World News

Fig. 1 Simone Anders, a team leader at the Advanced Light Source, uses a newly commissioned photoemission electron microscope with 20 nm of spatial resolution to study orientation of magnetic domains in a computer memory-storage device.

Fig. 1 Simone Anders, a team leader at the Advanced Light Source, uses a newly commissioned photoemission electron microscope with 20 nm of spatial resolution to study orientation of magnetic domains in a computer memory-storage device.

A newly completed photoemission electron microscope with 20-nm resolution went on-line in April for outside users at the Advanced Light Source (ALS) in Lawrence Berkeley National Laboratory (LBNL; Berkeley CA). The LBNL and Arizona State University (Tempe, AZ) researchers have also started design work toward the next resolution goal of 2 nm.

The current device (called PEEM2) was completed last summer after a three-year effort, under a Cooperative Research and Develop ment Agreement (CRADA), by re searchers at LBNL, IBM (Almaden, CA), and Arizona State Univer sity, according to Simone Anders, a team leader in the ALS Experimental Systems Group headed by Howard Padmore. The CRADA effort was funded 50% by the US Department of Energy and 50% by IBM.

The IBM funding was motivated largely by an interest in nonvolatile magnetic memory materials and computer hard-disk drives, Anders said. "PEEM2 can distinguish the magnetic properties of individual layers in multilayer structures," she said. "Magnetic multilayers may soon be used in magnetic memories, replacing electronic memory chips for various applications."

To characterize a sample, the researchers use electrons emitted from it by x-ray or extreme-ultraviolet-light excitation from the ALS. Then PEEM2 focuses the emitted electrons on a phosphor screen. The visible light image that PEEM2 produces on the phosphor screen is acquired by a charge-coupled-device camera.

A primary factor in the usefulness of PEEM2 is that the light it receives from the ALS can be selected to be either linearly polarized or right or left circularly polarized light. Linearly polarized light is useful for determining the orientation of bonds in thin films, and circularly polarized light can be used to determine the circular dichroism in regions of a magnetic sample. So with a beam tuned to a characteristic absorption edge of a magnetic material, such as cobalt, PEEM2 can look at the surface of a computer disk and distinguish the orientation of its magnetic bits.

Currently, IBM experiments take up 50% of available time on PEEM2, and the other 50% is available for outside users that range from local researchers at LBNL and the University of California-Berkeley campus to research projects from as far away as North Carolina, Canada, and Germany. In addition to the magnetic applications mentioned, the device has also been used to examine phase separation in polymers that determines the characteristics of end products such as paint and photographic film.

"Because it can distinguish between chemical bonds, PEEM2 can map areas of diamond-like bonds versus areas of graphite-like bonds in thin films of amorphous carbon, which are of potential use in flat-panel displays," Anders said. "And because PEEM2 can distinguish molecular orientations, it can be used to examine the mechanisms of polymer alignment," which are important factors in the development of liquid-crystal displays.

The original team of Anders, Joachim St?hr of IBM, Michael Scheinfein of Arizona State University, and a team of ALS engineers led by Ron Duarte hope to improve the resolution by an order of magnitude, down to 2 nm in PEEM3, during the next year and a half, Anders said. Currently the 20-nm resolution of PEEM2 is close to the theoretical limit for a photoemission electron microscope. But the team hopes to bypass that limit by designing an electrostatic mirror and magnetic separator to correct for the resolution-limiting chromatic aberration caused by the varying energy distribution of emitted electrons.

The higher resolution is expected to significantly improve the characterization of magnetic materials. "The width of a domain wall between two magnetic domains or the size of small magnetic domains might be in the nanometer range that we cannot see right now," Anders said. She expects PEEM3 to actually show the interface between bits in a single magnetic layer. "We will continue on magnetic materials and also on polymers, because with nanometer resolution we get into almost the molecular size."

BP Amoco goes solar at service stations - BP Amoco (London, England) announced in April that about 200 of its service stations worldwide will incorporate solar power. As part of the company`s Plug in the Sun program-the largest single project of its kind ever undertaken-solar panels will be installed on the roofs of many of its newest petrol stations (see photo).

During the first phase of the two-year, $50 million, 3.5-MW project, up to 400 solar panels will be installed on the canopy over the fuel pumps at service stations located in 11 countries. BP Amoco claims this will save around 3500 tonnes of carbon dioxide emissions every year. As a result of this project, the company will become one of the world`s largest users of solar power. It is already one of the world`s largest manufacturers of solar cells and modules.

