Bayesian regression for
multi-level machine-learned potentials
Subproject P03
The first-principles description of the properties of multi-component metal oxides is an exceedingly challenging problem. The reasons are that the configurational space grows exponentially with the number of species and standard Density Functional Theory (DFT) is often not accurate enough. The long-term objective of P03 is to accelerate first-principles calculations by developing machine-learning approaches for the description of the interatomic forces, Born effective charges, and other tensorial properties of multivalent oxides. The project will rely on kernel-based methods and Bayesian inference to implement fully automatic “on-the-fly” learning.
In the first project period, we will develop machine-learned force fields (MLFF) for DFT and DFT+U, whereby the number of components in the FF will be gradually increased. A concise framework for learning tensorial properties will be implemented. We will use this to simulate infrared spectra of oxide materials, which can be readily compared to the finite-temperature spectra measured by the experimental groups.
The difference between DFT and hybrid functionals will be machine-learned to go beyond semi-local functionals (Delta-learning). The long-term perspective is to extend this approach to highly accurate beyond-DFT methods, such as the random phase approximation and quantum chemistry (coupled cluster) methods. Although kernel-based methods are exceedingly accurate, they are often less efficient than NN. We will collaborate with other projects to recast the on-the-fly trained FF into NN potentials to address this issue.
Expertise
The main research efforts of the group are directed towards the development of quantum-mechanical tools for atomic-scale simulations of properties and processes in materials and the application of these methodologies to key areas of condensed matter physics and materials research. An important pillar of the research is the Vienna Ab initio Simulation Package (VASP), a general-purpose ab initio code for solving the many-electron Schrödinger equation. The code is among the world leaders in its field, with more than 3500 licensees worldwide. We have expertise with simulations for a vast number of properties using many different techniques:
- Density functional theory (DFT), including spin and non-collinear DFT
- Linear response theory to calculate phonons and dielectric properties
- Hartree-Fock techniques and many flavors of hybrid functionals
- Many-body perturbation theory, including GW and Bethe-Salpeter
- Wavefunction-based correlated methods (Møller-Plesset perturbation theory)
- Surface science, including growth and oxide formation
- Simulation of nanostructures
- Semiconductor physics: charge trapping, polarons
- Electronic excitations
- Defect energies in extended systems
For TACO, we will adapt our machine-learning techniques to tensorial properties and correlated wavefunction techniques. These techniques are directly integrated into VASP and allow to accelerate finite-temperature simulations by many orders of magnitudes.
Team
Former Members
Publications
2022
Schmid, Michael; Rath, David; Diebold, Ulrike
Why and How Savitzky–Golay Filters Should Be Replaced
Journal ArticleOpen AccessIn: ACS Measurement Science Au, vol. 2, no. 2, pp. 185–196, 2022.
Abstract | Links | BibTeX | Tags: P02
@article{ACSMEASURE2022,
title = {Why and How Savitzky–Golay Filters Should Be Replaced},
author = {Michael Schmid and David Rath and Ulrike Diebold},
url = {https://pubs.acs.org/doi/10.1021/acsmeasuresciau.1c00054},
doi = {10.1021/acsmeasuresciau.1c00054},
year = {2022},
date = {2022-02-17},
urldate = {2022-02-17},
journal = {ACS Measurement Science Au},
volume = {2},
number = {2},
pages = {185--196},
abstract = {Savitzky–Golay (SG) filtering, based on local least-squares fitting of the data by polynomials, is a popular method for smoothing data and calculations of derivatives of noisy data. At frequencies above the cutoff, SG filters have poor noise suppression; this unnecessarily reduces the signal-to-noise ratio, especially when calculating derivatives of the data. In addition, SG filtering near the boundaries of the data range is prone to artifacts, which are especially strong when using SG filters for calculating derivatives of the data. We show how these disadvantages can be avoided while keeping the advantageous properties of SG filters. We present two classes of finite impulse response (FIR) filters with substantially improved frequency response: (i) SG filters with fitting weights in the shape of a window function and (ii) convolution kernels based on the sinc function with a Gaussian-like window function and additional corrections for improving the frequency response in the passband (modified sinc kernel). Compared with standard SG filters, the only price to pay for the improvement is a moderate increase in the kernel size. Smoothing at the boundaries of the data can be improved with a non-FIR method, the Whittaker–Henderson smoother, or by linear extrapolation of the data, followed by convolution with a modified sinc kernel, and we show that the latter is preferable in most cases. We provide computer programs and equations for the smoothing parameters of these smoothers when used as plug-in replacements for SG filters and describe how to choose smoothing parameters to preserve peak heights in spectra.},
keywords = {P02},
pubstate = {published},
tppubtype = {article}
}
2021
Jakub, Zdenek; Meier, Matthias; Kraushofer, Florian; Balajka, Jan; Pavelec, Jiri; Schmid, Michael; Franchini, Cesare; Diebold, Ulrike; Parkinson, Gareth S.
