Open Access
Add to BibSonomyAbstract
Waveguide integrated MSM (metal-semiconductor-metal) Germanium (Ge) photodetectors (PDs) with a SiGe capping layer were exploited for silicon photonics integration. Under optimized epitaxial growth conditions, the capping layer passivated the Ge surface, resulting in sufficiently low dark current of the PDs. In addition, the PDs exhibited a narrower distribution of the dark current than PDs with a Si capping layer, probably due to the lower surface leakage current. Low-noise differential receivers with uniform MSM Ge PDs exhibiting 10 Gbps data transmission were realized.
© 2013 Optical Society of America
1. Introduction
In today’s broadband era, there is growing demand for higher data traffic in communication systems. In conventional inter-chip and intra-chip communication systems, however, copper interconnects are facing serious problems due to a limitation of serial data rates and an increase of power consumption [1]. Silicon photonics, in which optical devices are integrated with Si-based electronics, is a promising solution to this problem because of its potential to substantially increase bandwidth density and reduce power consumption [2]. Several optoelectronic transceivers have been demonstrated on Si using Si photonics technology [3–7]. We previously demonstrated high-density optical interconnects integrated with laser diodes, Si optical modulators, germanium (Ge) photodetectors (PDs), and Si optical waveguides on single Si with a transmission density of 3.5 Tbps/cm2 [8].
Further increases in the transmission density of optical interconnects, along with further shrinkage in the integrated chip size, will aggravate electrical crosstalk between the modulator and the photodetector, and signal noise will become severe in the integrated chip [6]. One solution to suppress the crosstalk is to use an optical circuit, such as differential receiver circuit [9–12]. Maximizing the effects of crosstalk suppression requires an optimization of the devices used in the circuit. Especially, the uniformity of the pair of PDs used in the differential receiver is important.
In the work reported here, we focused on optimizing Ge PDs in an optical circuit. Among the various types of PDs [13–20], we used metal-semiconductor-metal (MSM) PDs [19, 20] as they have two particular advantages over p-i-n type PDs with respect to process integration. One is the simplicity of the fabrication process because MSM PDs do not need impurity implantation and subsequent activation annealing. This also simplifies the process designing and shortens the turn-around time (TAT) of Si manufacturing. The other advantage is the simpler metal layout of the MSM PDs, which enables the PD to be arranged more flexibly than a p-i-n type PD [21]. The layout flexibility is particularly important for highly integrated optical circuits.
However, MSM PDs generally suffer from large dark current [22]. This is because a high electron field is applied to the PD’s Ge surface, where a large number of carrier recombination centers exist [23]. The leakage current at the Ge surface is thus much larger than that of a p-i-n PD. In addition, strong Fermi-level pinning of the metal/Ge interface to the valence band results in a low Schottky barrier height (SBH) [24], which further enlarges the dark current. Various methods have been proposed for suppressing the dark current, i.e., insertion of amorphous Ge (a-Ge) or amorphous Si (a-Si) film between the metal and Ge [25, 26], using an asymmetric layer structure of metal contacts [27, 28], and passivating the Ge surface [29]. Insertion of a-Ge or a-Si film seems the most suitable solution because it results in a sufficient SBH and passivates the Ge surface at the same time. However, these amorphous films are deposited non-selectively over the whole surface area, and film above the Si optical waveguide causes optical propagation loss due to light scattering. Hence, additional lithography and etching are needed to remove the amorphous film outside the PD area. This increase in the number of process steps reduces the simplicity advantage of MSM PD integrated silicon optoelectronic process. The asymmetric metal layer structure [27, 28] and Ge surface passivation [29] could be combined to suppress the dark current, but this would increase the complexity of the process.
