Open Access
Add to BibSonomyAbstract
We present a technique for enhancing the light output power of quantum cascade lasers (QCLs) by tilting of the front facet, which leads to a change of the modal reflectivity, resulting in an asymmetric light intensity distribution along the laser cavity. This asymmetry provides most of the light being emitted through one facet of the laser. An experimental study of threshold current, slope efficiency and light output power as a function of the front facet angles were performed and compared to conventional QCLs. The lasers with a front facet angle of 8° shows a 20% improved power output from the front facet.
©2013 Optical Society of America
1. Introduction
Quantum cascade lasers (QCLs) are semiconductor heterostructure lasers built from multiple quantum wells and are reliable optical light sources in the mid-infrared (MIR) and terahertz (THz) regions of the electromagnetic spectrum [1, 2]. They have many attractive features, such as freely designable emission wavelengths, and the emission of multiple photons per electron [3]. A wide range of applications for medical, optical communication, and environment monitoring have been realized with QCLs [4–7].
Most applications of semiconductor lasers require the emission from only one facet of the laser cavity. If identical facets are used, one facet emits only half of the total light power. Usually, the other facet is covered by high reflectivity materials to improve the feedback of the optical modes into the cavity and to insure most of the light escapes from only one facet, resulting in an improved optical output power. Under these circumstances, the enhancement of the optical power performance can be achieved by optimizing the reflectivity of the laser facets. An anti-reflection (AR) coating at the front facet and a high-reflection (HR) coating at the back facet are the most established techniques to manipulate the reflectivity [8, 9].
Tilting the laser facet is also a useful approach to tune the reflectivity [10]. However, the tilted facet produces light losses, since a portion of reflected light at the tilted facet is not coupled back into the laser mode [11]. For this reason, the tilted facet techniques have been less desirable for optical power improvement of light sources and, instead, used limitedly for laser applications such as traveling wave amplifiers (TWA) and superluminescent light emitting diodes (SLED) [11, 12]. Also, the tilted facet was introduced to achieve spatial beam quality control, without the power improvement, for broad area QCLs [13]. However, tilting one facet of two laser facets was successfully demonstrated as a competent approach to achieve light escaping predominantly from one facet [14]. The difference in the reflectivities between tilted and normal facets induces an asymmetric light intensity distribution along the laser cavity, resulting in asymmetric light power emission from the facets. When the ratio of the reflectivities is high enough, most of the light will emit through the facet which has the lower reflectivity. This effect is now applied to QCLs and will be discussed in further detail later in this paper.
In this work, we investigate the effect of the tilted front facet on the power performance of QCLs. The tilted facets were fabricated by focused ion beam (FIB) milling. The FIB technique has been successfully used to demonstrate advanced optical semiconductor devices [15, 16]. This technique provides an important advantage for experimental studies that it is possible to record the characteristics of the very same device before and after FIB milling. If a more scalable approach is demand, the same structure could be also achieved, without a usage of FIB process, by using the combination of a cleaved front facet and an etched back facet at a tilted waveguide.
