We report tunable VCSELs emitting around 1060 nm, enabled by high-contrast grating (HCG) mirror. Single-mode continuous-wave (CW) operation up to 110 °C is demonstrated, with room-temperature single-mode output power >1.3 mW at a very low threshold of ~300 µA. The obtained thermal resistance of 0.88 °C/mW is low for VCSELs with an oxide-confined laser aperture. A wide, continuous tuning range up to 40 nm was achieved with electrostatic and thermal tuning, at a fast tuning speed up to 1.15 MHz. In addition, we developed transverse-mode control designs of HCGs to greatly improve the single-mode yield of oxidized VCSELs. The cost-effective, wafer-scale fabrication makes these VCSELs promising as tunable light sources for swept-source optical coherent tomography (SS-OCT) and LiDAR applications.
© 2017 Optical Society of America
Vertical-cavity surface-emitting lasers (VCSELs) are key light sources in a variety of applications, including but not limited to data communications, optical interconnects, high-resolution displays, as well as laser sensing and printing . Compared with conventional edge-emitting lasers, the nature of the vertical optical cavity endues VCSELs with numerous advantages. First of all, the relatively short vertical cavity makes it less challenging to achieve single-mode laser emission. The high quality beam emitted from top surface can easily be coupled into fibers and other optical elements. Secondly, VCSELs are fabricated at a wafer scale, facilitating high-yield and low-cost mass production. Furthermore, the small footprint of VCSELs leads to low power consumption, thus high energy efficiency .
Wavelength tunable lasers are critical enabling key components for wavelength-division multiplexing (WDM) optical communication [3,4]. Tunable VCSELs using microelectromechanical structures (MEMS) possess one most desirable attribute, i.e. continuous, mode-hop-free and deterministic tuning characteristics. This characteristic is not only important for low-cost dense WDM (DWDM) systems, but also presents the only solution for continuously, widely tunable sources. The recent emergence of swept-source optical coherent tomography (SS-OCT) [5–8] as well as light detection and ranging (LiDAR) [9,10] makes MEMS-VCSEL attractive for sensing applications where most conventional tunable edge-emitting lasers cannot be used due to their lack of continuous tuning.
The wavelength window around 1060nm (typically 1000-1100 nm) has been recognized as the most suitable for ophthalmic OCT (optical coherence tomography) imaging because it balances water absorption and scattering effects in the eye thus enables visualization of the choroidal vascular structure . Although MEMS tunable VCSELs at other wavelengths (i.e. 850nm, 980nm, 1550 nm) have been extensively developed [12–14], the first electrically-pumped MEMS-VCSEL around 1060 nm was not demonstrated until 2013 . The work adopted an HCG as the top mirror and anti-reflection coating design in the cavity. Unfortunately, it showed a somewhat high threshold current of ~1 mA and a limited static tuning range of around 16 nm. Another 1050-nm MEMS-VCSEL was reported recently with excellent tuning performance ranging up to ~58.8 nm statically. However, it showed limited output power and tuning speed , as well as compromised thermal performance constrained by the dielectric DBRs. In addition, it also poses high requirements for epitaxial growth and fabrication processes, which may hinder mass production at large scales with low cost.
The simple-structured high-contrast grating (HCG) with a subwavelength period suspended by two MEMS beams works as a broadband high-reflectivity mirror and a thin lightweight MEMS actuator . Tunable HCG-VCSELs at 850 nm and 1550 nm have been previously demonstrated by our group [18,19]. Here we report, for the first time, an electrically pumped MEMS VCSEL lasing in the 1060-nm wavelength window showing simultaneously a wide tuning range and a fast tuning speed, which can enable real-time display of 3D image with high resolution. With the monolithic HCG as the tunable mirror and wet oxidation to provide current confinement, it leads to a wafer-scale fabrication process at low cost for mass production and commercialization .
