This special issue of Optical Engineering is the second of three consecutive issues on charge-coupled devices. Papers included in this issue present some of the recent major advancements in the manufacture of CCDs intended mainly for scientific applications, both as imaging detectors and as signal processing components. In addition, as with the other issues in this series, papers are included that describe several specific CCD-based instruments, illustrating the wide range of fields in which CCDs now play a role.

The paper describes the design and development of a CCD shift register that can take analog samples of an electrical signal at rates up to 1 x 108 samples/s and provide temporary storage for 1024 samples. It is intended for operation in a fast-in, slow-out mode for capturing high speed electrical transients. The requirements of fast clocking and moderate clock driver power led to the choice of a serial-parallel-serial design. The CCD chip contains four storage arrays that form two differential channels. Differential operation provides good linearity. The two channels can be used independently or multiplexed for sampling at the highest rate. The CCD is of the buried-channel type and uses four-phase clocking on all registers. It is fabricated using a modified NMOS process, with two polysilicon layers to provide an overlapping transfer gate structure. Operating characteristics of the CCD are described, together with its behavior in an instrument environment. For optimum performance the CCD chip is mounted on a hybrid, together with clock driver ICs. Such devices are used in the signal acquisition section of a digital storage instrument that was developed recently.

Tektronix is currently fabricating two very large area charge-coupled-device sensors intended for astronomical and other high performance scientific imaging applications, e.g., medical imaging, fluoroscopy, x-ray imaging, spectrometry, particle detectors, etc. In this paper we discuss the performance requirements for scientific-quality CCDs and then focus on the design of the two Tektronix devices and discuss the progress toward achieving the desired performance. These devices are intended for rear-illuminated applications and have 512 X512 and 2048 X2048 pixel formats. The thinned 10 to 20 pm thick Si membrane is fully supported by a unique glass ceramic substrate. Quantum efficiencies of >70% at 700 nm and >40% at wavelengths <400 nm have been measured on a test device. Dark currents as low as 6 pA/cm2 also have been measured recently.

This paper describes the design and performance of a scientific CCD array for use in NASA's Shuttle Image Spectrometer Experiment. The device is a four-phase, buried-channel CCD structure that operates in the frame-transfer mode. Thesensor consists of 64 X404 pixels (each 52µm X 52 um), has a 100% fill factor, and operates in the visible and near-infrared spectral regions. In operation, the 404 horizontal elements provide spatial information, while the 64 vertical elements give spectral information covering the wavelength range of 400 to 1000 nm in 10 nm increments. The high full-well capacity of each pixel (>2 X106 electrons) and low noise floor (<30 electrons rms) yield a dynamic range of more than 95 dB. In addition, the device has been designed to have good linearity characteristics. The unique dual-output structure allows a horizontal row to be read out to the right or to the left, or it can be split from the middle to both right and left output circuits simultaneously for high speed applications. The power dissipation of the device is about 60 mW.

The introduction of the flash gate has made possible the fabrication of backside-illuminated CCDs with high sensitivity and stability throughout a wide range of ultraviolet and visible wavelengths (100 to 5000 A ). It had been determined previously that the characteristics of the oxide layer beneath the gate are critical to the ultimate achievable CCD performance. However, by creating an improved oxide layer in conjunction with the flash gate, we are now able to consistently produce CCDs with near-ideal UV performance. In the interest of transferring flash technology to industry, in this paper we present recent results and related background theory that optimize the flash gate specifically for application in the UV.

This paper presents an updated version of an earlier paper describing the Texas Instruments three-phase 800 X800 charge-coupled-device image sensor ["800 X800 charge-coupled device image sensor," Opt. Eng. 22(5), 607-614 (1983)]. Although this device was originally designed to be used as the sensor for the Wide Field/Planetary Camera on the Hubble Space Telescope, it is now being used as the detector of choice on many ground-based telescopes. We review the performance of the device and indicate the important contributions it has made to the understanding of general CCD performance.

We present results of a program to enhance the ultraviolet and extreme ultraviolet response of charge-coupled devices. The ultimate goal of our program is to develop a large format device with both high and stable quantum efficiency from 100 to 3000 A that can be used as a windowless imaging detector in a space environment. Ultraviolet quantum efficiency measurements have been made for several ion-implanted and laser-annealed test CCDs specially fabricated by Tektronix for this program. Quantum efficiencies as high as 22% at 2500 A , where the absorption depth in silicon is -55 A, have been observed in one such test CCD. Quantum efficiency measurements of standard back-illuminated RCA and Tektronix CCDs are also presented.

