We review previously reported results on the fabrication and operation of spatial light modulators which use electroabsorption in multiple-quantum-well (MQW) structures. These devices are electrically addressed by clocking a pattern of charge packets into a charge-coupled device (CCD). At each CCD pixel the amount of stored charge determines the electric field across the underlying MQW structure, which in turn determines the optical transmission of the device near the band-edge quantum-well excitors peak. The light is incident normal to the wafer surface and enters the device through semi-transparent CCD gates. Both 16-stage and 32-stage one-dimensional devices and 16-by-16 two-dimensional devices have been fabricated on epitaxial layers containing GaAs/AlGaAs MOW structures for operation in the 0.85-μm region and also on epitaxial layers containing InGaAs/GaAs MQW structures for operation above 0.9 μm. The one-dimensional devices have been operated as spatial light modulators at low CCD clock frequencies (up to 1 MHz); they display modulation depths far superior to CCD spatial light modulators fabricated on bulk GaAs.

In this paper recent theoretical and experimental work in the area of multiple quantum well modulators is presented. The theoretical work includes the application of the effective mass approximation to compositional MQW structures and the use of a two-band tight-binding approximation to doping modulated Nipi structures. The theoretical calculations are used to obtain electric-field-dependent absorption and refractive index in the above MQW structures. Experimental electroabsorption in compositional MQW structures is described, as well as preliminary optical characterization of Nipi structures. Concepts of photoactivated and electrically addressed MQW-spatial light modulators are presented. Finally, theoretical evaluation of quantum dot arrays and their potential use in spatial light modulators is discussed.

A liquid crystal light valve (LCLV) incorporating a GaAs photoconductor has been developed. The device structure is based upon a microdiode array and hybrid field effect twisted nematic liquid crystal cell. The active area of 2.54 x 2.54 cm (1 in. x 1 in.) is defined by approximately 40 million, 2 μm square Schottky diodes at 4 μm pitch. Two device types are reported, both employing 2 in. GaAs wafers of semi-insulating LEC grown material. The first device has a photoconductor thinned to 70 μm. An ohmic back contact is established with Au/Ge Ni alloy. Results obtained from this device are presented. A second device is currently under investigation featuring a thinner photo-conductor <20 μm and a Se ion implanted heavily doped n+ back contact. Construction details are given for both LCLVs and reasons for initial design modifications are included. A brief overview of the present status of the second generation GaAs LCLV is presented and planned developments discussed.

The design, fabrication and performance of an improved 128 X 128 deformable mirror device (DMD) is presented. The DMD is a fast, low power, analog response, spatial light modulator. The improved DMD consists of a monolithically formed array of cantilever beams which are electrostatically deflected by an underlying MOS transistor address matrix. The on-chip address circuitry supports a single video input to the chip with all demux operations performed on-chip. The original DMD (1,2) differs in two respects from the improved version. Instead of monolithically formed cantilever beams, the original DMD utilizes a hybrid architecture wherein a continuous metalized membrane is placed over an underlying address chip by a mechanical lay-down process. In addition, instead of the single video input of the improved DMD, the original version has 128 inputs corresponding to the 128 columns of the x-y transistor matrix. The deformable mirrors of the improved DMD are cantilever beams formed by two layers of high reflectivity aluminum alloy. A thin layer forms the hinge and a thicker layer forms the cantilever beam and the surrounding support metal. The beam, hinge and support metalare formed over an underlying planarizing spacer layer, which is selectively removed by plasma etching to form an air gap underlying each cantilever beam pixel. This paper describes the design, fabrication and performance of the improved 128 X 128 cantilever beam DMD, including preliminary data on the Fourier plane behavior of this device and a discussion of its application to optical correlation. Dynamic range of the DMD in the Fourier plane at the Nyquist frequency has been measured at greater than 45 db and is limited by the CCD image detector used to record the intensity distribution.

Current developments in ferroelectric liquid-crystal device physics are reviewed and experimental work in shutter and photoaddressed spatial light modulators is presented.

This paper discusses the present state of development of the Hughes charge-coupled-device-addressed liquid crystal light valve (CCD-LCLV). The device is based on the photo-activated silicon LCLV, and features a serial electrical input that is converted to a truly parallel optical output. The device can utilize both coherent and incoherent readout light sources, with a spectrum extending from the visible to the IR. The complete array has been activated, at serial data rates exceeding 6.5 MHz, corresponding to 100 Hz frame rates, using external drive electronics. Design limiting resolution has been observed, with up to 50:1 contrast ratios. Real-time video has been demonstrated. The CCD-LCLV is suitable for a range of applications, including adaptive optics and optical data processing systems.

