Improvement of crystallinity by post-annealing and regrowth of Ge layers on Si substrates

Thin Solid Films 550 (2014) 509–514

Contents lists available at ScienceDirect

Thin Solid Films journal homepage: www.elsevier.com/locate/tsf

Improvement of crystallinity by post-annealing and regrowth of Ge layers on Si substrates Katsuya Oda 1, Kazuki Tani, Shin-ichi Saito, Tatemi Ido Photonics Electronics Technology Research Association, Japan Institute for Photonics–Electronics Convergence System Technology, Japan Central Research Laboratory, Hitachi, Ltd., 1-280 Higashi-Koigakubo, Kokubunji, Tokyo 185-8601, Japan

a r t i c l e

i n f o

Article history: Received 4 February 2013 Received in revised form 21 October 2013 Accepted 21 October 2013 Available online 30 October 2013 Keywords: Germanium Light-emitting Epitaxial growth Post-annealing Regrowth Lattice strain Photoluminescence

a b s t r a c t Post-annealing and regrowth of Ge were investigated to improve the crystallinity and to control the lattice strain of Ge layers directly grown on a Si substrate by low-temperature epitaxial growth. The post-annealing at a higher temperature was an effective way of improving the surface morphology and the crystallinity of the Ge layers. Furthermore, the lattice strain changed from compressive to tensile in b 110N crystal orientation when the postannealing temperature was increased, and the tensile strain of 0.19% was achieved at the annealing temperature of 700 °C. Consequently, the photoluminescence (PL) intensity increased with the increasing post-annealing temperature and a red-shift of the PL spectra could be observed due to reduction of direct bandgap energy at Г-point with the tensile strain. Although regrowth of the Ge layers had little impact on the lattice strain at a relatively low regrowth temperature, a thick Ge layer with high crystallinity was formed at 700 °C and a favorable PL spectrum was obtained. These results indicate that this combined technique can improve the performance of Ge light-emitting devices. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Silicon photonics are attracting attention as a key technology for chip-to-chip or on-chip optical interconnections to overcome the density of transmitted data and power consumption in metal lines in largescale integrated circuits [1,2]. Optical devices such as light emitters [3–6], optical receivers [7,8], waveguides [9], modulators [10,11], and photo-multiplexers/de-multiplexers [12] can be stably fabricated by fine Si processes. Since Si and Ge, which are commonly used in standard Si processing, are indirect bandgap materials, the fabrication of the on-chip light source has been a great challenge. Various methods have been attempted for integrated light sources that can be fabricated with complementally metal-oxide-semiconductor logic circuits by using the quantum size effect [13], heterostructures [14], and other processes [15–17]. One of the most promising devices of these approaches involves germanium lasers for improving the quantum efficiencies with highly n-type doping and tensile strain in the Ge active regions [18,19]. However, since the lattice constant of Ge is 4.2% larger than that of Si, if a Ge layer is grown on a Si substrate, crystal defects such as dislocation are generated at the Ge/Si interface due to the strain relaxation. The Ge epitaxial growth on Si substrate without any thick buffer layers has recently been achieved for metal-oxide-

1

E-mail address: [email protected] (K. Oda). Tel.: +81 42 323 1111.

0040-6090/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.tsf.2013.10.136

semiconductor field effect transistors [20,21] and for photodetectors [7,8], and improved crystallinity in Ge layers has been studied using cyclic annealing [22], high temperature annealing [23], and low temperature buffer layers [24]. However, the threading dislocations and nonradiative recombination centers must be further decreased to improve luminous efficiency. Furthermore, although highly n-type doping has been recognized as an effective way of improving the direct transitions at the Г-point, due to filling the conduction band minimum with electrons at the L-point [18,19], the crystallinity of the Ge layers degrades when doping concentration is high. Thus, the doping concentration is limited to about the order of 1019 cm−3 with acceptable crystallinity and surface morphology [25]. Another attempt involves improving the quantum efficiency with the tensile strain to the Ge layer [26], and although it has been reported that the Ge layers contain a tensile strain after high temperature annealing due to the difference of thermal expansion coefficients [26–28], systematic studies have not yet been carried out. In this study, we investigated the effects of a combination of Ge epitaxial growth and post-annealing on the crystallinity and the lattice strain in order to improve the optical properties of the Ge layers. 2. Experimental details The Ge was epitaxially grown by using a cold-wall rapid thermal chemical vapor deposition system. Germane (GeH4) was used as a source gas, which was supplied with H2 carrier gas. An eight-inch Si

