Holographic drives set for long-awaited debut

08 Jul 2004

Holographic data storage is about to come of age. Jacqueline Hewett speaks to InPhase Technologies, the US firm promising to have a 200 GB drive on the market next year.

From Opto & Laser Europe July/August 2004

Holographic storage drives capable of recording up to 10 times more data than next-generation DVDs are set to become commercially available in 2005. The company behind their development is InPhase Technologies of the US, whose first product will store 200GB of data - the equivalent of 98 million printed pages, 200,000 1MB photos or more than 1500h of high-quality music. What's more, the Colorado-based company is confident that its technology will lead to the production of a 1.6TB drive by 2010.

InPhase was spun out of Lucent's Bell Laboratories at the end of 2000 and now has a staff of 56. "We started an activity within Lucent to explore holographic data storage and the most important innovation was really the media development," Bill Wilson, InPhase's chief scientist and a company founder, told Opto & Laser Europe. "We developed the media technology and incubated it, but got to the point where we needed to build a drive. That's when we decided to create a new company."

Immediate advantages The attraction of holographic storage is that hundreds of separate holograms, known as pages, can be recorded through the full depth of the storage medium. Unlike related technologies, which record one data bit at a time onto the surface of a disc, holography allows one million bits of data to be written and read out in a single flash of light. This means that a postage-stamp-sized piece of media could be used to store 2GB of data, and transfer rates in excess of 20MB/s are possible. Storing data through the depth of the medium has the added bonus that it makes the content very difficult to pirate.

Couple these facts with InPhase's claim that it can produce a reliable optical medium at low cost, and holographic storage begins to look like an attractive alternative to current optical techniques. According to Wilson, the first beneficiaries of the new technique are likely to be the film and broadcast industries. The move to high-definition formats will see a huge explosion in the amount of content that has to be archived. Currently these industries use tape to archive content, but this approach has its drawbacks.

"The problem with tape is that it is not an archival medium, in the sense that you have to re-record it every 5-7 years," explained Wilson. "Once the industry moves to high-definition formats, there is no way tape can support that. Holographic recording is a natural fit for them. We predict our media's archival life to be of the order of 50 years. The cost of owning an archive will drop dramatically."

The development of high-density holographic storage drives has been plagued by the lack of both a suitable optical storage medium and a clever technique for writing data to it. But after 10 years of work, InPhase believes it has the answers. The challenges surrounding the hunt for the right medium were twofold: the right chemical composition had to be ascertained; and a method of producing optically flat media was needed to ensure distortion-free holograms.

The firm began by looking at photopolymers used in the display industry. Known as 1-chemistry systems, these compounds consist of a monomer dissolved in a polymer matrix. "The problem with those materials early on, though, was that they were designed for writing a really strong hologram into a thin piece of material," explained Wilson. "And this is completely the opposite of what you want to do in storage: write a relatively large number of really weak holograms."

Properties such as the medium's sensitivity, dynamic range and ability to retain its shape when heated are crucial, says Wilson. "There are lots of things you have to do to satisfy the criteria for a storage media. And what happened in our exploration of 1-chemistry systems was that it was impossible to decouple all of the parameters."

InPhase's breakthrough came when it developed what it calls a 2-chemistry system, comprising two independent but compatible polymers. The first polymer, which is thermally cured, acts as a solid matrix that can be optimized for the manufacturing process. The second is a photopolymer which is dissolved in the matrix. This photosystem can be tweaked to optimize the recording dynamics. The recording material that resulted from this system was christened "Tapestry".

"We started doing all our development work recording at green wavelengths," said Wilson. "But now that blu-ray lasers are starting to come online, we are able to transition our media from green to blue with only minor changes. We'll do the same thing when we go to the red for low-cost applications."

