This paper reviews the progress in the development of infrared image sensors with Schottky- barrier detectors (SBDs). Schottky-barrier focal plane arrays (FPAs) are infrared imagers that are fabricated by a well established silicon VLSI process, therefore, at the present time they represent the most advanced technology for large-area high-density focal plane arrays for many SWIR (1 to 3 micrometers ) and MWIR (3 to 5 micrometers ) applications. SBD line sensing arrays with up to 4096 X 4 elements and 2048 X 16-TDI elements were developed and SBD staring (area) arrays with up to 1040 X 1040 elements have been reported. PtSi SBDs represent the most established SBD technology for applications in the SWIR and MWIR bands. At an operating temperature of 77 K, the dark current density of PtSi SBDs is in the range of 1.0 to 4.0 nA/cm². Pd₂Si SBDs were developed for operation with passive cooling at 120 K in the SWIR band. IrSi SBDs have also been investigated to extend the application of Schottky-barrier focal plane arrays (FPAs) into the LWIR (8 to 10 micrometers ) spectral range. Because of very low readout noise, the IR-CCD imagers with PtSi SBDs which have quantum efficiency of .5% to 1% at 4.0 micrometers are capable of 300 K thermal imaging with a noise equivalent temperature (NE(Delta) T) from 0.033 to 0.15 K for operation at 30 frames/s and f/1.0 to f/2.8 optics.

Noise equivalent temperature difference (NE(Delta) T) values in a PtSi Schottky-barrier (SB) infrared CCD image sensor are calculated with variations in spectral response, device operating temperature, and internal radiation intensity generated from a lens barrel, using measured values from a recently developed 648 X 487 pixel SB infrared CCD image sensor. The barrier height is allowed to vary, whereas it was fixed before. NE(Delta) T values were calculated as a function of barrier height and were found to have a minimum. The optimum barrier height depends on quantum efficiency coefficient, device operating temperature, and internal radiation intensity.

A 512 pixel truly linear infrared (IR) charge coupled device (CCD) with Schottky barrier sensors and buttable edges has been developed, incorporating three different silicides working in the front-side illumination mode. The main differences between these silicides are the cut- off wavelength and the operating temperature. CoSi₂ and NiSi show a cut-off wavelength of about 2.8 micrometers allowing an operating temperature of 150 K and passive cooling, whereas PtSi has a cut-off wavelength of about 6 micrometers working at 77 K. All three devices present a good photoresponse uniformity along the sensor row and also a good noise behavior.

We have improved the performance of a 512 X 512 element PtSi Schottky-barrier infrared image sensor (512 X 512 IRCSD) by increasing the fill factor and saturation signal level. The sensor consists of 26 micrometers X 20 micrometers pixels in a 512 X 512 array format and has a large fill factor of 71% obtained with 1.2 micrometers minimum design rules and the charge sweep device (CSD) readout architecture. The improved 512 X 512 IRCSD was designed to be operated in either a field or frame integration interlace mode. The saturation signal level of the CSD imager is determined by the storage capacity of the Schottky-barrier detector (SBD). We optimized the structure and impurity concentration of the isolation region of the SBD in order to increase the large storage capacity. For an SBD reset voltage of 4 V, a saturation signal level and differential temperature response at 300 K were 2.9 X 10⁶ electrons and 3.2 X 10⁴ electrons/K, respectively. The noise equivalent temperature difference (NETD) at 300 K is estimated as 0.033 K with an f/1.2 cold shield.

The Hughes Aircraft Company CRC-365 is a hybrid infrared focal plane array that features a 256 X 256 array of platinum silicide (PtSi) photodiodes. The device was tested in a laboratory environment to determine its performance limits, in the context of MWIR imaging. Test results included the dc transfer function of the multiplexer, power dissipation, dark current, thermal barrier height, Schottky barrier height, blackbody response, saturation charge, signal contrast, Schottky quantum yield, temporal noise, NETD, response nonuniformity, temporal drift, and MRTD. Overall, performance was found to be excellent. Using two-point nonuniformity correction, background-noise-limited photodetection was achieved for a 40 degree target temperature range. Notable problem areas included bandwidth limitations and electrical crosstalk.