Solar electricity will help meet the power needs of all new BP Amoco service stations to be built in the UK, Australia, Germany, Austria, Switzerland, the Netherlands, Japan, Portugal, and Spain. Solar installations will also be incorporated into prototype sites in France and the USA as part of an extended pilot program. The target countries have been chosen for a number of reasons, including the ease of connecting to the national electricity transmission systems, industry infrastructure, and light levels.

The solar installation at each site will have a maximum power output of 20 kW. The level of power actually generated will vary from site to site. But at each, the solar panels on the canopy above the pumps will generate more clean energy than is consumed by the site`s lighting needs and the power requirements of the pumps below. The installations, which will be connected to the local electricity networks, will allow any excess electricity to be exported during the day and the shortfall imported at night. A 40-kW solar system will also be installed at three new office buildings as part of the redevelopment of BP`s facility in Sunbury, England.

Plug in the Sun unites two of BP Amoco`s major business areas. The company is the world`s leading solar energy company, having recently invested $45 million in solar power with the purchase of the 50% share of Solarex that it did not already own. The new company BP Solarex is headquartered in Maryland, USA. It will build on the current business activities of Solarex and BP Solar and has manufacturing sites in four countries and combined sales of more than $150 million.

"Our own use of solar power is an example of BP Amoco`s commitment to tackling the issue of climate change," says BP Amoco chief executive Sir John Browne. "Not only will BP Amoco be one of the largest producers of solar photovoltaic cells in the world, but it will also be one of the largest single users of solar power. By installing solar panels at such a large number of sites across the world we will also learn and add to expertise in handling issues of grid connection, contribute to the standardization of equipment, and drive down costs for all of our solar customers."

Multimode fibers can link free-space systems - While some free-space optical interconnect systems have used multimode fibers for their output, multimode types have rarely, if ever, been used for the input-but there could be several advantages to doing so. While working at NEC Research Institute (Princeton, NJ), engineers David Neilson and Eugen Schenfeld developed a prototype system that does just that. Neilson, who is now with Lucent Technologies (Holmdel, NJ), reports that multimode fibers have an advantage over single-mode fibers in both ease of alignment at connectors and ease of coupling at sources and detectors. Another significant factor influencing their use is the large installed base of 50- and 62.4-?m-core fibers for local-area-network (LAN) applications. The ability to integrate multimode fibers with free-space optics would allow the construction of switching and routing fabrics for LANs.

A major consideration when designing a free-space optical system linked by multimode fibers is the large numerical aperture of the fibers and the need for the optics to accommodate multiple spatial modes. To resolve this issue, the scientists modeled the beams from a multimode fiber in free space using the M2 parameter. This allowed treating the fibers as pseudo-Gaussian beams, but with larger waists. For example, at a wavelength of 850 nm, a fiber with a 62.5-?m core and an NA of 0.275 could be considered as a Gaussian beam with M2 = 32.5. According to Schenfeld, who is now with IBM (Yorktown Heights, NY), it was necessary to use microlenses with focal lengths of 2-5 mm to provide enough space between the relay lenses to allow inserting a beamsplitting surface for routing signals in the optical domain. This led to the need to reduce the angular beam divergence from that present in the fiber, which in turn produced an expansion of the waist size. Another benefit was that the system now had increased tolerance to lateral misalignment.

Fig. 1 Experimental free-space optical relay, which relies on off-the-shelf devices integrated with easy-to-fabricate plastic optomechanical components, could open up opportunities to construct large repeaters and switches for multigigabit Ethernet applications.

To build the relay modules, the researchers machined transparent poly-methyl methacrylate blocks into 25.4-mm-long cubes, to which plastic microlens arrays were cemented. According to Neilson, the microlenses, which were compression-molded in a 1.5-mm-thick acrylic sheet, had a 500-?m pitch with a 5.11-mm focal length at 850 nm. The lenses also had a multilayer coating with a reflectivity of less than 0.5% at that wavelength. Modules were joined with steel dowel pins. The 4 ? 4 fiber arrays were made up of 62.5-?m-core fibers on a 500-?m pitch. They had lenses with 600-?m focal lengths attached one focal length away from the end of the fiber. This spacing allowed the beam to expand for launch into the free-space optic relay (see figure on p. 18).

Neilson reports that the experimental system produced a data transmission rate of 2.5 Gbit/s from multimode fiber to multimode fiber over the plastic free-space optical relay.1 Equally important, it did so with commercial components when possible, including vertical-cavity surface-emitting lasers. The engineers designed other components so that they could be manufactured by mass-production techniques such as plastic injection molding and then could be aligned passively. If taken to commercialization, the low-cost, high-density interconnects demonstrated by Neilson and Schenfeld could enhance fiber-based LANS by integrating switching and routing functions.

