Ab Initio Quality Research Articles

Toolkit for the Construction of Reproducing Kernel-Based Representations of Data: Application to Multidimensional Potential Energy Surfaces.

In the early days of computation, slow processor speeds limited the amount of data that could be generated and used for scientific purposes. In the age of big data, the limiting factor usually is the method with which large amounts of data are analyzed and useful information is extracted. A typical example from chemistry are high-level ab initio calculations for small systems, which have nowadays become feasible even if energies at many different geometries are required. Molecular dynamics simulations often require several thousand distinct trajectories to be run. Under such circumstances suitable analytical representations of potential energy surfaces (PESs) based on ab initio calculations are required to propagate the dynamics at an acceptable cost. In this work we introduce a toolkit which allows the automatic construction of multidimensional PESs from gridded ab initio data based on reproducing kernel Hilbert space (RKHS) theory. The resulting representations require no tuning of parameters and allow energy and force evaluations at ab initio quality at the same cost as empirical force fields. Although the toolkit is primarily intended for constructing multidimensional potential energy surfaces for molecular systems, it can also be used for general machine learning purposes. The software is published under the MIT license and can be downloaded, modified, and used in other projects for free.

Ab Initio Quality NMR Parameters in Solid-State Materials Using a High-Dimensional Neural-Network Representation.

Nuclear magnetic resonance (NMR) spectroscopy is one of the most powerful experimental tools to probe the local atomic order of a wide range of solid-state compounds. However, due to the complexity of the related spectra, in particular for amorphous materials, their interpretation in terms of structural information is often challenging. These difficulties can be overcome by combining molecular dynamics simulations to generate realistic structural models with an ab initio evaluation of the corresponding chemical shift and quadrupolar coupling tensors. However, due to computational constraints, this approach is limited to relatively small system sizes which, for amorphous materials, prevents an adequate statistical sampling of the distribution of the local environments that is required to quantitatively describe the system. In this work, we present an approach to efficiently and accurately predict the NMR parameters of very large systems. This is achieved by using a high-dimensional neural-network representation of NMR parameters that are calculated using an ab initio formalism. To illustrate the potential of this approach, we applied this neural-network NMR (NN-NMR) method on the (17)O and (29)Si quadrupolar coupling and chemical shift parameters of various crystalline silica polymorphs and silica glasses. This approach is, in principal, general and has the potential to be applied to predict the NMR properties of various materials.

The modified quasi-quantum treatment of rotationally inelastic NO(X)–He scattering

A modified quasi-quantum treatment (MQQT) of molecular scattering has been developed to account for the softness of the repulsive part of the anisotropic atom-molecule PES. A contour of the PES is chosen such that the barrier height is just large enough to reflect the incoming kinetic energy, directed anti-parallel to the hard shell normal at the site of impact. The resulting rotationally inelastic quantum state resolved DCSs and ICSs of He + NO(X) at Ecol = 508 cm(-1) are compared to those obtained from regular QQT and from quantum mechanically exact calculations performed on the full highest quality ab initio Vsum PES. The MQQT parity changing DCSs for Δj ≤ 4 exhibit much better agreement with the QM DCSs than is obtained using regular QQT, particularly in the forward scattered direction. The improvements upon the remaining MQQT DCSs with respect to the regular QQT were minor, due to the near incompressible hard shell character of the n ≠ 1 or 3 anisotropic Legendre polynomial terms of the PES.

MoD-QM/MM Structural Refinement Method: Characterization of Hydrogen Bonding in the Oxytricha nova G-Quadruplex.

A generalization of the Moving-Domain Quantum Mechanics/Molecular Mechanics (MoD-QM/MM) hybrid method [Gascon, J. A.; Leung, S. S. F.; Batista, E. R.; Batista, V. S. J. Chem. Theory Comput. 2006, 2, 175-186] is introduced to provide a self-consistent computational protocol for structural refinement of extended systems. The method partitions the system into molecular domains that are iteratively optimized as quantum mechanical (QM) layers embedded in their surrounding molecular environment to obtain an ab initio quality description of the geometry and the molecular electrostatic potential of the extended system composed of those constituent fragments. The resulting methodology is benchmarked as applied to model systems that allow for full QM optimization as well as through refinement of the hydrogen bonding geometry in Oxytricha nova guanine quadruplex for which several studies have been reported, including the X-ray structure and NMR data. Calculations of (1)H NMR chemical shifts based on the gauge independent atomic orbital (GIAO) method and direct comparisons with experiments show that solvated MoD-QM/MM structures, sampled from explicit solvent molecular dynamics simulations, allow for NMR simulations in much improved agreement with experimental data than models based on the X-ray structure or those optimized using classical molecular mechanics force fields.

