The latest generation of solid-state arrays, including charge-coupled devices and charge injection devices, holds great promise for revolutionizing spectrochemical methods of analysis. However, successful application of these new technologies will require more than substituting new detectors for old. Different optical system configurations and operational approaches are considered and their impact on a number of spectroscopic techniques are discussed. Specific examples for efficiently employing x-y array detectors, including aberration corrected holographic grating, crossed interferometric dispersive and echelle spectrometers, are discussed. Present and future trends for applying array detector technology to advanced spectroscopic analysis are considered.
A detector with a 114 mm aperture, based on a charge-coupled device (CCD), has been designed for x-ray diffraction studies in protein crystallography. The detector was tested on a beamline of the National Synchrotron Light Source at Brookhaven National Laboratory with a beam intensity greater than 109 x-ray photons/s. A fiber-optic taper, an image intensifier, and a lens demagnify, intensify, and focus the image onto a CCD having 512 X 512 pixels. A detective quantum efficiency (DQE) of 0.36 was obtained by evaluating the statistical uncertainty in the detector output. The dynamic range of a 4 X 4 pixel resolution element, comparable in size to a diffraction peak, was 10 4. The point-spread function shows FWHM resolution of approximately 1 pixel, where a pixel on the detector face is 160 micrometers . A complete data set, consisting of forty-five 1 degree(s) rotation frames, was obtained in just 36 s of x-ray exposure to a crystal of chicken egg-white lysozyme. In a separate experiment, a lysozyme data set consisting of 495 0.1 degree(s) frames, was processed by the MADNES data reduction program, yielding symmetry R-factors for the data of 3.2- 3.5. Diffraction images from crystals of the myosin S1 head (a equals 275 angstroms) were also recorded. The Bragg spots, only 5 pixels apart, were resolved but were not sufficiently separated to process these data. Changes in the detector design which will improve the DQE and spatial resolution are outlined. The overall performance showed that this type of detector is well suited for x-ray scattering investigations with synchrotron sources.
Atherosclerosis is an arterial disorder characterized by the development of arterial plaques which reduce the distensibility of the artery and obstruct blood flow. Little is known about the mechanisms which initiate the plaques and cause them to grow; however, it is generally agreed that hemodynamic factors are associated with the development of atherosclerosis. To study this disease it is essential to know not only the geometry of the arterial lumen but also the shape of the intimal surface in order to assess the importance of hemodynamic effects. The authors constructed a table-top volume CT scanner with high resolution in all 3 dimensions, which can be used to analyze human arterial specimens in vitro. This system uses an x-ray image intensifier optically coupled to a TDI CCD sensor to obtain low-noise, low-scatter projection digital radiographs from many angles. A slot beam of radiation is scanned across the sample to reduce the detection of scattered radiation without causing excess x-ray tube heating. Objects to be imaged are placed on a computer-controlled stage and projections are obtained as the specimen is rotated through 180 degree(s). CT reconstructions of the resulting data produces volume images with 0.12 X 0.12 X 0.15 mm3 volume resolution.
High-resolution digital medical images are produced by solid-state linear x- ray detectors. The detectors analyze the x-ray images line by line. They are associated with a mechanical scanning of the image. Multilinear detectors sense several lines at the same time and operate a TDI (time delay and integration) digital process. The TDI process is useful to increase the image signal-to-noise ratio. Linear or multilinear detectors make imaging systems possible which have a very wide image dynamic range, lack of geometrical distortion, and direct digital output signal. The output signal can be easily interfaced with any kind of digital processing system.
This paper describes the results of coupling a Tektronix TK512 CCD imager to a Technical Instruments K2Bio Nipkow disk real-time confocal microscope. The characteristics of the selected TK512 CCD imager are as follows: the chip is a thinned, back-illuminated device which has an anti-reflection coating. The quantum efficiency of the device is 86.4 at 400 nm. The CCD imager is a TK512, silicon CCD which is sensitive from the UV to the near-IR. The sensor is a full-frame area imager, with CCDs fabricated using a buried channel, three-level polysilicon gate process with very high charge-transfer efficiency ($2 0.9999) and low dark current. Low noise on-chip amplifiers provide and interface to external preamplifiers with readout noise typically
Particle-induced displacement damage degrades performance parameters in state-of-the-art solid state imagers due to charge transfer inefficiency, increased dark current, and dark current spikes. This paper reviews some recent radiation damage measurements and discusses imager degradation in the context of a general understanding of how displacement damage alters semiconductor properties. An approach to predict the response of a sensor to a given space environment and shielding configuration is presented and the authors briefly discuss other displacement damage concerns such as secondary neutron production in shielding. Using an example based on limited experimental input, the authors demonstrate how higher (10 MeV) energy protons are responsible for over 90 of the damage in heavily shielded imager applications.