Design of the HCG-VCSEL structure here is depicted in Fig. 1(a). The as-grown 1060-nm epitaxial wafer consists of ~30-40 pairs of DBRs as the bottom mirror. Multiple strained InGaAs quantum wells provide the optical gain. The optical and electrical confinement is realized by oxidizing an layer, forming a small laser aperture. The suspended top HCG mirror offers high reflectivity benefitting from its high index-contrast with the surrounding air and its subwavelength dimensions. This structure is formed of n-p-n junctions, with the p-n junction forward biased to achieve lasing. By applying a voltage to the tuning contact to reverse bias the n-p junction, the HCG MEMS structure can be actuated electrostatically thus tuning the wavelength of the laser.
The general design of a VCSEL epitaxial wafer involves two important aspects: the optical design to maximize the overlap of the standing waves with the active region, and electrical design to provide efficient current flow into QWs and reduce the parasitic resistance. Optimization of the composition and thickness of each layer is performed during the optical cavity design. The goal is to achieve the largest confinement factor or overlap between the standing waves and the QWs in the active region. This is the key to achieving low threshold and high output power for VCSELs. The confinement factor is calculated with the transfer matrix method . A value of ~1.3% is achieved here, which is limited by a rather large penetration depth into the bottom DBRs. Through replacing the bottom DBRs by material pairs with a larger refractive index contrast, the confinement factor can be improved. To fulfill the round-trip condition of the VCSEL cavity, the thickness of sacrificial layer needs to be designed carefully to match the reflection phase carried by the selected HCG design.
Using the analytical tool from , the reflectivity contour map was simulated to facilitate the HCG design. In Fig. 1(b), the reflectivity distribution is plotted versus the HCG thickness () and wavelength () (both normalized by HCG period ) for TM incidence (light polarized perpendicularly to the grating bars). A thickness of ~300 nm is chosen for = 1060 nm with good fabrication tolerance. We simulated the reflectivity versus HCG airgap (a) and period () at this wavelength and thickness. A range of dimensions that can be used to define the HCG are then obtained, with period ~480-505 nm and airgap a~120-200 nm, as shown in Fig. 1(c).
The scanning electron microscopy (SEM) image in Fig. 2(a) shows the top-view of a finished HCG-VCSEL device. The zoomed-in image Fig. 2(b) highlights the HCG MEMS mirror fully suspended and surrounded by the air. Being supported by the MEMS arms on both sides, the whole HCG mirror can be actuated with electrostatic tuning to realize wavelength shift. Figure 2(c) displays an HCG array imaged by 3D confocal microscope, demonstrating the wafer-scale mass production.
3.2 LIV and spectra
The TM HCG-VCSEL maintains single transverse-mode lasing under CW electrical operation. A Keithley current source is used to bias the VCSEL using electrical probing. Figure 3(a) shows the light-current-voltage (LIV) characteristics of a typical device measured at 20°C, providing a low threshold current of ~300 and output power of 1.3 mW at a pump current of 4 mA. A camera captured the near-field emission of the device through a 100X objective. While showing a small spot of spontaneous emission below the lasing threshold (), strong speckle patterns were observed above the lasing threshold, which is the classic signature of a lasing resonance. The laser has a voltage of ~8 V at 4 mA, indicating a series resistance of around 2 . The series resistance can be improved by annealing the metal contacts and optimizing the doping levels in the current spreading layer. The wall plug efficiency is about 4%, which is mainly limited by the relatively high voltage and series resistance.
Various temperatures are applied to the copper heat sink directly in contact with the VCSEL chip. At room temperature, the devices lase at ~1080 nm instead of ~1060 nm. This is caused by epitaxial growth imperfections that can be corrected by more calibrations of the layer thicknesses during mass production, which is beyond the scope of this paper. Single-mode lasing was sustained beyond 110 °C heat sink temperature. Even at 110 °C, the output power is ~0.2 mW. Figure 3(b) illustrates the thermal behavior of the LIV characteristics as a function of heat sink temperature. While the output power is reduced with the increased temperature, the threshold current does not increase much. The VCSELs exhibit a thermal rollover with increasing current bias due to gain spectrum red-shifting more rapidly than the resonant cavity spectrum, an effect typically seen in VCSELs .
