Open Access
Add to BibSonomyAbstract
We demonstrate germanium (Ge) microdisks surrounded by highly reflective circular Bragg gratings on highly n-doped germanium-on-insulator (GOI) substrate. The GOI substrate is fabricated by wafer bonding from Ge grown on Si substrate, and n-type doping concentration of 2.1×1019 cm−3 is achieved by phosphorus diffusion from a spin-on-dopant source. Very sharp Fabry-Perot resonant peaks with high contrast fringes and Q-factors up to 400 are observed near the direct band gap of Ge in photoluminescence spectra. The reflectivity of gratings are enhanced by a factor larger than 3 in a wide wavelength range from 1.57 to 1.82 µm, compared with that of Ge/SiO2 interfaces in normal microdisks without circular Bragg gratings. The surface emission intensity of the devices is found to be increased by the grating period. Our results indicate that GOI microdisk with circular Bragg grating is a promising optical resonator structure suitable for realizing low threshold, compact Ge lasers integrated on Si substrate.
© 2017 Optical Society of America
1. Introduction
Germanium (Ge) has attracted great attention recently as a promising candidate for realization of monolithic-integrated laser sources on silicon (Si) substrate [1, 2]. Due to the relatively small energy difference between its direct and indirect band gaps, Ge can be engineered close to or into a direct band gap semiconductor by using methods such as applying tensile strain [3–5], n-type doping [6, 7], and alloying with tin (Sn) [8]. Significant progress has been made recently, leading to demonstration of preliminary Ge [9–11] and GeSn [12–14] lasers. In order to achieve fundamental direct band gap, relatively high tensile strain (~1.6–2.0% in the case of biaxial) or high Sn composition (~10%) is needed. These will impose technical difficulties in material preparation and device fabrication. On the other hand, by incorporating heavy n-type doping, the required tensile strain for realization of lasers can be largely reduced. Moreover, the emission wavelength still lies in the telecommunication band, which has great advantage that little modification is needed for other photonic devices already well developed.
In most reports, so far, the active Ge layer was usually directly grown on Si substrate, in which a low-temperature buffer layer was necessary to accommodate the large lattice mismatch between Si and Ge [15, 16]. Large amount of dislocations are inevitably created at Ge/Si interface, which behave as non-radiative recombination centers, and thus limiting the quantum efficiency of light emitting devices. On the other hand, the Ge layer near the top is less affected by the interface dislocations, which is more favorable as device layer. By bonding the face-down Ge epitaxial layer with another handle substrate, e.g, Si substrate with SiO2 layer on top, a Ge-on-Insulator (GOI) substrate is formed and it is thus possible to remove the low-temperature buffer layer from top [17, 18]. Moreover, the burred oxide layer can provide strong optical confinement in Ge in vertical direction. It is thus easy to fabricate high Q-factor optical resonators, which are inevitable to realize laser devices. Compared with III–V materials bonded to Si substrate in which high performance laser devices have been demonstrated [19, 20], GOI has the advantages of being compatible with mature complementary metal-oxide-semiconductor (CMOS) process and capability of large wafer fabrication, thus is highly promising to realize low cost electronicphotonic integrated chips.
Among variety of optical resonators, microdisks are very attractive due to their miniaturized footprints and low optical loss. Two types of resonant modes exist in microdisks, namely, Fabry-Perot (FP) modes formed from the reflection between two opposite sidewalls of microdisks, and whispering-gallery-modes (WGMs) traveling along the perimeter of microdisks by total internal reflection [21, 22]. WGMs have inherently higher Q-factors due to large refractive index contrast between Ge and cladding materials. However, the sidewall roughness induced scattering loss may degrade the Q-factor significantly. Also, the mode profiles of WGMs are located near the microdisk edges, where strain relaxation is maximum due to the formation of free surface and injected carriers are largely affected by surface recombination. So far, WGM-related resonance has only been observed in undoped Ge microdisks, with Q-factors up to 700 near direct band gap, and over than 4000 near indirect band gap [16,23–26]. On the other hand, FP resonant modes are less affected by the sidewalls, and have been observed in most of Ge microdisks, including both tensile-strained [27, 28] and/or n-doped [29]. However, due to the low reflectivity at the Ge/cladding interface, the resonant peaks were usually with low contrast and low Q-factors. In order to enhance the reflectivity, one of the most promising solutions is to surround the microdisk with Bragg gratings [30, 31]. M. El Kurdi et al recently demonstrated highly tensile-strained Ge microdisks with circular Bragg reflectors and observed cavity resonance with Q-factor as high as 2000 around 2 µm [32]. This work represents a significant achievement towards realization of practical Ge lasers since two very important aspects including high tensile strain and high Q-factor cavity have been encompassed. However, the n-type doping concentration was still relatively low. Also, in order to apply sufficiently high and uniform tensile strain, free-standing structure was used, which might cause thermal issues under high excitation level.