In this paper, we present a method for suppressing the dark current that does not increase the number of process steps. A single crystal silicon germanium (SiGe) capping layer was epitaxially grown on the Ge surface to passivate it. It passivates Ge surface by forming covalent bonds on it. This reduces the Fermi-level pinning effect and also suppresses the surface leakage current. Since the SiGe capping layer can be grown selectively on Ge, lithography and etching are not needed as they are in application of an amorphous passivation layer. The key virtue of using the SiGe capping layer is the simplicity of the selective growth compared to the growth of other single-crystal materials such as Si [30]. For example, Si selective growth requires strict growth conditions such as high growth temperature [31], using a corrosive gas such as dichlorosilane (SiH2Cl2) [32], adding an etching gas such as HCl during the growth [33], or limiting a growth time within an incubation period for Si deposition on SiO2 [34]. The above conditions for Si selective growth cause drawbacks to the optoelectronic integration or Si manufacturing. A high temperature growth inhibits a fine control of the doping profiles in Si optical modulators [35] integrated with PD due to severe impurity diffusion. Usage of the corrosive gas shortens the maintenance period of the epitaxial growth equipment. Addition of HCl causes undesirable etching of Ge underneath, and a growth using the incubation period tends to result in a rough surface [34]. In contrast, HCl-free SiGe selective growth can be easily achieved by using conventional low-pressure chemical vapor deposition (LP-CVD) with mono-silane (SiH4) as the source gas of the Si, which is a common condition in Si manufacturing.
This paper describes process optimization of a waveguide integrated MSM Ge PD with a SiGe capping layer and its characteristics such as very low dark current. It explains how the SiGe capping layer effectively improves the uniformity of PD performance. Finally, a differential receiver circuit with highly uniform MSM Ge PDs is described.
2. Structure and fabrication of MSM Ge PD
Figure 1 shows the schematic structure of the MSM Ge PD with a SiGe capping layer. It has an evanescent structure in which the Ge light absorption layer and SiGe capping layer are selectively grown on a Si optical waveguide. The waveguide is formed by patterning the Si-on-insulator (SOI) layer, and the growth area is defined by patterning silicon dioxide (SiO2) deposited on the waveguide. The growth is performed by LP-CVD with a growth pressure of 5 Torr. The source gases are SiH4 and mono-germane (GeH4), and the carrier gas is hydrogen (H2). The growth temperature of the Ge light absorption layer is fixed at 600 °C. The layers are grown without using any etching gas (HCl or Cl2) during the growth nor performing any etch back after the growth. The thickness of the Ge layer is fixed at 1 μm. The Ge composition of the SiGe capping layer is set to about 10%, and its thickness varies from 10 nm to 30 nm. Above the SiGe capping layer, a TiN/Al/TiN/Ti metal layer is sputtered and patterned to form a double schottky junction. The distance between the metal-SiGe contacts varies from 0.8 μm to 2 μm.
3. Process optimization
The epitaxial growth conditions were optimized to reduce the dark current. When a Si or SiGe capping layer is grown on Ge, it is generally difficult to obtain a flat surface because the Ge tends to interdiffuse into the capping layer [36]. Moreover, in the case of SiGe growth, Ge atoms in the SiGe are likely to degrade the surface flatness further, because the surface migration of Ge is larger than that of Si [37] and Ge atoms can easily migrate to the position where the strain between the SiGe and Ge can be effectively released.
Figure 2 shows SEM images of selectively grown Ge light absorption layer and SiGe capping layer. The thickness of the SiGe capping layer was 20 nm. Layers grown under the standard condition (shown on the left side of Fig. 2) reveal that the surface of the SiGe capping layer is quite large to the extent that the Ge underneath it is partially exposed to the surface. To improve the surface flatness, we reduced the growth temperature to suppress the interdiffusion between the Ge and the SiGe capping layer. Figure 3 shows the growth rate of the SiGe capping layer as a function of the inverse growth temperature. The cross and circle marks represent the flatness of the SiGe layer surface. The surface roughness of the SiGe capping layer improved with a decrease in the growth temperature. However, the growth rate decreased significantly. Moreover, the low-temperature growth induces non-selective growth because the low growth rate increases the time that the surface exposed to the source gases and the small migration length of adatoms increases nucleation density on the oxide [31]. In this study, the temperature regime where the SiGe surface became smooth was pretty close to the temperature region in which non-selective growth occurs. We therefore optimized an additional parameter, the H2 flow rate, to achieve a flat surface at relatively high growth temperatures. When the H2 flow rate was reduced from 20 l/min. to 10 l/min., the surface flatness improved, as indicated in Fig. 3. Since this improvement was accompanied by an increase in the growth rate, we attributed the improvement to an H2 limited growth mechanism. If the H2 flow rate is high, surface hydrogen limits the SiGe growth [38] and enhances the SiGe/Ge interdiffusion and Ge migration, resulting in large surface roughness.
