The angled-grating broad-area laser is a promising candidate for high power, high brightness diode laser source. The key point in the design is the angled gratings which can simultaneously support the unique snake-like zigzag lasing mode and eliminate the direct Fabry-Perot (FP) feedback. Unlike a conventional laser waveguide mode, the phase front of the zigzag mode periodically changes along the propagation direction. By use of the mirror symmetry of the zigzag mode, we propose and demonstrate the folded cavity angled-grating broad-area lasers. One benefit of this design is to reduce the required wafer space compared to a regular angled-grating broad-area laser, especially in a long cavity laser for high power operation. Experimental results show that the folded cavity laser exhibits good beam quality in far field with a slightly larger threshold and smaller slope efficiency due to the additional interface loss.
© 2013 OSA
High power and high brightness operation of semiconductor lasers requires a large emitting aperture to avoid catastrophic optical mirror damage (COMD) and effective transverse mode control mechanism to maintain a single lobe (near) diffraction-limited far field. To meet these requirements, the angled-grating broad-area laser has been proposed and demonstrated to deliver over 1W power with diffraction-limited beam quality [1–4]. In the design, the lowest loss Bragg mode confined by the transverse gratings is favored as the lasing mode [5–7]. The angled grating is the key point to support the snake-like lasing mode and to eliminate the direct FP feedback between two end facets. In a conventional waveguide laser, the phase front of the guided mode does not change along the propagation direction. But for angled-grating broad-area lasers, the phase front exhibits a zigzag pattern with a large angle (equal to the grating tilt angle) with respect to the propagation direction, which provides an effective method to fold the cavity by use of mirror symmetry. In this paper, we propose and experimentally demonstrate a folded cavity design for angled-grating broad-area lasers. The folded design can be also interpreted as two cascaded angled-grating broad-area lasers with opposite tilting directions. We show that with a careful selection of the cavity length, a low loss lasing mode can be obtained. Compared to the regular angled-grating broad-area laser, we experimentally demonstrate that the folded cavity laser provides similar beam quality, but with a slightly higher threshold and smaller slope efficiency due to the additional loss at the interface.
One instant benefit of the new design is to occupy less wafer space than the regular angled-grating broad-area laser. In a regular angled-grating broad-area laser, the tilt angle is usually from 10° to 20° and the length of cavity can be as long as 4mm. As a result, angled-grating broad-area lasers can use 10 times wafer space more than conventional straight cavity broad-area lasers. The proposed design only takes 50% wafer space of the regular angled-grating broad-area laser with the same cavity length. And in principle, multiple foldings can be used to further reduce the required wafer space for a long cavity laser.
2. Laser design and fabrication
Figure 1 shows the schematic plot of a folded cavity angled-grating broad-area laser. The laser consists of two cascaded angled-grating broad-area lasers that tilt to opposite directions. The waveguide mode in each half is the same as that in an unfolded angled-grating cavity as shown in Fig. 2(a). The snake-like zigzag mode consists of two planewave-like components in resonance with the gratings, denoted as R1 and R2 in Fig. 2(b). The wavevectors of R1, R2 and the gratings satisfy the phase matching condition: k⃗R1 + k⃗G = k⃗R2 as shown in the inset of Fig. 2(b). The angle between k⃗R1(k⃗R2) and the grating axis is equal to θ, the grating tilt angle. A detailed modal gain analysis of the grating confined modes can be found in  and the mode theory of an unfolded angled-grating broad-area laser is well explained in . At the two end facets in Fig. 2(c), the propagation direction of R1(R′2) component is perpendicular to the facet and that of R2(R′1) is titled by 2θ. When reflected, R1(R′2) component will be efficiently fed back to the cavity and R2(R′1) component will be lost. Therefore, at the facet, the preferred cavity mode will have the maximum perpendicular component (R1, R′2 in this example) and minimum tilted component (R2, R′1 in this example). At the interface between two halves, the cavity mode with the maximum transmissivity, T, from one half to another will be preferred. According to [7, 9], the transmissivity T can be expressed as:Fig. 2(c). The maximum transmissivity is obtained when θB = 0. This means that the maximum transmission happens between two perpendicular components on both sides (R1, R′2 in this example). This is also indicated by the k-vector relationship in Fig. 2(c). It shows that the transmission from R2 to R′2/R′1 and from R1 to R′1 suffers high loss due to the large wavevector mismatch. Since the phase front and mode profile of R1 and R′2 are identical, it allows for efficient mode coupling. As a result, the length of one half (L/2) should be integer times of 2Lc, where Lc is the coupling length of R1 and R2, defined as the length during which the power of one component is fully coupled into the other component. The coupling length can be calculated through Lc = π/|β1 − β2|, where β1 and β2 are the propagation constants of the first two lowest modal loss Bragg modes confined by the gratings. If this condition is not satisfied, the mode will suffer high loss at the interface. Figure 2(d) and (e) show the mode profiles in a symmetrically folded angled-grating cavity when L = 4NLc and L = (4N + 2)Lc, respectively, where N is an integer. When the correct cavity length condition is satisfied in Fig. 2(d), we can obtain a low loss folded cavity mode with efficient mode coupling at the middle interface. But strong diffraction loss is observed in Fig. 2(e) when this condition is not satisfied. There is another constraint condition on the cavity length L. To suppress the direct FP feedback between two end facets, we have to eliminate that on both halves, which means L must satisfy tan(θ)(L/2) > W. All the mode profiles are obtained by the FDTD method.
