Optical gain enhancement is demonstrated in a standard mid-infrared quantum cascade laser in pulse operation, using a near infrared illumination on the laser facet. An increase in the laser emission is observed, as well as greater dynamic range, threshold reduction, and a blue shift in the laser cavity modes. The optically induced gain increase allows for optical switching of the laser. All the changes have a nonlinear dependency on the illumination optical power and are attributed to the free carrier concentration increase and the electron transport change in the active region due to the near infrared illumination.
© 2009 OSA
The quantum cascade laser (QCL)  is a promising mid-infrared (MIR) source with attractive applications in spectroscopy  and free-space optical communication . Since its first demonstration, a major focus of research is the improvement of the QCL performance towards high power room temperature operation. Increased doping concentration can increase the laser dynamic range, however, this also leads to an increase in the free carrier absorption and higher threshold and induces a V-shape build-in electrical field which typically decreases gain and limits the dynamic range . QCL performance is also improved by a better thermal dissipation , high reflectivity coating on the facet , and plasmon enhanced waveguide . To improve the QCL voltage efficiency, injectorless structures , shortened injector , and heterogeneous injector  have been employed. Careful design of the QCL with enhanced upper laser level confinement was used to improve the threshold and slope efficiency .
A different way to influence output power was recently demonstrated by C. Zervos et al., who reported their observation of improved performance in the QCL emission power and threshold by illuminating the active region with 60 mW near infrared (NIR) laser pulses through a 10 µm × 50 µm wide window etched in the top contact . However, there was no report on the gain change, only a direct increase of emission power was observed. Furthermore, this approach changes the laser structure and is not compatible with QCL overgrow processing for room temperature operation.
In this paper, we present our experimental observation of an improvement of a standard MIR QCL performance by illuminating its front facet with about 1 mW NIR beam, much smaller power than the previously reported case and without any additional processing to change the laser structure. The NIR excites electrons from the valence band to the conduction band and generates free carriers. The change in carrier concentration and electron transport lead to changes in intersubband laser gain, the lasing emission power, and the slope efficiency as well as cavity modes wavelength. Neither of the latter four was reported by the previous work. The photon-generated free carriers may also have effects on the refractive index of the cavity, the cavity optical confinement, the reflectivity of the laser facet and the cavity temperature. However, as we will show, neither of them makes a major contribution to the observed optical enhancement. Besides the ability of fast NIR to MIR signal conversion , our approach might also allow for more efficient room temperature operation, which is the focus of this paper.
2. Experimental setup
Gain enhancement was observed in several QC lasers, but in this paper, we report only results from a standard 35-stage type-I In0.52Al0.48As/In0.53Ga0.47As four-level multimode Fabry-Perot QCL based on a two-phonon resonant design, with a central wavelength of 7.48 μm, an active region of 2 μm × 15 μm, a laser cavity length of 1.358 mm, and uncoated facets. The QCL is mounted on the cold finger of a closed-cycle Helium cryostat held at 30 K. It is driven by a current pulse source (20 ns pulse duration, repetition rate 5 KHz) monitored by a high speed current loop sensor. Using two f/4 ZnSe lenses, the QCL’s MIR emission is collected and then focused on a fast MCT infrared photodetector. To evaluate the refractive index change and obtain insight into thermal effects, the QCL emission spectrum is also recorded using a FTIR spectrometer. A Ti:sapphire NIR beam with central wavelength 820 nm, pulse width 100 fs and repetition rate 83 MHz is focused down to a 20 μm spot on the QCL front facet with an incident angle about 30 degrees to the QCL MIR beam.
3. Experimental results and discussion
3.1 Current-light and current-voltage characters
Under external NIR illumination, the QCL shows a clear increase in its MIR emission power at any given current above threshold. In Fig. 1(a) , the MIR power (taking into account the loss on the two ZnSe lenses and the ZnSe window in the cryostat) without NIR illumination (dash line) and with 1mW average NIR illumination (solid line) is plotted against the current. It can be clearly seen that the net power enhancement increases with the bias current and peaks with about 35% enhancement near the roll-over point at 566 mA. The illumination also shifts the I-L curve roll-over point towards a higher current value, extending not only the output power but also increasing the dynamic range of the laser. Additionally, the slope efficiency above threshold is increased by about 16% from 0.19 W/A to 0.22 W/A. The threshold current is reduced by 7 mA from 230 mA to 223 mA, which indicates the potential for optically switching a QCL. This is visualized in the inset of Fig. 1(a), where the MIR emissions from the QCL driven below threshold with illumination (solid line) and without (dash line) are compared and the switching effect becomes quite obvious.
