Search

We develop a stochastic resolution of identity approach to the real-time second-order Green's function (real-time sRI-GF2) theory, extending our recent work for imaginary-time Matsubara Green's function [ Takeshita et al. J. Chem. Phys. 2019 , 151 , 044114 ]. The approach provides a framework to obtain the quasi-particle spectra across a wide range of frequencies and predicts ionization potentials and electron affinities. To assess the accuracy of the real-time sRI-GF2, we study a series of molecules and compare our results to experiments as well as to a many-body perturbation approach based on the GW approximation, where we find that the real-time sRI-GF2 is as accurate as self-consistent GW. The stochastic formulation reduces the formal computatinal scaling from O(Ne5) down to O(Ne3) where Ne is the number of electrons. This is illustrated for a chain of hydrogen dimers, where we observe a slightly lower than cubic scaling for systems containing up to Ne ≈ 1000 electrons.

We develop a range-separated stochastic resolution of identity (RS-SRI) approach for the four-index electron repulsion integrals, where the larger terms (above a predefined threshold) are treated using a deterministic RI and the remaining terms are treated using a SRI. The approach is implemented within a second-order Green's function formalism with an improved O(N3) scaling with the size of the basis set, N. Moreover, the RS approach greatly reduces the statistical error compared to the full stochastic version [T. Y. Takeshita et al., J. Chem. Phys. 151, 044114 (2019)], resulting in computational speedups of ground and excited state energies of nearly two orders of magnitude, as demonstrated for hydrogen dimer chains and water clusters.

Generalized Kohn–Sham density functional theory is a popular computational tool for the ground state of extended systems, particularly within range-separated hybrid (RSH) functionals that capture the long-range electronic interaction. Unfortunately, the heavy computational cost of the nonlocal exchange operator in RSH-DFT usually confines the approach to systems with at most a few hundred electrons. A significant reduction in the computational cost is achieved by representing the density matrix with stochastic orbitals and a stochastic decomposition of the Coulomb convolution (J. Phys. Chem. A 2016, 120, 3071). Here, we extend the stochastic RSH approach to excited states within the framework of linear-response generalized Kohn–Sham time-dependent density functional theory (GKS-TDDFT) based on the plane-wave basis. As a validation of the stochastic GKS-TDDFT method, the excitation energies of small molecules N2 and CO are calculated and compared to the deterministic results. The computational efficiency o...

The projector augmented wave (PAW) method of Bl\"ochl makes smooth but non-orthogonal orbitals. Here we show how to make PAW orthogonal, using a cheap transformation of the wave-functions. We show that the resulting Orthogonal PAW (OPAW), applied for DFT, reproduces (for a large variety of solids) band gaps from the ABINIT package. OPAW combines the underlying orthogonality of norm-conserving potentials with the large grid spacings and small spectral range in PAW. The OPAW framework can also be combined with other electronic structure theory methods., Comment: 8 pages, 2 figures, 2 tables

A stochastic orbital approach to the resolution of identity (RI) approximation for 4-index electron repulsion integrals (ERIs) is presented. The stochastic RI-ERIs are then applied to second order Moller–Plesset perturbation theory (MP2) utilizing a multiple stochastic orbital approach. The introduction of multiple stochastic orbitals results in an O(NAO3) scaling for both the stochastic RI-ERIs and stochastic RI-MP2, NAO being the number of basis functions. For a range of water clusters we demonstrate that this method exhibits a small prefactor and observed scalings of O(Ne2.4) for total energies and O(Ne3.1) for forces (Ne being the number of correlated electrons), outperforming MP2 for clusters with as few as 21 water molecules.

Author(s): Cytter, Y; Rabani, E; Neuhauser, D; Preising, M; Redmer, R; Baer, R | Abstract: The Kubo-Greenwood (KG) formula is often used in conjunction with Kohn-Sham (KS) density functional theory (DFT) to compute the optical conductivity, particularly for warm dense matter. For applying the KG formula, all KS eigenstates and eigenvalues up to an energy cutoff are required and thus the approach becomes expensive, especially for high temperatures and large systems, scaling cubically with both system size and temperature. Here, we develop an approach to calculate the KS conductivity within the stochastic DFT framework, which requires knowledge only of the KS Hamiltonian but not its eigenstates and values. We show that the computational effort associated with the method scales linearly with system size and reduces in proportion to the temperature, unlike the cubic increase with traditional deterministic approaches. In addition, we find that the method allows an accurate description of the entire spectrum, including the high-frequency range, unlike the deterministic method which is compelled to introduce a high-frequency cutoff due to memory and computational time constraints. We apply the method to helium-hydrogen mixtures in the warm dense matter regime at temperatures of ∼60kK and find that the system displays two conductivity phases, where a transition from nonmetal to metal occurs when hydrogen atoms constitute ∼0.3 of the total atoms in the system.