The development of a Platinum Silicide (PtSi) array for use in ground-based astronomy has been the goal of a joint effort between the National Optical Astronomy Observatories (NOAO) and the Hughes Technology Center (HTC). This has been accomplished by the introduction of the SWIR PtSi hybrid array for astronomy. The resulting array has an optical cavity tuned to enhance the internal quantum efficiency at 1.7 microns and the use of hybrid technology produces an array with near 100% effective fill factor. Dark currents are much less than 10 electrons sec^-1 and the read noise is 55 - 60 e rms. Although the quantum efficiency is not as high as the exotic materials, the high uniformity, stability, and low cost make this an array worth considering for many astronomy applications.

This paper describes the construction of a high-speed PtSi infrared charge-coupled device (IRCCD) camera system funded by the U.S. Army Tank-Automotive Command (TACOM) for recording periodic events in the infrared. A brief discussion of PtSi infrared detection, and an analysis of expected signal contrast and NE(Delta) T of the device are presented. Integration times to less than 1 ms and the ability to record frames based on an external trigger allow this setup to achieve effective frame rates of over 1000 frames per second. The use of a personal computer (PC) as a controller for CCD clocking and video capture results in a flexible camera system design.

The design and performance of a 640 (H) X 480 (V) element PtSi Schottky-barrier infrared image sensor employing a low-noise MOS X-Y addressable readout multiplexer and on-chip low-noise output amplifier is described. The imager achieves an NEDT equals 0.10 K at 30 Hz frame rates with f/1.5 optics (300 K background). The MOS design provides a measured saturation level of 1.5 X 10⁶ electrons (5 V bias) and a noise floor of 300 rms electrons per pixel. A multiplexed horizontal/vertical input address port and on-chip decoding is used to load scan data into CMOS horizontal and vertical scanning registers. This allows random access to any sub-frame in the 640 X 480 element focal plane array. By changing the digital pattern applied to the vertical scan register, the FPA can be operated in either an interlaced or non-interlaced format, and the integration time may be varied over a wide range (60 microsecond(s) to > 30 ms, for RS 170 operation) resulting in `electronic shutter' variable exposure control. The pixel size of 24 micrometers X 24 micrometers results in a fill factor of 38% for 1.5 micrometers process design rules. The overall die size for the IR imager is 13.7 mm X 17.2 mm. All digital inputs to the chip are TTL compatible and include ESD protection.

For years, thermal model developers have promoted the approach of using simulated data validated by measurements as the best method of analyzing all aspects of a thermal signature -- target, background, atmosphere, and imager. Recent advances in high speed CPUs and high performance graphic workstations have allowed for improved proficiency in these thermal signature models and has helped convince `the user' that the modeled approach is viable. Now, targets and backgrounds can be modeled more quickly and with better realism and imagers of all types can be simulated in practical runtimes. These improved capabilities increase the temptation to look to modeling as the panacea for all difficulties encountered in infrared imaging as sensor designers, smart weapons designers, and vehicle concept designers all realize the cost and practical limitations of using measured data only. This paper examines the benefits and pitfalls experienced by U.S. Army modelers particularly in the target and background modeling area and provides some guidelines for future modeling directions.

Focal plane applications demand a high degree of linearity in the detector response function (voltage out vs. photon flux in). For calibrating radiometric data and for correcting channel-to- channel nonuniformities in nonradiometric data, the response function of the focal plane must be correctable to within 0.1%. This specification requires either significant improvement in focal plane technologies or in methods to correct for it. Two-point calibration is often used to correct for nonuniformities across a focal plane array (FPA), as well as for calibration. Because the input-output curves of FPA channels are nonlinear, two-point calibration produces a systematic calibration error as a function of flux, and the channel-to-channel variations of this calibration error leave a significant post-correction nonuniformity. A simple physical model of the detector nonlinearity is used to illustrate these points. The sensor degradation due to nonlinearities is predicted from the pixel-to-pixel variations in nonlinearity after two-point correction. Variations of only 0.2% can result in significant degradations of the array D^*.

