Active-learning for global optimization of Ni-Ceria nanoparticles: The case of Ce4-xNixO8- x (x = 1, 2, 3).

Ni-CeO2 nanoparticles (NPs) are promising nanocatalysts for water splitting and water gas shift reactions due to the ability of ceria to temporarily donate oxygen to the catalytic reaction and accept oxygen after the reaction is completed. Therefore, elucidating how different properties of the Ni-Ceria NPs relate to the activity and selectivity of the catalytic reaction, is of crucial importance for the development of novel catalysts. In this work the active learning (AL) method based on machine learning regression and its uncertainty is used for the global optimization of Ce(4-x)NixO(8-x) (x = 1, 2, 3) nanoparticles, employing density functional theory calculations. Additionally, further investigation of the NPs by mass-scaled parallel-tempering Born-Oppenheimer molecular dynamics resulted in the same putative global minimum structures found by AL, demonstrating the robustness of our AL search to learn from small datasets and assist in the global optimization of complex electronic structure systems.

Reinforcement learning for in silico determination of adsorbate-substrate structures.

Reinforcement learning (RL) methods have helped to define the state of the art in the field of modern artificial intelligence, mostly after the breakthrough involving AlphaGo and the discovery of novel algorithms. In this work, we present a RL method, based on Q-learning, for the structural determination of adsorbate@substrate models in silico, where the minimization of the energy landscape resulting from adsorbate interactions with a substrate is made by actions on states (translations and rotations) chosen from an agent's policy. The proposed RL method is implemented in an early version of the reinforcement learning software for materials design and discovery (RLMaterial), developed in Python3.x. RLMaterial interfaces with deMon2k, DFTB+, ORCA, and Quantum Espresso codes to compute the adsorbate@substrate energies. The RL method was applied for the structural determination of (i) the amino acid glycine and (ii) 2-amino-acetaldehyde, both interacting with a boron nitride (BN) monolayer, (iii) host-guest interactions between phenylboronic acid and ß-cyclodextrin and (iv) ammonia on naphthalene. Density functional tight binding calculations were used to build the complex search surfaces with a reasonably low computational cost for systems (i)-(iii) and DFT for system (iv). Artificial neural network and gradient boosting regression techniques were employed to approximate the Q-matrix or Q-table for better decision making (policy) on next actions. Finally, we have developed a transfer-learning protocol within the RL framework that allows learning from one chemical system and transferring the experience to another, as well as from different DFT or DFTB levels.

How important is the amount of exact exchange for spin-state energy ordering in DFT? Case study of molybdenum carbide cluster, Mo4C2.

Since the form of the exact functional in density functional theory is unknown, we must rely on density functional approximations (DFAs). In the past, very promising results have been reported by combining semi-local DFAs with exact, i.e. Hartree-Fock, exchange. However, the spin-state energy ordering and the predictions of global minima structures are particularly sensitive to the choice of the hybrid functional and to the amount of exact exchange. This has been already qualitatively described for single conformations, reactions, and a limited number of conformations. Here, we have analyzed the mixing of exact exchange in exchange functionals for a set of several hundred isomers of the transition metal carbide, Mo4C2. The analysis of the calculated energies and charges using PBE0-type functional with varying amounts of exact exchange yields the following insights: (1) The sensitivity of spin-energy splitting is strongly correlated with the amount of exact exchange mixing. (2) Spin contamination is exacerbated when correlation is omitted from the exchange-correlation functional. (3) There is not one ideal value for the exact exchange mixing which can be used to parametrize or choose among the functionals. Calculated energies and electronic structures are influenced by exact exchange at a different magnitude within a given distribution; therefore, to extend the application range of hybrid functionals to the full periodic table the spin-energy splitting energies should be investigated.

QMLMaterial─A Quantum Machine Learning Software for Material Design and Discovery.