Linear scaling density functional theory is important for understanding electronic structure properties of nanometer scale systems. Recently developed stochastic density functional theory can achieve linear or even sublinear scaling for various electronic properties without relying on the sparsity of the density matrix. The basic idea relies on projecting stochastic orbitals onto the occupied space by expanding the Fermi-Dirac operator and repeating this for Nχ stochastic orbitals. Often, a large number of stochastic orbitals are required to reduce the statistical fluctuations (which scale as Nχ -1/2) below a tolerable threshold. In this work, we introduce a new stochastic density functional theory that can efficiently reduce the statistical fluctuations for certain observable and can also be integrated with an embedded fragmentation scheme. The approach is based on dividing the occupied space into energy windows and projecting the stochastic orbitals with a single expansion onto all windows simultaneously. This allows for a significant reduction of the noise as illustrated for bulk silicon with a large supercell. We also provide theoretical analysis to rationalize why the noise can be reduced only for a certain class of ground state properties, such as the forces and electron density.

The stochastic density functional theory (DFT) [Phys. Rev. Lett. 111, 106402 (2013)] is a valuable linear scaling approach to Kohn-Sham DFT that does not rely on the sparsity of the density matrix. Linear (and often sub-linear) scaling is achieved by introducing a controlled statistical error in the density, energy and forces. The statistical error (noise) is proportional to the inverse square root of the number of stochastic orbitals and thus decreases slowly, however, by dividing the system to fragments that are embedded stochastically, the statistical error can be reduced significantly. This has been shown to provide remarkable results for non-covalently bonded systems, however, the application to covalently bonded systems had limited success, particularly for delocalized electrons. Here, we show that the statistical error in the density correlates with both the density and the density matrix of the system and propose a new fragmentation scheme that elegantly interpolates between overlapped fragments. We assess the performance of the approach for bulk silicon of varying supercell sizes (up to $N_{e}=16384$ electrons) and show that overlapped fragments reduce significantly the statistical noise even for systems with a delocalized density matrix., 9 pages, 6 figures

Splittings of the translation–rotation (TR) eigenstates of the solid light-molecule endofullerenes M@C60 (M = H2, H2O, HF) attributed to the symmetry breaking have been observed in the infrared (IR) and inelastic neutron scattering spectra of these species in the past couple of years. In a recent paper [Felker et al., Phys. Chem. Chem. Phys., 2017, 19, 31274], we established that the electrostatic, quadrupolar interaction between the guest molecule M and the twelve nearest-neighbor C60 cages of the solid is the main source of the symmetry breaking. The splittings of the three-fold degenerate ground states of the endohedral ortho-H2, ortho-H2O and the j = 1 level of HF calculated using this model were found to be in excellent agreement with the experimental results. Utilizing the same electrostatic model, this theoretical study investigates the effects of the symmetry breaking on the excited TR eigenstates of the three species, and how they manifest in their simulated low-temperature (5–6 K) near-IR (NIR) and far-IR (FIR) spectra. The TR eigenstates are calculated variationally for both the major P and minor H crystal orientations. For the H orientation, the calculated splittings of all of the TR levels of these species are less than 0.1 cm−1. For the dominant P orientation, the splittings vary strongly depending on the character of the excitations involved. In all of the species, the splittings of the higher rotationally excited levels are comparable in magnitude to those for the j = 1 levels. For the levels corresponding to purely translational excitations, the calculated splittings are about an order of magnitude smaller than those of the purely rotational eigenstates. Based on the computed TR eigenstates, the low-temperature NIR (for M = H2) and FIR (for M = HF and H2O) spectra are simulated for both the P and H orientations, and also combined as their weighted sum (0.15H + 0.85P). The weighted sum spectra computed for M = H2 and HF match quantitatively the corresponding measured spectra, while for M = H2O, the weighted sum FIR spectrum predicts features that can potentially be observed experimentally.

