Projected Commutator DIIS Method for Accelerating Hybrid Functional Electronic Structure Calculations.

SQDFT: Spectral Quadrature method for large-scale parallel [formula omitted] Kohn–Sham calculations at high temperature

We present a subspace projection technique to conduct large-scale Kohn-Sham density functional theory calculations using spectral finite-element discretization. The proposed method treats both metallic and insulating materials in a single framework, and is applicable to both pseudopotential as well as all-electron calculations. The key ideas involved in the method include: (i) employing a higher-order spectral finite-element basis that is amenable to mesh adaption; (ii) using a Chebyshev filter to construct a subspace which is an approximation to the occupied eigenspace in a given self-consistent field iteration; (iii) using a localization procedure to construct a non-orthogonal localized basis spanning the Chebyshev filtered subspace; (iv) using a Fermi-operator expansion in terms of the subspace-projected Hamiltonian represented in the non-orthogonal localized basis to compute relevant quantities like the density matrix, electron density and band energy. We demonstrate the accuracy and efficiency of the approach on benchmark systems involving pseudopotential calculations on metallic aluminum nano-clusters up to 3430 atoms and on insulating alkane chains up to 7052 atoms, as well as all-electron calculations on silicon nano-clusters up to 3920 electrons. The benchmark studies revealed that accuracies commensurate with chemical accuracy can be obtained, and a subquadratic-scaling with system size was observed for the range of materials systems studied. In particular, for the alkane chains---close to linear-scaling is observed, whereas, for aluminum nano-clusters---the scaling is observed to be $\mathcal{O} (N^{1.46})$. For all-electron calculations on silicon nano-clusters, the scaling with the number of electrons is computed to be $\mathcal{O} (N^{1.75})$. Furthermore, significant computational savings have been realized with the proposed approach with respect to reference calculations.

Alkaline earth dimers have small bond energies (less than 5 kcal/mol) that provide a difficult challenge for electronic structure calculations. They are especially challenging for Kohn-Sham density functional theory (KS-DFT) using generalized gradient approximations (GGAs) as the exchange-correlation density functional because GGAs often do not provide accurate results for weak interactions. Here we treat alkaline earth dimers from six different rows of the periodic table. We show that the dominant correlating configurations are the same in all six dimers. We also show that multiconfiguration pair-density functional theory (MC-PDFT) using a fully translated GGA as the on-top density functional not only performs much better than KS-DFT with GGAs in predicting equilibrium distances and dissociation energies but also performs better than the more computationally demanding complete active space second-order perturbation theory (CASPT2) with large basis sets and performs even better than CASPT2 with smaller basis sets.

A hybrid analytical-numerical integration scheme is introduced to accelerate the Fock build in self-consistent field (SCF) and time-dependent density functional theory (TDDFT) calculations. To evaluate the Coulomb matrix J[D], the density matrix D is first decomposed into two parts, the superposition of atomic density matrices D⊕A and the rest DR = D-D⊕A. While J[D⊕A] is evaluated analytically, J[DR] is evaluated fully numerically [with the multipole expansion of the Coulomb potential (MECP)] during the SCF iterations. Upon convergence, DR is further split into those of near (DRC) and distant (DRL) atomic orbital (AO) pairs, such that J[DRC] and J[DRL] are evaluated seminumerically and fully numerically (with MECP). Such a hybrid J-build is dubbed "analytic-MECP" (aMECP). Likewise, the analytic evaluation of K[D⊕A] and seminumerical evaluation of K[DR] are also invoked for the construction of the exchange matrix K[D] during the SCF iterations. The chain-of-spheres (COSX) algorithm [Chem. Phys. 356, 98 (2009]) is employed for K[DR] but with a revised construction of the S-junctions for overlap AO pairs. To distinguish from the original COSX algorithm (which does not involve the partition of the density matrix D), we denote the presently revised variant as COSx. Upon convergence, DR is further split into those of near (DRC) and distant (DRL) AO pairs followed by a rescaling, leading to and , respectively. and are then evaluated analytically and seminumerically (with COSx), respectively. Such a hybrid K-build is dubbed "analytic-COSx" (aCOSx). Extensive numerical experimentations reveal that the combination of aMECP and aCOSx is highly accurate for ground state SCF calculations ( error in energy) and is particularly efficient for calculations of large molecules with extended basis sets. As for TDDFT excitation energies, a medium grid for MECP and a coarse grid for COSx are already sufficient.