The Piezo-Electric Elasto-Optic (PEEO) effect was studied previously by G.Sirat, N.Ben-Yosef and Y.Fisher . The main problem is the technological one .i.e. the manufacture of the modulation array. The requirements of the array are a large number of piezoelectric elastooptic crystals vibrating at different resonant frequencies.The mono-lithic approach was introduced previously by using a configuration of three well separated electrodes . The present work will describe a fabrication technique of a Monolithic Image Plane modulator. Crystal thickness and pixel definition were optimized for maximum pixel density and minimum crosstalk.The optimization algorithm solves the acoustical wave equation with the boundary conditions for a three electrode configuration. Iterating with the length, thickness and density of the electrodes,a well defined standing wave beneath the central electrode is achieved.In order to assess the appropriate density and conversion values , the wave envelope may be graphically displayed.The common electrode deposition technique is replaced by a structural etching method . Previously a metalized electrode was used as the mass load . It is replaced by etching a quasi electrode on the crystal substrate,thus obtaining a modulated surface . The unidimensional monolithic frequency multiplexer is defined on a single intrinsic GaAs crystal.The experimental results agree with the numerical predictions.A maximum signal to noise ratio of 2.6 and a minimum band width of 49 Hz has been achieved.

Microscopic descriptions from recently developed many-electron theory of second and third order nonlinear optical susceptibilities are reviewed for several important conjugated organic structures. Electron correlation effects result in a symmetry controlled mechanism of highly correlated, charge separation in n-electron densities during virtual excitation processes.

Quantum wells (QWs), which can be made from alternate thin layers of two semiconductor materials, show a large electroabsorption called the quantum-confined Stark effect (QCSE). This made possible the realization of electrically controlled optical modulators, and optically controlled self electro-optic effect devices (SEED), that can operate at high speed and low energy density. We will review the various materials systems that have shown QCSE and some of the different devices that have already been fabricated, the latest among them being an optically controlled spatial light modulator, an optically addressed dynamical memory, and an optical S-R latch.

The electro-optical properties of n-i-p-i doping superlattices are very promising for device applications, Two distinct approaches towards modulating the absorption and the refraction by applying external voltages are identified, and calculated results for GaAs n-p superlattices are presented. The first method uses the injection and extraction of non-equilibrium charge carriers into the superlattice via selective contacts to the n- and p-type layers. The second approach makes use of the possibility to apply a uniform external electric field parallel to the growth direction via sandwich electrodes. It is shown that the two schemes differ substantially in their dynamic response.

A ferroelectric polarization P (permanent, spontaneous electric dipole moment) is predicted for certain non-racemic liquid crystals based upon simple symmetry arguments. The magnitude of this polarization is a major factor in determining the eletro-optic response times achievable in ferroelectric liquid crystal (FLC) spatial light modulators. While the symmetry arguments are compelling, insight into the actual molecular origins of P is necessary if new materials with high polarization and fast electro-optic response are to be designed in a directed way. Based upon the concept that ferroelectricity in FLCs is a manifestation of a novel form of molecular recognition occurring in the FLC phase, we have developed a simple stereochemical model for P in principle allowing prediction of the sign (handedness) and magnitude of the polarization for specific compounds. The design, synthesis and ferroelectric properties of new materials will be described in the context of this model.

An experiment was performed to demonstrate the holographic phase conjugation characteristics of a PLZT Integrated Spatial Light Modulator (ISLM) device. The ISLM was addressed with angularly separated reference and object beams. The aberrations of a colli-mated laser beam were stored in an ion-implanted sample of antiferroelectric (AFF) phase PLZT in the form of a fringe pattern resulting from the interference of the object beam with the reference beam. The angle of incidence between the beams was 0.4° and resulted in a resolution of approximately 15 line pairs per millimeter. The -1 diffracted order was directed back through the path of the object beam and the aberrations were cancelled. It was concluded that AFE-phase PLZT has sufficient diffraction efficiency for phase conjugation at higher resolutions which result from addressing with angularly separated beams.

This paper deals with the fundamental limit of speed of the photorefractive effect at a given power level. Such a limit is derived by counting the number of photons needed for the formation of a photorefractive index grating.