Recent analytical and experimental work has provided new insights into the production of damage sites in silicon Charge-Coupled Devices (CCDs) by energetic particles and into the effects of these sites on CCD performance. An approximate correlation is presented between experimental results and a prediction of proton-induced displacement damage, and possible explanations for remaining inconsistencies are discussed. As a consequence of this agreement, it is now possible to predict the effect of complicated space proton environments upon CCD charge transfer efficiency and other CCD performance parameters. This prediction requires evaluation of the damage resulting from only a small number (
This paper presents preliminary results on the performance of n-channel, backside-thinned charge-coupled devices (CCDs) as electron-bombarded-semiconductor (EBS) imagers for the detection of 1-10 keV electrons. The devices exhibit average EBS gains ranging from approximately 50 at 1 keV to 1600 at 10 keV. Device radiation tolerance has been investigated by exposing normally-clocked devices to 6 keV electron doses up to 0.01 Coulombs/cm2. Room temperature pre- and post-irradiation results are presented for these key device parameters: full well capacity, dark current, and charge transfer efficiency (CTE). At the maximum dose of 0.01 Coulombs/cm2, full well capacity decreases 9 from an initial value of 680,000 e-, and dark current increases from 2 to approximately 50 nA/cm2. There are no measurable changes in large signal CTE up to the maximum dose. Radiation damage at energies other than 6 keV is estimated by measurement of the x-ray generation efficiency of silicon as a function of electron energy. Device stability after temperature cycling has been studied by subjecting packaged devices to vacuum bakes of 24 hours at 300 degree(s)C. Full well, CTE, EBS gain, and output amplifier performance are unchanged after the extended temperature cycle, while dark current decreases slightly by 15. In summary, these initial results indicate that the CCD can function as both an efficient and robust electron imager.
The charge-coupled device (CCD) camera of the Soft X-ray Telescope (SXT) for the Japanese Solar-A Mission utilizes a 1024 X 1024 virtual phase CCD manufactured by Texas Instruments in Japan. This sensor will be subject to radiation in the form of trapped protons from the earth''s radiation belts and soft x-rays (0.2-4 keV) in the solar image. Proton damage produces ''dark spikes'' or pixels of enhanced dark current. This can be characterized in terms of the average increase in dark current as a function of proton fluence and predicted through proton transfer calculations. During the preparation of this camera it has been discovered that exposure to soft x-rays creates ''permanent'' ionization damage in the gate insulator, resulting in flat-band shift, dark current increase, loss of charge transfer efficiency, and, ultimately, total unpinning of the sensor. It has been found that ultra-violet, and to a lesser degree, visible-light flooding photo-emits free electrons into the gate oxide which ''anneals'' the damage, restoring proper operation of the CCD.
A simplified model of the back surface of the CCD is discussed. The model incorporates parameters which account for the important features of the back surface: a surface recombination velocity, an electric field which can assist in or oppose the collection of signal charge, and a field free region. Calculations using the model equations are presented to illustrate the effect of these parameters on the distribution of excess carriers and on the resulting quantum efficiency. The results indicate that only moderate fields are required to achieve high quantum efficiencies even at short wavelengths.