ULTRAFAST LASERS

Tabletop laser fuses nuclei

For many people, the idea of controlled nuclear fusion conjures up visions of inexhaustible energy. Although the goal of "breakeven" may only be achieved-if ever-by massive, billion-dollar projects, fusion on a smaller scale may also have its benefits, both scientific and other wise. By exploiting properties of atomic clusters of deuterium, re searchers at Lawrence Livermore National Laboratory (LLNL; Livermore, CA) have used a tabletop femtosecond laser to produce significant amounts of fusion within the confines of a room. The resulting high yield of neutrons from their apparatus may, if enhanced further, result in a compact neutron source with many practical uses.

The researchers form the atomic clusters by expanding a deuterium gas jet cooled to -170°C into a vacuum, causing the gas to condense into clusters about 50 ? in diameter, each containing about 1000 atoms. Using a Ti:sapphire laser, they bombard the clusters with 120-mJ, 35-fs pulses at a wavelength of 820 nm and a repetition rate of 10 Hz (see photo on p. 26). In the resulting ionization, electrons undergo rapid collisional heating for the 1-ps period before the clusters disintegrate into plasma. The intense laser light heats the electrons to high energies, driving the electrons away from the clusters and producing a strong radial electric field that accelerates the cluster ions. The laser energy is thus transferred to the atoms with nearly 90% efficiency, with atoms receiving energies of up to tens of kiloelectronvolts-enough for fusion to occur. This contrasts with a mere 5% efficiency of energy transfer into an uncooled deuterium gas jet that contains no clusters.

Each deuterium-deuterium (D-D) fusion event produces a neutron with 2.45 MeV of energy. Because the speed of a neutron is proportional to its energy, a time-of-flight measurement provides a good estimate of the energy spectrum of the ejected neutrons. By setting up neutron detectors at 1, 2.5, and 3.2 m and measuring the neutron flux as a function of time elapsed from the laser pulse, the researchers detect a sharp peak of neutrons emitted at a 2.45-MeV energy-exactly that produced by D-D fusion. Approximately 1 ? 104 neutrons per laser pulse are produced, which translates into about 1 ? 105 neutrons per joule of incident laser energy, a yield comparable to that from inertial-confinement-fusion implosions produced by large lasers such as the Nova laser at LLNL.

One potential way to further increase the neutron yield of the tabletop-fusion apparatus is to increase its ion energies by driving the clusters with two differing excitation wavelengths, according to Todd Ditmire, project leader. "Because the fusion cross section increases very rapidly with energy, this increase would yield more fusion events," he explains. "The cross section for D-D fusion really begins to turn on above 10 keV and increases quite nonlinearly up to a few hundred kiloelectronvolts."

In addition, deuterium-tritium (D-T) fusion produces neutrons with an even higher energy of 14.1 MeV. These and other options could lead to a high-flux neutron source suitable for materials science and neutron radiography. "We think the most promising application at this point is to use the neutrons from

D-T fusion for studying materials under very high neutron fluxes, [such as] materials that would one day be used in an actual fusion reactor," says Ditmire. "We will need to increase the neutron yield by many orders of magnitude if such an application is to be realized. However, a compact, affordable avenue to producing lots of fusion neutrons for such studies does not exist right now, so if this source could be scaled up, it could represent a viable option."

Fusion plasma at the exit of a cryogenically cooled jet of deuterium forms when atomic clusters are ionized by femtosecond laser pulses.

Large, light mirror moves Airborne Laser forward

A lightweight mirror for the US Air Force Airborne Laser (ABL) program has been delivered to Contraves Brashear Systems LP (Pittsburgh, PA). The company will spend more than a year polishing it to the required optical quality. The mirror, 62 in. in diameter and 8 in. thick, was built by Corning Incorporated (Corning, NY) to be light enough that it could be mounted on an airplane. Corning built a mirror with a core weighing only 330 lb by machining with water jets. Had it been solid, a mirror of the same size would have weighed nearly 2000 lb (see photo).

Once polishing is complete, the mirror will be mounted in a turret ball on the front of a modified 747 aircraft. The movable turret will allow the airplane`s crew to aim the mirror and direct a high-energy beam from the Air Force`s chemical oxygen iodine laser at fast-moving enemy missiles hundreds of kilometers away. The ABL team, managed by Boeing (Seattle, WA), is scheduled to test a "shoot-down" of a missile in 2003, with a development and manufacturing program to begin as early as 2004 (see Laser Focus World, May, 1999, p. 108).

