Holograms: a new dimension in data storage

17 Jun 2002

As optical storage technologies reach their limits, holography promises to meet the demands of next-generation systems. However, after 30 years of research there are still no commercial holographic devices. Nadya Anscombe finds that European start-ups are set to change this.

From Opto & Laser Europe January/February 2001

Bayer's orange disc

After many years of empty promises, holographic data storage is about to make its market debut. German start-up company Optostor has developed what it claims is the first ever commercially available holographic mass-storage set-up. The firm plans to unveil the system in March.

Conventional optical storage systems hold information as bits on the surface of a disc. Optostor's set-up stores pages of data as holograms through the volume of a crystal that is half the size of a floppy disk and has a storage capacity of several terabytes.

The idea of storing data in holograms has been around for more than 30 years and researchers all over the world have been competing with each other to make the technology commercially viable. However, optical-engineering and materials-development problems have slowed progress.

Now, the availability of inexpensive chip components, such as liquid-crystal displays for spatial light modulators and CCD camera chips for use in detector arrays, has accelerated development. It seems that the little-known firm Optostor has beaten the pack. The company will be showing its product at the CeBIT show in Hanover, Germany, following a development time of only two years - in 1998 the firm gained the exclusive rights to the work on holographic storage carried out by Theo Woike's group at the University of Cologne's Institute of Crystallography.

Wolfgang Flakowski, managing director of Optostor, told OLE: "Our write-once-read-many system is designed for the mass-storage market. Between May and the end of 2001 we will be working with five firms that will test the stability and performance of the system. We hope to be selling commercial quantities of the set-up by 2002."

Optostor uses a photorefractive crystal - lithium niobate - as the storage medium. This material has been the mainstay of holographic data storage for many years, but it has weaknesses. For example, the light that is used to write the holograms also erases them during reading. This means that the holograms have to be fixed into the crystal, usually by heating.

Flakowski is reluctant to give details on how Optostor's system solves this problem. "We have about 20 patents being processed and until these are published we are keeping our system under wraps." Swedish company Optilink is ahead in the patenting race. It possesses many national and international patents covering its technology and it too is planning to debut its product at the CeBIT show. While both companies' technologies are based on holograms, the two techniques are different.

Optilink's set-up is based on creating holograms in a thin layer of a photochromic polymer, which means that the storage mechanism relies on the orientation of the polymeric side-chains.

Optilink's research is divided between the Risø National Laboratory in Denmark and the Budapest University of Technology and Economics in Hungary. As well as funding researchers, the company also has full-time employees working at both sites.

One of these is Gregg van Volkenburgh, chief technology officer. He told OLE: "There has been much hype about holographic data storage, but our materials actually work. They can be put onto various supporting structures and our first system will target the smart-card market. We hope to be showing it this spring."

Van Volkenburgh claims that the prototype will have a capacity of 4 Mbyte. However, up to 1 Gbyte can be stored on a smart card using his company's technology.

Optilink's system

Optilink's holographic system is based on a liquid-crystalline polyester that contains azobenzene. This group reorientates itself when it is irradiated with polarized light.

Raman Ramanujam, one of the company's researchers working in Denmark, says that the advantage of this set-up is that one wavelength of light can be used to read and write the holograms.

Writing holograms on photochromic materials can be slow. However, Ramanujam said: "We've made holograms using a picosecond laser and have found that it is the number of photons that is important and not the length of time of illumination."

Currently, the data density is more than 1 bit/µm². Optilink expects to pass the limit of 10 bit/µm² with the introduction of phase-coded multiplexing, which will give a capacity of more than 2 Gbyte if half of the card area is used as the active surface.

Ramanujam believes that Optilink's technology will complement rather than compete with Optostor's system. "Our method is simple and cheap and we only use one laser. Our first product is aimed at a very different market to Optostor's set-up."

For many years materials have been the Achilles' heel of holographic storage. Now, thanks to materials advances made by companies such as Optostor and Optilink, it seems that the tables have turned and the difficulties lie in optical engineering.

Ramanujam said: "The challenge now is to improve and miniaturize all of the optics and electronics that are needed for a system. We cannot get hold of good spatial light modulators and this has been a limiting factor in the development of holographic systems. Devices of high enough resolution are not commercially available." Geoff Burr, a member of IBM's holographic data-storage research team in the US, agrees. He said: "This is a typical chicken-and-the-egg problem. People will not develop parts for holographic data storage until they have seen the set-up working at impressive specifications. However, we cannot achieve this without the components."

Holographic data-storage

Unlike Optostor and Optilink, IBM's team has not restricted itself to one technology. It is investigating crystal and polymer systems. The team has demonstrated nearly 400 bits/µm², which Burr believes to be state of the art (a compact disc can store 0.7 bits/µm² and a DVD 4.5 bits/µm²).

To extend this high density into high capacity, IBM scientists are building a research platform to test the idea of using two beams of different wavelengths to read and write data, with one acting as a gating beam. In two-colour gated holography the crystal is illuminated with a gating beam - for example, incoherent ultraviolet light - during the recording process. This permits the storage of data that, subsequently, can be read in the absence of ultraviolet light without erasing the holograms.

This seems the most promising approach to solving the problem of a photorefractive medium, where the light that is used to read the hologram can destroy it. However, for this idea to become a reality the right material needs to become available. Currently, IBM is using single-colour lithium niobate to test its approach. A buffer hologram is employed to grab the original object beam and produce its phase-conjugate for transfer into the crystal.

Burr said: "This is one of the things that we're excited about. We're going to use a phase-conjugate readout to replace the expensive imaging optics. An invention of ours carries out the phase-conjugation outside of the storage material. This means that by the time the data-bearing object beam is recorded in the holographic system it has already been phase-conjugated."

Data-pattern search

This will enable the same reference beam to be used for recording and retrieving data and will also increase the potential storage capability. The entire system remains stationary, so this vast storage capacity can be combined with fast access.

IBM has been investigating holographic data storage off and on for more than 30 years. However, despite the advances that have been made in the last six years of concentrated effort, the company is still some way away from building a prototype.

Data imprint

Burr said: "We have moved away from building large platforms with small capabilities and no scalability to small testbeds that not only have an impressive capacity but can also be scaled down in size. If we wanted to we could put the systems that we are building now into the volume of a PC tower simply by folding the optics that are currently laid out on a two-dimensional table into three dimensions."

While this sounds simple, the final system size will be limited by the dimensions of the laser and the fact that the optical beams must be expanded to the size of the spatial light modulator. Now that many of the materials-science difficulties have been overcome, it is up to the optical engineers to make holographic data storage work.