Analytic gradient and molecular dynamics simulations using the fragment molecular orbital method combined with effective potentials

The completely analytic energy gradients are derived and implemented for the two-body fragment molecular orbital (FMO2) method combined with the model core potentials (MCP) and effective fragment potentials (EFP). The many-body terms in EFP require solving coupled-perturbed Hartree-Fock equations, which are derived and implemented. The molecular dynamics (MD) simulations are performed using the FMO2/MCP method for the capped alanine decamer and with the FMO2/EFP method for the zwitterionic conformer of glycine tetramer immersed in the water layer of 6.0 Å (135 water molecules). The results of the MD simulations using the FMO2/EFP and FMO2/MCP gradients show that the total energy is conserved at the time steps less than 1 fs.     

Optimizing conical intersections of solvated molecules: the combined spin-flip density functional theory/effective fragment potential method

Solvent effects on a potential energy surface crossing are investigated by optimizing a conical intersection (CI) in solution. To this end, the analytic energy gradient has been derived and implemented for the collinear spin-flip density functional theory (SFDFT) combined with the effective fragment potential (EFP) solvent model. The new method is applied to the azomethane-water cluster and the chromophore of green fluorescent protein in aqueous solution. These applications illustrate not only dramatic changes in the CI geometries but also strong stabilization of the CI in a polar solvent. Furthermore, the CI geometries obtained by the hybrid SFDFT/EFP scheme reproduce those by the full SFDFT, indicating that the SFDFT/EFP method is an efficient and promising approach for understanding nonadiabatic processes in solution.     

Fully analytic energy gradient in the fragment molecular orbital method

The Z-vector equations are derived and implemented for solving the response term due to the external electrostatic potentials, and the corresponding contribution is added to the energy gradients in the framework of the fragment molecular orbital (FMO) method. To practically solve the equations for large molecules like proteins, the equations are decoupled by taking advantage of the local nature of fragments in the FMO method and establishing the self-consistent Z-vector method. The resulting gradients are compared with numerical gradients for the test molecular systems: (H(2)O)(64), alanine decamer, hydrated chignolin with the protein data bank (PDB) ID of 1UAO, and a Trp-cage miniprotein construct (PDB ID: 1L2Y). The computation time for calculating the response contribution is comparable to or less than that of the FMO self-consistent charge calculation. It is also shown that the energy gradients for the electrostatic dimer approximation are fully analytic, which significantly reduces the computational costs. The fully analytic FMO gradient is parallelized with an efficiency of about 98% on 32 nodes.     

Electrostatic energy in the effective fragment potential method: theory and application to benzene dimer

Evaluation of the electrostatic energy within the effective fragment potential (EFP) method is presented. The performance of two variants of the distributed multipole analysis (DMA) together with two different models for estimating the charge penetration energies was studied using six homonuclear dimers. The importance of damping the higher order multipole terms, i.e. charge dipole, was also investigated. Damping corrections recover more than 70% of the charge penetration energy in all dimers, whereas higher order damping introduces only minor improvement. Electrostatic energies calculated by the numerical DMA are less accurate than those calculated by the analytic DMA. Analysis of bonding in the benzene dimer shows that EFP with inclusion of the electrostatic damping term performs very well compared to the high-level coupled cluster singles, doubles, and perturbative triples method. The largest error of 0.4 kcal/mol occurs for the sandwich dimer configuration. This error is about half the size of the corresponding error in second order perturbation theory. Thus, EFP in the current implementation is an accurate and computationally inexpensive method for calculating interaction energies in weakly bonded molecular complexes.     

Molecular-orbital-free algorithm for the excited-state force in time-dependent density functional theory

Starting from the equation of motion in the density matrix formulation, we reformulate the analytical gradient of the excited-state energy at the time-dependent density functional theory level in the nonorthogonal Gaussian atom-centered orbital (AO) basis. Analogous to the analytical first derivative in molecular-orbital (MO) basis, a Z-vector equation has been derived with respect to the reduced one-electronic density matrix in AO basis, which provides a potential possibility to exploit quantum locality of the density matrix and avoids the matrix transformation between the AO and the MO basis. Numerical tests are finished for the excited-state geometry optimization and adiabatic excitation energy calculation of a series of small molecules. The results demonstrate the computational efficiency and accuracy of the current AO-based energy gradient expression in comparison with the MO-based scheme.     