An analytical model has been developed for predicting the spectral response of thinned, p+-doped back-illuminated charge-coupled device (CCD) imagers. The device is divided into two regions: a thin, uniformly doped p+ layer used to passivate the illuminated back surface from external electrical effects, and a p- region that extends from the p+ region across the approximately 10-micrometers thickness of the device to the potential well in the buried channel. The one-dimensional steady-state continuity equation for low-injection conditions has been solved analytically for the surface p+ region, which is characterized by electron diffusion length and coefficients appropriate for the doping level and a surface recombination velocity Sn that represents the loss of photoelectrons at the surface. All photoelectrons generated in the p- region are assumed to be collected in the buried channel because of the long diffusion length and the presence of a field sweeping the carriers into the CCD channel. The effect of multiple internal reflections on photoabsorption at long wavelengths is included. The quantum efficiency of this device is calculated as a function of the depth and recombination velocity of the p+ surface layer, using Sn as the only independent fitting parameter, and matches experimental results well over the wavelength range from 360 to 1100 nm.
In order to support rigorous requirements for tracking, surveillance, and astronomical applications, SAIC pursues megapixel focal plane development programs which require high performance in areas such as quantum efficiency, dynamic range, charge transfer efficiency, and total noise. To provide CCDs with these properties, special backside thinning, post- processing, anti-reflection coating, and high-speed readout capabilities have been developed in-house at SAIC. The authors discuss the thinning efforts to date on SAIC 1024 X 1024 arrays, including flattening mounts and anti-reflection coatings. Test and characterization results are presented for various chips which have been passivated using either flashgate or p+ substrate remnant. Finally, preliminary results in two related areas are discussed: initial thinning efforts on SAIC/LORAL 18 micrometers 2048 X 2048 CCDs, and electron imaging with thinned chips in a Digicon tube.
The authors use charge-coupled devices (CCDs) from a lot of Loral wafers containing 3072 X 1024 and 800 X 1200 pixel detectors. These CCDs were specifically designed for astronomical spectrographic applications with the intention of thinning them at Steward Observatory for backside illumination. The thinning procedure and how it relates to the packaging method developed to keep the devices flat for use in fast optical beams are described. Initial results with a new oxidation procedure to allow better backside charging is also described. This oxidation method is especially effective with biased gate-charging techniques which apply a voltage directly over the backside oxide.
A high-speed 512 X 512 charge injection device with selectable one to four video ports has been developed, fabricated, and tested beyond the designed speed of operation. The imager has four independently controllable video ports allowing for all possible combinations. This is accomplished by having each port hard wired to one out of every four rows sequentially. Each port is selected via a multiplexer in the sequence desired. The horizontal scanner was designed to operate up to 30 MHz. The device was tested at the wafer level to 42 Mhz element rate per port. This element rate allows a maximum of 168 MHz element rate with four ports operating in parallel.
A third-generation SAGE instrument is about to be designed as part of the NASA Earth Observational System. Previous instruments have used individual diodes as detectors. The new instrument will use a custom design CCD to dramatically enhance the study of the gas and aerosol components of the upper atmosphere. The CCD is a 3 X 400 imaging array that has a single serial register and an exposure control drain. It will be used at the focal plane of a spectrometer covering the spectral range from 288 nm to 1.02 micrometers .
Results obtained in fabricating and testing of large CCD image sensors are reported. The emphasis is on high quantum efficiency, excellent charge transfer efficiency at low signal level, large pixel count, low readout noise, and very low dark current. The focus is on the use of the devices for optical astronomy where these parameters are most important. Test results for CCDs fabricated by Ford Aerospace and by EG Experiments to demonstrate the feasibility of a reproducible biased-gate using transparent indium tin oxide as a conducting layer over a silicon oxide insulating layer are discussed. Quantum efficiency of bias-gate thinned CCDs is compared with results obtained from a phosphor-coated front- illuminated CCD.
CIDTEC has developed a PC-based instrumentation camera incorporating a preamplifier per row CID imager and a microprocessor/LCA camera controller. The camera takes advantage of CID X-Y addressability to randomly read individual pixels and potentially overlapping pixel subsets in true nondestructive (NDRO) as well as destructive readout modes. Using an oxy- nitride fabricated CID and the NDRO readout technique, pixel full well and noise levels of approximately 1*106 and 40 electrons, respectively, were measured. Data taken from test structures indicates noise levels (which appear to be 1/f limited) can be reduced by a factor of two by eliminating the nitride under the preamplifier gate. Due to software programmability, versatile readout capabilities, wide dynamic range, and extended UV/IR capability, this camera appears to be ideally suited for use in spectroscopy and other scientific applications.
