Signatures of Conical Intersection Dynamics in the Time-Resolved Photoelectron Spectrum of Furan: Theoretical Modeling with an Ensemble Density Functional Theory Method.

The non-adiabatic dynamics of furan excited in the ππ* state (S2 in the Franck-Condon geometry) was studied using non-adiabatic molecular dynamics simulations in connection with an ensemble density functional method. The time-resolved photoelectron spectra were theoretically simulated in a wide range of electron binding energies that covered the valence as well as the core electrons. The dynamics of the decay (rise) of the photoelectron signal were compared with the excited-state population dynamics. It was observed that the photoelectron signal decay parameters at certain electron binding energies displayed a good correlation with the events occurring during the excited-state dynamics. Thus, the time profile of the photoelectron intensity of the K-shell electrons of oxygen (decay constant of 34 ± 3 fs) showed a reasonable correlation with the time of passage through conical intersections with the ground state (47 ± 2 fs). The ground-state recovery constant of the photoelectron signal (121 ± 30 fs) was in good agreement with the theoretically obtained excited-state lifetime (93 ± 9 fs), as well as with the experimentally estimated recovery time constant (ca. 110 fs). Hence, it is proposed to complement the traditional TRPES observations with the trXPS (or trNEXAFS) measurements to obtain more reliable estimates of the most mechanistically important events during the excited-state dynamics.

Low-dimensional projection approach for efficient sampling of molecular recognition and polymer aggregation.

The one-dimensional projection (ODP) approach is extended to two-dimensional umbrella sampling (TDUS) and is applied to three different complex systems in combination with a reactive force field (ReaxFF). TDUS is capable of showing detailed features of the free-energy surface (FES) of the double-proton transfer of the acetic acid dimer. It also revealed the direct relationship between the types of hydrogen bonding and binding strengths in the case of adrenaline molecular recognition by SIVSF (Serine, Isoleucine, Valine, Cysteine, and Phenylalanine). The study of polymer aggregation using TDUS shows that aggregation is preferred with a less-polar solvent, which is also consistent with the experimental observation of a tape-casting process. Therefore, TDUS can be generally useful in FES explorations from simple chemical reactions to complex processes of molecular recognition and polymer aggregation.

Structural or population dynamics: what is revealed by the time-resolved photoelectron spectroscopy of 1,3-cyclohexadiene? A study with an ensemble density functional theory method.

Time-resolved photoelectron spectra during the photochemical ring-opening reaction of 1,3-cyclohexadiene (CHD) are modeled by an ensemble density functional theory (eDFT) method. The computational methodology employed in this work is capable of correctly describing the multi-reference effects arising in the ground and excited electronic states of molecules, which is important for the correct description of the ring-opening reaction of CHD. The geometries of molecular species along the non-adiabatic molecular dynamics (NAMD) trajectories reported in a previous study of the CHD photochemical ring-opening were used in this work to calculate the ionization energies and the respective Dyson orbitals for all possible ionization channels. The obtained theoretical time-resolved spectra display decay characteristics in a reasonable agreement with the experimental observations; i.e., the decay (and rise) of the most mechanistically significant signals occurs on the timescale of 100-150 fs. This is very different from the excited state population decay characteristics (τS1 = 234 ± 8 fs) obtained in the previous NAMD study. The difference between the population decay and the decay of the photoelectron signal intensity is traced back to the geometric transformation that the molecule undergoes during the photoreaction. This demonstrates the importance of including the geometric information in interpretation of the experimental observations.

Computation of Molecular Electron Affinities Using an Ensemble Density Functional Theory Method.

