3D optical data storage

3D Optical Data Storage is characterized by the ability to inscribe data within the volume of a data storage medium with three-dimensional resolution, as opposed to the two-dimensional resolution afforded by, for example, magnetic tape or CD. This innovation potentially allows very high data densities, but requires addressing techniques whereby points in the media may be read and written to through many other data points. No commercial product has yet come to market.

Overview

Current optical storage media, such as the CD and DVD store data as a series of reflective marks on an internal surface of a disc. In order to increase storage capacity, it is possible for discs to hold two or even more of these data layers, but their number is severely limited since the addressing laser interacts with every layer that is passes through on the way to and from the addressed layer. These interactions cause noise that limits the technology to perhaps ~10 layers. 3D optical data storage methods circumvent this issue by using addressing methods where only the specifically addressed data interacts substantially with the addressing light. This necessarily involves nonlinear data reading and writing methods, in particular nonlinear optics. 3D optical data storage is related to (and competes with) holographic data storage, but operates on different principles.

In this way, a disc the size and shape of a DVD can be envisaged to hold 100 or more individually addressable layers of information, each at a density similar to that of a DVD. Thus, a storage capacity of 1 Terabyte on a single disc seems easily within grasp.

History

The origins of the field date back to the 1950s, when Hirshberg developed the photochromic spiropyrans and suggested their use in data storage. In the 1970s, Barachevskii demonstrated that this photochromism could be excited by two-photon absorption, and finally at the end of the 1980s Peter T. Rentzepis showed that this could lead to three-dimensional data storage. The development of the laser was critical to the timing of these advances. Addressing in Rentzepis' system was provided by two intersecting lasers, such that two-photon excitation of the chromophores only took place at the intersection, and the bulk of the material was transparent to each laser individually. Data inscription was achieved by exciting the photoisomerization of the spiropyran chromophore, and readback was achieved by exciting fluorescence, which was collected and measured. The fluorescence behavior of the material depended on the form of the chromophore (written or unwritten). This proof-of-concept system had many technical limitations that precluded it from commercialization, but it stimulated a great deal of research and development in the following decades to solve these issues.

In the years since the conceptualization of the field, many academic and commercial groups have worked on 3D optical data storage products and technologies. Most of the developed systems are based to some extent on the original idea of Rentzepis, but a wide range of physical phenomena for data reading and writing have been investigated, large numbers of chemical systems for the medium have been developed and evaluated, and extensive work has been carried out in solving the problems associated with the optical systems required for the reading and writing of data. Currently, several groups remain working on solutions with various levels of development and interest in commercialization (see below), but no product has been generally released.

Processes for Creating Written Data

Data writing in a 3D optical storage medium requires that a change take place in the medium upon irradiation. This change is generally a chemical reaction of some sort, although other possibilities exist. Chemical reactions that have been investigated include photoisomierzations, photodecompositions, and polymerization initiation. Most investigated havebeen photochromic compounds, which include azobenzene, spiropyran, stilbene, and diarylethene. If the photochemical change is reversible, then rewritable data storage may be achieved, at least in principle.

Writing by Multiphoton Absorption

Although there are many nonlinear optical phenomena, only multiphoton absorption is capable of injecting into the media the significant energy required to electronically excite molecular species and cause chemical reactions. Two-photon absorbance is the strongest multiphoton absorbance by far, but still it is a very weak phenomenon, leading to low media sensitivity. Therefore, much research has been directed at providing chromophores with high two-photon absorption cross-sections. This is seen as a critical requirement for the development of a commercial product.

Another approach to improving media sensitivity has been to employ resonant two-photon absorption. Nonresonant two-photon absorption (as is generally used) is weak since in order for excitation to take place, the two exciting photons must arrive at the chromophore at almost exactly the same time. This is because the chromophore is unable to interact with a single photon alone. However, if the chromophore has an energy level corresponding to the (weak) absorption of one photon then this may be used as a stepping stone, allowing more freedom in the arrival time of photons and therefore a much higher sensitivity. However, this one-photon absorbance is a linear process, and therefore risks compromising the 3D resolution of the system.

Data Inscription During Manufacturing

Data may also be created in the manufacturing of the media, as is the case with most optical disc formats for commercial data distribution. In this case, the user can not write to the disc - it is a ROM format. Data may be written by a nonlinear optical method, but in this case the use of very higher powered, expensive lasers can be envisaged, so media sensitivity becomes less of an issue. Alternatively, the data may be physically written into the disc, like the molded pits and lands of a CD or DVD, either by molding or printing technologies.

Processes for Reading Data

The reading of data from 3D optical memories has been carried out in many different ways. While some of these rely on the nonlinearity of the light-matter interaction to obtain 3D resolution, others use methods that spatially filter the media's linear response. Reading methods include:

• Two photon absorbance (resulting in either absorbance or fluorescence). This method is essentially two-photon microscopy.
• Linear excitation of fluorescence with confocal detection. This method is essentially confocal laser scanning microscopy. It offers excitation with much lower laser powers than does two-photon absorbance, but has some potential problems because the addressing light interacts with many other data points in addition to the one being addressed.
• Measurement of small differences in the refractive index between the two data states. This method usually employs a phase contrast microscope. No absorption of light is necessary, so there is no risk of damaging data while reading, but the required refractive index mismatch in the disk may limit the thickness (i.e. number of data layers) that the media can reach.

