Open Access
Add to BibSonomy
Abstract
We have developed a membrane buried-heterostructure (BH) distributed feedback (DFB) laser consisting of an optically coupled III-V/Si waveguide and SiN surface grating. A 230-nm-thick membrane III-V layer enables us to construct an optical supermode in a 220-nm-thick Si waveguide and control the optical confinement factor in both the III-V and Si layers by changing Si waveguide width. This makes it possible to use a conventional Si photonics platform because the Si waveguides widely used on it are around 220-nm thick. To fabricate the BH—the key component for constructing a membrane laser with a lateral current-injection structure—we used direct wafer bonding and regrowth by metalorganic vapor phase epitaxy. Light output from the DFB laser is transferred to the Si waveguide through a short inverse-taper InP waveguide. A fiber-chip interface constructed by using inverse-taper Si waveguides and SiOx waveguides provides 2-dB fiber coupling loss. Fiber coupling power of 7.9 mW is obtained with a λ/4-shifted DFB laser with a 500-µm-long cavity. Single-mode lasing with a side-mode suppression ratio of 50 dB and lasing up to 120°C are also demonstrated.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
With the rapid growth of data traffic, the demand for high-capacity optical transmitters with reduced cost, size, and power consumption is increasing. Photonic integrated circuits (PICs) in which lasers and Mach–Zehnder modulators (MZMs) are integrated on a single chip will be essential components of transmitters in networks using advanced modulation formats. A Si photonics platform is an excellent candidate for the mass production of PICs because it enables us to fabricate photonic devices on large Si wafers with high density and high yield using mature Si complementary metal-oxide-semiconductor (CMOS) technology [1]. A variety of compact optical devices, such as filters [2,3], polarization rotators [4], spot-size converters (SSCs) [5], MZMs [6], and photodetectors (PDs) [5], have been developed with Si photonics technology on large Si wafers.
Since Si is an indirect-transition semiconductor, the main issue with Si photonics platforms has been the integration of lasers on Si substrates. Recently, III-V devices such as lasers and semiconductor optical amplifiers (SOAs) have been heterogeneously integrated on a Si substrate [7–14] by using direct wafer-bonding or die-bonding techniques. Tunable external-cavity lasers or narrow-linewidth distributed feedback (DFB) lasers consisting of an active III-V waveguide with a low-loss Si waveguide have also been developed [9,10]. Moreover, Si MZMs have been integrated with lasers through a Si waveguide [11]. All these results were obtained by using III-V layers with a vertical p-i-n junction, in which the total thickness of the III-V layers exceeded 2 µm to ensure a sufficient distance between the optical mode field and contact metals. To control the optical mode between the Si and III-V waveguides, the integrated Si waveguides needed to be as thick as 400–500 nm so that the waveguides would have comparable effective refractive indices [11–13].
However, such thick Si waveguides have a large effective refractive index mismatch with many Si photonic devices, in which Si waveguide thickness of around 220 nm is widely used to attain a small bending radius and efficient optical modulation in MZMs. Therefore, integrating conventional III-V devices using a vertical p-i-n junction on a Si photonics platform, involves an additional etching or growth of the Si layer to change its thickness [11,13]. However, these approaches increase the thickness variation and surface roughness of the Si layer, resulting in degradation of device performance due to increased propagation loss and lower filtering extinction ratios, and they increase process complexity and cost as well. For the above reasons, lasers suitable for integration on a Si photonics platform are strongly desired.
The thickness of the III-V layers has to be reduced to make the effective refractive index of the III-V waveguide comparable to that of a Si waveguide with thickness of around 220 nm. Membrane III-V photonic devices are quite suitable for this purpose, and we have already developed devices with III-V layer thickness of 350 nm or less, such as SOAs [15], lasers [16–20], and optical phase shifters [21,22]. In previous devices, owing to the large optical confinement in the III-V active region, we have demonstrated high-modulation efficiency for directly modulated lasers (DMLs) and phase shifters. In this work, we demonstrated another important feature of membrane photonic devices, namely integration with a 220-nm-thick Si waveguide. In this case, a DFB laser will be used as a bias light source for MZMs. Thus, the optical confinement factor in the III-V layer should be reduced and the laser cavity length should be increased to achieve relatively large output power compared with our previously developed DMLs, whose cavity lengths are typically less than 100 µm. Thus, we use an optical supermode in III-V/Si waveguides to reduce the optical confinement factor in the III-V layer and increase it in the Si layer, where the optical confinement factor can be easily controlled by changing the width of the optically coupled Si waveguide [23].