Analytic energy gradient in combined time-dependent density functional theory and polarizable force field calculation

Formulas for evaluating analytic energy gradient are derived for combined time-dependent density functional theory (TDDFT) and polarizable force field methods that incorporate dipole polarizability tensors and linearly induced point dipoles. The Z-vector method for determining relaxed one-particle difference density matrix in regular TDDFT methods is extended to include induced dipoles. The analytic gradient of the mutual polarization energy of the force field and the TDDFT excited state can be formulated by using the TDDFT difference density-induced dipoles and the transition state density-induced dipoles. All the forces and torques involving induced dipoles can be efficiently evaluated using standard electrostatic formulas as if the induced dipoles were permanent dipoles. The formulas are given in the most general form and are applicable to various flavors of polarizable force fields. Implementation and tests with a polarizable five-point water model show that the formulas are rigorous. The carbonyl vibration modes and infrared spectrum intensities of a cluster formed by acetone and two water molecules are studied.     

Polarizable continuum model with the fragment molecular orbital-based time-dependent density functional theory

The polarizable continuum model (PCM) for describing the solvent effect was combined with the fragment molecular orbital-based time-dependent density functional theory (TDDFT). Several levels of the many-body expansion were implemented, and the importance of the many-body contributions to the singlet-excited states was discussed. To calibrate the accuracy, we performed a number of the model calculations using our method and the regular TDDFT in solution, applying them to phenol and polypeptides at the long-range corrected BLYP/6-31G* level. It was found that for systems up to 192 atoms the largest error in the excitation energy was 0.006 eV (vs. the regular TDDFT/PCM of the full system). The solvent shifts and the conformer effects were discussed, and the scaling was found to be nearly linear. Finally, we applied our method to the lowest singlet excitation of the photoactive yellow protein (PYP) in aqueous solution and determined the excitation energy to be in reasonable agreement with experiment. The excitation energy analysis provided the contributions of individual residues, and the main factors as well as their solvent shifts were determined.     

Multithreaded parallelization of the energy and analytic gradient in the fragment molecular orbital method

The fragment molecular orbital method in GAMESS is parallelized in a multithreaded OpenMP implementation combined with the MPI version of the two-level generalized distributed data interface. The energy and analytic gradient in gas phase and the polarizable continuum model of solvation are parallelized in this hybrid three-level scheme, achieving a large memory footprint reduction and a high parallel efficiency on Intel Xeon Phi processors. The parallel efficiency is demonstrated on the Stampede2 and Theta supercomputers using up to 2048 nodes (262 144 threads).     

Real-time, local basis-set implementation of time-dependent density functional theory for excited state dynamics simulations

We present a method suitable for large-scale accurate simulations of excited state dynamics within the framework of time-dependent density functional theory (DFT). This is achieved by employing a local atomic basis-set representation and real-time propagation of excited state wave functions. We implement the method within SIESTA, a standard ground-state DFT package with local atomic basis, and demonstrate its potential for realistic and accurate excited state dynamics simulations using small and medium-sized molecules as examples (H(2), CO, O(3), and indolequinone). The method can be readily applied to problems involving nanostructures and large biomolecules.     

Time-dependent density functional theory based upon the fragment molecular orbital method

Time-dependent density functional theory (TDDFT) was combined with the two-body fragment molecular orbital method (FMO2). In this FMO2-TDDFT scheme, the system is divided into fragments, and the electron density for fragments is determined self-consistently. Consequently, only one main fragment of interest and several fragment pairs including it are calculated by TDDFT. To demonstrate the accuracy of FMO2-TDDFT, we computed several low-lying singlet and triplet excited states of solvated phenol and polyalanine using our method and the standard TDDFT for the full system. The BLYP functional with the long-range correction (LC-BLYP) was employed with the 6-31G(*) basis set (some tests were also performed with 6-311G(*), as well as with B3LYP and time-dependent Hartree-Fock). Typically, FMO2-TDDFT reproduced the full TDDFT excitation energies within 0.1 eV, and for one excited state the error was about 0.2 eV. Beside the accurate reproduction of the TDDFT excitation energies, we also automatically get an excitation energy decomposition analysis, which provides the contributions of individual fragments. Finally, the efficiency of our approach was exemplified on the LC-BLYP6-31G(*) calculation of the lowest singlet excitation of the photoactive yellow protein which consists of 1931 atoms, and the obtained value of 3.1 eV is in agreement with the experimental value of 2.8 eV.     

