We present the design and realization of short-wavelength (λ = 4.53 μm) and buried-heterostructure quantum cascade lasers in a master oscillator power amplifier configuration. Watt-level, singlemode peak optical output power is demonstrated for typical non-tapered 4 μm wide and 5.25 mm long devices. Farfield measurements prove a symmetric, single transverse-mode emission in TM00-mode with typical divergences of 25° and 27° in and perpendicular to growth direction, respectively. We demonstrate singlemode tuning over a range of 7.9 cm−1 for temperatures between 263K and 313K and also singlemode emission for different driving currents. The side mode suppression ratio is measured to be higher than 20 dB.
© 2013 OSA
Since the first demonstration of distributed feedback (DFB) quantum cascade (QC) lasers in 1997 , tremendous progress has been made in the development of singlemode emitting QC lasers [2–4]. Such devices are of particular interest because prominent molecules like CO2, N2O, CH4 and many more have their fundamental roto-vibrational absorptions in the mid-infrared (3–12 μm) spectral region. Pulsed and continuous wave DFB-QC laser based trace gas sensors could already show detection levels in the parts-per-billion range or better, proving the excellent suitability of such systems for high-resolution spectroscopy [5–8]. As has also been shown recently [9, 10] pulsed, singlemode QC laser sources emitting in the wavelength range around 2200 cm−1 are of particular interest in meteorological science e.g. for the isotope-specific spectroscopy of N2O .
In order to overcome limitations in wavelength-tuning of typical DFB QC lasers, widely tunable external-cavity QC laser systems have been developed in the past couple of years  as well as very recently in-situ tunable DFB QC laser devices, where only the applied current has to be varied [12, 13].
Another issue, besides increasing the spectral tuning range, which concerns singlemode devices is increasing the emitted optical output power. To account for this different approaches have been proposed. The most obvious approach is to increase the length of a normal DFB device . Therefore the coupling strength  of the grating has to be reduced accordingly, where L is the periodicity of the grating and Dn and neff are the refractive index step and effective refractive index of the mode, respectively. Otherwise the κ · l (l = device length) product is too high which leads to a reduced optical output power . On the contrary, a too low κ · l value results in a bad mode discrimination.
Figure 1(a) shows the theoretical transmission-curve (left-hand scale) and modal losses (right-hand scale) of a 5.25 mm long DFB QC laser (black) with a periodicity of 717 nm and a coupling strength κ of 1.4 cm−1 simulated with a transmission matrix approach. The κ-value is similar to the one used in , where the loss component of the coupling was used to lift the degeneracy between the two modes at the edge of the stopband .
This device shows a weakly pronounced stopband with a spectral width of 0.62 cm−1 and a mode discrimination of 0.1 cm−1 and 0.14 cm−1 towards the next mode which is not lying at the edge of the stopband and the Fabry-Pérot (FP) mode level, respectively.
To simulate the robustness of such an approach towards fabrication fluctuations, a perturbation of the effective refractive index with a Gaussian distribution was included every 10 μm along the waveguide (green curve). This perturbation represents small laser ridge width fluctuations on the order of 130 nm of a 4 μm wide laser ridge which can easily occur during device fabrication. As can be seen such a small perturbation completely washes out the stopband and removes any singlemode behavior.
An important parameter in designing DFB lasers is the spectral gain margin for singlemode emission. It indicates the maximum spectral distance from the gain peak for which singlemode emission can still be realized by e.g. designing a DFB grating at this wavelength. Going beyond this spectral position will lead to multimode emission because the FP modes reach their lasing threshold before the DFB mode does. The gain margin is of particular interest when different wavelengths have to be realized within one process. One good example for this is the high power array by Rauter et al. where each individual device within the array is addressing a different wavelength .
The condition for maintaining singlemode emission reads as follows:Fig. 1(a) (αDFB = 2.36 cm−1, αm = 2.5 cm−1) that the gain ratio has to be gDFB/g(λmax) > 0.94 for singlemode emission. Including the perturbation no singlemode emission can be achieved any more for the DFB device.