A fabricated λ/4 shifted DFB laser with a 500-µm-long cavity and an optically coupled III-V/Si waveguide exhibits single-mode lasing with a side-mode suppression ratio (SMSR) of 50 dB and lasing at up to 120°C. Light output from the DFB laser is transferred to the Si waveguide by using an inverse-taper InP waveguide. The Si layer thickness is 220 nm on the entire platform. In addition, a SiOx waveguide was integrated by using an inverse-taper Si waveguide, which provide a 2-dB fiber coupling loss, and the fiber coupling power of 7.9 mW is demonstrated.
2. Device structure and design
First, we describe the impact of III-V layer thickness on the integration with a 220-nm Si waveguide. Figure 1(a) shows the effective refractive indices of 220-nm and 500-nm-thick stripe Si waveguides versus Si core width. The effective refractive indices of a 2-µm-thick and 2-µm-wide III-V waveguide [16] and a 230-nm-thick membrane III-V waveguide [9] are also shown. The membrane III-V waveguide consists of a 600-nm-wide multiple quantum well (MQW) buried in the InP clad layer, i.e., buried heterostructure (BH), in which the BH is sandwiched between the SiO2 layer. The effective refractive indices of the 2-µm and 230-nm-thick III-V waveguides are 3.20 and 2.64, respectively. These results are effective indices only for fundamental transverse electric (TE) mode. The effective refractive index of the Si waveguides increases with core width until it reaches about 1 µm and then saturates. As shown in Fig. 1(a), in the case of a 220-nm Si waveguide, there is no Si waveguide core width where the effective refractive index is comparable to that of the 2-µm-thick III-V waveguide. Consequently, it is difficult to perform optical mode engineering and construct a transition between the 220-nm-thick Si and conventional 2-µm thick III-V layer. To overcome this difficulty, a Si waveguide with a thickness of about 500 nm is used to ensure their effective refractive index matching. On the other hand, since the effective refractive index of the 230-nm-thick membrane III-V waveguide is comparable to that of the 220-nm-thick Si waveguide, optical mode engineering and constructing a transition between 220-nm-thick Si waveguide and 230-nm thick III-V waveguide are possible by controlling the widths of both waveguides.
Fig. 1. (a) Calculated effective refractive indices of Si and III-V waveguides for TE mode. Schematics of a (b) bird’s eye view and (c) cross section of membrane BH DFB laser consisting of heterogeneous III-V/Si waveguide with a SiN grating. (d) Schematic of top view of device containing two output waveguides and HNAF for device characterization.
Figure 1 shows schematics of a bird’s eye view (c) and cross section (d) of the proposed membrane BH DFB laser consisting of optically coupled III-V/Si waveguides with a SiN surface grating on a Si substrate. We use a lateral current-injection (LCI) structure. This structure enables us to reduce total III-V layer thickness and reduce its effective refractive index as mentioned above, and the optical absorption in the contact materials can be neglected when the optical mode field is far enough away from the contact materials. The distance between the metal pad and the waveguide center is designed to be larger than 1.9 µm to suppress the absorption loss in the metal, because the 1/e2 width of the mode field for horizontal direction is 0.87 µm. The MQWs are buried in an InP cladding layer, which has a wider band gap and lower refractive index than the MQWs. The BH enables us to improve the carrier confinement in the active region and reduce the absorption in the p-type InP layer. In our previous works, we designed membrane lasers to reduce the operating energy for direct modulation applications. Therefore, we did not place waveguides underneath the membrane lasers to increase the optical confinement in the active region, and the cavity length was designed to be less than 100 µm [16–20]. On the other hand, the laser used as the bias light source for the MZM should be designed to achieve high-output power. Therefore, the cavity length must be increased and internal loss must be decreased. For this purpose, we can use the supermode in the III-V/Si waveguide to reduce the optical confinement factor in the III-V layers and increase it in the low-loss Si layer [23]. To construct the single-mode DFB laser, a λ/4-shifted SiN grating is formed on the top surface of the InP cladding layer. The supermode also enables us to reduce the coupling coefficient of the SiN surface grating to suppress the spatial hole-burning effect. In this study, the cavity length was set to 500 µm, which is relatively long compared with that in our reported membrane DFB lasers [16–20].