In this work, we have fabricated highly n-doped GOI substrate with doping concentration up to 2.1×1019 cm−3 by direct wafer bonding and spin-on-dopant (SOD) diffusion. Ge microdisks surrounded by circular Bragg gratings are realized by standard lithography and dry etching process. FP resonant peaks with high contrast and high Q-factors are observed from photoluminescence (PL) spectra at both room and low temperatures. Compared with normal Ge microdisks, the reflectivity of Bragg grating is found to be increased by a factor of >3 in a wide wavelength range covering the emission spectra of Ge. We also investigate the dependence of emission spectra on the grating period by comparing experimental and simulation results. The pump power dependence is finally studied and the route towards realization of lasers is discussed.
2. Material preparation and device fabrication
GOI substrate was fabricated by direct wafer bonding of a handle Si substrate with thermally grown SiO2 layer of 650-nm-thick and a donor Si substrate with 740-nm-thick tensile-strained Ge layer grown by solid-source molecular beam epitaxy. The material growth and wafer bonding procedures were described in [16] and [25] in detail. The low-temperature Ge buffer layer was removed by chemical mechanical polishing and the Ge device layer was thinned down to about 220 nm. GOI substrate was then n-doped by phosphorus diffusion from a SOD source and a subsequent rapid thermal annealing at 800 °C for 30 seconds [7]. Figure 1(a) shows the transmission electron microscope (TEM) image of cross-section of a fabricated GOI sample. The stacking faults and misfit dislocations near Ge/Si interface can be completely removed in GOI, and only very few threading dislocations are visible in the image. The doping profile of phosphorus measured by secondary ion mass spectrometry (SIMS) is shown in Fig. 1(b). The sharp peak around 200 nm indicates that phosphorus was piled up at Ge/SiO2 interface. In Ge layer, uniform doping concentration of about 9×1019 cm−3 was achieved. In order to obtain activated doping concentration, room-temperature PL for both undoped and n-doped GOI was measured. The PL spectra with pump laser power of 1 mW are shown in Fig. 1(c). As expected, the PL intensity of n-doped GOI is significantly enhanced, by a factor of 4.8, compared with that of undoped GOI. The PL peak is also red-shifted by ~34 meV, due to the band gap narrowing effect caused by heavy doping. This corresponds to an activated doping concentration of about 2.1×1019 cm−3 [33], indicating that only 23% phosphorus was activated under current annealing condition.
Fig. 1 (a) Cross-section TEM image of n-doped GOI substrate. (b) Phosphorus doping profile measured by SIMS. Inset shows the doping profile in linear scale. (c) Room-temperature PL spectra of undoped and n-doped GOI. The pump laser is a 980 nm laser and the power is 1 mW.
Devices including normal microdisks (MDs) and microdisks with circular Bragg gratings (MDCBGs) were then fabricated on the n-doped GOI substrate by electron beam lithography and inductively coupled plasma reactive-ion etching. For MDCBGs, microdisks were surrounded by 9 and half periods of concentric gratings and Ge was fully etched down to the buried oxide (BOX) in the grating trenches. This is only possible because the whole device structure is formed on solid substrate. The gratings were designed and fabricated with periods ranging from 280 to 400 nm, and filling factor of 0.5, such that the corresponding stop bands for TE polarization cover the light emission wavelength range of Ge. Finally, SiO2 cladding layer with thickness of about 650 nm was deposited on top of the substrate by plasma-enhanced chemical vapor deposition in order to ensure symmetric optical confinement along vertical direction. Figure 2 shows scanning electron microscope (SEM) images of a fabricated MD and MDCBG before SiO2 deposition, as well as zoomed-view of the grating.
