Publications

See Volker's Google Scholar profile for a full list and up-to-date metrics.

An automated framework for exploring and learning potential-energy surfaces

We introduce autoplex, an automation framework for MLIPs, and demonstrate its usefulness in random structure searching.

Nat. Commun. 2025, 16, 7666

Synthetic pre-training for neural-network interatomic potentials

We show that synthetic data, obtained at scale with an existing ML potential, constitute a useful pre-training task for graph-network potentials.

Mach. Learn.: Sci. Technol. 2024, 5, 015003

Indirect learning and physically guided validation of interatomic potential models

We introduce the idea of distilling an accurate, but slow "teacher" MLIP into a less complex, faster "student" model.

J. Chem. Phys. 2022, 157, 104105

Signatures of paracrystallinity in amorphous silicon from machine-learning-driven molecular dynamics

We revisit the long-standing question whether "paracrystalline" regions can exist in the otherwise disordered network of a-Si.

Nat. Commun. 2025, 16, 2360

Geometrically frustrated interactions drive structural complexity in amorphous calcium carbonate

We simulate amorphous calcium carbonate (ACC) using Hybrid Reverse Monte Carlo refinement, and study the nature of interactions in the ACC network.

Nat. Chem. 2024, 16, 36–41

Cluster Fragments in Amorphous Phosphorus and their Evolution under Pressure

We describe the structure of amorphous red phosphorus with machine-learning-driven simulations.

Adv. Mater. 2022, 34, 2107515

Accelerated First-Principles Exploration of Structure and Reactivity in Graphene Oxide

We describe the atomic structure of graphene oxide (GO) using ML-driven simulations, comparing our simulations to earlier experiments, and providing the GO-MACE-23 model for future studies in the field.

Angew. Chem. Int. Ed. 2024, 63, e202410088

Device-scale atomistic modelling of phase-change memory materials

We report "device-scale" atomistic simulations of phase-change memory materials with a bespoke MLIP model.

Nat. Electron. 2023, 6, 746–754

Coarse-grained versus fully atomistic machine learning for zeolitic imidazolate frameworks

We create ML models for energetics in zeolitic imidazolate framework (ZIF) structures, discussing to what extent chemical information can be “coarse-grained” in hybrid framework materials.

Chem. Commun. 2023, 59, 11405–11408

The amorphous state as a frontier in computational materials design (Nature Reviews Materials, 2025)
Data as the next challenge in atomistic machine learning (Nature Computational Science, 2024)
How to validate machine-learned interatomic potentials (The Journal of Chemical Physics, 2023)
Simulations in the era of exascale computing (Nature Reviews Materials, 2023)
Gaussian Process Regression for Materials and Molecules (Chemical Reviews, 2021)

Publications

Key publications: ML interatomic potentials

An automated framework for exploring and learning potential-energy surfaces

Synthetic pre-training for neural-network interatomic potentials

Indirect learning and physically guided validation of interatomic potential models

Key publications: Amorphous structures

Signatures of paracrystallinity in amorphous silicon from machine-learning-driven molecular dynamics

Geometrically frustrated interactions drive structural complexity in amorphous calcium carbonate

Cluster Fragments in Amorphous Phosphorus and their Evolution under Pressure

Key publications: Functional materials by design

Accelerated First-Principles Exploration of Structure and Reactivity in Graphene Oxide

Device-scale atomistic modelling of phase-change memory materials

Coarse-grained versus fully atomistic machine learning for zeolitic imidazolate frameworks

Review, perspective, and commentary articles