The laser has two output waveguides comprising III-V/Si, Si, and SiOx waveguides, as shown in Fig. 1(d). An InP taper waveguide is used to suppress the loss and reflection at the interface between the III-V/Si and Si waveguides. A Si taper waveguide is used at the interface between the Si and SiOx waveguides. The SiOx waveguide was designed to achieve low coupling loss when it is coupled with a single-mode high-numerical-aperture fiber (HNAF).
Figure 2(a) shows the calculated Si width dependence of the filling factors in the p-type InP, MQWs (six periods), and Si for the optically coupled III-V/Si waveguides. In the calculation, the MQW width, total III-V layer thickness, Si layer thickness, and interlayer SiO2 thickness between the III-V and Si waveguides were 600, 230, 220, and 100 nm, respectively. With increasing Si width, the filling factor in Si increases because of the increased effective refractive index of the Si waveguide, as described in Fig. 1(a). Even though the Si width is increased, the optical field is still confined in the MQWs, which is enough to ensure sufficient optical gain. Large optical loss in the cavity degrades the basic characteristics of lasers, such as the threshold current, output power, SMSR, and linewidth [24]. Since the optical loss in p-type InP is much larger than that in Si, it is important to reduce the filling factor in the p-InP. As shown in Fig. 2(a), the filling factor in p-type InP has a minimum at the Si width of 1 µm. With increasing Si width, it dramatically decreases and then increases again because the optical mode field spreads in the lateral direction along the wide Si. On the other hand, we found that the III-V/Si waveguides with Si over 0.9-µm wide has a horizontal high-order mode, which means the laser with the wide Si can lase in undesirable high-order modes when the lasing conditions are satisfied. According to these calculations, we designed the Si width to be 0.8 µm for lasing in the fundamental mode with lower cavity loss.
Fig. 2. (a) Calculated filling factors in p-InP, QWs, and Si for heterogeneous III-V/Si waveguide. (b) Calculated transmittance between III-V/Si waveguide and Si waveguide via inverse InP taper.
Figure 2(b) shows the calculated transmittance between the III-V/Si waveguide and Si waveguide via the inverse InP taper on the Si waveguide. The taper tip width is 100 nm. A taper with such a narrow tip can be easily fabricated because the membrane III-V layer provides a small aspect ratio in taper formation. The InP taper on the Si waveguide constructs an adiabatic mode transformer based on two coupled waveguides [25]. When the taper length is 10 µm or less, ripple behavior occurs due to the odd-mode coupling on the coupled-waveguide system [25]. The calculation results show that the optical mode in the laser section (i.e., the III-V/Si waveguide) is transferred to the Si waveguide with a small loss of less than 0.1 dB and that a taper length of 50 µm is long enough for low-loss coupling. According to the calculations, we designed the taper length to be 50 µm with a sufficient margin.