510

K. Oda et al. / Thin Solid Films 550 (2014) 509–514

(100) CZ wafer was used as the substrate, which was initially treated by RCA cleaning, HF wet etching of the SiO2 layer, and hydrogen termination of the wafer surface by rinsing in de-ionized water. Then, the wafer was annealed in a hydrogen atmosphere at 850 °C for 2 min. as a pre-cleaning process just before the Ge growth. It was confirmed that there were no contaminants such as oxygen and carbon on the wafer surface after the pre-cleaning process. As the starting point of improving the crystallinity and controlling the lattice strain, Ge layers with good surface morphology were grown at 420 °C under relatively higher pressure of 7000 Pa. Then, the Ge layers were annealed in the same H2 atmosphere to improve the crystallinity. Although heat-up rate for the annealing was controlled in about 5 °C/s, the wafer was simply cooled in H2 flow after the annealing, so the maximum cool-down rate of − 20 °C/s was obtained just after the end of the annealing. Surface morphology of the annealed Ge layers was evaluated by an atomic force microscope (AFM) [Veeco, Multimode in tapping mode with 1–10 Ω cm P-doped Si cantilever]. Crystallinity and lattice strain of the Ge layers were evaluated by high-resolution X-ray diffraction (XRD) analysis [Panalytical, X'Pert Pro with Bragg-Brentano geometry] and micro-Raman spectroscopy [Tokyo Instruments, Nanofinder]. We use a Cu Kα1 X-ray source for the XRD measurements. Moreover, the crystallinity and optical properties of the Ge layers were also investigated by micro-photoluminescence (PL) spectroscopy [Tokyo Instruments, Nanofinder]. Raman and PL spectroscopies were carried out with an Ar laser in which the pumping wavelength was 457.9 nm.

3. Results and discussions Fig. 1 shows a reciprocal space map (RSM) of XRD (XRD-RSM) from the 130-nm-thick Ge layer directly grown on the Si substrate. An AFM image of the Ge layer is also shown in the inset. A relatively good surface morphology of the Ge layer was achieved: the root mean square (RMS)

2μm

10 nm

Qy (rlu) 0

0.43

Si sub. (113) 0 0

2μm

0.42

0.41

Ge (113) 0.4

-0.2

-0.19

Qx (rlu)

Fig. 1. XRD-RSM of (-1-13) diffraction from as-grown Ge layer on Si substrate. The inset shows AFM image of Ge layer grown on Si substrate.

