High-power laser-diode array is monolithically fabricated
Engineers at Quintessence Photonics (Sylmar, CA) have developed a 2-D monolithic high-power laser-diode array composed of elements that are surface-emitting lasers, but not VCSELs (vertical-cavity surface-emitting lasers; VCSELs are low-power devices). Instead, the laser cavities are in-plane, just as for ordinary edge-emitting laser diodes. The light is coupled vertically out by 45° monolithically integrated total-internal-reflection (TIR) mirrors formed by a wet-etch process (seen here in a scanning-electron-micrograph perspective view); the design allows wafer-scale fabrication. A prototype array of 75 laser diodes (three rows of 25 each) exceeded a 100-W optical output at 985 nm.
The second cavity mirror for each laser is an unpumped distributed Bragg reflector (DBR). The TIR beam leaves the laser through an antireflection-coated surface. The laser has an angular emission of 5° × 35°. Increasing the DBR-grating strengths of the lasers will raise their efficiencies to that of lasers in conventional edge-emitting arrays. The company plans to ship prototypes for commercial evaluation by the fourth quarter of 2004. "We are currently showing devices for demonstration to customers," says George Lintz, chief operating officer. "We will achieve more than a kilowatt with our modules over the next 12 months." Contact George Lintz at [email protected].
Birefringent lens reduces aberration effects
Researchers at Shizuoka University (Hamamatsu City, Japan) and the University of Calcutta (Calcutta, India) have calculated that, for certain cases, the incorporation of a birefringent lens into a multielement lens design can reduce the effect of off-axis aberrations to a practical degree. The effect also occurs for single-element birefringent lenses (and, in fact, was the researchers' model system), but the requirement for two polarizers and a very low-level birefringence make single-element versions less practical.
A +45° polarizer, a birefringent lens element, and another +45° polarizer form the basic version that was compared to a conventional lens. Primary astigmatism with a coefficient of 1λ viewed at the image plane between sagittal and tangential foci resulted in the well-known "circle of least confusion" for the conventional lens, but resulted in a spot with a smaller spread for the birefringent lens. Primary coma with a coefficient of 2λ showed similar results. The researchers had previously calculated that the effects of spherical aberration were also reduced. The concept would probably find best use in optical systems that already rely on polarized light. Contact Sucharita Sanyal at [email protected].
Fabry-Perot wavelength-stabilization method has ideal error-signal shape
When quasi-monochromatic light passes through a Fabry-Perot etalon at a fixed angle, the exiting intensity can vary sharply versus wavelength; this property can be used for laser-wavelength stabilization. For this setup, the signal—simply the output of a photodetector—always has a positive value. An optimum error signal, however, would be one that had a value that was approximately linear within a certain range and that passed through zero at the center of that range. Now, researchers at National Chiao Tung University and National Tsing Hua University (both of Hsinchu, Taiwan) have developed a straightforward way to obtain such a signal from a Fabry-Perot etalon.
The laser beam under test is split into two separated beams (A and B), each entering at a slightly different angle to produce two intensity peaks occurring at slightly different wavelengths. Each beam is measured by its own photodetector. When one peak is subtracted from the other (A minus B), an optimum error-signal shape results. In an experiment at 657.46 nm producing two peaks 1.1 GHz wide, drift for a single-peak test was 0.28 GHz in the first three hours, while the double-peak method produced no observable drift in the same time frame. Contact Mao-Sheng Huang at [email protected].
Bent substrate with linear Fresnel-zone plate focuses x-rays
Optics that focus hard x-rays are based on familiar approaches such as refractive lenses, mirrors, and Fresnel-zone plates—but are, in one way or another, pushed to the extreme (see Laser Focus World, September 1999, p. 24). For example, two crossed cylindrical mirrors at grazing incidence can, together, focus x-rays in two axes. Now, researchers at the University of California–Santa Barbara, Copenhagen University (Copenhagen, Denmark), and the European Radiation Synchrotron Facility (Grenoble, France) have developed a simple alternative—a linear Bragg-Fresnel lens on an initially flat multilayer substrate, in which the lens focuses in one axis while the substrate, when bent, focuses in the other.
A multilayer mirror made of 200 layers of molybdenum and silicon is deposited on a 10 × 60-mm glass substrate. Linear Fresnel zones are patterned in a gold layer on the substrate. Bending the substrate by applying moments at the ends creates a cylindrical shape. A 250 × 200-µm 12.4-keV x-ray beam was focused to a 1.6 × 12-µm spot, with the larger-than-expected 12-µm spot axis the result of a nonoptimum mirror bender producing a noncylindrical shape. Contact Youli Li at [email protected].
Organic solid-state laser emits deep-blue light
Although amplified spontaneous emission had previously been observed in the 392- to 444-nm region for organic solid-state molecules, lasing had not. Now, scientists at the Technische Universität Braunschweig and the Physikalisch-Technische Bundesanstalt Braunschweig (both of Braunschweig, Germany) and Covion Organic Semiconductors (Frankfurt am Main, Germany) have created an optically pumped deep-blue laser from an organic spirobifluorene derivative.