"The lightweighting of this mirror represents an important technological achievement that enables us to meet the unique environments associated with an airborne platform," said Paul Shattuck, ABL program manager at Lockheed Martin Missiles & Space (Sunnyvale, CA), which is developing the control system for the ABL.

In January, a $2.5 million telescope built by Contraves Brashear was installed at a new Air Force Research Laboratory site at White Sands Missile Range (White Sands, NM) for a system designed to track and destroy missiles. That 1-m telescope will be used to send and receive laser light between it and Salinas Peak, 35 miles away. Researchers will measure how the atmosphere distorts the beam and use deformable mirrors to compensate for the distortions. Like the airborne mirror, the telescope can be aimed-it moves up and down 5° and revolves through a full circle. In tests scheduled to start in June, the system will use a 300-W tracking laser, a 50- to 200-W adaptive optics beacon laser, and a 40-W scoring laser.

The atmosphere in which these tests will be conducted, at elevations from 8000 to 9000 ft, is denser than the air the ABL will have to deal with, as it flies above 40,000 ft. But the researchers believe the data they collect will be applicable to the flying system as well.

Neil Savage

Tunable VCSEL provides 2-mW CW output

Small size, capability for cost-efficient high-volume production, and hop-free tuning capability make VCSELS, such as this 1550-nm tunable device, ideal for fiberoptic communications.

Tunable lasers may become enabling technologies for a new generation of communications systems called all-optical networks. In such networks, routing and switching functions are performed in the optical domain by both tunable lasers and filters-for a cost some say is one-tenth that of electronic switching systems. This alone is producing fierce competition among companies developing vertical-cavity surface-emitting-laser (VCSEL) technology.

For example, CoreTek Inc. (Burlington, MA) has demonstrated a 1550-nm VCSEL with a maximum of 2 mW of tunable CW single-mode output. According to its CEO, Parviz Tayebati, this is reportedly the first microelectromechanical VCSEL that is continuously and rapidly tunable throughout the dense-wavelength-division-multiplexing (DWDM) wavelength band of 1514-1620 nm. In the future, the firm predicts, such devices will deliver more than 6 mW of power with tuning ranges greater than 100 nm (see figure).

CoreTek originally developed the laser technology for defense applications under contracts supported by the Ballistic Missile Defense Organization and the Defense Advanced Research Projects Agency. According to the firm`s newly appointed chief technology officer Daryoosh Vakhshoori, microelectromechanical VCSELs promise mode-hop-free, single-mode, and wide-wavelength-tuning operation. These characteristics make them ideal transmitters for DWDM systems.

"The need for tunable lasers that can be set at designated wavelengths with speed and precision is driven by the

continuing demand for more-closely packed transmission channels to achieve more bandwidth," says Tayebati. As the number of transmitted channels increases from 8 to 400 or more over the next five years, he believes manufacturing of scalable and cost-effective DWDM systems based on lasers with fixed wavelength will be a serious challenge. The cost of stocking lasers will increase proportionally. Tunable lasers can relieve the manufacturing bottlenecks in high-channel-count DWDM systems, as well as reduce their costs. Ultimately, this could accelerate the expansion of high-bandwidth networks into metro and local area networks.

CoreTek`s 1550-nm tunable device is a microelectromechanical structure in which the top curved mirror of the laser cavity is displaced by voltage-induced electrostatic force. The cavity is designed for operation under optical pumping at 980 nm. When tuning voltage is ramped from 0 to 39 V, 50 nm of continuous tuning is possible.

In the system, a 1.55-?m strain-compensated multiple-quantum-well system provides the gain for the light that is bouncing back and forth between the bottom flat mirror and the top curved mirror to produce lasing action. According to Vakhshoori, such a stable resonator results in extremely low cavity output and also defines and stabilizes the size of the fundamental mode under different injection and wavelength-tuning conditions.

"Our device`s lasing spot size is

7 ?m," he notes. "Control of this dimension is essential for long-term stability and efficiency of lasing power coupled into single-mode fiber. In our case, direct fiber butt coupling would result in less than 1-dB laser-to-fiber coupling loss, he says.

"The advantage of using an optical pump as the input power source is that the overlap of the pump source with the fundamental mode of our half-symmetric cavity guarantees lasing of TEM00 mode. This is the only spatial mode of interest in VCSELs and is necessitated by the requirement of stable and efficient single-mode laser-fiber coupling."

