In the present study, we adopt a comprehensive approach of ab-initio simulations by using a commercial software package VASP 6.5.1 based on Density Functional Theory (DFT) using a projector augmented wave (PAW) method with generalized gradient approximation (GGA) along with Perdew-Burke-Ernzerhof (PBE) exchange-correlation functional and calculations are performed on a grid of points in real space, which allowed us to calculate the full set of piezoelectric and mechanical (acoustic) properties of Al1−xScxN (0≤x≤0.5, AlScN) and Al1−xBxN (0≤x≤0.5, AlBN) at 0K, as well as from 200K to 500K [1, 2]. The measured phase velocities and piezoelectric coupling coefficients, kt2 from Al1−xScxN (0.31≤x≤0.38, AlScN) based Thin Film Bulk Acoustic Wave Devices at 5-7 GHz allowed good consistency when comparing with ab-initio simulated piezoelectric and stiffness tensors. Two-layer Graph Atomic Cluster Expansion (GRACE-2L) Universal Machine Learning Interatomic Potential (uMLIP) ) trained on the Meta Open Materials 2024 (OMat24) dataset, was utilized in predicting e33, k2, C33 and lattice densities from 200K to 500K temperature range for Al1−xBxN (0≤x≤0.5, AlBN).

In this training, we will highlight the capabilities of MedeA for performing configurational sampling of alloys and doped structures. These methods involve random substitutions, special quasi-random structures, cluster expansion methods, molecular dynamics of doped supercells using machine-learned potentials, and how electronic properties can be predicted for large doped supercells using machine learning.

Density functional theory (DFT) calculations of charged surfaces and molecules are essential for advancing electrocatalysis and energy materials, yet they are traditionally hindered by the requirement of charge neutrality under three-dimensional (3D) periodic boundary conditions. To overcome this limitation, we introduce a recent methodological advancement in the Vienna ab initio simulation package (VASP) that enables 0D and 2D open boundary conditions. By utilizing a Coulomb kernel truncation method combined with a highly efficient padding approach, we systematically eliminate unphysical long-range vacuum interactions and selectively subtract unwanted periodic artifacts. The computational efficiency and robustness of this approach are demonstrated through large supercell calculations of a charged chlorine defect on an NaCl(001) surface and extensive molecular dynamics simulations of a stepped Au(211) water electrode-electrolyte interface.

Building on these methodological foundations, this work investigates the free solvation energies of sodium and fluorine ions at an unreconstructed Au(111)-water interface. Both charged and uncharged slabs are considered. To achieve the necessary timescales, we employ on-the-fly machine learning force field (MLFF) training in VASP, utilizing large supercells containing approximately 100 water molecules. These force fields are subsequently refitted using GRACE-2layer models.

Leveraging these accelerated models, we perform extensive thermodynamic free energy simulations using the weighted histogram analysis method (WHAM) and metadynamics to extract accurate free energy profiles.

Crucially, we address the fundamental challenge of simulating ions close to an electrode at the point of zero charge using finite slab models.

Our simulations demonstrate that periodic slab calculations for ions dissolved in water must be performed in charged states to correctly capture the macroscopic, infinite point-of-zero-charge limit.

Discover More of What MedeA Can Do

Join us on October 14th for a live MedeA software demonstration, open to both current customers and those interested in learning more about MedeA.

MedeA offers a broad range of capabilities, and even experienced users may be familiar with only a portion of what the platform has to offer. This session will showcase how VASP and LAMMPS are integrated within MedeA, along with advanced tools, high-throughput calculations, and integrated computational workflows.

Whether you’re a current MedeA user looking to expand how you use the platform or you’re exploring MedeA for the first time, this session will provide a practical look at its capabilities and how they work together.

The demonstration will be followed by a 20-minute live Q&A, giving you the opportunity to ask questions and learn more about the tools and workflows most relevant to your work.

In this training, we will go over a robust procedure for training machine-learned forcefields (MLFFs) and how to apply these MLFFs on large-scale systems simulated in the LAMMPS molecular dynamics engine to compute various material properties. Additionally, we will cover new features related to NMR calculations and Transition State Search calculations using MedeA VASP.

First-principles simulations of atomic and electronic structure in solids offer a powerful route to predict and understand material properties. (1) This is particularly relevant in the case of point defects which dramatically affect material properties yet present many challenges for experimental characterisation.

Recent years have seen significant advances in both computational methodologies (2–4) and associated toolkits (5–9) for modelling defect behaviour. I will discuss our collaborative efforts in this area; including the continued development of the open-source doped defect simulation package5, approaches for exploring defect energy surfaces (including MLIP accelerations) (10,11) and remaining challenges in this area. (12)

Atomic-scale defects play a critical role in determining the functional properties and performance of many solid-state materials, with important implications for semiconductor doping, catalytic active sites, charge-carrier recombination and efficiency in solar photovoltaics, and ionic and electronic conductivity in battery materials. However, the dilute concentrations of point defects often make their experimental characterization extremely challenging. First-principles calculations provide a powerful complementary approach for investigating the structure, energetics, and electronic properties of defects at the atomic scale.
This training session will provide a systematic demonstration of density functional theory (DFT)-based workflows for investigating point defects in solids using MedeA. In particular, the session will cover: (1) exploration of defect configurational landscapes, including vacancies, interstitials, and antisites in multiple charge states using the ShakeNBreak approach; (2) DFT calculations of defect-containing supercells using MedeA VASP and the doped package integrated within MedeA; and (3) post-processing workflows for calculating and analyzing defect formation-energy diagrams, charge-transition levels, and defect and carrier concentrations. Cadmium telluride (CdTe), a technologically relevant solar photovoltaic material, will be used as a representative material system.

Machine learning interatomic potentials (MLIPs) are moving from narrowly parameterized models toward expressive foundation models that cover much of the periodic table. In this talk I summarize this development through the lens of the Graph Atomic Cluster Expansion (GRACE). GRACE provides a formally complete basis for many-body atomic interactions and, on that basis, a unified model hierarchy: many recent MLIPs, from local descriptor-based potentials to semilocal message-passing networks, emerge as specific limits of the GRACE formalism.

The framework also brings computational advantages. GRACE avoids the combinatorial growth in basis size that usually accompanies multi-component systems, retaining linear scaling with system size while scaling favorably with the number of chemical elements and the complexity of the basis.

Finally, I present applications that span general-purpose foundation models and fine-tuned models for specific problems: predicting melting temperatures, diffusion mechanisms and barriers, and grain boundary structures, optimizing surface compositions of multi-component alloys, and simulating the reduction of iron by hydrogen. I also show how recently developed atom-resolved uncertainty estimates make it possible to detect extrapolation on the fly, so that simulations of millions of atoms can be run with confidence.

In this training, we will discuss various benefits of running a GRACE potential for determining material properties that are otherwise diXicult to assess with DFT alone. In particular, the ability to determine the temperature-dependence of non-electronic and electronic properties will be discussed.