We demonstrate a submicrometer-scale hydrogenated amorphous silicon (a-Si:H) waveguide with a record low propagation loss of 0.60 ± 0.02 dB/cm because of the very low infrared optical absorption of our low defect a-Si:H film, the optimized waveguide structure and the fabrication process. The waveguide has a core with a thickness of 440 nm and a width of 780 nm that underlies a 100-nm-thick ridge structure, and is fabricated by low-cost i-line stepper photolithography and with low-temperature processing at less than 350°C, making it compatible with the backend process of complementary metal oxide semiconductor (CMOS) fabrication.
© 2014 Optical Society of America
Plasma-deposited hydrogenated amorphous silicon (a-Si:H) has already found a wide range of applications, including solar cells, thin film transistors, X-ray image sensors and integrated fluorescence detectors for microfluidic biochemical analysis [1,2]. The next application of this industry-proven and mature material could be to silicon photonic integrated circuits (PICs). Thanks to its high refractive index and high transparency in the telecommunication wavelength range, a-Si:H could form the submicrometer-scale cores of compact optical waveguides for high-density PICs. The high confinement of optical mode field in the submicrometer-scale waveguide combined with higher optical nonlinearity and lower nonlinear absorption of a-Si:H than the counterparts in crystalline silicon (c-Si) makes a-Si:H an attractive material for optical nonlinear devices [3–5]. Various examples include parametric amplifiers, all-optical signal processing devices and all-optical modulators [6–10]. Furthermore, the low-temperature deposition of device-grade a-Si:H films allows for the vertical stacking of optical circuit layers simply by repeating the process of a-Si:H core fabrication clad with an SiO2 layer, and thus constructing a three-dimensional (3D) optical interconnect (OI) without thermal damage to the underlying large-scale integrated (LSI) circuit. While the conventional silicon-on-insulator (SOI) approach is inherently challenging for stacking of optical layers, and thereby ultimately limits the integration density, the 3D architecture could provide higher integration density and low crosstalk at the waveguide crossings because of the excellent optical isolation provided by the SiO2 layer .
The increasing need for revolutionary on-chip 3D OIs originates from one of the biggest problems that modern LSIs are facing today: the increasing signal delays and power consumption of electrical interconnects (EIs). In particular, this is very serious for the global wiring layers, which are the topmost layers among the multiple stacked electrical layers above CMOS devices, because of their comparatively long signal conduction distances. One possible solution is to replace the EIs with 3D OIs that provide higher speeds, broader bandwidths and lower power consumption. The only missing passive device required to complete this solution is the vertical optical signal transfer device located between the upper and lower waveguides. Very recently, we have demonstrated a knife-edge tapered a-Si:H waveguide that works very well, not only for spot-size converters but also for low-loss vertical interlayer coupling [12,13]. In addition, the existing CMOS backend process used for the global EI can also be deployed to reduce the fabrication costs of the on-chip OIs. Overall, the a-Si:H based on-chip 3D OI is very promising for the next generation of LSIs. Therefore, it is imperative to continuously improve the performance of each device.
The low-loss a-Si:H optical waveguides are particularly important, because they are the most fundamental devices and the propagation loss is directly related to the power dissipation of the OIs as well as it greatly influences the performance of nonlinear devices. Optical excitation of a-Si:H in the telecommunication wavelength range around 1.55 μm happens to be associated with a transition involving a Si dangling bond defect state (hereafter called defect absorption), so it is imperative to minimize the defect density to obtain high transparency at 1.55 μm. Although they exhibited a very low propagation loss of 1.0-1.2 dB/cm [8, 14], our low defect a-Si:H wire waveguides were fabricated using an electron beam lithography (EBL), making it particularly challenging for high volume production with low cost. Other research groups have also reported submicrometer-scale a-Si:H waveguides [15–23]. However, the reported losses are not low enough for “optical global wiring” on a few-cm scale.
In this paper, we first investigated the propagation losses of wire and ridge waveguides with standard 220-nm-thick a-Si:H and c-Si cores clad with plasma-deposited SiO2, which was fabricated using i-line stepper photolithography, and then confirmed that the propagation losses due to the absorption of a-Si:H are negligible, based on the sub-gap absorption spectrum of the a-Si:H film measured by a constant photocurrent method (CPM) calibrated by transmittance and reflection (T&R) spectroscopy. Because a comparison between the a-Si:H and c-Si waveguides implies that the top surface roughness of a-Si:H contributes to the propagation loss, we planarized the top surface of the a-Si:H film by chemical mechanical polishing (CMP) and fabricated thicker a-Si:H ridge waveguides such that the optical mode field was far from the surface. As a result, we achieved a propagation loss of 0.60 dB/cm. To the best of our knowledge, this is the lowest reported value among the submicrometer-scale a-Si:H waveguides [15–23].