Paula M. Noaker

Call for Entries

Commercial Technology

Achievement Awards

see p. 109

ASTRONOMY

Tiny wavelength shifts reveal giant planets

Doppler measurements of the star Upsilon Andromedae show subtle oscillations in velocity, revealing the existence of three Jupiter-sized planets, the first planetary system ever discovered around a Sun-like star. Astronomers now have spectrometers that are sensitive enough and that have been operating long enough to let them see the previously unseeable-the tiny wobbles in a star that betray the presence of multiple planets. Teams from San Francisco State University (SFSU; San Francisco, CA) and the Harvard-Smithsonian Center for Astrophysics (CFA; Cambridge, MA) each used a different spectrometer to come up with results similar enough to convince them of the existence of the giant planets orbiting Upsilon Andromedae, a star 1.2 times the mass of our Sun, lying about 44 light years away in the direction of Andromeda (see image).

In 1996, astronomers Geoffrey Marcy of SFSU and Paul Butler of the Anglo-Australian Observatory (Epping, NSW, Australia) discovered a planet of at least three-quarters the mass of Jupiter. Their data showed it orbiting the star every 4.6 days at a distance of 0.06 astronomical unit (AU)-one AU equals the average distance of Earth from the Sun, about 93 million miles. More observations gave them enough data to realize that a second planet, and then a third, were also tugging on the star.

The middle planet is at least twice the mass of Jupiter and orbits every 242 days, at a distance equivalent to the distance of Venus from the Sun. The farthest planet is at least four times the mass of Jupiter and orbits every 3.5 to 4 years at 2.5 AU from the star.

The SFSU team used the Hamilton Echelle Spectrometer, which collects light from the 3- and 0.6-m telescopes at the Lick Observatory (Mt. Hamilton, CA). The CFA team used its Advanced Fiber Optic Echelle (AFOE), a fiber-coupled, bench-mounted spectrometer attached to the 1.5-m telescope at the Whipple Observatory (Mt. Hopkins, AZ). Echelle spectrometers pass light through a grating, correction lenses, a mirror, and a field flattener to produce a square-shaped spectrum on a charge-coupled-device camera (see Laser Focus World, Jan. 1998, p. 30). Both instruments use an iodine absorption cell, allowing scientists to calibrate the wavelengths, and both have a resolution of about 50,000. By measuring Doppler shifts in the light, researchers can see changes in the velocity of the star as the orbiting planets pull it slightly in one direction, then another.

Because even giant planets are so much smaller than stars, the pull they exert is tiny-Jupiter causes a wobble in the Sun`s velocity of only 12 m/s, too small for astronomers to measure if they were observing from Upsilon Andromedae. And astronomers need to watch more than one orbit to get an accurate sense of the planet`s motion, stretching observations over years. The Lick Observatory observed the star 89 times-five times in the mid-1980s and starting again in 1992. It wasn`t until 1994, when the Hamilton spectrometer was upgraded, that the resolution and wavelength coverage were enough to show the wobble. The AFOE, which observed the star 52 times, did not achieve the required precision until 1995.

Timothy Brown of the National Center for Atmospheric Research (Boulder, CO) said it is only the increasing precision of spectrometers that has allowed astronomers to measure the Doppler shift, the technique that has shown all 20 of the planets discovered so far.

Neil Savage

Blue light originates from a polymer solution placed between indium tin oxide-coated glass sheets and electrically biased. Onset of photon emission occurs at a bias of 6 V.

Electricity makes light of the solution

Researchers at the University of California, Los Angeles (UCLA; Los Angeles, CA), and SRI International (Menlo Park, CA) have demonstrated the production of strong electrically generated luminescence from polymer solutions encapsulated between two glass sheets in a closed-cell configuration. The resulting low-voltage device has potential as an illumination source and, when pixelated, as a flat-panel display.

At the heart of the solution-based light-emitting device (SLED) is a fluid made of poly[9,9-bis(3,6-dioxaheptyl)-fluorene-2,7-diyl], or BDOH-PF, dissolved in dichloro-benzene to a concentration ranging from 1% to 20%. When subjected to an electric field, BDOH-PF emits light through a process called electrogenerated chemiluminescence; the higher the concentration of BDOH-PF in the solution, the more light is produced. The solution has a photoluminescent quantum efficiency of 77%. To make a SLED, a quantity of BDOH-PF is mixed with glass beads 1-2 ?m in size and a few drops of the solution are placed on a glass substrate coated with indium tin oxide (ITO), a transparent conducting film. Another ITO-coated piece of glass is positioned on top and pressure is applied, squeezing the extra polymer solution out and creating a BDOH-PF film of the same thickness as the bead diameter. The dichloro-benzene solvent evaporates along the perimeter of the SLED, creating a seal solid enough for research purposes. The whole procedure takes a few minutes. Due to its construction, a SLED is naturally free of pinhole defects.