roughness was 1.02 nm for the measured area of 2 μm × 2 μm. In the XRD-RSM, an intense Si (-1-13) peak was observed, which represented the diffraction from the Si substrate under the Ge layer. Since the XRDRSM was measured by using semiconductor array detectors, errors in the counts occur if the diffraction intensity is very high; therefore, the streak line observed around the Si (-1-13) peak does not represent any actual diffraction. Although the thicknesses of the Ge layer was only 130 nm, a diffraction peak from Ge (-1-13) could be clearly observed, which means that the single crystalline Ge layer was obtained by using low-temperature epitaxial growth. Furthermore, since the position of the diffraction peak from the Ge layer was close to that from the unstrained Ge, it was confirmed that the lattice strain in the Ge layer originated from the lattice mismatch between Si and Ge was almost completely relaxed. However, the displacement of the diffraction peak shows that the as-grown Ge layer still contained a compressive strain just after the low temperature epitaxial growth at 420 °C due to the larger lattice constant of Ge compared to that of the Si substrate. 3.1. Effect of post-annealing Several approaches have been investigated to improve the crystallinity of Ge layers grown on the Si substrates [22–24], and it has been reported that cyclic annealing at a relatively higher temperature can reduce the threading dislocation density [22] in the Ge layers. This has led to studies on the effect of annealing on the crystallinity and the lattice strain of Ge layers. After the low temperature epitaxial growth of the Ge layers at 420 °C, the temperature was increased to the annealing temperature in the same H2 atmosphere as that during the epitaxial growth, and the Ge layers were then annealed at certain temperatures for 10 min. XRD-RSM and AFM images of the Ge layers annealed at various temperatures (Ta) after the low temperature epitaxial growth are shown in Fig. 2. Although the surface morphology of the Ge layer annealed at 500 °C (Fig. 2a) was almost the same as the as-grown Ge layer (Fig. 1), the short-range roughness had vanished and long-range undulations appeared after annealing at 600 °C (Fig. 2b). Moreover, some dimples were observed on the surface in addition to the long-range undulations after annealing at 700 °C (Fig. 2c). Despite having almost the same RMS value of the surface roughness, it was obvious that the short-range roughness could transform into the long-range undulations by the post-annealing at temperatures above 600 °C, which increased the area ratio of the relatively flat surface. As shown in the XRD-RSMs, the Ge (-1-13) diffraction peaks became steeper and the peak intensity increased when the annealing temperature was increased, indicating that the crystallinity of the Ge layers was increased by the post-annealing. Fig. 3 shows the lattice strain in the Ge layers in the b001N and b110N crystal orientations as a function of the annealing temperature. Since the growth time for the low temperature epitaxial growth was long enough relative to the time of post-annealing, the lattice strain of the Ge layer annealed at Ta = 420 °C was the same as that of the as-grown Ge layer. The lattice strain in the b 110N crystal orientation increased as Ta increased, and the strain in the b001N crystal orientation showed an opposite dependence. Although the Ge layer contained the compressive strain in the b110N crystal orientation at Ta = 420 °C, this strain started decreasing when Ta was increased, and the Ge was completely un-strained at Ta = 530 °C. Furthermore, the direction of the lattice strain changed from compressive to tensile after annealing at Ta N 530 °C, and the tensile strain at Ta = 700 °C reached 0.19%. This result is consistent with previous studies [29]. Normally, a grown layer with a larger lattice constant compared to a substrate contains a compressive strain within the growth plane. However, since the Ge layers grown on the Si substrate were almost completely relaxed even after low temperature growth, the Ge lattice could be dislocated at the Ge/Si interface by post-annealing, and the lattice strain of the Ge layer was relaxed during annealing at the relatively higher temperature with the volume determined by the thermal

K. Oda et al. / Thin Solid Films 550 (2014) 509–514

Qy (rlu)

10 nm

2μm

511

Qy (rlu)

(a)

10 nm

2μm

(b) 0

0.43

0

0.43

Si sub. (113)

Si sub. (113) 0

0 0

0

2μm

0.42

0.42

0.41

0.41

Ge (113) 0.4

Ge (113) 0.4

Ta = 500°C -0.2

Qy (rlu)

-0.19

2μm

Qx (rlu)

2μm

Ta = 600°C -0.2

-0.19

Qx (rlu)

10 nm

(c) 0

0.43

Si sub. (113)

0 0

2μm

0.42

0.41

Ge (113) Ta = 700°C

0.4

-0.2

-0.19

Qx (rlu)

Fig. 2. XRD-RSMs of (-1-13) diffraction from Ge layers after post-annealing at (a) 500 °C, (b) 600 °C, and (c) 700 °C. The insets show AFM images of each Ge layers after post-annealing.

expansion coefficients [30,31]. After annealing, the volume of the Ge layer and the Si substrate both shrunk as the temperature decreased, and there was barely any change to the lattice alignment at low

temperature. The volume of the Si substrate could be returned to its original value because it was thick enough. However, the volume of the Ge layer could not be returned due to its larger thermal expansion

512

K. Oda et al. / Thin Solid Films 550 (2014) 509–514

0.3

PL intensity (arb. units)

Lattice strain (%)

<110>

<001>

0.2

0.1

Tensile 0

-0.1

Peak wavelength (nm)

1600

1550

Ta = 700°C Bulk Ge

1500

650°C Compressive Tensile

1450 300 400 500 600 700 800 T (°C)

pump

600°C

= 457.8 nm

@RT Bulk Ge

-0.2

500°C 420°C (as-grown)