2. 220-nm-thick waveguides
In this work, two types of a-Si:H waveguide were tested: fully etched wire waveguides and partially etched ridge waveguides. The propagation loss of the ridge waveguide is influenced more by the waveguide material properties, such as the absorption and scattering of the core. For reference, the propagation losses of the c-Si waveguides on SOI wafers were also evaluated.
220-nm-thick a-Si:H films were deposited by plasma decomposition of a source gas mixture of SiH4 and H2 at 250°C on a Si wafer with a 2-μm-thick thermal oxide, while the SOI wafer comprises a 220-nm-thick top Si layer and an underlying 2-μm-thick buried oxide layer on a bulk Si substrate. The patterns of the 420-nm-wide wire waveguides and the 780-nm-wide ridge waveguides were transferred to photoresists on the a-Si:H wafer and the SOI wafer using an i-line stepper (NSR-2205i12D, Nikon, Japan). For the wire waveguide patterning, which is close to the diffraction limit, a bottom anti-reflection coating (BARC) was inserted underneath the photoresist to ensure the uniformity of the waveguide width. After the photoresist was developed, the BARC was etched with inductively coupled plasma (ICP) of a CF4 and O2 gas mixture and the Si was subsequently etched using an ICP of a SF6 and C4F8 gas mixture. In contrast, the ridge waveguides were formed by CF4-based capacitively coupled plasma-reactive ion etching. A 1.5-μm-thick silicon dioxide layer was finally deposited as the upper cladding by plasma decomposition of a gas mixture of tetraethyl orthosilicate (TEOS) and O2 at a deposition temperature of 350°C.
Figure 1 shows cross-sectional scanning electron microscope (SEM) images of the fabricated waveguides together with the main electrical field (Ex) profile of the quasi-transverse electric (q-TE) mode, as simulated by the finite difference method (FDM). The waveguide dimensions are summarized in Table 1, and were extracted from the SEM images. Although both c-Si and a-Si:H were etched under the same process conditions, the etching depth of a-Si:H is slightly deeper when compared with c-Si, meaning that the etching rate of a-Si:H is slightly higher. The difference in the etching depths, however, is too small to affect the waveguide performance. We therefore ignored this difference in the following discussion.
To evaluate the propagation losses of the fabricated waveguides using a standard cut-back method, 12 wire waveguides were formed on a chip, and were then classified into three sets containing four waveguides, with lengths ranging from 0.5 to 1.7 cm. Similarly, 16 ridge waveguides were formed, and were then classified into four sets containing four ridge waveguides with different lengths. To laterally couple light to/from all waveguides through the common edge of the chip, the waveguides were arranged in a convoluted manner with a curve radius of 10 μm for the wire waveguides and 400 μm for the ridge waveguides. Near the chip edge, the wire waveguides were widened from 420 to 1500 nm to enhance their coupling efficiencies by expanding the optical spot-size. These wire and ridge waveguides satisfy the single mode conditions with negligible bending losses for the q-TE mode at the wavelength of 1.55 μm; and these results were confirmed by numerical analysis based on FDM.
The losses were evaluated for the q-TE mode only, because the ridge waveguides do not support a quasi-transverse magnetic (q-TM) mode. The TE-polarized light at 1.55 μm was fed into the waveguides, and the output light was measured using an optical power meter. Figure 2 shows the measured transmittances of the waveguides plotted against the waveguide length, and the propagation losses were extracted from the slopes of the linear fittings.
Next, the dependence of the transmittance on the wavelength in the C-band, ranging from 1530 to 1565 nm, was measured for the 17-mm-long waveguides, as shown in Fig. 3. TE-polarized broadband light generated by an erbium-doped fiber amplifier was coupled into the waveguide. The output light was measured using an optical spectrum analyzer.
Despite the fact that both the a-Si:H and c-Si waveguides were fabricated under the same conditions, the a-Si:H waveguides have higher propagation losses than their c-Si counterparts. These differences are likely to originate from something that is relevant to the material properties. To elucidate the cause, we characterized the absorption coefficient of a-Si:H in the telecommunications wavelength range.