The thermal resistance is critical for optimizing the device efficiency, which is defined by the following ratio,Fig. 3(c), a fitted ratio of 0.061 nm/°C can be extracted. This number is determined by index change versus temperature despite of type of laser. The wavelength shift versus dissipated thermal power was then characterized by increasing the injection current at a fixed temperature of 20°C. Multiplying the bias current with the corresponding voltage and subtracting the optical output power, the dissipated thermal power can be calculated. The fitted ratio results in ~0.054 nm/mW. Taking the two numbers to obtain the ratio, a thermal resistance of around 0.88 °C/mW is obtained. Compared with similar oxidized VCSELs, the thermal resistance is smaller [24, 25]. This shows great promise for the thermal design in our epitaxial wafer.
3.3 Tuning characteristics
Wavelength tuning of the HCG-VCSEL is realized through electrical actuation of the HCG MEMS structure. The HCG structure and the semiconductor layer beneath it form a parallel capacitive transducer. By biasing the isolating junction between the tuning contact and laser contact, HCG MEMS moves downward, thus reducing the optical cavity length. Continuous wavelength tuning of ~20 nm towards the shorter wavelength is firstly obtained with 0 to 14.26 V of external applied voltage, shown in Fig. 4(a). At 1064 nm, under an injection current of 2 mA, the laser stops lasing due to gain and/or reflectivity falling below the necessary level. Above 15.57 V at 2 mA, the device starts lasing again with a higher order longitudinal mode around 1098 nm, and continuously tunes for another 14 nm over the applied voltage range of 15.57-17.8 V. The total tuning range through pure mechanical MEMS actuation is 34 nm. Similar mode-switching during tuning was also observed in [18, 26–28]. It provides invaluable information about the free spectral range (FSR) of the VCSEL which directly affects the tuning range. With a large FSR, the mode-switching can easily be corrected by changing only the tuning gap size, so a full continuous single longitudinal-mode tuning range can be achieved. If we increase the injection current at the same time as tuning the voltage, tuning due to thermal heating can push the longer wavelength limit to 1104 nm. Thus, single-mode continuous tuning accounting for both mechanical and thermal approaches reaches 40 nm. In addition, the tuning voltage used here is relatively small compared with other MEMS structures in VCSEL [16, 29], due to the light weight of our HCG structure. This allows for easy integration with low-voltage CMOS electronics for energy efficiency.
In order to investigate the limiting factor of the tuning range, the rigorous coupled-wave analysis (RCWA) method  and the transfer matrix method are used together to simulate the reflectivity of the mirrors (including both the compound HCG mirror and bottom DBR mirror) versus tuning airgap and wavelength, as shown in Fig. 4(b). The resonance wavelength of the corresponding cavity at each tuning step is also simulated and overlaid on top of the reflectivity contour. This simulation correlates with the measurement data in Fig. 4(a), and a theoretical tuning range of 47 nm is predicted for this HCG-VCSEL structure. This physical limit is determined by the free spectral range (FSR) between two longitudinal modes, ~50 nm from the simulation. Therefore, the wavelength tuning range could be further increased by improving the designs of the cavity structure for a wider FSR. More simulation details can be found in .
When applying voltage to the tuning contact, the electrostatic force will reach an equilibrium balance with the elastic force from the MEMS spring. This will result in a new position of the HCG MEMS structure which is closer to the VCSEL body than the original position. This relation can be expressed by the following equation,19]. The wavelength versus tuning voltage for the first longitudinal mode measured in Fig. 4(a) is plotted in Fig. 4(c), and the fitted relation between and (red curve) is overlaid on top of the experimental data (black dots), which is generated utilizing Eq. (2) and the calculation in Fig. 4(b). The well-matched trend proves that the HCG MEMS structure here indeed behaves as a micro-scale parallel capacitive transducer. The tuning efficiency (wavelength shift over tuning gap shift) is approximately 0.06 nm/nm, which is mainly constrained by the large penetration depth into the DBRs. A larger tuning efficiency can be achieved by replacing the 2 pairs of top DBRs with an anti-reflection layer of equivalent optical thickness .