2. Light intensity distribution along a laser cavity
A semiconductor laser consists of a gain medium in a cavity, including front and back facets with reflectivity Rfront and Rback. The optical modes, which are confined to the cavity, bounce back and forth between the facets, thereby enhancing the amplification process through the feedback mechanism with a constructive interference of the modes. The amplification process will finally reach a state where the optical gain equals the optical losses in the system. This is a threshold condition and the corresponding threshold current density Jth can be written as
where G, Γ, αw and L are indicating the gain coefficient, optical confinement factor, waveguide loss and cavity length, respectively.2.1 As-cleaved facets and a coated front facet
Figure 1(a) shows a sketch for the symmetric light intensity distribution along the cavity of a Fabry-Pérot laser (Rfront = Rback). The dashed and solid arrows are indicating the light reflection and transmission T at the laser facets. The P1 and P3 are indicating the reflected light intensities at the front and back facet, respectively. The P2 and P4 are indicating the exponentially amplified light intensities after forward and backward propagation arriving at the other facet. The light intensities P1, P2, P3 and P4 are related to each other by optical net gain (g) as P1 × egL = P2 and P3 × egL = P4 and by the reflectivity as P2 × Rfront = P3 and P4 × Rback = P1. According to this dependence, one can derive that the ratio of the power arriving at the facets:
The emitted power from the facets is the power arriving at the facet times the transmission T (T = 1 – R), shown in Eq. (3) [17, 18].Figure 2(b) shows a sketch of the asymmetric light intensity distribution along the cavity of the laser with an AR coated front facet (Rfront ç ≠ Rback). According to Eq. (3), the light power emission ratio between the front and back facet consists of two parts: first, the ratio of arriving at the facets given by (Rback / Rfront)1/2, and second, the ratio of the transmission of the facets (Tfront / Tback). For example, let us suppose that Rfront = 0.1 and Rback = 0.3. Then the values of (Rback / Rfront)1/2 and (Tfront / Tback) are 1.73 and 1.28, respectively. This illustrates that the power emission ratio is strongly influenced by the ratio of power arriving at the facets. This fact is important to explain the power emission ratio with tilted front facet in the next section.
Fig. 1 Sketch of the (a) symmetric and (b) asymmetric light intensity distribution inside a laser cavity. The arrows (dash) and arrows (line) are indicating the light reflection and transmission at the laser facet.
Fig. 2 Sketch of the asymmetric light intensity distribution inside a laser cavity for the tilted front facet. The arrows (dash), arrows (line) and symbols ( × —–) are indicating the light reflection, emission at the laser facets and reflected light at the tilted front facet which is not coupled back into the cavity, respectively.
2.2 A tilted front facet
Figure 2 shows a sketch of the light intensity distribution along the cavity of the laser with a tilted front facet. The back facet is left as-cleaved (Rfront ≠ Rback). The symbols ( × —–) are indicating the reflected light at the tilted front facet, but they do not couple back into the laser mode. This additional light loss results in a modification of the power transmission to T < (1 − R), Eq. (4). Nevertheless, the Eq. (1) is valid for the tilted facet since the reflectivity is interoperated as modal reflectivity. The high emission ratio still can be expected by the high ratio of the power arriving at the facets (Eq. (2) since the ratio of (Tfront / Tback) is less influential for the emission ratio as mentioned above. Although the models are drastically simplified, this approximation is helpful to show the main enhancement mechanism of the power emission from the tilted front facet.
Based on Snell’s law, the critical angle of the incident light at the interface between our semiconductor structure and air is 18.2°, with an effective reflective index (neff) of 3.2 for the semiconductor and 1 for the air. Therefore, for a plane wave, the transmission of the tilted facet would go to zero when the tilting angle is larger than the critical angle. In our case, a narrow confined optical mode is no longer a plane wave and, hence, the situation is more complex.3. Fabrication, measurement set up and focused ion beam (FIB)
The QCL heterostructure in this work was grown by molecular beam epitaxy (MBE). It consists of thirty-five periods of an In0.53Ga0.47As/In0.52Al0.48As two-phonon resonance active core [19]. The low-doped n-type (n = 2 × 1017 cm−3) InP substrate is acting as the lower waveguide cladding layer. The active region is sandwiched between two 500 nm InGaAs (n = 5 × 1016 cm−3) core layers. The upper waveguide cladding consists of a 1500 nm InAlAs layer (n = 1 × 1017 cm−3) followed by a 800 nm InAlAs-layer (n = 2 × 1017 cm−3) grown on top. The structure is covered by a 350 nm highly doped InGaAs layer (n = 4 × 1018 cm−3). A 100 nm InGaAs (n = 2 × 1019 cm−3) contact layer completed the QCL growth [20]. The QCL ridges were defined from a single die using contact lithography and then SiCl4/Ar reactive ion etching. Thereafter, a 300 nm thick silicon nitride (SiN) insulating layer was deposited by plasma-enhanced chemical vapor deposition. The SiN was then removed from the top of the ridges. A 500 nm Ti/Au contact finished the topside processing and a backside ohmic contact Ge/Au/Ni/Au (15/30/14/150 nm) completed the standard QCL fabrication. Ridges with cleaved facets were soldered with indium to copper heat sinks, wire bonded, and installed on a Peltier cooler to stabilize the heat sink at 293 K. The length for all the fabricated QCLs is 2.4 mm. To suppress higher order lateral modes, the width of the fabricated QCLs was chosen to be 10 µm since narrower ridges exhibit a larger discrimination against higher order lateral modes.