Research of vertebrate and invertebrate vision systems has revealed them to be remarkable assemblies of simple cells performing collectively various image processing and analysis tasks. Among these are counted edge enhancement, noise suppression, dynamic range compression, and motion and object orientation detection. These functions are achieved due to the massively parallel structure of these systems and appropriate non-linear inter-cell interactions, among them lateral inhibition. The high degree of connectivity existent in the vertebrate retina is currently beyond reach of integrated implementations; however, even its approximations applied to focal plane arrays can result in enhanced and more sophisticated performance. These approximations are discussed mathematically by means of methods developed for analysis of neural networks. A photoreceptor lateral interaction network, Grossberg's shunting neural network, and a novel modified version of the latter are compared in their effect on spatial nonuniformity noise and edge enhancement. These two qualities are of special interest in the case of infrared imaging. The modified shunting network combines an adaptive lateral signal spread amongst photodetectors with non-linear, multiplicative lateral inhibition. The first effect serves to reduce the effects of spatial noise, while the second, by its differentiating nature, removes low spatial frequencies and enhances high spatial frequency components inherent to the image.

This paper describes the computer simulation of a hybrid infrared focal plane array (IRFPA) readout multiplexer. The device under study is a switched field effect transistor (SWIFET) readout multiplexer that utilizes a source follower per detector (SFD) unit cell amplifier. The objective of this study is to determine if the limiting component of the IRFPA nonlinearity is the infrared detector array or the readout multiplexer. This study was performed by developing a computer simulation within which the material, fabrication, and operational parameters of the SFD readout multiplexer could be varied. The computer simulation is able to predict experimental data from a SFD readout multiplexer thereby providing validation that the device models used in the computer simulation are correct. The most significant result determined by this study is that the nonlinearity associated with the SFD readout multiplexer is approximately an order of magnitude lower than the nonlinearity from a state-of-the-art PtSi infrared detector array. The nonlinearity in the SFD readout multiplexer is dominated by the body-effect in the source follower amplifiers that comprise the SFD readout multiplexer.

Digital signal processing (DSP) on focal plane array (FPA) is attractive for large focal planes for reducing the amount of data output to no more than that which is of interest and for simplifying the IO to a simple digital bus. However, semiconductor circuits dissipate too much power for use on the FPA, overloading the cooling system capability, and requiring too much system cooling power for many applications. On the other hand, superconductive circuits (SC) offer an attractive alternative because they dissipate only about 0.1% the power of semiconductor circuits for an equivalent circuit function. SC 12 bit A/D converter and SC shift registers demonstrated in Nb at 4 K are readily convertible to NbN at 10 K. As the development of active devices in YBa₂Cu₃O₇ matures, these and a full complement of logic devices should be possible as high as 80 K. Scene signal and detector leakage current considerations require that long wavelength IR/FPA using quantum detectors must operate at cryogenic temperatures (< 80 K). It is no significant burden to use SC circuits at these cryogenic temperatures. SC circuits operate so much faster than semiconductor circuits and SC memory circuits are so relatively limited in size that DSP architecture has to be restructured. The derived benefit in terms of system capability will warrant this investment.

A long wavelength infrared imaging system is under development at Loral Infrared & Imaging Systems. The imager features a solid state pyroelectric focal plane array sensitive to 8 - 14 micrometers radiation operating at ambient temperature; it needs no cooling or temperature control to obtain performance comparable to conventional cooled FLIR systems. The focal plane array is fabricated as a hybrid structure with a fully reticulated lithium tantalate detector array bump mounted to a CMOS multiplexer. The optical, thermal, and electrical design of the focal plane array is described in detail. The prototype imaging system uses a focal plane array of 192 X 128 pixels on 50 micrometers centers. The next version will use a focal plane with approximately 330 X 240 pixels for full NTSC (U.S. standard television) compatibility. The design of the prototype imaging system is described. The predicted noise equivalent temperature difference (NETD) for these systems is less than 0.1 degree(s)C with f/1 optics at a 30 Hz display rate.