We explore a new formalism to study the nonlinear electronic density response based on Kohn–Sham density functional theory (KS-DFT) at partially and strongly quantum degenerate regimes. It is demonstrated that the KS-DFT calculations are able to accurately reproduce the available path integral Monte Carlo simulation results at temperatures relevant for warm dense matter research. The existing analytical results for the quadratic and cubic response functions are rigorously tested. It is demonstrated that the analytical results for the quadratic response function closely agree with the KS-DFT data. Furthermore, the performed analysis reveals that currently available analytical formulas for the cubic response function are not able to describe simulation results, neither qualitatively nor quantitatively, at small wavenumbers q < 2qF, with qF being the Fermi wavenumber. The results show that KS-DFT can be used to describe warm dense matter that is strongly perturbed by an external field with remarkable accuracy. Furthermore, it is demonstrated that KS-DFT constitutes a valuable tool to guide the development of the nonlinear response theory of correlated quantum electrons from ambient to extreme conditions. This opens up new avenues to study nonlinear effects in a gamut of different contexts at conditions that cannot be accessed with previously used path integral Monte Carlo methods.

The vibrational self-consistent field (VSCF) method yields anharmonic states and spectra for molecular vibrations, and it serves as the starting point for more sophisticated correlated-vibration methods. Convergence of the iterative, non-linear optimization in VSCF calculations can be erratic or altogether unsuccessful, particularly for chemical systems involving low-frequency motions. In this work, a vibrational formulation of the Direct Inversion of the Iterative Subspace method of Pulay is presented and investigated. This formulation accounts for distinct attributes of the vibrational and electronic cases, including the expansion of each single-mode vibrational wavefunction in its own basis set. The resulting Direct Inversion of the Iterative Subspace method is shown to substantially accelerate VSCF convergence in all convergent cases as well as rectify many cases where Roothaan-based methods fail. Performance across systems ranging from small, rigid molecules to weakly bound molecular clusters is investigated in this analysis.

The accuracy of non-resonant and resonant (resonant inelastic X-ray scattering) X-ray emission spectra simulated based upon Kohn-Sham density functional theory is assessed. Accurate non-resonant X-ray emission spectra with the correct energy scale are obtained when short-range corrected exchange-correlation functionals designed for the calculation of X-ray absorption spectroscopy are used. It is shown that this approach can be extended to simulate resonant inelastic X-ray scattering by using a reference determinant that describes a core-excited state. For this spectroscopy, it is found that a standard hybrid functional, B3LYP, gives accurate spectra that reproduce the features observed in experiment. However, the ability to correctly describe subtle changes in the spectra arising from different intermediate states is more challenging and requires averaging over conformations from a molecular dynamics simulation. Overall, it is demonstrated that accurate non-resonant and resonant X-ray emission spectra can be simulated directly from Kohn-Sham density functional theory.

The structures, binding energies, and enthalpies of small molecular clusters incorporating a single lithium cation and up through six waters have been determined with extended Gaussian basis sets using Hartree–Fock and post-Hartree–Fock methods. The resulting properties are analyzed with respect to both basis set completeness and degree of correlation recovery, including core–core and core–valence effects. Although the lithium–water interaction is largely electrostatic in nature, small basis sets, lacking in polarization and near-valence diffuse functions, drastically overestimate the strength of the bond (by 20 kcal/mol or more) and underestimate the Li+...O distance by up to 0.1 Å. Their poor performance is attributable to inherent errors in describing the electric moments and polarizability of water and to large basis set superposition errors. Thus, the accuracy with which the fundamental lithium–water interaction could be modeled was primarily dependent on the quality of the Gaussian basis set and not upon the level of correlation recovery. Basis set enlargement and correlation effects both tend to reduce the strength of the Li+(H2O) bond, but produce corrections of opposite sign for the Li+...O bond length. Although correlation effects play a minor role in describing the lithium–water interaction, as the size of the cluster increases and the number of waters involved in multiple hydrogen bonds grows, correlation recovery can become significant.

KSSOLV (Kohn-Sham Solver) is a MATLAB (Matrix Laboratory) tool-box for solving the Kohn-Sham density functional theory (KS-DFT) with the plane-wave basis set. In the KS-DFT calculations, the most expensive part is commonly the diagonalization of Kohn-Sham Hamiltonian in the self-consistent field (SCF) scheme. To enable a personal computer to perform medium-sized KS-DFT calculations that contain hundreds of atoms, we present a hybrid CPU-GPU implementation to accelerate the iterative diagonalization algorithms implemented in KSSOLV by using the MATLAB built-in Parallel Computing Toolbox. We compare the performance of KSSOLV-GPU on three types of GPU, including RTX3090, V100, and A100, with conventional CPU implementation of KSSOLV respectively and numerical results demonstrate that hybrid CPU-GPU implementation can achieve a speedup of about 10 times compared with sequential CPU calculations for bulk silicon systems containing up to 128 atoms.