Figure 4 shows the sub-gap absorption spectrum of a 1-μm-thick a-Si:H film deposited under the same conditions that was measured using the CPM, which was again calibrated with the T&R spectrum. The thick film is necessary to remove any interference effects and properly evaluate the absorption coefficient in the sub-gap region. The absorption coefficient of the thick film was 10−2 cm−1 at a wavelength of 1550 nm, corresponding to 0.04 dB/cm. The defect density that was evaluated from the integrated defect absorption is 4.2 × 1015 cm−3, based on a conversion factor of 1.9 × 1016 cm−2eV−1 , indicating that the film has a low defect density. It is well known that the neutral Si dangling bond defect density measured by electron spin resonance (ESR) shows a linear dependence on the film thickness, but with significantly high extrapolated intercepts at a thickness of zero . This means that the bulk defect density of the a-Si:H film is independent of thickness, but also means that the surface defect density is high. Therefore, we conclude that the absorption due to bulk defects is negligible relative to the propagation losses of the waveguide. Table 2 shows a comparison of the deposition temperatures and the material losses of the submicrometer-scale a-Si:H waveguides reported to date, showing that the material losses of our a-Si:H film might be the lowest. Because of its thermodynamically non-equilibrium nature, the properties of a-Si:H are influenced by the deposition parameters. In particular, the deposition temperature is the most critical parameter in determining the film quality. As the deposition temperature increases, precursor SiH3 radicals may find energetically favorable sites because of the enhanced surface diffusion of SiH3 on the hydrogen-covered growth surface, resulting in a reduced defect density. Above 250°C, the thermal desorption of hydrogen occurs, increasing the defect density. As a result of the balance of these two reactions, the lowest defect density is usually achieved at a process temperature of around 250°C .
The other potential origin to contribute propagation loss would be the presence of microvoids in the a-Si:H waveguide. It is argued that the dihydride SiH2 and trihydride SiH3 make an a-Si:H film heterogeneous, creating H clustering or microvoids in the film . Fourier transform infrared (FTIR) absorption spectroscopy of our low defect a-Si:H film deposited on a intrinsic float zone c-Si substrate, however, shows stretching mode of only monohydride SiH around 2000 cm−1 (7-8% hydrogen content); even no shoulder at higher wavenumber region is observed. Therefore, the density of the microvoids would be sufficiently low that the light scattering generated at the microvoids is negligible.
In contrast, the top surface of the as-grown a-Si:H is approximately three times as rough as that of the SOI wafer, as observed by atomic force microscope (AFM), and as shown in Fig. 5. Because the absorption in the upper and lower claddings is also negligible, the loss differences between the a-Si:H and c-Si waveguides, i.e. ~1.8 dB/cm for the wire waveguide and ~0.5 dB/cm for the ridge waveguide, originate from this top surface roughness. Therefore, reduction of the top surface roughness is likely to improve the propagation loss, although we cannot rule out the possibility that absorption by interface defects may also contribute to the propagation loss.
It should also be noted that the ridge waveguides exhibited lower losses than the wire waveguides. This is because of the smaller overlap of the optical field with the rough etched sidewalls for the ridge waveguides compared with that in the wire waveguides (see Fig. 1(e) and (f)). The overlap can be controlled by changing the a-Si:H thickness, but this is not the case with the c-Si waveguide. An increase in the thickness of the a-Si:H waveguide would move the optical field away from the etched sidewalls and the top surface, thus reducing the scattering losses at the interfaces.
3. Loss reduction
A propagation loss of less than 1 dB/cm is preferable for these waveguides . To further reduce the losses in the a-Si:H waveguide and pursue a limit due to the material loss in the waveguide structure, we planarized the top surface of the a-Si:H film by CMP and fabricated a ridge waveguide with a thicker core. The fabrication procedure is as follows.
We first deposited a 500-nm-thick a-Si:H film on a 2-μm-thick thermal oxide top layer on a Si wafer, followed by the removal of an approximately 60-nm-thick a-Si:H layer by CMP. Figure 6(a) shows the smooth top surface of the polished a-Si:H film. The ridge waveguides were then fabricated with a ridge height of 100 nm. Figure 6(b) shows a cross-sectional image of the fabricated waveguide, and its dimensions are summarized in Table 3. Note that the fabricated waveguide only supports a fundamental q-TE mode and that the bending loss at a radius of 400 μm is again negligible.