The computation of electron attachment energies (electron affinities) was implemented in connection with an ensemble density functional theory method, the state-interaction state-averaged spin-restricted ensemble-referenced Kohn-Sham (SI-SA-REKS or SSR) method. With the use of the extended Koopmans' theorem, the electron affinities and the respective Dyson orbitals are obtained directly for the neutral molecule, thus avoiding the necessity to compute the ionized system. Together with the EKT-SSR (extended Koopmans' theorem-SSR) method for ionization potentials, which was developed earlier, EKT-SSR for electron affinities completes the implementation of the EKT-SSR formalism, which can now be used for obtaining electron detachment as well as the electron attachment energies of molecules in the ground and excited electronic states. The extended EKT-SSR method was tested in the calculation of several closed-shell molecules. For the molecules in the ground states, the EKT-SSR energies of Dyson's orbitals are virtually identical to the energies of the unoccupied orbitals in the usual single-reference spin-restricted Kohn-Sham calculations. For the molecules in the excited states, EKT-SSR predicts an increase of the most positive electron affinity by approximately the amount of the vertical excitation energy. The electron affinities of a number of diradicals were calculated with EKT-SSR and compared with the available experimental data. With the use of a standard density functional (BH&HLYP), the EKT-SSR electron affinities deviate on average by ca. 0.2 eV from the experimental data. It is expected that the agreement with the experiment can be improved by designing density functionals parametrized for ionization energies.

Recent developments in the general atomic and molecular electronic structure system.

A discussion of many of the recently implemented features of GAMESS (General Atomic and Molecular Electronic Structure System) and LibCChem (the C++ CPU/GPU library associated with GAMESS) is presented. These features include fragmentation methods such as the fragment molecular orbital, effective fragment potential and effective fragment molecular orbital methods, hybrid MPI/OpenMP approaches to Hartree-Fock, and resolution of the identity second order perturbation theory. Many new coupled cluster theory methods have been implemented in GAMESS, as have multiple levels of density functional/tight binding theory. The role of accelerators, especially graphical processing units, is discussed in the context of the new features of LibCChem, as it is the associated problem of power consumption as the power of computers increases dramatically. The process by which a complex program suite such as GAMESS is maintained and developed is considered. Future developments are briefly summarized.

Development of a new parameter optimization scheme for a reactive force field based on a machine learning approach.

Reactive molecular dynamics (MD) simulation is performed using a reactive force field (ReaxFF). To this end, we developed a new method to optimize the ReaxFF parameters based on a machine learning approach. This approach combines the k-nearest neighbor and random forest regressor algorithm to efficiently locate several possible ReaxFF parameter sets. As a pilot test of the developed approach, the optimized ReaxFF parameter set was applied to perform chemical vapor deposition (CVD) of an α-Al2 O3 crystal. The crystal structure of α-Al2 O3 was reasonably reproduced even at a relatively high temperature (2000 K). The reactive MD simulation suggests that the (11 2 ¯ 0) surface grows faster than the (0001) surface, indicating that the developed parameter optimization technique could be used for understanding the chemical reaction in the CVD process. © 2019 Wiley Periodicals, Inc.

Simulations of infrared and Raman spectra in solution using the fragment molecular orbital method.

Analytic second derivatives of the energy with respect to nuclear coordinates are derived for the fragment molecular orbital method combined with the polarizable continuum model. Harmonic frequencies, infrared intensities and normal Raman activities of large molecular systems in solution can be evaluated. Periodic trends on SN2 chemical reactions are elucidated. The accuracy of the developed method is established in comparison to full calculations without fragmentation. The method is applied to ionic liquids and crambin (PDB: ). Solvent effects on the vibrational frequencies are discussed.

Efficient implementations of analytic energy gradient for mixed-reference spin-flip time-dependent density functional theory (MRSF-TDDFT).

