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

Estiaque H. Shourov, Ryan Jacobs, Wyatt A. Behn, Zachary J. Krebs, Chenyu Zhang, Patrick J. Strohbeen, Dongxue Du, Paul M. Voyles, Victor W. Brar, Dane D. Morgan, and Jason K. Kawasaki

Phys. Rev. Materials 4, 073401 – Published 1 July 2020

Abstract

The electronic, magnetic, thermoelectric, and topological properties of Heusler compounds (composition $X Y Z$ or $X_{2} Y Z$ ) are highly sensitive to stoichiometry and defects. Here we establish the existence and experimentally map the bounds of a semi-adsorption-controlled growth window for semiconducting half-Heusler FeVSb films, grown by molecular beam epitaxy (MBE). We show that due to the high volatility of Sb, the Sb stoichiometry is self-limiting for a finite range of growth temperatures and Sb fluxes, similar to the growth of III-V semiconductors such as GaSb and GaAs. Films grown within this window are nearly structurally indistinguishable by x-ray diffraction (XRD) and reflection high energy electron diffraction (RHEED). The highest electron mobility and lowest background carrier density are obtained towards the Sb-rich bound of the window, suggesting that Sb vacancies may be a common defect. Similar semi-adsorption-controlled bounds are expected for other ternary intermetallics that contain a volatile species $Z =$ {Sb, As, Bi}, e.g., CoTiSb, LuPtSb, GdPtBi, and NiMnSb. However, outstanding challenges remain in controlling the remaining Fe/V ( $X / Y$ ) transition metal stoichiometry.

Received 12 March 2020
Revised 19 April 2020
Accepted 16 June 2020

DOI:https://doi.org/10.1103/PhysRevMaterials.4.073401

Physics Subject Headings (PhySH)

Heusler alloy

Density functional theory Molecular beam epitaxy Photoemission spectroscopy Scanning transmission electron microscopy Scanning tunneling microscopy

Condensed Matter, Materials & Applied Physics

Authors & Affiliations

Estiaque H. Shourov ¹, Ryan Jacobs¹, Wyatt A. Behn², Zachary J. Krebs², Chenyu Zhang¹, Patrick J. Strohbeen¹, Dongxue Du¹, Paul M. Voyles¹, Victor W. Brar², Dane D. Morgan¹, and Jason K. Kawasaki ^1,*

¹Materials Science and Engineering, University of Wisconsin–Madison, Madison, Wisconsin 53706, USA
²Department of Physics, University of Wisconsin–Madison, Madison, Wisconsin 53706, USA

^*jkawasaki@wisc.edu

Article Text (Subscription Required)

Click to Expand

Supplemental Material (Subscription Required)

Click to Expand

References (Subscription Required)

Click to Expand

Issue

Vol. 4, Iss. 7 — July 2020

Reuse & Permissions

Access Options

Author publication services for translation and copyediting assistance advertisement

