Diffusion coefficients of multi-principal element alloys from first principles

Damien K. J. Lee, Anirudh Raju Natarajan · 2026

Read Diffusion coefficients of multi-principal element alloys from first principles by Damien K. J. Lee, Anirudh Raju Natarajan (2026) 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-2603-24228.txt.

Introduction

Lee and Natarajan compute diffusion rates for alloys of many main elements. Vacancy-driven diffusion in such alloys is poorly understood. Their embedded local cluster expansion links first-principles data with kinetic Monte Carlo runs. They apply it to a six-element alloy: vanadium, chromium, niobium, molybdenum, tantalum, and tungsten. Local barriers, not thermodynamics, mainly control diffusion, they find. They screen other mixes and flag some with fast, anti-sluggish diffusion.

Catalog record

CREATORDamien K. J. Lee, Anirudh Raju Natarajan
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE50113
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.24228

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Diffusion coefficients of multi-principal element alloys from first principles Damien K. J. Lee, Anirudh Raju Natarajan Abstract Vacancy-mediated diffusion in multi-principal element alloys (MPEAs) remains poorly understood. Existing computational methods face challenges in connecting electronic structure to macroscopic transport coefficients due to the large number of chemical elements. To address this, we introduce the embedded local cluster expansion (eLCE), which bridges first-principles calculations with kinetic Monte Carlo simulations to compute the matrix of multicomponent diffusion coefficients. Applying this approach to refractory MPEAs in the V–Cr–Nb–Mo–Ta–W system, we evaluate the complete mobility and diffusion tensors of a six-component alloy at finite temperatures. We find that local kinetic barriers, rather than thermodynamics or vacancy correlation factors, primarily control diffusion in these materials. Whether diffusion is sluggish or anti-sluggish depends on the mean vacancy migration barrier relative to the rule-of-mixtures estimate and on the availability of percolating pathways of fast-diffusing species. We use this insight to screen the senary composition space and identify compositions with anti-sluggish diffusion. This study presents a predictive, first-principles approach for computing non-dilute transport coefficients and designing MPEAs with targeted transport properties. I Introduction Diffusion governs microstructural evolution, phase stability, creep resistance, and high-temperature performance in alloys Ardell and Ozolins 2005 ; Miracle and Senkov 2017 ; Darling et al. 2016 ; Hinrichs et al. 2025 ; Wang et al. 2025 . Multi-principal element alloys (MPEAs) are of growing interest for high-temperature applications Yeh et al. 2004 ; Miracle and Senkov 2017 , yet diffusion in these systems remains poorly understood. Diffusion in MPEAs has been widely described as “sluggish” Tsai et al. 2013 ; Hsu et al. 2024 , but the microscopic origin of this claim has not been established, largely because a quantitative, first-principles evaluation of diffusion in these alloys has not been achieved. Quantifying diffusion in multicomponent alloys is challenging from both experimental and computational perspectives. Experimentally, coupled fluxes and cross-correlations between species complicate the extraction of transport coefficients Paul et al. 2014 ; DeHoff and Kulkarni 2002 ; Dash et al. 2022 . Computationally, modeling vacancy-mediated diffusion requires resolving the combinatorial complexity of local chemical environments that govern migration barriers Xing et al. 2024 ; Goiri et al. 2019 . Furthermore, simulations must access timescales long enough to yield macroscopic transport coefficients. Most existing studies rely on tracer diffusivities Zhang et al. 2022a ; Sen et al. 2023 or pseudo-binary approximations Kirkaldy and Lane 1966 ; Esakkiraja et al. 2019 ; Dash et al. 2020 . A rigorous evaluation of the complete multicomponent diffusion tensor from first-principles statistical mechanics has not been achieved beyond simple alloys Dash et al. 2022 ; Van der Ven et al. 2020 . At the atomistic scale, diffusion in substitutional alloys proceeds through thermally activated vacancy hops Van der Ven et al. 2010 ; Balluffi et al. 2005 ; Kirkaldy et al. 1987 ; Shewmon 1958 . In multicomponent systems, the migration barrier for each hop depends on the local chemical environment, producing a broad spectrum of activation energies even in binary alloys Goiri et al. 2019 ; Behara and Van der Ven 2024 . A rigorous description of diffusion therefore requires both an accurate model of migration energies across diverse local environments and sufficient statistical sampling of vacancy transport over long timescales. Cluster expansion techniques coupled with kinetic Monte Carlo (kMC) simulations can in principle provide both ingredients Deng et al. 2022 ; Jadidi et al. 2023 ; Canepa 2023 . Although cluster expansions have been highly successful for modeling