Analytic energy gradients of individual singlet and triplet states with respect to nuclear coordinates are derived and implemented for the collinear mixed-reference spin-flip time-dependent density functional theory (MRSF-TDDFT), which eliminates the problematic spin-contamination of SF-TDDFT. Dimensional-transformation matrices for the singlet and triplet response spaces are introduced, simplifying the subsequent derivations. These matrices enable the general forms of MRSF-TDDFT equations to be similar to those of SF-TDDFT, suggesting that the computational overhead of singlet or triplet states for MRSF-TDDFT is nearly identical to that of SF-TDDFT. In test calculations, the new MRSF-TDDFT yields quite different optimized structures and energies as compared to SF-TDDFT. These differences turned out to mainly come from the spin-contamination of SF-TDDFT, which are largely cured by MRSF-TDDFT. In addition, it was demonstrated that the clear separation of singlet states from triplets dramatically simplifies the location of minimum energy conical intersection. As a result, it is clear that the MRSF-TDDFT has advantages over SF-TDDFT in terms of both accuracy and practicality. Therefore, it can be a preferred method, which is readily applied to other "black-box" type applications, such as the minimum-energy optimization, reaction path following, and molecular dynamics simulations.

Analytic second derivatives for the efficient electrostatic embedding in the fragment molecular orbital method.

The analytic second derivatives of the energy with respect to nuclear coordinates are developed for restricted Hartree-Fock and density functional theory, based on the two-body fragment molecular orbital method (FMO) and combined with the electrostatic embedding potential, self-consistently determined by point charges for far separated fragments and electron densities for near fragments. The accuracy of the method is established with respect to FMO using the exact embedding potential based on electron densities and to full calculations without fragmentation. The computational efficiency of parallelization is measured on the K supercomputer and the method is applied to simulate infrared spectra of two proteins, Trp-cage (PDB: 1L2Y) and crambin (1CRN). The nature of the vibrations in the Amide I peak of crambin and the Tyr symmetric stretch peak in Trp-cage are analyzed in terms of localized vibrations. © 2018 Wiley Periodicals, Inc.

Efficient Geometry Optimization of Large Molecular Systems in Solution Using the Fragment Molecular Orbital Method.

The analytic gradient is derived for the frozen domain formulation of the fragment molecular orbital (FMO) method combined with the polarizable continuum model. The accuracy is tested in comparison to full FMO calculations for a representative set of systems in terms of the gradient accuracy, protein-ligand binding energies, and optimized structures. The frozen domain method reproduced geometries optimized with full FMO within 0.03-0.09 Å in terms of reduced mean square deviations, whereas a single-point gradient calculation is accelerated by the factor of 38 (Trp-cage protein in explicit solvent, PDB: 1L2Y ) and 12 (crambin, PDB: 1CRN ). The method is applied to a geometry optimization of the K-Ras protein-ligand complex (4Q03) using two domain definitions, and the optimized structures are consistent with experiment. Pair interaction analysis is used to identify residues important in binding the ligand.

Analytic second derivative of the energy for density-functional tight-binding combined with the fragment molecular orbital method.

The analytic second derivative of the energy is developed for the fragment molecular orbital (FMO) method combined with density-functional tight-binding (DFTB), enabling simulations of infrared and Raman spectra of large molecular systems. The accuracy of the method is established in comparison to full DFTB without fragmentation for a set of representative systems. The performance of the FMO-DFTB Hessian is discussed for molecular systems containing up to 10 041 atoms. The method is applied to the study of the binding of α-cyclodextrin to polyethylene glycol, and the calculated IR spectrum of an epoxy amine oligomer reproduces experiment reasonably well.

Analytic second derivative of the energy for density functional theory based on the three-body fragment molecular orbital method.

Analytic second derivatives of the energy with respect to nuclear coordinates have been developed for spin restricted density functional theory (DFT) based on the fragment molecular orbital method (FMO). The derivations were carried out for the three-body expansion (FMO3), and the two-body expressions can be obtained by neglecting the three-body corrections. Also, the restricted Hartree-Fock (RHF) Hessian for FMO3 can be obtained by neglecting the density-functional related terms. In both the FMO-RHF and FMO-DFT Hessians, certain terms with small magnitudes are neglected for computational efficiency. The accuracy of the FMO-DFT Hessian in terms of the Gibbs free energy is evaluated for a set of polypeptides and water clusters and found to be within 1 kcal/mol of the corresponding full (non-fragmented) ab initio calculation. The FMO-DFT method is also applied to transition states in SN2 reactions and for the computation of the IR and Raman spectra of a small Trp-cage protein (PDB: 1L2Y). Some computational timing analysis is also presented.