Low light level imaging with commercially available CCDs has been limited by CCD availability as well as by poor performance at low signal levels by many devices constructed for consumer applications. These performance limitations usually take the form of poor charge-transfer efficiency at the low temperatures and/or high noise levels associated with charge-detection schemes that normally need to operate only near room temperature. This paper describes a commercial virtual-phase CCD and associated operating mode that allow high performance low light level imaging at low temperatures. This capability makes the described device ideally suited to both scientific and military low light level imaging applications.

We have designed and fabricated a 420 X420 pixel front-illuminated CCD imager. The device was designed to be used both individually and as part of a four-chip focal plane consisting of four chips closely abutted in a 2 X 2 array, for a total resolution of 840 X 840 pixels. Particular attention has been paid to achieving low-noise operation, and noise levels of less than 20 e- at a 1.0 MHz data rate have been achieved routinely. We describe special focal plane assembly techniques with which we have achieved 1 um relative positional accuracy between the chips.

Driven by a requirement for an 80 mm square sensor, we have combined CCD and tapered fiber optic technologies to create a large area focal plane. Seven custom low noise CCD imagers are assembled onto four tapered fiber optic bundles to match the physical characteristics of a tube camera used in an operational system. Two taper ratios are used in the focal plane to meet requirements for broad search and high accuracy measurements in a single device. The entire focal plane is cooled to -40°C to reduce dark current to insignificance at an integration time of 0.6 s. The focal plane has been integrated into a camera head that has an all-digital interface. The camera preserves the low noise characteristics of the focal plane in a multichannel assembly.

A slow-scan, cooled CCD camera system, similar to that used for astronomical observations, was constructed at Toulouse Observatory and used for testing CCD chips. Several devices were tested extensively. This paper describes the design and performance of some Thomson-CSF solid-state area-array CCD sensors. These sensors use a frame-transfer organization adapted to operate in a double-interlaced-field readout mode with a memory zone (sensor TH 7861) or adapted to operate in a single-field mode without a memory zone (sensors TH 7882, TH 7883, TH 7884). We briefly describe the system and the evaluation of the photometric performance of these chips, including quantum efficiency, readout noise, transfer efficiency, linearity, dark current, and field uniformity. Some projected developments of Thomson-CSF chips for scientific applications, such as the TH X31156 (1024 X1024 pixels) and the buttable TH 7882 (TH X31157), are also presented.

With the increasing use of charge-coupled devices for astronomical observations, it is vital to obtain the highest quantum efficiency possible due to the limited amount of observing time available on large telescopes. One method of increasing the response of back-illuminated CCDs is to apply antireflection coatings. Because silicon's index of refraction is extremely high in the ultra-violet, a gain of more than 100% in quantum efficiency can be achieved at a wavelength of 0.37 Am with the proper choice of dielectric thin-film materials. We present a study of materials suitable for silicon CCD antireflection coatings for use in the 0.32 to 1.0 Am spectral range. Constraints on these materials, such as refractive index, absorption coefficient, application method, tempera-ture requirements, internal stress, and radioactivity, are discussed. We find hafnium oxide, lead fluoride, and aluminum oxide to be among the most suitable thin-film materials available. We also present a number of single- and multiple-layer antireflection coating designs optimized for CCDs, with particular emphasis on the blue and ultraviolet.

A new class of charge-coupled devices called charge-coupled-computing devices is described. These analog circuits perform arithmetic functions such as addition, subtraction, and magnitude comparison in the charge domain. The circuits are compact and are designed to be insensitive to rail voltages, simplifying their utilization. These devices, in conjunction with input, output, and analog memory circuits, can be combined to form a simple but general-purpose and fully programmable charge-coupled computer. A prototype charge-coupled computer has been fabricated and tested. Prospects for forming a large array of computers (e.g., 1000 to 10,000) on a single chip for spatially parallel image preprocessing are discussed. Such image plane preprocessing of data would find use in real-time mobile robot vision systems, in which low power, lightweight computing is critical for economical viability.