Fig. 2 SEM images of (a) a normal microdisk and (b) a microdisk with circular Bragg grating, both with diameters of 8 µm. (c) is the zoomed-view of the grating with a period of 400 nm.
3. Measurement results and discussions
The light emission properties of the devices were characterized by micro-PL measurement, in which a 980 nm continuous wave semiconductor laser was focused onto the devices by an objective lens (100×, NA=0.50), and the light emission was collected by the same objective lens and detected by an extended InGaAs detector. Figure 3 shows PL spectra of a MD and MDCBG measured at room temperature (RT) and low temperature (LT) of 29 K. The disk diameters are both 16 µm, and the grating period for MDCBG is 400 nm. Resonant fringes were observed near the direct band gap of Ge in both devices. The resonant peaks appear only in a narrow wavelength range below the direct band gap at RT, while span over much wider wavelength range at LT, and can be even observed above the direct band gap. The Q-factors and fringe contrast, defined by (Imax + Imin)/(Imax − Imin), where Imax and Imin are peak and valley intensities in PL spectra respectively, of resonant peaks in LT PL spectra are also much larger than those in RT PL spectra. These can be attributed to the fact that free carrier absorption induced by heavy doping and optically injected carriers and residual material absorption of Ge are significantly reduced at LT [32].
Fig. 3 Room-temperature PL spectra of n-doped GOI (a) MD and (b) MDCBG, with pump laser power of 10 mW; Low-temperature (T=29 K) PL spectra of n-doped GOI (c) MD and (d) MDCBG, with pump laser power of 30 mW.
The spectra from both devices consist of a broadband background and several sharp resonant peaks. The broadband background corresponds to the light emission coupled to the radiation modes in Ge slab waveguide and contains no resonant feature, while resonant peaks correspond to the light emission coupled to the FP resonant modes formed between two opposite mirrors (Ge/SiO2 interface in MD or Bragg grating in MDCBG) [30]. From the resonant peaks in the PL spectrum shown in Fig. 3(c) and 3(d), we calculated the group refractive index to be about 4.674 by assuming the cavity roundtrip length to be twice of the disk diameter. This is consistent with the group refractive index of Ge slab waveguide (∼4.554 around wavelength of 1.64 µm), indicating there resonant peaks indeed correspond to FP resonant modes, rather than WGMs with different cavity length. These FP modes are guided modes propagating along the in-plane direction, and can only be detected from the lens overhead through the scattering of Ge/SiO2 interface in MD or the diffraction of the grating in MDCBG. Due to the larger diffraction efficiency of the grating, we can observe that the resonant peaks are much more intense in MDCBG than those in MD.
Moreover, compared with those in MD, the resonant peaks in MDCBG are much sharper at both RT and LT. For example, Q-factor of resonant peak around 1.67 µm is increased from 199 to 395 in LT PL spectra. The fringe contrast is also much larger in MDCBG. The improvement was observed in all of the fabricated devices with diameters of 4, 8, 12 and 16 µm. This is due to the much higher reflectivity of Bragg grating that that of Ge/SiO2 interface. By comparing PL spectra of devices with different diameters, we noticed that there was no visible resonance splitting caused by different azimuthal numbers [32] for 16 µm microdisks, most likely due to the fact that the lateral dimension of devices is much larger than wavelength in Ge. Simple FP theory thus can be used to extract the reflectivity and cavity loss. The PL spectra, after subtracting the broadband background, were fitted by an universal transmittance function of FP resonator
where C is a constant, R the mirror reflectivity, α the loss coefficient of medium inside FP resonator, n the refractive index of medium inside resonator, D the diameter of MD, and λ the wavelength. Half round trip transmittance (HRTT) Rexp(−αD) of the resonators at LT, extracted through the fitting, are shown in Fig. 4. The HRTT in MDCBG is much larger than that in MD over a broadband wavelength range. Since the disks in MD and MDCBG have the same dimension and pumping condition, the total material loss α, including the residual material loss and free carrier absorption loss, is same for both devices. The enhancement is thus directly related the reflectivity. Around 1.77 µm, the reflectivity of Bragg grating is 3.8 times larger than that of Ge/SiO2 interface. The threshold gain needed for lasing, represented by thus can be reduced by a factor of 3.1, indicating that our devices are promising structure for realization of low threshold Ge lasers. In a broad wavelength range from 1.57 to 1.82 µm, we could obtain reflectivity enhancement factor of larger than 3. This range is also consistent with the simulation result that the designed grating has a wide stop band from 1.5 to 2.3 µm.Fig. 4 Extracted half round trip transmittance Rexp(−αD) of n-doped GOI MD and MDCBG, from LT PL spectra.