High-purity BaTiO₃ single crystals (with transition-metal ion impurity levels well below 1016 cm-3 ) and a series of iron-doped BaTiO₃ crystals with iron concentrations ranging from 5 to 1000 ppm were grown. These crystals were annealed in oxidizing and reducing atmospheres and their photorefractive properties characterized. A thermodynamic defect model was used to predict changes in the valence state of the iron as a function of annealing temperature, oxygen partial pressure and iron concentration. It was found that the "pure" crystals were photorefractive and that the saturation diffraction efficiency of gratings established in both the doped and the undoped crystals increased when the crystals were annealed in reducing atmospheres, but did not scale with the concentration of iron in the crystals. The evidence suggests that variable-valent iron ions are not the dominant photorefractive centers in these samples.

A method for implementing general linear transformations on one dimensional optical field distributions in integrated optics is described. This method uses the Bragg selectivity of volume holograms to replace one dimension of the bulk optical vector matrix multiplier. A means of writing appropriate matrices optically using unguided light is also described. Experimental results are presented which demonstrate the diffraction of guided light by holograms written with unguided beams.

Many functions involving manipulations of symbolic information are capable of being performed in a parallel fashion. With the advent of new parallel multiprocessor architectures, the approach to symbolic computing is experiencing a change from complex heuristic approaches accomplished on serial machines to more brute-force, but much faster, approaches on parallel machines. Due to the heavy degree of interconnection required in such architectures and to their parallel nature, numerous optical processing techniques are under serious consideration for symbolic computing. This paper describes the multiprocessor approach to symbolic computing and discusses the central role of two-dimensional SLMs for optical symbolic computing.

Neural networks are highly interconnected systems of simple analog processing elements which are intrinsically highly parallel in nature and exhibit abilities such as leatniny and generalization. bince optical systems are inherently highly paialiel, in that spatially modulated wavefronts can contain large amount of information and nonlinear materials such as photorefractive crystals can be used to pertoLm processing functions on these wavefronts, optical implementations of neural network architectures are a natural development. We refer to optical implementations of neural network systems as "Optical Neuromorphs."

The capacity of the Vander Lugt correlator, defined as the maximum number of separate images that can be recognized, is estimated. The increase in capacity that results from the use of a volume hologram in place of the commonly used planar hologram is derived. The effects of binarizing the reference filter and the shift invariant properties of the two classifying systems are also analyzed.

Iterative processing techniques can often provide useful solutions to problems, especially problems of inversion, when deterministic solutions are impracticable or too costly. These iterative techniques usually require a feedback path so some measure of the output can affect an input of the system in order to guide it towards the correct solution. This feedback path often needs to be a sophisticated manipulation of the output state, such as linear transformation of a vector. However, when the processing can be properly designed, the necessary dynamic range of this feedback path doesn't need to be as high as the desired precision of the solution. Such feedback operations can be performed quite well by analog multichannel optical processing systems. We show how we have used optical systems in a simulated-annealing inversion algorithm to invert a system of linear equations, such as deconvolution for recovering an intensity distribution from an image formed by an aberrated or indirect imaging system. We show that we can significantly increase the execution speed over the digital processor, but we do not sacrifice the ultimate precision of the solution by using the analog optical system.

Many optical computing schemes employ either intensity-based logic or polarization-based logic. Intensity-based logic utilizes the absence or presence of light to represent the binary logic states 1 and 0, thus dissipating light after each operation. Polarization-based logic uses orthogonal states of polarization to represent its two binary states. This has the advantage that no light is lost in performing cascaded logic operations. Surface stabilized ferroelectric liquid crystal (SSFLC) devices, which can function as half-wave plates, can be used to implement polarization-based parallel optical logic gates. Under the influence of an applied electric field, a properly oriented SSFLC device rotates the polarization of incident light through 90°. By using SSFLC matrix arrays, the light passing through selected pixels can be rotated independently of adjacent pixels. Operating characteristics of a 4 X 4 SSFLC matrix array including contrast ratio, switching speed, switching voltage, and rotary transmission are presented. A matrix addressing scheme is discussed and implemented with a SSFLC. Using two SSFLC matrix arrays, we demonstrate the XOR/XNOR logic operations along with measurements of the percentage of polarized light rotated into the output binary states.

The inverted-cloverleaf cantilever beam version of the Texas Instruments Deformable Mirror Device (DMD) has become available for use as a spatial light modulator in hybrid optical correlators. In this form the DMD is active principally in phase rather than amplitude, though both quantities vary. The DMD can be used in both the input and filter planes of a correlator. The optical effects of addition in phase differ substantially from addition in electric intensity or even amplitude, so the theory developed for phase-only filtering of amplitude objects does not wholly apply. We have done simulations of the phase-encoded/phase-filtered correlation and show the expected correlation results for example images. Sensitivity to noise, scale, rotation and gain have been examined. Experimental work has been started to incorporate DMD's in an optical correlator.

