Newsbreaks

Microdisk-cavity LEDs promise brighter output; Etalon needs no temperature stabilization; PICASO paints a new picture of ultrashort pulse measurements; and more.

Jan 1st, 2000

Microdisk-cavity LEDs promise brighter output

High-brightness light-emitting diodes (LEDs) have been making commercial inroads, particularly with the development of devices based on III-nitrides. At the 1999 Fall Meeting of the Materials Research Society (Nov. 29-Dec. 3; Boston, MA) Jingyu Lin and colleagues from Kansas State University (Manhattan, KS) reported that replacing conventional broad-area LEDs with micro-sized LED arrays can enhance the total light output for a fixed luminous area. The group fabricated microdisk- and microring-cavity LEDs from metal-organic chemical-vapor deposition grown gallium nitride p-n junctions and indium gallium nitride/gallium nitride quantum wells. Lin and colleagues reported that the effects of the small size produced quantum efficiencies that were significantly greater than conventional LEDs. Current density in microdisk-cavity LEDs was six times greater than in standard LEDs, they said. The group said that the findings could have important implications for the design of tiny ultraviolet- or blue-emitting optoelectronic devices, such as micro-LEDs, microdisk and microring laser diodes, micro-vertical-cavity LEDs, and arrays for micro-size displays. Contact Hongxing Jiang at jiang@ksu.edu.

Etalon needs no temperature stabilization

In dense wavelength-division multiplexing applications, closely spaced optical wavelengths must be combined or extracted—operations well suited for a Fabry-Perot etalon. However, conventional etalons consisting of a slab of glass have a peak transmission wavelength that varies with temperature due to a change in thickness, requiring temperature stabilization. Masataka Shirasaki, a scientist at the Massachusetts Institute of Technology (Cambridge, MA), has developed a solid etalon of simple construction that is insensitive to temperature.

A thin glass etalon with high-reflection coatings is sandwiched between two thicker slabs of glass and bonded to them. The two slabs are of a different material from that of the etalon and have a higher coefficient of thermal expansion. When the assembly heats up, the slabs stretch the etalon in two lateral dimensions, decreasing the thickness of the etalon and canceling its thickness increase due to thermal expansion. The etalon itself is 50-100 µm thick—giving it a free spectral range of 8-16 nm—and the slabs are each 1 mm thick. The design completely cancels the first-order wavelength shift over a 90°C temperature range, leaving only a minor °0.01-nm quadratic wavelength shift. Contact Masataka Shirasaki at mshir@mit.edu.

PICASO paints a new picture of ultrashort pulse measurements

Researchers at the University of New Mexico (Albuquerque, NM) and Los Alamos National Laboratory (Los Alamos, NM) have developed an algorithm for measuring the amplitude and phase of ultrashort optical pulses that makes use of a dispersive element in one arm of an autocorrelator. Dubbed Phase and Intensity from Correlation and Spectrum Only (PICASO), the algorithm combines a nonlinear (second-order) intensity cross-correlation from an unbalanced Michelson interferometer with the pulse spectrum to obtain the pulse intensity and phase.

The cross-correlation, obtained between the original pulse and one stretched by a fused-silica plate, makes the algorithm sensitive to both the sign of the pulse chirp and the asymmetry of the pulse envelope. Stretching the pulse in one arm of the correlator also helps remove a temporal ambiguity problem encountered in second-order correlation techniques. In addition, the researchers made use of multiphoton current detectors to overcome difficulties associated with propagating short pulses through bulk nonlinear crystals. A comparison of PICASO with single-shot, second-harmonic-generation frequency-resolved optical gating in measuring a 300-fs pulse from an optical parametric oscillator produced good agreement between the two methods. Contact Jeffrey Nicholson at jeffn@phys.unm.edu.

Microsphere lases green at a low threshold

A powerful, low-cost laser emitting in the green could prove very useful for telecommunications, optical data storage, and medicine. At the Laboratoire d'Optronique (Lannion, France) and l'Ecole Normale Supérieure (Paris, France) researchers have demonstrated a diode-pumped microsphere that emits around 540 nm, with a lasing threshold of 30 µW of absorbed power, about 300 times lower than the lowest previously observed. The researchers used ZBLAN, a heavy-metal fluoride glass, doped with Er3+, to form spheres from 60 to 120 µm in diameter. A series of total internal reflections forms light into a thin ring at the equator of the dielectric sphere, and light can be coupled into or out of the sphere by a prism with a high refractive index. An 801-nm pump beam from a diode laser is delivered to the sphere via a single-mode fiber. The researchers found that the laser emission remained linear even at ten times the lasing threshold. Because these microspheres are easy and inexpensive to produce, they hold promise for a variety of applications. Contact Jean-Michel Raimond at jmr@physique.ens.fr.

Antenna resonates to visible light

An ongoing effort at the Center for Research and Education in Optics and Lasers (CREOL; Orlando, FL) to develop microscopic lithographically fabricated antennas combined with diodes for light detection has resulted in many interesting devices, all operating in the mid-infrared (IR). Now, CREOL researchers have observed resonant sensitivity to visible light in a lithographic antenna, paving the way for heterodyne detection in the visible and for other applications.