The light emission properties of MDCBGs with different grating periods were then investigated and analyzed by numerical simulation based on finite-difference time-domain (FDTD) method [34]. As we mentioned above, the lateral dimension the the microdisk is much larger than wavelength, the devices thus can be modeled by 2D structures, only considering the cross-section geometry across the diameter of the devices. Figure 5(a) shows LT PL spectra of MDCBGs with different grating periods, together with calculated grating reflectivity spectra. During simulation, only Bragg grating regions were included in the computation domain. A pulse field source, with spatial distribution of fundamental mode of Ge slab waveguide and temporal distribution consisting of a Gaussian envelope function multiplying a sinusoidal carrier, was used to excite the structure. A field monitor located behind the excitation source was used to record the reflected electromagnetic field and Fourier transformation was then performed to obtain the frequency/wavelength response of the device (e.g., reflectivity spectrum). For gratings with periods of 320, 360, and 400 nm, their stop bands can fully cover the light emission spectrum of n-doped Ge. Therefore, high contrast FP fringes were all observed. The ultrabroad stop band is owing to the high index contrast of the fully etched grating structure. The extracted HRTT of the resonant peaks barely changes with grating period, which is consistent with the simulation result that the reflectivity in stop band has weak dependence on grating period. This indicates that the devices are quit tolerant against fabrication discrepancy, which is another important advantage.
Fig. 5 (a) LT PL spectra of MDCBGs (solid line) and simulated reflectivity spectra (dashed line) of Bragg gratings with different grating periods. (b) Calculated majority electric field intensity profiles of the resonant modes at 1.64 µm in the form of log10(E2). (c)–(f) are calculated far-field patterns of the corresponding resonant modes for different grating periods: (c) a=280 nm, (d) a=320 nm, (e) a=360 nm, and (f) a=400 nm. A zoomed view of the far-field pattern is also shown in the inset of (c).
Besides, the intensity of broadband background in the PL spectra is nearly the same for MDCBGs with grating periods of 320, 360, and 400 nm, while the intensity of resonant peaks is significantly increased against grating period. Since the background is related to the radiation modes of Ge slab waveguides, it is thus independent on the grating structure. On the other hand, the resonant peaks, corresponding to the FP resonant modes, are strongly modified by the resonator structure. Figure 5(b) plots the calculated majority electric field intensity profiles of one of the resonant modes at 1.64 µm of each device. Figure 5(c)–(f) plot the corresponding far-field patterns in the vertical half-space. For these simulations, the whole device region was included in the computation domain. An impulse field source, with Gaussian function spatial distribution and delta-function temporal distribution, was used to excite the structure. After obtaining the mode profiles, the far-field patterns were calculated from the near field intensity at the device surface through a near-to-far-field projection based on Fourier transformation. As the grating period increases, the resonant mode is strongly diffracted towards vertical direction by the grating. The far-field intensity is also found to be increased as the grating period, which is consistent with the experimental results. Therefore, as long as the stop band of grating could cover the light emission spectrum of Ge, it is possible to optimize the surface emission intensity by increasing grating period, while keeping grating reflectivity as high as enough.