In contrast to a simple additional photon current inside the laser active region (which is not measured by the current sensor), the changes in the I-L curves indicate a possible higher optical gain of the QCL, because photon current effect will just shift the I-L curve to lower current values. In fact, assuming complete absorption of the 1 mW NIR beam (taking into account the optical loss on the optics surfaces), the current values should be lowered by the equivalent current of about 0.64mA. Yet the observed 7 mA reduction in the threshold current is ten times higher hence pointing towards a more complex process. The dynamic range increase implies an increase in the carrier concentration in the QCL active region . But, unlike the free carrier generated by doping, which always leads to threshold increase due to the free carrier absorption, here, the photon generated free carriers increase the dynamics range but decrease the threshold at the same time.
Figure 1(a) also gives the corresponding I-V curves plotted for illumination (solid line) and non-illumination (dash line) cases. Under the illumination, the voltage measured across the laser structure is reduced for any given current, which can be explained by the optical induced free carriers. The observation agrees with both theoretical and experimental results with increased carrier concentration due to doping [14,15]. Last, but not least, we want to stress that the voltage decrease can also be found far below the lasing threshold. This implies that the dominant reason for the observed modulation is not based on any optical mechanism associated with a change in reflectivity, optical confinement, and absorption, as all of them would have only a marginal effect on the carrier transport across the laser structure below the threshold current and cannot explain the observed strong changes
3.2 Incident power dependency of the optical enhancement
To study the optical enhancement further, the modulation dependency on the incident power is obtained for different bias current around the threshold current. Figure 1(b) gives the corresponding MIR peak power values plotted against the illumination average power at 225 mA (solid square), 230 mA (threshold without illumination, solid circle) and 235 mA (solid triangle), respectively. All three cases show a qualitatively similar nonlinear dependency on the incident NIR power. The optical emission increases very fast with the incident power and flattens out around an illumination power smaller than 1 mW. Similar behavior is observed for higher injection current far above the threshold. It is noticed that, at threshold at 235mA current, a 50 µW incident NIR can increase the QCL MIR peak power already by 7 times, showing the ability of switching MIR lasing with only tiny NIR optical power.
3.3 Spatial dependency of the optical enhancement
As previous results indicate that an increase of the carrier concentration within the active region is the source of the optical enhancement, a drastic dependency on the spot position is expected and observed. The inset of Fig. 2(a) gives the geometry diagram of the variation of the spot, while the laser was operated at threshold without laser emission. As shown in the main part of Fig. 2(a), the QCL MIR emission (solid squares) reaches its maximum only when the spot aligns with the active region and it decreases as the spot moves away. Yet it should be noticed that the MIR lasing behavior is still affected by 50% when the NIR laser spot hits on the substrate 500 μm away from the QCL active region. The exponential fit (solid curve) gives an estimated photon-generated free carrier in-plane traveling distance of 144 μm. This indicates that a large part of free carriers generated outside the active region will move into the laser cavity before recombination and clearly contribute to the enhancement.
3.4 Cavity mode spectrum change
Figure 2(b) shows the QCL wavelength shift of a given mode around 7.485 μm at different incident NIR power. Like the nonlinear dependency observed previously, the blue shift increases with incident NIR power and starts to saturate at about 0.1 mW. This cavity mode wavelength blue shift eliminates heating effect as major reason for the optical enhancement. As shown in the inset of Fig. 2(b), temperature increase results in about 0.16 nm/K red shift of the cavity modes, opposite from the observed blue shift, which is attributed to the photon-generated free carrier induced refractive index reduction.
3.5 Front facet reflectivity change
To evaluate the contribution of a possible front facet reflectivity change to the optical enhancement, the average refractive index is calculated based on cavity length and the measured wavelengths of two neighboring modes, and , as given in Eq. (1).
