Photoferroelectric Image Storage in PLZT Ceramics

Authors Paul Peercy and Cecil Land check image resolution at new PFE device. New method of storing images in PLZT eliminates the need for photoconductive films on the ceramic. A new lead lanthanum zirconate titanate (PLZT) ceramic device uses a recently discovered photo ferroelectric (PFE) effect to store high-resolution, high-contrast, nonvolatile images1.. The PFE effect eliminates the need for photoconductive films which characterize all previous PLZT image storage devices.2.-8. The new image storage device consists simply of a thin flat plate of PLZT sandwhiched between two transparent electrodes. Image storage is achieved by exposing the image on a surface of the plate, using near-UV light at the band gap energy of 3.35 eV (3700Å), and, at the same time, switching the ferroelectric polarization through a portion of the the hysteresis loop.9. Images are stored both as spatial distributions of light scattering centers in the bulk of the PLZT and as surface deformations which form a relief pattern of the image on the exposed surface. Both image storage phenomena are related to spatial distributions of ferroelectric domain orientations9 introduced in the PLZT during the image storage process. Stored images can be read out using either transmitted or reflected light. Transmitted light is scattered selectively so that the image can be viewed directly or projected onto a screen using a collimated light source and a Schlieren optical system to enhance the image contrast. For projection, the PFE device is placed in one focal plane of a converging lens and an aperture is located at the opposite focal point. Reconstruction of the stored image by reflected light utilizes the surface defromations on the storage surface. Collimated light is diffracted and scattered by the surface deformations so that the image can be projected onto a screen using Schlieren optics.5. Reflectivity of the storage surface can be increased by depositing on the surface a dichroic film that transmits the near-UV write-in light and reflects the visible readout light. Either total or selective erasure of stored images is accomplished by uniformly illuminating the area to be erased with near-UV light and simultaneously switching the ferroelectric polarization back to its initial state prior to image storage. Experiments have shown that erasure of the stored image can also be accomplished by heating the PFE device to the ferroelectric-penferroelectric phase transition temperature(T1) at which all domain structure disappears.10., 11. For the PLZT 7/65/35 (7 at % La, 65/35 zirconium to titanium ratio) ceramics used in PFE devices, Tt is about 100°C.10. Erasure by heating has also been achieved with I2R heating obtained by passing a sheet current through one of the transparent electrodes of the PFE device.11. The presently known characteristics of the PFE device along with its attractive simplicity of fabrication suggest several potential applications. Temporary image storage and display is an important potential use. Various types of image and optical information processing are also suggested by device characteristics discussed later. Using scanning techniques to write-in and read out images could also accomplish transmission of high-resolution images over various types of communication channels.12. A major advantage of the PFE device over previous PLZT image storage devices is its simplicity.1. The device consists of a polished plate (0.2 to 0.3 mm thick) of coarse-grained (grain size >3 μm), rhombohedral-phase PLZT ceramic7. with low-resistance, transparent indium-tin oxide, In2–xSnxO3–y, (ITO) electrodes13. sputter deposited on the two major surfaces as shown in Fig 1. The device shown in Fig 1 is designed for readout of the stored image using transmitted light. For image readout with reflected light, a dichroic film should be deposited on the ITO electrode surface exposed to write-in light. The dichroic film transmits the write-in light and reflects the readout light. With transmitted readout light (scattering mode of operation), the maximum range of image contrast (exceeding 100:1) and maximum resolution (about 40 line pairs/mm) obtained to date were with a device using PLZT 7/65/35 with approximately 5 μm average grain size. ITO electrodes with low sheet resistivities of 2 to 3 ohms per square are used to minimize switching strains and allow fast switching speeds. All previous PLZT image storage devices required a transparent photoconductive film deposited on one or both surfaces of the ceramic. These photoconducting films were then overcoated with transparent electrode(s).2.-8. In those PLZT ferroelectric-photoconductor (FE-PC) devices which use organic photoconductors (e.g., polyvinyl carbazole, PVK), image storage is inherently slow (1 to 10 seconds) because the conduction mechanism in PVK is a hopping process. Furthermore, the PVK is not uniformly transparent at all visible wavelengths which adds insertion loss and undesirable color in the device. Using sputter-deposited cadmium sulphide14. or zinc-cadmium sulphide13. photoconductive films achieves much faster image storage (e.g. 10 to 30 ms). These films usually have lower optical insertion losses than the PVK films, which makes them additionally attractive for most device applications. Unfortunately, CdS and ZnCdS films are technologically difficult to sputter with predictable and reproducible results. Difficulties and limitations imposed by photoconductive films emphasize the need for eliminating these films. PFE devices thus have an advantage over previous FE-PC image storage devices. Photoferroelectric PLZT ceramics are both photosensitive and ferroelectric. Visible light can photoexcite carriers from trapping centers in the band gap, and near-UV light at or near the band gap energy (3.35 eV or 3700 Å) can photoexcite carriers both from trapping centers in the band