Plastic versus elastic strain relaxation in heteroepitaxy of InAs/GaAs(110)

Kenji Shiraishi,*1 Norihisa Oyama,2 Ko Okajima,3 Nori Miyagishima,3 Kyozaburo Takeda,3 HiroshiYamaguchi,1 Tomonori Ito,4 and Takahisa Ohno2

1NTT Basic Research Laboratories, Atsugi, Kanagawa 243-0198, Japan

2National Research Institute for Metals, Tsukuba, Ibaragi 305-0047, Japan

3Waseda University, Shinjuku, Tokyo 169-8555, Japan

4Mie University, Tsu, Mie 514-8507, Japan

*Phone: +81-46-240-3407, Fax: +81-46-240-4317, E-mail: siraisi@will.brl.ntt.co.jp

It has been a crucial issue to clarify the strain relaxation mechanism during heteroepitaxy. Although strain relaxation depends on factors such as surface energy and stoichiometry in atomic scale, several experiments indicate that the generation of misfit dislocations plays an important role in relaxing lattice strain, and also in determining the heteroepitaxial growth mode. In this study, we first formulate the free energy of the heterointerface system based on the elastic continuum theory to investigate the heteroepitaxial growth mode. It is found that our formulation can describe the heteroepitaxial growth mode when some physical parameters, such as formation energies of misfit dislocations and effective elastic constants, are determined appropriately. Moreover, these parameters are successfully determined by combining microscopic first-principles calculations and macroscopic phenomenological theory via total energy of the system. As a result, the formation energy of the 90o perfect misfit dislocation is revealed to be 0.96 eV/A in InAs/GaAs(110) systems. By combining the determined parameters, our theory also predicts that the critical thickness is 2.35 ML. In order to confirm our prediction, we also studied InAs growth on GaAs(110) substrate using scanning tunneling microscopy (STM). After 2.8 ML deposition, a 90o misfit dislocation began to be generated in the [001] direction. These results are in good agreement with the theoretical prediction, and justify our theoretical investigation [1,2].

  1. N. Oyama, E. Ohta, K. Takeda, K. Shiraishi, and H. Yamaguchi, J. Cryst. Growth 201/202, 256 (1999).
  2. K. Okajima, K. Takeda, N. Oyama, E. Ohta, K. Shiraishi, and T. Ohno, Jpn. J. Appl. Phys. 39, L917 (2000).