Fig. 3 Growth rate of SiGe capping layer as function of inverse growth temperature. Cross and circle marks indicate flatness of capping layer: cross marks represent rough surface, and circles represent flat surface.
Therefore, the surface flatness can be improved by reducing the H2 flow rate, which suppresses the interdiffusion and Ge migration. Using the reduced H2 flow condition, we obtained a flat SiGe surface in the temperature regime ensuring selective growth, as shown on the right side of Fig. 2. We also examined the SiGe/Ge interface and Ge composition of the SiGe layer. Figures 4(a) and 4(b) respectively show a TEM image and SIMS profile of a SiGe capping layer grown under the optimized growth condition. An atomically flat SiGe/Ge interface is evident in Fig. 4(a), and a step-like abrupt Ge profile is evident in Fig. 4(b), both of which reflect the suppression of interdiffusion and Ge migration. Figure 4(a) also indicates good quality of the SiGe capping layer. Although several twin boundaries or stacking faults were observed in the layer due to the strain relaxation of SiGe, their number was small and their effects on the diode properties is expected to be negligible.
4. Characteristics of MSM Ge PD
Using the optimized growth condition, we fabricated an MSM Ge PD for characterization purposes.
We first examined the schottky characteristics between the metal and SiGe capping layer. Figure 5(a) shows the I-V characteristics of the metal-SiGe schottky junction at various temperatures. The test structure, shown in the inset of Fig. 5(a), consisted of a bulk Si substrate, a non-selectively grown Ge light absorption layer, a non-selectively grown SiGe capping layer, and patterned metal contacts on the SiGe surface. The bias was applied at the surface metal, and the ground was connected to the backside metal. The rectified I-V characteristics at various temperatures indicate the formation of a schottky junction. Since the negative bias corresponds to the forward bias of the diode, holes are considered to be the majority carriers. This is because dislocations at Ge/Si interface act as acceptors and the Ge light absorption layer exhibits p-type characteristics [39]. Then, the hole schottky barrier height was derived from the temperature dependence of the I-V characteristics and schottky diode equations written below.
Fig. 5 (a) I-V characteristics of SiGe capped Ge schottky structure. Schematic image of test structure is shown in inset. I-V curves were measured at various temperatures. (b) Diode current divided by the squared absolute temperature as a function of inverse temperature. Data were taken at several forward biases from Fig. 5(a). The inset of the figure shows the value of ΦBp – qV/n as a function of the absolute bias.
Symbols in the equations correspond to the parameters listed in Table 1.
Figure 5(b) shows diode current divided by the squared absolute temperature as a function of inverse temperature. Data were taken at several forward biases and fitted to Eq. (1). From the fitting, the value of ΦBp – qV/n was derived at each bias, as shown in the inset of Fig. 5(b). ΦBp was then obtained by the linear fitting of the plot shown in the inset. The value of ΦBp was estimated to be 0.45 eV.
Using the value of ΦBP with the band gap of SiGe and Ge, we derived an energy band structure of the MSM as shown in Fig. 6. The schottky barrier height for electron (ΦBn) was calculated to be 0.56 eV. Both the electron and hole barriers were found to be sufficiently high for double schottky junctions.
Fig. 6 Energy band structure of MSM Ge PD with SiGe capping layer. Hole energy barrier (ΦBp) was extracted from temperature dependence of forward current shown in Fig. 5(b). Electron energy barrier (ΦBn) was estimated from ΦBp and band gap of SiGe.
We next fabricated MSM patterns on a bare Ge layer with a SiGe capping layer and characterized the photo current and dark current. Figure 7(a) shows the photo and dark currents as a function of the applied voltage. A schematic image of the test structure is shown at the top of the figure. Incident light with a wavelength of 1.55 μm was illuminated vertically onto the surface. The photo current was sufficiently high and the dark current was quite low for this simple MSM Ge PD structure with a SiGe capping layer. The photo and dark currents of an MSM Ge PD with a Si capping layer are also plotted in Fig. 7(a) for comparison. Both were comparable to those of the MSM Ge PD with the SiGe capping layer. This indicates that the SiGe capping layer is as effective as the Si capping layer. The slightly higher dark current with the SiGe capping layer is attributed to the lower hole schottky barrier height.