The two-dimensional electron gas charge-coupled device (2DEG-CCD) structure is an outgrowth of recent advances in 2DEG-FET structures for digital logic circuitry and microwave devices. The 2DEG-FET structures, which are known by severalacronyms such as HEMT, SDHT, TEGFET, HIGFET, and MODFET, utilize the abrupt heterointerface between two semiconductor materials and consequent conduction band discontinuity to confine electrons . Due to confinement in the direction perpendicular to the interface, the resultant electron distribution is known as a two-dimensional electron gas. Because of the confinement dimensions and the low electron effective mass usually associated with group 111-V materials, quantum mechanics plays an important role in broadening the spatial electron distribution and defining allowed energy states. However, quantum effects typically play a minor role in understanding device behavior at temperatures above 77 K. 2DEG-FET devices have several attributes which include very high mobility of the channel charge (typically in excess of 5,000 cm2/V-sec at room temperature), high transconductance, and low voltage swing requirements. The 2DEG structure is attractive for CCD applications for several reasons. First, the high low-field mobility and use of a semi-insulating substrate suggest very high speed device operation. Second, the charge-handlin capabilty of the 2DEG-CCD structure is large compared with MESFET-type CCDs, and can exceed 1x101 carriers/cm2. Third, the lattice-matched heterointerface has a potentially lower interface trap density than the intrinsically mismatched silicon-silicon dioxide interface, as well as improved radiation hardness. Finally, the useful operating temperature of the 2DEG-CCD is expected to be lower than that of the silicon CCD. Other features of the 2DEG-CCD include fabrication compatibility with high-performance, low noise 2DEG-FET output circuitry and an anti-blooming gate structure.
An update on research activities at Columbia University in the area of focal-plane image processing is presented. Two thrust areas have been pursued: image reorganization for image compression and image half-toning. The image reorganization processor is an integration of a 256 X 256 frame-transfer CCD imager with CCD-based circuitry for pixel data reorganization to enable difference encoding for hierarchical image compression. The reorganization circuitry occupies 2 of the total chip area and is performed using three parallel-serial-parallel (SP3) registers, a pixel resequencing block, and a sampling block for differential output. The chip has achieved a CTE of 0.99994 in this new SP3 architecture, at an output rate of 83 X 103 pixels/sec. (0.9996 at 2 X 106 pixels/sec) and an overall output amplifier sensitivity of 3.2 (mu) V/electron. The half-toning chip design has been described previously, and consists of a 256 X 256 frame transfer imager, a pipeline register, and comparator circuit. Functional testing of these elements is reported at this time.
A large-format CCD imager to be used in a 2 X 2 mosaic array has been designed and fabricated. Each quadrant is an independent imager of 2048 X 2048 15 micrometers pixels, designed to be edge-butted on two sides. After sawing and mounting the individual dice in custom, buttable packages, the authors assembled a 4096 X 4096 mosaic array measuring more than 61 mm on a side with just 400 microns dead space between imaging areas. With this buttable package design, each quadrant of the mosaic can be separately tested, optimized, and, if necessary, replaced. Also described are smaller 2688 X 512 15 micrometers pixel CCD imagers designed for spectrographic applications that were fabricated using space in the chords of the same 100 mm silicon wafer containing the 2048 X 2048 edge-buttable devices.
This paper describes the design and performance of a large format, 1280(H) X 1024(V) pixel, full-frame CCD image sensor. A lateral-overflow drain (LOD) is incorporated for antiblooming control while providing high responsivity and photographic speed. The use of an LOD also provides high optical overload protection, and an extremely linear photoresponse compared to that of its vertical-overflow drain (VOD) counterpart owing to the near-unity nonideality factor of the LOD structure. Optical overloads in excess of 43,000 times saturation have been measured without blooming, and the photoresponse nonlinearity at saturation is less than 1. The LOD structure is also much easier to fabricate than a typical VOD structure and is much less sensitive to processing and drain bias variations. The device''s two-phase CCD pixels measure 16 micrometers on a side. The LOD structure occupies roughly 30 of this area using conservative design rules. Since much of this area is required for overlap of the various layers of the color-filter array (CFA), the incorporation of the LOD does not result in any significant additional loss of fill factor. To further enhance the detector''s sensitivity, a transparent gate electrode process has been implemented which replaces one of the CCD''s polysilicon electrodes with indium-tin oxide (ITO). The ITO gate electrode increases the quantum efficiency of the detector from 1.5 at a wavelength of 400 nm. The responsivity of the transparent gate electrode structure over the visible spectrum is 210 nA/lux. This is more than 38 higher than that of the all-polysilicon gate structure.
