SCINE-Software for chemical interaction networks.

The software for chemical interaction networks (SCINE) project aims at pushing the frontier of quantum chemical calculations on molecular structures to a new level. While calculations on individual structures as well as on simple relations between them have become routine in chemistry, new developments have pushed the frontier in the field to high-throughput calculations. Chemical relations may be created by a search for specific molecular properties in a molecular design attempt, or they can be defined by a set of elementary reaction steps that form a chemical reaction network. The software modules of SCINE have been designed to facilitate such studies. The features of the modules are (i) general applicability of the applied methodologies ranging from electronic structure (no restriction to specific elements of the periodic table) to microkinetic modeling (with little restrictions on molecularity), full modularity so that SCINE modules can also be applied as stand-alone programs or be exchanged for external software packages that fulfill a similar purpose (to increase options for computational campaigns and to provide alternatives in case of tasks that are hard or impossible to accomplish with certain programs), (ii) high stability and autonomous operations so that control and steering by an operator are as easy as possible, and (iii) easy embedding into complex heterogeneous environments for molecular structures taken individually or in the context of a reaction network. A graphical user interface unites all modules and ensures interoperability. All components of the software have been made available as open source and free of charge.

Lifelong Machine Learning Potentials.

Machine learning potentials (MLPs) trained on accurate quantum chemical data can retain the high accuracy, while inflicting little computational demands. On the downside, they need to be trained for each individual system. In recent years, a vast number of MLPs have been trained from scratch because learning additional data typically requires retraining on all data to not forget previously acquired knowledge. Additionally, most common structural descriptors of MLPs cannot represent efficiently a large number of different chemical elements. In this work, we tackle these problems by introducing element-embracing atom-centered symmetry functions (eeACSFs), which combine structural properties and element information from the periodic table. These eeACSFs are key for our development of a lifelong machine learning potential (lMLP). Uncertainty quantification can be exploited to transgress a fixed, pretrained MLP to arrive at a continuously adapting lMLP, because a predefined level of accuracy can be ensured. To extend the applicability of an lMLP to new systems, we apply continual learning strategies to enable autonomous and on-the-fly training on a continuous stream of new data. For the training of deep neural networks, we propose the continual resilient (CoRe) optimizer and incremental learning strategies relying on rehearsal of data, regularization of parameters, and the architecture of the model.

Insights into lithium manganese oxide-water interfaces using machine learning potentials.

Unraveling the atomistic and the electronic structure of solid-liquid interfaces is the key to the design of new materials for many important applications, from heterogeneous catalysis to battery technology. Density functional theory (DFT) calculations can, in principle, provide a reliable description of such interfaces, but the high computational costs severely restrict the accessible time and length scales. Here, we report machine learning-driven simulations of various interfaces between water and lithium manganese oxide (LixMn2O4), an important electrode material in lithium ion batteries and a catalyst for the oxygen evolution reaction. We employ a high-dimensional neural network potential to compute the energies and forces several orders of magnitude faster than DFT without loss in accuracy. In addition, a high-dimensional neural network for spin prediction is utilized to analyze the electronic structure of the manganese ions. Combining these methods, a series of interfaces is investigated by large-scale molecular dynamics. The simulations allow us to gain insights into a variety of properties, such as the dissociation of water molecules, proton transfer processes, and hydrogen bonds, as well as the geometric and electronic structure of the solid surfaces, including the manganese oxidation state distribution, Jahn-Teller distortions, and electron hopping.

Predicting oxidation and spin states by high-dimensional neural networks: Applications to lithium manganese oxide spinels.

Lithium ion batteries often contain transition metal oxides such as LixMn2O4 (0 ≤ x ≤ 2). Depending on the Li content, different ratios of MnIII to MnIV ions are present. In combination with electron hopping, the Jahn-Teller distortions of the MnIIIO6 octahedra can give rise to complex phenomena such as structural transitions and conductance. While for small model systems oxidation and spin states can be determined using density functional theory (DFT), the investigation of dynamical phenomena by DFT is too demanding. Previously, we have shown that a high-dimensional neural network potential can extend molecular dynamics (MD) simulations of LixMn2O4 to nanosecond time scales, but these simulations did not provide information about the electronic structure. Here, we extend the use of neural networks to the prediction of atomic oxidation and spin states. The resulting high-dimensional neural network is able to predict the spins of the Mn ions with an error of only 0.03 ℏ. We find that the Mn eg electrons are correctly conserved and that the number of Jahn-Teller distorted MnIIIO6 octahedra is predicted precisely for different Li loadings. A charge ordering transition is observed between 280 K and 300 K, which matches resistivity measurements. Moreover, the activation energy of the electron hopping conduction above the phase transition is predicted to be 0.18 eV, deviating only 0.02 eV from experiment. This work demonstrates that machine learning is able to provide an accurate representation of both the geometric and the electronic structure dynamics of LixMn2O4 on time and length scales that are not accessible by ab initio MD.

From Molecular Fragments to the Bulk: Development of a Neural Network Potential for MOF-5.