Figure 7 shows that the a-Si:H ridge waveguide (AR440) exhibited a propagation loss of 0.60 ± 0.02 dB/cm for the q-TE mode at a wavelength of 1.55 μm. This record low propagation loss was achieved by combining the low absorption of our low defect a-Si:H film with the thicker core of the ridge waveguide and the CMP-planarized top surface. It should be noted here that the polished top surface of the a-Si:H film is as smooth as that of the SOI wafer, as shown in Fig. 6(a). Since the propagation loss for the q-TM mode is dominated by the light scattering generated at the top surface, the propagation loss of the CMP-planarized a-Si:H wire waveguide would be comparable to that of the c-Si wire waveguide. We have not fabricated it yet, though. Concerning the q-TE mode, we expect that the propagation loss of the CMP-planarized a-Si:H wire waveguide would be somewhat higher compared to the c-Si wire waveguide. We speculate that subtle difference of the sidewall roughness as a result of the aforementioned different etching properties might produce non-trivial difference of propagation losses between a-Si:H and c-Si wire waveguides for the q-TE mode. This argument will be tested by fabricating and evaluating the CMP-planarized a-Si:H wire waveguide and the results will be published in near future.
For optical nonlinear devices, several research groups are concerned about photo-induced degradation (PID). No increase of the propagation loss in our a-Si:H waveguide, however, was observed not only in this work but also in our recent four wave mixing experiment . An optical gap determined by Tauc plot was 1.75 eV in our film, which is higher compared to the previous work , so that two photon absorption would be significantly reduced, suppressing non-radiative recombination of photo-generated carriers and photo-induced degradation. Recently, light-induced hydrogen diffusion was observed in reference to PID . Although we did not investigate hydrogen stability during the infrared laser exposure, there might be correlation between the light-induced hydrogen diffusion and PID. Nevertheless, the defect absorption would be more relevant to the propagation loss, whether it is degraded or not.
Also, the dependence of the transmittance on the wavelength in the C-band is within a 0.5 dB range for the 17-mm-long waveguide, as shown in Fig. 7.
According to , the CMP process may create new Si dangling bonds at the surface, significantly increasing the propagation losses. While the CMP process has reduced the losses in this work, the passivation of surface defects may be critical for further reduction of the propagation losses.
We achieved a propagation loss of 0.60 ± 0.02 dB/cm in a submicrometer-scale a-Si:H waveguide because of the low absorption of our a-Si:H film, the reduced overlap of the optical field with the top and sidewall surfaces obtained by increasing the a-Si:H core thickness, and the smoother top surface produced by CMP processing. To the best of our knowledge, this is the lowest propagation loss obtained for submicrometer-scale a-Si:H waveguides. The propagation loss does not reach the limit imposed by the material properties of the low defect a-Si:H. We therefore believe that it is possible to further reduce the propagation losses of these a-Si:H waveguides by optimizing the fabrication process and improving the passivation of the surface defects.
This research was funded by the Japan Society for the Promotion of Science (JSPS) through the “Funding Program for World-Leading Innovative R&D on Science and Technology (FIRST Program),” initiated by the Council for Science and Technology Policy (CSTP). Some of this work was conducted at the AIST Nano-Processing Facility, which is supported by the “Nanotechnology Support Project.”