Stochastic density functional theory (sDFT) is becoming a valuable tool for studying ground state properties of extended materials. The computational complexity of describing the Kohn-Sham orbitals is replaced by introducing a set of random (stochastic) orbitals leading to linear and often sub-linear scaling of certain ground-state observable at the account of introducing a statistical error. Schemes to reduce the noise are essential, for example, for determining the structure using the forces obtained from sDFT. Recently we have introduced two embedding schemes to mitigate the statistical fluctuations in the electron density and resultant forces on the nuclei. Both techniques were based on fragmenting the system either in real-space or slicing the occupied space into energy windows, allowing for a significant reduction of the statistical fluctuations. For chemical accuracy further reduction of the noise is required, which could be achieved by increasing the number of stochastic orbitals. However, the convergence is relatively slow as the statistical error scales as $1/\sqrt{N_\chi}$ according to the central limit theorem, where $N_\chi$ is the number of random orbitals. In this paper we combined the aforementioned embedding schemes and introduced a new approach that builds on overlapped fragments and energy windows. The new approach significantly lowers the noise for ground state properties such as the electron density, total energy, and forces on the nuclei, as demonstrated for a G-center in bulk silicon., Comment: The following article has been submitted to the Journal of Chemical Physics

Absorption cross-section spectra for gold nanoparticles were calculated using fully quantum Stochastic Density Functional Theory and a classical Finite-Difference Time Domain (FDTD) Maxwell solver. Spectral shifts were monitored as a function of size (1.3-3.1 nm) and shape (octahedron, cubeoctahedron, and truncated cube). Even though the classical approach is forced to fit the quantum TDDFT at 3.1nm, at smaller sizes there is a significant deviation as the classical theory is unable to account for peak splitting and spectral blue shifts even after quantum spectral corrections. We attribute the failure of classical methods at predicting these features to quantum effects and low density of states in small nanoparticles. Classically, plasmon resonances are modeled as collective conduction electron excitations, but at small nanoparticle size these excitations transition to few or even individual conductive electron excitations, as indicated by our results.

We introduce the concept of sparse stochastic compression, an efficient stochastic sampling of any general function. The technique uses sparse stochastic orbitals (SSOs), short vectors that sample a small number of space points. As a first demonstration, SSOs are applied in conjunction with simple direct projection to accelerate our recent stochastic $GW$ technique; the new developments enable accurate prediction of ${G}_{0}{W}_{0}$ quasiparticle energies and gaps for systems with up to ${N}_{e}g10,000$ electrons, with small statistical errors of $\ifmmode\pm\else\textpm\fi{}0.05\phantom{\rule{0.28em}{0ex}}\mathrm{eV}$ and using less than 2000 core CPU hours. Overall, stochastic $GW$ scales now linearly (and often sublinearly) with ${N}_{e}.$

Simulations in the warm dense matter regime using finite temperature Kohn-Sham density functional theory (FT-KS-DFT), while frequently used, are computationally expensive due to the partial occupation of a very large number of high-energy KS eigenstates which are obtained from subspace diagonalization. We have developed a stochastic method for applying FT-KS-DFT, that overcomes the bottleneck of calculating the occupied KS orbitals by directly obtaining the density from the KS Hamiltonian. The proposed algorithm scales as $O\left(N{T}^{\ensuremath{-}1}\right)$ and is compared with the high-temperature limit scaling $O\left({N}^{3}{T}^{3}\right)$ of the deterministic approach, where $N$ is the system size (number of electrons, volume, etc.) and $T$ is the temperature. The method has been implemented in a plane-waves code within the local density approximation (LDA); we demonstrate its efficiency, statistical errors, and bias in the estimation of the free energy per electron for a diamond structure silicon. The bias is small compared to the fluctuations and is independent of system size. In addition to calculating the free energy itself, one can also use the method to calculate its derivatives and obtain the equations of state.