Media Design

The active part of 3D optical storage media is usually an organic polymer either doped or grafted with the photochemically active species.

Media Form Factor

Media for 3D optical data storage have been suggested in several form factors:

• Disc. A disc media offers a progression from CD/DVD, and allows reading and writing to be carried out by the familiar spinning disc method.
• Card. A credit card form factor media is attractive from the point of view of portability and convenience, but would be of a lower capacity than a disc.
• Crystal or cube. Several science fiction writers have suggested small solids that store massive amounts of information, and at least in principle this could be achieved with 3D optical data storage.

Media Manufacturing

The simplest method of manufacture - the molding of a disk in one piece - is a possibility for some systems. However, it may be advantageous for the media to be constructed layer by layer. For example, this can allow features such as 'buffer zones' to exist between the data layers. In the case of ROM disks where the data is produced in manufacturing, this is a necessity. However, this layer-by-layer construction need not be the sandwiching of many layers together. Another alternative is to create the medium in the form of a roll of tape.

Drive Design

A drive designed to read and write to 3D optical data storage media may have a lot in common with CD/DVD drives, particularly if the form factor and data structure of the media is similar to that of CD or DVD. However, there are a number of notable differences that must be taken into account when designing such a drive, including:

• Laser. High-powered lasers are required that may be bulky, power-hungry, and require significant cooling. In addition, the optics must be able to withstand the high powers and the whole device must be safe for operation. In some examples, even more than one laser is required, meaning that beam combination optics are required.
• Variable Spherical Aberration Correction. Because the system must address different depths in the medium, and at different depths the spherical aberration induced in the focus is different, a method is required to dynamically account for these differences. Possible methods include optical elements that swap in and out of the optical path, moving elements, and adaptive optics.
• Detection. The detection system is very different from that in a CD or DVD, and requires operation with much lower signals. When fluorescence is used for reading, special light collection optics may be used to maximize the signal.
• Data Tracking. Layers of DVD-like data may be accessed and tracked in similar ways to DVD discs, but moving between layers (and remaining on the correct layer) must be addressed. The possibility of using parallel or page-based addressing has also been suggested. This would allow much faster data transfer rates, but would add a significant degree of complexity over DVD drives, requiring components such as spatial light modulators, the imaging of signals, and complex data handling.

Development Issues

Despite the highly attractive nature of 3D optical data storage, the development of commercial products has taken a significant length of time. This is the result of several technical issues that many systems have suffered from to lesser or greater extents, for example:

• Destructive reading. Since both the reading and the writing of data are carried out with laser beams, there is a potential for the reading process to cause a small amount of writing. In this case, the repeated reading of data may serve to erase it. This issue has been addressed by many methods, such as the use of different absorbance bands for each process, or the use of a reading method that does not involve the absorption of energy.
• Thermodynamic Stability. Many chemical reactions that appear not to take place in fact happen very slowly. In addition, many reactions that appear to have happened can slowly reveres themselves. Since most 3D media are based on chemical reactions, there is therefore a risk that either the unwritten points will slowly become written or that the written points will slowly revert to being unwritten. This issue is particularly serious for the spiropyrans, but extensive research was conducted to find more stable chromophores for 3D memories.
• Media Sensitivity. As we have noted, 2-photon absorption is a weak phenomenon, and therefore powerful lasers are usually needed to produce it. In particular, Ti-sapphire laser or frequency-doubled Nd:YAG laser excitation is used, but these instruments are expensive and unsuitable for consumer products.

Academic Development

Much of the development of 3D optical data storage has been carried out in universities. The groups that have provided valuable input include:

• Peter T. Rentzepis was the originator of the field.
• Watt W. Webb developed the two-photon microscope.
• Masahiro Irie developed the diarylethene family of photochromic materials.
• Yoshimasa Kawata developed data manipulation systems.
• Kevin C Belfield is developing photochemical systems for 3D optical data storage by the use of resonant energy transfer between molecules.

Commercial Development

In addition to the academic research, several companies have been set up to commercialize 3D optical data storage:

• Call/Recall was founded by Peter Rentzepis and continues to operate with funding by sbir grants. They have demonstrated non-destructive readout with over 1 billion read cycles before any noticeable change in the readout signal quality.
• Constellation 3D developed the Fluorescent Multilayer Disc at the end of the 1990s, which was a ROM disk, manufactured layer by layer. The company failed, but the IP was acquired by D-Data Inc., who are attempting to introduce it as the Digital Multilayer Disk.
• Mempile are developing a commercial system under the name TeraDisc. The company has demonstrated the recording and readback of 100 layers of information on a 0.6 mm thick disc, as well as low crosstalk, high sensitivity, and thermodynamic stability.
• The Romanian scientist Eugen Pavel has patented a doped glass-based 3D media named the Hyper CD-ROM, but serious development of the idea does not appear to have begun.
• Landauer inc. are developing a media based on resonant 2-photon absorption, but have no public plans to commercialize it. Since the media is based on DVD-sized single crystals of sapphire, it is unlikely that such a system could be economically viable.