SiGe technology has lately attracted much interest due to a unique combination of outstanding high-frequency characteristics of Si/SiGe heterostructures with their compatibility with well-established Si and Complementary Metal Oxide Semiconductor (CMOS) technologies. Rising of SiGe growth rate and benefits of using the commercially available equipment drive attention of process developers to high pressure and temperature ranges, up to atmospheric pressure and 1000 ^oC. Simulation and optimization of such processes require development of versatile models of SiGe epitaxial growth, valid in a wide range of operating conditions.

We have developed a predictive model of SiGe epitaxial growth from SiH₄-GeH₄-H₂ accounting for different factors affecting the growth, including adsorption and desorption of the reactive species, surface diffusion of hydrogen adatoms, Ge surface segregation, and elastic strains in the epilayer. The gaseous species are assumed to decompose at the surface into separate atoms, H atoms covering the surface and blocking free adsorption sites while Si and Ge atoms quickly incorporating into the crystal. Fig.1 illustrates the surface kinetic mechanism we employ in the model. Note that there are many unrelated models of separate surface processes in the literature (see, for example, [1-5]), however, they do not seem to have ever been linked within a unified approach. We have unified these models using a general quasi-equilibrium approach.

Fig.1. Kinetic mechanism of SiGe epitaxial growth
from SiH₄-GeH₄-H₂

We started with a steady version of the model for growth of pure Si or Ge from SiH₄-H₂ or GeH4-H2, respectively, and then extended it to SiGe alloy. The model was carefully verified using a wide range of experimental data available in the literature. The computations reproduce major features of SiGe epitaxial growth and allow understanding of physical mechanisms responsible for the effects observed. The model verification for growth of pure Si and Ge is presented in Fig. 2. Fig. 2a shows good agreement of the computed and experimental dependencies of the growth rate on temperature for growth of pure Si or Ge (data of Li et al [6]). Agreement of the computed and experimental data on the surface coverage with H adatoms in growth of pure Si is displayed in Fig. 2b (data of Liehr et al [7]).

Fig.2. Steady growth of pure Si or Ge from SiH₄-H₂ or GeH₄-H₂, respectively:
growth rates vs. temperature compared to data of Li et al [6] (a)
and coverage of Si surface with H-adatoms vs. temperature
compared to data of Liehr et al [7] (b)

Extension of the model to growth of SiGe alloy is illustrated in Fig.3. The computed growth rate as a function of the Ge fraction in the crystal is presented in Figs. 3 a and b in comparison with the data of Robbins et al [8] and Bozzo et al [9], respectively. It is seen that the computations reproduce the observed non-monotonic behavior of the growth rate with the Ge fraction. To illustrate the effects of different physical factors on the growth rate, Fig. 2b shows also the growth rate vs. the Ge fraction computed in successive "switching off" of some factors. Surface segregation is seen to exert the major effect on the growth rate and to be responsible for the non-monotonic behavior. Good agreement of the computed and experimental dependencies of the Ge fraction in the crystal on that in the gas phase is presented in Fig. 3c (data of Bozzo et al [9]).

Fig.3. Steady growth of SiGe alloy from SiH₄-GeH₄-H₂:
growth rate vs. Ge content in the alloy
compared to data of Robbins et al [8] (a) a Bozzo et al (b)
and Ge content in the alloy vs. that in the gas phase
compared to data of Bozzo et al [9] (c)

Currently, we are extending the model of SiGe growth from SiH₄-GeH₄-H₂ to unsteady processes to describe formation of the component profiles in Si/SiGe heterostructures. With the current version of the unsteady model, we have successfully described the Ge profile in a Si capping layer. The corresponding computed and experimental profiles are compared in Fig. 4 (data of Tok et al [4]). The effects of different physical factors on the Ge profile are illustrated in Fig. 4b where the profiles are again computed in successive "switching off" of some factors.

Fig.4. Unsteady growth of SiGe alloy from SiH₄-GeH₄-H₂:
Ge proviles in the Si capping layer compared to data of Tok et al [4] (a)
and Ge profiles computed in "switching of" of some physical factors (b)

References

1. H. Kim, N. Taylor, T.R. Bramblett, and J.E. Greene, J. Appl. Phys., 84, 6372 (1998).
2. S. Fukatsu, K. Fujita, H. Yaguchi, Y. Shiraki, and R. Ito, Surf. Sci., 267, 79 (1992).
3. N.M. Russell and W.G. Breiland, J. Appl. Phys., 73, 3525 (1992).
4. E.S.Tok, N.J. Woods, and J. Zhang, J. Cryst. Growth, 209, 321 (2000).
5. A. Fisher, H.-J. Osten, and H. Richter, Solid-State Electronics, 44, 869 (2000).
6. C. Li, S. John, E. Quinones, and S. Banerjee, J. Vac. Sci. Technol., A 14, 170 (1996).
7. M. Liehr, C.M. Greenlief, S.R. Kasi, and M. Offenberg, Appl. Phys. Lett. 56, 629 (1990).
8. D.J. Robbins, J.L. Glasper, A.G. Gullis, and W.Y. Leong, J. Appl. Phys., 69, 3729 (1991).
9. S. Bozzo, J.-L. Lazzari, C. Coudreau, A. Ronda, F. Arnaud d'Avitaya, J. Derrien, S. Mesters, B. Hollaender, P. Gergaud, and O. Thomas, J.Cryst.Growth, 216, 171 (2000).

Our publications on SiGe

1. A.S. Segal, S.Yu. Karpov, A.P. Sid'ko, and Yu.N. Makarov, Journal of Crystal Growth, 225, 268 (2001).
2. A.S. Segal, A.P. Sid'ko, S.Yu. Karpov, and Yu.N. Makarov, in "Fundamental Gas-Phase and Surface Chemistry of Vapor Deposition II/ Process Control, Diagnostics and Modeling in Semiconductor Manufacturing", ed. M.D. Allendorf, M.T. Swihart, and M. Meyyappan, Electrochemical Society Proceedings, 2001-13, 229 (2001).
3. A.S. Segal, A.P. Sid'ko, S.Yu. Karpov, Yu.N. Makarov, in "Semiconductor Silicon 2002 (9th International Symposium)", ed. M.T. Swihart, M.D. Allendorf, and M. Meyyappan, Electrochemical Society Proceedings, 2002-2, 567 (2002).