The PL spectra of MDCBG under different pump power are shown in Fig. 6(a). Although sharp resonant emission peaks have been observed in our devices, we don’t consider them as lasing behavior since linewidth narrowing is obviously not observed as the pump power increases. As shown in Fig. 6(b), the extracted HRTT continuously decreases against pump power. Since the grating reflectivity is not affected by the pump laser, the total loss of Ge is increasing, indicating that free carrier absorption is still dominant in our devices. Besides, the red-shift of resonant peaks indicates that pump laser induced heating couldn’t be ignored. The raised temperature also leads to a red-shift of the band gap of Ge, and this shift is faster than that of the resonant peaks [26, 35]. Therefore, the residual material absorption is also increased near the band edge, indicating that the temperature raising also partly contributes to the linewidth broadening. Based on the temperature-dependent refractive index of Ge [36], we calculated by FDTD simulation that the temperature coefficient of the resonant wavelength shift around 1.64 µm is about 0.19 nm/K. From experimental results shown in Fig. 6(b), it can be estimated that the device temperature was raised by nearly 100 K by increasing pump power from 10 to 60 mW. This is much lower than that in free-standing microdisks, in which only 3 mW pump laser induced a temperature increasing of 200 K [32]. It is attributed to the much larger thermal conductivity of SiO2 than air. The thermal resistance of the devices can be further improved by using materials with even larger thermal conductivity such as Al2O3 as the BOX layer during GOI fabrication [37].
Fig. 6 (a) Low-temperature (29 K) PL spectra of n-doped GOI MDCBG under different pump power. The spectra were each offset by 2000 along the vertical axis for visibility. (b) Resonant wavelength and extracted Rexp(−αD) of the resonant peak around 1.65 µm under different pump power.
Our results indicate that the linewidth narrowing and optical gain couldn’t be obtained in Ge under current tensile strain and n-type doping levels. It is thus necessary to further increase the strain and/or doping concentration. The tensile strain could be increased by depositing stressor layer like Si3N4 on top of the devices [5]. Due to the higher n-type doping concentration, the tensile strain required for optical gain could be reduced, thus we expect that top stressor might be enough to apply sufficient strain while not using free-standing structure. Also, in our device structure, the active medium is the whole disk region, thus nonuniform strain distribution near the disk edge induced by top stressor is not a concern. For n-type doping, our SIMS measurement result has shown that the phosphorus incorporation could be as high as 9×1019 cm−3 and the doping concentration was only limited by the activation rate. This can be solved by using some advanced dopant activation techniques, such as flash-lamp annealing [38] and pulse laser annealing [39]. Moreover, since the grating region in our devices is also heavily doped and not pumped, the free carrier absorption could degrade the reflectivity. A selective drive-in diffusion process, in which the grating region can be masked by SiO2, can be performed to increase the reflectivity, thus decreasing the threshold gain furthermore.
4. Conclusion
Ge microdisks with circular Bragg gratings have been realized on highly n-doped GOI substrate fabricated by wafer bonding and spin-on-dopant diffusion. Sharp resonant emission peaks, corresponding to FP resonant modes, were observed at both room and low temperatures. The Q-factors, FP fringe contrast and wavelength ranges where resonant peaks appear were all enhanced as the temperature was reduced. Compared with that of Ge/SiO2 interfaces in normal microdisks without circular Bragg gratings, the reflectivity of gratings was increased by a factor larger than 3 in wide wavelength ranges and found to be tolerant to the grating period. From the experimental and simulation results, we also found that the surface emission intensity increased against grating period. These allow us to optimize the emission intensity as well as keeping the reflectivity high. Our device structure is thus a promising optical resonator for low threshold and miniaturized Ge lasers. Pump power dependent measurement showed that optical gain was not obtained in current devices, indicating tensile strain and dopant activation ratio should be increased further in order to realize Ge lasers.
Funding
MEXT Supported Program for the Strategic Research Foundation at Private Universities 2015–2019; Grant-in-Aid for Scientific Research (B) (26286044).
Acknowledgments
This work was partly supported by Nanotechnology Research Center for Analysis of Tokyo City University, Japan.
