Changing Mid-Infrared Wavelengths in Milliseconds

Breck Hitz

Because most molecules have unique vibrational spectra in the mid-infrared region (mid-IR is often defined as the region from 3 to 30 μm), spectroscopy in that region has many applications, from remote chemical sensing to biomedicine to trace-gas detection. Reliable wavelength-tunable mid-IR lasers are required for spectroscopy in this region. Moreover, these lasers are of intense interest in military applications because many modern “smart” weapons depend on mid-IR sensors, which can be disabled by powerful lasers at the appropriate mid-IR wavelength.

Recently, scientists at Tohoku University in Sendai, Japan, demonstrated a laser system that generates tens of kilowatts of peak power in the mid-IR and that tunes from one wavelength to another in milliseconds. Their system comprises a YAG laser followed by a pair of optical parametric oscillators (OPOs).

Previous investigators — including those at Tohoku — have accessed the mid-IR with YAG-pumped OPOs. What is unique about the current approach is that the scientists tuned the mid-IR wavelength of the second OPO not by tuning that oscillator, but by tuning the wavelength of the first OPO that pumped the second one. This approach avoids the wavelength-dependent walk-off and other deleterious effects that accompany angle-tuning the mid-IR OPO.

Their experimental setup began with a commercial Nd:YAG laser from Spectra-Physics (Figure 1). The 1.06-μm output of that laser pumped the first OPO, which contained two KTP crystals, each mounted on its own galvanoscanner from Harmonic Drive Systems.

Figure 1. Scientists tuned mid-IR output of a second (ZGP) OPO by tuning the wavelength of the first (KTP) OPO. Reprinted with permission of Optics Letters.

The advantage of a double-crystal OPO is that the second crystal exactly compensates the walk-off and beam displacement of the first, so that the output is a round beam that remains stationary as the crystals are angle-tuned. These two crystals were oriented for type-II phase matching, so that the signal and the idler emerged from the resonator in orthogonal polarizations.

The scientists discarded the ordinary polarization with a beamsplitter, leaving the extraordinary beam that was tunable between 1.87 and 2.4 μm and that contained as much as 52 mJ at 2 μm. This beam then pumped the second OPO, which contained a ZnGeP₂ crystal supplied by MolTech GmbH.

Figure 2. By tuning the pump wavelength between ~2 and 2.3 μm (X-axis), the scientists tuned the mid-IR OPO between ~5.0 and 9.8 μm (Y-axis). The different curves are for different fixed phase-matching angles in the second OPO. The data points are experimental, and the solid curves are calculations based on the relevant Sellmeier equations.

The scientists could tune the output of the second OPO continuously and randomly between 5.0 and 9.8 μm (Figure 2). Because the KTP crystals in the first OPO were mounted on fast galvanoscanners, the mid-IR output of the second OPO could be tuned from one wavelength to another in about a millisecond. At the low end of the mid-IR spectrum (i.e., 5 μm), the tuning was limited by the reflectivity of the second OPO’s mirrors. At the long-wavelength end of the spectrum, the situation was more complex.

When the beam from the first OPO was normal to the face of the ZnGeP₂ crystal in the second OPO, the mid-IR output could be tuned to wavelengths as long as 9.8 μm and was ultimately limited by the OPO’s high oscillation threshold. This is indicated by the curve labeled 51° in Figure 2, where 51° is the internal angle relative to the optic axis that corresponds to normal incidence at the crystal face. When the ZnGeP₂ was angle-tuned for nonorthogonal incidence angles, the Fresnel loss from its surfaces increased rapidly and limited the tuning to wavelengths shorter than 9 μm.

Figure 3. Peak powers as high as 150 kW were generated in the mid-IR spectral region. The solid curve represents results obtained by tuning the first OPO while the second OPO was fixed. The broken curve represents the reverse situation.

Peak power and energy outputs from the mid-IR OPO varied between about 10 and 150 kW, and 200 and 1000 μJ, respectively, as the scientists tuned it between 5 and 10 μm (Figure 3). The linewidth of the mid-IR output was approximately 6 cm^–1, and the M-square of the beam was 3.7. The measured conversion efficiency from the 2-μm pump to the 7.7-μm mid-IR was nearly 7.5 percent, corresponding to a quantum (photon) efficiency of 28 percent.

Optics Letters, Feb. 1, 2007, pp. 274-276.