Carrier mobilities and electron-phonon interactions beyond DFT
Aleksandr Poliukhin, Nicola Colonna, Francesco Libbi, Samuel Poncé, Nicola Marzari · 2025
Read Carrier mobilities and electron-phonon interactions beyond DFT by Aleksandr Poliukhin, Nicola Colonna, Francesco Libbi, Samuel Poncé, Nicola Marzari (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-14852.txt.
Introduction
Aleksandr Poliukhin, Nicola Colonna, Francesco Libbi, Samuel Poncé, and Nicola Marzari work out how electrons and lattice vibrations interact, with tools beyond DFT once thought too costly. Their work came out in 2025. Electron-phonon coupling governs transport, superconductivity, and light absorption. Standard density functional theory gives close values for it, but not exact ones. Hybrid functionals, Koopmans functionals, and GW theory promise better numbers. What was missing was one general way to draw couplings out of all of them.
Their finite-difference framework works with any method that gives energies and states. A projection scheme based on energy gaps gets past the old limits of direct finite difference work. Symmetry cuts the number of atomic displacements the method must test, so costs stay manageable. Carrier mobility values can now climb beyond the standard level, and models of real solids gain much sharper inputs.
Catalog record
| CREATOR | Aleksandr Poliukhin, Nicola Colonna, Francesco Libbi, Samuel Poncé, Nicola Marzari |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 65 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-14852
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Carrier mobilities and electron-phonon interactions beyond DFT Aleksandr Poliukhin, Nicola Colonna, Francesco Libbi, Samuel Poncé, Nicola Marzari Abstract Electron-phonon coupling is a key interaction that governs diverse physical processes such as carrier transport, superconductivity, and optical absorption. Calculating such interactions from first-principles with methods beyond density-functional theory remains a challenge. We introduce here a finite-difference framework for computing electron-phonon couplings for any electronic structure method that provides eigenvalues and eigenvectors, and showcase applications for hybrid and Koopmans functionals, and G W GW many-body perturbation theory. Our approach introduces a novel projectability scheme based on eigenvalue differences and bypasses many of the limitations of the direct finite difference methods. It also leverages symmetries to reduce the number of independent atomic displacements, thereby keeping computational costs manageable. This approach enables seamless integration with established first-principles codes for generating displaced supercells, performing Wannier interpolations, and evaluating transport properties. Applications to silicon and gallium arsenide show that advanced electronic-structure functionals predict different electron-phonon couplings and modify band curvatures, resulting in much more accurate estimates of intrinsic carrier drift mobilities and effective masses. In general, our method provides a robust and accessible framework for exploring electron-phonon interactions in complex materials with state-of-the-art electronic structure methods. Electron-phonon interactions play a central role in determining fundamental materials’ properties, such as electron and hole mobilities Li et al. 2019a ; Poncé et al. 2020 ; Claes et al. 2025 , superconductivity Bercx et al. 2025 , band renormalization Antonius et al. 2014 and non-adiabatic effects Lazzeri and Mauri 2006 . Accurate modeling of these interactions is essential for advancing technologies ranging from efficient electronic devices to novel superconducting materials. First-principles calculations of electron-phonon couplings based on density-functional theory (DFT) have reached a maturity level that even allows high-throughput calculations Antonius et al. 2014 . A key step in these approaches is the evaluation of electron-phonon matrix elements, which quantify the effective interactions between electrons and phonons Giustino 2017 ; Bercx et al. 2025 . Computationally, two approaches exist for this task: density-functional perturbation theory (DFPT) Baroni et al. 2001 ; Gonze and Lee 1997 and the finite difference (FD) method Chang et al. 1985 ; Kresse et al. 1995 ; Vandenberghe and Fischetti 2014 . Recent work suggests machine learning as another promising alternative, as it allows accurate results when trained on relatively small datasets Zhong et al. 2024 ; Li et al. 2024a . Although DFPT is generally a more computationally efficient approach than FD methods, as it offers favorable scaling and access to arbitrary phonon wavevectors \mathbf{q} , it requires dedicated and involved implementations. For this reason, perturbative approaches based on DFPT have only recently been applied to specific beyond-DFT methods, such as DFT+U Anisimov et al. 1991 ; Zhou et al. 2021 ; Yang et al. 2025 and G_{0}W_{0} Hybertsen and Louie 1985 ; Li et al. 