The development of first-principles-quality reactive atomistic potentials for organic-inorganic hybrid materials is still a substantial challenge because of the very different physics of the atomic interactions-from covalent via ionic bonding to dispersion-that have to be described in an accurate and balanced way. In this work we used a prototypical metal-organic framework, MOF-5, as a benchmark case to investigate the applicability of high-dimensional neural network potentials (HDNNPs) to this class of materials. In HDNNPs, which belong to the class of machine learning potentials, the energy is constructed as a sum of environment-dependent atomic energy contributions. We demonstrate that by the use of this approach it is possible to obtain a high-quality potential for the periodic MOF-5 crystal using density functional theory (DFT) reference calculations of small molecular fragments only. The resulting HDNNP, which has a root-mean-square error (RMSE) of 1.6 meV/atom for the energies of molecular fragments not included in the training set, is able to provide the equilibrium lattice constant of the bulk MOF-5 structure with an error of about 0.1% relative to DFT, and also, the negative thermal expansion behavior is accurately predicted. The total energy RMSE of periodic structures that are completely absent in the training set is about 6.5 meV/atom, with errors on the order of 2 meV/atom for energy differences. We show that in contrast to energy differences, achieving a high accuracy for total energies requires careful variation of the stoichiometries of the training structures to avoid energy offsets, as atomic energies are not physical observables. The forces, which have RMSEs of about 94 meV/ a0 for the molecular fragments and 130 meV/ a0 for bulk structures not included in the training set, are insensitive to such offsets. Therefore, forces, which are the relevant properties for molecular dynamics simulations, provide a realistic estimate of the accuracy of atomistic potentials.

A full additive QM/MM scheme for the computation of molecular crystals with extension to many-body expansions.

An additive quantum mechanics/molecular mechanics (QM/MM) model for the theoretical investigation of molecular crystals (AC-QM/MM) is presented. At the one-body level, a single molecule is chosen as the QM region. The MM region around it consists of a finite cluster of explicit MM atoms, represented by point charges and Lennard-Jones potentials, with additional background charges to mimic periodic electrostatics. Cluster charges are QM-derived and calculated self-consistently to ensure a polarizable embedding. We have also considered the extension to many-body QM corrections, calculating the interactions of a central molecule to neighboring units in the crystal. Full gradient expressions have been derived, also including symmetry information. The scheme allows for the calculation of molecular properties as well as unconstrained optimizations of the molecular geometry and cell parameters with respect to the lattice energy. Benchmarking the approach with the X23 reference set confirms the convergence pattern of the many-body extension although a comparison to plane-wave density functional theory reveals a systematic overestimation of cohesive energies by 6-16 kJ mol-1. While the scheme primarily aims to provide an inexpensive and flexible way to model a molecule in a crystal environment, it can also be used to reach highly accurate cohesive energies by the straightforward application of wave function correlated approaches. Calculations with local coupled cluster with singles, doubles, and perturbative triples, albeit limited to numerical gradients, show an impressive agreement with experimental estimates for small molecular crystals.

Strained hydrogen bonding in imidazole trimer: a combined infrared, Raman, and theory study.

In this work, a careful analysis of anharmonic couplings in NH and some CH stretch modes of imidazole is carried out. This includes IR and Raman spectra of the isolated molecule and aggregates up to the trimer, together with two different theoretical approaches to the calculation of anharmonic shifts and absolute band positions. The imidazole dimer is vibrationally characterized for the first time in vacuum isolation under supersonic jet conditions, showing substantial shifts from previous helium droplet experiments and evidence for Fermi resonance for the hydrogen-bonded NH stretch. The most stable imidazole trimer structure is unambiguously shown to be cyclic with three non-equivalent, highly strained hydrogen bonds. This contrasts the helium droplet observation of a chain trimer involving two unstrained hydrogen bonds. These experimental conclusions are strongly corroborated by theory, including vibrational perturbation theory and anharmonic normal mode analysis. Systematic error compensation in some of these methods is emphasized. Intramolecular anharmonic coupling constants from perturbation theory are validated by Raman hot band jet spectroscopy of the monomer. Imidazole aggregation is shown to provide valuable benchmarking opportunities for electronic structure and in particular for anharmonic vibrational methods, covering the field of strong and strongly distorted hydrogen bonding.

The Guinness molecules for the carbohydrate formula.

A systematic review and analysis of the most stable spatial arrangements of n carbon, n oxygen, and 2n hydrogen atoms including vibrational zero-point energy up to n = 5 shows that small-molecule aggregates win, typically followed by thermally unstable molecules, before kinetically stable molecules and finally carbohydrates are found. Near n ≈ 60 a crossover to carbon allotropes and ice as the global minimum structure is expected and the asymptotic limit is most likely graphite and ice. Implications for astrochemical and fermentation processes are discussed. Density functionals like B3LYPD3 are found to describe these energy sequences quite poorly, mostly due to an overestimated stability of carbon in high oxidation states.