 1 x-ray photon), thus validating the detector's efficacy for protein crystallography. The design of an in-house a-Si:H TFT pixel array for integration with an avalanche a-Se layer is detailed. Results obtained using single pixel are promising and highlight the feasibility of a-Si:H pixels coupled with avalanche a-Se layer for protein crystallography application.