The folded cavity angled-grating broad-area lasers are fabricated in an InP-based multiple quantum well (MQW) epitaxy wafer. The details of the epitaxy wafer is described in . The tilting angle θ is set to be 10° and accordingly, the grating period is calculated to be 1.368μm. The etch depth of 900nm is chosen to obtain a grating coupling coefficient around 0.1/μm and 100 periods of gratings are etched, resulting in a total width around 140μm. The grating coupling length is calculated to be about 150μm which matches well with our previous experiments on the conventional angled-grating broad-area lasers. Accordingly, the final length of the folded cavity is selected to be around 1.8mm which is about 12 times of the calculated coupling length. The fabrication process consists of a series of steps of lithography, etching, planarization, metallization and ion-implantation. The detailed fabrication process is described in . Right after the p-contact deposition, the laser chips are ion-implanted in the area outside the metal contact to diminish the lateral current leakage. The effect of ion-implantation on angled-grating broad-area laser is described in [11–13]. At last, the chip is cleaved to the length of about 1.8mm and the laser diodes are bonded and wired on a c-mount for measurement. Figure 3 shows the scanning electron microscope pictures of the devices.
3. Measurement results and discussion
We measure the folded and unfolded cavity angled-grating broad-area laser with the same cavity length for comparison. The two types of devices are fabricated together with the same fabrication processes and parameters. All the measurements are carried out with a CW current source in a cryostat with the temperature set at 250K. Figure 4(a) shows the near field profiles of the folded cavity and unfolded cavity angled-grating broad-area laser. The widths of aperture are 155.3μm and 156.1μm for folded and unfolded cavity, respectively. The width should be similar because both cavities have the same grating design. The far field profiles and camera images are shown in Fig. 4(b). The divergence angles are 0.96° and 0.78° for the folded cavity and unfolded cavity, respectively. We also include the theoretical far field calculated from a planewave aperture with the same width as what we measured. The simulated result is shown in green dash-dotted line in Fig. 4(b). The divergence angle of the folded cavity is a little bit larger than that of the unfolded cavity due to the existence of the middle interface. The angled gratings can be considered as a waveguide where the desired Bragg mode is filtered. The folded design needs longer cavity or stronger coupling to eliminate the undesired components induced by the interface. This means with the same etching depth and cavity length as in our situation, the folded cavity will have a slightly larger divergence angle along the slow axis. However, the difference is small in our experiments and both devices can be considered as near diffraction-limited. Figure 4(c) shows the light-current (LI) curves of the folded and unfolded cavity. For the unfolded cavity, the threshold and slope efficiency are 365.5mA and 0.21W/A, respectively. Compared to the unfolded cavity, the folded cavity has a larger threshold of 436.4mA and smaller slope efficiency of 0.14W/A. The differences in LI curves are due to the extra optical loss induced by the interface in the folded cavity. The relatively strong spontaneous emission before lasing is due to the scattering loss induced by the deeply etched gratings. We believe that the output power is mainly limited by thermal management and the epitaxy wafer which is not designed or optimized for high power applications. The interface may also be the reason for the different lasing wavelength in the two devices as shown in Fig. 4(d). The lasing wavelengths for the folded and unfolded cavity are 1542.7nm and 1521.8nm, respectively. It is possible that extra heat is generated because of the extra loss induced by the interface in the folded cavity. This reduces the efficiency and the wavelength is red-shifted due to the heat. Since the gratings are only resonant with the transverse wave vector, multiple longitudinal modes are allowed in the angled-grating broad-area lasers . The existence of the interface also changes the longitudinal mode resonance condition, resulting in the small peak at 1550nm in the spectrum of the folded cavity laser.