Quantum master equation method based on the broken-symmetry time-dependent density functional theory: application to dynamic polarizability of open-shell molecular systems

A novel method for the calculation of the dynamic polarizability (α) of open-shell molecular systems is developed based on the quantum master equation combined with the broken-symmetry (BS) time-dependent density functional theory within the Tamm-Dancoff approximation, referred to as the BS-DFTQME method. We investigate the dynamic α density distribution obtained from BS-DFTQME calculations in order to analyze the spatial contributions of electrons to the field-induced polarization and clarify the contributions of the frontier orbital pair to α and its density. To demonstrate the performance of this method, we examine the real part of dynamic α of singlet 1,3-dipole systems having a variety of diradical characters (y). The frequency dispersion of α, in particular in the resonant region, is shown to strongly depend on the exchange-correlation functional as well as on the diradical character. Under sufficiently off-resonant condition, the dynamic α is found to decrease with increasing y and/or the fraction of Hartree-Fock exchange in the exchange-correlation functional, which enhances the spin polarization, due to the decrease in the delocalization effects of π-diradical electrons in the frontier orbital pair. The BS-DFTQME method with the BHandHLYP exchange-correlation functional also turns out to semiquantitatively reproduce the α spectra calculated by a strongly correlated ab initio molecular orbital method, i.e., the spin-unrestricted coupled-cluster singles and doubles.     

Excited state geometry optimizations by analytical energy gradient of long-range corrected time-dependent density functional theory

An analytical excitation energy gradient of long-range corrected time-dependent density functional theory (LC-TDDFT) is presented. This is based on a previous analytical TDDFT gradient formalism, which avoids solving the coupled-perturbed Kohn-Sham equation for each nuclear degree of freedom. In LC-TDDFT, exchange interactions are evaluated by combining the short-range part of a DFT exchange functional with the long-range part of the Hartree-Fock exchange integral. This LC-TDDFT gradient was first examined by calculating the excited state geometries and adiabatic excitation energies of small typical molecules and a small protonated Schiff base. As a result, we found that long-range interactions play a significant role even in valence excited states of small systems. This analytical LC-TDDFT gradient was also applied to the investigations of small twisted intramolecular charge transfer (TICT) systems. By comparing with calculated ab initio multireference perturbation theory and experimental results, we found that LC-TDDFT gave much more accurate absorption and fluorescence energies of these systems than those of conventional TDDFTs using pure and hybrid functionals. For optimized excited state geometries, LC-TDDFT provided fairly different twisting and wagging angles of these small TICT systems in comparison with conventional TDDFT results.     

The Gaussian and augmented-plane-wave density functional method for ab initio molecular dynamics simulations

A new algorithm for density-functional-theory-based ab initio molecular dynamics simulations is presented. The Kohn–Sham orbitals are expanded in Gaussian-type functions and an augmented-plane-wave-type approach is used to represent the electronic density. This extends previous work of ours where the density was expanded only in plane waves. We describe the total density in a smooth extended part which we represent in plane waves as in our previous work and parts localised close to the nuclei which are expanded in Gaussians. Using this representation of the charge we show how the localised and extended part can be treated separately, achieving a computational cost for the calculation of the Kohn–Sham matrix that scales with the system size N as O(NlogN). Furthermore, we are able to reduce drastically the size of the plane-wave basis. In addition, we introduce a multiple-cutoff method that improves considerably the performance of this approach. Finally, we demonstrate with a series of numerical examples the accuracy and efficiency of the new algorithm, both for electronic structure calculations and for ab initio molecular dynamics simulations. Received: 15 December 1998 /Accepted: 18 February 1999 /Published online: 14 July 1999     

Analytical approach for the excited-state Hessian in time-dependent density functional theory: formalism, implementation, and performance

The paper presents the formalism, implementation, and performance of the analytical approach for the excited-state Hessian in the time-dependent density functional theory (TDDFT) that extends our previous work [J. Liu and W. Z. Liang, J. Chem. Phys. 135, 014113 (2011)] on the analytical Hessian in TDDFT within Tamm-Dancoff approximation (TDA) to full TDDFT. In contrast to TDA-TDDFT, an appreciable advantage of full TDDFT is that it maintains the oscillator strength sum rule, and therefore yields more precise results for the oscillator strength and other related physical quantities. For the excited-state harmonic vibrational frequency calculation, however, full TDDFT does not seem to be advantageous since the numerical tests demonstrate that the accuracy of TDDFT with and without TDA are comparable to each other. As a common practice, the computed harmonic vibrational frequencies are scaled by a suitable scale factor to yield good agreement with the experimental fundamental frequencies. Here we apply both the optimized ground-state and excited-state scale factors to scale the calculated excited-state harmonic frequencies and find that the scaling decreases the root-mean-square errors. The optimized scale factors derived from the excited-state calculations are slightly smaller than those from the ground-state calculations.     