The wavelength tuning speed of the VCSEL is mainly determined by the resonance frequency of the MEMS structure. The frequency response of the mechanical tuning was measured and fitted with a harmonic oscillator model [Fig. 4(d)], yielding a 3-dB bandwidth of 1.15 MHz with a damped resonance frequency ~600 kHz. The Q-factor for the HCG MEMS is fitted to be around 1.5, mainly constrained by air damping and the small spring constant. Using COMSOL, a finite element method (FEM) tool, the mechanical resonance is simulated to be ~540 kHz, close to the measured value. The spatial displacement of the fundamental eigenmode is shown in Fig. 4(e). The simulation parameters used for the HCG MEMS include Young’s modulus, Poisson’s ratio, and density. While ~1.15 MHz is already larger than the scanning rate requirement of most commercially available SS-OCT systems, it can be further improved. As is proportional to , where is the spring constant and is the mass of the MEMS structure, the ultralight weight HCG MEMS structure can potentially have 10 times faster tuning than its DBR counterpart. The spring constant here is intentionally designed to be smaller for wider tuning range, and can be tailored to suit specific applications. We have indeed demonstrated tuning speed >27 MHz with a smaller HCG size and optimized spring constant .
4. Transverse model control
As introduced above, the laser aperture definition for GaAs-based VCSELs generally utilizes the wet oxidation process which is highly selective to with high Al-concentration. Although this technique is relatively mature in the VCSEL fabrication industry with expensive in situ supervision systems, it is yet extremely challenging to control under a laboratory environment. Any fluctuation in the experimental conditions such as furnace temperature, vapor concentration, and gas flow, will greatly impact the yield and uniformity of VCSELs on a wafer scale. To obtain single transverse-mode VCSELs especially, an aperture of <5 is usually required, while oxidation rate increases as the aperture closes up . Another degree of freedom to control transverse mode of VCSEL is therefore desired.
Previous studies have shown the effects of HCG on the transverse mode of a cavity, such as its size , its dispersion [35, 36], and the mode spatial profile [37, 38]. Here we report another approach, utilizing the dependence of HCG reflectivity on the light incident angles (related to and ) . The schematics of an HCG receiving light at an oblique incidence angle are depicted in Fig. 5(a). Figure 5(b) illustrates the reflectivity contour (for R>99.5%) for HCG airgap versus incidence angle, under a fixed HCG period (~505 nm). Two designs are selected with distinct angle-dependent behavior. Design I with a~140 nm shows high reflectivity up to , while the reflectivity of design II with a~105 nm is drastically reduced beyond . Thus for design II, higher-order transverse modes emitting at a higher angle will be suppressed due to the reflectivity falling below threshold.
The two designs were fabricated into HCG-VCSEL devices with similar oxidation apertures. Figures 5(c) and 5(d) show a comparison of the results, with (c) for design I (~505 nm, a~140 nm) and (d) for design II (~505 nm, a~105 nm). An IR camera was used to examine the aperture size right after wet oxidation, resulting in a square aperture with side length of around (201) for both devices. We then measured the laser spectra of the devices with CW operation under a series of pump powers. As shown in Fig. 5(a-iii) and (b-iii), device with HCG design I delivers multiple transverse modes for all pump levels, not surprisingly. However, device with HCG design II demonstrates a single transverse mode lasing, despite the large oxidation aperture. Various transverse-mode-control designs were carried out on multiple HCG-VCSEL devices, which all showed similar effects. Therefore, we have validated that the angular dependence of HCG can be used to efficiently control transverse modes of VCSEL and improve the yield of single-mode lasers. Similar effects were also shown in . This discovery is also very meaningful for pattern design of VCSEL emission in the future.