We examine the possibility of using a Metropolis algorithm for computing the exchange energy in a large molecular system. Following ideas set forth in a recent publication (Baer, Neuhauser, and Rabani, Phys. Rev. Lett. 111, 106402 (2013)) we focus on obtaining the exchange energy per particle (ExPE, as opposed to the total exchange energy) to a predefined statistical error and on determining the numerical scaling of the calculation achieving this. For this we assume that the occupied molecular orbitals (MOs) are known and given in terms of a standard Gaussian atomic basis set. The Metropolis random walk produces a sequence of pairs of three-dimensional points (x,x'), which are distributed in proportion to ρ(x,x')(2), where ρ(x,x') is the density matrix. The exchange energy per particle is then simply the average of the Coulomb repulsion energy υC(|x-x'|) over these pairs. To reduce the statistical error we separate the exchange energy into a short-range term that can be calculated deterministically in a linear scaling fashion and a long-range term that is treated by the Metropolis method. We demonstrate the method on water clusters and silicon nanocrystals showing the magnitude of the ExPE standard deviation is independent of system size. In the water clusters a longer random walk was necessary to obtain full ergodicity as Metropolis walkers tended to get stuck for a while in localized regions. We developed a diagnostic tool that can alert a user when such a situation occurs. The calculation effort scales linearly with system size if one uses an atom screening procedure that can be made numerically exact. In systems where the MOs can be localized efficiently the ExPE can even be computed with "sublinear scaling" as the MOs themselves can be screened.

The stochastic formulation of multiexciton generation (MEG) rates is extended to provide access to MEG efficiencies in nanostructures containing thousands of atoms. The formalism is applied to a series of CdSe/CdS seeded nanorod heterostructures with different core and shell dimensions. At energies above 3Eg (where Eg is the band gap), the MEG yield increases with decreasing core size, as expected for spherical nanocrystals. Surprisingly, this behavior is reversed for energies below this value, and is explained by the dependence of the density of states near the valence band edge, which increases with the core diameter. Our predictions indicate that the onset of MEG can be shifted to lower energies by manipulating the density of states in complex nanostructure geometries.

For metallic nanoparticles less than 10 nm in diameter, localized surface plasmon resonances (LSPRs) become sensitive to the quantum nature of conduction electrons. In this regime, experimental probes of size-dependent LSPRs are particularly challenging, and contradictory results are often reported. Unfortunately, quantum mechanical simulations based on time-dependent Kohn-Sham density functional theory (TD-KSDFT) are computationally too expensive to tackle metal particles larger than 2 nm. Herein, we present a time-dependent orbital-free density functional theory (TD-OFDFT) that accurately captures the dynamic response of electrons in the presence of realistic ionic potentials. The TD-OFDFT method offers a comparable accuracy as TD-KSDFT but with a much lower computational cost. Using TD-OFDFT, we study size-dependent LSPRs on Na nanoparticles with diameters from 0.7 to 12.3 nm. The optical absorption spectra exhibit a nonmonotonic behavior from blue shift to red shift and back to blue shift as the particle size decreases. Three principal plasmon modes are identified, and their physical origins are elucidated. Competing physical mechanisms responsible for the nonmonotonic size dependence are discussed. The TD-OFDFT provides a unified theoretical framework that bridges the gap between classical electromagnetic theory and quantum mechanical theory for plasmonics and nanophotonics.

A fast method is developed for calculating the random phase approximation (RPA) correlation energy for density functional theory. The correlation energy is given by a trace over a projected RPA response matrix, and the trace is taken by a stochastic approach using random perturbation vectors. For a fixed statistical error in the total energy per electron, the method scales, at most, quadratically with the system size; however, in practice, due to self-averaging, it requires less statistical sampling as the system grows, and the performance is close to linear scaling. We demonstrate the method by calculating the RPA correlation energy for cadmium selenide and silicon nanocrystals with over 1500 electrons. We find that the RPA correlation energies per electron are largely independent of the nanocrystal size. In addition, we show that a correlated sampling technique enables calculation of the energy difference between two slightly distorted configurations with scaling and a statistical error similar to that of the total energy per electron.