References and links
1. J. Liu, X. Sun, D. Pan, X. Wang, L. C. Kimerling, T. L. Koch, and J. Michel, “Tensile-strained, n-type Ge as a gain medium for monolithic laser integration on Si,” Opt. Express 15(18), 11272–11277 (2007). [CrossRef] [PubMed]
2. S. Saito, A. Z. Al-Attili, K. Oda, and Y. Ishikawa, “Towards monolithic integration of germanium light sources on silicon chips,” Semicond. Sci. Technol. 31, 043002 (2016). [CrossRef]
3. J. R. Jain, A. Hryciw, T. M. Baer, D. A. B. Miller, M. L. Brongersma, and R. T. Howe, “A micromachining-based technology for enhancing germanium light emission via tensile strain,” Nat. Photonics 6(6), 398–405 (2012). [CrossRef]
4. M. J. Suess, R. Geiger, R. A. Minamisawa, G. Schiefler, J. Frigerio, D. Chrastina, G. Isella, R. Spolenak, J. Faist, and H. Sigg, “Analysis of enhanced light emission from highly strained germanium microbridges,” Nat. Photonics 7(6), 466–472 (2013). [CrossRef]
5. G. Capellini, C. Reich, S. Guha, Y. Yamamoto, M. Lisker, M. Virgilio, A. Ghrib, M. El Kurdi, P. Boucaud, B. Tillack, and T. Schroeder, “Tensile Ge microstructures for lasing fabricated by means of a silicon complementary metal-oxide-semiconductor process,” Opt. Express 22(1), 399–410 (2014). [CrossRef] [PubMed]
6. M. El Kurdi, T. Kociniewski, T.-P. Ngo, J. Boulmer, D. Debarre, P. Boucaud, J. F. Damlencourt, O. Kermarrec, and D. Bensahel, “Enhanced photoluminescence of heavily n-doped germanium,” Appl. Phys. Lett. 94, 191107 (2009). [CrossRef]
7. X. Xu, X. Wang, K. Nishida, K. Takabayashi, K. Sawano, Y. Shiraki, H. Li, J. Liu, and T. Maruizumi, “Ultralarge transient optical gain from tensile-strained, n-doped germanium on silicon by spin-on dopant diffusion,” Appl. Phys. Express 8, 092101 (2015). [CrossRef]
8. R. Chen, H. Lin, Y. Huo, C. Hitzman, T. I. Kamins, and J. S. Harris, “Increased photoluminescence of strain-reduced, high-Sn composition Ge1−x Snx alloys grown by molecular beam epitaxy,” Appl. Phys. Lett. 99, 181125 (2011). [CrossRef]
9. J. Liu, X. Sun, R. Camacho-Aguilera, L. C. Kimerling, and J. Michel, “Ge-on-Si laser operating at room temperature,” Opt. Lett. 35(5), 679–681 (2010). [CrossRef] [PubMed]
10. R. E. Camacho-Aguilera, Y. Cai, N. Patel, J. T. Bessette, M. Romagnoli, L. C. Kimerling, and J. Michel, “An electrically pumped germanium laser,” Opt. Express 20(10), 11316–11320 (2012). [CrossRef] [PubMed]
11. R. Koerner, M. Oehme, M. Gollhofer, M. Schmid, K. Kostecki, S. Bechler, D. Widmann, E. Kasper, and J. Schulze, “Electrically pumped lasing from Ge Fabry-Perot resonators on Si,” Opt. Express 23(11), 14815–14822 (2015). [CrossRef] [PubMed]
12. S. Wirths, R. Geiger, N. von den Driesch, G. Mussler, T. Stoica, S. Mantl, Z. Ikonic, M. Luysberg, S. Chiussi, J. M. Hartmann, H. Sigg, J. Faist, D. Buca, and D. Grutzmacher, “Lasing in direct-bandgap GeSn alloy grown on Si,” Nat. Photonics 9(2), 88–92 (2015). [CrossRef]