Starting from a characterization of the used pyroelectric materials LiNbO₃ and LiTaO₃ the paper describes the hybrid arrangement and essential properties of realized single- element detectors and arrays. It shows that the design and the production technique of pyroelectric chips have a strong influence on the thermal and spatial resolution of the detectors. This technique includes both the thinning process and reticulation of the chips, for which ion beam milling as a universal method was optimized and used. Single-element detectors with extremely thin, self-supporting LiTaO₃ chips (dp < 2 micrometers ) were produced. With a responsive element of 2 X 2 mm² in area, they have a specific detectivity D^* (500 K; 10 Hz; 25 degree(s)C) > 1 X 10⁹ cmHz^1/2W^-1. Linear arrays with 128 responsive elements of 90 X 100 micrometers ² element size, 100 micrometers pitch, integrated readout circuit, and coated germanium window have a noise equivalent power (NEP) (500 K; 40 Hz; 25 degree(s)C) of 4 nW. The modulation transfer function MTF_S (40 Hz; 31 p/mm; 25 degree(s)C) is 0.15 for pyroelectric chips without isolating grooves and was increased up to 0.45 by means of ion- beam milling of 10 micrometers wide isolating grooves between the responsive elements. First results of two-dimensional arrays with 128 X 128 elements, of 50 micrometers pitch and integrated CCD-readout circuit are presented.

In an attempt to combine the properties of high temperature superconductors with the high thermal conductivity and low specific heat of diamond, we have explored the deposition of in- situ YBa₂Cu₃O_7-(delta ) (YBCO) superconducting films on polycrystalline diamond thin films. We demonstrate for the first time superconducting YBCO films on diamond employing multiple layer buffer layer systems. Three different composite buffer layer systems were explored for this purpose: (1) Diamond/Zr/YSZ/YBCO, (2) Diamond/Si₃N₄/YSZ/YBCO, and (3) Diamond/SiO₂/YSZ/YBCO. Adherent thin Zr films were deposited by dc sputtering on the diamond films at 450 to 820 degree(s)C. The yttria stabilized zirconia (YSZ) was deposited by reactive RF sputtering at 680 to 750 degree(s)C. The Si₃N₄ and SiO₂ were also deposited by on-axis RF sputtering at 400 to 700 degree(s)C. YBCO films were grown on the buffer layers by off-axis RF sputtering at substrate temperatures between 690 degree(s)C and 750 degree(s)C. In all cases, the as-deposited YBCO films were superconducting above 77 K. This demonstration enables the fabrication of low heat capacity, fast response time bolometric far IR detectors and paves the way for the use of HTSC as a high frequency interconnect metallization on thick diamond film based multichip modules.

Electrolyte electroreflectance (EER) is used to investigate the near surface properties of HgCdTe. We used the technique of EER coupled with electrochemical etching. Annealing of B ion-implanted HgCdTe with ZnS encapsulation and anode sulfide film is studied. The results of fitting parameters show that a highly disordered surface layer exists after implantation, and an obvious recovery occurs after the sample is annealed at 200 degree(s)C. It also shows that there are no obvious improvements when the sample is annealed at 300 degree(s)C. The sample encapped with CdS film is better than uncapped CdS film. We found that the composition of Cd at the surface changes due to the chemical interaction of anode sulfide film. It also shows that the sulfide active CdS film can improve the adherence of ZnS to the MCT substrate and make the sample more stable.

Recent developments in HgCdTe photovoltaic detector technology are reviewed. The status of related areas in China are introduced. Some aspects of research work on device physics and technology conducted in the authors' laboratories are discussed. These include: the performance of HgCdTe photodiodes for IR fiber communication; the effects of field- enhanced generation-recombination and imperfections of the pn junction on HgCdTe photodiode I-V characteristics; and an analysis of the dependence of energy gap of HgCdTe on temperature and composition.