Using a thinned, three-phase Texas Instruments 800 X800 CCD (T1 3PCCD) and inexpensive plastic Polaroid material, we have constructed an imaging polarimeter for astronomical data collection. A unique method of image shifting, shuttering, and Polaroid rotation enables us to circumvent the sky transmission fluctuations that normally plague polarimetry. The method is possible only because of the exceedingly low dark current and exceedingly high charge-transfer efficiency of this CCD. Our observations of standard polarimetry stars have verified the operation of our imaging polarimeter and have demonstrated that polarimetry can be accurately performed below the 1% polarization level with well-behaved CCDs.

Techniques and instruments developed for extracting precise positional information from CCD images of point-source and extended optical targets are described. With the use of thinned, backside-illuminated devices, performance levels close to those expected from straightforward geometric considerations are obtained. For ideal point sources (ie., stars outside of the earth's atmosphere), centerfinding accuracy of 1/100 pixel and measurement jitter of less than 1/250 pixel have been measured. Detailed tests showed that the main factor limiting tracker accuracy is small variation in the optical image shape rather than CCD noise or response nonuniformity. Methods of searching the entire field for the desired targets are described, along with windowing techniques for tracking. Also discussed are three specific examples of flight instruments previously developed or now being developed at the Jet Propulsion Laboratory.

A star mapper using a Fairchild 2048-element linear CCD array (CCD 143) has been developed for the Indian Remote Sensing Satellite (IRS-1A), which is scheduled for launch in 1987. The star mapper is designed to scan stars up to 5th magnitude by using the orbital motion of the sun synchronous satellite, with a swath of 8° across the scan direction. Design details of the star mapper along with some test results of the qualification model are presented.

Selenium films obtained by vacuum deposition and hot-wall epitaxy (HWE) were investigated for structural, morphological, and holographic characteristics. Vacuum-deposited Se films, which are generally amorphous, were found to crystallize under an intense beam of an electron microscope. The transmission electron diffraction results indicated that the crystallized phase was b-monoclinic. These films have high transmittance in the visible spectrum and exhibit good exposure characteristics for recording holograms. Parameters such as diffraction efficiency, modulation transfer function, and spatial frequency, determining the exposure characteristics, were evaluated. Films grown by hot-wall epitaxy, however, have altogether different properties. The surface morphology showed growth to needlelike single crystals 5 kall in size, with preferred growth in the (100) direction. These films were gray and exhibited poor transmittance and scattering centers, thus losing their recording properties. X-ray diffraction results of Se films grown by HWE indicated that the crystallized phase was hexagonal in structure.

Solar-selective black nickel absorbing films have been deposited on galvanized iron substrates by a simple dip-and-dry method using a mixture of nickel chloride and thiourea solution (0.1 M concentration). The films developed a surface structure with well-defined globular particles distributed over the entire surface and exhibited good selective properties with solar absorptance = 0.81, thermal emittance = 0.10, and selectivity = 8.10.

The frequency response characteristics of a mechanical surface profilometer were examined. The treatment was based on a computational model that tracks the motion of a spherical stylus as it traces a mathematically defined surface. The stylus was found to possess a nonlinear response to the surface profile. The effects of the nonlinearity on the frequency response and spatial frequency bandwidth of the stylus profilometer were examined. The model assumes that no external force acts on the stylus; therefore, effects such as surface deformation were not investigated.

Changes in surface and bulk properties of United Detector Technology pin-05D photodiodes as a result of 1.3 Mrad gamma irradiation are compared. As suspected by previous investigators but not verified until now, changes in surface properties are seen experimentally to significantly alter overall device characteristics. Changes in device properties include increases in surface conductivity, improved quantum efficiency at visible wavelengths, decreased dark current at very low reverse bias, decreased diode ideality factor, decreased infrared response, and decreased minority carrier lifetime. The first four results are new and permit differentiation between surface and bulk effects. A model consistent with all of these measurements is presented to explain the changes. The model is based upon gamma ray photodesorption of surface impurities.

In an editorial entitled "Status of Optical Engineering" that appeared in this space one year ago, I presented some publication information for 1986 and compared it with similar information from the previous year. Because I think that many readers of Optical Engineering may find such information of interest, decided to do it again this year. In the table below, the figures in the 1986 column are actual figures for that year, whereas those in the 1987 column are projections based on publication data available at the time this editorial was written.

Until relatively few years ago, random behavior in physical systems was thought to result either from quantum mechanical uncertainties or from simple complexity. The motion of atoms in a Maxwellian gas is an example of the latter process, which on a macro scale we describe only on a statistical basis. The mechanics of turbulent fluids had been another such subject: there were just too many degrees of freedom present to enable a meaningful solution of Navier-Strokes equation.