Electron-phonon origins of unconventional resistivity in moderately correlated perovskite oxides

Abstract

Transition-metal perovskite oxides exhibit moderately correlated metallic phases, several of which exhibit a $T^2$ resistivity scaling up to temperatures far exceeding the regime where Fermi-liquid electron-electron scattering is expected to dominate. Some of these materials, such as SrMoO$_3$, also exhibit unexplained ultra-low room-temperature resistivity. We demonstrate that in SrMoO$_3$, SrWO$_3$, SrTaO$_3$, SrNbO$_3$, and SrVO$_3$ electron-phonon scattering results in quadratic-scaling resistivity due to the shape of the Fermi surface and the thermal activation of optical phonons. We also reveal that the origin of the low resistivity of SrMoO$_3$ is an overall low electron-phonon coupling strength, and identify SrWO$_3$ and SrTaO$_3$ as other possible low-resistivity oxides. Additionally, we find that the strength of electron-phonon coupling is sensitive to structural distortions, energies of optical phonons, and the treatment of electronic correlations. This suggests design principles for finding other ultra-high conductivity transition-metal oxides, and has significant implications for theoretical interpretation of direct-current resistivity in transition-metal oxides and beyond.

Figures (4)

• Figure 1: DC resistivity versus temperature of $Pm\overline{3}m$ SrMoO$_3$, SrWO$_3$, SrTaO$_3$, SrNbO$_3$ (with 1% tensile hydrostatic strain), and SrVO$_3$, calculated using an iterative solution to the BTE from electron-phonon scattering (indicated by points) as well as fits to the Kukkonen model Kukkonen1978 (indicated by solid lines). The gray dashed line demonstrates $T^2$ scaling, and the red (SrMoO$_3$) and purple (SrNbO$_3$) show the sheet of the Fermi surface which is comprised of interpenetrating cylinders.
• Figure 2: Phonon dispersion for (a) SrMoO$_3$ and (c) SrVO$_3$ with and without Hubbard $U$. Panels (b) and (d) show the el-ph matrix elements averaged over the Fermi surface [Eq. \ref{['eq:g_fermi']}] (solid curves) and isotropic transport el-ph coupling strength $\lambda_{\text{tr}}$ [Eq. \ref{['eq:lambda']}] resolved over phonon frequencies (dashed curves) versus phonon energy for SrMoO$_3$ and SrVO$_3$, respectively.
• Figure 3: DC resistivity versus temperature for SrMoO$_3$. The solid red and blue lines are el-ph resistivities from iterative BTE solutions using DFT+$U$ data for the $Pm\overline{3}m$ and $Imma$ structures, respectively. The red dashed line further contains the el-el resistivity for the $Pm\overline{3}m$ structure, taken from Ref. Our_PRL and separately shown in yellow. Light blue points are experimental measurements for thin films Wang2001Lekshmi2005Cappelli2022Radetinac2016, and light red for the single crystal Nagai2005. Residual resistivities are subtracted as in Ref. Our_PRL. Gray dotted curves are a guide to the eye indicating $T^5$ and $T^2$ dependencies.
• Figure 4: Ab-initio DFT+$U$+BTE temperature-dependent el-ph resistivities (black circles) fit to the model of Ref. Kukkonen1978, Eq. (\ref{['eq:kukkonen']}), (magenta solid curves) which assumes a cylindrical Fermi surface, and the conventional Bloch–Grüneisen formula, Eq. (\ref{['eq:BG']}) (green solid curves), which assumes a spherical Fermi surface. Dashed (dotted) magenta and green curves are contributions from Debye acoustic (Einstein optical) modes to the Kukkonen and Bloch fits, respectively. For SrVO$_3$, the Kukkonen fit required an additional optical mode indicated by the dot-dash curve in panel (e). Gray lines are guides to the eye for $T^2$ scalings and blue shaded regions indicate the temperature range of $T^2$ scaling. Panel (f) shows the onset of the el-ph $T^2$ regime as a function of the diameter of the Fermi-surface cylinder.