Images

Figure 1
Bulk and surface structure of FeVSb (001) films. (a) Cross-sectional HAADF-STEM image of the FeVSb/MgO interface. Crystallographic directions are referenced to the FeVSb unit cell which is outlined in red. Here, individual columns of Sb, Fe, and V atoms appear as a merged bright spot elongated along [001]. Inset: Structural model of FeVSb/MgO (001), with the unit cell outlined in black. (b) Empty states STM image (500 mV sample bias, 3 nA tunnel current) of the FeVSb surface showing a $(2 \times 1)$ surface reconstruction. The surface unit cell is indicated by the white and black rectangles on the STM image and the model crystal structure, respectively. The empty states contrast is expected to correspond to partially filled Sb dangling bonds or partially filled V $3 d$ orbitals [1]. These samples were grown at a temperature of $\approx 450^{\circ} C$ and Sb/V flux ratio of 2.2. Inset: Model of the surface reconstruction, characterized by Sb-Sb dimerization. (c) Sb $4 d$ core level evolution as a function of photon energy, showing evidence for surface Sb-Sb dimerization. The estimated photoelectron mean free path $λ$ is derived from the universal curve [2]. Shaded curves show a Voigt fit to the $h ν = 500$ eV data.
Reuse & Permissions
Figure 2
Thermodynamics of FeVSb and GaSb adsorption-controlled growth. (a),(b) Ellingham diagrams for GaSb and FeVSb. In both plots, the upper curves (blue) represent the change in Gibbs free energy for antimony sublimation Sb(s) $\Leftrightarrow \frac{1}{4} {Sb}_{4}$ (g). The lower curves (red) are the change in free energies for decomposition of solid GaSb or FeVSb, respectively. The shaded regions bounded by Sb sublimation and FeVSb (GaSb) decomposition define the growth window, in which solid FeVSb or GaSb are in equilibrium with antimony vapor, hence the Sb stoichiometry of the solid is self-limited. See text for further details. (c) Crystal structures for FeV and FeVSb, with an expected $2 a_{B 2} \approx a_{h H}$ epitaxial relationship.
Reuse & Permissions
Figure 3
Growth window determination by in-situ surface diffraction. RHEED patterns for beam oriented along the [110] azimuth of FeVSb films grown at $500^{\circ} C$ (bottom) and at $560^{\circ} C$ (top) with varying Sb/V flux. Red arrows mark the bulk $1 \times$ reflections. The Sb deficient region corresponds to mixed phases of FeVSb (half-Heusler) and FeV (B2), which are epitaxial with one another. The Sb rich region corresponds to excess Sb islands formed on the FeVSb surface. Within the growth window, the RHEED patterns exhibit a streaky ( $2 \times 1$ ) pattern characteristic of half-Heusler surfaces.
Reuse & Permissions
Figure 4
Structural growth window. Left panels: $θ – 2 θ$ x-ray diffraction scans (Cu $K α$ ) for samples grown at (a) $560^{\circ} C$ and (b) $500^{\circ} C$ as a function of Sb/V atomic flux ratio. Substrate reflections are marked by asterisks. $00 l$ -type reflections are observed corresponding to an epitaxial FeVSb film. For very Sb-rich conditions, reflections from a precipitated Sb phase are observed at $2 θ = 43 . 93^{\circ}$ and $96 . 83^{\circ}$ . Right panels: higher resolution scans of the FeVSb 002 reflection, tracking changes in the out-of-plane lattice parameter.
Reuse & Permissions
Figure 5
Structural and electrical transport window. (a) Out-of-plane lattice parameter (circles, left axis) and Sb/V composition (squares, right axis) as a function of Sb/V atomic flux ratio, at growth temperatures of $560^{\circ} C$ and $500^{\circ} C$ . The out-of-plane lattice parameter was extracted from the peak position of the 002 reflection in x-ray diffraction. The Sb/V ratio was determined by energy dispersive x-ray spectroscopy measurements, which are calibrated to a known standard by Rutherford backscattering spectrometry (RBS). We estimate the error bars for the lattice parameter and Sb/V ratio to be $Δ c \approx 0.002$ Å and $Δ Sb$ /V $\approx 0.026$ , respectively, based on the standard deviation for samples grown at $500^{\circ} C$ for flux ratios from 6 to 12, which are well inside the growth window. The standard deviation in the lattice parameter is on par with typical standard deviation observed for other adsorption-controlled ternary compounds grown by MBE [37]. (b) Electron mobility (left axis) and density (right axis) at 300 K extracted from the Hall effect at field greater than 2 T.
Reuse & Permissions
Figure 6
Experimental phase diagram as a function of Sb/V flux ratio and inverse growth temperature. Filled circles denote growth within the structural window as determined by RHEED, XRD, and EDS. Unfilled circles are outside the growth window. The size of the circles scales with the magnitude of the majority carrier mobility at 300 K. Dotted lines are guides to the eye.
Reuse & Permissions

Physical Review Materials