Structural elucidation of chemical compounds is challenging experimentally, and theoretical chemistry methods have added important insight into molecules, nanoparticles, alloys, and materials geometries and properties. However, finding the optimum structures is a bottleneck due to the huge search space, and global search algorithms have been used successfully for this purpose. In this work, we present the quantum machine learning software/agent for materials design and discovery (QMLMaterial), intended for automatic structural determination in silico for several chemical systems: atomic clusters, atomic clusters and the spin multiplicity together, doping in clusters or solids, vacancies in clusters or solids, adsorption of molecules or adsorbents on surfaces, and finally atomic clusters on solid surfaces/materials or encapsulated in porous materials. QMLMaterial is an artificial intelligence (AI) software based on the active learning method, which uses machine learning regression algorithms and their uncertainties for decision making on the next unexplored structures to be computed, increasing the probability of finding the global minimum with few calculations as more data is obtained. The software has different acquisition functions for decision making (e.g., expected improvement and lower confidence bound). Also, the Gaussian process is available in the AI framework for regression, where the uncertainty is obtained analytically from Bayesian statistics. For the artificial neural network and support vector regressor algorithms, the uncertainty can be obtained by K-fold cross-validation or nonparametric bootstrap resampling methods. The software is interfaced with several quantum chemistry codes and atomic descriptors, such as the many-body tensor representation. QMLMaterial's capabilities are highlighted in the current work by its applications in the following systems: Na20, Mo6C3 (where the spin multiplicity was considered), H2O@CeNi3O5, Mg8@graphene, Na3Mg3@CNT (carbon nanotube).

GAMaterial-A genetic-algorithm software for material design and discovery.

Genetic algorithms (GAs) are stochastic global search methods inspired by biological evolution. They have been used extensively in chemistry and materials science coupled with theoretical methods, ranging from force-fields to high-throughput first-principles methods. The methodology allows an accurate and automated structural determination for molecules, atomic clusters, nanoparticles, and solid surfaces, fundamental to understanding chemical processes in catalysis and environmental sciences, for instance. In this work, we propose a new genetic algorithm software, GAMaterial, implemented in Python3.x, that performs global searches to elucidate the structures of atomic clusters, doped clusters or materials and atomic clusters on surfaces. For all these applications, it is possible to accelerate the GA search by using machine learning (ML), the ML@GA method, to build subsequent populations. Results for ML@GA applied for the dopant distributions in atomic clusters are presented. The GAMaterial software was applied for the automatic structural search for the Ti6 O12 cluster, doping Al in Si11 (4Al@Si11 ) and Na10 supported on graphene (Na10 @graphene), where DFTB calculations were used to sample the complex search surfaces with reasonably low computational cost. Finally, the global search by GA of the Mo8 C4 cluster was considered, where DFT calculations were made with the deMon2k code, which is interfaced with GAMaterial.

Automatic structural elucidation of vacancies in materials by active learning.

Finding the optimum structures of non-stoichiometric or berthollide materials, such as (1D, 2D, 3D) materials or nanoparticles (0D), is challenging due to the huge chemical/structural search space. Computational methods coupled with global optimization algorithms have been used successfully for this purpose. In this work, we have developed an artificial intelligence method based on active learning (AL) or Bayesian optimization for the automatic structural elucidation of vacancies in solids and nanoparticles. AL uses machine learning regression algorithms and their uncertainties to take decisions (from a policy) on the next unexplored structures to be computed, increasing the probability of finding the global minimum with few calculations. The methodology allows an accurate and automated structural elucidation for vacancies, which are common in non-stoichiometric (berthollide) materials, helping to understand chemical processes in catalysis and environmental sciences, for instance. The AL vacancies method was implemented in the quantum machine learning software/agent for material design and discovery (QMLMaterial). Also, two additional acquisition functions for decision making were implemented, besides the expected improvement (EI): the lower confidence bound (LCB) and the probability of improvement (PI). The new software was applied for the automatic structural search for graphite (C36) with 3 (C36-3) and 4 (C36-4) carbon vacancies and C60 (C60-4) fullerene with 4 carbon vacancies. DFTB calculations were used to build the complex search surfaces with reasonably low computational cost. Furthermore, with the AL method for vacancies, it was possible to elucidate the optimum oxygen vacancy distribution in CaTiO3 perovskite by DFT, where a semiconductor behavior results from oxygen vacancies. Throughout the work, a Gaussian process with its uncertainty was employed in the AL framework using different acquisition functions (EI, LCB and PI), and taking into account different descriptors: Ewald sum matrix and sine matrix. Finally, the performance of the proposed AL method was compared to random search and genetic algorithm.