Influence of Surface Symmetry on the Onset of Nitrogen Eley–Rideal Recombination on Tungsten

Classical trajectory simulations, using potential energy surfaces of ab initio quality, are performed to investigate the influence of crystal symmetry on the Eley–Rideal abstraction dynamics of N atoms colliding, under normal incidence, N-preadsorbed tungsten (100) and (110) surfaces. Low-energy reactivity (<0.5 eV collision energy) results much higher for the (110) crystallographic plane. Such a feature stems from the topology of the interactions upon approach of the N-gas atom toward the N adsorbate: the strong lateral corrugation responsible for a significant threshold for recombination on the (100) surface is much smoother on the (110) plane, allowing low-energy incident atoms to react. Temperature is found to only slightly affect reactivity.

Nuclear Quantum Effects in Water: A Multiscale Study.

We outline a method to investigate the role of nuclear quantum effects in liquid water making use of a force field derived from ab initio simulations. Starting from a first-principles molecular dynamics simulation, we obtain an effective force field for bulk liquid water using the force-matching technique. After validating that our effective model reproduces the key structural and dynamic properties of the reference system, we use it to perform path integral simulations to investigate the role played by nuclear quantum effects on bulk water, probing radial distribution functions, vibrational spectra, and hydrogen bond fluctuations. Our approach offers a practical route to derive ab initio quality molecular models to study quantum effects at a low computational cost.

Efficient N2 Formation on Ag(111) by Eley–Rideal Recombination of Hyperthermal Atoms

Using molecular dynamics simulations and potential energy surfaces of ab initio quality, we show that direct pickup of N adsorbates by gas-phase N is a highly efficient channel for N2 formation on Ag(111). This recombination process, called Eley–Rideal, was traditionally associated with lighter projectiles and regarded as marginal, but here, we obtain reactivities for N of ≳35% even at incident energies of a few eV. The good agreement found between simulated and published experimental energies of the desorbed N2 is a fingerprint of this otherwise elusive recombination.

Transport Properties of Water and Sodium Dodecyl Sulfate

In this work, results from atomistic molecular dynamics studies investigating the effect of surfactant concentration on the transport properties of bulk surfactant aqueous solutions, focusing on the anionic surfactant sodium dodecyl sulfate (SDS), are reported. The surfactant self-diffusion and the thermal conductivity of bulk aqueous SDS solutions were computed at a range of concentrations at room and boiling temperatures. Additionally, MP2f (Akin-Ojo et al., 2008, “Developing Ab Initio Quality Force Fields From Condensed Phase Quantum-Mechanics/Molecular-Mechanics Calculations Through the Adaptive Force Matching Method,” J. Phys. Chem., 129, p. 064108), one of a new generation water potentials is assessed for its suitability in reproducing the transport and thermal properties of bulk water. The thermal conductivity of MP2f water model was found to be: 0.64 W/(m⋅K) at 298 K and 0.66 W/(m⋅K) at 373 K, in much better agreement with the experimental values compared to both the rigid and the flexible TIP3P water model.

Fast and accurate generation of ab initio quality atomic charges using nonparametric statistical regression

We introduce a class of partial atomic charge assignment method that provides ab initio quality description of the electrostatics of bioorganic molecules. The method uses a set of models that neither have a fixed functional form nor require a fixed set of parameters, and therefore are capable of capturing the complexities of the charge distribution in great detail. Random Forest regression is used to build separate charge models for elements H, C, N, O, F, S, and Cl, using training data consisting of partial charges along with a description of their surrounding chemical environments; training set charges are generated by fitting to the b3lyp/6-31G* electrostatic potential (ESP) and are subsequently refined to improve consistency and transferability of the charge assignments. Using a set of 210 neutral, small organic molecules, the absolute hydration free energy calculated using these charges in conjunction with Generalized Born solvation model shows a low mean unsigned error, close to 1 kcal/mol, from the experimental data. Using another large and independent test set of chemically diverse organic molecules, the method is shown to accurately reproduce charge-dependent observables--ESP and dipole moment--from ab initio calculations. The method presented here automatically provides an estimate of potential errors in the charge assignment, enabling systematic improvement of these models using additional data. This work has implications not only for the future development of charge models but also in developing methods to describe many other chemical properties that require accurate representation of the electronic structure of the system.

A Minimal Implementation of the AMBER-PAICS Interface for Ab Initio FMO-QM/MM-MD Simulation

Abstract For the purpose of providing a realistic description of the reaction mechanisms in large molecular systems, we propose a quantum mechanical/molecular mechanical (QM/MM) method combined with the ab initio fragment molecular orbital (FMO) method, i.e., the ab initio FMO-QM/MM method. By connecting a molecular dynamics (MD) program AMBER with an FMO program PAICS, we have implemented an AMBER-PAICS interface (AP-IF). Using the AP-IF, we demonstrate three example applications: (a) a hydrogen fluoride and water molecular clusters, (b) an alanine dipeptide in aqueous solution, and (c) a prion protein–GN8 complex. From these results, it is confirmed that the FMO-QM/MM method offers a good compromise between chemical accuracy and computational cost and enables us to obtain in ab initio quality the inter- and intramolecular interaction energies between molecules or residues in large molecular systems such as solution and biomolecule, by using the dynamics-based interfragment interaction energy (IFIE) analysis.

