Structure of levitated Si-Ge melts studied by high-energy x-ray diffraction in combination with reverse Monte Carlo simulations

Irina Pozdnyakova, Oleksandr Roik, James W E Drewitt , Aleksei Bytchkov, Florian Kargl, Sandro Jahn, Séverine Brassamin, Louis Hennet*

*Corresponding author for this work

Research output: Contribution to journalArticle (Academic Journal)peer-review

3 Citations (Scopus)
27 Downloads (Pure)

Abstract

The short-range order in liquid Si, Ge and binary Six-Ge1-x alloys (x = 0.25, 0.50, 0.75) was studied by x-ray diffraction and reverse Monte Carlo simulations. Experiments were performed in the normal and supercooled liquid states by using the containerless technique of aerodynamic levitation with CO2 laser heating, enabling deeper supercooling of liquid Si and Si-Ge alloys than that previously reported. The local atomic structure of liquid Si and Ge is found to resemble the β-tin structure. The first coordination numbers of about 6 for all compositions are found to be independent of temperature indicating the supercooled liquids studied retain this high-density liquid (HDL) structure. However, there is evidence of developing local tetrahedral ordering, as manifested by a shoulder on the right side of the first peak in S(Q) which becomes more prominent with increasing supercooling. This result is potentially indicative of a continuous transition from the stable HDL β-tin (high pressure) phase, towards a metastable low-density liquid (LDL) phase, reminiscent of the diamond (ambient pressure) structure.
Original languageEnglish
Article number244002
Number of pages11
JournalJournal of Physics Condensed Matter
Volume33
Issue number24
DOIs
Publication statusPublished - 14 May 2021

Bibliographical note

Publisher Copyright:
© 2021 IOP Publishing Ltd.

Keywords

  • aerodynamic levitation
  • SiGe melts
  • liquid structure
  • synchrotron x-ray diffraction
  • RMC

Fingerprint

Dive into the research topics of 'Structure of levitated Si-Ge melts studied by high-energy x-ray diffraction in combination with reverse Monte Carlo simulations'. Together they form a unique fingerprint.

Cite this