A new active learning approach for adsorbate-substrate structural elucidation in silico.

Adsorbate interactions with substrates (e.g. surfaces and nanoparticles) are fundamental for several technologies, such as functional materials, supramolecular chemistry, and solvent interactions. However, modeling these kinds of systems in silico, such as finding the optimum adsorption geometry and energy, is challenging, due to the huge number of possibilities of assembling the adsorbate on the surface. In the current work, we have developed an artificial intelligence (AI) approach based on an active learning (AL) method for adsorption optimization on the surface of materials. AL uses machine learning (ML) regression algorithms and their uncertainties to make a decision (based on a policy) for the next unexplored structures to be computed, increasing, though, the probability of finding the global minimum with a small number of calculations. The methodology allows an accurate and automated structural elucidation of the adsorbate on the surface, based on the minimization of the total electronic energy. The new AL method for adsorption optimization was developed and implemented in the quantum machine learning software/agent for material design and discovery (QMLMaterial) program and was applied for C60@TiO2 anatase (101). It marks another software extension with a new feature in addition to the automatic structural elucidation of defects in materials and of nanoparticles as well. SCC-DFTB calculations were used to build the complex search surfaces with a reasonably low computational cost. An artificial neural network (NN) was employed in the AL framework evaluated together with two uncertainty quantification methods: K-fold cross-validation and non-parametric bootstrap (BS) resampling. Also, two different acquisition functions for decision-making were used: expected improvement (EI) and the lower confidence bound (LCB).

Selective binding of choline by a phosphate-coordination-based triple helicate featuring an aromatic box.

In nature, proteins have evolved sophisticated cavities tailored for capturing target guests selectively among competitors of similar size, shape, and charge. The fundamental principles guiding the molecular recognition, such as self-assembly and complementarity, have inspired the development of biomimetic receptors. In the current work, we report a self-assembled triple anion helicate (host 2) featuring a cavity resembling that of the choline-binding protein ChoX, as revealed by crystal and density functional theory (DFT)-optimized structures, which binds choline in a unique dual-site-binding mode. This similarity in structure leads to a similarly high selectivity of host 2 for choline over its derivatives, as demonstrated by the NMR and fluorescence competition experiments. Furthermore, host 2 is able to act as a fluorescence displacement sensor for discriminating choline, acetylcholine, L-carnitine, and glycine betaine effectively.The choline-binding protein ChoX exhibits a synergistic dual-site binding mode that allows it to discriminate choline over structural analogues. Here, the authors design a biomimetic triple anion helicate receptor whose selectivity for choline arises from a similar binding mechanism.

Accurate DFT-D3 Calculations in a Small Basis Set.

Calculations of interaction energies of noncovalent interactions in small basis sets are affected by the basis set superposition error and dispersion-corrected DFT-D methods and are thus usually parametrized only for triple-ζ and larger basis sets. Nevertheless, some smaller basis sets could also perform well. Among many combinations tested, we obtained excellent results with the DZVP-DFT basis and newly parametrized D3 dispersion correction. The accuracy of interaction energies and geometries is close to significantly more expensive calculations.

Unraveling the Structure-Affinity Relationship between Cucurbit[n]urils (n = 7, 8) and Cationic Diamondoids.