A distributed-feedback Bragg grating was patterned via electron-beam lithography and dry-etched into a 100-nm-thick film of silicon dioxide on a silicon substrate. A thin film of the organic material was then deposited on the grating. A pulsed nitrogen laser emitting at 337 nm was used as the pump source. Depending on the grating period (which varied across the substrate and which could be chosen by directing the pump beam), the laser emitted anywhere from 401.5 to 434.2 nm. The laser threshold energy density was 83 µJ/cm2. The spectral width (full-width at half-maximum) of the laser lines ranged from 0.11 to 0.4 nm; the beam itself was polarized with a polarization ratio of 31:1 and was highly anisotropic. Contact Thomas Riedl at [email protected].
White OLED for LCD backlighting reaches 18.4-lm/W efficiency
Researchers at Universal Display (Ewing, NJ) and Toyota (Aichi, Japan) have demonstrated a red-green-blue (RGB) organic-light-emitting-diode (OLED) device that emits white light at efficiencies of 18.4 lm/W and 36 cd/A and a 16% external quantum efficiency at a luminance of 1000 cd/m2. Rather than the fluorescent emission of conventional small-molecule OLEDs, this emitter relies on more-efficient phosphorescent emission for all three RGB outputs. The color-rendering index of 0.79 is in the acceptable range for white-light illumination, but since the intended purpose of the device is liquid-crystal-display (LCD) backlighting, the close match of the output with the RGB LCD-filter bandpasses ensures high display-color quality.
Because the red subpixel in flat-panel displays usually uses the most power, the researchers decided to use a slightly less-saturated red-emitter that was more efficient; the substitution is estimated to reduce display power consumption by 25%. In contrast, they predict that as more-saturated blue phosphorescent dopants are developed, display efficiency will go up. Lifetimes of the red and green phosphorescent OLEDs are projected to exceed 15,000 and 25,000 hours respectively; the lifetime of the blue emitter is still being improved to meet requirements for a commercial product. Contact Yeh-Jiun Tung at [email protected].
Nonlinear QWIP makes autocorrelation measurements of ultrafast pulses
A quantum-well IR photodetector (QWIP) optimized for nonlinear behavior has been developed by researchers at the Fraunhofer Institut für Angewandte Festkörperphysik (Freiburg, Germany) and the National Research Council (Ottawa, Ont., Canada). The detector's three equidistant energy levels produce a giant resonant linearity that leads to a quadratic power dependence of the photocurrent down to light intensities of 0.1 W/cm2. The two-photon effect is useful for autocorrelation measurements of ultrashort picosecond IR pulses.
The gallium arsenide–based QWIP has a 20-period active region and is sensitive in two bands centered on 8 and 10 µm. Test light from a 10.3-µm-emitting carbon dioxide laser produced a nonlinear absorption coefficient of 1.3 × 107 cm/GW—six orders of magnitude larger than for bulk material. A Ti:sapphire laser and an optical parametric oscillator produced mid-IR pulses of a few picojoules in energy and 165 fs in length for the autocorrelation experiment. An interferometric setup with an oscillating mirror allowed averaging of phase and the acquisition of an accurate intensity autocorrelation. At high bias voltages, the signal shows a linear dependence on intensity resulting from a tunneling process. Contact Thomas Maier at [email protected].
Quantum-cascade laser is designed for third-harmonic emission
A quantum-cascade laser developed by scientists at Lucent Technologies' Bell Laboratories (Murray Hill, NJ) and Texas A&M University (College Station, TX) has an active region designed to simultaneously emit its fundamental wavelength and third harmonic. The indium gallium arsenide/aluminum indium arsenide device has a three-well diagonal-transition active region containing a third-order nonlinear oscillator.
The fundamental emission is at approximately 11.1 µm and the third-order wavelength is 3.7 µm; the device also emits some second-order 5.4-µm light. The laser was operated in pulsed mode with 50-nm pulses at a duty cycle of less than 1% at a temperature of 6 K. Calculations show that a fundamental (pump) peak power of 100 mW produces 30 to 50 nW of third-order light, a level in agreement with the low tens of nanowatts measured by a liquid-nitrogen-cooled indium antimonide photovoltaic detector (the fundamental was measured by a liquid-nitrogen-cooled mercury cadmium telluride detector). The researchers aim to improve the efficiency of the wavelength conversion, in particular by incorporating phase-matching into the waveguide design. Contact Claire Gmachl (who is now at Princeton University) at [email protected].
Water penetration rate is measured in silica and sapphire fibers
Silica optical fibers are used for sensing strain, temperature, and other parameters, even in harsh environments. Sapphire optical fibers have been demonstrated as a sturdier alternative to silica, but details on both fibers' environmental performances are needed to make it easier to choose between the fiber types for specific applications. Researchers at the Virginia Polytechnic Institute and State University (Blacksburg, VA) have tested the penetration rate of water into both types of fiber under elevated temperature and pressure.
Commercially available single-mode silica and single-crystal multimode unclad sapphire fibers were compared under broadband illumination and measured by an optical spectrum analyzer over a 400- to 1700-nm spectral range. Two 70-cm silica fiber segments and two 30-cm sapphire fiber segments were tested in sealed chambers held at 280°C and an equilibrium water-vapor pressure for water of 932 psi absolute. For the silica fiber, absorption of light began to rise after 2000 hours, which corresponded to the time that water penetrated through the cladding to the fiber's core. The sapphire fiber, though unclad, showed no increased light absorption over the 3739-hour duration of the test. Contact Anbo Wang at [email protected].