Newsbreaks

Tailored harmonics produce attosecond pulses; Optical switch reaches record throughput; Complex emitter layer quadruples OLED efficiency; and more.

Tailored harmonics produce attosecond pulses

By tailoring short-wavelength harmonics generated by an ultrafast pulse passing through argon (Ar), researchers at the Foundation for Research and Technology-Hellas (Crete, Greece) have created a train of subfemtosecond pulses led by a single, strong pulse only 60 attoseconds (as) in duration. The optical spectrum of such a train of pulses is a series of harmonics spanning the 40-50-eV region and a coherent continuum background spanning 25-60 eV. The latter is associated with the single 60-as pulse. The filter used to tailor the harmonics was 150-nm-thick aluminum silicon (1%). A Ti:sapphire laser supplied an 800-nm, 60-fs pulse that was then passed through an autocorrelator, creating two time-delayed pulses; these were focused into the exit pinhole of an Ar laminar-flow cell. A microchannel-plate detector acquired the signal exiting the filter. Intensity was measured as a function of time delay between the two pulses, producing an autocorrelation trace. The 60-as spike occurred at the region of strong overlap of the two 800-nm pulses. Such a short spike may someday allow the probing of electron motions in matter. Contact Nektarios Papadogiannis atpapadogi@iesl.forth.gr.

Optical switch reaches record throughput

An optical crossconnect switch that can route any of 112 input channels to any of 112 outputs has been developed by Lucent Technologies. Based on micro-electromechanical systems (MEMS) mirrors, it is "the largest fully functional, fully configured optical crossconnect" yet demonstrated, said David Neilson of Bell Labs (Holmdel, NJ) in a postdeadline paper presented at the 2000 Optical Fiber Communication conference (Baltimore, MD) last March. A microlens array focuses light from input fibers onto a two-dimensional array of MEMS mirrors that can tilt ±5° on each of two axes. Each micromirror bounces the input beam to a fixed reflector that directs it back to another micromirror in the array. That mirror, in turn, reflects the light to the output fiber. Neilson said switching times were less than 10 ms, and insertion loss was 7.5 ±2.5 dB. A major advantage is that the number of active switching elements increases with the port count N, not the number of possible connections N2. "It should be scalable to port counts greater than 1000," said Neilson. By testing the switching array with a 320-Gbit/s data stream at one wavelength, they achieved a switching rate of 35.8 Tbit/s. Contact David Neilson at neilson@lucent.com.

Complex emitter layer quadruples OLED efficiency

Researchers at Princeton University (Princeton, NJ) and the University of Southern California (Los Angeles, CA) have used a phosphorescent sensitizer to excite a fluorescent dye and nearly quadruple the efficiency of a fluorescent red organic light-emitting device (OLED). The new design replaces the simple emissive layer in a conventional OLED with a complex layer composed of alternating phosphorescent and fluorescent materials. In the conventional design, only excitons with singlet spin configuration go on to emit light while wasting the energy of excitons with triplet spin configuration and yielding an efficiency of about 25%. The complex layer quadruples the efficiency of the device by enabling light emissions from excitons with triplet spin configuration also. Stephen Forrest, head of the research team, which is also collaborating with Universal Display Corp. (Ewing, NJ), said the new technique would be available to electronics manufacturers within six months for applications such as car stereo displays. He added that it might eventually lead to use of OLEDs in cell phones and laptop computers with substantially increased energy efficiency. Contact Stephen Forrest atforrest@ee.princeton.edu.

Crystal polarizes light with 73% efficiency

By blocking one polarization component, an ordinary linear polarizer converts unpolarized light to polarized light with a maximum theoretical efficiency of 50%; actual linear polarizers do less well than this. Researchers at the University of Rochester (Rochester, NY), however, have built a polarizer that works at greater than 50% efficiency—and, at first glance, seems to violate the second law of thermodynamics.

The device is based on a photorefractive crystal within which two interfering laser beams create a grating that causes the first order of the first beam to constructively interfere with the zeroth order of the second and the first order of the second to destructively interfere with the zeroth order of the first. The result is that both beams are combined. The two beams can result from an unpolarized beam split into two polarized components with one component rotated 90°. Theoretical efficiency of the device is 100%, and actual efficiency has reached 73%. Future photorefractive materials may do the same with incoherent light. The researchers presume that the decrease in entropy caused by the device is offset by an increase of entropy either in the crystal or in the beam intensity. Contact Robert Boyd atboyd@optics.rochester.edu.

Small-area GaN detector produces large gain

Researchers at the United States Military Academy (West Point, NY) and the University of Texas at Austin (Austin, TX) have fabricated a gallium nitride (GaN) photodetector that has avalanche gain. Because avalanche photodiodes have low noise, they provide an attractive alternative to higher-noise metal-semiconductor-metal GaN detectors.

The device is constructed on sapphire and contains five epitaxial layers. Avalanche gain is achieved when the gain region is subjected to a high electrical field. When driven by a high electrical field, GaN can experience the creation of microplasmas at defect locations; because GaN has a large number of defects, the resulting microplasmas can induce breakdown at a voltage below that needed to reach high gain. To solve this problem, the researchers reduced the detector diameter to 25 —m, producing gain areas with relatively defect-free regions. Such detectors can be driven by an electric field of up to 3 MV/cm and reach a gain of 25 for a 363-nm wavelength without forming microplasmas. Contact John Carrano atjcc@mail.utexas.edu.

