Control of polymorphism during epitaxial growth of hyperferroelectric candidate LiZnSb on GaSb (111)B

Abstract

A major challenge for ferroelectric devices is the depolarization field, which competes with and often destroys long-range polar order in the limit of ultrathin films. Recent theoretical predictions suggest a new class of materials, termed hyperferroelectics, that should be robust against the depolarization field and enable ferroelectricity down to the monolayer limit. Here we demonstrate the epitaxial growth of hexagonal LiZnSb, one of the hyperferroelectric candidate materials, by molecular-beam epitaxy on GaSb (111)B substrates. Due to the high volatility of all three atomic species, we find that LiZnSb can be grown in an adsorption-controlled window, using an excess zinc flux. Within this window, the desired polar hexagonal phase is stabilized with respect to a competing cubic polymorph, as revealed by X-ray diffraction and transmission electron microscopy measurements. First-principles calculations show that for moderate amounts of epitaxial strain and moderate concentrations of Li vacancies, the cubic LiZnSb phase is lower in formation energy than the hexagonal phase, but only by a few meV per formula unit. Therefore we suggest that kinetics plays a role in stabilizing the desired hexagonal phase at low temperatures. Our results provide a path towards experimentally demonstrating ferroelectricity and hyperferroelectricity in a new class of ternary intermetallic compounds.

On the microscopic mechanisms behind hyperferroelectricity

“…unscreened depolarization field under open circuit boundary conditions. Such features thus make hyperFE systems suitable for applications as low dimensional functional materials [3][4][5][6][7]; moreover, the existence of a switchable electric polarization in hyperFEs can be allowed in metals and not restricted to only insulators and semiconductors as for standard FEs [4,8,9].…”

Semi-adsorption-controlled growth window for half Heusler FeVSb epitaxial films

“…It remains an outstanding challenge to control the Fe/V stoichiometry, which is not self-limited when using atomic Fe and V fluxes from effusion cells. Re-cent demonstrations MBE-grown LiZnSb, in which all three atomic species are volatile, suggest that it may be possible to control the full stoichiometry of a ternary Heusler compound [42]. However, for transition metal based Heuslers such as FeVSb, control of the X/Y transition metal ratio may require replacing one or both of the elemental transition metal sources with a volatile metalorganic precursor.…”