To summarize, we have demonstrated a widely and fast tunable VCSEL emitting around 1060 nm, enabled by the HCG as an integrated tuning mirror. The HCG VCSELs show single mode emission under CW electrical operation. The output power can go beyond 1.3 mW with a threshold current of ~300 A. The laser provides excellent thermal characteristics, and it can lase with heat sink temperatures above 110 . Utilizing the temperature-dependent and power-dependent spectra, the thermal resistance is calculated to be 0.88 /mW, outperforming other VCSEL structures with even larger laser apertures. The wavelength can be tuned through electrostatic actuation of the HCG MEMS structure, which behaves like a capacitive transducer plate. A total tuning range of 40 nm is obtained, including 34-nm of MEMS tuning and 6-nm of thermal tuning. A theoretical tuning limit of 50-nm results from the simulation, which can be improved by designing the epitaxial layer structure to obtain wider FSR. Although the MEMS structure is optimized for wider tuning range rather than fast tuning speed, a 3-dB bandwidth of 1.15 MHz can still be obtained from the MEMS mechanical response measurement, which is sufficient for SS-OCT application. In addition, transverse-mode control is realized by utilizing the angular-dependence of HCG reflectivity. This saves VCSELs from the traditionally challenging wet oxidation process, and improves the yield of single-mode lasers even on top of large laser apertures. Comparing with the other available 1060-nm tunable VCSELs, our work here demonstrates the combination of wide tuning range, fast tuning speed, and most importantly the ease of fabrication with high yield at low cost. The devices reported here are promising for commercialization in applications such as SS-OCT, LiDAR, and short-link optical communication.
1. R. Szweda, “VCSEL applications diversify as technology matures,” III–Vs Review 19(1), 34–38 (2006). [CrossRef]
2. C. J. Chang-Hasnain, J. P. Harbison, C.-E. Zah, M. W. Maeda, L. T. Florez, N. G. Stoffel, and T.-P. Lee, “Multiple wavelength tunable surface-emitting laser arrays,” IEEE J. Quantum Electron. 27(6), 1368–1376 (1991). [CrossRef]
3. C. J. Chang-Hasnain, M. W. Maeda, N. G. Stoffel, J. P. Harbison, L. T. Florez, and J. Jewell, “Surface emitting laser arrays with uniformly separated wavelengths,” Electron. Lett. 26(13), 940–942 (1990). [CrossRef]
4. C. J. Chang-Hasnain, “Tunable VCSEL,” IEEE J. Sel. Top. Quantum Electron. 6(6), 978–987 (2000). [CrossRef]
5. D. Huang, E. A. Swanson, C. P. Lin, J. S. Schuman, W. G. Stinson, W. Chang, M. R. Hee, T. Flotte, K. Gregory, C. A. Puliafito, and et, “Optical coherence tomography,” Science 254(5035), 1178–1181 (1991). [CrossRef] [PubMed]
7. B. Potsaid, V. Jayaraman, J. Y. Jiang, P. J. S. Heim, I. Grulkowski, J. G. Fujimoto, and A. E. Cable, “1065nm and 1310nm MEMS tunable VCSEL light source technology for OCT imaging,” in SPIE Biomedical Optics & Medical Imaging (2012), paper 10.1117.