Another approach for increasing the optical output power is to build a two-section cavity. One section consists of a DFB waveguide which acts as a singlemode seed. The second part is a FP cavity which acts as an amplifier and which can be scaled up in length to increase the optical output power without influencing the spectral purity of the emission. The mayor advantage of such a configuration is, that it improves the mode discrimination significantly. This is due to the use of a short DFB section with strong grating coupling coefficient which is more robust towards defects and fabrication fluctuations. An additional positive side-effect of such a two-section geometry is that spatial hole burning, which is one of the main reasons for multimode emission in QC lasers , is reduced significantly. Therefore an anti-reflection (AR) coating is put at the end-facet of the FP section. This transforms the standing wave in the FP amplifier to a traveling mode suppressing spatial hole burning in this section. The AR-coating is also needed to suppress self-lasing of the FP cavity and additional FP modes which are present within the stopband region of this two-section configuration as can be seen in Fig. 1(b).
Such so-called master-oscillator power-amplifier (MOPA) devices have been demonstrated based on a QC laser first by Troccoli et al.  in 2002. Recently, results have been published for single emitters without AR-coating which can be tuned quasi-continuously with mode-hopes on the FP modes of the amplifier section , and arrays, which by tapering the amplifier section up to >100 μm ridge width, result in Watt-level peak optical output power at a low duty-cycle of 0.025% .
In contrast to those devices an asymmetric approach is presented in this paper with a short DFB section and a long FP cavity. Figure 1(b) shows the simulated transmission curve (left-hand scale) and the modal losses (right-hand scale) for a MOPA device (black) with 1.25 mm long DFB section and a 4 mm long amplifier including front-facet AR-coating. In this configuration the stopband has a spectral width of 4.2 cm−1 (κ ∼ 35 cm−1). Due to the asymmetric device geometry the resulting modal losses are asymmetric favoring one of the two DFB modes with a discrimination of 0.5 cm−1. The corresponding mode with the lowest losses is lying at 2.7 cm−1 and the discrimination towards the next mode which is not lying at the edge of the stopband and the FP modes is about 1.6 cm−1 and 2.8 cm−1, respectively. These values are more than one order of magnitude higher compared to the normal DFB grating proving the significantly improved mode selection mechanism of the MOPA geometry. The gain ratio for maintaining singlemode emission of the MOPA device yields 0.49 (αDFB = 2.7 cm−1, αm = 5.5 cm−1) which is significantly lower than for the DFB device. Including the perturbation, which is done in the green curve of Fig. 1(b), barely influences the effect of the grating. The transmission and modal losses are basically unchanged as well as the gain ratio for singlemode emission. This result again shows the improved singlemode behavior for the MOPA device when compared to the normal DFB.
2. Sample fabrication
The devices presented in section 3 of this paper are based on a strain-balanced In0.635Ga0.365As/Al0.665In0.335As active region design grown on InP substrate by molecular beam epitaxy. Details on the design, which is the result of a fully automatized genetic optimization algorithm, can be found in a previous publication . The only difference is that the average doping in the active region was reduced by 33% to 6.7x1010cm−2. After growth the layer was fabricated in a buried heterostructure configuration to result in low-loss narrow ridges [4, 21]. Into a 2.5 mm long section along the waveguide a first-order DFB grating was incorparated whereas the second section consists of a up to 5 mm long straight FP cavity. The maximum length of the amplifier section was a compromise between the maximum current the pulse driver used for the measurements can supply, the fabrication design which should include several wavelengths in one fabrication process and a reasonable device length for mounting. Figure 2 shows the picture of a mounted MOPA device with 2.5 mm long DFB (upper part) and 4 mm long FP section (lower part). For an high thermal conductance  enabling high duty-cycle operation up to continuous-wave mode which is needed for different spectroscopic techniques like e.g. the intra-pulse  or the QEPAS  scheme, no tapering was added to the amplification section. Moreover a non-tapered device geometry results in a more symmetric farfield and non-astigmatic beam. As will be shown in the next section a symmetric farfield pattern could be achieved with this approach.
A single-layer Al2O3-coating (refractive index n = 1.647 at 4.5 μm ) with a thickness of 687 nm was put as AR-coating on the front-facet. This results in a simulated residual front-facet reflectivity of 0.4% which was measured to be 0.6% on the final devices. This value is low enough to eliminate lasing from the FP-cavity.
3. Sample characterization
A typical device with 4 μm wide front-facet and 1.25 mm long DFB seeding section followed by a 4 mm long straight FP amplifier is presented in this section.