Compressive -0.3 300

400

550°C

500

600

700

1300

800

Fig. 3. Annealing temperature dependences of lattice strain in Ge layers grown on Si substrate along b001N and b110N crystal orientations. Dotted line indicates lattice strain in b110N crystal orientation calculated with difference between thermal expansion coefficients of Si and Ge.

coefficients compared with that of Si. Therefore, the tensile lattice strain remained only in the Ge layers after cooling [27]. Ideal lattice strain in b110N crystal orientation was also plotted in Fig. 3, which was calculated with only the difference of the thermal expansion coefficients between Si and Ge, so these values indicate the maximum lattice strain. Since, there are large discrepancies between calculation and measured values, it seems that relaxation ratio has a large effect on the lattice strain even at the lower temperatures. Raman spectra from the annealed Ge layers are shown in Fig. 4. The peak position of un-strained Ge was also shown in the figure. There are no obvious differences in the shape of the spectra, but the peak position shifted to small wave number as the annealing temperature increased. The inset of Fig. 4 shows the relation between the lattice strain derived from the XRD measurements and the peak position of Raman spectra.

1400

1500

1600

Wavelength (nm)

Ta (°C)

Fig. 5. Photoluminescence spectra from Ge layers annealed with different temperatures. The inset shows peak wavelength of photoluminescence from Ge layers as a function of annealing temperature.

The peak position was normalized as a difference from that corresponding to the un-strained Ge. Since the Raman shift reflects the energy levels of phonons, the peak position moved to large wave number with compressive strain in the Ge layer, and vice versa [22]. Since there is an obvious linear dependence between the Raman shift and the lattice strain derived from the XRD-RSM, micro-Raman measurements can be used for a quantitative analysis of local lattice strain in the uniform Ge layers. PL spectra from the post-annealed Ge layers with various annealing temperatures are shown in Fig. 5. Although Ge is an indirect bandgap material and the L-valley has the lowest energy level in the conduction 2μm

Qy (rlu)

10 nm

0

0.43

Si sub. (113)

0.3 Ta = 700°C

Lattice strain (%)

0.2

Ta = 700°C

0

650°C

0.1

0

550°C

0

Compressive

Tensile -0.1

500°C

0.42 as-grown

-0.2

Intensity (arb. units)

2μm

600°C

-0.3

-1 0 1 Variation of Raman shift (cm-1)

650°C 0.41

600°C 550°C 500°C

Bulk Ge 290

300

Ge (113)

420°C (as-grown) 310

320

330

0.4

Raman shift (cm-1) -0.2 Fig. 4. Raman spectra from Ge layers on Si substrate annealed at different temperatures after epitaxial growth. The inset shows relationship between lattice strain and variation of Raman shift from Ge layers after post-annealing.

-0.19

Qx (rlu)

Fig. 6. XRD-RSM of (-1-13) diffraction from regrown Ge layer at 600 °C on post-annealed Ge layer at 600 °C. The inset shows AFM image of regrown Ge layer at 600 °C.

K. Oda et al. / Thin Solid Films 550 (2014) 509–514

band, we were able to observe the recombination between electrons and holes at the Г-valley as the luminescence even from the bulk Ge at a wavelength of 1550 nm (dashed line in Fig. 5). A comparison with the post-annealed Ge layers shows that although the spectrum was very weak and broad for the as-grown Ge layer, obvious peaks could be observed from annealed samples at Ta N 600 °C. Moreover, the PL intensity increased and the peak shape became sharper as the annealing temperature was increased. Generally, PL spectrum is strongly affected by crystallinity, because non-radiative recombination was significantly increased with defects such as dislocation and stacking faults. Therefore, these results suggest that the crystallinity of the Ge layers was significantly improved by the post-annealing. The dependence of peak wavelength on Ta is shown in the inset of Fig. 5. The peak was observed at a shorter wavelength from the Ge layer annealed at 500 °C compared with that from bulk Ge, and a red shift of the PL peaks occurred after post-annealing at a higher temperature. In addition, the peak wavelength from the unstrained-Ge was 1550 nm, which is almost the same value as that from the bulk Ge. These results show that the bandgap energy at the Г-point was varied by the lattice strain in the Ge layer [32–34], which is consistent with the XRD and Raman spectroscopy measurements. These results indicate that, in the range of this study, the most favorable PL characteristic can be obtained from the Ge layer after post-annealing at higher temperature. However, the density of dimples increased and the undulations were emphasized when the annealing temperature was increased to over 750 °C, so we determined the optimum annealing temperature to be 700 °C due to improving the crystallinity and the effectively applied tensile strain. 3.2. Regrowth of Ge The crystallinity of an epitaxially grown layer significantly depends on the quality and cleanliness of the growth surface. We therefore investigated the regrowth of Ge after improving the crystallinity of the Qy (rlu)