Having found the right 2-chemistry recipe, the next hurdle was making a medium that was almost interferometrically flat. To surmount this, InPhase teamed up with Maxell and Imation and together they created the simple "Zerowave" method."It's similar to making DVDs. You start off with two substrates and our 2-chemistry material is the bond in between," Wilson explained. InPhase says its Zerowave process routinely fabricates media with better than λ/4/cm² flatness.

With these two elements of the puzzle in place, the Tapestry medium hit the market. InPhase is in its second year of selling green-wavelength-recording material (Tapestry HDS 3000), while blue media (Tapestry HDS 5000) launched in February this year. Red-sensitive material is also in the pipeline.

The firm sells its media in 2x3 inch slides, 3x3 inch coupons, 5.25 inch-diameter discs or customer-specified formats with thicknesses ranging from 200µm to 2mm. It is currently supplying the media and test equipment to optical drive companies that are investigating their own holographic drives. Although they are already low-cost, InPhase hopes to further reduce the cost of its media to 25 cents per GB ($50, or €41, for 200GB).

High-density storage There was still one final piece of the jigsaw missing, however: how best to write high densities of data to the medium. In the simplest approach, holograms are recorded using light from a single laser beam. This is split into a signal beam, which carries the data, and a reference beam. The hologram is formed where the two beams intersect in the recording medium.

The data are encoded onto the signal beam using a spatial light modulator. This converts electronic data (a stream of noughts and ones) into a 2D checked pattern of light and dark pixels. This pattern represents a page of data containing around one million bits. To read out the data page, the reference beam deflects off the hologram and reconstructs the stored information. The result is projected onto a CMOS detector which reads the data out in parallel.

High-density storage requires stacks of data pages, called books, to be written to different locations throughout the medium. Typically each book contains up to 1000 pages and has a width of 3mm, with approximately 1500 books stored in each disc. The key issue was finding the most efficient technique for recording the data.

InPhase teamed up with Sony, Alps Electric and an undisclosed optical mechanical assembly partner to work on this. The collaboration started by building an angle multiplex drive. This involves writing one data page and then sweeping the reference beam over a large number of angles to write successive data pages in the same volume.

Although the technique worked well, the angular sweep needed to write many pages was large and a significant amount of material was wasted between book locations. "To get reasonable capacities you have to store anything from 600 to 1000 holograms in the same volume, and this requires an angular sweep of 30-40 degrees or more," said Wilson. "Designing the optics to scan a beam over that angular space and keep the phase of the beam flat is a tough optics problem."

To eliminate the large angle sweep and the wasted space between the books, InPhase started overlapping the stacks of data pages. The company's innovation was an idea it calls polytopic recording and filtering. When a reference beam is shone onto overlapping books, light is diffracted from adjacent books as well as the central one. To filter out the unwanted information, InPhase places a carefully configured aperture just beyond the media or at the beam waist. Only the image of the central book passes through the filter to the detector, while the rest is blocked by the aperture. When the full book has been read out, the medium moves and the beam falls onto another series of books.

Wilson believes this elegant architecture will be sufficient to produce a 1.6TB drive. To ramp up capacity, InPhase plans to increase the angular sweep of the reference beam and improve the resolution of the spatial light modulator and CMOS detector.

In the meantime, InPhase is gearing up to release its Tapestry 200R in 2005. This will boast a storage capacity of 200GB, a transfer rate of 20MB/s and 96 pages per book. It will also use InPhase's Tapestry HDS 5000 medium in tandem with a 405nm 60mW laser diode. Targeted at the video and data archival sectors, the device will be priced between $7000 and $10,000 and will be the same size as a high-end tape drive.

The next big question is whether InPhase can adapt to produce a rewritable material and drive. Wilson says it can, and that significant progress has already been made. "Our hope is that a rewritable version of our drive will come online within 18 months of the release of our first drive," he said. "The only change will be to add an erase cycle into the drive, but you will have to do a global erase."

It remains to be seen whether InPhase can keep pace with its product roadmap, but with the main technological hurdles out of the way, it is hard to see what could stop the advance of holographic data storage.