Staring 128 X 128 hybrid HgCdTe FPAs have been demonstrated with very good sensitivity and operability at temperatures compatible with thermoelectric cooling (> 160 K). The FPAs consist of HgCdTe/sapphire (PACE-I; producible alternative to CdTe for epitaxy) detector arrays hybridized to a CMOS readout having a gate modulation input circuit. FPAs with SWIR (2.5 micrometers at 78 K) and MWIR (4.56 micrometers at 180 K) cutoff wavelengths ((lambda) _co) were made and evaluated. The SWIR arrays were ZnS passivated; the MWIR arrays were CdTe-passivated. Though the (lambda) _co of the MWIR devices was not specifically optimized for terrestrial imaging at TE-cooled temperatures in the preferred 3.4 to 4.1 micrometers band, very good sensitivity was achieved, particularly relative to other technologies at temperatures >= 120 K. Mean laboratory noise equivalent temperature differences (NE(Delta) T) at 120 K, 170 K, and 180 K were 0.0048 K, 0.053 K, and 0.061 K respectively, for the MWIR device. While the NE(Delta) T was measured without a spectral filter, the sensitivity for 3.4 to 4.1 micrometers bandpass extrapolates to camera NE(Delta) T <EQ 0.05 K, if f/1.5 or faster optics are used. Near BLIP Detectivity (D^*) of 1.62 X 10¹³ cm-Hz^1/2/W and mean NE(Delta) T of 0.04 K were measured on the SWIR hybrid at 22.5 msec integration time and operating temperatures <EQ 162 K. Imagery of corresponding quality was subsequently generated. Since the CMOS multiplexer dissipates little power and needs a minimum of support circuitry, a viable thermoelectrically cooled FPA technology is implied.

This paper presents the first results obtained at LETI/LIR on 12.5 micrometers HgCdTe (MCT) infrared photodetectors. The objective of this program is to develop long buttable linear arrays of a few thousand photosites for earth observation from satellites. In a preliminary phase, a prototype with three photovoltaic sub-modules has been achieved and presents promising performances which are described in this paper.

A principal concern with the performance of GaAs/AlGaAs multiple quantum well long- wavelength detectors is obtaining high detectivity at relatively high temperatures. We have fabricated GaAs/AlGaAs quantum well infrared photodetectors (QWIPs) with graded barrier regions by varying the aluminum concentration during the barrier growth to improve high temperature performance. The effect of barrier grading has been characterized by the measurement of dark current and responsivity, leading to an evaluation of the temperature dependent detectivity. Detectivity of 2 X 10⁹ cm(root)Hz/W at 100 K has been measured on graded barrier QWIPs. We also report on the investigation of QWIPs with cut- off wavelengths ((lambda) _c) greater than 12 micrometers . A QWIP with (lambda) _c approximately 13.8 micrometers was grown with peak responsivity (R_p) of 1.2 A/W at 10 K. Both R_p and (lambda) _c were weakly bias dependent. The maximum detectivity of 6 X 10¹² cm(root)Hz/W was at 10 K at 11.7 micrometers and (lambda) _c equals 14.5 micrometers .

A metal grating coupled step-bound-to-miniband (SBTM) transition multiquantum well long wavelength infrared photodetector (LWIP) using a lightly strained In_0.07Ga_0.93As quantum well with a short-period Al_0.4Ga_0.6As -GaAs superlattice barrier structure has been developed. The new structure created a potential `step' in the superlattice barrier to block the undesirable electron tunneling current from the heavily doped ground state in the quantum well, which results in a significant reduction in the device dark current. The measured absorbance spectra and photocurrent are in good agreement with our theoretical predictions. The responsivity R_(lambda ) at V_b equals 6 V and T equals 77 K was found equal to 0.2, 0.15 A/W for the backside, and front normal incident illumination, respectively.