The lasers were operated in pulsed mode with a pulse length of 100 ns at a repetition rate of 5 kHz (0.05% duty-cycle). Optical output for the power and spectrum were measured with a calibrated deuterated triglycine sulfate detector in a Fourier transform infrared spectrometer. The emission wavelength (λ) of the QCLs is ~8 µm at 293 K. The far field measurements were carried out using a liquid-nitrogen cooled mercury cadmium telluride (MCT) detector mounted on a computer-controlled rotational stage. The active size of the detector is 1 mm. The distance between the detector and laser facet is 50 mm. All the measurements were performed at room temperature.
After the fundamental characterization steps are completed, tilted facets with angles of 4, 8, 12, 17 and 22° where fabricated (Fig. 3). Only the front facet of the QCLs was milled, while the back facet was left as cleaved. The FIB milling was performed using a ZEISS NEON 40ESB with a Ga + liquid ion source. The ion beam was accelerated at the energy of 30 keV with a beam current of 300 pA, focused onto the surface of the ridge end and raster scanned in the areas to be milled. The angle of the milled facet is defined by a computerized rotational milling pattern control system. After the FIB process, The angle of the milled facets was measured by a scanning electron microscope (SEM) imaging system. We consider that an error of ± 0.5° is possible. The FIB process is similar to the one presented in [21], Z. Y. Zhang et al. An example of a milled QCL facet is shown in Fig. 3(b), where the facet has been tilted by 12°.
Fig. 3 (a) Sketch of a QCL with a tilted front facet. Only the front facet of the QCLs was milled, while the back facet was left as cleaved. (b) A scanning electron microscope picture shows the fabricated tilted front facet (θF = 12°). ( + ) and (–) signs are defined for the far field measurement in the following section.
Typical drawbacks of the FIB milling process are an implantation of the Ga+ ions and the damage of the surface, attributed to the exposure of the surface to high dosage Ga+ ions during the milling process. However, in our work, the direction of focused ion beam is parallel to the laser facet where the laser beam emits. Hence, the laser facet is less affected from the Ga+ ions implantation and the damage. In other words, the most damaged part is the substrate of the QCL which is not influencing to the laser performance. In order to study the effect of the tilted facets, all measurements are done on the emission from the tilted facets at room temperature before and after milling. To check for damage at the surface by the FIB milling, we compared the output power performance between cleaved and milled 0° facet angle (θF). They show nearly the same output power performance. Therefore, the FIB induced surface damage is negligible in this experiment.
4. Result and discussion
4.1 Modal reflectivity of a tilted front facet
In Fig. 4(a), the ratios of the threshold current densities after FIB milling to before FIB milling (≥ 1) are shown for all given facet angles. The modal reflectivity of the tilted front facet decreases as a function of the facet angle, resulting in a gradual increase of the threshold current density as a function of the facet angle. In Fig. 4(b), the modal reflectivity of the tilted front facet (Rfront) is determined by the measured threshold current density using Eq. (1) [15]. For this, we used a constant waveguide loss of αw = 13.2 cm−1 obtained by measuring Jth as a function of the cavity length of a conventional QCL [20]. Generally, the reflectivity of cleaved facet of a conventional QCL is 0.27 ( = Rback). The determined modal reflectivities of the tilted front facet are 0.21, 0.12, 0.073, 0.011 and 0.0021 for the facet angle of 4°, 8°, 12°, 17°, and 22°, respectively.
Fig. 4 (a) The ratio of the threshold current densities of QCLs after FIB milling to before FIB milling for all given facet angles. (b) The experimentally determined modal reflectivity (triangles) of the tilted facet as a function of the facet angles.