According to the obtained spectrum and Eq. (1), the calculated refractive index is about 3.3793 without optical illumination. For the observed 0.4 nm blue shift, the refractive index in the whole cavity has to be reduced by about 2.6 × 10−4 or 0.08 ‰, assuming a uniform change in the cavity. However, this small change will neither give substantially better optical confinement nor change the facet reflectivity by more than 0.012%.17], R = 0.3 and αm2 = 4.4637 cm −1 respect to a refraction index of 3.4, d = 0.1358 cm, and a typical range of γo from 25.5 cm−1 to 200 cm−1, the reflectivity R has to roughly be doubled to be responsible for an optical power increase of 35%, and thus easily experimentally observable. Yet, in corresponding experiments measuring the front facet MIR reflectivity under NIR illumination, less than 1% changes were observed. Therefore, the reflectivity change can be ruled out as major effect contributing to the gain enhancement. Actually, the photon-generated free carriers only reduce the refractive index, which can lead to facet and cavity losses and in turn increase the threshold, opposite to the observed threshold reduction.
3.6 Indirect gain change measurement
Based on the above experimental results, it is evident that the NIR illumination caused MIR optical enhancement can only be explained on the basis of a gain increase. Actually, the threshold reduction and the slope efficiency increase imply a decrease in the value of τ2/ τ32, which in turn increase the gain coefficient , where τ2 is the lower laser subband life time and τ32 is the nonradiative transition rate from upper laser level to the lower laser level. This gain change can be evaluated with the method described in reference . According to the equations for the threshold current and the slope efficiency , we have Eq. (3),Eq. (3) is independent on waveguide loss. As mentioned above, only marginal changes in Г and αm, are observed and hence we treat as a constant. So the gain coefficient g changes in the same way as ηIth at different incident power. For this purpose, I-L curves are obtained at different illumination powers, and then the corresponding threshold current and slope efficiency are deduced. As shown in Fig. 3 , the ηIth value increases with the illumination power and shows a similar nonlinearity as the QCL MIR power. This clearly indicates an increase in QCL optical gain coefficient under illumination.
In conclusion, a standard MIR QCL performance is improved by the front facet NIR illumination. The photon-generated free carriers lead to a change in the electron concentration, the electron transport and in turn a laser gain coefficient increase. Optical emission enhancement, switching-on below threshold, dynamic range increase, slope efficiency increase, blue shift in laser modes wavelength and nonlinear behavior are observed. As a wavelength converter, this optical approach can be used to translate NIR signal of the conventional fiber communication system into the MIR signal for the free space communication application. It might be extended to QCLs at different spectral range.
The authors would like to thank Scott S. Howard and Zhijun Liu in Princeton University for the QCL preparation, as well as Prof. Claire Gmachl for her support and helpful discussions. The authors also want to acknowledge Prof. Edward Whittaker for supporting equipment and Seong-wook Park and I-Chun Anderson Chen for assistance in Ti: sapphire laser.
References and links
1. J. Faist, F. Capasso, D. L. Sivco, C. Sirtori, A. L. Hutchinson, and A. Y. Cho, “Quantum Cascade Laser,” Science, New Series 264, 553–556 (1994).