Fig. 7 (a) Photo and dark current characteristics of simple MSM PD structures. Schematic image of test structure is shown at top. (b) Dark current of a single pair of double Schottky junctions with SiGe capping layer and with Si capping layer. The current is normalized to the current of the SiGe capped sample which has the smallest metal spacing. (c) Cumulative probability of MSM Ge PD’s dark current distribution.
The key feature of the MSM Ge PD with a SiGe capping layer is its dark current behavior at relatively high voltage: it does not exhibit a rapid increase in the dark current as is the case with the Si capping layer. To clarify the reason for this, we examined dark current dependence on the metal spacing using a single pair of double Schottky junctions. As shown in Fig. 7(b), the dark current of the double Schottky junctions with a SiGe capping layer was virtually insensitive to the metal spacing whereas that of the MSM Ge PD with a Si capping layer increased as the metal spacing was reduced. We attribute this difference in behavior to the difference in defect density at the interface between the Ge layer and the capping layer. The lattice mismatch between the Ge layer and the SiGe capping layer was smaller than that between the Ge layer and the Si capping layer, resulting in a lower defect density. Therefore, in the MSM Ge PD with a SiGe capping layer, there was less defect-related surface recombination than the one with a Si capping layer, resulting in weaker dependence of the dark current on the electric field.
Since the dark current of the MSM Ge PD with a SiGe capping layer was found to be virtually insensitive to the metal spacing, this PD would have uniform characteristics. Figure 7(c) shows the cumulative probability of the MSM Ge PD’s dark current distribution with the SiGe capping layer and with the Si capping layer. The distribution for the PD with the SiGe capping layer was much narrower than the one with the Si capping layer, reflecting the dark current insensitiveness of the PD with the SiGe capping layer to the metal spacing. The narrower distribution of the dark current was also seen in the PD with the SiGe capping layer when the capping layer thickness was 10 nm and 30 nm (not shown). Thus, the SiGe capping layer is potentially valid for producing MSM Ge PDs with highly uniform characteristics.
Finally, a waveguide integrated MSM Ge PD with a SiGe capping layer (structure shown in Fig. 1) was fabricated. The Ge area size was 7 × 30 μm2. Thickness of a SiGe capping layer was set to 20 nm by taking into account the tradeoff between the dark current and high-speed characteristics. Dark current of the PD was found to increase with decreasing the capping layer thickness. Prominent increase of the dark current was seen when the thickness was 10 nm, while sufficiently low dark current was obtained when the thickness was larger than 20 nm. On the other hand, high-speed characteristics were expected to degrade with increasing the capping layer thickness because electric field is consumed in the capping layer and the electric field applied to the Ge light absorption layer becomes week. Therefore, 20 nm is an optimal thickness which enables both low dark current and high-speed characteristics. Figure 8 shows the characteristics of PD: photo and dark current (Fig. 8(a)), frequency response characteristics (Fig. 8(b)), and eye diagrams (Fig. 8(c)). All the measurements shown here were taken using a laser light source with a 1.55 μm wavelength. As shown in Fig. 8(a), a sufficiently low dark current (70 nA at 1 V) and sufficient photo current were obtained. From the photocurrent, the responsivity was estimated to be 1.0 A/W. From the frequency response shown in Fig. 8(b), the 3 dB cutoff frequency was 8.5 GHz at 5 V dc bias and 12.5 GHz at 10 V. In Fig. 8(c), the eye diagrams were obtained at 20 Gbps with a 27-1 pseudorandom binary sequence (PRBS). Clear eye openings were obtained. Table 2 summarizes the characteristics of the MSM Ge PD with a SiGe capping layer.
Fig. 8 (a) Photo and dark current characteristics of waveguide integrated MSM Ge PD with SiGe capping layer. (b) Frequency response characteristics of MSM Ge PD. (c) Eye diagrams of MSM Ge PD at 20 Gbps.
Table 2. Characteristics of waveguide integrated MSM Ge PD with SiGe capping layer
The PD has several rooms to improve the performance concerning a viewpoint of power consumption, because the eye diagrams were taken at relatively high voltage (7 V) and the voltage dependence of the 3 dB cutoff frequency was large. This is because the PD had relatively large metal spacing (0.8 μm) which weakened electric field inside the Ge light absorption layer. The high-speed performance at low operating voltage would be improved by using advanced CMOS process in which the metal spacing is around 0.3 μm [20].