Rapid oxygen exchange between hematite and water vapor
Journal ArticleOpen AccessIn: Nature Communications, vol. 12, iss. 1, no. 6488, 2021.
Abstract | Links | BibTeX | Tags: P02, P04, P07
@article{Jakub2021,
title = {Rapid oxygen exchange between hematite and water vapor},
author = {Zdenek Jakub and Matthias Meier and Florian Kraushofer and Jan Balajka and Jiri Pavelec and Michael Schmid and Cesare Franchini and Ulrike Diebold and Gareth S. Parkinson},
doi = {10.1038/s41467-021-26601-4},
year = {2021},
date = {2021-11-10},
journal = {Nature Communications},
volume = {12},
number = {6488},
issue = {1},
publisher = {Springer Science and Business Media LLC},
abstract = {Oxygen exchange at oxide/liquid and oxide/gas interfaces is important in technology and environmental studies, as it is closely linked to both catalytic activity and material degradation. The atomic-scale details are mostly unknown, however, and are often ascribed to poorly defined defects in the crystal lattice. Here we show that even thermodynamically stable, well-ordered surfaces can be surprisingly reactive. Specifically, we show that all the 3-fold coordinated lattice oxygen atoms on a defect-free single-crystalline “r-cut” (1-102) surface of hematite (α-Fe_{2}O_{3}) are exchanged with oxygen from surrounding water vapor within minutes at temperatures below 70 °C, while the atomic-scale surface structure is unperturbed by the process. A similar behavior is observed after liquid-water exposure, but the experimental data clearly show most of the exchange happens during desorption of the final monolayer, not during immersion. Density functional theory computations show that the exchange can happen during on-surface diffusion, where the cost of the lattice oxygen extraction is compensated by the stability of an HO-HOH-OH complex. Such insights into lattice oxygen stability are highly relevant for many research fields ranging from catalysis and hydrogen production to geochemistry and paleoclimatology.},
keywords = {P02, P04, P07},
pubstate = {published},
tppubtype = {article}
}
Franceschi, Giada; Schmid, Michael; Diebold, Ulrike; Riva, Michele
Two-dimensional surface phase diagram of a multicomponent perovskite oxide: La0.8Sr0.2MnO3 (110)
Journal ArticleIn: Physical Review Materials, vol. 5, no. 9, pp. L092401, 2021.
Abstract | Links | BibTeX | Tags: P02
@article{Franceschi2021,
title = {Two-dimensional surface phase diagram of a multicomponent perovskite oxide: La_{0.8}Sr_{0.2}MnO_{3} (110)},
author = {Giada Franceschi and Michael Schmid and Ulrike Diebold and Michele Riva},
doi = {10.1103/physrevmaterials.5.l092401},
year = {2021},
date = {2021-09-24},
urldate = {2021-09-24},
journal = {Physical Review Materials},
volume = {5},
number = {9},
pages = {L092401},
publisher = {American Physical Society (APS)},
abstract = {The many surface reconstructions of (110)-oriented lanthanum strontium manganite (\textbf{La_{0.8}Sr_{0.2}MnO}3}}, LSMO) were followed as a function of the oxygen chemical potential (\textit{\textbf{μ}_{O}}) and the surface cation composition. Decreasing \textit{\textbf{μ}_{O}} causes Mn to migrate across the surface, enforcing phase separation into \textit{\textbf{A}}-site-rich areas and a variety of composition-related, structurally diverse \textit{\textbf{B}}-site-rich reconstructions. The composition of these phase-separated structures was quantified with scanning tunneling microscopy, and these results were used to build a two-dimensional phase diagram of the LSMO(110) equilibrium surface structures.},
keywords = {P02},
pubstate = {published},
tppubtype = {article}
}