4.2 A comparison of slope efficiencies between the front and back facets
Figure 5(a) shows an example of light-current-voltage (LIV) characteristics for the emission from the front facet (tilted) and the back facet (as-cleaved) of the QCL with a front facet angle of 17°. In this case, 87% of the total output power is out-coupled through the tilted front facet, showing successfully the asymmetry of the light intensity distribution. In Fig. 5(b), the slope efficiency (ηs = dP / dI) ratios of front (dPfront / dI) to back (dPback / dI) facets as a function of facet angles are shown (triangles). The Pfront and Pback are light output power from the front and back facets, respectively. And they are compared to the ratios (stars) of the power arriving at the laser facets (Eq. (2)). Equation (2) can be valid for the tilted facet since (Tfront / Tback) is negligible as mentioned in a section 2.2. In order to avoid additional phenomenon in the comparison of the slope efficiencies, such as gain saturation induced by higher current injection, only the linear parts (just above threshold) of the LI curves are taken for the comparison. Up to 17°, the measured ratio of the slope efficiencies follows the ratio of power arriving at the facets. However, for the facet angle of 22°, the ratio of the slope efficiency is decreased. The emission from the tilted front facet is decreased strongly by the reduced transmission of the front facet. In the case of a plane wave, the transmission would already be zero.
Fig. 5 (a) A comparison of LIV characteristics (the linear part of the LI curve just above threshold) for the emission from front and back facets of the QCL with a facet angle of 17°. (b) Ratios of the slope efficiencies between front and back facet (triangles), as a function of facet angles compared to the ratios (Eq. (2) of the power arriving at the laser facets (stars).
4.3 Enhancement of light output power by a tilted front facet
In Fig. 6(a), the light-current density-voltage (LJV) characteristics of the QCLs with a facet angle of 8° (after FIB milling) are compared to the conventional QCL (before FIB milling). The LJV characteristics are determined by the light collected from the front facets only. The slope efficiency and peak power of the milled facets are improved by 20% and 21%, respectively. The threshold current density of the QCLs is increased by 8.7% due to the decreased modal reflectivity. The experimentally determined modal reflectivity of the facet is 0.12. The ratio of slope efficiencies, as a function of the facet angle, is shown in Fig. 6(b). The facet angle of 8° shows the most improved slope efficiency, compared to the other given facet angles. Up to the 17° facet angle, the slope efficiencies are comparable to the conventional QCLs. For the laser with a facet angle of 22°, the transmission at the front facet is significantly low, resulting in a poor laser power performance.
Fig. 6 (a) A comparison of LJV characteristics between before and after FIB milling for QCLs with a facet angle of 8°. (b) The ratios, between before and after FIB milling, of ηs for the QCLs as a function of the facet angles, based on the emitted light from the front facet.
4.4 Lateral far field profiles
In Fig. 7(a), the lateral far field mode profiles of the QCLs are shown. The profiles were measured nearly at the peak optical power along the lateral (Y) direction [see Fig. 3(a)]. In order to compare precisely the emitted beam angles from the tilted facets, all measured lasers are located side by side on the same chip. The data have been normalized to allow for a comparison. The dash line at 0° represents the ridge normal direction. Positive and negative angles are defined in Fig. 3(b). The far field profiles reveal two remarkable behaviors. First, the profiles from all given facet angles are single-lobed with decreasing a full width at half maximum (FWHM) for larger facet angles, showing improved beam quality up to 17°. Second, the beam emission angles from the ridge direction are small for the facet angles up to 17°.
Fig. 7 (a) Far field profiles measured along the lateral (Y) direction [see Fig. 3(a)] of various facet angles of QCLs, driven nearly at the peak optical power. The dash line at 0° represents the ridge normal direction. (b) The beam emission angle from the ridge normal (circles) and FWHM (squares) as a function of the facet angles are shown.