5. Differential optical receivers with MSM Ge PDs
Taking the advantage of the good uniformity of PD characteristics, we applied MSM Ge PDs with a SiGe capping layer to a differential receiver system of balanced PDs that requires equalized characteristics of a PD pair. Figure 9(a) shows the configuration of the differential optical circuit. It comprises arrayed lasers, Si optical modulators, and MSM Ge PDs, all of which are linked each other via Si waveguides [12]. The optical modulator was a Mach-Zehnder interferometer composed of phase shifters and multimode interference (MMI) couplers [40]. CW light from a laser diode was launched into the optical modulator. Differential RF input signals were pre-emphasized by a differentiator and input to the modulator. The voltage amplitude after pre-emphasis was 3.4 V peak to peak. The modulated optical signals were input to the PD array and converted into electrical signals which were then amplified by as RF amplifier. An optical image of a MSM PD pair in the circuit is shown in Fig. 9(b).
Fig. 9 (a) Schematic configuration of differential optical receiver circuit. (b) Optical image of MSM PD pair in optical circuit.
Figure 10 shows eye diagrams of PD output at 10 Gbps with a 27-1 pseudo-random binary sequence (PRBS). The eye diagram of differential PDs is shown along with that of individual PDs in each pair. The discrepancy in output amplitude between two individual PDs was due to a difference in the applied voltage because the bias applied to a phase shifter is varied to maximize the extinction ratio. Using differential PDs resulted in clear eye openings with eye height almost twice that of those when individual PDs were used. The sigma value of the amplitude jitter, on the other hand, did not show a significant increase. As a consequence, the differential optical circuit showed a higher signal-to-noise ratio (S/N). We consider that the high-uniformity of the MSM PD pairs effectively contributes to the S/N enhancement. We also showed that the highly distorted eye diagrams in the individual PDs were corrected in the differential PDs when the crosstalk between the modulators and the PDs was large. These results show that using MSM Ge PDs with a SiGe capping layer in a differential circuit effectively minimizes the effects of crosstalk. A faster data rate for future data rate performance would be achieved with these SiGe capped Ge MSM PDs if we shrink the metal spacing by utilizing advanced Si process.
Fig. 10 Eye diagrams of differential PD pairs in optical receiver circuits. Eye diagrams of individual PDs in each pair are also shown for comparison. The eye diagrams were taken at 7 V.
6. Summary
We have developed and tested waveguide integrated MSM (metal-semiconductor-metal) Ge photodetectors (PDs) with a SiGe capping layer for use in silicon photonics integration. The growth conditions for the capping layer were optimized, and a very smooth capping layer surface was obtained in a selective growth conditions. The obtained SiGe capping layer had a sufficiently high barrier height for both electrons and holes in an MSM Ge PD. A fabricated MSM Ge PD with the SiGe capping layer exhibited sufficiently low dark current (70 nA for a 7 × 30 μm2 contact area) and high responsivity (1.0 A/W). In addition, it exhibited a much narrower distribution of the dark current than an MSM Ge PD with a Si capping layer, indicating that a Ge PD with highly-uniform characteristics can be realized. Finally, highly-uniform MSM Ge PDs were used to fabricate a differential optical receiver circuit that exhibited both a higher signal-to-noise ratio and reduced signal distortion at 10 Gbps. A differential optical receiver circuit containing MSM Ge PDs with a SiGe capping layer effectively suppressed the crosstalk between the Si optical modulator and PDs.
Acknowledgment
This research is granted 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). Part of the fabrication was performed at TIA-SCR, AIST.