High-resolution solid-state image sensors have become readily available due to continuing advances in VLSI technology. The authors have developed a 1.6 megapixel full-frame CCD image sensor (KAF-1600) with a 3:2 aspect ratio to meet industrial and scientific applications. The high-resolution sensor, measuring 1.55 cm X 1.0 cm, consists of 1536 X 1024 pixels. The pixel size is 9 microns X 9 microns. The sensor has a single-readout register with the capability of a 2-to-1 line charge summing. The architecture and results of the megapixel image sensor are presented.
Development of a 20482 CCD for a second-generation space telescope instrument has produced some very encouraging devices. The first experimental lot of 10 devices have very few defects, dark currents of less than 12 electrons/pixel/hour at -80 degree(s), readout noise levels of less than 4 electrons rms, and excellent charge transfer efficiency at signal levels of less than 10 electrons.
The Tektronix CCD Manufacturing Group has applied their thinning back surface enhancement and anti-reflection coating processes to produce a 1024 by 1024 chargecoupled device imager with high quantum efficiency (QE) from 350 to 1 100 nm. The TK1O24AB device designed for scientific imaging applications features low noise wide dynamic range excellent charge transfer efficiency and low dark current. The quad-output architecture permits the simultaneous readout of each quarter of the device reducing the time to read out the CCD to that of a 5 12 by 512 device. This paper summarizes the test results from several lots of TK 1024AB runs. The subjects covered include: QE the on-chip amplifier characteristics dark current measurement CTE and characterization of various defects. The paper will also describe the test hardware and procedures used to evaluate the performance of the devices.
A technique for improving the performance oflarge area high resolution Charge-Coupled Device (CCD) imagers will be described. Adding an additional doped channel down the center of a CCD register provides for charge confinement. This leads to improved charge transfer efficiency and resistance to radiation damage. Two dimenslonal theoretical analysis will be shown along with measured device performance.
This device has a number of potential applications. It may be used in a direct view mode as a thermal night sight. It may be integrated with a fiber-optic faceplate visible charge-couple device (CCD) and then used as a TV-type thermal imager. Alignment of the sensor to the CCD is not very critical. In a slightly modified version, the output image may be a spatial light modulator serving as input to an optical processor. Examples of the above implementations are given. The converter can be tailored to accept either the 3 to 5 micrometers or the 8 to 14 micrometers infrared bands. The quantum efficiency is on the order of 30, but the low noise performance and starting mode of operation readily offset this. The unique design greatly reduces the impact of structural defects. Pixel-to-pixel uniformity of the device is very good. The simple design offers a potential for high-resolution large-area imagers. The paper discusses the supporting theory and device design. The manufacturing process, the results to date, and the performance are also discussed.
The paper describes the system consisting of a highly sensitive remote CCD camera head and a processing/control module. Specially designed virtual phase CCD has staggered pixel arrangement ensuring the vertical and horizontal resolution about 250 TV lines with the chip size of 2, 8 X 3 mm2. The system utilizes successive field color encoding (field sequential color method). The signal processing/control module comprises a real-time digital interpolator and video-frame memory. The image processor hardware and software provide image acquisition, treatment, distribution, and other functions.
The CCD imaging system for use in ground telescopes was manufactured and tested. The two- level controller provides great flexibility in system operation. The results of system performance at the Crimea Observatory are presented.
A family of VP CCD image sensors for different industrial and scientific applications was designed, fabricated, and tested. All of them share the common concept of a 2.5-phase photosensitive cell, combining the advantages of known 1.5-phase VP devices (reduced dark current and increased quantum efficiency) with simpler fabrication process and extended functions.