The device consists of crossed metal whiskers that form a dipole antenna integrated with a metal-oxide-metal diode. It has a minimum feature size of 0.2 µm, a length of 6.7 µm, and is optimized for the mid-IR. The researchers rotated the polarization of polarized 633-nm light chopped at 400 Hz from a helium-neon laser, while monitoring the output signal of the detector. They observed a polarization dependence, thus proving that a portion of the output signal resulted from antenna resonance and not photoconductive or thermal effects. Scanning the laser spot across the antenna showed that visible frequencies were strongly attenuated, a problem that must be overcome for efficient operation. Optimizing the device geometry and materials for visible light should decrease attenuation, according to the researchers. Contact Cristophe Fumeaux at cfumeaux@mail.creol.ucf.edu.

Subsonic COIL produces 25% chemical efficiency

Since the late 1980s, researchers have studied supersonic operation of chemical oxygen-iodine lasers (COILs) extensively. One drawback for industrial applications has been a vacuum-pump capacity requirement that is much higher than in the subsonic regime because of the addition of buffer gas. Researchers at Tokai University (Japan) recently demonstrated high-pressure subsonic operation of a COIL, producing 448 W of output power for a chlorine input rate of 19.7 mmol/s. Chemical efficiency is 25%, comparable to the highest reported chemical efficiency of COIL using nitrogen as the buffer gas. In addition, the 3.5-J/l specific energy produced by the system is more than five times higher than that of the researchers' supersonic COIL device and reportedly is comparable to the highest specific energy obtained by a helium-diluent supersonic COIL. In the subsonic experiment, the laser cavity utilized singlet delta oxygen generated by a liquid-jet-type singlet oxygen generator without supersonic expansion—operating pressure reached 0.80 kPa. Cooled nitrogen gas was added to the singlet oxygen flow to enhance the output power. Contact Fumio Wani at wani_f@khi.co.jp.

Two-dimensional profiler sees tiny index variations

Researchers at the National Institute of Standards and Technology (Boulder, CO) have developed an apparatus for profiling the refractive index of optical fibers and waveguides in two dimensions. The technique relies upon the refracted-ray method, in which the specimen to be measured is immersed in index-matching fluid and light is passed through a corner of the specimen. A focused spot is scanned across the specimen; light emerging from the focus is refracted out the side of the specimen and collected by an optical system with a central pupil block, resulting in an intensity variation at a second focus that is proportional to refractive index at the first focus.

Collimated light from a 635-nm-emitting temperature-controlled diode laser is mechanically chopped at 500 Hz and directed into a microscope objective with a numerical aperture of 0.65. The objective moves on an x-y-z stage having a 0.1-µm resolution; the optical system itself is limited to a lateral resolution of 0.35 µm. The setup results in a refractive-index resolution of 4.3 x 10-5, an order of magnitude better than the absolute measurement uncertainty of the system. The high resolution allows the researchers to visualize deposition layers in a graded-index fiber. Contact Matt Young at myoung@mines.edu.

Polarization analyzer fabricated on planar lightwave circuit

Researchers at NTT Photonics Laboratories (Ibaraki, Japan) and Japan Women's University (Tokyo, Japan) have fabricated an integrated optical polarization analyzer on a planar lightwave circuit. The first device stage consisted of a polarization splitter in a balanced Mach-Zehnder interferometer configuration with an amorphous silicon film on one of the two arms and a heater to adjust phase bias on the other arm. The second stage consisted of a polarization converter fabricated by inserting a polyimide half waveplate into a waveguide gap. The third stage was a 3 x 3 multimode interference coupler. The polarization splitter divided the input signal into transverse electric (TE) and transverse magnetic (TM) signals. Half of each signal was extracted for monitoring from the converter stage before the remaining half of the TM component was converted to TE polarization and fed into the coupler with the remaining half of the TE component. The researchers observed interference signals at the output ports, which differed in phase because the coupler functioned as a 120° hybrid. Overall device size was 10 x 65 mm2 with a relative index difference of 0.75% and a 6 x 6-µm2 core size. The device had a polarization extinction ratio of greater than 12 dB over the 1530- to 1560-nm wavelength range. Contact Takashi Saida at sai@iba.iecl.ntt.co.jp.

Beamsplitting ball lens functions like five separate components

A beamsplitting ball lens developed at NEC Research Institute (Princeton, NJ) and Radiant Research Inc. (Austin, TX) combines the functions of four imaging lenses and a beamsplitter. According to project researchers, the lens could also be formed with its midplane consisting of a wavelength filter, a polarization filter, or an amplitude-phase mask or grating. The integrated functionality may simplify the parallel-channel optical circuitry for future short-distance optical interconnection needs.

Ball lenses were produced by cutting a conventional ball lens in two, forming a beamsplitter filter at the midplane, and cementing the halves together. For volume production, ball halves could be formed by polymer injection molding. For unity magnification, all tested beamsplitting balls resolved greater than 30 line pairs/mm, with no apparent geometric distortion when their input apertures were set to one half of the ball radius. With a radius of 3 mm or smaller aperture stops, the observed resolution for the on-axis exceeded 50 line pairs per millimeter. For power measurements, most ball lenses delivered a splitting ratio with a variation of °6% from the designed 50/50 splitting ratio. Contact Yao Li at yao@research.nj.nec.com.

More in Research