In conclusion, the folded cavity angled-grating broad-area laser is proposed and demonstrated. Compared to the conventional angled-grating broad-area laser, the folded design takes advantage of the mirror symmetry of the zigzag mode and can save 50% wafer space. The length of cavity should be correctly chosen to reduce the interface loss and avoid direct FP feedback between the facets. The measurement results show that the folded design can provide near diffraction-limited beam quality with a slightly higher threshold and smaller slope efficiency.
The authors acknowledge funding support from a DARPA Young Faculty Award ( N66001–10–1–4038), ARO Young Investigator Award ( W911NF–11–1–0519), and DURIP Award ( W911NF–11–1–0312). The authors also acknowledge the use of the Gatech Nanotechnology Research Center Facility and associated support services in the completion of this work.
References and links
1. S. D. Demars, K. M. Dzurko, R. J. Lang, D. Welch, D. R. Scifres, and A. Hardy, “Angled-grating distributed feedback laser with 1 W cw single-mode diffraction-limited output at 980nm,” in “Lasers and Electro-Optics, 1996. CLEO ’96., Summaries of papers presented at the Conference on,” (1996), 77–78.
2. V. V. DWong, S. D. DeMars, A. Schoenfelder, and R. J. Lang, “Angled-grating distributed-feedback laser with 1.2 W cw single-mode diffraction-limited output at 10.6μm,” in “In Laser and Electro-Optics, 1998. CLEO ’98., Summaries of papers presented at the Conference on,” (1998), 34–35.
3. K. Paschke, R. Guther, J. Fricke, F. Bugge, G. Erbert, and G. Trankle, “High power and high spectral brightness in 1060 nm alpha-dfb lasers with long resonators,” Electron. Lett. 39, 369–370 (2003). [CrossRef]
4. R. E. Bartolo, W. W. Bewley, I. Vurgaftman, C. L. Felix, J. R. Meyer, and M. J. Yang, “Mid-infrared angled-grating distributed feedback laser,” Appl. Phys. Lett. 76, 3164–3166 (2000). [CrossRef]
5. R. J. Lang, K. Dzurko, A. A. Hardy, S. Demars, A. Schoenfelder, and D. F. Welch, “Theory of Grating-Confined Broad-Area Lasers,” IEEE J. Quantum Electron. 34, 2196–2210 (1998). [CrossRef]
6. R. Guther, “Beam propagation in an active planar waveguide with an angled bragg grating (α laser),” J. Mod. Optic. 45, 1537–1546 (1998). [CrossRef]
7. A. Sarangan, M. Wright, J. Marciante, and D. Bossert, “Spectral properties of angled-grating high-power semiconductor lasers,” IEEE J. Quantum Electron. 35, 1220–1230 (1999). [CrossRef]
8. L. Zhu, A. Scherer, and A. Yariv, “Modal Gain Analysis of Transverse Bragg Resonance Waveguide Lasers With and Without Transverse Defects,” IEEE J. Quantum Electron. 43, 934–940 (2007). [CrossRef]
9. D. Marcuse, “Reflection loss of laser mode from tilted end mirror,” J. Lightwave. Technol. 7, 336–339 (1989). [CrossRef]
11. K. Paschke, A. Bogatov, F. Bugge, A. E. Drakin, J. Fricke, R. Güther, A. A. Stratonnikov, H. Wenzel, G. Erbert, and G. Tränkle, “Properties of ion-implanted high-power angled-grating distributed-feedback lasers,” IEEE J. Sel. Top. Quantum Electron 9, 1172–1178 (2003). [CrossRef]
12. Y. Zhao and L. Zhu, “Improved beam quality of coherently combined angled-grating broad-area lasers,” Photonics Journal, IEEE 5, 1500307–1500307 (2013). [CrossRef]
13. S. J. Pearton, “Ion implantation for isolation of III–V semiconductors,” Mater. Sci. Rep. 4, 313–363 (1990). [CrossRef]