10 nm

2μm

513

Ge base layers by a combination of the low temperature epitaxial growth and the post-annealing at higher temperature. Since there is a concern about degradation of the surface morphology on the regrown Ge layer at high growth temperature, the measurements of surface roughness and crystallinity of the Ge layers were done by AFM and XRD, respectively, at various regrowth temperatures (Tr). Fig. 6 shows a XRD-RSM and a AFM image of the Ge layer with the regrowth of the 230-nm-thick Ge layer at Tr = 600 °C on the post-annealed 150-nmthick Ge layer at Ta = 600 °C. The surface roughness was reduced and the surface morphology was sufficiently improved by the Ge regrowth compared with the post-annealed Ge layer without regrowth shown in Fig. 2b. The RMS roughness of the regrown Ge layer was 0.58 nm for the measured area of 2 μm × 2 μm. In the XRD-RSM, although the tensile strain in b110N crystal orientation was slightly decreased to 0.05% after regrowth, the shape of the Ge (-1-13) diffraction peaks is quite similar for both samples, and the intensity of the peak from the regrown Ge layer was four times larger than that corresponding to the sample without the regrowth. Furthermore, the FWHM values of the diffraction peaks were 77% and 69% decreased after regrowth in the b110N and b001N crystal orientations, respectively. Although the thickness of the total Ge layer increased at 2.4 times larger than that of the Ge layer without the regrowth, the diffraction intensity increased at 4.4 times larger, so it seems that the crystallinity of the regrown Ge layer was sufficiently improved. The regrowth was tried after post-annealing at a higher temperature, Ta = 700 °C. Fig. 7 shows the XRD-RSM and AFM images of the Ge layers with the Ge regrowth at Tr = 600 °C or 700 °C grown on the post-annealed 160-nm-thick Ge layer at Ta = 700 °C. Thickness of the regrown Ge layers at Tr = 600 and 700 °C was 230 and 200 nm, respectively. The surface morphology was improved by the Ge regrowth, and the RMS roughness of the regrown Ge layers at Tr = 600 °C and 700 °C was 0.79 and 0.82 nm, respectively. As compared with Fig. 2c, the shapes of the Ge (-1-13) diffraction peaks were similar and the tensile strain in b110N crystal orientation was almost the same for these 10 nm

2μm

Qy (rlu)

(b)

(a)

0

0

0.43

0.43

Si sub. (113)

Si sub. (113)

0 0

0 0

2μm

0.42

0.42

0.41

0.41

Ge (113)

2μm

Ge (113)

0.4

0.4

-0.2

-0.19

Qx (rlu)

-0.2

-0.19

Qx (rlu)

Fig. 7. XRD-RSMs of (-1-13) diffraction from regrown Ge layers at (a) 600 °C, and (b) 700 °C on Ge layer annealed at 700 °C. The insets show AFM images of regrown Ge layers.

514

K. Oda et al. / Thin Solid Films 550 (2014) 509–514

Acknowledgments

PL intensity (arb. units)

Ta = 700°C λpump = 457.8 nm @RT

Tr = 700°C

References

600°C

1200

We thank Prof. Y. Arakawa, Prof. S. Iwamoto, and Dr. S. Kako for enlightening discussions. 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).