We report the measurement of the binding constants (Ka) for cucurbit[n]uril (n = 7, 8) toward four series of guests based on 2,6-disubstituted adamantanes, 4,9-disubstituted diamantanes, 1,6-disubstituted diamantanes, and 1-substituted adamantane ammonium ions by direct and competitive 1H NMR spectroscopy. Compared to the affinity of CB[7]·Diam(NMe3)2, the adamantane diammonium ion complexes (e.g., CB[7]·2,6-Ad(NH3)2 and CB[7]·2,6-Ad(NMe3)2) are less effective at realizing the potential 1000-fold enhancement in affinity due to ion-dipole interactions at the second ureidyl C═O portal. Comparative crystallographic investigation of CB[7]·Diam(NMe3)2, CB[7]·DiamNMe3, and CB[7]·1-AdNMe3 revealed that the preferred geometry positions the +NMe3 groups ≈0.32 Å above the C═O portal; the observed 0.80 Å spacing observed for CB[7]·Diam(NMe3)2 reflects the simultaneous geometrical constraints of CH2···O═C close contacts at both portals. Remarkably, the CB[8]·IsoDiam(NHMe2)2 complex displays femtomolar binding affinity, placing it firmly alongside the CB[7]·Diam(NMe3)2 complex. Primary or quaternary ammonium ion looping strategies lead to larger increases in binding affinity for CB[8] than for CB[7], which we attribute to the larger size of the carbonyl portals of CB[8]; this suggests routes to develop CB[8] as the tightest binding host in the CB[n] family. We report that alkyl group fluorination (e.g., CB[7]·1-AdNH2Et versus CB[7]·1-AdNH2CH2CF3) does not result in the expected increase in Ka value. Finally, we discuss the role of solvation in nonempirical quantum mechanical computational methodology, which is used to estimate the relative changes in Gibbs binding free energies.

A Nexus between Theory and Experiment: Non-Empirical Quantum Mechanical Computational Methodology Applied to Cucurbit[n]uril⋠Guest Binding Interactions.

A training set of eleven X-ray structures determined for biomimetic complexes between cucurbit[n]uril (CB[7 or 8]) hosts and adamantane-/diamantane ammonium/aminium guests were studied with DFT-D3 quantum mechanical computational methods to afford ΔGcalcd binding energies. A novel feature of this work is that the fidelity of the BLYP-D3/def2-TZVPP choice of DFT functional was proven by comparison with more accurate methods. For the first time, the CB[n]⋅guest complex binding energy subcomponents [for example, ΔEdispersion , ΔEelectrostatic , ΔGsolvation , binding entropy (-TΔS), and induced fit Edeformation(host) , Edeformation(guest) ] were calculated. Only a few weeks of computation time per complex were required by using this protocol. The deformation (stiffness) and solvation properties (with emphasis on cavity desolvation) of cucurbit[n]uril (n=5, 6, 7, 8) isolated host molecules were also explored by means of the DFT-D3 method. A high ρ2 =0.84 correlation coefficient between ΔGexptl and ΔGcalcd was achieved without any scaling of the calculated terms (at 298 K). This linear dependence was utilized for ΔGcalcd predictions of new complexes. The nature of binding, including the role of high energy water molecules, was also studied. The utility of introduction of tethered [-(CH2 )n NH3 ]+ amino loops attached to N,N-dimethyl-adamantane-1-amine and N,N,N',N'-tetramethyl diamantane-4,9-diamine skeletons (both from an experimental and a theoretical perspective) is presented here as a promising tool for the achievement of new ultra-high binding guests to CB[7] hosts. Predictions of not yet measured equilibrium constants are presented herein.

Representative Amino Acid Side-Chain Interactions in Protein-DNA Complexes: A Comparison of Highly Accurate Correlated Ab Initio Quantum Mechanical Calculations and Efficient Approaches for Applications to Large Systems.