References and links
1. R. G. Beausoleil, P. J. Kuekes, G. S. Snider, S.-Y. Wang, and R. S. Williams, “Nanoelectronic and Nanophotonic Interconnect,” Proc. IEEE 96(2), 230–247 (2008). [CrossRef]
2. D. A. B. Miller, “Optical Interconnects to Silicon,” IEEE J. Sel. Top. Quantum Electron. 6(6), 1312–1317 (2000). [CrossRef]
3. I. A. Young, E. Mohammed, J. T. S. Liao, A. M. Kern, S. Palermo, B. A. Block, M. R. Reshotko, and P. L. D. Chang, “Optical I/O Technology for Tera-Scale Computing,” IEEE J. Solid-State Circuits 45(1), 235–248 (2010). [CrossRef]
4. G. Li, X. Zheng, J. Lexau, Y. Luo, H. Thacker, T. Pinguet, P. Dong, D. Feng, S. Liao, R. Shafiiha, M. Asghari, J. Yao, J. Shi, I. N. Shubin, D. Patil, F. Liu, K. Raj, R. Ho, J. E. Cunningham, and A. V. Krishnamoorthy, “Ultralow-power silicon photonic interconnect for high-performance computing systems,” Proc. SPIE 7607(1), 760703 (2010). [CrossRef]
5. A. Mekis, S. Gloeckner, G. Masini, A. Narasimha, T. Pinguet, S. Sahni, and P. De Dobbelaere, “A Grating-Coupler-Enabled CMOS Photonics Platform,” IEEE J. Sel. Top. Quantum Electron. 17(3), 597–608 (2011). [CrossRef]
6. X. Zheng, F. Y. Liu, J. Lexau, D. Patil, G. Li, Y. Luo, H. D. Thacker, I. Shubin, J. Yao, K. Raj, R. Ho, J. E. Cunningham, and A. V. Krishnamoorthy, “Ultralow Power 80 Gb/s Arrayed CMOS Silicon Photonic Transceivers for WDM Optical Links,” J. Lightwave Technol. 30(4), 641–650 (2012). [CrossRef]
7. X. Zheng, D. Patil, J. Lexau, F. Liu, G. Li, H. Thacker, Y. Luo, I. Shubin, J. Li, J. Yao, P. Dong, D. Feng, M. Asghari, T. Pinguet, A. Mekis, P. Amberg, M. Dayringer, J. Gainsley, H. F. Moghadam, E. Alon, K. Raj, R. Ho, J. E. Cunningham, and A. V. Krishnamoorthy, “Ultra-efficient 10 Gb/s hybrid integrated silicon photonic transmitter and receiver,” Opt. Express 19(6), 5172–5186 (2011). [CrossRef] [PubMed]
8. Y. Urino, T. Shimizu, M. Okano, N. Hatori, M. Ishizaka, T. Yamamoto, T. Baba, T. Akagawa, S. Akiyama, T. Usuki, D. Okamoto, M. Miura, M. Noguchi, J. Fujikata, D. Shimura, H. Okayama, T. Tsuchizawa, T. Watanabe, K. Yamada, S. Itabashi, E. Saito, T. Nakamura, and Y. Arakawa, “First demonstration of high density optical interconnects integrated with lasers, optical modulators, and photodetectors on single silicon substrate,” Opt. Express 19(26), B159–B165 (2011). [CrossRef] [PubMed]
9. C.-S. Li and H. S. Stone, “Differential Board/Backplane Optical Interconnects for High-speed Digital Systems Part I: Theory,” J. Lightwave Technol. 11(7), 1234–1249 (1993).
10. M. Aamer, A. Griol, A. Brimont, A. M. Gutierrez, P. Sanchis, and A. Håkansson, “Increased sensitivity through maximizing the extinction ratio of SOI delay-interferometer receiver for 10G DPSK,” Opt. Express 20(13), 14698–14704 (2012). [CrossRef] [PubMed]
11. B. Mikkelsen, C. Rasmussen, P. Mamyshev, and F. Liu, “Partial DPSK with excellent filter tolerance and OSNR sensitivity,” Electron. Lett. 42(23), 1363–1364 (2006). [CrossRef]
12. J. Fujikata, Y. Urino, S. Akiyama, T. Shimizu, N. Hatori, M. Okano, M. Ishizaka, T. Yamamoto, T. Baba, T. Akagawa, T. Usuki, D. Okamoto, M. Miura, M. Noguchi, D. Shimura, H. Okayama, T. Tsuchizawa, T. Watanabe, K. Yamada, S. Itabashi, E. Saito, K. Wada, T. Nakamura, and Y. Arakawa, “Differential signal transmission in silicon-photonics integrated circuit for high density optical interconnects,” Proc of 8th IEEE International Conference on Group IV Photonics (GFP) 365–367 (2011). [CrossRef]