1300

As-annealed

1400

1500

1600

Wave length (nm) Fig. 8. Photoluminescence spectra from Ge layers annealed at 700 °C and regrowth of Ge at 600 °C or 700 °C.

samples. Moreover, the intensities of the diffraction peaks from the Ge layers regrown at Tr = 600 °C and 700 °C were 2.4 and 3 times larger, respectively, than that corresponding to the sample without the regrowth. Additionally, the FWHM values of the diffraction peaks decreased to 92% and 80% in the b 110N crystal orientation and to 79% and 78% in the b 001N crystal orientation, respectively. The PL spectra from these samples are shown in Fig. 8. There are no obvious differences in the peak wavelength due to the similar lattice strain. However, the peak intensity from the Ge layer regrown at Tr = 700 °C was higher than that corresponding to Tr = 600 °C. Moreover, the peak shape of the PL spectrum from the Ge layer regrown at Tr = 700 °C was slightly steeper than the others. These results indicate that the lattice strain of the Ge layers was determined by the thermal process with the highest temperature after the low temperature epitaxial growth, and that the crystallinity was increased as the regrowth temperature increased. 4. Conclusion In order to realize monolithic integrated Ge lasers suitable for advanced Si processes, we investigated how to improve the crystallinity and tensile strain of the Ge layers by a combination of lowtemperature epitaxial growth, post-annealing, and Ge regrowth. Increasing the temperature for the post-annealing was effective for improving the surface morphology and the crystallinity of the Ge layers. Furthermore, the lattice strain changed from compressive (−0.19%) to tensile (0.19%) in the b 110N crystal orientation. The PL intensity thus increased when the post-annealing temperature increased and the red-shift of the PL spectra could be observed due to reducing the direct bandgap energy at the Г-point by increasing the tensile strain. The optimum post-annealing temperature was 700 °C from the viewpoint of the crystallinity and the lattice strain. The regrowth of Ge layers also improved the crystallinity, and the favorable PL spectrum was therefore obtained from the regrown Ge layer at 700 °C. These results indicate that the proposed method can improve the performances of Ge-based light-emitting devices for high-speed optical communications.

