For any storage technology to remain competitive over time, it is critical that its access time, system volume, and cost be kept constant or preferably reduced while its capacity and data rate are increased. This requires that an increasing amount of data should be accessed by low-cost pickup sensors that can move quickly and accurately. Mechanical constraints dictate that fast and accurate movements can only be achieved over short distances; this consideration leads to the conclusion that data must be kept as local as possible with respect to the pickup heads. Historically, this consideration has driven the increase in areal densities, allowing much larger amounts of data to be stored, accessed, and retrieved without an increase in access time and system cost.
However, as optical areal densities approach optical diffraction limits, researchers have started seeking new solutions. On the one hand, solutions may entail further increasing the areal density by combating the diffraction limits of optics, using, for example, near-field optics. On the other hand, solutions may take advantage of additional available dimensions such as proposed for various 3-D optical storage concepts. Indeed, data residing in a volume may be considered as being local to the pickup sensors if the performance cost and actual cost of accessing it in 3-D is affordable. This consideration has led to several research programs in 3-D optical storage. (It is interesting that some electronic DRAM designers are seeking similar solutions by packaging DRAM ICs in 3-D.)
Several approaches to 3-D optical storage are being actively investigated in both the United States and Japan. These approaches include extending present disk system storage technologies to (1) a 3-D layered format; (2) volume holographic encoding, where data is recorded in a distributed fashion in the volume; and (3) wavelength encoding persistent hole-burning, where wavelength or time can be used as the third dimension. Bit-oriented layered disk media promises to impact the field sooner than holographic storage or spectral hole-burning.
Layered 3-D optical storage
Layered 3-D optical storage is a natural extension of present optical disk systems that has the potential to increase media capacity without significantly affecting the drive costs. In this case, volumetric density (Mbit/in. 3) becomes critical. Volumetric density can be limited by the linear and track densities, as discussed earlier in the context of the areal density of 2-D optical disks. Volumetric density can also be limited by the layer density (layers/in.), which represents a density metric along the third dimension. The layer density can be limited by the resolution of the media, the depth of focus of the pickup lens, and the accuracy with which the lens can be positioned in the third dimension. Therefore, the volumetric density is governed by the effective volume of the spot, which in turn is limited by the volumetric resolution of the media, the numerical aperture of the optics, the wavelength, and the positional accuracy of the pickup head in the third dimension.
Layered optical disk storage is presently being investigated (using different approaches and at different levels of development) by Toshiba, Matsushita, and Sony in Japan, and by IBM and Call/ Recall, Inc. in the United States.
Double-layer optical disks. The capacity of present optical disk systems can be immediately doubled with a single drive by having two active storage layers, one on each side of the disk substrate. Toshiba's proposed Double Video Disk (DVD) approach addresses each layer with different optical heads, one on each side. Sony proposes a similar two-layer structure that is accessed by a single optical pickup head. For computing applications, single-sided access is preferred. These approaches may well accelerate the introduction and acceptance of new optical video disk products.
Hybrid multilayer 3-D optical disks. In the United States, IBM is investigating methods of addressing disks of four to twenty active layers with a single optical head. In this approach (Fig. 3.15), several conventional CD-ROM disks, each with an active layer, are stacked together, and a dynamic focusing lens records or reads bits from a desired layer. The speed, depth of focus, and resolution in the third dimension of the dynamic focusing are critical to the success of this method. Focusing speed may limit the seek time of the system; depth of focus and positional accuracy in the third dimension may limit the spot size; and ultimately, crosstalk noise between bits at different layers will limit the total number of layers. Although the layer densities of this approach may be limited (each layer thickness ~100 mm), it may be developed rather quickly, since it is based on conventional media.
Fig. 3.15. Multilayer optical disk stack (IBM).
Monolithic multilayer 3-D optical disks. Two approaches, one investigated by Call/Recall, Inc., in the United States and the other by Matsushita, promise to bring layer densities to a level comparable to track densities and eliminate the crosstalk between layers. Both approaches rely on new organic polymer storage media and new schemes of data access.