Acidity constants from vertical energy gaps: density functional theory based molecular dynamics implementation

The question of how to compute acidity constants (pK(a)) treating solvent and solute at the same level of theory remains of some interest, for example in the case of high or low pH conditions. We have developed a density functional theory based molecular dynamics implementation of such a method. The method is based on a half reaction scheme computing free energies of dissociation from the vertical energy gaps for insertion or removal of protons. Finite system size effects are important, but largely cancel when half reactions are combined to full reactions. We verified the method by investigating a series of organic and inorganic acids and bases spanning a wide range of pK(a) values (20 units). We find that the response of the aqueous solvent to vertical protonation/deprotonation is almost always asymmetric and correlated with the strength of the hydrogen bonding of the deprotonated base. We interpret these observations in analogy with the picture of solvent response to electronic ionization.     

Fitting properties from density functional theory based molecular dynamics simulations to parameterize a rigid water force field

In the quest towards coarse-grained potentials and new water models, we present an extension of the force matching technique to parameterize an all-atom force field for rigid water. The methodology presented here allows to improve the matching procedure by first optimizing the weighting exponents present in the objective function. A new gauge for unambiguously evaluating the quality of the fit has been introduced; it is based on the root mean square difference of the distributions of target properties between reference data and fitted potentials. Four rigid water models have been parameterized; the matching procedure has been used to assess the role of the ghost atom in TIP4P-like models and of electrostatic damping. In the former case, burying the negative charge inside the molecule allows to fit better the torques. In the latter, since short-range interactions are damped, a better fit of the forces is obtained. Overall, the best performing model is the one with a ghost atom and with electrostatic damping. The approach shown in this paper is of general validity and could be applied to any matching algorithm and to any level of coarse graining, also for non-rigid molecules.     

Gradients of the polarization energy in the effective fragment potential method

The effective fragment potential (EFP) method is an ab initio based polarizable classical method in which the intermolecular interaction parameters are obtained from preparative ab initio calculations on isolated molecules. The polarization energy in the EFP method is modeled with asymmetric anisotropic dipole polarizability tensors located at the centroids of localized bond and lone pair orbitals of the molecules. Analytic expressions for the translational and rotational gradients (forces and torques) of the EFP polarization energy have been derived and implemented. Periodic boundary conditions (the minimum image convention) and switching functions have also been implemented for the polarization energy, as well as for other EFP interaction terms. With these improvements, molecular dynamics simulations can be performed with the EFP method for various chemical systems.     

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.     

Dynamical quantum-electrodynamics embedding: Combining time-dependent density functional theory and the near-field method

We develop an approach for dynamical (ω > 0) embedding of mixed quantum mechanical (QM)∕classical (or more precisely QM∕electrodynamics) systems with a quantum sub-region, described by time-dependent density functional theory (TDDFT), within a classical sub-region, modeled here by the recently proposed near-field (NF) method. Both sub-systems are propagated simultaneously and are coupled through a common Coulomb potential. As a first step we implement the method to study the plasmonic response of a metal film which is half jellium-like QM and half classical. The resulting response is in good agreement with both full-scale TDDFT and the purely classical NF method. The embedding method is able to describe the optical response of the whole system while capturing quantum mechanical effects, so it is a promising approach for studying electrodynamics in hybrid molecules-metals nanostructures.     

Noncovalent interactions in extended systems described by the effective fragment potential method: theory and application to nucleobase oligomers

The implementation of the effective fragment potential (EFP) method within the Q-CHEM electronic structure package is presented. The EFP method is used to study noncovalent π-π and hydrogen-bonding interactions in DNA strands. Since EFP is a computationally inexpensive alternative to high-level ab initio calculations, it is possible to go beyond the dimers of nucleic acid bases and to investigate the asymptotic behavior of different components of the total interaction energy. The calculations demonstrated that the dispersion energy is a leading component in π-stacked oligomers of all sizes. Exchange-repulsion energy also plays an important role. The contribution of polarization is small in these systems, whereas the magnitude of electrostatics varies. Pairwise fragment interactions (i.e., the sum of dimer binding energies) were found to be a good approximation for the oligomer energy.