SPECTROSCOPY - QC lasers achieve performance milestones
Since their invention at Bell Laboratories, the R&D arm of Lucent Technologies (Murray Hill, NJ), five years ago, quantum-cascade (QC) lasers have surpassed a number of important performance and reliability milestones and are yielding promising results in applications-oriented laboratory tests. Researchers from Bell Labs and Pacific Northwest National Laboratory (PNNL; Richland, WA) reported the latest device characteristics and performance levels in chemical-sensing spectroscopy application
Since their invention at Bell Laboratories, the R&D arm of Lucent Technologies (Murray Hill, NJ), five years ago, quantum-cascade (QC) lasers have surpassed a number of important performance and reliability milestones and are yielding promising results in applications-oriented laboratory tests. Researchers from Bell Labs and Pacific Northwest National Laboratory (PNNL; Richland, WA) reported the latest device characteristics and performance levels in chemical-sensing spectroscopy applications in July at the 44th annual meeting and exhibition of the International Society for Optical Engineering (SPIE) in Denver, CO.
The range of possible wavelengths for QC Fabry-Perot lasers currently extends from 3.5 to 17 µm, according to Claire Gmachl, who spoke at the meeting on behalf of a Lucent research team led by QC laser inventor Frederico Capasso and including molecular-beam-epitaxy (MBE) process inventor Alfred Cho. Single-mode, tunable distributed-feedback lasers have been fabricated at 5-, 8- and 11-µm wavelengths with 150-nm tuning ranges, greater than 30 dB of side-mode suppression and less than 1 MHz of intrinsic linewidth.
To date, the QC lasers have also exceeded the 1-mW output power range available with lead-salt diodes by an order of magnitude, with an average room-temperature power in pulsed operation on the order of 15 mW, Gmachl said. Power milestones achieved to date for devices in the 5- and 8-µm ranges include 0.54- and 1.35-W peak powers for pulsed operation at room temperature and liquid-nitrogen temperature, respectively. Continuous-wave power marks are 5 mW, 100 mW, and 0.2 W at maximum operating temperature of 175 K, 120 K, and liquid-nitrogen temperature, respectively. Gmachl also reported the development of multiple wavelength lasers emitting at 6.6 and 7.9 µm simultaneously. She said that approximately 15 different spectroscopy groups are testing the devices.
The PNNL researchers have worked so far only in continuous-wave mode at liquid-nitrogen temperature primarily in the 5- and 8-µm ranges, and they have found particular value in the narrow linewidths and wavelength stability. "If you take a gas sample and operate at reduced pressures-25 to 30 torr-you get rid of pressure broadening, and many gas molecules will exhibit extremely sharp narrow-spectral-absorption features," said John Hartman, a member of a testing team at PNNL that includes Richard Williams, Jim Kelly, and Steven Sharpe.
Low noise characteristics facilitate signal averaging with quantum-cascade lasers without linewidth broadening artifacts produced by laser jitter, as illustrated by similarity between average of two scans and 1000 scans.
Gas-absorption lines are on the order of 100 to 200 MHz, which are easily accommodated by the nominal 1-MHz linewidth of the QCs. In fact, at the CLEO meeting in May the PNNL team reported a 12-kHz linewidth achieved by locking the laser wavelength to a nitrous oxide transition, which opens up applications outside of spectroscopy, such as windshear measurements and laser vibrometers, Kelly said.
"We've found that the quantum cascades are very robust," Williams added. "We've been operating some of these devices for a year and a half now, and if you take them back to the same operating current and temperature, they operate at the same wavelength that they did a year and a half ago." The wavelength reproducibility in the QC lasers can reduce system complexity by eliminating the need for a gas reference cell. The reproducibility is maintained after exposure to moisture, electrostatic environments, and multiple temperature cycles that would be detrimental for lead salts.
Traditional lead salt diodes provide a 3- to 30-µm operating range, he noted. But traditional devices lack the power, stability, and environmental robustness needed for sensing specific target vapors in environments ranging from US Department of Energy (DoE) cleanup sites to semiconductor-fabrication processes. "The tunable diode technology that we've used in the past has left us with capability-want lists," Hartman said. "The quantum cascades have already established their capability to meet some of those requirements."
The PNNL team has also found that the QC lasers have very quiet noise characteristics. "With devices that have been used previously, if you took a thousand scans across a narrow linewidth and accumulated those in a storage scope, you would see an absorption linewidth that broadened as the number of scans was increased," Hartman said. "And that broadening would be due to the jitter of the laser emission."
Using QC lasers, the researchers compared data sets from two scans averaged in a digital scope and a thousand scans averaged in a digital scope and found no observable difference (see figure). "The lasers aren't drifting, and their wavelength is very stable," Williams said. "So you can get phenomenal signal-to-noise with a very simple system-just a digital scope rather than a big computer and bunch of hardware."
The PNNL team also presented preliminary results in Denver on heterodyne FM spectroscopy with which they explored the potential for increasing sensitivity by putting an RF modulation on the lasers and recovering the signal in a heterodyne mode. Because the QC lasers work in different voltage regimes and current levels than the lead-salt devices, the PNNL researchers have had to develop circuits tailored to QC requirements.