2019b . Up to now, most of the calculations on the electron-phonon-related properties have been performed using DFPT with semilocal DFT functionals and this method is considered to be the state of the art. Notwithstanding its success, discrepancies persist between first-principles predictions based on DFPT and experimental observations for many electron-phonon related properties Antonius et al. 2014 ; Tran et al. 2024 ; Yuzbashyan et al. 2022 . This is partially due to the fact that Kohn-Sham (KS) DFT does not provide reliable quasiparticle energies and other excited-state properties Perdew and Levy 1983 ; Godby et al. 1987 ; Marzari et al. 2021 . To address this, more advanced methods have been developed to improve the description of quasiparticle energies. These include the incorporation of a fraction of exact exchange Becke 1993 ; Heyd et al. 2003 , the enforcement of piecewise linearity conditions Dabo et al. 2010 ; Colonna et al. 2019 ; Borghi et al. 2014 ; Nguyen et al. 2018 ; Wing et al. 2021 ; Li et al. 2017 , and the inclusion of many-body effects Hybertsen and Louie 1985 ; Onida et al. 2002 ; Golze et al. 2019 . Improved electronic structure approaches were reported to affect electron-phonon properties Li et al. 2019b ; Li et al. 2024b , but their impact on carrier mobility needs to be better understood. For these reasons, there is a growing need for a more straightforward and general method to calculate electron-phonon couplings. For a long time, finite differences have been successfully used to predict phonon properties with any functional, and today, well-established codes exists that implement these developments Togo and Tanaka 2015 ; Hjorth Larsen et al. 2017 ; Lin et al. 2025 . However, the absence of community codes to perform a similar task for electron-phonon properties underscores the need for a more accessible and general approach. One of the primary objectives of this work is to establish an approach based on FD to calculate electron-phonon matrix elements starting from any arbitrary electronic-structure method that provides eigenvectors and eigenvalues of the Hamiltonian for displaced systems. This approach and the code that implements it aim to fill the gap in the current software landscape. We illustrate our approach using Quantum ESPRESSO Giannozzi et al. 2009 ; Giannozzi et al. 2017 , Koopmans Linscott et al. 2023 , and Yambo Marini et al. 2009 as a b-initio electronic structure engines. Furthermore, we provide an interface to the EPW software Poncé et al. 2016 ; Lee et al. 2023 and compute the transport properties of Si and GaAs to assess the effect that using hybrid and Koopmans functionals, and GW many-body perturbation theory have on transport properties. I Results The central quantity that describes electron-phonon interactions arises from the expansion of the effective potential that acts on the electrons due to the motion of atoms, which is given by: \hat{H}^{\rm e-ph}=\sum_{\kappa\alpha l}\frac{\partial\hat{V}}{\partial\tau_{\kappa\alpha l}}\delta\tau_{\kappa\alpha l}, (1) where \hat{V} is the effective potential, \tau_{\kappa\alpha l}=\mathbf{R}_{l}+\tau_{\kappa\alpha} is the \alpha Cartesian component of the position of the atom \kappa in a unit cell located at position \mathbf{R}_{l} . The matrix element of this operator describes the probability that an electron transitions from an initial to a final state because of the interaction with the atomic displacement \delta\tau_{\kappa\alpha l} . Expanding the atomic displacement \delta\tau_{\kappa\alpha l} on the phonon basis, it is possible to define the electron-phonon matrix element g_{mn\nu}(\textbf{k},\textbf{q}) as: g_{mn\nu}(\textbf{k},\textbf{q})\equiv\\ \sum_{\kappa\alpha l}\sqrt{\dfrac{\hbar}{2M_{\kappa}\omega_{\textbf{q}\nu}}}e^{i\textbf{q}\cdot\textbf{R}_{l}}e_{\kappa\alpha\textbf{q}\nu}\bra{\psi_{m\textbf{k}+\textbf{q}}}\dfrac{\partial V}{\partial\tau_{\kappa\alpha l}}\ket{{\psi_{n\textbf{k}}}}, (2) where \omega_{\textbf{q}\nu} , e_{\kappa\alpha\textbf{q}\nu} are the phonon frequencies and eigenvectors respectively, \psi_{n\textbf{k}} is the one-particle wave function, and \partial V/\partial\tau_{\kappa\alpha l} is the change of potential with respect to the displacement. Standard FD approaches require one to evaluate the change in effective potential due to atomic displacement by using finite differences in a displaced supercell commensurate with the phonon q -vector of interest. However, in pseudopotential codes, there is a non-local part of the potential that is treated analytically and thus would require an additional implementation when different types of pseudopotential are used, such as USPP Vanderbilt 1990 or PAW Blöchl 1994 . For norm-conserving pseudpotentials the expression is given in Sec. S1, Eq. ( S2 ) of the SI sup 2025 . Moreover, in beyond-DFT approaches, such as hybrids, Koopmans, or G_{0}W_{0} , the effective potential is also a more complex operator that requires additional treatment, such as saving the unique potential for