The approach used by Call/Recall relies on recording bits in a volume by using two-photon recording (Hunter et al. 1990). As Figure 3.16 shows, a spot is written in the volume of a molded organic polymer only at points of temporal and spatial intersection of two beams with sufficient photon energies, one carrying information and the other specifying location. The recorded bits are read by fluorescence when excited by single green photons absorbed within the written spot volume. Using this method, Call/Recall has demonstrated multiple image storage in ROM configuration with a portable player unit. The results indicate no crosstalk between layers and excellent stability of the written bits at room temperature. Spot size is limited by the recording wavelengths through diffraction effects, as well as by the sensitivity and integration time of the readout detector. The approach promises low-cost, high-volumetric-density ROM disk media with a thousand or more layers for image storage, and also low-cost compact disk player drive units employing semiconductor green lasers. In addition to the extremely large volumetric densities achievable with this monolithic 3-D disk approach, stored data can be accessed in parallel by taking advantage of low-cost digital camera technology, possibly eliminating the need for a dynamic focusing lens.
Fig. 3.16. Principle of two-photon recording in 3-D (Call/Recall, Inc.).
The multifrequency optical memory Matsushita is investigating relies on a layered organic media where the absorption wavelength of the written spots in each layer can be programmed during fabrication. Polymers doped with spiropyran molecules are used in each layer. When exposed to ultraviolet light and an appropriate heat treatment, the sensitized molecules at a given layer exhibit a sharp and distinct absorption band at the selected wavelength. A wavelength-tunable laser can then read the bits by tuning its wavelength to the wavelength associated with the desired layer. The capacity and density of such a memory are ultimately determined by how sharply the absorption bands are synthesized and how well the relative positions in the spectra are controlled. Layers are presently deposited by Langmuir-Blodget technique. Matsushita is investigating the approach at the fundamental research level.
Potential impact of multilayer disk formats. The potential impact of layered 3-D optical disks on the capacity of optical disks can be much greater than the impact of, for example, the blue lasers. This is because the growth factor in capacity is directly proportional to the number of layers. Assuming that the areal density is not affected, the 3-D layering provides the potential for realizing optical disks with capacities exceeding 100 GB slightly beyond the turn of the millennium, as described in Figure 3.17. In this area, research in Japan seems to be somewhat behind that of the United States.
Fig. 3.17. Potential impact of 3-D multilayer optical storage.
In bit-oriented memories, if any portion of the storage media is damaged or blocked, the data stored in that region is lost. This is not the case for holographic storage, where the information about each stored bit is distributed throughout a large region. If a portion of the holographic storage media is damaged or blocked, instead of causing catastrophic loss of some of the data, all of the data are partially degraded. For common types of damage, such as surface dust or smudges, holograms are remarkably robust, which makes holographic data storage an attractive option. Despite requirements for more complex optical systems, high-cost media, and sources with good spatial coherence, there has been continued research in the field since the early 1960s (Solymar and Cooke 1981), particularly in the United States.
Volume-phase holograms. Holograms are created by recording the interference pattern of two optical wavefronts. The storage media can record the fringes as index and/or amplitude modulation. When the recording is illuminated by one of the wavefronts (the reference beam), the other wavefront (the object beam) is reproduced. Since the diffraction efficiency of amplitude holograms is limited to 6-7%, and the efficiency of volume-amplitude holograms is limited to 33%, volume-phase holograms with diffraction efficiency approaching 100% are best suited for storage applications. In addition, the Bragg selectivity of volume-phase holograms allows the volume multiplexing (by angle or wavelength multiplexing) of many pages of data with little crosstalk between pages.