3. Fabrication
Figure 3 shows the fabrication procedure for the membrane BH DFB laser consisting of the III-V/Si waveguide with the SiN surface grating on a Si substrate. First, a Si waveguide was formed by dry etching using a silicon-on-insulator (SOI) substrate with a 220-nm-thick Si layer and a 2-µm-thick buried oxide (BOX) layer. A SiO2 film was deposited on the Si substrate and then polished by chemical mechanical polishing (CMP) for wafer bonding. The InP substrate containing the six-period InGaAsP-based QWs (photoluminescence (PL) peak wavelength of 1.52 µm) and an etch stop layer was directly bonded to the SOI substrate containing the Si waveguide by oxygen-plasma assisted bonding [Fig. 3(a)]. After the wafer bonding, the InP substrate was removed by using the etch stop layer. Removing the InP substrate leaves a membrane III-V layer on the Si waveguide [Fig. 3(b)]. After that, we used alignment marks in the SOI layer to align the BH position. The MQW layer was etched by dry and wet-chemical etching to form its core [Fig. 3(c)]. After that, an undoped InP layer was grown to form the BH, and then an InGaAs layer was grown as the contact material by metalorganic vapor phase epitaxy (MOVPE) [Fig. 3(d)]. Since the thermal expansion coefficients of Si, SiO2, and InP are different, thermal strain and crystal-quality degradation are induced at the process temperature. However, negligible degradation should appear when the III-V layer thickness is smaller than the critical thickness (typically less than ∼430 nm) [26]. Zn thermal diffusion and Si ion implantation were used to form the p- and n-type doping regions. Then, the InGaAs layer was patterned by wet etching with the contacts remaining. As a surface grating material, SiN was deposited to a thickness of 20 nm on the top-surface of the InP layer [Fig. 3(e)]. We used electron cyclotron resonance (ECR) plasma sputtering [27] for low-temperature deposition of a smooth, thin SiN film with high uniformity. The SiN grating was formed by dry etching, where the λ/4 shift was positioned at the center of the grating. The grating depth observed with an atomic force microscope is 20 nm, which corresponds to the thickness of the deposited SiN film. After that, InP tapers with a tip width of 100 nm were fabricated by dry etching using a regrown undoped InP film. Then, the metal electrodes were formed. Finally, the SiOx waveguide coupled with the Si taper was formed in the back-end process [5] [Fig. 3(f)]. A scanning electron microscope (SEM) image of a cross section of the fabricated laser is shown in Fig. 4. The Si is embedded in the SiO2 film, and its position is aligned with the MQW. The thickness of the SiO2 interlayer between the Si and III-V and the total III-V layer thickness are 100 and 230 nm, respectively. The widths of the MQW and Si are 600 and 840 nm, respectively.
4. Experimental results and discussion
A fabricated laser with a cavity length of 500 µm was performed by continuous-wave operation. HNAFs were physically contacted to both SiOx waveguide facets, as shown in Fig. 1(d). Figure 5(a) shows the measured output power versus injection current (L-I curve) at a stage temperatures of 25°C. Also shown is the L-I curve measured with a PD placed in front of the SiOx waveguide facet. The vertical axis is the sum of the output power emitted from both facets. By comparing the L-I curves obtained with and without the HNAFs, the fiber coupling loss was estimated to be 2 dB, which is comparable to previous results [18]. The threshold current of 3.5 mA and the maximum fiber output power of 7.9 mW were obtained at a stage temperature of 25°C. We measured the optical spectrum of the light emitted from one facet with an optical spectrum analyzer. The optical spectrum with an injection current of 40 mA is shown in Fig. 5(b). Single-mode lasing was obtained, where the lasing wavelength and SMSR were 1552 nm and 50 dB, respectively. No other lasing peaks in the entire 300-nm-wavelength span were observed. The inset of Fig. 5(b) shows the measured optical spectrum near the threshold current, with the injection current of 5 mA. The measured spectrum is in good agreement with the spectrum calculated by using the transfer matrix model, where the grating coupling coefficient was estimated to be 33 cm−1 from optical mode field simulations. These results indicate another advantage of the membrane III-V structure, namely that unwanted reflections in output waveguides are successfully suppressed when using a 220-nm-thick Si waveguide.
Fig. 5. (a) Optical output power versus injection current at a stage temperature of 25°C with and without HNAF. (b) Lasing spectrum at stage temperature of 25°C and injection current of 40 mA. Inset: lasing spectrum at stage temperature of 25°C and injection current of 5 mA and spectrum calculated based on transfer matrix model.
Compared with lasers on InP substrate, lasers on Si substrate have relatively high thermal impedance due to the low-thermal conductivity BOX layer [28–30], and this remains a common issue for high-temperature operation of lasers on Si substrate. The thermal impedance can be reduced by reducing the BOX layer thickness [28], but the optical leakage loss of the SiOx waveguide used for the fiber coupling will increase exponentially—the calculated leakage losses of the SiOx waveguide with 1- and 2-µm-thick BOX layer were 2 and 0.1 dB/mm, respectively. Therefore, to integrate the SiOx waveguides, we used an SOI substrate with an 2-µm-thick BOX layer although its thermal impedance becomes large. To achieve lasing despite the high thermal impedance, power consumption at threshold must be small. In this context, membrane BH lasers are excellent for low-threshold operation because of their small active area. We measured the L-I curves and optical spectra for stage temperatures ranging from 25 to 120°C. As shown in Fig. 6, we achieved high-temperature operation up to 120°C. With increasing stage temperature, the threshold current increased and maximum fiber output power decreased, as shown in Fig. 7. Even at the stage temperature of 80°C, the maximum fiber output power of 3.3 mW with a threshold current of 7.5 mA was obtained.