gap and across the band gap. Because they are ferroelectric, PLZT ceramics can be poled to a saturation remanent polarization by application of an electric field of sufficient magnitude. Their remanent polarization can be reversed or reoriented totally or incrementally by a properly oriented electric field.9. These ferroelectric properties are illustrated by the polarization P versus electric field E hysteresis loop shown in Fig 2. The net polarization is zero in the virgin or thermally depoled state (A). Applying a negative electric field increases the polarization (negatively) until it saturates at the maximum applied field. Removing the field, poles the ceramic to negative saturation remanence (B). Remanent polarization can be switched incrementally through intermediate values, e.g., (C), (D) and (F), to positive saturation remanence at (B') by application of a positive electric field. In the PLZT compositions used for PFE image storage, the remanent polarization (See Fig 2) is directly related to corresponding ferroelectric domain orientations. When a voltage is applied to the electrodes of a PFE device (Fig 1), the resulting electric field in the ceramic plate can be spatially modulated by exposing a surface of the device to band gap light of spatially nonuniform intensity. In areas illuminated by band gap light, the electric field is affected by the presence of photoexcited carriers; in the nonilluminated areas, the electric field is uniform. If the polarity of the applied voltage is such that the partially illuminated surface is negative with respect to the opposite surface, the electric field which produces domain switching in the illuminated areas will be greater than that in the dark areas. In this case, an applied voltage just large enough to switch the ferroelectric remanent polarization in the illuminated areas will produce substantially less polarization change in the dark areas. This selective domain switching results in spatial nonuniformity of the ferroelectric remanent polarization and a corresponding spatial nonuniformity in the light scattering characteristics and surface deformation. Aex is the area exposed to band gap light, and Ad is the dark area. As will be discussed in the next section, ΔPr(ex) depends on the intensity of the band gap light. After image storage (switching from (F) to (G) in Fig 3) the UV light is removed and the net polarization is switched to some point such as (H) in Fig 3 to obtain a high-contrast positive of the input image or to (J) to obtain a high-contrast negative of the input image. The net polarization may be subsequently switched from (J) to (H) through intermediate values to obtain various amounts of baseline subtraction and corresponding variations of gray and contrast. It should be noted that the hysteresis loops of Fig. 2, 3 are taken from actual data on the same PFE device while storing a photographic image, and the curves of Fig. 2, 3 are reproduced with the same scale factor. The data of Fig. 2, 3 show that the effective coercive field9. during image storage was approximately 30% less than the coercive field measured from Fig 2. Therefore, the presence of photoexcited carriers or space charge fields resulting from retrapping of photoexcited carriers caused a net average increase of the domain switching field of about 30% during image storage. Illuminating the originally exposed surface of the PFE device with band gap light of uniform intensity (about 15 mW/cm2), and switching the net remanent polarization to saturation remanence at (B') – as shown by the dashed lines of Fig 3 – erases the stored image. As stated in the preceding section, a voltage applied to the electrodes of a PFE device produces an electric field in the PLZT ceramic which can be spatially modulated by exposing a surface of the device to band gap light of spatially non-uniform intensity. Absorbed photons in PLZT grains at or near the surface of illuminated areas photoexcite carriers which are transported by the applied electric field to new trapping sites. The resulting charge separation within these grains produces space charge fields (Esc) which oppose the applied field (EA). The charge redistribution gives rise to an effective spatial variation of the internal domain switching fields (Ei) in the illuminated areas. Fig 4 illustrates schematically the PFE image storage mechanism. Fig 4 represents a portion of a cross section of a PFE device, including the ceramic grain structure with grain dimensions greatly exaggerated for the purpose of illustration. At the beginning of the image storage process illustrated in Fig 3, a portion of the PFE device surface is illuminated with band gap light as shown in Fig 4. A voltage of the polarity indicated in Fig 4 is applied to the ITO electrodes. Photoexcited carriers in the grain layers near the surface exposed to UV radiation are transported by the applied field to new trapping sites within the grains or at the grain boundaries. Fig 4 shows charge separation in only one grain layer, but, depending on the nature of the absorption characteristics of PLZT15. and the presence of longer wavelength illumination, absorption and carrier photoexcitation may occur over several grain layers, beginning at the illuminated suface. The space charge fields Esc oppose the applied field EA in the absorbing grains. Since the charge separation depends on the electric field within these grains, the maximum attainable value of Esc is equal to −EA. The net effect of compensation of EA by Esc in the absorbing grains is that the applied voltage VA is dropped across the the nonabsorbing grains beneath the illuminated area, which results in an increased electric field E,. If the thickness of the ceramic plates is t and the thickness of the absorbing layer is tA, then in the limit of Esc=−EA. Since EA = VA/t, it is obvious that EI > EA. The field EI causes domains to switch at lower applied voltage