When emitting, the device appears blue (see photo). Its emission spectrum lies entirely within the visible, cutting on sharply at 420 nm and cutting off more gradually between 500 and 600 nm. Quantum efficiency of a SLED in units of photons per electron is estimated to be 1%. Device lifetime, currently at several hours, is limited by residual evaporation of the organic solvent through the dried seal. "We are still searching for a proper way to seal the device," says Yang Yang, assistant professor at UCLA. "Epoxy is one possibility."

Typical SLEDs are about 3 ? 3 mm square, according to Yang, although they have been made up to 20 ? 20 mm in size. "It is quite simple to make large-area devices," says Yang. The brightness of a SLED ranges from 20 cd/m2 for an electrical bias of 10 V to 100 cd/m2 for a bias of 20 V-although when the spacing of the glass sheets is widened to

2 ?m using larger glass beads, the luminescence virtually disappears. The transparency of a SLED is above 90% across the visible spectrum, allowing the devices to be vertically stacked. Turn-on time is between 20 and 40 ms, comparable to that for liquid-crystal displays (LCDs). The researchers want to improve both device efficiency and turn-on speed.

A potential early use for the technology is in efficient flat-panel backlighting for small LCDs, for which the traditional fluorescent lamp used in larger LCDs is too bulky. "However, our long-term goal is to make displays," says Yang. He notes that his group is obtaining patents on pixelated SLED technology. In addition, the researchers are aiming to develop electrically pumped solution-based polymer lasers.

John Wallace

OPTICAL COATINGS

Multilayer coatings boost grating efficiency

High-dispersion diffraction gratings are required for the observation of spectral line profiles and Doppler shifts in solar, astrophysical, and laboratory radiation sources. According to researchers John Seely and Jack C. Rife at the Naval Research Laboratory (NRL; Washington, DC), this type of grating may be part of the payload on a future satellite developed to study the dynamics of the Sun`s corona, photosphere, and transition regions. This high-dispersion grating stands out from the crowd because, in addition to traveling into space, it may be covered with a dual-bandpass molybdenum/silicon (Mo/Si) multilayer coating instead of the more-typical gold coating.

The NRL researchers and colleagues have demonstrated what is reportedly the first enhancement of the normal-incidence efficiency of a grating with 4800 grooves/mm in the extreme ultraviolet region by use of such a multilayer coating. Other study participants are with the National Astronomical Observatory and Tokyo Metropolitan University (both in Tokyo, Japan), as well as SFA Inc. (Landover, MD).

The scientists studied the normal-incidence efficiencies of two replicas of a ruled master grating, one with the multilayer coating, the other with a gold coating.1 The master grating had 4800 grooves/mm and a nominal blaze angle of 3.7°. Each replica grating had a grooved area of 40 ? 40 mm and an aluminum surface upon which coatings were layered.

Seely reports that both replicas, as characterized by atomic force microscopy, had bumpy surface features larger than the nominal height of their grooves. Microroughness values were 12-20 ? rms in the 5-18 ?m-1 frequency range for the grating with the multilayer coating and 22-32 ? rms for the gold grating.

The normal-incidence efficiencies of the replicas were measured with synchrotron radiation in the 12.5-32.5-nm wavelength region. The peak re flectance of the multilayer coating was found to be 22% in the first Bragg order near 32.5 nm and 28% in the second Bragg order near 12.6 nm. The grating`s peak efficiency was 2.6% in the first diffraction order near 22.5 nm and 0.3% in the second diffraction order near

12.5 nm. At both 22.5 and 12.5 nm, the grating with the multilayer coating had higher normal-incidence efficiencies than the efficiencies of the gold-coated grating (see figure). The efficiency of the Mo/Si-coated grating was basically enhanced by the relatively high reflectance of that coating in the wavelength regions corresponding to its first and second Bragg orders.

The large surface roughness and groove irregularities, however, reportedly lowered efficiencies of both replicas. For instance, the Mo/Si-coated replica grating recorded efficiencies smaller than those that would be produced by multilayer gratings fabricated using the holographic ion-etched blazed technique. The 4800-grooves/mm master grating was an initial test ruling, though; researchers expect subsequent Mo/Si-coated gratings to have improved microroughness and groove profile regularity.

Paula M. Noaker

REFERENCE

1. John Seely et al., Appl. Opt. 38, 1920 (April 1999).

MATERIALS RESEARCH

U of A launches lab for optical materials

TUCSON, AZ-The University of Arizona held a dedication ceremony last month for a new Optical Materials and Technology (OMAT) laboratory that will eventually be included within a 30,000-sq ft University Research Center. The re search center is under construction at the University of Arizona Science and Technology Park located southwest of the main campus.