References and links
1. R. A. Street, Technology and Application of Amorphous Silicon, (Springer, 2000).
2. T. Kamei, B. M. Paegel, J. R. Scherer, A. M. Skelley, R. A. Street, and R. A. Mathies, “Integrated hydrogenated amorphous Si photodiode detector for microfluidic bioanalytical devices,” Anal. Chem. 75(20), 5300–5305 (2003). [CrossRef] [PubMed]
3. K. Narayanan and S. F. Preble, “Optical nonlinearities in hydrogenated-amorphous silicon waveguides,” Opt. Express 18(9), 8998–9005 (2010), http://www.opticsinfobase.org/oe/abstract.cfm?uri=oe-18-9-8998. [CrossRef] [PubMed]
4. Y. Shoji, T. Ogasawara, T. Kamei, Y. Sakakibara, S. Suda, K. Kintaka, H. Kawashima, M. Okano, T. Hasama, H. Ishikawa, and M. Mori, “Ultrafast nonlinear effects in hydrogenated amorphous silicon wire waveguide,” Opt. Express 18(6), 5668–5673 (2010), http://www.opticsinfobase.org/oe/abstract.cfm?uri=oe-18-6-5668. [CrossRef] [PubMed]
5. C. Lacava, P. Minzioni, E. Baldini, L. Tartara, J. M. Fedeli, and I. Cristiani, “Nonlinear characterization of hydrogenated amorphous silicon waveguides and analysis of carrier dynamics,” Appl. Phys. Lett. 103(14), 141103 (2013). [CrossRef]
6. B. Kuyken, H. Ji, S. Clemmen, S. K. Selvaraja, H. Hu, M. Pu, M. Galili, P. Jeppesen, G. Morthier, S. Massar, L. K. Oxenløwe, G. Roelkens, and R. Baets, “Nonlinear properties of and nonlinear processing in hydrogenated amorphous silicon waveguides,” Opt. Express 19(26), B146–B153 (2011), http://www.opticsinfobase.org/oe/abstract.cfm?uri=oe-19-26-b146. [CrossRef] [PubMed]
7. K.-Y. Wang, K. G. Petrillo, M. A. Foster, and A. C. Foster, “Ultralow-power all-optical processing of high-speed data signals in deposited silicon waveguides,” Opt. Express 20(22), 24600–24606 (2012), http://www.opticsinfobase.org/oe/abstract.cfm?uri=oe-20-22-24600. [CrossRef] [PubMed]
8. S. Suda, K. Tanizawa, Y. Sakakibara, T. Kamei, K. Nakanishi, E. Itoga, T. Ogasawara, R. Takei, H. Kawashima, S. Namiki, M. Mori, T. Hasama, and H. Ishikawa, “Pattern-effect-free all-optical wavelength conversion using a hydrogenated amorphous silicon waveguide with ultra-fast carrier decay,” Opt. Lett. 37(8), 1382–1384 (2012). [CrossRef] [PubMed]
9. K. Narayanan, A. W. Elshaari, and S. F. Preble, “Broadband all-optical modulation in hydrogenated-amorphous silicon waveguides,” Opt. Express 18(10), 9809–9814 (2010), http://www.opticsinfobase.org/oe/abstract.cfm?uri=oe-18-10-9809. [CrossRef] [PubMed]
10. B. Kuyken, S. Clemmen, S. K. Selvaraja, W. Bogaerts, D. Van Thourhout, P. Emplit, S. Massar, G. Roelkens, and R. Baets, “On-chip parametric amplification with 26.5 dB gain at telecommunication wavelengths using CMOS-compatible hydrogenated amorphous silicon waveguides,” Opt. Lett. 36(4), 552–554 (2011). [CrossRef] [PubMed]
11. K. Furuya, R. Takei, T. Kamei, Y. Sakakibara, and M. Mori, “Basic study of coupling on three-dimensional crossing of Si photonic wire waveguide for optical interconnection on inter or inner chip,” Jpn. J. Appl. Phys. 51(4S), 04DG12 (2012). [CrossRef]
12. R. Takei, M. Suzuki, E. Omoda, S. Manako, T. Kamei, M. Mori, and Y. Sakakibara, “Silicon knife-edge taper waveguide for ultralow-loss spot-size converter fabricated by photolithography,” Appl. Phys. Lett. 102(10), 101108 (2013). [CrossRef]
13. R. Takei, E. Omoda, M. Suzuki, S. Manako, T. Kamei, M. Mori, and Y. Sakakibara, “Low-loss optical interlayer transfer for three-dimensional optical interconnect,” in Proceedings of 10th International Conference on Group IV Photonics (Seoul, South Korea, 2013), pp. 91–92. [CrossRef]
14. K. Furuya, K. Nakanishi, R. Takei, E. Omoda, M. Suzuki, M. Okano, T. Kamei, M. Mori, and Y. Sakakibara, “Nanometer-scale thickness control of amorphous silicon using isotropic wet-etching and low loss wire waveguide fabrication with the etched material,” Appl. Phys. Lett. 100(25), 251108 (2012). [CrossRef]
15. A. Harke, M. Krause, and J. Mueller, “Low-loss singlemode amorphous silicon waveguides,” Elec. Lett. 41(25), 1377–1379 (2005). [CrossRef]
16. D. K. Sparacin, R. Sun, A. M. Agarwal, M. A. Beals, J. Michel, and L. C. Kimerling, “ Low loss amorphous silicon channel waveguides for integrated photonics,” in Proceedings of 3rd International Conference on Group IV Photonics (Ottawa, Canada, 2006), pp. 255–257. [CrossRef]
17. B. Han, R. Orobtchouk, T. Benyattou, P. R. A. Binetti, S. Jeannot, J. M. Fedeli, and X. J. M. Leijtens, “Comparison of optical passive integrated devices based on three materials for optical clock distribution,” in Proceedings of ECIO 07 (Copenhagen, Denmark, 2007), pp. 1–4.