Neural network interatomic potential for the phase change material GeTe

GeTe is a prototypical phase change material of high interest for applications in optical and electronic non-volatile memories. We present an interatomic potential for the bulk phases of GeTe, which is created using a neural network (NN) representation of the potential-energy surface obtained from reference calculations based on density functional theory. It is demonstrated that the NN potential provides a close to ab initio quality description of a number of properties of liquid, crystalline and amorphous GeTe. The availability of a reliable classical potential allows addressing a number of issues of interest for the technological applications of phase change materials, which are presently beyond the capability of first principles molecular dynamics simulations.

Structure and Vibrational Motion of Insulin from Raman Optical Activity Spectra

The Raman optical activity (ROA) spectroscopic technique has been applied in the past to many biologically relevant systems including peptides, proteins, sugars, and even viruses. However, theoretical interpretation of the spectra relies on lengthy quantum-chemical computations, which are difficult to extend to larger molecules. In the present study, ROA and Raman spectra of insulin under a range of various conditions were measured and interpreted with the aid of the Cartesian-coordinate tensor transfer (CCT) method. The CCT methodology yielded spectra of insulin monomer and dimer of nearly ab initio quality, while at the same time reproducing the experimental data very well. The link between the spectra and the protein structure could thus be studied in detail. Spectral contributions from the peptide backbone and the amino acid side chains were calculated. Likewise, specific intensity features originating from the α-helical, coil, β-sheet, and 3(10)-helical parts of the protein could be deciphered. The assignment of the Raman and ROA bands to intrinsic molecular coordinates as based on the harmonic force field calculation revealed their origin and degree of locality. Alternatively, the relation of the structural flexibility of insulin to the inhomogeneous broadening of spectral bands was studied by a combination of CCT and molecular dynamics (MD). The present study confirms the sensitivity of the ROA technique to some subtle static and dynamic changes in molecular geometry, and many previous ad hoc or semiempirical spectral-structure assignments could be verified. On the other hand, a limitation in longer-range tertiary structure sensitivity was revealed. Unlike for smaller molecules with approximately equal contributions of the electric dipole (α), quadrupole (A), and magnetic dipole (G') polarizabilities, only the electric dipolar polarization (α) interactions seem to dominate in the protein ROA signal. The simulations concern the largest molecule for which such spectra were interpreted by a priori procedures and significantly enhance protein folding studies undertaken by this technique.

Nucleation mechanism for the direct graphite-to-diamond phase transition

Graphite and diamond have comparable free energies, yet forming diamond from graphite in the absence of a catalyst requires pressures that are significantly higher than those at equilibrium coexistence. At lower temperatures, the formation of the metastable hexagonal polymorph of diamond is favoured instead of the more stable cubic diamond. These phenomena cannot be explained by the concerted mechanism suggested in previous theoretical studies. Using an ab initio quality neural-network potential, we carried out a large-scale study of the graphite-to-diamond transition assuming that it occurs through nucleation. The nucleation mechanism accounts for the observed phenomenology and reveals its microscopic origins. We demonstrate that the large lattice distortions that accompany the formation of diamond nuclei inhibit the phase transition at low pressure, and direct it towards the hexagonal diamond phase at higher pressure. The proposed nucleation mechanism should improve our understanding of structural transformations in a wide range of carbon-based materials.

Theoretical treatment of the electronic excited states of the DMSO molecule: A challenge for current theoretical methods

We present here a computational study of the vertical excitation energies of the dimethylsulfoxide (DMSO) molecule. This molecule is the simplest among the sulfoxides and it is here used as a model for testing the performance of current high quality ab initio computations and to asses the possibility of studying larger molecules containing the sulfoxide group. We shall present a detailed comparison of the results obtained by different methodologies with recent experimental data.

Ab initioquality neural-network potential for sodium

An interatomic potential for high-pressure high-temperature (HPHT) crystalline and liquid phases of sodium is created using a neural-network (NN) representation of the ab initio potential energy surface. It is demonstrated that the NN potential provides an ab initio quality description of multiple properties of liquid sodium and bcc, fcc, cI16 crystal phases in the P-T region up to 120 GPa and 1200 K. The unique combination of computational efficiency of the NN potential and its ability to reproduce quantitatively experimental properties of sodium in the wide P-T range enables molecular dynamics simulations of physicochemical processes in HPHT sodium of unprecedented quality.