We present a new multichannel optical correlator/convolver architecture which uses an acoustooptic light modulator (AOLM) for the input channel and a Semetex magnetooptic spatial light modulator (MOSLM) for the set of parallel reference channels. Details of the anamorphic optical system are discussed. Experimental results illustrate use of the system as a convolver for performing digital multiplication by analog convolution (DMAC). A limited gray scale capability for data stored by the MOSLM is demonstrated by implementing this DMAC algorithm with trinary logic. Use of the MOSLM allows the number of parallel channels for the convolver to be increased significantly compared with previously reported techniques while retaining the capability for updating both channels at high speeds.

The optical implementation of all parallel logic gates using inexpensive LCTVs is demonstrated and basic experimental results are presented.

An all-optical fiber-optic crossbar switch capable of interconnecting N input fibers to M output fibers with an arbitrary interconnect pattern has been constructed. The switch can use any 1-D or 2-D spatial light modulator capable of realizing NxM pixels, and can be configured for permutations, full or partial broadcast, or "wired oring" of several inputs to a given output. Values of N and M achievable depend on the light efficiency of the switch components and the data rate. A 4x4 switch has been built; in a computer switching environment, the approach is capable of realizing a 16x16 switch at 1 Gb/s and a 32x32 switch at 100 Mb/s. The development of the switch has included construction of both a PLZT switch array and a magneto-optic switch array. We will report on the investigations of both technologies. Reconfiguration times in the range of a few microseconds to a few tens of microseconds are of interest.

An optical crossbar switch is proposed which uses a semiconductor-addressed, deformable-mirror spatial light modulator to direct gigabit per second optical transmissions to several receivers. The switch performance is described in terms of mirror element size, deflection range and number of modulator elements, as well as emitter, detector and lens dimensions. These parameters are incorporated in a link energy budget and also provide estimates of system dimensions. Maximum-sized 128 to 1000 channel crossbars appear feasible based on the performance of various lightwave components.

Applications of Gallium Phosphide (GaP) acousto-optic devices, at wavelengths from 570nm - 1.06um seem to be ideal for fiber optic modulators, scanners, deflectors, frequency shifters, Q-switches and mode lockers. One of the major applications are for RF spectrometers in early warning radar receivers and auto-correlators. Longitudinal GaP acousto-optic Bragg cells which have respectively operational frequencies in the range of 200 MHz - 3 GHz and diffraction efficiencies in the range of 120%/RF watt to 1%/RF watt have recently been fabricated. Comparatively, shear GaP devices which have operational frequencies in the range of 200 MHz to 2 GHz and diffraction efficiencies from 80%/RF watt to 7%/RF watt have also been constructed.

This paper reviews the current state of development of a high-performance infrared image transducer for use in advanced real-time infrared scene simulation. This transducer is based upon the silicon liquid crystal light valve (LCLV) technology developed at Hughes Research Laboratories. Results are presented for simulations in the 3 to 5 Am and the 8 to 12 μm spectral regions.

After the aberrations of a collimated laser beam are stored in a hologram, the -1 order or conjugate wave can be directed back through the original aberration and will emerge as an unaberrated wave. We perform this type of aberration correction using a new type of PLZT (lead lanthanum zirconate titanate) as an instant storing holographic media. After a review of holographic aberration and storage of images in PLZT, we describe a laboratory experiment which proves this concept. Star tests and interferograms demonstrate the success of the experiment.

A real time hologram generated in the liquid crystal layer of a leaky mirror silicon light valve is read by the same low intensity beams used to record it. For a particular geometry, one of the diffracted beams propagating away from the device is the optical phase conjugate of one of the original beams, thus enabling correction of path propagation aberrations. Experimental results show correction of tilt, focusing error, and random phase aberrations.

A real-time polychromatic image correlator that uses a magneto-optic spatial light modulator (SLM) device for pattern recognition based on both the color and shape of an input object is presented. The proposed system utilizes a multi-channel spectral matched spatial filter employed in a binary coherent optical correlator. Input color images are transformed into binary color coded coherent images by a color grating. The color coded images are read-out by a charge coupled device interfaced with a magneto-optic SLM. The color coded binary images are then processed by a multi-channel joint spectral matched spatial filter synthesized by monochromatic light. Pattern recognition experiments for naturally illuminated real color objects are presented.