Figure 3 shows the symmetric and single-lobe farfield of the MOPA QC laser in TM00-mode along and perpendicular to the growth direction of the device. For this measurement a pyro-electric detector mounted on a 2-axis goniometer was used, which can move in a half-sphere around the front-facet of the device. The MOPA was driven at its peak optical output power, i.e. at 1500 mA by 50 ns long current pulses at 1% duty cycle and the temperature was fixed to 293K. The full width at half maximum is measured to be 25° in growth direction and 27° perpendicular to that direction.
Figure 4 shows the light-current-voltage characteristics of this device for 100 ns long pulses at 1% duty cycle measured between 313K and 263K. A peak optical output-power ranging from 0.6 W up to 1 W is obtained for this temperature range. Note that Watt-level optical output power could be achieved for non-tapered devices in this case.
Figure 5(a) shows the spectral emission for different driving currents at a constant temperature of 293K and a pulselength of 50 ns at 1% duty cycle. Depending on the current applied to the device the emission wavelength can be tuned between 2208.9 cm−1 and 2209.8 cm−1. For values close to the lasing threshold Ith, i.e. 700 mA and 800 mA corresponding to 1.05*Ith and 1.2*Ith, the wavelength stays constant at 2209.8 cm−1. Further increasing the current up to 1500 mA (= 2.25*Ith) leads to singlemode tuning due to intra-pulse thermal heating of the device with a total tuning range of about 1 cm−1. The side-mode surpression-ratio is better than 23 dB. Figure 5(b) shows the spectral emission for a constant injection current of 1500 mA (50 ns, 1% duty cycle), which is at the peak optical output power and varied temperatures between 263K and 313K. The total singlemode tuning spans a range of 7.9 cm−1 between 2206 cm−1 and 2213.9 cm−1.
In conclusion the design, fabrication and characterization of buried-heterostructure and short-wavelength MOPA QC lasers is presented together with a simulation-based discussion and comparison with standard DFB QC lasers. Typical narrow and long (4 μm × 5.25 mm) devices, show a symmetric TM00 farfield and singlemode peak optical output power of up to 1 W. The devices were singlemode in the entire investigated temperature range between 263K and 313K with a side-mode suppression-ratio of better than 20 dB.
The authors would like to thank P. Jouy, A. Bismuto and Y. Bonetti for fruitful discussions and the help in process development and M. Ebnöther for expert technical assistance. Financial support by the Swiss National Science Foundation (SNSF) within the framework of the NCCR Quantum Photonics project is gratefully acknowledged.
References and links
1. J. Faist, C. Gmachl, F. Capasso, C. Sirtori, D. L. Sivco, J. N. Baillaergeion, and A. Y. Cho, “Distributed feedback quantum cascade lasers,” Appl. Phys. Lett. 70, 2670–2672 (1997). [CrossRef]
2. A. Wittmann, M. Giovannini, J. Faist, L. Hvozdara, S. Blaser, D. Hofstetter, and E. Gini, “Room temperature, continuous wave operation of distributed feedback quantum cascade lasers with widely spaced operation frequencies,” Appl. Phys. Lett. 89, 141116 (2006). [CrossRef]
3. Q. Y. Lu, Y. Bai, N. Bandyopadhyay, S. Slivken, and M. Razeghi, “2.4 W room temperature continuous wave operation of distributed feedback quantum cascade lasers,” Appl. Phys. Lett. 98,181106 (2011). [CrossRef]
4. B. Hinkov, A. Bismuto, Y. Bonetti, M. Beck, S. Blaser, and J. Faist, “Singlemode quantum cascade lasers with power dissipation below 1W,” El. Lett. 48, 646–647 (2012). [CrossRef]
7. A. A. Kosterev and F. K. Tittel, “Chemical sensors based on quantum cascade lasers,” J. Quantum Electron. 38, 582–591 (2002). [CrossRef]
8. A. Manninen, B. Tuzson, H. Looser, Y. Bonetti, and L. Emmenegger, “Versatile multipass cell for laser spectroscopic trace gas analysis,” Appl. Phys. B 109, 461–466 (2012). [CrossRef]
9. J. R. Köster, R. Well, B. Tuzson, R. Bol, K. Dittert, A. Giesemann, L. Emmenegger, A. Manninen, L. Cádenas, and J. Mohn, “Novel laser spectroscopic technique for continuous analysis of N2O isotopomers - application and intercomparison with isotope ratio mass spectrometry,” Rapid Comm. Mass Spectrom. 27, 216–222 (2013). [CrossRef]
10. P. Wunderlin, M. F. Lehmann, H. Siegrist, B. Tuzson, A. Joss, L. Emmenegger, and J. Mohn, “Isotope signatures of N2O in a mixed microbial population system: constraints on N2O producing pathways in wastewater treatment,” Environ. Sci. Technol. 47, 1339–1348 (2013).