13. D. Stange, S. Wirths, R. Geiger, C. Schulte-Braucks, B. Marzban, N. Driesch, G. Mussler, T. Zabel, T. Stoica, J. Hartmann, S. Mantl, Z. Ikonic, D. Grutzmacher, H. Sigg, J. Witzens, and D. Buca, “Optically pumped GeSn microdisk lasers on Si,” ACS Photon. 3(7), 1279–1285 (2016). [CrossRef]
14. S. Al-Kabi, S. A. Ghetmiri, J. Margetis, T. Pham, Y. Zhou, W. Dou, B. Collier, R. Quinde, W. Du, A. Mosleh, J. Liu, G. Sun, R. A. Soref, J. Tolle, B. Li, M. Mortazavi, H. A. Naseem, and S. Yu, “An optically pumped 2.5 µ m GeSn laser on Si operating at 110K,” Appl. Phys. Lett. 109, 171105 (2016). [CrossRef]
15. H. Luan, D. R. Lim, K. K. Lee, K. M. Chen, J. G. Sandland, K. Wada, and L. C. Kimerling, “High-quality Ge epilayers on Si with low threading-dislocation densities,” Appl. Phys. Lett. 75(19), 2909–2911 (1999). [CrossRef]
16. K. Nishida, X. Xu, K. Sawano, T. Maruizumi, and Y. Shiraki, “Highly n-doped, tensile-strained Ge grown on Si by molecular beam epitaxy,” Thin Solid Films 557, 66–69 (2014). [CrossRef]
17. J. R. Jain, D. Ly-Gagnon, K. C. Balram, J. S. White, M. L. Brongersma, D. A. B. Miller, and R. T. Howe, “Tensile-strained germanium-on-insulator substrate fabrication for silicon-compatible optoelectronics,” Opt. Mater. Express 1(6), 1121–1126 (2011). [CrossRef]
18. Y. Hoshi, K. Sawano, K. Hamaya, M. Miyao, and Y. Shiraki, “Formation of tensilely strained germanium-on-insulator,” Appl. Phys. Express 5, 015701 (2012). [CrossRef]
19. A. W. Fang, H. Park, O. Cohen, R. Jones, M. J. Paniccia, and J. E. Bowers, “Electrically pumped hybrid AlGaInAs-silicon evanescent laser,” Opt. Express 14(20), 9203–9210 (2006). [CrossRef] [PubMed]
20. J. Van Campenhout, P. Rojo-Romeo, P. Regreny, C. Seassal, D. Van Thourhout, S. Verstuyft, L. Di Cioccio, J.-M. Fedeli, C. Lagahe, and R. Baets, “Electrically pumped InP-based microdisk lasers integrated with a nanophotonic silicon-on-insulator waveguide circuit,” Opt. Express 15(11), 6744–6749 (2007). [CrossRef] [PubMed]
21. J. S. Xia, Y. Ikegami, K. Nemoto, and Y. Shiraki, “Observation of whispering-gallery modes in Si microdisks at room temperature,” Appl. Phys. Lett. 90, 141102 (2007). [CrossRef]
22. X. Xu, S. Narusawa, T. Chiba, T. Tsuboi, J. Xia, N. Usami, T. Maruizumi, and Y. Shiraki, “Silicon-based light-emitting devices based on Ge self-assembled quantum dots embedded in optical cavities,” IEEE J. Sel. Top. Quantum Electron. 18(6), 1830–1838 (2012). [CrossRef]
23. P. H. Lim, Y. Kobayashi, S. Takita, Y. Ishikawa, and K. Wada, “Enhanced photoluminescence from germanium-based ring resonators,” Appl. Phys. Lett. 93, 041103 (2008). [CrossRef]
24. G. Shambat, S. Cheng, J. Lu, Y. Nishi, and J. Vuckovic, “Direct band Ge photoluminescence near 1.6 µ m coupled to Ge-on-Si microdisk resonators,” Appl. Phys. Lett. 97, 241102 (2010). [CrossRef]
25. X. Xu, H. Hashimoto, K. Yoshida, K. Sawano, and T. Maruizumi, “High Q-factor resonant photoluminescence from Ge-on-Insulator microdisks,” Proc. SPIE 9891, 98910V (2016). [CrossRef]