2. K. Frank, Tittel, Yury A. Bakhirkin, Robert F. Curl, Anatoliy A. Kosterev, Matthew R. McCurdy, Stephen G. So and Gerard Wysocki, “Laser Based Chemical Sensor Technology: Recent Advances and Applications” in Advanced Environmental Monitoring, Young J. Kim and Ulrich Platt Editor, Springer Netherlands (2008)
3. R. Martini and E. A. Whittaker, “Quantum Cascade Laser Based Free Space Optical Communications,” J. Opt. Fiber. Commun. Rep. 2(4), 279–292 (2005). [CrossRef]
4. V. D. Jovanović, D. Indjin, N. Vukmirović, Z. Ikonić, P. Harrison, E. H. Linfield, H. Page, X. Marcadet, C. Sirtori, C. Worrall, H. E. Beere, and D. A. Ritchie, “Mechanisms of dynamic range limitations in GaAs/AlGaAs quantum-cascade lasers: Influence of injector doping,” Appl. Phys. Lett. 86(21), 211117 (2005). [CrossRef]
5. A. Tsekoun, R. Go, M. Pushkarsky, M. Razeghi, and C. K. Patel, “Improved performance of quantum cascade lasers through a scalable, manufacturable epitaxial-side-down mounting process,” Proc. Natl. Acad. Sci. U.S.A. 103(13), 4831–4835 (2006). [CrossRef] [PubMed]
6. H. Page, P. Collot, A. de Rossi, V. Ortiz, and C. Sirtori, “High reflectivity metallic mirror coatings for mid-infrared (λ ≈ 9 μm) unipolar semiconductor lasers,” Semicond. Sci. Technol. 17(12), 1312–1316 (2002). [CrossRef]
7. C. Sirtori, J. Faist, F. Capasso, D. L. Sivco, A. L. Hutchinson, and A. Y. Cho, “Quantum cascade laser with plasmon-enhanced waveguide operating at 8.4 µm wavelength,” Appl. Phys. Lett. 66(24), 3242 (1995). [CrossRef]
8. D. Dey, W. Wu, O. G. Memis, and H. Mohseni, “Injectorless quantum cascade laser with low voltage defect and improved thermal performance grown by metal-organic chemical-vapor deposition,” Appl. Phys. Lett. 94(8), 081109 (2009). [CrossRef]
9. M. D. Escarra, A. J. Hoffman, K. J. Franz, S. S. Howard, R. Cendejas, X. Wang, J.-Y. Fan, and C. Gmachl, “Quantum cascade lasers with voltage defect of less than one longitudinal optical phonon energy,” Appl. Phys. Lett. 94(25), 251114 (2009). [CrossRef]
10. A. J. Hoffman, S. Schartner, S. S. Howard, K. J. Franz, F. Towner, and C. Gmachl, “Low voltage-defect quantum cascade laser with heterogeneous injector regions,” Opt. Express 15(24), 15818–15823 (2007). [CrossRef] [PubMed]
11. P. T. Keightley, L. R. Wilson, J. W. Cockburn, M. S. Skolnick, J. C. Clark, R. Grey, G. Hill, and M. Hopkinson, “Improved performance from GaAs-AlGaAs quantum cascade lasers with enhanced upper laser level confinement,” Physica E 7(1-2), 8–11 (2000). [CrossRef]
12. C. Zervos, M. D. Frogley, C. C. Phillips, D. O. Kundys, L. R. Wilson, M. Hopkinson, and M. S. Skolnick, “All-optical switching in quantum cascade laser,” Appl. Phys. Lett. 90(5), 053505 (2007). [CrossRef]
13. G. Chen, C. G. Bethea, R. Martini, P. D. Grant, R. Dudek, and H. C. Liu, “high speed all-optical modulation of a standard quantum cascade laser,” Appl. Phys. Lett. 95(10), 101104 (2009). [CrossRef]
14. T. Aellen, M. Beck, N. Hoyler, M. Giovannini, J. Faist, and E. Gini, “Doping in quantum cascade lasers. I. InAlAs–InGaAs/InP midinfrared devices,” J. Appl. Phys. 100(4), 043101 (2006). [CrossRef]
15. J. Mc Tavish, D. Indjin, and P. Harrison, “Aspects of the internal physics of InGaAs/InAlAs quantum cascade lasers,” J. Appl. Phys. 99(11), 114505 (2006). [CrossRef]
16. B. E. A. Saleh, and M. C. Teich, Fundamentals of Photonics (John Wiley & Sons Inc. USA, 1991), Chap. 16.
17. Z. Liu, D. Wasserman, S. Howard, A. J. Hoffman, and C. Gmachl et al.., “Room-Temperature Continuous-Wave Quantum Cascade Lasers Grown by MOCVD Without Lateral Regrowth,” IEEE Photon. Technol. Lett. 18(12), 1347–1349 (2006). [CrossRef]
18. Carlo Sirtori and Roland Teissier, “Quantum cascade lasers: overview of basic principles of operation and state of the art” in Intersuband transitions in quantum structures, Roberto Paiella Editor (McGraw-Hill New York, 2006), 15.
19. C. Sirtori, S. Barbieri, P. Kruck, V. Piazza, M. Beck, J. Faist, U. Oesterle, P. Collot, and J. Nagle, “Influence of DX centers on the performance of unipolar semiconductor lasers based on GaAs-AlxGa1-xAs,” IEEE Photon. Technol. Lett. 11(9), 1090–1092 (1999). [CrossRef]