11. A. Hugi, R. Terazzi, Y. Bonetti, A. Wittmann, M. Fischer, M. Beck, J. Faist, and E. Gini, “External cavity quantum cascade laser tunable from 7.6 to 11.4 μm,” Appl. Phys. Lett. 95, 061103 (2009). [CrossRef]
12. T. S. Mansuripur, S. Menzel, R. Blanchard, L. Diehl, C. Pflügl, Y. Huang, J. H. Ryou, R. D. Dupuis, M. Loncar, and F. Capasso, “Widely tunable mid-infrared quantum cascade lasers using sampled grating reflectors,” Opt. Express 20, 23339–23348 (2012). [CrossRef] [PubMed]
13. S. Slivken, N. Bandyopadhyay, S. Tsao, S. Nida, Y. Bai, Q. Y. Lu, and M. Razeghi, “Sampled grating, distributed feedback quantum cascade lasers with broad tunability and continuous operation at room temperature,” Appl. Phys. Lett. 100, 261112 (2012). [CrossRef]
14. J. Faist, Quantum Cascade Lasers (Oxford University, 2013) [CrossRef]
15. A. Wittmann, “High-performance quantum cascade laser sources for spectroscopic applications,” PhD. thesis 18363, ETH Zürich (2009).
16. P. Rauter, S. Menzel, A. K. Goyal, C. A. Wang, A. Sanchez, G. W. Turner, and F. Capasso, “High-power arrays of quantum cascade laser master-oscillator power-amplifiers,” Opt. Express 21, 4518–4530 (2013). [CrossRef] [PubMed]
17. A. Gordon, C. Y. Wang, L. Diehl, F. X. Kärtner, A. Belyanin, D. Bour, S. Cozine, G. Höfler, H. C. Liu, H. Schneider, T. Maier, M. Troccoli, J. Faist, and F. Capasso, “Multimode regimes in quantum cascade lasers: From coherent instabilities to spatial hole burning,” Phys. Rev. A 77, 053804 (2008). [CrossRef]
18. M. Troccoli, C. Gmachl, F. Capasso, D. L. Sivco, and A. Y. Cho, “Mid-infrared (lambda ∼ 7.4 μ m) quantum cascade laser amplifier for high power single-mode emission and improved beam quality,” Appl. Phys. Lett. 80, 4103–4105 (2002). [CrossRef]
19. P. Fuchs, J. Friedl, S. Höfling, J. Koeth, A. Forchel, L. Worschech, and M. Kamp, “Single mode quantum cascade lasers with shallow-etched distributed Bragg reflector,” Opt. Express 20, 3890–3897 (2012). [CrossRef] [PubMed]
20. A. Bismuto, R. Terazzi, B. Hinkov, M. Beck, and J. Faist, “Fully automatized quantum cascade laser design by genetic optimization,” Appl. Phys. Lett. 101, 021103 (2012). [CrossRef]
21. A. Bismuto, R. Terazzi, M. Beck, and J. Faist, “Electrically tunable high performance quantum cascade laser,” Appl. Phys. Lett. 96, 141105 (2010). [CrossRef]
22. J. Manne, A. Lim, J. Tulip, and W. Jäger, “Sensitive detection of acrolein and acrylonitrile with a pulsed quantum-cascade laser,” Appl. Phys. B 107, 441–447 (2012). [CrossRef]
23. A. Kosterev, G. Wysocki, Y. Bakhirkin, S. So, R. Lewicki, M. Fraser, F. Tittel, and R. F. Curl, “Application of quantum cascade lasers to trace gas analysis,” Appl. Phys. B 90, 166–176 (2008). [CrossRef]
24. E. Palik, Handbook of Optical Constants of Solids II (Academic, 1998).