than EA the field in the dark area of the ceramic. The result – uniform localized domain switching — is essential to the image storage process. Assuming that the quantum efficiency is uniform over the surface of the ceramic plate, the number of photocarriers excited in any illuminated area depends approximately linearly on the localized intensity of the band gap radiation. It follows, therefore, that the magnitude of the localized space charge fields Esc and, hence, the localized domain switching field EI depend approximately linearly on the intensity of the band gap radiation provided |EI| > |EA|in the absorbing grains. This intensity dependence of E, is responsible for the apparently linear gray scale reproduction in stored photographic images. Other physical mechanisms have been suggested16 to explain nonuniform domain switching during image storage. These mechanisms involve both photoconductivity and photoassisted domain switching. Glass, et al.,17 have reported the presence of both photoconductive and photovoltaic effects at wavelengths of 4500 Å and less for a PLZT 8/65/35 composition. Further study of PFE image storage mechanisms will probably reveal that all the physical effects enumerated above contribute in varying degrees to the image storage capabilities of PLZT. The erasure process described in the preceding section frefrequently requires more light energy than the storage process. Rouchon, et al.,18 report that erasure of photoinduced changes in refractive index in PLZT 9/65/35 requires four times the storage light energy density. We also find that images stored for a number of days are more difficult to erase than those stored for only a few hours. Better understanding of the storage mechanism will undoubtedly lead to more efficient erasure processes. At present, the erasure mechanism is thought to consist of photoexcitation of carriers and subsequent recombination of electron-hole pairs to restore the original polarization state of the ceramic prior to image storage. Images are stored in PFE devices as spatial distributions of scattering centers in the bulk of the PLZT and as surface deformations which form a relief pattern of the image on the exposed surface. Light scattering in coarse-grained, rhombohedral-phase PLZT ceramics varies with the magnitude of the remanent polarization Pr measured parallel to the light propagation direction.1., 4., 7. thus, from a practical standpoint, the transmittance or insertion loss of a PLZT plate measured by a fixed-aperture detector depends on the remanent polarization of the PLZT. Also, PR is the saturation remanent polarization after application of an electric field Es equal to three times the coercive field Ec, i.e., Es=3Ec. Points (A) through (G) on the insertion loss curve of Fig 5 correspond approximately to similarly designated points on Fig. 2, 3, except that the sense of the polarization axis is reversed in Fig 5. From the discussion of manent polarization of the PLZT. Fig 5 shows the insertion loss in dB plotted as a function of the normalized remanent polarization Pr/PR for a PLZT 7/65/35 ceramic plate. Note that insertion loss in dB can be divided by 10 to obtain optical density. Also, PR is the saturation remanent polarization after application of an electric field Es equal to three times the coercive field Ec, i.e., Es=3Ec. Points (A) through (G) on the insertion loss curve of Fig 5 correspond approximately to similarly designated points on Fig. 2, 3, except that the sense of the polarization axis is reversed in Fig 5. From the discussion of the image storage an erasure processes relating to Fig. 2, 3, you can determine the associated optical insertion losses from Fig 5. For example, if a positive of an input image is stored as a spatial distribution of scattering states between (B) and (D), it can be converted to a negative by simply switching the remanent polarization states so that they are distributed between (D) and (B‘). In this example, localized areas with an original insertion loss at (D) would be switched to (B‘), and areas originally at (B) would be switched to (D). This illustrates how positive-tonegative image conversion occurs and how you can obtain baseline substraction to control the contrast of the stored image by switching to intermediate polarization states, an important capability of PFE devices for contrast enhancement and image processing. The variation of light scattering with changes in remanent polarization is useful only in devices which use transmitted light to view or project the stored image. Reflective mode devices employ a similar variation of surface strain with changes in remanent polarization to store images as spatial variations of surface deformations.2., 7. Fig. 6 shows two photographic images stored using a PFE dedevice with a 0.25 mm thick PLZT 7/65/35 ceramic plate. The images were projected onto Polaroid color film through a Schlieren optical system with an effective angular aperture of about 5° using visible light transmitted by the PFE device. The images were stored using light from a Hg-vapor lamp of about 50 mW/cm2 intensity and 700 Å bandwith centered at the 3650 Å Hg line. Similar image storage has been obtained with light of about 15 mW/cm2 intensity and 200 Å bandwith centered at the 3650 Å Hg line. Maximum contrast of the stored images is about 20 dB, and maximum resolution is between 30 and 40 line pairs/mm. Because the images were photographed on Polaroid color film, the apparent contrast and resolution in the reproductions of Figs 6, 7 are considerably less than in the ceramic plate. The images of Figs 6, 7 can also be viewed or projected by light reflected from the surface exposed to the UV storage light. Indentations or deformations in the locally switched areas of the exposed surface form a relief pattern which can be used to diffract light reflected from this surface and reconstruct the image using Schlieren optics.5.