26. A. Z. Al-Attili, S. Kako, M. K. Husain, F. Y. Gardes, N. Higashitarumizu, S. Iwamoto, Y. Arakawa, Y. Ishikawa, H. Arimoto, K. Oda, and S. Saito, “Whispering galler mode resonances from Ge micro-disks on suspended beams,” Front. Mater. 2, 00043 (2015). [CrossRef]
27. A. Ghrib, M. El Kurdi, M. de Kersauson, M. Prost, S. Sauvage, X. Checoury, G. Beaudoin, I. Sagnes, and P. Boucaud, “Tensile-strained germanium microdisks,” Appl. Phys. Lett. 102, 221112 (2013). [CrossRef]
28. R. W. Millar, K. Gallacher, J. Frigerio, A. Ballabio, A. Bashir, I. MacLaren, G. Isella, and D. J. Paul, “Analysis of Ge micro-cavities with in-plane tensile strains above 2 %,” Opt. Express 24(5), 4365–4374 (2016). [CrossRef]
29. X. Xu, K. Sawano, T. Maruizumi, and Y. Shiraki, “Resonant light emission from n-doped Germanium-on-Insulator microdisks at room-temperature,” in Proceedings of 11th International Conference on Group IV Photonics (IEEE, 2014) pp. 231–232.
30. M. Y. Su and R. P. Mirin, “Enhanced light extraction from circular Bragg grating coupled microcavities,” Appl. Phys. Lett. 89, 033105 (2006). [CrossRef]
31. X. Sun and A. Yariv, “Surface-emitting circular DFB, disk-, and ring-Bragg resonator lasers with chirped gratings: a unified theory and comparative study,” Opt. Express 16(12), 9155–9164 (2008). [CrossRef] [PubMed]
32. M. El Kurdi, M. Prost, A. Ghrib, A. Elbaz, S. Sauvage, X. Checoury, G. Beaudoin, I. Sagnes, G. Picardi, R. Ossikovski, F. Boeuf, and P. Boucaud, “Tensile-strained germanium microdisks with circular Bragg reflectors,” Appl. Phys. Lett. 108, 091103 (2016). [CrossRef]
33. R. Camacho-Aguilera, Z. Han, Y. Cai, L. C. Kimerling, and J. Michel, “Direct band gap narrowing in highly doped Ge,” Appl. Phys. Lett. 102, 152106 (2013). [CrossRef]
34. RSoft FullWAVE, Synopsys, Inc., https://optics.synopsys.com/rsoft/.
35. X. Xu, K. Nishida, K. Sawano, T. Maruizumi, and Y. Shiraki, “Resonant photoluminescence from microdisk based on n-doped, tensile-strained Ge on Si,” presented at International Conference on Solid State Devices and MaterialsFukuoka, Japan, 24–27, Sept. 2013
36. R. F. Potter, “Germanium (Ge),” in Handbook of Optical Constants of Solids, E. D. Palik, ed. (Academic, 1985). [CrossRef]
37. Y. Moriyama, K. Ikeda, S. Takeuchi, Y. Kamimuta, Y. Nakamura, K. Izunome, A. Sakai, and T. Tezuka, “Ultrathin-body Ge-on-insulator wafers fabricated with strongly bonded thin Al2O3/SiO2 hybrid buried oxide layers,” Appl. Phys. Express 7, 086501 (2014). [CrossRef]
38. S. Prucnal, F. Liu, M. Voelskow, L. Vines, L. Rebohle, D. Lang, Y. Berencen, S. Andric, R. Boettger, M. Helm, S. Zhou, and W. Skorupa, “Ultra-doped n-type germanium thin films for sensing in the mid-infrared,” Sci. Rep. 6, 27643 (2016). [CrossRef] [PubMed]
39. R. Milazzo, G. Impellizzeri, D. Piccinotti, D. De Salvador, A. Portavoce, A. La Magna, G. Fortunato, D. Mangelinck, V. Privitera, A. Carnera, and E. Napolitani, “Low temperature deactivation of Ge heavily n-type doped by ion implantation and laser thermal annealing,” Appl. Phys. Lett. 110, 011905 (2017). [CrossRef]