configurational thermodynamics and vacancy migration in simpler alloys, their extension to kinetic barriers in concentrated multicomponent systems remains computationally demanding and has long been considered intractable Dash et al. 2022 ; Xing et al. 2024 . Here, we introduce the embedded local cluster expansion (eLCE) formalism, which enables rapid and accurate evaluation of kinetically resolved activation barriers across complex multicomponent environments. By coupling this formalism with kMC simulations and first-principles electronic structure calculations, we study substitutional diffusion in the senary V–Cr–Nb–Mo–Ta–W system, a refractory MPEA under active investigation for its mechanical and high-temperature properties Cook et al. 2024 ; Kumar et al. 2024 ; George et al. 2019 ; Mak et al. 2021 ; Li et al. 2020 . We rigorously compute the full non-dilute diffusion coefficient tensor from first-principles statistical mechanics and analyze diffusion in terms of its collective eigenmodes. Our results reveal that sluggish diffusion in this class of alloys is governed by the structure of the vacancy barrier spectrum and the emergence of percolating low-energy pathways, rather than chemical complexity alone. This work establishes a predictive framework for diffusion in multicomponent alloys, rigorously connecting atomistic migration energies to macroscopic transport coefficients. II Results II.1 Kinetics of multicomponent alloys We begin by outlining the formalism required to compute kinetic properties from first principles. In a perfect crystal, the number of lattice sites, including vacancies, is conserved. Diffusion is driven by gradients in diffusion potentials Kehr et al. 1989 ; Balluffi et al. 2005 , defined as the difference between the chemical potential of species i and that of vacancies: \tilde{\mu}_{i}=\mu_{i}-\mu_{Va} (1) The diffusion fluxes of the c chemical elements in the multicomponent alloy are related to the diffusion potentials through the Onsager transport coefficients: \begin{bmatrix}\vec{J_{1}}\\ \vec{J_{2}}\\ \vdots\\ \vec{J_{c}}\end{bmatrix}=-\begin{bmatrix}L_{11}&L_{12}&\cdots&L_{1c}\\ L_{21}&L_{22}&\cdots&L_{2c}\\ \vdots&\vdots&\ddots&\vdots\\ L_{c1}&L_{c2}&\cdots&L_{cc}\end{bmatrix}\begin{bmatrix}\nabla\tilde{\mu}_{1}\\ \nabla\tilde{\mu}_{2}\\ \vdots\\ \nabla\tilde{\mu}_{c}\end{bmatrix} (2) where L_{ij} are the elements of the Onsager transport coefficient matrix \mathbf{L} . Since the total number of sites is conserved, only the fluxes of migrating species need to be tracked, with \sum_{i}\vec{J_{i}}+\vec{J}_{Va}=0 . The Onsager transport coefficients can be obtained using Kubo-Green linear response theory Green 1954 ; Kubo 1957 : \displaystyle L_{ij} \displaystyle=\frac{1}{\Omega k_{B}T}\tilde{L}_{ij} (3) \displaystyle\tilde{L}_{ij} \displaystyle=\frac{1}{2d\,t\,M}\left\langle\left(\sum_{\zeta}\Delta\vec{R}_{i}^{\zeta}(t)\right)\cdot\left(\sum_{\zeta}\Delta\vec{R}_{j}^{\zeta}(t)\right)\right\rangle (4) in which \Omega is the volume per substitutional site, d is the dimensionality of the diffusion network, and M is the number of substitutional sites in the crystal. The vectors \Delta\vec{R}_{i}^{\zeta}(t) connect the endpoints of the trajectory of atom \zeta of species i after time t . These trajectories can be obtained from simulations in which atoms exchange with vacancies at frequencies given by transition state theory Vineyard 1957 : \Gamma=\nu^{*}\exp(\frac{-\Delta E}{k_{B}T}) (5) where \nu^{*} is a vibrational prefactor and \Delta E is the migration barrier of a hop. When \Delta E is known for each possible hop, kMC simulations can be used to evolve the system and compute the transport coefficients from eq. 4 Binder and Heermann 2010 . Gradients in chemical potentials are not readily accessible in experiments. Instead, the diffusion coefficient matrix \mathbf{D} , which relates fluxes to concentration gradients, is more commonly used. The non-dilute diffusion coefficients can be derived from the Onsager transport coefficients Balluffi et al. 2005 ; Van der Ven et al. 2010 ; Kirkaldy et al. 1987 ; Shewmon 1958 : \mathbf{D}=\begin{bmatrix}D_{11}&D_{12}&\cdots&D_{1c}\\ D_{21}&D_{22}&\cdots&D_{2c}\\ \vdots&\vdots&\ddots&\vdots\\ D_{c1}&D_{c2}&\cdots&D_{cc}\end{bmatrix}=\mathbf{\tilde{L}}\mathbf{\tilde{\Theta}} (6) where the thermodynamic factor matrix \bm{\tilde{\Theta}} is proportional to the Hessian of the free energy, and can be calculated from composition fluctuations within the semi-grand canonical ensemble: \tilde{\Theta}^{-1}_{ij}=\frac{1}{M}\langle N_{i}N_{j}\rangle-\langle N_{i}N_{j}\rangle (7) Both \mathbf{\tilde{L}} and \mathbf{\tilde{\Theta}} are computed under boundary conditions of fixed overall composition, vacancy concentration, and temperature. Additional metrics characterize different aspects of diffusion. The correlation factor f_{i} measures the degree of correlation between successive hops: f_{i}=\frac{\sum_{\zeta}\Delta\vec{R}_{i}^{\zeta}(t)}{\sum_{\zeta}N^{\zeta}(t)\Delta\vec{r}^{2}} (8) where N^{\zeta}(t) is the number of hops performed by

Read the complete text