Langmuir evaporation of GaAs(1 1 1)A and GaAs(1 1 1)B: Monte Carlo simulation

A. A. Spirina, V. L. Alperovich, N. L. Shwartz

Research output: Contribution to journalArticlepeer-review

1 Citation (Scopus)

Abstract

The Langmuir evaporation (LE) of GaAs vicinal surfaces with various crystallographic orientations is analyzed at the atomistic level by Monte Carlo simulation. The evolution of the surface morphology and LE characteristics (gallium and arsenic evaporation rates, congruent evaporation temperature Tc, effective activation energy of Ga desorption) are studied under the increasing annealing temperature for GaAs(1 1 1)A and GaAs(1 1 1)B singular, and vicinal surfaces with various atomic step density. At low temperatures T < Tc, the step-terraced morphology of a vicinal surface is preserved in the LE process. At high temperatures T > Tc, the characteristics and crystallographic orientation specificity of surface morphology are determined by the formation of liquid Ga droplets. It is shown by simulation that, as compared to the (1 1 1)B surface, on the (1 1 1)A surface, the congruent evaporation temperature is substantially higher and is more sensitive to the atomic step density, as well as to the concentration of intentionally introduced defects on the terraces of the initial step-terraced surface. These observations are explained by two morphological peculiarities of LE on (1 1 1)A and B surfaces: (i) on the (1 1 1)A surface, Ga droplets penetrate deep into the substrate; (ii) on the (1 1 1)B surface, vacancy islands are more readily formed on terraces.

Original languageEnglish
Article number148281
Number of pages6
JournalApplied Surface Science
Volume540
DOIs
Publication statusPublished - 28 Feb 2021

Keywords

  • Atomic step
  • Ga droplet
  • GaAs
  • Langmuir evaporation
  • Monte Carlo simulation
  • Surface
  • GA-AS
  • DROPLET EPITAXY
  • VAPOR-PRESSURES
  • NANOSTRUCTURES
  • GALLIUM

Fingerprint Dive into the research topics of 'Langmuir evaporation of GaAs(1 1 1)A and GaAs(1 1 1)B: Monte Carlo simulation'. Together they form a unique fingerprint.

Cite this