In Fig. 7(b), the beam emission angles (θE) from the ridge direction for θF of 4°, 8°, 12° and 17° are −4°, −3°, −6°, and −4°, respectively (−8°, −11°, −18°, and −21° from the facet normal). Based on Snell’s law, using the emission angles in the air and neff of 3.2 for the device, the incident angles into the tilted facets inside the cavity are calculated to be around 2.5°, 3.4°, 5.5° and 6.4° for θF of 4°, 8°, 12° and 17°, respectively. The far field for the θF of 22° shows an emission angle of −40° from the ridge direction (−62° from the facet normal), which would results in an incident angle of 16° into the tilted facets. These results indicate that the mode inside the cavity no longer propagates linearly along the waveguide due to the nonparallel facets. For a better understanding of the beam emission angle from the tilted front facet, better knowledge of the optical mode profile inside the laser cavity would be required. However, this is beyond the scope of this paper.
5. Conclusion
In summary, we have demonstrated that the technique of the tilted front facet not only reduces modal reflectivity but also increases output power performance. For the facet angle of 8°, the emitted light power is increased by 20%, compared to the power of a conventional as cleaved QCLs. For the facet angle of 17°, 87% of the total output power is out-coupled through the tilted front facet. Up to a 17° facet angle, the QCLs show comparable or increased slope efficiencies to conventional QCLs. The far field profiles show a single-lobe and decreasing FWHM as a function of the facet angle. To achieve enhanced light output power, the tilted front facet technique looks promising since the technique provides higher beam quality and an independence from the coating techniques.
Acknowledgments
The authors acknowledge the support by the Austrian projects IR-ON (Austrian Science Fund (FWF): F2503-N17), the Austrian Nanoinitiative project PLATON, the “Gesellschaft für Mikro- und Nanoelektronik” GMe.
References and links
1. J. Faist, F. Capasso, D. L. Sivco, C. Sirtori, A. L. Hutchinson, and A. Y. Cho, “Quantum Cascade Laser,” Science 264(5158), 553–556 (1994). [CrossRef] [PubMed]
2. S. Kumar, “Recent progress in Terahertz Quantum Cascade Lasers,” IEEE J. Quantum Electron. 17(1), 38–47 (2011). [CrossRef]
3. Y. Yao, A. J. Hoffman, and C. F. Gmachl, “Mid-infrared quantum cascade lasers,” Nat. Photonics 6(7), 432–439 (2012). [CrossRef]
4. M. Hannemann, A. Antufjew, K. Borgmann, F. Hempel, T. Ittermann, S. Welzel, K. D. Weltmann, H. Völzke, and J. Röpcke, “Influence of age and sex in exhaled breath samples investigated by means of infrared laser absorption spectroscopy,” J. Breath. Res. 5(2), 027101 (2011). [CrossRef] [PubMed]
5. F. Capasso, R. Paiella, R. Martini, R. Colombelli, C. Gmachl, T. L. Myers, M. S. Taubman, R. M. Williams, C. G. Bethea, K. Unterrainer, H. Y. Hwang, D. L. Sivco, A. Y. Cho, A. M. Sergent, H. C. Liu, and E. A. Whittaker, “Quantum Cascade Lasers: Ultrahigh-speed operation, optical wireless communication, narrow linewidth, and far-infrared emission,” IEEE J. Quantum Electron. 38(6), 511–532 (2002). [CrossRef]
6. G. Wysocki, R. Lewicki, R. F. Curl, F. K. Tittel, L. Diehl, F. Capasso, M. Troccoli, G. Hofler, D. Bour, S. Corzine, R. Maulini, M. Giovannini, and J. Faist, “Widely tunable mode-hop free external cavity quantum cascade lasers for high resolution spectroscopy and chemical sensing,” Appl. Phys. B 92(3), 305–311 (2008). [CrossRef]