The OMAT staff is currently headed by Joe Perry, a physical chemist formerly with the NASA Jet Propulsion Laboratory, and Seth Marder, a synthetic chemist formerly with the California Institute of Technology (both in Pasadena, CA). The OMAT laboratory has already begun to work closely with the Optical Sciences Center on the main campus and will focus primarily on the development of materials based on organic molecules and polymers to facilitate the storage, display, and transfer of information.

For instance, while still at CalTech, Marder collaborated with Bernard Kippelen and Nasser Peyghambarian at the Optical Sciences Center on the development of infrared photorefractive polymers for biological imaging.1 The researchers will also use advanced laser technology to manipulate the optical properties of these materials and to pattern them into three-dimensional micro device structures. Along these lines, Marder and Perry have already collaborated with a number of researchers from various institutions on the design of organic molecules with two-photon absorption cross sections large enough to enable efficient 3-D microfabrication processes (see figure). 2,3

During a technical presentation on this work last month at the spring meeting of the Materials Research Society in San Francisco, CA, Marder said that the switchable nonlinear absorption properties of such materials have potential applications in optical limiting. The University of Arizona group will focus on commercial applications, however, that include 3-D microfabrication, 3-D optical memory, and biomedical applications such as imaging within tissues and photodynamic therapy, he said.

In addition to the multidisciplinary focus that the laboratory will bring to the design and synthesis of new materials, the various laser and optical systems at OMAT will also provide a vertical capability for testing, developing, and production of prototype materials. The vertical integration of the lab and its staff will provide "a capability to go from discovery and research to applicaions not found in many other multidisciplinary projects," Perry said. "Part of the educational process here is that measurement capabilities are included as a factor in the design of molecules," added Marder.

Richard Powell, University of

Arizona vice president for research and former head of the Optical Sciences Center, has high hopes for the OMAT lab. "[It] will bring world-class optical researchers to the University of Arizona Science and Technology Park," he said. "Its presence supports Tucson`s growing international reputation as `optics valley` and augments the university`s established expertise in this area." The 1340-acre Science and Technology Park includes nearly 2 million sq ft of laboratory and office space that currently houses 15 high-technology companies including IBM, Raytheon Missile Systems, Edmund Scientific Co., and NP Photonic Technologies.

Hassaun Jones-Bey

REFERENCES

1. B. Kippelen et al., Science 279, 54, (2 Jan. 1998).

2. M. Albota et al., Science 281, 1653 (11 Sept. 1998).

3. B. H. Cumpston et al., Nature 398, 51 (4 March 1999).

Charts illustrate the difference in efficiency of the multilayer (top) and gold-coated high-dispersion gratings (bottom), with each measured at a 10° angle of incidence and a wavelength of 225 ? . The outside (M<0) diffraction orders are identified.

Among 3-D microstructures produced by two-photon induced polymerization are a photonic bandgap structure (a, with magnified top view of the structure in (b)), a tapered waveguide structure (c), and an array of cantilevers (d).

MEDICAL IMAGING

FDA approves faster ophthalmic imager

A new system that produces a single, high-resolution color picture of almost the entire retina was granted US Food and Drug Administration (FDA) approval last February. Produced by Optos plc (Dunfermline, Scotland), the Panoramic 200 is the product of six years of research and development.

Conventional retinal-imaging techniques typically capture only a small portion of the retina-about 5%. The current standard techniques, such as direct or indirect ophthalmoscopy, also require dilation of the pupil, are very time-consuming, and limit office productivity. More-specialized retinal-imaging methods, such as fundus photography, also have drawbacks-field of view is limited, patient imaging is uncomfortable, and a trained operator is required. The Panoramic 200, however, can image the full retina in just 0.25 s without the need for dilation, contact with the cornea or scleral depression, or stressful and potentially harmful levels of illumination (see photos).

How it works

The device is based on red- and green-emitting lasers-each gives information about different structures in the eye. Light from the lasers is combined into a single beam that is scanned quickly in the vertical and horizontal axes. The imaging system is key to getting the high-resolution images of the entire retina at low light levels and without pupil dilation. A patented, ultrawide-field, ellipsoidal mirror is at the center of a virtual scanning system that allows the retina to be scanned over a 200° ? 200° area (measured from the center of the eye). The scanning laser light passes through the pupil and is pivoted about the virtual point, which is coincident with the eye`s pupillary point. It is the small size and stability of this virtual point of scan that gives the system its high-quality imaging properties.