On the Importance of Lennard–Jones Parameter Calibration in QM/MM Framework: Reaction Path Tracing via Free Energy Gradient Method for Ammonia Ionization Process in Aqueous Solution

Abstract The free energy gradient (FEG) method combined with the QM/MM–MD calculation has realized full-atomic structural optimization of the solute NH3–H2O molecule pair in aqueous solution, exploring successful tracing of the reaction path on the free energy surface from the neutral state (H3N–H2O) to the ionized (H4+N–O−H) through the transition state (H3N···H+···O−H). To reproduce accurately the structural change of the hydrated molecule pair, the LJ parameter calibration in the QM/MM nonelectrostatic interaction was essential for four kinds of QM solute species, i.e., OH−, H2O, NH3, and NH4+, treated in the NDDO-SSRP framework with the ab initio quality. Two linear interpolations of these calibrated LJ parameters optimized at the NSS and ISS were utilized in the path-dependent FEG calculation, as is common in the free energy perturbation treatment. After including the effects of the intramolecular entropic contribution and quantum tunneling, the free energies of activation and reaction were satisfactorily estimated to be 13.8 and 9.8 kcal mol−1 (1 kcal = 4.184 kJ), respectively. It is expected that far more agreement with the experimental values should be reasonably attained if the QM region were extended to three surrounding water molecules to take in the electron delocalization effect adequately.

Graphite-diamond phase coexistence study employing a neural-network mapping of theab initiopotential energy surface

An interatomic potential for the diamond and graphite phases of carbon has been created using a neural-network (NN) representation of the ab initio potential energy surface. The NN potential combines the accuracy of a first-principles description of both phases with the efficiency of empirical force fields and allows one to perform a molecular-dynamics study, of ab initio quality, of the thermodynamics of graphite-diamond coexistence. Good agreement between the experimental and calculated coexistence curves is achieved if nuclear quantum effects are included in the simulation.

Novel Computational Methods for Nanostructure Electronic Structure Calculations

Experimentally relevant nanocrystals often contain a few thousands to hundreds of thousands of atoms. Yet, to understand their electronic structures, surface and impurity effects, atomic relaxations, interior electric fields, carrier dynamics, and transports, it is often necessary to carry out atomistic simulations. Owing to the advance of recent algorithm developments and improved supercomputer powers, it is now possible to calculate such nanocrystals based on ab initio methods. In this review, we discuss the numerical algorithms (the plane-wave pseudopotential method and the real-space finite-difference method) used in conventional density-functional-theory calculations, which enable the simulations of systems up to one or two thousand atoms. We also introduce methods designed specifically for nanostructure calculations. These methods [the charge-patching method (CPM) and the linear scaling three-dimensional fragment method (LS3DF)] can be used to calculate systems with hundreds of thousands of atoms. Whereas CPM is an approximation with ab initio quality, the LS3DF method is an O(N) method with essentially the same results as the direct methods. The computational aspects of the algorithms, especially for their parallelization scalability, are also emphasized in the review.

Potential Energy Surfaces Fitted by Artificial Neural Networks

Molecular mechanics is the tool of choice for the modeling of systems that are so large or complex that it is impractical or impossible to model them by ab initio methods. For this reason there is a need for accurate potentials that are able to quickly reproduce ab initio quality results at the fraction of the cost. The interactions within force fields are represented by a number of functions. Some interactions are well understood and can be represented by simple mathematical functions while others are not so well understood and their functional form is represented in a simplistic manner or not even known. In the last 20 years there have been the first examples of a new design ethic, where novel and contemporary methods using machine learning, in particular, artificial neural networks, have been used to find the nature of the underlying functions of a force field. Here we appraise what has been achieved over this time and what requires further improvements, while offering some insight and guidance for the development of future force fields.

Revised electrostatics from invariom refinement of the 18-residue peptaibol antibiotic trichotoxin A50E

Detailed functional understanding of the action of biological macromolecules is reliant on high-quality, accurate data describing their atomic and electrostatic structural properties. We present the application of cutting-edge modeling techniques to the 18-residue peptaibol antibiotic trichotoxin A50E [J. K. Chugh, H. Bruckner and B. A. Wallace, Biochemistry, 2002, 41, 12934] for which low temperature X-ray data (T = 150 K) at subatomic resolution were retrieved from the Protein Databank. Use of aspherical scattering factors (invarioms) enables refinement of individual coordinates and anisotropic displacement parameters unconstrained and standard uncertainties can be estimated by inversion of the full least-square matrix. However, application of the invariom method to model the electron density of trichotoxin A50E has not only improved the structural model in terms of geometry, figures of merit and anisotropic displacement parameters, but also results in the direct calculation of an ab initio quality electrostatic potential and associated dipole moment. Disagreement with previously published electrostatics and possible functional implications are discussed.