18. S. K. Selvaraja, E. Sleeckx, M. Schaekers, W. Bogaerts, D. V. Thourhout, P. Dumon, and R. Baets, “Low-loss amorphous silicon-on-insulator technology for photonic integrated circuitry,” Opt. Commun. 282(9), 1767–1770 (2009). [CrossRef]
19. R. Sun, J. Cheng, J. Michel, and L. Kimerling, “Transparent amorphous silicon channel waveguides and high-Q resonators using a damascene process,” Opt. Lett. 34(15), 2378–2380 (2009). [CrossRef] [PubMed]
20. S. Zhu, G. Q. Lo, and D. L. Kwong, “Low-loss amorphous silicon wire waveguide for integrated photonics: effect of fabrication process and the thermal stability,” Opt. Express 18(24), 25283–25291 (2010), http://www.opticsinfobase.org/oe/abstract.cfm?uri=oe-18-24-25283. [CrossRef] [PubMed]
21. J. Kang, Y. Atsumi, M. Oda, T. Amemiya, N. Nishiyama, and S. Arai, “Low-loss Amorphous Silicon Multilayer Waveguides Vertically Stacked on Silicon-on-Insulator Substrate,” Jpn. J. Appl. Phys. 50(12R), 120208 (2011). [CrossRef]
22. T. Lipka, O. Horn, J. Amthor, and J. Müller, “Low-loss multilayer compatible a-Si:H optical thin films for photonic applications,” J. Europ. Opt. Soc. Rap. Public. 7, 12033 (2012). [CrossRef]
23. S. Zhu, G. Q. Lo, W. Li, and D. L. Kwong, “Effect of cladding layer and subsequent heat treatment on hydrogenated amorphous silicon waveguides,” Opt. Express 20(21), 23676–23683 (2012), http://www.opticsinfobase.org/oe/abstract.cfm?uri=oe-20-21-23676. [CrossRef] [PubMed]
24. S. Rao, F. G. D. Corte, and C. Sumonte, “Low-loss amorphous silicon waveguides grown by PECVD on indium tin oxide,” J. Europ. Opt. Soc. Rap. Public. 5, 10039s (2010). [CrossRef]
25. T. Kamei, N. Hata, A. Matsuda, T. Uchiyama, S. Amano, K. Tsukamoto, Y. Yoshioka, and T. Hirao, “Deposition and extensive light soaking of highly pure hydrogenated amorphous silicon,” Appl. Phys. Lett. 68(17), 2380 (1996). [CrossRef]
26. T. Shimizu, H. Kidoh, M. Matsumoto, A. Morimoto, and M. Kumeda, “Photo-created defects in a-Si:H as elucidated by ESR, LESR and CPM,” J. Non-Crys. Solids 114, 630–632 (1989).
28. R. Suzuki, Y. Kobayashi, T. Mikado, A. Matsuda, P. J. Mcelheny, S. Mashima, H. Ohgaki, M. Chiwaki, T. Yamazaki, and T. Tomimasu, “Characterization of Hydrogenated Amorphous Silicon Films by a Pulsed Positron Beam,” Jpn. J. Appl. Phys. 30(10), 2438–2441 (1991). [CrossRef]
29. P. Dong, W. Qian, S. Liao, H. Liang, C.-C. Kung, N.-N. Feng, R. Shafiiha, J. Fong, D. Feng, A. V. Krishnamoorthy, and M. Asghari, “Low loss shallow-ridge silicon waveguides,” Opt. Express 18(14), 14474–14479 (2010), http://www.opticsinfobase.org/oe/abstract.cfm?uri=oe-18-14-14474. [CrossRef] [PubMed]
30. F. Kail, S. Fellah, A. Abramov, A. Hadjadj, and P. R. I. Cabarrocas, “Experimental evidence for extended hydrogen diffusion in silicon thin films during light-soaking,” J. Non-Crys. Solids 362, 1083–1086 (2006).