[1] B. Jalali, S. Fathpour, J. Lightwave Technol. 24 (2006) 4600. [2] R. Soref, Silicon, vol. 2, Springer, 2010. 1. [3] S. Saito, T. Takahama, K. Tani, M. Takahashi, T. Mine, Y. Suwa, D. Hisamoto, Appl. Phys. Lett. 98 (2011) 261104. [4] H. Rong, A. Liu, R. Jones, O. Cohen, D. Hak, R. Nicolaescu, A. Fang, M. Paniccia, Nature 433 (2005) 292. [5] D. Liang, A.W. Fang, H.-W. Chen, M.N. Sysak, B.R. Koch, E. Lively, O. Raday, Y.-H. Kuo, R. Jones, J.E. Bowers, Appl. Phys. A 95 (2009) 1045. [6] R.E. Camacho-Aguilera, Y. Cai, N. Patel, J.T. Bessette, M. Romagnoli, L. Kimerling, J. Michel, Opt. Express 20 (2012) 11316. [7] L. Vivien, J. Osmond, J.-M. Fédéli, D. Marris-Morini, Paul Crozat, J.-F. Damlencourt, E. Cassan, Y. Lecunff, S. Lava, Opt. Express 17 (2009) 6252. [8] T. Yin, R. Cohen, M.M. Morse, G. Sarid, Y. Chetrit, D. Rubin, M.J. Paniccia, Opt. Express 15 (2007) 13965. [9] K. Yamada, T. Shoji, T. Tsuchizawa, T. Watanabe, J. Takahashi, S. Itabashi, Opt. Lett. 28 (2003) 1663. [10] W.M.J. Green, M.J. Rooks, L. Sekaric, Y.A. Vlasov, Opt. Express 15 (2007) 17106. [11] S. Akiyama, T. Baba, M. Imai, T. Akagawa, M. Takahashi, N. Hirayama, H. Takahashi, Y. Noguchi, Hideaki Okayama, T. Horikawa, T. Usuki, Opt. Express 20 (2012) 2911. [12] S. Park, K.-J. Kim, I.-G. Kim, G. Kim, Opt. Express 19 (2011) 13531. [13] R. Apetz, L. Vescan, A. Hartmann, C. Dieker, H. Lüth, Appl. Phys. Lett. 66 (1995) 445. [14] J.C. Sturm, H. Manoharan, L.C. Lenchyshyn, M.L.W. Thewalt, N.L. Rowell, J.-P. Noël, D.C. Houghton, Phys. Rev. Lett. 66 (1991) 1362. [15] A.G. Cullis, L.T. Canham, Nature 353 (1991) 335. [16] B. Zheng, J. Michel, F.Y.G. Ren, L.C. Kimerling, D.C. Jacobson, J.M. Poate, Appl. Phys. Lett. 64 (1994) 2842. [17] D. Leong, J. Harry, K.J. Reeson, K.P. Homewood, Nature 387 (1997) 686. [18] J. Liu, X. Sun, D. Pan, X. Wang, L.C. Kimerling, T.L. Koch, J. Michel, Opt. Express 15 (2007) 11272. [19] X. Sun, J. Liu, L.C. Kimerling, J. Michel, Appl. Phys. Lett. 95 (2009) 011911. [20] J. Mitard, B. De Jaeger, F.E. Leys, G. Hellings, K. Martens, G. Eneman, D.P. Brunco, R. Loo, J.C. Lin, D. Shamiryan, T. Vandeweyer, G. Winderickx, E. Vrancken, C.H. Yu, K. De Meyer, M. Caymax, L. Pantisano, J. Meuris, M.J. Heyns, IEDM Tech. Dig. (2008) 873. [21] M. Halbwax, D. Bouchier, V. Yam, D. Débarre, L.H. Nguyen, Y. Zheng, P. Rosner, M. Benamara, H.P. Strunk, C. Clerc, J. Appl. Phys. 97 (2005) 064907. [22] H.-C. Luan, D.R. Lim, K.L. Lee, K.M. Chen, J.G. Sandland, K. Wada, L.C. Kimerling, Appl. Phys. Lett. 75 (1999) 2909. [23] V. Terzieva, L. Souriau, M. Caymax, D.P. Brunco, A. Moussa, S. Van Elshocht, R. Loo, F. Clemente, A. Satta, M. Meuris, Thin Solid Films 517 (2008) 172. [24] Z. Zhiwen, C. Li, H. Lai, S. Chen, J. Yu, J. Cryst. Growth 310 (2008) 2508. [25] G.D. Dilliway, R. van den Boom, B. van Daele, F.E. Leys, T. Clarysse, B. Parmentier, A. Moussa, C. Defranoux, A. Benedetti, O. Richard, H. Bender, E. Simoen, M. Meuris, ECS Trans. 3 (2006) 599. [26] J. Liu, D.D. Cannon, K. Wada, Y. Ishikawa, D.T. Danielson, S. Jongthammanurak, J. Michel, L.C. Kimerling, Phys. Rev. B 70 (2004) 155309. [27] Y. Ishikawa, K. Wada, J. Liu, D.D. Cannon, H.-C. Luan, J. Michel, L.C. Kimerling, J. Appl. Phys. 98 (2005) 013501. [28] M. El Kurdi, H. Bertin, E. Martincic, M. de Kersauson, G. Fishman, S. Sauvage, A. Bosseboeuf, P. Boucaud, Appl. Phys. Lett. 96 (2010) 041909. [29] D.D. Cannon, J. Liu, Y. Ishikawa, K. Wada, D.T. Danielson, S. Jongthammanurak, J. Michel, L.C. Kimerling, Appl. Phys. Lett. 84 (2004) 906. [30] H.P. Singh, Acta Crystallogr. Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr. 24 (1968) 469. [31] Y. Okada, Y. Tokumaru, J. Appl. Phys. 56 (1984) 314. [32] F. Cerdeira, C.J. Buchenauer, Fred H. Pollak, M. Cardona, Phys. Rev. B 5 (1972) 580. [33] C.G. Van de Walle, Phys. Rev. B 39 (1989) 1871. [34] C.G. Van de Walle, R.M. Martin, Phys. Rev. B 34 (1986) 5621.