24 Aug 2005
Near-field optical data storage requires a lens with a numerical aperture of greater than one and an air gap of 30 nm between the lens and disc. Jacqueline Hewett visited Philips in the Netherlands to find out how the electronics giant has dealt with these challenges.
Third-generation optical storage formats, such as Blu-ray, are just beginning to enter the consumer market, but researchers are already developing successors. One promising contender is near-field optical recording, which has the potential to store 150 GB of data on a standard 12 cm disc.
As the popularity of archiving digital photographs, high-definition TV recordings and other such items grows, so does the need for storage capacity. New standards have been released to meet this insatiable demand, from CDs that store 680 MB to single-layer Blu-ray discs that hold 25 GB (see box). Today, firms like Philips believe there is a real need to develop removable media with capacities of hundreds of gigabytes.
Over 400 delegates gathered at the International Symposium on Optical Memory and Optical Data Storage, held in Hawaii in July, to hear the latest developments in this thriving field. It was at this four-day forum that Philips presented details of two proof-of-principle experiments that show the potential of near-field optical technology.
Blu-ray uses a 405 nm laser diode and a lens with a numerical aperture (NA) of 0.85. Increasing the NA of the lens above one opens the door to higher capacities, but also poses the daunting problem that the lens must be held just 30 nm from the disc.
Although this sounds like a tough challenge, Philips has come up with two solutions, each based on a 405 nm laser diode. The first is an NA 1.9 system that can store up to 150 GB on the first (top) surface of a disc. The second uses an NA 1.45 lens, and information is written through a polymer cover-layer and stored in multiple layers within the disc, each holding up to 75 GB.
Both systems make use of a solid immersion lens (SIL) - a miniature hemispherical lens, allowing NAs of greater than one - coupled to an objective lens. Philips' starting point was the NA 1.9 lens system, which consists of a super-hemispherical SIL with a radius of curvature of 0.5 mm and an NA 0.45 objective lens (see figure 1).
"We make our SIL from LaSF35 glass. This is the highest refractive index glass [n = 2.086 at 405 nm] available, which is transparent to blue light," explained Coen Verschuren, a senior scientist at Philips Research Laboratories. "The NA 0.45 objective has the strength of a CD lens and is made with conventional technology."
The SIL is made from spheres of 1 mm in diameter that are ground and polished into a truncated cone shape (essentially a cone with a flat tip, out of which the laser light exits the SIL). According to Verschuren, the challenge is manufacturing the SIL to an exact thickness. "If you want to have a good lens, the thickness tolerance is only 0.2 μm," he said. "This kind of accuracy cannot yet be reached in mass-production of lenses today."
Instead, Philips uses interferometry to make an accurate measurement of the SIL's thickness. If it is not within the required tolerances, both the polishing and interferometry steps are repeated. While this is not ideal for mass-production, it suits work at a proof-of-principle stage.
Having produced high-precision optics, the next problem was how to accurately assemble both lenses. "We centre the SIL and glue it into a lens holder, then use interferometry to position and glue the objective lens with micron accuracy," explained Verschuren.
And, as if producing this lens was not enough of a challenge, the next step was to come up with a way to hold the lens just 30 nm from the disc. Several years ago, Philips started this work in a cleanroom environment and used a slider-based servo to control the motion of the lens holder. "We wanted to make a system for removable discs, but even in a cleanroom the slider was not sufficiently robust," explained Verschuren. "If the slider hit something on the disc, it would often be damaged beyond repair."
Learning from this experience, and motivated by similar work being done at Sony, Philips changed tack and began working in a standard laboratory, using an actuator found in a DVD system. This proved to be a smart move as the actuator was much more robust than the slider and was capable of maintaining the 30 nm air gap.
Philips uses a term called the gap error signal (GES) to control the air gap between the lens and the disc. The air gap is measured optically by detecting the polarization of light reflected from the disc.
In what Philips terms the "pull-in process", the servo system brings the lens close to the disc from a relatively long distance, without hitting it. At this point, the GES starts changing rapidly and the system switches to an air-gap control mode. To account for the waviness (non-flatness) of the plastic disc and to ensure that the lens never collides with the disc, the GES is continuously fed into the servo system, which reacts by moving the lens.
"Our goal is to maintain the air gap with an accuracy of ±2 nm, and we can achieve this easily with the actuator," said Verschuren. "We used a Blu-ray substrate from our pre-production lines in our first tests. We can deal with rotation speeds of 3000 rpm on a disc and still maintain the accuracy of the air gap to a couple of nanometres."
Bearing in mind that this work is done in a standard laboratory, which is plagued by dust, air conditioning and vibrations, this is an impressive result.
"Dust is the least of our worries because it is too big," said Verschuren. "Fingerprints can be more of a problem in the long run. The servo system is sufficiently robust that it can push the lens through a fingerprint, but the oil from it will collect on the lens, which is a problem." He adds that the data is written onto the top surface of the disc, so even a small scratch results in data being lost.
To get around contamination issues and the tight tolerances placed on the thickness of the SIL, Philips came up with an alternative NA 1.45 system. This approach has the distinct advantage that the disc is covered by a polymer layer three microns thick.
Data is written through the cover-layer into a layer that has a capacity of 75 GB. "You also have the possibility of going to a multiple-layer system, so, despite the lower NA, you can get the same capacity, or maybe more than the NA 1.9 system," commented Verschuren.
The NA 1.45 lens system is made up of a hemispherical SIL with a radius of curvature of 0.5 mm and an NA 0.7 objective. While the SIL is created using the same glass spheres and method as the first surface system, the NA 1.45 option not only benefits from a hemispherical SIL (as opposed to a super-hemispherical SIL in the NA 1.9 system), but also a thickness tolerance as large as 10 μm.
Although Philips is not planning mass-manufacturing at the moment, this improvement in tolerance would make the lens far cheaper to fabricate. It also comes within the realms of what is commercially available now.While these factors make a big difference to the practicality of the system and its robustness, one disadvantage is that the NA of the lens needs to be smaller than the refractive index of the cover-layer. This means that the cover-layer cannot be produced by sputter deposition, where Verschuren says a refractive index of around two is possible.
"You have to go to spin-coatable polymers that have a refractive index of between 1.5 and 1.6, and this is why we have an NA of 1.45," he said. "If chemical companies could produce spin-coated polymers with higher refractive indices, then the NA of the lens would increase and the capacity go up."
Another change on the cards is replacing the current glass objective lens with an equivalent plastic lens. "Most of the mass is in the objective lens," said Verschuren. "If the moving mass is smaller, then the servo has less work to do and that helps a lot. We simply haven't done this yet, but I am sure it will happen in the future."
At the moment, Philips is working on both the NA 1.9 and NA 1.45 configurations in parallel. "We have been promoting the cover-layer approach for reasons of improved robustness," said Verschuren. "The NA 1.9 system works amazingly well in the lab, but we think the cover-layer is more practical."
Although these results highlight that near-field optical recording is feasible, it may be a while before it hits the commercial world. Philips will release its first Blu-ray product - a recorder for a PC drive that can record Blu-ray, DVD and CD - in 2006. After that, market uptake and the adoption of Blu-ray will affect the launch of any post-Blu-ray technology. Despite the uncertainty in time-scale, it is clear that near-field optical recording is robust and will be waiting in the wings to take over when consumers demand extra capacity.