8. I. Grulkowski, J. J. Liu, B. Potsaid, V. Jayaraman, C. D. Lu, J. Jiang, A. E. Cable, J. S. Duker, and J. G. Fujimoto, “Retinal, anterior segment and full eye imaging using ultrahigh speed swept source OCT with vertical-cavity surface emitting lasers,” Biomed. Opt. Express 3(11), 2733–2751 (2012). [CrossRef] [PubMed]
9. P. F. McManamon, “Review of ladar: a historic, yet emerging, sensor technology with rich phenomenology,” Opt. Eng. 51(6), 060901 (2012). [CrossRef]
10. B. Behroozpour, N. Quack, P. Sandborn, S. Gerke, W. Yang, C. Chang-Hasnain, M. C. Wu, and B. E. Boser, “Method for increasing the operating distance of MEMS LIDAR beyond Brownian noise limitation,” in Conference on Lasers and Electro-Optics (IEEE 2014), paper AW3H.2. [CrossRef]
11. B. Povazay, B. Hermann, V. Kajic, B. Hofer, and W. Drexler, “High speed, spectrometer based optical coherence tomography at 1050 nm for isotropic 3D OCT imaging and visulization of retinal and choroidal vasculature,” in Proc. Biomed. Opt. OCT Opthalmic Appl. (2008), pp. 2733–2751.
12. F. Koyama, “Advances and new functions of VCSEL photonics,” Opt. Rev. 21(6), 893–904 (2014). [CrossRef]
13. E. C. Vail, G. S. Li, W. Yuen, and C. J. Chang-Hasnain, “High performance and novel effects of micromechanical tunable vertical-cavity lasers,” IEEE J. Sel. Top. Quantum Electron. 3(2), 691–697 (1997). [CrossRef]
14. C. Gierl, T. Gruendl, P. Debernardi, K. Zogal, C. Grasse, H. A. Davani, G. Böhm, S. Jatta, F. Küppers, P. Meissner, and M.-C. Amann, “Surface micromachined tunable 1.55 μm-VCSEL with 102 nm continuous single-mode tuning,” Opt. Express 19(18), 17336–17343 (2011). [CrossRef] [PubMed]
15. T. Ansbaek, I.-S. Chung, E. S. Semenova, and K. Yvind, “1060-nm tunable monolithic high index contrast subwavelength grating VCSEL,” IEEE Photonics Technol. Lett. 24, 455–457 (2013).
16. D. D. John, C. B. Burgner, B. Potsaid, M. E. Robertson, B. K. Lee, W. J. Choi, A. E. Cable, J. G. Fujimoto, and V. Jayaraman, “Wideband electrically pumped 1050-nm MEMS-tunable VCSEL for ophthalmic imaging,” J. Lightwave Technol. 33(16), 3461–3468 (2015). [CrossRef] [PubMed]
17. C. J. Chang-Hasnain and W. Yang, “High-contrast gratings for integrated optoelectronics,” Adv. Opt. Photonics 4(3), 379–440 (2012). [CrossRef]
18. M. C. Y. Huang, Y. Zhou, and C. J. Chang-Hasnain, “A nanoelectromechanical tunable laser,” Nat. Photonics 2(3), 180–184 (2008). [CrossRef]
19. Y. Rao, W. Yang, C. Chase, M. C. Y. Huang, D. P. Worland, S. Khaleghi, M. R. Chitgarha, M. Ziyadi, A. E. Willner, and C. J. Chang-Hasnain, “Long-wavelength VCSEL using high contrast grating,” IEEE J. Sel. Top. Quantum Electron. 19(4), 1701311 (2013). [CrossRef]
20. K. Li, C. Chase, Y. Rao, and C. J. Chang-Hasnain, “Widely tunable 1060-nm high-contrast grating VCSEL,” in Compound Semiconductor Week (CSW, IEEE2016), paper MoC4–2.
21. S. L. Chuang, Physics. of Photonic Devices (Wiley, 2009), p. 411.
22. W. Yang and C. J. Chang-Hasnain, “High contrast grating solver package,” University of California at Berkeley (2014), https://light.eecs.berkeley.edu/cch/hcgsolver.html.