Fig. 6. (a) Optical output power versus injection current at stage temperatures ranging from 25 to 120°C. (b) Lasing spectrum at stage temperatures ranging from 25 to 120°C and injection current of 35 mA.
Fig. 7. Threshold current (a) and maximum fiber output power (b) versus stage temperatures ranging from 25 to 120°C.
Next, we estimated the thermal impedance of the fabricated laser. The thermal impedance should be indicated by the increase of active III-V region temperature as a function of electrical power consumption. The thermal impedance can be calculated from the lasing wavelength shifts induced by both temperature and power consumption. The lasing wavelength is determined by not only the temperature of the III-V region but also the temperature of the Si waveguide. Here, we simulated the temperature distribution for the laser. Figure 8(a) shows the temperature distribution, where we assumed that a 100-mW heat source was embedded in the p-InP layer for simplicity. Figure 8(b) shows the temperature increases in the Si waveguide core and the active region, which are almost the same with a 7% difference. Therefore, we estimated the thermal impedance of the fabricated laser from the lasing wavelength shift versus electrical power and stage temperature. As shown in Fig. 9, the lasing wavelength shift with increasing electrical power and stage temperature were 0.0162 nm/mW and 0.0854 nm/K, respectively. Then, the thermal impedance was calculated to be 189 K/W from these values. This thermal impedance is relatively large, however, we achieved lasing up to 120°C, as shown in Fig. 6(a). This is because the membrane BH laser achieves low-threshold operation [16–20,23,26,30,31] and low heat generation at the threshold current. From the measured thermal impedance and power consumption at the threshold current, the temperature increase for the stage temperature of 25°C was calculated to be only 0.6°C. The extremely low power consumption of the membrane BH laser compensates for the relatively large thermal impedance and suppresses the temperature increase, resulting in the high-temperature operation. Thus, it was demonstrated that the laser with the membrane BH achieves high-temperature operation as well as efficient optical coupling with 220-nm-thick Si waveguides. These results indicate that the membrane BH laser is suitable for integration on a Si photonics platform.
Fig. 8. (a) Calculated temperature distribution for the laser with a 100-mW heat source in the p-InP region. (b) Calculated temperature increment for active and Si cores.
5. Conclusion
We demonstrated a membrane BH DFB laser consisting of an optically coupled III-V/Si waveguide and a SiN surface grating. The laser was fabricated by using CMP, direct wafer bonding, and regrowth by MOVPE. The laser was integrated on a Si photonics platform using Si and SiOx waveguides, and the Si layer thickness is 220 nm on the entire platform. The maximum fiber coupling power of 7.9 mW and single-mode lasing with a SMSR of 50 dB were demonstrated in a fabricated 500-µm cavity length laser. Although the thermal impedance of the fabricated laser was estimated to be 189 K/W, high-temperature operation up to 120°C was demonstrated thanks to the low-threshold membrane BH laser. These results indicate that the membrane BH laser has the potential for use in large-scale PICs including lasers and MZMs on a Si platform.
Funding
National Institute of Information and Communications Technology.
Disclosures
The authors declare no conflicts of interest.
