Artem Samtsevych

Artem Samtsevych

Auto-generated profile, based on publicly available information

Artem Samtsevych

Fritz Haber Institute of the Max Planck Society

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
21 April 2022 in Nature Chemistry
Verified
Journal Article
Closed access
PDF file
YWYu Wang
Yu Wang
MBMaxim Bykov
Maxim Bykov
ICIlya Chepkasov
Ilya Chepkasov
ASArtem Samtsevich
Artem Samtsevich
EBElena Bykova
Elena Bykova
AGAlexander Goncharov
Alexander Goncharov

12 June 2023 in The Journal of Chemical Physics
Verified
Journal Article
Open access
PDF file
CPChiara Panosetti
Chiara Panosetti
YLY. Lee
Y. Lee
ASA. Samtsevych
A. Samtsevych
CSChristoph Scheurer
Christoph Scheurer
The increasing popularity of machine learning (ML) approaches in computational modeling, most prominently ML interatomic potentials, opened possibilities that were unthinkable only a few years ago—structure and dynamics for systems up to many thousands of atoms at an ab initio level of accuracy. Strictly referring to ML interatomic potentials, however, a number of modeling applications are out of reach, specifically those that require explicit electronic structure. Hybrid (“gray box”) models based on, e.g., approximate, semi-empirical ab initio electronic structure with the aid of some ML components offer a convenient synthesis that allows us to treat all aspects of a certain physical system on the same footing without targeting a separate ML model for each property. Here, we compare one of these [Density Functional Tight Binding with a Gaussian Process Regression repulsive potential (GPrep-DFTB)] with its fully “black box” counterpart, the Gaussian approximation potential, by evaluating performance in terms of accuracy, extrapolation power, and data efficiency for the metallic Ru and oxide RuO2 systems, given exactly the same training set. The accuracy with respect to the training set or similar chemical motifs turns out to be comparable. GPrep-DFTB is, however, slightly more data efficient. The robustness of GPRep-DFTB in terms of extrapolation power is much less clear-cut for the binary system than for the pristine system, most likely due to imperfections in the electronic parametrization.

1 May 2026 in Journal of Chemical Theory and Computation
Verified
Journal Article
Open access
ASArtem Samtsevych
Artem Samtsevych
YSYihua Song
Yihua Song
THTammo van der Heide
Tammo van der Heide
BABalint Aradi
Balint Aradi
BHBenjamin Hourahine
Benjamin Hourahine
CPChiara Panosetti
Chiara Panosetti
Density functional tight binding (DFTB) offers a computationally efficient alternative to ab initio methods, bridging the accuracy of density functional theory (DFT) and the speed of semiempirical models. The approximate nature of DFTB makes its reliability highly dependent on parameter quality. While recent advancements have significantly improved the parametrization of the so-called repulsive potential, the parametrization of the electronic part of the DFTB interaction remains relatively simplistic and underdeveloped. We present DFTB Slater-Koster Optimizer (DSKO), a novel framework that aims at producing accurate and transferable electronic parameter sets under rigorous physical constraints. Incorporating robust optimization algorithms and physics-informed loss functions, DSKO generates DFTB electronic parameters that yield electronic properties, such as density of states and band structures, closely matching DFT reference data. The versatility of DSKO facilitates the wide application of DFTB to materials science challenges, paving the way for routine high-fidelity semiempirical simulations.

2 October 2022 in arXiv
Verified
Preprint
Open access
PDF file
AKAlexey Korovin
Alexey Korovin
IHInnokentiy S. Humonen
Innokentiy S. Humonen
ASArtem I. Samtsevich
Artem I. Samtsevich
RERoman A. Eremin
Roman A. Eremin
AVArtem I. Vasilyev
Artem I. Vasilyev
Semen BudennyySemen Budennyy
Semen Budennyy
The discovery of new catalysts is one of the significant topics of computational chemistry as it has the potential to accelerate the adoption of renewable energy sources. Recently developed deep learning approaches such as graph neural networks (GNNs) open new opportunity to significantly extend scope for modelling novel high-performance catalysts. Nevertheless, the graph representation of particular crystal structure is not a straightforward task due to the ambiguous connectivity schemes and numerous embeddings of nodes and edges. Here we present embedding improvement for GNN that has been modified by Voronoi tesselation and is able to predict the energy of catalytic systems within Open Catalyst Project dataset. Enrichment of the graph was calculated via Voronoi tessellation and the corresponding contact solid angles and types (direct or indirect) were considered as features of edges and Voronoi volumes were used as node characteristics. The auxiliary approach was enriching node representation by intrinsic atomic properties (electronegativity, period and group position). Proposed modifications allowed us to improve the mean absolute error of the original model and the final error equals to 651 meV per atom on the Open Catalyst Project dataset and 6 meV per atom on the intermetallics dataset. Also, by consideration of additional dataset, we show that a sensible choice of data can decrease the error to values above physically-based 20 meV per atom threshold.

17 April 2026 in The Journal of Physical Chemistry C
Verified
Journal Article
Open access
MPMateusz Pokora
Mateusz Pokora
JGJakub Goclon
Jakub Goclon
JMJohannes T. Margraf
Johannes T. Margraf
CPChiara Panosetti
Chiara Panosetti
ASArtem Samtsevych
Artem Samtsevych
KWKrzysztof Winkler
Krzysztof Winkler
Semilocal density functional theory (DFT) and semiemprical GFN1/GFN2-xTB, as well as selected DFTB/3ob calculations, were performed to examine acidity in metal–organic frameworks (MOFs) using pyridine (Py) as a probe molecule. We studied Py adsorption on both Lewis and Bro̷nsted acid sites in ten distinct MOF structures. Our study provides detailed insight into their structures, thermodynamic properties, and vibrational spectra in selected ranges. All geometrical and energetic parameters obtained at the GFN1/GFN2-xTB levels were compared with those calculated using DFT. The results suggest that GFN1-xTB can only be used as a low-cost substitute for DFT for some systems. The adsorption energy prediction usually leads to unsatisfactory findings. GFN2-xTB and DFTB/3ob perform better than GFN1-xTB. Importantly, we identified areas where the GFN1/GFN2-xTB methods could be improved to yield more accurate results. By analyzing the density overlap region indicator (DORI) versus the adsorption energies, we revealed the nature and type of interactions in Py@MOFs.

8 December 2025 in ChemRxiv
Verified
Preprint
Open access
PDF file
ASArtem Samtsevych
Artem Samtsevych
YSYihua Song
Yihua Song
THTammo van der Heide
Tammo van der Heide
BABalint Aradi
Balint Aradi
BHBenjamin Hourahine
Benjamin Hourahine
CPChiara Panosetti
Chiara Panosetti
Density Functional Tight-Binding (DFTB) offers a computationally efficient alternative to {\em ab initio} methods, bridging between the accuracy of DFT and the speed of semiempirical models. The approximate nature of DFTB makes its reliability highly dependent on parameter quality though. While recent advancements have significantly improved the parametrization of the so-called repulsive potential, the parametrization of the electronic part of the DFTB interaction remains relatively simplistic and underdeveloped. We present our in-house DFTB Slater-Koster Optimizer (DSKO), a novel framework that aims at producing accurate and transferable electronic parameter sets under rigorous physical constraints. Incorporating robust optimization algorithms and physics-informed loss functions, DSKO generates DFTB electronic parameters that yield electronic properties, such as density of states and band structures, closely matching DFT reference data. The versatility of DSKO facilitates the wide application of DFTB to materials science challenges, paving the way for routine high-fidelity semiempirical simulations.