23. M. W. Maeda, C. J. Chang-Hasnain, A. V. Lehmen, H. Izadpanah, C. Linda, M. Z. Iqbal, L. T. Florez, and J. P. Harbison, “Mluti-gigabit/s operation of 16-wavelength vertical cavity surface emitting laser array,” IEEE Photonics Technol. Lett. 3(10), 863–865 (1991). [CrossRef]
24. B. Weigl, M. Grabherr, C. Jung, R. Jager, G. Reiner, R. Michalzik, D. Sowada, and K. J. Ebeling, “High-performance oxide-confined GaAs VCSELs,” IEEE J. Sel. Top. Quantum Electron. 3(2), 409–415 (1997). [CrossRef]
25. K. Lascola, “Master degree dissertation: modeling of vertical cavity surface emitting lasers,” University of California at Berkeley (1997).
26. B. Kogel, P. Debernardi, P. Westbergh, J. S. Gustavsson, Å. Haglund, E. Haglund, J. Bengtsson, and A. Larsson, “Integrated MEMS-tunable VCSELs using a self-aligned reflow process,” IEEE J. Quantum Electron. 48(2), 144–152 (2012). [CrossRef]
27. M. Nakahama, H. Sano, S. Inoue, T. Sakaguchi, A. Matsutani, and F. Koyama, “Tuning characteristics of monolithic MEMS VCSELs with oxide anti-reflection layer,” IEEE Photonics Technol. Lett. 25(18), 1747–1750 (2013). [CrossRef]
28. C. Gierl, T. Gruendl, P. Debernardi, K. Zogal, C. Grasse, H. A. Davani, G. Böhm, S. Jatta, F. Küppers, P. Meissner, and M.-C. Amann, “Surface micromachined tunable 1.55 μm-VCSEL with 102 nm continuous single-mode tuning,” Opt. Express 19(18), 17336–17343 (2011). [CrossRef] [PubMed]
29. C. Lam, H. Liu, B. Koley, X. Zhao, V. Kamalov, and V. Gill, “Fiber optic communication technologies: What’s needed for datacenter network operations,” IEEE Commun. Mag. 48(7), 32–39 (2010). [CrossRef]
31. P. Qiao, G.-L. Su, Y. Rao, M. C. Wu, C. J. Chang-Hasnain, and S. L. Chuang, “Comprehensive model of 1550 nm MEMS-tunable high-contrast-grating VCSELs,” Opt. Express 22(7), 8541–8555 (2014). [CrossRef] [PubMed]
33. P. C. Ku and C. J. Chang-Hasnain, “Thermal oxidation of AlGaAs: modeling and process control,” IEEE J. Quantum Electron. 39(4), 577–585 (2003). [CrossRef]
36. A. Taghizadeh, J. Mork, and I. Chung, “Vertical-cavity in-plane heterostructures: physcis and applications,” Appl. Phys. Lett. 107(18), 181107 (2015). [CrossRef]
37. M. Gębski, O. Kuzior, M. Dems, M. Wasiak, Y. Y. Xie, Z. J. Xu, Q. J. Wang, D. H. Zhang, and T. Czyszanowski, “Transverse mode control in high-contrast grating VCSELs,” Opt. Express 22(17), 20954–20963 (2014). [CrossRef] [PubMed]
38. W. Yang, J. Ferrara, K. Grutter, A. Yeh, C. Chase, Y. Yue, A. E. Willner, M. C. Wu, and C. J. Chang-Hasnain, “Low loss hollow-core waveguide on a silicon substrate,” Nanophotonics 1(2012), 23–29 (2012).
39. K. Li, Y. Rao, C. Chase, W. Yang, and C. J. Chang-Hasnain, “Beam-Shaping Single-Mode VCSEL With A High-Contrast Grating Mirror,” in Conference on Lasers and Electro-Optics (IEEE 2016), paper SF1L. [CrossRef]
40. F. Koyama, “Engineering of angular dependence of high-contrast grating mirror for transverse mode control of VCSELs,” Proc. SPIE 8995, 89950H (2014). [CrossRef]