Stationary points for the OH − + CH 3F → CH 3OH + F − potential energy surface

Machine learning (ML) of kinetic energy functionals (KEFs), in particular kinetic energy density (KED) functionals, is a promising way to construct KEFs for orbital-free density functional theory (DFT). Neural networks and kernel methods including Gaussian process regression (GPR) have been used to learn Kohn-Sham (KS) KED from density-based descriptors derived from KS DFT calculations. The descriptors are typically expressed as functions of different powers and derivatives of the electron density. This can generate large and extremely unevenly distributed datasets, which complicates effective application of ML techniques. Very uneven data distributions require many training datapoints, can cause overfitting, and can ultimately lower the quality of an ML KED model. We show that one can produce more accurate ML models from fewer data by working with smoothed density-dependent variables and KED. Smoothing palliates the issue of very uneven data distributions and associated difficulties of sampling while retaining enough spatial structure necessary for working within the paradigm of KEDF. We use GPR as a function of smoothed terms of the fourth order gradient expansion and KS effective potential and obtain accurate and stable (with respect to different random choices of training points) kinetic energy models for Al, Mg, and Si simultaneously from as few as 2000 samples (about 0.3% of the total KS DFT data). In particular, accuracies on the order of 1% in a measure of the quality of energy-volume dependence B'=EV0-ΔV-2EV0+E(V0+ΔV)ΔV/V02 (where V0 is the equilibrium volume and ΔV is a deviation from it) are obtained simultaneously for all three materials.

Ab initio calculations have been carried out on the water molecule, using the IBMOL program, in order to investigate some possible uses and applications of small Gaussian basis sets. Firstly, calculations have been carried out in which emphasis was placed upon an improved description of the valence orbitals and, secondly, a study has been made of the feasibility of performing calculations in which a large basis set is used in the evaluation of the more important integrals and a small basis set for the remaining integrals. The two investigations indicate ways in which the calculated values of certain molecular properties may be improved upon without extending the computation time significantly.

This paper presents techniques for Fock matrix construction that are designed for high performance on shared and distributed memory parallel computers when using Gaussian basis sets. Four main techniques are considered. (1) To calculate electron repulsion integrals, we demonstrate batching together the calculation of multiple shell quartets of the same angular momentum class so that the calculation of large sets of primitive integrals can be efficiently vectorized. (2) For multithreaded summation of entries into the Fock matrix, we investigate using a combination of atomic operations and thread-local copies of the Fock matrix. (3) For distributed memory parallel computers, we present a globally accessible matrix class for accessing distributed Fock and density matrices. The new matrix class introduces a batched mode for remote memory access that can reduce the synchronization cost. (4) For density fitting, we exploit both symmetry (of the Coulomb and exchange matrices) and sparsity (of 3-index tensors) and give a performance comparison of density fitting and the conventional direct calculation approach. The techniques are implemented in an open-source software library called GTFock.

On the structure of the 3B 1 excited state of water

The performance of multistate density functional theory (MSDFT) with nonorthogonal state interaction (NOSI) is assessed for 100 vertical excitation energies against the theoretical best estimates extracted to the full configuration interaction accuracy on the database developed by Loos et al. in 2018 (Loos2018). Two optimization techniques, namely, block-localized excitation and target state optimization, are examined along with two ways of estimating the transition density functional (TDF) for the correlation energy of the Hamiltonian matrix density functional. The results from the two optimization methods are similar. It was found that MSDFT-NOSI using the spin-multiplet degeneracy constraint for the TDF of spin-coupling interaction, along with the M06-2X functional, yields a root-mean-square error (RMSE) of 0.22 eV, which performs noticeably better than time-dependent density functional theory (DFT) at an RMSE of 0.43 eV using the same functional and basis set on the Loos2018 database. In comparison with wave function theory, NOSI has smaller errors than CIS(D∞), LR-CC2, and ADC(3) all of which have an RMSE of 0.28 eV, but somewhat greater than STEOM-CCSD (RMSE of 0.14 eV) and LR-CCSD (RMSE of 0.11 eV) wave function methods. In comparison with Kohn-Sham (KS) DFT calculations, the multistate DFT approach has little double counting of correlation. Importantly, there is no noticeable difference in the performance of MSDFT-NOSI on the valence, Rydberg, singlet, triplet, and double-excitation states. Although the use of another hybrid functional PBE0 leads to a greater RMSE of 0.36 eV, the deviation is systematic with a linear regression slope of 0.994 against the results with M06-2X. The present benchmark reveals that density functional approximations developed for KS-DFT for the ground state with a noninteracting reference may be adopted in MSDFT calculations in which the state interaction is key.