About 1500 ophthalmologists in the UK and the USA have expressed interest in the new instrument, which will be on sale in the USA later this year. Meanwhile, the company recently opened an office in Boston, MA, to service the US market.

Anderson believes that the Panoramic 200 could revolutionize eye care. "We believe we have a ground-breaking new product in terms of the effect it will have on the way eye care is provided. The ease with which the new system can detect retinal disease at an early stage means that vision loss is reduced, and any intervention is less traumatic. As health-care systems move from treatment to prevention, patients demand a more-friendly basic examination, and eye-care professionals need an easy, quick, and complete view of the retina. The Panoramic 200 fulfills this need."

More information on Optos can be found on its Web site: www.optos.com.

Bridget R. Marx

Japanwatch

Incorporating news from O plus E magazine, Tokyo

Organic dye and excimer

laser pattern silica glass

FIGURE 1. A layer of pyrene (dissolved in

then ultraviolet (UV) light from an excimer laser is transmitted through a mask and a lens that

projects a reduced version of the mask pattern on the silica. The laser light heats the boundary between the glass and the pyrene solution,

causing the surface of the glass to melt in the mask pattern.

WAKO-Shun Ou and colleagues at the National Institute of Materials and Chemical Research have created an optical microfabrication process that uses the light-to-heat energy transfer of the organic dye molecule pyrene to create patterns in silica glass.

Pyrene has long been in use, but only recently have its absorption properties been thoroughly investigated. Pyrene absorbs 248-nm laser light exceedingly well, and it efficiently converts the light energy into heat by inducing molecular vibrations. In other words, it functions as a laser-excited "molecular heater" (see Fig. 1).

The physical and chemical details of this mechanism are not fully understood; the researchers believe, however, that there is a repeated transition from light to heat between two of the electronic excitation levels of the pyrene. The process has created features with a depth of 3.5 ?m while leaving the pyrene-covered glass surface smooth and free of cracks (see Fig. 2).

The pyrene solution used in this experiment is highly concentrated and absorption is so high that the light from a krypton fluoride laser can penetrate only about 1 ?m into the solution after it passes through the glass. Any heat produced by the process is, therefore, concentrated near the surface of the glass. The researchers estimate that the temperature around the molecules becomes as high as 2000 K. The melting point of silica glass is 1700 K.

Although many methods of microfabrication have been used for silica glasses, this method may prove superior because atmospheric fabrication is possible, it is a single-step method, and the laser power required is 1/40 of that required for direct excimer laser abrasion. In addition, the technique causes very few cracks or other imperfections, and there is good depth control (to within nanometers). The process can be used for sapphire, calcium fluoride, polymer fluorides, and other materials.

The group is currently investigating more closely the theoretical basis of this process. It is also aiming to make resolution finer while expanding the area that can be covered per process. The group believes that as far as resolution is concerned, submicron levels will be possible.

Courtesy of O plus E magazine, Tokyo

FIGURE 2. Scanning electron microscope reveals pattern on the pyrene-covered glass

surface after it is exposed to 248-nm light (left). The depth of the features is approximately 3.5 ?m with a clear edge. The surface of the glass is very smooth, and there are no cracks (right).

BIOMEDICAL SCREENING

Scattering protein shows disease potential

Red laser diode beam passing through a heated cell containing urine produces detectable

scattering if protein is present.

OSAKA-Matsushita Electric Industrial Co. Ltd. has developed an optical method of detecting protein in urine that does not require disposable materials such as test papers or fluids. Protein exists only in trace amounts in the urine of healthy people; elevated levels of protein in urine, however, indicate a higher risk of kidney disease. This new uric-protein-detection technology can be automated to fit into home toilets, thus making regular measurement possible.

Current uric-protein-detection kits for home use require test papers, which makes collection and analysis of data cumbersome and difficult to automate. Storage of the papers, handling during testing, and disposal involve complex mechanical functions on the part of the sampling equipment.

The theory behind the new method is simple. Urine is heated in a clear container and irradiated by a red-emitting semiconductor laser. If there is protein in the urine, the heating causes it to coagulate and the solution becomes cloudy. By measuring the degree of scattering with an optical sensor, it is possible to accurately determine the quantity of protein in the urine (see figure).

Modulation of the laser light improves detection sensitivity, and the stability of this method is being verified by measuring not only the scattered light intensity after the protein in the urine coagulates, but also the change in scattering intensity as the urine is slowly heated. The current system requires 1.5 ml of urine and 60 s to detect any protein (concentration above 10 mg/

deciliter). The test results are digitized so the storage and analysis of data are simple and medical trends can be tracked over an extended period.

More in Research