Domain Boundaries in a Metallic Distortive Polar Metal

TL;DR

The paper investigates domain boundaries in a metallic distortive polar metal, Mn$_5$Al$_8$, revealing a symmetry-lowering transition from cubic $\\gamma$-brass to polar rhombohedral $\gamma_2$ that creates a herringbone multi-variant domain structure with H–H and T–T boundaries. Using EBSD, TEM, and 4D-STEM DPC, the authors map the four polar variants and the characteristic interlocking Y boundary pattern, then link boundary character to surface reactivity through copper deposition, XPS, and EFM measurements. They show that boundaries can locally modulate the density of states and work function, despite metallic screening, suggesting bound-charge-like effects in metals and offering functionalizable boundaries for tunable surface chemistry. A broader ICSD-based chemical-space analysis positions Mn$_5$Al$_8$ within a growing set of metallic DPMs, motivating further exploration of symmetry-lifting transitions and boundary-driven phenomena in polar metals for advanced electronic materials and devices.

Abstract

Polar metals are an underexplored material class combining two properties that are typically incompatible, namely a polar crystal structure and reasonable electrical conductivity. These intriguing materials offer a unique combination of properties, potentially relevant to optoelectronics, catalysis, memory devices, among other applications. The distortive polar metal (DPM) subclass forms through a symmetry-lifting phase transformation into a non-centrosymmetric polar crystal structure. In the process, domains with uniform geometric polar directions form, oftentimes separated by domain boundaries with polarity discontinuities arranged in "charged" head-to-head (H-H) or tail-to-tail (T-T) morphologies. To date, only metallic oxide DPM microstructures have been studied. Here we reveal, in the intermetallic DPM Mn$_{5}$Al$_{8}$, different surface interactions and electron transfer reactivity at domain boundaries depending on their H-H or T-T character. Variable surface reactivity suggests localized changes in electronic work functions due to an increase (H-H) or decrease (T-T) in the electronic density of states. These findings suggest that metallic DPMs may offer functionalizable domain boundaries and deserve increased attention, given that they allow tunable chemistries and various thermomechanical processing or transformation protocols. Ultimately, this study proposes unconventional metal physics, propelling the discovery and design of advanced electronic materials and devices.

Figures (4)

• Figure 1: Structural and electronic assessment of polar metal state.a, Mn-Al phase diagram showing suggested high-temperature (cubic, $\gamma$-brass, red) and low temperature (rhombohedral, $\gamma_2$, blue) phases. The arrow in this diagram highlights the composition chosen and transformation pathway. b, Room temperature and in-situ $1000^{\circ}$C high temperature XRD data of the cubic $\gamma_2$–brass and polar $\gamma_2$ phase, with the cubic/pseudo-cubic respective unit cells viewed along $[100]$ and $[111]$. c, Electrical resistivity vs temperature data of Mn$_5$Al$_8$ is plotted alongside many studied polar metal compounds of different classifications hockoxyoung2023Ali2014 and some metals hockoxyoung2023Desai1984ZHANG2001, and the left inset confirms positive correlation between resistivity and temperature. The inset on the right shows a photograph of as-cast sample pieces. d, The calculated electronic band structure and density of states are given, confirming metallic behavior.
• Figure 2: Microstructure analysis and herringbone domain boundary characterization.a, Backscattered electron (BSE) SEM micrograph of the as-cast Mn$_{51}$Al$_{49}$ microstructure, showcasing one large grain (grain boundary marked with red dotted line) and the diverse “colony” morphologies that make of up Rank-2 Laminate Twins. b, EBSD micrograph of characteristic herringbone domains, showing 4 different variants oriented $\sim90^{\circ}$ to each other with the domain boundaries marked with black line and the twin traces of $\{100\}_{PC}$ and $\{011\}_{PC}$ are indicated by orange and white lines, respectively; the corresponding pole figure is given in the inset. c, High-resolution HAADF-STEM micrograph taken of a $(100)_{PC}$-twin, viewed along the $[21\bar{3}\bar{2}]||[001]_{PC}$ zone axis. The twin boundary is marked with a dashed line as a guide to the eye, and polarization $\langle111\rangle$ axes given (going out of the plane) in light-blue arrows.
• Figure 3: Direct observation of surface effects at Mn$_5$Al$_8$ domain boundaries.a, 2D schematic articulating metallic DPM domain boundaries, where surplus electrons (free charge carriers) can accumulate at H-H domain boundaries, commensurate with local “bound charge” deviations due to geometrical polarization. The variation in the DOS in turn yields a decrease of the work function for H-H domain boundaries, and the inverse is expected for T-T domain boundaries (increase in work function). b, A schematic illustration depicting the repeating ‘interlocking Y pattern’ outlining the herringbone morphology, with yellow/black lines corresponding to H-H/T-T domain boundaries, respectively c, SEM low voltage (0.5 kV) micrograph showing bright contrast at one set of (H-H) domain boundaries. d, BSE-SEM micrograph showing decoration of domain boundaries by deposition of copper particles. In this micrograph the herringbone morphology, and an adjacent morphology exhibiting linear “charged” boundaries (i.e. ‘backgammon’ domain morphology) is decorated with copper, though this patterning is not discussed herein (see Fig. S7c,d).
• Figure 4: Chemical sorting of polar compounds and metallic polar metals.a, A chemical sorting of all compounds in one of the ten polar crystal classes (point groups 1, 2, $m$, $mm2$, 3, $3m$, 4, $4mm$, 6, $6mm$) in the ICSD ($\sim$2600 unique compounds): compound containing oxygen are listed as oxides, whereas complex oxides here include any oxyanion-species containing oxide group (e.g., carbonates, sulfates). Compounds containing other non-metal species (e.g. carbides, nitrides, sulfides) are listed as such. Compounds listed in the metalloid category contain one of these six species (B, Si, Ge, As, Sb, Te,), and metallic compounds comprise only metallic alkaline, alkaline earth, transition, post-transition, and lanthanide species (5.1% or $\sim$134 in total). Using the common notion of chemical elements as metals, metalloids, and nonmetals provides a simple, but useful classification of the predominant bonding type by chemical composition. b, Further sorting of metallic compounds into subcategories of rare earth, precious, toxic/radioactive, and common (of which Mn$_5$Al$_8$ is one). Hatched regions correspond to compounds containing species in both adjacent groups.