Large area, very long wavelength infrared focal plane arrays will be required for future astronomy, defense, commercial, and other applications. Rockwell International has developed and extended the state-of-the-art of these arrays, resulting in the world's first 128 X 128 element blocked impurity band (BIB) detector focal plane arrays. Rockwell's advanced epitaxial process control has made possible the fabrication of large area, uniform BIB detector arrays, having pixel spacings as small as 75 micrometers . Using standard indium bumping techniques, these detectors are mated to multiplexers which incorporate a wide variety of design features. Characterization of these arrays has shown low intrinsic noise, excellent response uniformity, response linearity over a wide dynamic range, and over 97% pixel operability. The results achieved demonstrate the feasibility and versatility of very long wavelength infrared (to 28 micrometers ) focal plane array technology based on BIB detectors.

Plasma reflectivity edge in infrared reflectivity spectra measured on narrow-gap semiconductors has been found to be most sensitive to the variation of carrier concentration based upon which a new way of infrared modulation has been studied. The modulation gain in power is impressively high. Thus a less sensitive room-temperature infrared detector could be as sensitive as the existing cooled infrared detectors in detecting middle-infrared radiation. The detectivity has been calculated.

The relationship between the Fermi level of MCT and the impurity concentration and temperature was calculated with considering the special characteristics of nonparabolic bands of MCT material. And then an imitative formula which coincide very well with the values of log log plot of NDA versus T(m/m0) the curves for constant reduced Fermi level are partially straight lines only have been obtained. Key words: Fermi level, nonparabolic band, Burstein—Moss effect

This paper details a high performance MWIR scanning array with 30 time delay and integration stages (TDI), forward and reverse scan operation, less than 80 noise electrons per stage, and on focal plane blooming control and dynamic range compression. The 92 column by 30 TDI device is fabricated using hybrid technology with the detector fabricated in InSb and indium bumped to a VLSI/CMOS/CCD silicon readout fabricated in 1.25 micrometers CMOS with two poly and three metal interconnect levels.

A new approach to the fabrication of monolithic infrared focal plane arrays is presented and examined in this paper. The array is based on photovoltaic diodes, parallel integration capacitors, and MIS field effect transistors (FET). The photodiode is connected directly to the integrating capacitor while the MISFET serves as a pass gate to the video line. This configuration is operated in the pseudo-staring mode. The array was implemented in InSb, in a process based on a new passivation in which a photo chemical oxidation of InSb is followed by a conventional photo chemical SiO₂ growth. A two-level metallization process was developed serving both for electrical connection and optical coverage. Two configurations were tested for the layout of the two metal layers. In addition, the lower metallization was implemented in Cr, Ti, and Al. The optimal structure is a planar array with Cr as the first metal layer which forms the source and drain contacts.

As the emphasis in infrared detector research shifts toward larger and more complicated arrays the amount of time spent on simple single-element and small arrays is decreasing. One set of applications where discrete detectors and arrays are still finding use is in satellites. In addition, scanned imaging arrays based on single element detectors and small arrays are still being manufactured. Discussion here is for small arrays and single element detectors. One of the aspects of detector operation that always needs to be addressed is amplification. Often detectors are attached to amplifiers through rather long leads. Such systems are subject to unwanted microphonic response as a result of the motion of the leads relative to each other or to the ground plane. This sort of microphonic response can many times be eliminated through careful wiring and routing techniques, however, in some severe environments it is not possible to eliminate all microphonic response. A commonly used solution to this problem is to hybridize the detector with a JFET front end to reduce the effective output resistance of the detector circuit relative to the amplifier input. The TIA in such configurations is completed off the focal plane at room temperature. This means that half the circuit is operating at cryogenic temperatures while the other part is operating at room temperature some distance away. Ideally it would be more convenient, if not better, to include the amplifier on the focal plane with the detector. (Of course this hybridization is necessary for large two-dimensional arrays.) Data have been acquired to show some of the limitations and opportunities for such an approach. Typical bipolar operational amplifiers (OP-27, OP-37, LM108) will not operate well at cryogenic temperatures. CMOS operational amplifiers generally will operate at cryogenic temperatures but suffer from high front-end voltage noise. The TLC2201 from Texas Instruments is a CMOS op-amp manufactured for low voltage noise. A discussion of its applicability to IR detector operation is presented herein.