Photorefractive crystals (PRC) and photopolymers are typically used at universities both in the United States and Europe in demonstration experiments. These materials are recorded at room temperature, and the inorganic PRCs offer high photocyclicity. Using PRCs, 10,000 images have been successfully stored and retrieved (Mok and Stoll 1992). Holographic storage, like two-photon storage, allows for parallel read-write of the data and offers the potential for very high data rates. The main drawbacks of the approach are the high cost of the PRCs, the decrease in diffraction efficiency with increasing number of multiplexed holograms and readout cycles, and the relatively small and slow changes in the index of refraction. To circumvent these bottlenecks, researchers recently have made progress using recording schedules to improve the uniformity of the recorded data, fixing procedures to permanently store data in a ROM fashion, and employing new recording geometries to minimize crosstalk. Most recently, photorefractive polymers have been demonstrated that exhibit large index-of-refraction changes under very large electric fields.
Most of the above-mentioned progress has been achieved in United States, where the National Storage Industry Consortium carries out a research program with the participation of several U.S. corporations and universities to develop holographic storage media and systems. In Japan, there is less interest in holographic storage, although several university laboratories and Matsushita are involved with it at the fundamental research level. Several Japanese managers the JTEC panel talked with seemed somewhat skeptical and asked, "What has changed in holographic storage since the '60s to justify renewed interest? The materials are still small in dimensions, difficult to produce, and expensive."
Persistent spectral hole-burning
Another 3-D storage system option is persistent spectral hole-burning (PSHB), which uses wavelength as the third dimension (Moerner 1988). PSHB results from the photo-induced transformations between the various ground states of a molecule. The types of molecules being investigated for PSHB experiments have a large, inhomogeneously broadened absorption band comprised of a large number of narrow absorption lines.
To record information into PSHB materials, a tunable wavelength, narrow bandwidth optical source is used to illuminate the material. The illuminating light induces photophysical transformations, which dramatically modify the absorption profile near the source energy. After recording, the originally smooth profile is altered into one with sharp dips near the recording frequencies.
The extent to which the wavelength can be used to store multiple bits of information in a PSHB material is governed by the number of distinct holes that can be burned into the absorption profile. This is given by the ratio of the inhomogeneous and homogeneous linewidths. This ratio can be as high as 10 6, but only at extremely low temperatures (< 4 degrees K). This is because, in general, the homogeneous linewidth increases with temperature. Recent research efforts have focused on developing new materials that have a large number of spectral holes at close to room temperature.
The original memory system proposed for this technology was a bit-oriented memory. The spectral holes were burned using an intense beam (write operation) and detected using a beam with much lower laser intensity (read operation). One problem with this system is that it is quite difficult to determine if a spectral hole is present without using a variable threshold detector. Recently, absorption holography has been implemented to reduce the background intensity and greatly improve the contrast ratios. Other recent advances have been made to reduce crosstalk by applying an external electric field and sweeping the optical frequency and phase while recording the hologram. A 20-second movie has been recorded with this procedure (Kohler et al. 1991).
PSHB is an active area of research at both Japanese and U.S. universities, despite the requirement of low temperature, because of the expectation that it will provide a storage capacity improvement of three to four orders of magnitude over that of present optical disk systems. It is hoped that future advances will raise the operating temperature. Basic research is being carried out at Mitsubishi, Sony, and NTT. SRI is studying a variation of PSHB, photon echo, where time rather than wavelength is used as the third dimension.
One of the novel optical storage technologies that may have an impact in the next millennium is near-field scanning microscopy (NSOM). In its essence, this method is very similar to that used for magnetic hard disks, in that the pickup heads must be placed very near (about 50 nm above) the media. A near-field probe can produce spots as small as 40 nm in diameter and conceptually can achieve areal densities on the order of 50 to 100 Gbits/sq. in. The problem is that the probe must be in near contact with the medium, making it difficult to prevent head crashes and support media removability.
Near-field optical storage techniques are being investigated both in Japan and the United States, mostly as methods for testing and diagnosis of current optical and magnetic storage media, because of the high resolution capability this approach provides.