Electro-optic birefringence controls GaAs VCSEL polarization

To incorporate vertical-cavity surface-emitting lasers (VCSELs) into polarization-sensitive systems such as magneto-optic disks or coherent detection systems, VCSELs with a single preferential polarization are desirable. Most VCSELs without intentional polarization selectivity show orthogonal polarization states at and above the threshold and unstable polarization switching with increased current. Although researchers have produced a single dominant polarization mode using features such as anisotropic transverse cavities, and even non [001] gallium arsenide (GaAs) structures, changing the dominant polarization mode was impossible once the device was fabricated. Researchers at the Korea Advanced Institute of Science and Technology and the Electronics and Telecommunications Research Institute (both in Taejon, Korea), working with a colleague at the University of California, Berkeley (Berkeley, CA), have developed a polarization-control method for VCSELs based on a [001] GaAs substrate using electro-optic birefringence. Birefringence was induced at the top distributed Bragg reflector by applying an electric field along the [001] direction. The cavity resonance of the polarized light along the [110] and [110] directions shifted to shorter and longer wavelengths, depending on the direction of the applied field. By varying the direction and strength of the electric field, the scientists actively controlled VCSEL polarization, with the dominant mode in the [110] direction for a negative field and in the [110] direction for a positive field. Contact Byung Tae Ahn atbtahn@cais.kaist.ac.kr.

Light slows to a crawl

Researchers at the Rowland Institute of Science (Cambridge, MA) and Harvard University (Cambridge, MA) have been radically slowing light by passing it through a specially prepared Bose-Einstein condensate (BEC; a cooled cloud of atoms all in the same quantum state). Now they have beaten their own record, reducing the velocity of light further to a speed of 0.44 m/s, or 1 mph. They use both a coupling and a probe laser beam to achieve electromagnetically induced transparency in a sodium BEC, which produces a slowing of the probe beam. The new results come from the use of two separate wavelengths for the two beams, rather than one, allowing the use of the sodium D2 line for coupling and the sodium D1 line for the probe and thus reducing losses.

The next step will be to slow the light in the BEC down to 1 cm/s, says Lene Hau, principal investigator. This is the point where the speeds of sound and light are equal, allowing atoms to ride along on the light pulses. Such a phenomenon could permit the creation of low-power nonlinear optical devices. By combining such high nonlinearities with nanostructures, tiny optical switches might be made that change states when influenced by as few as two photons, says Hau. Contact Lene Hau athau@physics.harvard.edu.

Diode-pumped Q-switched laser is controlled using fluorescence feedback

In conventional active Q-switching, triggering the output pulse is done by electronic control with a pulse generator at a preset pulse-repetition frequency. The cavity's Q is switched by dead reckoning from the pulse generator without examination of the true state of population inversion of the pumped medium. Stable laser performance requires a stable pump laser and stabilized cavity conditions. According to researchers at the Photonics Research Group at Nanyang Technological University in Singapore, an alternative electronic control method could remove the uncertainty in the actual inversion level at the onset of Q-switching. Their technique incorporates an active optical-feedback loop based on the monitoring of fluorescence intensity outside the laser cavity of an acousto-optic Q-switched Nd:YVO4 (vanadate) laser. With this method, they can select the level of fluorescence intensity that triggers the Q-switch. When the initial inversion level indicated by the detected fluorescence reaches that predetermined value, Q-switching is initiated. Researchers can then vary the reproducibility of the output-pulse peak power and pulsewidth and reduce the shot-to-shot variations of both parameters. Contact Wenjie Xie atP145520398@ntu.ed.sg.

Dual-focus Fresnel-lens modulator produced from GaAs/AlGaAs

Researchers at the University of Michigan (Ann Arbor, MI) and the University of Texas (Austin, TX) have fabricated and tested a dual-focus Fresnel-lens modulator using a gallium arsenide/aluminum gallium arsenide (GaAs/AlGaAs) multiple-quantum-well heterostructure. The heterostructure was grown using molecular-beam epitaxy in the form of a PIN (MQW) diode with 60 periods of undoped GaAs/AlGaAs sandwiched between 20 periods of GaAs/AlGaAs superlattice cladding on either side. The alternate opaque and transparent rings of a Fresnel lens were simulated using a quantum-confined Stark effect over alternate rings etched in the heterostructure. A light-source wavelength coincident with the zero-bias MQW absorption peak and a 2pi light-path phase difference between rings produced a near-opaque condition at zero bias and transparency at negative bias.

The device was tested using an 859-nm diode-laser source. No focusing of the light took place upon passing through the lens with a zero bias voltage applied, and a focused spot was observed beyond the lens when a bias of -12 V was applied. Light intensity at a central focal spot beyond the lens was observed to increase by a factor of seven with the bias voltage applied. Contact Pallab Bhattacharya atpkb@eecs.umich.edu.

More in Research