A high performance IR detector/amplifier featuring high speed and high sensitivity is described. The design, developed at Cincinnati Electronics Corporation includes a novel JFET input configuration and cryogenic hybrid circuit implementation. Prototypes of this sensor assembly have demonstrated near background-limited performance at a bandwidth in excess of 600 kHz.

During the photodiode manufacture process of Mg+ ion implantation on N type InSb, it has been found that the high sheet resistance caused by implantation damage affects current-voltage (I-V) characteristics of the photodiode and there are three segments appeared in the I-V curve. Photodiodes with such high sheet resistances have excellent Auto-Gain-Control (AGC) characteristics.The structure and working mechanism of this photodiode are analyzed and its equivalent model is established. Computer simulation is found to be in good correspondence with the I-V characteristics of the real photodiode.The Auto-Gain-Control characteri tics curves of the photodiodes, which have different high sheet resistances, are presented in this paper.The possible applications of such InSb photodiode in the field of infrared(IR) systems are also discussed.

Based on our previous study of the single-gate charge-injection-device (CID) model we found that a serial connection of a diode and a capacitor can be used to emulate the full function of the single-gate CID under charge-sharing mode operation. With an additional capacitor, this model can be further extended to emulate the dual-gate CID which is necessary in developing the two-dimensional (2-D) focal-plane-array (FPA). These capacitors and diodes can be monolithically integrated together to replace the CID array. In the novel `CID emulator' array, in addition to preservation of the performance of the CID array, more advantages can be obtained. The major advantage is that the high quality oxide-InSb interface required in the CID array becomes unnecessary, and the restriction on the selection of the dielectric material is thus lifted. By choosing a dielectric material with a large dielectric constant, larger capacity and longer integration time become possible. The readout characteristics of the single-gate and dual-gate InSb-based CID emulator and an Si-based 16 X 1 CID emulator linear array are presented.

This paper describes a resolution improvement for a mercury cadmium telluride (MCT) infrared charge-coupled device (IRCCD) using a new microscanner. To keep the microscanner compact, the scanner prism is placed between the lens and the IRCCD, and driven by a bimorph piezoelectric actuator. By compensating the drive pulse and optimizing the prism thickness, the microscanner gives a fast and stable response with low distortion. The medium wavelength imager developed for the microscanner gives a 128 X 128 pixel thermal image in the 3 to 5 micrometers band using a 64 X 64 element IRCCD. The Nyquist frequency was increased from 10 to 20 cycles/mm to reduce aliasing without decreasing the S/N ratio. The noise equivalent temperature difference (NETD) was measured at 0.06 K with an f-1.7 lens. The modulation transfer function (MTF) was estimated by considering the scan distortion and the pixel aperture profile. By optimizing the aperture profile of the photodiode, the MTF at the Nyquist frequency was increased from 0 to 20%.

Single pixels and linear arrays of microbolometers employing the high-T_c superconductor YBa₂Cu₃O₇ have been fabricated by silicon micromachining techniques. The substrates are 3 in. diameter silicon wafers upon which buffer layers of Si₃N₄ and yttria-stabilized zirconia (YSZ) have been deposited. The YBa₂Cu₃O₇ was deposited by ion beam sputtering upon the yttria-stabilized zirconia (YSZ), then photolithographically patterned into serpentines 4 micrometers wide. Anisotropic etching in KOH removed the silicon underlying each pixel, thereby providing the necessary thermal isolation. When operated at 70 degree(s)K with 1 (mu) A dc bias, the D^* is 7.5 X 10⁸ cm Hz^1/2/Watt with a thermal response time of 24 msec.