Representative pairs of amino acid side chains and nucleic acid bases extracted from available high-quality structures of protein-DNA complexes were analyzed using a range of computational methods. CCSD(T)/CBS interaction energies were calculated for the chosen 272 pairs. These reference interaction energies were used to test the MP2.5/CBS, MP2.X/CBS, MP2-F12, DFT-D3, PM6, and Amber force field methods. Method MP2.5 provided excellent agreement with reference data (root-mean-square error (RMSE) of 0.11 kcal/mol), which is more than 1 order of magnitude faster than the CCSD(T) method. When MP2-F12 and MP2.5 were combined, the results were within reasonable accuracy (0.20 kcal/mol), with a computational savings of almost 2 orders of magnitude. Therefore, this method is a promising tool for accurate calculations of interaction energies in protein-DNA motifs of up to ∼100 atoms, for which CCSD(T)/CBS benchmark calculations are not feasible. B3-LYP-D3 calculated with def2-TZVPP and def2-QZVP basis sets yielded sufficiently good results with a reasonably small RMSE. This method provided better results for neutral systems, whereas positively charged species exhibited the worst agreement with the benchmark data. The Amber force field yielded unbalanced results-performing well for systems containing nonpolar amino acids but severely underestimating interaction energies for charged complexes. The semiempirical PM6 method with corrections for hydrogen bonding and dispersion energy (PM6-D3H4) exhibited considerably smaller error than the Amber force field, which makes it an effective tool for modeling extended protein-ligand complexes (of up to 10,000 atoms).

Correction: The strength and directionality of a halogen bond are co-determined by the magnitude and size of the σ-hole.

Correction for 'The strength and directionality of a halogen bond are co-determined by the magnitude and size of the σ-hole' by Michal Koláret al., Phys. Chem. Chem. Phys., 2014, 16, 9987-9996.

Influence of hydrophobic residues on the binding of CB[7] toward diammonium ions of common ammonium···ammonium distance.

We report the binding constants of CB[7] toward a series of naphthalene diammonium and 4,4'-dipiperidinium derivatives and compare the results with those obtained previously for CB[7]·3b by (1)H NMR and X-ray crystallography. The nature of binding in the host·guest complexes was investigated using quantum mechanical tools.

Large-Scale Quantitative Assessment of Binding Preferences in Protein-Nucleic Acid Complexes.

The growing number of high-quality experimental (X-ray, NMR) structures of protein­DNA complexes has sufficient enough information to assess whether universal rules governing the DNA sequence recognition process apply. While previous studies have investigated the relative abundance of various modes of amino acid­base contacts (van der Waals contacts, hydrogen bonds), relatively little is known about the energetics of these noncovalent interactions. In the present study, we have performed the first large-scale quantitative assessment of binding preferences in protein­DNA complexes by calculating the interaction energies in all 80 possible amino acid­DNA base combinations. We found that several mutual amino acid­base orientations featuring bidentate hydrogen bonds capable of unambiguous one-to-one recognition correspond to unique minima in the potential energy space of the amino acid­base pairs. A clustering algorithm revealed that these contacts form a spatially well-defined group offering relatively little conformational freedom. Various molecular mechanics force field and DFT-D ab initio calculations were performed, yielding similar results.

The strength and directionality of a halogen bond are co-determined by the magnitude and size of the σ-hole.

The σ-holes of halogen atoms on various aromatic scaffolds were described in terms of their size and magnitude. The electrostatic potential maps at the CAM-B3LYP-D3(bj)/def2-QZVP level were calculated and the σ-holes of more than 100 aromatic analogues were thoroughly analysed to relate the σ-holes to the binding preferences of the halogenated compounds. Both the size and magnitude of the σ-hole increase when passing from chlorinated to iodinated analogues. Also, the σ-hole properties were studied upon chemical substitution of the aromatic ring as well as in the aromatic ring. Further, the angular variations of the interactions were investigated on a selected set of halogenbenzene complexes with argon and hydrogen fluoride (HF). In order to analyse interaction energy components, DFT-SAPT angular scans were performed. The interaction energies of bromobenzene complexes were evaluated at the CCSD(T)/complete basis set level providing the benchmark energetic data. The strength of the halogen bond between halogenbenzenes and Ar atoms and HF molecules increases while its directionality decreases when passing from chlorine to iodine. The decrease of the directionality of the halogen bond is larger for a HF-containing complex and is caused by electrostatic and exchange-repulsion energies. These findings are especially valuable for protein-halogenated ligand-binding studies, applied in the realm of rational drug development and lead optimisation.