13. K.-W. Ang, T.-Y. Liow, M.-B. Yu, Q. Fang, J. Song, G.-Q. Lo, and D.-L. Kwong, “Low Thermal Budget Monolithic Integration of Evanescent-Coupled Ge-on-SOI Photodetector on Si CMOS Platform,” IEEE J. Sel. Top. Quantum Electron. 16(1), 106–113 (2010). [CrossRef]
14. T. Yin, R. Cohen, M. M. Morse, G. Sarid, Y. Chetrit, D. Rubin, and M. J. Paniccia, “40Gb/s Ge-on-SOI waveguide photodetectors by selective Ge growth,” in Tech. Dig. Opt. Fiber Commun. Conf. (IEEE Photonics Society, 2008), paper OMK2. [CrossRef]
15. S. Liao, N.-N. Feng, D. Feng, P. Dong, R. Shafiiha, C.-C. Kung, H. Liang, W. Qian, Y. Liu, J. Fong, J. E. Cunningham, Y. Luo, and M. Asghari, “36 GHz submicron silicon waveguide germanium photodetector,” Opt. Express 19(11), 10967–10972 (2011). [CrossRef] [PubMed]
16. D. Feng, S. Liao, P. Dong, N.-N. Feng, H. Liang, D. Zheng, C.-C. Kung, J. Fong, R. Shafiiha, J. Cunningham, A. V. Krishnamoorthy, and M. Asghari, “High-speed Ge photodetector monolithically integrated with large cross-section silicon-on-insulator waveguide,” Appl. Phys. Lett. 95(26), 261105 (2009). [CrossRef]
17. L. Vivien, A. Polzer, D. Marris-Morini, J. Osmond, J.-M. Hartmann, P. Crozat, E. Cassan, C. Kopp, H. Zimmermann, and J. M. Fédéli, “Zero-bias 40Gbit/s germanium waveguide photodetector on silicon,” Opt. Express 20(2), 1096–1101 (2012). [CrossRef] [PubMed]
18. G. Li, Y. Luo, X. Zheng, G. Masini, A. Mekis, S. Sahni, H. Thacker, J. Yao, I. Shubin, K. Raj, J. E. Cunningham, and A. V. Krishnamoorthy, “Improving CMOS-compatible Germanium photodetectors,” Opt. Express 20(24), 26345–26350 (2012). [CrossRef] [PubMed]
19. L. Chen and M. Lipson, “Ultra-low capacitance and high speed germanium photodetectors on silicon,” Opt. Express 17(10), 7901–7906 (2009). [CrossRef] [PubMed]
20. S. Assefa, F. Xia, S. W. Bedell, Y. Zhang, T. Topuria, P. M. Rice, and Y. A. Vlasov, “CMOS-integrated high-speed MSM germanium waveguide photodetector,” Opt. Express 18(5), 4986–4999 (2010). [CrossRef] [PubMed]
21. J. Brouckaert, G. Roelkens, D. V. Thourhout, and R. Baets, “Compact InAlAs–InGaAs Metal–Semiconductor–Metal Photodetectors Integrated on Silicon-on-Insulator Waveguides,” IEEE Photon. Technol. Lett. 19(19), 1484–1486 (2007). [CrossRef]
22. J. D. Hwang and E. H. Zhang, “Effects of a a-Si:H layer on reducing the dark current of 1310 nm metal–germanium–metal photodetectors,” Thin Solid Films 519(11), 3819–3821 (2011). [CrossRef]
23. S. Assefa, F. Xia, and Y. A. Vlasov, “Reinventing germanium avalanche photodetector for nanophotonic on-chip optical interconnects,” Nature 464(7285), 80–84 (2010). [CrossRef] [PubMed]
24. A. Dimoulas, P. Tsipas, A. Sotiropoulos, and E. K. Evangelou, “Fermi-level pinning and charge neutrality level in germanium,” Appl. Phys. Lett. 89(25), 252110 (2006). [CrossRef]
25. J. Oh, S. K. Banerjee, J. C. Campbell, J. Oh, S. K. Banerjee, and J. C. Campbell, “Metal–Germanium–Metal Photodetectors on Heteroepitaxial Ge-on-Si With Amorphous Ge Schottky Barrier Enhancement Layers,” IEEE Photon. Technol. Lett. 16(2), 581–583 (2004). [CrossRef]