References
1. C. R. Doerr, “Silicon photonic integration in telecommunications,” Front. Phys. 3(37), 1–16 (2015). [CrossRef]
2. S.-H. Jeong, D. Shimura, T. Simoyama, T. Horikawa, Y. Tanaka, and K. Morito, “Si-nanowire-based multistage delayed Mach–Zehnder interferometer optical MUX/DeMUX fabricated by an ArF-immersion lithography process on 300 mm SOI wafer,” Opt. Lett. 39(13), 3702–3705 (2014). [CrossRef]
3. S. Cheung, T. Su, and K. Okamoto, “Ultra-compact silicon photonics 512×512 25 GHz arrayed waveguide grating router,” IEEE J. Sel. Top. Quantum Electron. 20(4), 310–316 (2014). [CrossRef]
4. H. Fukuda, K. Yamada, T. Tsuchizawa, T. Watanabe, S. Shinojima, and S. Itabashi, “Silicon photonic circuit with polarization diversity,” Opt. Express 16(7), 4872–4880 (2008). [CrossRef]
5. T. Tsuchizawa, K. Yamada, T. Watanabe, S. Park, H. Nishi, R. Kow, H. Shinojima, and S. Itabashi, “Monolithic integration of silicon-, germanium-, and silica-based optical devices for telecommunication applications,” IEEE J. Sel. Top. Quantum Electron. 17(3), 516–525 (2011). [CrossRef]
6. J. Witzens, “High-speed silicon photonics modulators,” Proc. IEEE 106(12), 2158–2182 (2018). [CrossRef]
7. M. L. Davenport, S. Skendzic, N. Volet, J. C. Hulme, M. J. R. Heck, and J. E. Bowers, “Heterogeneous Silicon/III-V semiconductor optical amplifiers,” IEEE J. Sel. Top. Quantum Electron. 22(6), 78–88 (2016). [CrossRef]
8. A. Abbasi, J. Verbist, J. V. Kerrebrouck, F. Lelarge, G.-H. Duan, X. Yin, J. Bauwelinck, G. Roelkens, and G. Morthier, “28 Gb/s direct modulation heterogeneously integrated C-band InP/SOI DFB laser,” Opt. Express 23(20), 26479–26485 (2015). [CrossRef]
9. T. Komljenovic, S. Srinivasan, E. Norberg, M. Davenport, G. Fish, and J. E. Bowers, “Widely tunable narrow-linewidth monolithically integrated external-cavity semiconductor lasers,” IEEE J. Sel. Top. Quantum Electron. 21(6), 214–222 (2015). [CrossRef]
10. C. T. Santis, S. T. Steger, Y. Vilenchik, A. Vasilyev, and A. Yariv, “High-coherence semiconductor lasers based on integral high-Q resonators in hybrid Si/IIIV platform,” Proc. Natl. Acad. Sci. U. S. A. 111(8), 2879–2884 (2014). [CrossRef]
11. T. Ferrotti, B. Blampey, C. Jany, H. Duprez, A. Chantre, F. Boeuf, C. Seassal, and B. B. Bakir, “Co-integrated 1.3 µm hybrid III-V/silicon tunable laser and silicon Mach-Zehnder modulator operating at 25Gb/s,” Opt. Express 24(26), 30379–30401 (2016). [CrossRef]
12. M. Tran, D. Huang, T. Komljenovic, J. Peters, A. Malik, and J. Bowers, “Ultra-low-loss silicon waveguides for heterogeneously integrated silicon/III-V photonics,” Appl. Sci. 8(7), 1139 (2018). [CrossRef]
13. P. Dong, T.-C. Hu, T.-Y. Liow, Y.-K. Chen, C. Xie, X. Luo, G.-Q. Lo, R. Kopf, and A. Tate, “Novel integration technique for silicon/III-V hybrid laser,” Opt. Express 22(22), 26854–26861 (2014). [CrossRef]
14. B. Jang, K. Tanabe, S. Kako, S. Iwamoto, T. Tsuchizawa, H. Nishi, N. Hatori, M. Noguchi, T. Nakamura, K. Takemasa, M. Sugawara, and Y. Arakawa, “A hybrid silicon evanescent quantum dot laser,” Appl. Phys. Express 9(9), 092102 (2016). [CrossRef]
15. T. Hiraki, T. Aihara, K. Takeda, T. Fujii, T. Tsuchizawa, T. Kakitsuka, and S. Matsuo, “Heterogeneously Integrated Low-power-consumption Semiconductor Optical Amplifier on Si Platform,” Proc. CLEO 2019, STh3N.1 (2019). [CrossRef]
16. S. Matsuo, T. Fujii, K. Hasebe, K. Takeda, T. Sato, and T. Kakitsuka, “Directly modulated buried heterostructure DFB laser on SiO2/Si substrate fabricated by regrowth of InP using bonded active layer,” Opt. Express 22(10), 12139–12147 (2014). [CrossRef]
