On saw-tooth variation of residual strain in Sn particles epitaxially grown on SnTe

Abstract

Vincent reported that Sn particles formed on SnTe above 140 °C show a saw-tooth dependece of the residual strain on the particle width, and Vincent and Jesser and van der Merwe interpreted the dependence as due to one-by-one introduction of misfit dislocations along the interface between growing “solid” Sn particles and the substrate. This paper reexamines the report and the interpretation. Electron microscopy of in situ growth reveals that the Sn particles nucleate and grow in liquid phase above 130 °C, in solid phase below 100 °C, and nucleate in liquid phase and solidify in the course of growth in the intermediate temperature range. Residual strains in the particles of different widths measured similarly to Vincent do not show a systematic saw-tooth like dependence on particle width in all the temperature ranges. The residual strain observed in situ for one and the same “growing” particle in the lowest temperature range, on the other hand, shows a certain saw-tooth variation. The moiré fringes in the (400) weak beam image, however, do not show the alternation of high and low peak contrast of the fringes which should appear if misfit dislocations did exist along the interface. With use of a linear atomic chain model for the growing particle, the observed saw-tooth variation is shown to be due not to the formation of misfit dislocations, but to the transitions from a series of even stable states to the next series of odd stable states, and so on. We mean by “odd” and “even” that the arrangements of atoms in the chain are symmetric around the central atom located at the bottom and at the top of the substrate potential, respectively. The effects of growth kinetics, which prevent the particles from faithfully pursuing the stable states necessary for the saw-tooth variation to be seen, are discussed in conjunction with observed details such as a notable difference in particle shapes depending on the formation temperatures.