26. J. D. Hwang, Y. H. Chen, C. Y. Kung, and J. C. Liu, “High Photo-to-Dark-Current Ratio in SiGe/Si Schottky-Barrier Photodetectors by Using an a-Si:H Cap Layer,” IEEE Trans. Electron. Dev. 54(9), 2386–2391 (2007). [CrossRef]
27. C. O. Chui, A. K. Okyay, and K. C. Saraswat, “Effective Dark Current SuppressionWith Asymmetric MSM Photodetectors in Group IV Semiconductors,” IEEE Photon. Technol. Lett. 15(11), 1585–1587 (2003). [CrossRef]
28. J. D. Hwang, W. T. Chang, Y. H. Chen, C. Y. Kung, C. H. Hu, and P. S. Chen, “Suppressing the dark current of metal–semiconductor–metal SiGe/Si heterojunction photodetector by using asymmetric structure,” Thin Solid Films 515(7–8), 3837–3839 (2007). [CrossRef]
29. M. Takenaka, K. Morii, M. Sugiyama, Y. Nakano, and S. Takagi, “Dark current reduction of Ge photodetector by GeO₂ surface passivation and gas-phase doping,” Opt. Express 20(8), 8718–8725 (2012). [CrossRef] [PubMed]
30. J. Fujikata, M. Miura, M. Noguchi, D. Okamoto, T. Horikawa, and Y. Arakawa, “Si Waveguide-Integrated Metal–Semiconductor–Metal and p–i–n-Type Ge Photodiodes Using Si-Capping Layer,” Jpn. J. Appl. Phys. 52(4), 04CG10 (2013). [CrossRef]
31. N. Afshar-Hanaii, J. M. Bonar, A. G. R. Evans, G. J. Parker, C. M. K. Starbuck, and H. A. Kemhadjian, “Thick selective epitaxial growth of silicon at-960°C using silane only,” Microelectron. Eng. 18(3), 237–246 (1992). [CrossRef]
32. A. Ishitani, H. Kitajima, K. Tanno, H. Tsuya, N. Endo, N. Kasai, and Y. Kurogi, “Selective silicon epitaxial growth for device-isolation technology,” Microelectron. Eng. 4(1), 3–33 (1986). [CrossRef]
33. J. L. Regolini, D. Bensahel, J. Mercier, and E. Scheid, “Silicon selective epitaxial growth at reduced pressure and temperature,” J. Cryst. Growth 96(3), 505–512 (1989). [CrossRef]
34. J. Murota, N. Nakamura, M. Kato, N. Mikoshiba, and T. Ohmi, “Lowtemperature silicon selective deposition and epitaxy on silicon using the thermal decomposition of silane under ultraclean environment,” Appl. Phys. Lett. 54(11), 1007–1009 (1989). [CrossRef]
35. X. Tu, T.-Y. Liow, J. Song, M. Yu, and G. Q. Lo, “Fabrication of low loss and high speed silicon optical modulator using doping compensation method,” Opt. Express 19(19), 18029–18035 (2011). [CrossRef] [PubMed]
36. N. Ozguven and P. C. McIntyre, “Silicon-germanium interdiffusion in high-germanium-content epitaxial heterostructures,” Appl. Phys. Lett. 92(18), 181907 (2008). [CrossRef]
37. S. J. Chey and D. G. Cahill, “Relaxation of Nanometer-Scale Surface Morphology,” in Dynamics of Crystal Surfaces and Interfaces, ed. P.M. Duxbury, and T.J. Pence, (Springer Science + Business Media, New York, 1997).
38. S. M. Jang and R. Reif, “Effects of hydrogen and deposition pressure on Si1-xGex growth rate,” Appl. Phys. Lett. 60(6), 707–709 (1992). [CrossRef]
39. G. Masini, L. Colace, G. Assanto, H.-C. Luan, and L. C. Kimerling, “High-Performance p-i-n Ge on Si Photodetectors for the Near Infrared: From Model to Demonstration,” IEEE Trans. Electron. Dev. 48(6), 1092–1096 (2001). [CrossRef]
40. S. Akiyama, T. Baba, M. Imai, T. Akagawa, M. Takahashi, N. Hirayama, H. Takahashi, Y. Noguchi, H. Okayama, T. Horikawa, and T. Usuki, “12.5-Gb/s operation with 0.29-V•cm V(π)L using silicon Mach-Zehnder modulator based-on forward-biased pin diode,” Opt. Express 20(3), 2911–2923 (2012). [CrossRef] [PubMed]
