QUANTUM-CASCADE LASERS: Single laser emits beams at two angles

A new member of the quantum-cascade laser (QCL) family has some impressive characteristics—like lasing at two or more wavelengths, or in multiple directions. The result is another in a growing line of QCL output-modification techniques developed by an international team of researchers from the Swiss Federal Institute of Technology (Zurich, Switzerland), Hamamatsu (Hamamatsu City, Japan), and the group headed by QCL co-inventor Federico Capasso at Harvard University (Cambridge, Massachusetts).

The trick—as with the team's prior creations—lies in plasmonics, the manipulation of the quantized, high-frequency free-electron oscillations known as plasmons.

Members of the same team previously introduced a subwavelength aperture into the QCL structure along with a metallic grating, the combination of which couples some of the laser energy into surface plasmons. They went on to show that, if the "plasmon grating" has a period about the same as the surface plasmon wavelength, the output divergence of the laser could be reduced by orders of magnitude as the waves from both structures interfere in the far field.¹

"Lasers emitting multiple beams are both interesting and useful so we have long envisioned creating them," says Nanfang Yu, the Harvard researcher who is lead author on the new work, which tackles the case in which the grating is not the same as the surface-plasmon wavelength.

Two gratings, two outputs

Here, a convenient consequence of interference again occurs: two different gratings of different periods can couple the plasmons out—and thus the laser output—in two different directions. The team built a QCL running at 8 µm with grating spacings of 7.8 and 6 µm, resulting in outputs separated in angle by nearly 20°.

The approach results in further control in terms of the intensity of the individual beams—the longer the grating is in terms of total number of grooves, the more intense the emitted beam.

And by building a stack of two different active regions with an appropriately designed single plasmon grating tuned for both, the team can create QCLs that operate at two wavelengths with spatially separate, collimated output. For this, the team created a stack operating at both 10.5 and 9.3 µm with a grating of period 8.5 µm, resulting in two beams spatially demultiplexed by an angle of about 9°.²

The authors note that the single-grating approach will always be a tradeoff in terms of collimation, as the different wavelengths have different optimum grating characteristics. Nevertheless, the prototype devices are a powerful manifestation of the idea. "These are proof-of-principle demonstrations; however, there is no obstacle preventing us from making lasers that emit many beams with the same or different colors," says Yu.

IR plasmons go farther

The devices also demonstrate a fundamental point about mid-IR surface plasmons. While surface plasmons in the visible are known to propagate for a distance of only a few wavelengths before being absorbed or scattered, their cousins in the IR were calculated to survive much longer—hundreds of wavelengths.

But until now, no one had ever proved it. "The successful operation of our devices largely depends on the spread of surface plasmons over a broad area on the device facet, which is a perfect manifestation of low-loss propagation of mid-IR surface plasmons," says Yu.

Yu says that the real power of the approach will lie in applications calling for high-throughput monitoring at a number of different wavelengths, such as environmental and atmospheric sensing. But further applications for small multidirectional and multiwavelength sources will abound.

"A device emitting two beams of the same color can be very useful for interferometry, which requires two coherent beams: a probe beam and a reference beam," says Yu. "A device emitting two beams of different colors is ideal for differential absorption LIDAR (light detection and ranging), where the difference in absorption between two wavelengths can provide a quantitative mapping of concentration."

However, many applications will call for collimation in two dimensions, rather than the single "knife-edge" collimation of the current work. The researchers are now looking into more sophisticated approaches to give them full control over the spreading of the surface plasmons.

REFERENCES

1. Yu et al., Nature Photonics 2, p. 564 (2008).
2. Yu et al., App. Phys. Lett. 95, p. 161108 (2009).