High-conversion-efficiency, diode-pumped continuous-wave Raman laser.

We demonstrate a diode-pumped cw Raman laser in H(2) with photon-conversion efficiency of (66+/-8)%. Pumped by an injection-locked diode laser at 792 nm, the Stokes laser produces a peak output power of ~16mW at 1180 nm. Accompanying the high Stokes power are deviations from the existing theory, which are believed to be caused by the thermal-lensing effect of the Raman gas.

Widely tunable continuous-wave Raman laser in diatomic hydrogen pumped by an external-cavity diode laser.

What is to the authors' knowledge the first experimental demonstration of a nonresonant cw Raman laser pumped by a tunable external-cavity diode laser (ECDL) is presented. The ECDL is phase-frequency locked to a high-finesse Raman laser cavity containing diatomic hydrogen (H(2)) by the Pound-Drever-Hall locking technique. The Stokes lasing threshold occurs at a pump power of 400 +/- 30 muW, and a maximum photon conversion efficiency of 12.0 +/- 1.3% is achieved at 1.6 mW of pump power. A 40-nm tuning range of the cw Stokes emission, 1174-1214 nm, is obtained by tuning of the wavelength of the ECDL pump source.

Laser diode facet modal reflectivity measurements.

A simple and accurate method for measuring the front facet modal reflectivity of a Fabry-Perot laser diode is presented. In this method, optical feedback from an external mirror of known reflectivity, R(ext), is used to alter the laser diode threshold current. The effect of the external mirror and front facet reflectivities on the threshold current then allows for a measurement of the front facet modal reflectivity of the laser diode and is theoretically and experimentally studied. This method was used to measure a facet reflectivity of R(2) = 0.0151(+0.0018/-0.0032) [R(2) = 0.00592(+0.00085/-0.00123)] for a commercially antireflection-coated facet of a laser diode with a center wavelength of 795 nm (935 nm). The results of the reflectivity measurements based on the threshold current as a function of the external mirror reflectivity are compared with the results of the reflectivity measurements based on modulation depth of the optical spectrum [IEEE J. Quantum Electron. QE-19, 493 (1983)].

Simple method for measuring frequency chirps with a Fabry-Perot interferometer.

A simple method for measuring the frequency-chirped response of lasers is presented. This method relates the deviation from the Lorentzian line shape of the transmission of a Fabry-Perot interferometer to the frequency chirp of the laser and allows a direct measurement of the frequency chirp. Two chirps produced by an external-cavity laser diode with an intracavity electro-optic crystal were measured. The first measurement was of a linear chirp of 800 MHz occurring in a time of 12.3 micros, and the second measurement was of eight repeated 800-MHz linear chirps each occurring in 337 micros. Agreement between the measured and the expected frequency-chirped response of the laser is shown.

Programmable frequency reference for subkilohertz laser stabilization by use of persistent spectral hole burning.

We report what is believed to be the first demonstration of laser frequency stabilization directly to persistent spectral holes in a solid-state material. The frequency reference material was deuterated CaF(2): Tm(3+) prepared with 25-MHz-wide persistent spectral holes on the H(6)(3)?H(4)(3) transition at 798 nm. The beat frequency between two lasers that were independently locked to persistent spectral holes in separate crystal samples showed typical root Allan variances of 780+/-120Hz for 20-50-ms integration times.

Diode-pumped nonresonant continuous-wave Raman laser in H2 with resonant optical feedback stabilization.

We demonstrate a nonresonant cw Raman laser pumped by an optically locked diode laser at 790 nm that produces cw Stokes (1178-nm) and coherent anti-Stokes (595-nm) emission. Considering the modest pump powers, relative low cost, and predicted spectral purity, we expect that frequency downconversion of tunable diode lasers through stimulated Raman scattering will provide an attractive source for remote sensing, spectroscopic, and atomic physics applications. The Stokes laser threshold is 240+/-19muW pump power, and emission is observed over a roughly 10-nm range by adjustment of the optical locking feedback phase. Photon-conversion efficiency rises throughout the pump-power region, with a peak value of 15+/-2% .

Continuous-wave Raman laser in H(2).

Recent developments in high-finesse cavities now make broadly tunable, continuous-wave Raman lasers possible. The design and preliminary characterization of what is to the authors' knowledge the first continuous-wave Raman laser in H(2) are presented. The threshold is currently at 2 mW of pump, making diode laser pumping possible. The maximum photon conversion efficiency observed was 35% at 7.6 mW of pump power.

Correcting an astigmatic, non-Gaussian beam.

An off-axis spherical mirror is used to correct an astigmatic, non-Gaussian beam from a pulsed, frequency-doubled Nd:YAG laser. The beam is then spatially filtered by a series of two pinholes to make the beam near Gaussian.

Frequency stability of high-finesse interferometers.

The effects of temperature and pressure on the stability of a high-finesse interferometer are considered, and the design of a high-finesse interferometer that minimizes these effects is presented. The high-finesse interferometer has a free spectral range of 23,600 MHz, a finesse of greater than 30,000, and a measured stability of better than 7 MHz/h (0.3 mfringes/h).

Low-frequency intensity noise in semiconductor lasers.

The low-frequency intensity noise at 25 MHz of a Fabry-Perot semiconductor laser is measured as a function of injection current. All the measurements are taken at room temperature and the laser is operated with a commercial current source (the conditions under which laser diodes are often used). At the highest injection current of twice threshold, the intensity noise is 5.5 dB above the shot-noise limit. When the longitudinal side mode suppression of the laser is 20 dB or larger, the intensity noise is modeled adequately by an expression derived from the single-mode, small-signal, linearized, semiclassical rate equations. All the parameters used in the theory are derived or referenced.

High-finesse interferometers.

A detailed plan for the construction and use of a nonconfocal cavity used as a high-finesse interferometer is presented. The interferometer has a free spectral range of l5 GHz, with a finesse of over 30,000.

Efficient seeding of a Raman amplifier with a visible laser diode.

Recent developments in tunable laser diodes now make it possible to do injection seeding of a quantum-limited Raman amplifier. We present data from an experiment in which a tunable laser diode in the visible was injected into a Raman amplifier with a coupling efficiency of 65% +/- 45%. This efficiency approaches the theoretical limit of 100% and is much larger than the 0.1-0.2% coupling efficiency reported in the literature for a comparable experiment when an infrared laser diode is used.

Cyclic shearing interferometer for collimating short coherence-length laser beams.

Until now there has not been an accurate method for measuring the radius of curvature, R, of a short coherence-length light source, such as a short-pulse or broadband laser. We show that the easily aligned cyclic shearing interferometer (CSI) solves this problem. The CSI produces a stable fringe pattern from which R can be determined and can be used on beams with short coherence times down to 300 fs because the two beams in the interferometer follow nearly the same path. Comparison with data from a broadband XeCl laser (30-ps coherence time) confirms that the CSI performs as theory predicts.

Efficient stimulated Raman scattering due to absence of second Stokes growth.

Stimulated Raman scattering of a XeCl laser at 308 nm in a high-pressure H(2) cell shows high conversion into first Stokes (S1) because of an unexpected holdoff of the second Stokes (S2) component. Specifically, a photon efficiency of 88% is obtained into S1. Comparison with a plane-wave model indicates that a theory including a spatially nonuniform gain and higher-order mode generation may be necessary to understand the holdoff of the S2 growth.