7. B. Schwarz, P. Reininger, H. Detz, T. Zederbauer, A. M. Andrews, S. Kalchmair, W. Schrenk, O. Baumgartner, H. Kosina, and G. Strasser, “A bi-functional quantum cascade device for same frequency lasing and detection,” Appl. Phys. Lett. 101(19), 191109 (2012). [CrossRef]
8. R. Maulini, A. Lyakh, A. Tsekoun, R. Go, C. Pflügl, L. Diehl, F. Capasso, C. Kumar, and N. Patel, “High power thermoelectrically cooled and uncooled quantum cascade lasers with optimized reflectivity facet coatings,” Appl. Phys. Lett. 95(15), 151112 (2009). [CrossRef]
9. Y. Bai, S. R. Darvish, N. Bandyopadhyay, S. Slivken, and M. Razeghi, “Optimizing facet coating of quantum cascade lasers for low power consumption,” J. Appl. Phys. 109(5), 053103 (2011). [CrossRef]
10. C. E. Zah, J. S. Osinski, C. Caneau, S. G. Menocal, L. A. Reith, J. Salzman, F. K. Shokoohi, and T. P. Lee, “Fabrication and performance of 1.5µm GaInAsP travelling-wave laser amplifiers with angled facets,” Electron. Lett. 23(19), 990 (1987). [CrossRef]
11. M. Troccoli, C. Gmachl, F. Capasso, D. L. Sivco, and A. Y. Cho, “Mid-infrared (λ≈7.4 µm) quantum cascade laser amplifier for high power single-mode emission and improved beam quality,” Appl. Phys. Lett. 80(22), 4103 (2002). [CrossRef]
12. E. A. Zibik, W. H. Ng, D. G. Revin, L. R. Wilson, J. W. Cockburn, K. M. Groom, and M. Hopkinson, “Broadband 6 µm < λ < 8 µm superluminescent quantum cascade light-emitting diodes,” Appl. Phys. Lett. 88(12), 121109 (2006). [CrossRef]
13. Y. Bai, S. Slivken, Q. Y. Lu, N. Bandyopadhyay, and M. Razeghi, “Angled cavity broad area quantum cascade lasers,” Appl. Phys. Lett. 101(8), 081106 (2012). [CrossRef]
14. C. F. Lin, “Superluminescent diodes with angled facet etched by chemically assisted ion beam etching,” Electron. Lett. 27(11), 968 (1991). [CrossRef]
15. A. O. Dirisu, G. Silva, Z. Liu, C. F. Gmachl, F. J. Towner, J. Bruno, and D. L. Sivco, “Reduction of facet reflectivity of quantum-cascade lasers with subwavelength grating,” IEEE Photon. Technol. Lett. 19(4), 221–223 (2007). [CrossRef]
16. N. Yu, R. Blanchard, J. Fan, F. Capasso, T. Edamura, M. Yamanishi, and H. Kan, “Small divergence edge-emitting semiconductor lasers with two-dimensional plasmonic collimators,” Appl. Phys. Lett. 93(18), 181101 (2008). [CrossRef]
17. M. Ettenberg, H. S. Sommers, H. Kressel, and H. F. Lockwood, “Control of facet damage in GaAs laser diodes,” Appl. Phys. Lett. 18(12), 571 (1971). [CrossRef]
18. K. Petermann, Laser diode modulation and noise, (KTH Scientific Publishers, Dordrecht, 1991), Chap. 2.4. “Lasing characteristic of Fabry-Pérot Type Laser”.
19. Z. Liu, D. Wasserman, S. S. Howard, A. J. Hoffman, C. F. Gmachl, X. Wang, T. Tanbun-Ek, L. Cheng, and F. S. Choa, “Room-temperature continuous-wave quantum cascade lasers grown by MOCVD without lateral regrowth,” IEEE Photon. Technol. Lett. 18(12), 1347–1349 (2006). [CrossRef]
20. E. Mujagić, M. Nobile, H. Detz, W. Schrenk, J. Chen, C. Gmachl, and G. Strasser, “Ring cavity induced threshold reduction in single-mode surface emitting quantum cascade lasers,” Appl. Phys. Lett. 96(3), 031111 (2010). [CrossRef]
21. Z. Y. Zhang, I. J. Luxmoore, C. Y. Jin, H. Y. Liu, Q. Jiang, K. M. Groom, D. T. Childs, M. Hopkinson, A. G. Cullis, and R. A. Hogg, “Effect of facet angle on effective facet reflectivity and operating characteristics of quantum dot edge emitting lasers and superluminescent light-emitting diodes,” Appl. Phys. Lett. 91(8), 081112 (2007). [CrossRef]
