Intrinsic defect intolerance in the ultra-pure metal PtSn$_4$

TL;DR

PtSn4 is shown to be intrinsically defect-intolerant, with typical crystals exhibiting RRR values exceeding 1000 and unable to drop below 100 even under extreme growth rates. The authors combine resistivity measurements, DFT calculations, and STM imaging to show that both Pt and Sn sublattices contribute roughly equally to transport, and that vacancy formation energies strongly penalize defect formation. This defect-free character underpins the material's extreme magnetoresistance and potential surface topology, while comparing related MSn4 compounds highlights PtSn4's unique resistance to defect formation. The work positions PtSn4 as a pristine platform for exploring intrinsic transport and disorder-free topological phenomena, with prospects for exfoliation to few-layer forms.

Abstract

Ultra-pure materials are highly valued as model systems for the study of intrinsic physics. Frequently, however, the crystal growth of such pristine samples requires significant optimization. PtSn$_4$ is a rare example of a material that naturally forms with a very low concentration of crystalline defects. Here, we investigate the origin of its low defect levels using a combination of electrical resistivity measurements, computational modeling, and scanning tunneling microscopy imaging. While typical flux-grown crystals of PtSn$_4$ can have residual resistivity ratios (RRRs) that exceed 1000, we show that even at the most extreme formation speeds, the RRR cannot be suppressed below 100. This aversion to defect formation extends to both the Pt and Sn sublattices, which contribute with equal weight to the conduction properties. Direct local imaging with scanning tunneling microscopy further substantiates the rarity of point defects, while the prohibitive energetic cost of forming a defect is demonstrated through density functional theory calculations. Taken together, our results establish PtSn$_4$ as an intrinsically defect-intolerant material, making it an ideal platform to study other properties of interest, including extreme magnetoresistance and topology.

Figures (4)

• Figure 1: Structural and electrical properties of PtSn$_4$. (a,b) The quasi two-dimensional crystal structure of PtSn$_4$ is composed of staggered layers, where each layer is composed of a Pt square lattice sandwiched between two layers of Sn. Within each layer, square antiprismatic PtSn$_8$ units have an edge-sharing coordination as can be seen when viewing the structure along the $b$-axis. Sn$_\text{a}$ and Sn$_\text{b}$ label two crystallographically equivalent positions that when exposed as the top layer through cleaving yield distinct surfaces. (c) The metallic self-flux method yields cm-scale crystals of PtSn$_4$ with well-defined rectangular facets along the crystallographic $b$-axis. The as-grown crystal is found to cleave easily in this direction. (d) Rietveld refinement of powder x-ray diffraction data confirms the orthorhombic Ccca (no. 68) structure and high crystalline quality. (e) These as-grown crystals of PtSn$_4$ exhibit remarkable residual resistivity ratios (RRRs) above 1000 for crystals grown with a standard cooling rate (60 hrs). Dramatically accelerated cooling (12 and 1.5 hrs) still yields crystals with RRR values over 100, indicative of very good crystal quality and very low levels of defects.
• Figure 2: The role of dimensionality in PtSn$_4$. (a) Normalized resistivity as a function of temperature for in-plane (i$\parallel$ ac) and out-of-plane (i$\parallel$ b) directions in as-grown PtSn$_4$ crystals plotted on a log-log scale. The absolute value of resistivity for the two directions is similar in magnitude, indicative of relatively isotropic electrical conductivity, typical for a three-dimensional material. For (i$\parallel$ b) the data below 10 K are at the signal detection threshold. (b) Projected band structures onto the Pt (purple) and Sn atoms (gray). (c) DOS with contributions from the Pt (purple) and Sn (gray). At the Fermi level, the total Sn weights are only slightly higher than the Pt weights. This demonstrates that both Pt and Sn layers contribute to the overall transport in PtSn$_4$. The insert shows the first BZ of the primitive unit cell of PtSn$_4$ and the chosen high-symmetry path (red).
• Figure 3: Comparison of the structure-property relationships within comparable members in the MSn$_4$ family. (a) Illustrates a periodic table representation of the members of the MSn$_4$ (M= Rh, Pd, Ir, Pt, or Au) family showing the respective space group of formation, along with experimentally determined residual resistivity ratio (RRR), and DFT calculated vacancy defect formation energy for M and Sn for the relevant crystal structure. The RRR value for RhSn$_4$ is taken from literature xing2016large. An Ag analog, indicated by the grayed-out rectangle, does not exist. (b) Temperature dependence of the in-plane (ip) resistivity (left panel) and resistivity normalized to room temperature (right panel) for AuSn$_4$ (yellow), IrSn$_4$ (teal), PdSn$_4$ (light purple), and PtSn$_4$ (purple). While the room temperature resistivity values are comparable across the four materials, the residual resistivity at base temperature differs by more than an order of magnitude.
• Figure 4: Categorization of point defects in PtSn$_4$ via low-temperature scanning tunneling microscopy (STM). (a) STM topography on pristine PtSn$_4$ surface ($\Delta$z = 30 pm) and (b) the Pt-lattice corrugation overlay shown as white outlined blue circles. Gallery of nine defect types observed in PtSn$_4$, including (c-e) three types of Sn-site defects and (f-k) six types of Pt-site defects. The black line on the defects indicates the corresponding mirror planes. Images are acquired at different sample bias and set point currents: (c,f,h,i,j) used (-400 meV, 500 pA), (d) used (400 meV, 600 pA), (e) used (-400 meV, 600 pA), (g) used (-500 meV, 500 pA), and (k) used (-400 meV, 500 pA). All measurements were performed at a temperature of 4.4 K.