17. S. Matsuo, T. Fujii, K. Hasebe, K. Takeda, T. Sato, and T. Kakitsuka, “Directly modulated DFB laser on SiO2/Si substrate for datacenter networks,” J. Lightwave Technol. 33(6), 1217–1222 (2015). [CrossRef]
18. H. Nishi, T. Fujii, K. Takeda, K. Hasebe, T. Kakitsuka, T. Tsuchizawa, T. Yamamoto, K. Yamada, and S. Matsuo, “Membrane distributed-reflector laser integrated with SiOx-based spot-size converter on Si substrate,” Opt. Express 24(16), 18346–18352 (2016). [CrossRef]
19. T. Fujii, K. Takeda, N.-P. Diamontopoulos, E. kanno, K. Hasebe, H. Nishi, R. Nakao, T. Kakitsuka, and S. Matsuo, “Heterogeneously integrated membrane lasers on Si substrate for low operating energy optical links,” IEEE J. Sel. Top. Quantum Electron. 24(1), 1–8 (2018). [CrossRef]
20. E. Kanno, K. Takeda, T. Fujii, K. Hasebe, H. Nishi, T. Yamamoto, T. Kakitsuka, and S. Matsuo, “Twin-mirror membrane distributed-reflector lasers using 20-µm-long active region on Si substrate,” Opt. Express 26(2), 1268–1277 (2018). [CrossRef]
21. T. Hiraki, T. Aihara, K. Hasebe, K. Takeda, T. Fujii, T. Kakitsuka, T. Tsuchizawa, H. Fukuda, and S. Matsuo, “Heterogeneously integrated III–V/Si MOS capacitor Mach–Zehnder modulator,” Nat. Photonics 11(8), 482–485 (2017). [CrossRef]
22. T. Hiraki, T. Aihara, K. Takeda, T. Fujii, T. Kakitsuka, T. Tsuchizawa, H. Fukuda, and S. Matsuo, “Membrane InGaAsP Mach-Zehnder modulator with SiN:D waveguides on Si platform,” Opt. Express 27(13), 18612–18619 (2019). [CrossRef]
23. T. Aihara, T. Hiraki, K. Takeda, K. Hasebe, T. Fujii, T. Tsuchizawa, T. Kakitsuka, and S. Matsuo, “Lateral current injection membrane buried heterostructure lasers integrated on 200-nm-thick Si waveguide,” Proc. OFC 2018, W3F.4 (2018).
24. T. L. Koch and U. Koren, “Semiconductor lasers for coherent optical fiber communications,” J. Lightwave Technol. 8(3), 274–293 (1990). [CrossRef]
25. X. Tun, H.-C. Liu, and A. Yariv, “Adiabaticity criterion and the shortest adiabatic mode transformer in a coupled-waveguide system,” Opt. Lett. 34(3), 280–282 (2009). [CrossRef]
26. T. Fujii, T. Sato, K. Takeda, K. Hasebe, T. Kakitsuka, and S. Matsuo, “Epitaxial growth of InP to bury directly bonded thin active layer on SiO2/Si substrate for fabricating distributed feedback lasers on silicon,” IET Optoelectron. 9(4), 151–157 (2015). [CrossRef]
27. T. Amazawa, T. Ono, M. Shimada, S. Matsuo, and H. Oikawa, “Ultrathin oxide films deposited using electron cyclotron resonance sputter,” J. Vac. Sci. Technol., B 17(5), 2222–2225 (1999). [CrossRef]
28. M. N. Sysak, H. Park, A. W. Fang, J. E. Bowers, R. Jones, O. Cohen, O. Raday, and M. Paniccia, “Experimental and theoretical thermal analysis of a Hybrid Silicon Evanescent Laser,” Opt. Express 15(23), 15041–15046 (2007). [CrossRef]
29. S. Stankovic, R. Jones, M. N. Sysak, J. M. Heck, G. Roelkens, and D. V. Thourhout, “Hybrid III V/Si distributed feedback laser based on adhesive bonding,” IEEE Photonics Technol. Lett. 24(23), 2155–2158 (2012). [CrossRef]
30. K. Doi, T. Shindo, J. Lee, T. Amemiya, N. Nishiyama, and S. Arai, “Thermal analysis of lateral-current-injection membrane distributed feedback laser,” IEEE J. Quantum Electron. 50(5), 321–326 (2014). [CrossRef]
31. T. Okamoto, N. Nunoya, Y. Onodera, S. Tamura, and S. Arai, “Low-threshold singlemode operation of membrane BH-DFB lasers,” Electron. Lett. 38(23), 1444–1446 (2002). [CrossRef]