Efficient molecular dynamics simulations of multiple radical center systems based on the fragment molecular orbital method.

The fully analytic energy gradient has been developed and implemented for the restricted open-shell Hartree-Fock (ROHF) method based on the fragment molecular orbital (FMO) theory for systems that have multiple open-shell molecules. The accuracy of the analytic ROHF energy gradient is compared with the corresponding numerical gradient, illustrating the accuracy of the analytic gradient. The ROHF analytic gradient is used to perform molecular dynamics simulations of an unusual open-shell system, liquid oxygen, and mixtures of oxygen and nitrogen. These molecular dynamics simulations provide some insight about how triplet oxygen molecules interact with each other. Timings reveal that the method can calculate the energy gradient for a system containing 4000 atoms in only 6 h. Therefore, it is concluded that the FMO-ROHF method will be useful for investigating systems with multiple open shells.

Unrestricted density functional theory based on the fragment molecular orbital method for the ground and excited state calculations of large systems.

We extended the fragment molecular orbital (FMO) method interfaced with density functional theory (DFT) into spin unrestricted formalism (UDFT) and developed energy gradients for the ground state and single point excited state energies based on time-dependent DFT. The accuracy of FMO is evaluated in comparison to the full calculations without fragmentation. Electronic excitations in solvated organic radicals and in the blue copper protein, plastocyanin (PDB code: 1BXV), are reported. The contributions of solvent molecules to the electronic excitations are analyzed in terms of the fragment polarization and quantum effects such as interfragment charge transfer.

Analytic second derivatives of the energy in the fragment molecular orbital method.

We developed the analytic second derivatives of the energy for the fragment molecular orbital (FMO) method. First we derived the analytic expressions and then introduced some approximations related to the first and second order coupled perturbed Hartree-Fock equations. We developed a parallel program for the FMO Hessian with approximations in GAMESS and used it to calculate infrared (IR) spectra and Gibbs free energies and to locate the transition states in SN2 reactions. The accuracy of the Hessian is demonstrated in comparison to ab initio results for polypeptides and a water cluster. By using the two residues per fragment division, we achieved the accuracy of 3 cm(-1) in the reduced mean square deviation of vibrational frequencies from ab initio for all three polyalanine isomers, while the zero point energy had the error not exceeding 0.3 kcal/mol. The role of the secondary structure on IR spectra, zero point energies, and Gibbs free energies is discussed.

Analytic Gradient for Time-Dependent Density Functional Theory Combined with the Fragment Molecular Orbital Method.

The analytic energy gradient of energy with respect to nuclear coordinates is derived for the fragment molecular orbital (FMO) method combined with time-dependent density functional theory (TDDFT). The response terms arising from the use of a polarizable embedding are derived. The obtained analytic FMO-TDDFT gradient is shown to be accurate in comparison to both numerical FMO-TDDFT and unfragmented TDDFT gradients, at the level of two- and three-body expansions. The gradients are used for geometry optimizations, molecular dynamics, vibrational calculations, and simulations of IR and Raman spectra of excited states. The developed method is used to optimize the geometry of the ground and excited electronic states of the photoactive yellow protein (PDB: 2PHY).

Toward Accurate Prediction of Ion Mobility in Organic Semiconductors by Atomistic Simulation.

A multiscale scheme (MLMS: Multi-Level Multi-Scale) to predict the ion mobility (µ) of amorphous organic semiconductors is proposed, which was successfully applied to the hole mobility predictions of 14 organic systems. An inverse relationship between µ and reorganization energy is observed due to local polaronic distortions. Another moderate inverse correlation between µ and distribution of site energy change exists, representing the effects of geometric flexibility. The former and the latter represent the intramolecular and intermolecular geometric effects, respectively. In addition, a linear correlation between transfer coupling and µ is observed, showing the importance of orbital overlaps between monomers. Especially, the highest hole mobility of C6-2TTN is due to its large transfer coupling. On the other hand, another high hole mobility of CBP turned out to come from the high first neighbor density (ρFND) of its first self-solvation, emphasizing the proper description of amorphous structural configurations with a sufficiently large number of monomers. In general, systems with either unusually high transfer coupling or high first neighbor density can potentially have high µ regardless of geometric effects. Especially, the newly suggested design parameter, ρFND, is pointing to a new direction as opposed to the traditional π-conjugation strategy.

New reaction model for O-O bond formation and O2 evolution catalyzed by dinuclear manganese complex.

A new mechanism of the oxygen evolving reaction catalyzed by [H(2)O(terpy)Mn(µ-O)(2)Mn(terpy)OH(2)](3+) is proposed by using density functional theory. This proton coupled electron transfer (PCET) model shows reasonable barriers. Because in experiments excess oxidants (OCl(-) or HSO(5)(-)) are required to evolve oxygen from water, we considered the Mn(2) complex neutralized by three counterions. Structure optimization made the coordinated OCl(-) withdraw a H(+) from the water ligand and produces the reaction space for H(2)O(2) formation with the deprotonated OH(-) ligand. The reaction barrier for the H(2)O(2) formation from OH(-) and protonated OCl(-) depends significantly on the system charge and is 14.0 kcal/mol when the system is neutralized. The H(2)O(2) decomposes to O(2) during two PCET processes to the Mn(2) complex, both with barriers lower than 12.0 kcal/mol. In both PCET processes the spin moment of transferred electrons prefers to be parallel to that of Mn 3d electrons because of the exchange interaction. This model thus explains how the triplet O(2) molecule is produced.

Unrestricted Hartree-Fock based on the fragment molecular orbital method: energy and its analytic gradient.

A consideration of the surrounding environment is necessary for a meaningful analysis of the reaction activity in large molecular systems. We propose an approach to perform unrestricted Hartree-Fock (UHF) calculations within the framework of the fragment molecular orbital (FMO) method (FMO-UHF) to study large systems with unpaired electrons. Prior to an energy analysis one has to optimize geometry, which requires an accurate analytic energy gradient. We derive the FMO-UHF energy and its analytic gradient and implement them into GAMESS. The performance of FMO-UHF is evaluated for a solvated organic molecule and a solvated metal complex, as well as for the active part of a protein, in terms of energy, gradient, and geometry optimization.

Accelerated Deep Learning Dynamics for Atomic Layer Deposition of Al(Me)3 and Water on OH/Si(111).

Knowledge of the detailed mechanism behind the atomic layer deposition (ALD) can greatly facilitate the optimization of the manufacturing process. Computational modeling can potentially foster the understanding; however, the presently available capabilities of the accurate ab initio computational techniques preclude their application to modeling surface processes occurring on a long time scale, such as ALD. Although the situation can be greatly improved using machine learning (ML), this technique requires an enormous amount of data for training datasets. Here, we propose an iterative protocol for optimizing ML training datasets and apply ML-assisted ab initio calculations to model surface reactions occurring during the Al(Me)3/H2O ALD process on the OH-terminated Si (111) surface. The protocol uses a recently developed low-dimensional projection technique (TDUS), greatly reducing the amount of information required to achieve high accuracy (ca. 1 kcal/mol or less) of the developed ML models. The resulting free energy landscapes reveal fine details of various aspects of the target ALD process, such as the surface proton transfer, zwitterionic surface configurations, elimination-addition/addition-elimination, and SN2 reactions as well as the role of the surface entropic and temperature effects. Simulations of adsorption dynamics predict that the maximum physisorption rate of ca. 70% is achieved at the incidence velocity urms of the reactants in the range of 15-20 Å/ps. Hence, the proposed protocol furnishes a very effective tool to study complex chemical reaction dynamics at a much reduced computational cost.