We show that a rigid scissors-like GW self-consistency approach, labeled here Δ¯GW0, can be trivially implemented at zero additional cost for large scale one-shot G0W0 calculations. The method significantly improves one-shot G0W0 and for large systems is very accurate. Δ¯GW0 is similar in spirit to evGW0 where the self-consistency is only applied on the eigenvalues entering Green’s function, while both W and the eigenvectors of Green’s function are held fixed. Δ¯GW0 further assumes that the shift of the eigenvalues is rigid scissors-like so that all occupied states are shifted by the same amount and analogously for all the unoccupied states. We show that this results in a trivial modification of the time-dependent G0W0 self-energy, enabling an a posteriori self-consistency cycle. The method is applicable for our recent stochastic-GW approach, thereby enabling self-consistent calculations for giant systems with thousands of electrons. The accuracy of Δ¯GW0 increases with the system size. For molecules, it is up to 0.4-0.5 eV away from coupled-cluster single double triple (CCSD(T)), but for tetracene and hexacene, it matches the ionization energies from both CCSD(T) and evGW0 to better than 0.05 eV. For solids, as exemplified here by periodic supercells of semiconductors and insulators with 6192 valence electrons, the method matches evGW0 quite well and both methods are in good agreement with the experiment.We show that a rigid scissors-like GW self-consistency approach, labeled here Δ¯GW0, can be trivially implemented at zero additional cost for large scale one-shot G0W0 calculations. The method significantly improves one-shot G0W0 and for large systems is very accurate. Δ¯GW0 is similar in spirit to evGW0 where the self-consistency is only applied on the eigenvalues entering Green’s function, while both W and the eigenvectors of Green’s function are held fixed. Δ¯GW0 further assumes that the shift of the eigenvalues is rigid scissors-like so that all occupied states are shifted by the same amount and analogously for all the unoccupied states. We show that this results in a trivial modification of the time-dependent G0W0 self-energy, enabling an a posteriori self-consistency cycle. The method is applicable for our recent stochastic-GW approach, thereby enabling self-consistent calculations for giant systems with thousands of electrons. The accuracy of Δ¯GW0 increases with the system size. For molecules, it ...

We develop an alternative formulation in the energy-domain to calculate the second order Moller−Plesset (MP2) perturbation energies. The approach is based on repeatedly choosing four random energies using a nonseparable guiding function, filtering four random orbitals at these energies, and averaging the resulting Coulomb matrix elements to obtain a statistical estimate of the MP2 correlation energy. In contrast to our time-domain formulation, the present approach is useful for both quantum chemistry and real-space/plane wave basis sets. The scaling of the MP2 calculation is roughly linear with system size, providing a useful tool to study dispersion energies in large systems. This is demonstrated on a structure of 64 fullerenes within the SZ basis as well as on silicon nanocrystals using real-space grids.

Charge carrier localization in extended atomic systems has been described previously as being driven by disorder, point defects or distortions of the ionic lattice. Here we show for the first time by means of first-principles computations that charge carriers can spontaneously localize due to a purely electronic effect in otherwise perfectly ordered structures. Optimally-tuned range-separated density functional theory and many-body perturbation calculations within the GW approximation reveal that in trans-polyacetylene and polythiophene the hole density localizes on a length scale of several nanometers. This is due to exchange-induced translational symmetry breaking of the charge density. Ionization potentials, optical absorption peaks, excitonic binding energies and the optimally-tuned range parameter itself all become independent of polymer length as it exceeds the critical localization scale. Moreover, lattice disorder and the formation of a polaron result from the charge localization in contrast to the traditional view that lattice distortions precede charge localization. Our results can explain experimental findings that polarons in conjugated polymers form instantaneously after exposure to ultrafast light pulses., Comment: 5 pages, 4 figures

A time-dependent formulation for electron-hole excitations in extended finite systems, based on the Bethe-Salpeter equation (BSE), is developed using a stochastic wave function approach. The time-dependent formulation builds on the connection between time-dependent Hartree-Fock (TDHF) theory and configuration-interaction with single substitution (CIS) method. This results in a time dependent Schr\"odinger-like equation for the quasiparticle orbital dynamics based on an effective Hamiltonian containing direct Hartree and screened exchange terms, where screening is described within the Random Phase Approximation (RPA). To solve for the optical absorption spectrum, we develop a stochastic formulation in which the quasiparticle orbitals are replaced by stochastic orbitals to evaluate the direct and exchange terms in the Hamiltonian as well as the RPA screening. This leads to an overall quadratic scaling, a significant improvement over the equivalent symplectic eigenvalue representation of the BSE. Application of the time-dependent stochastic BSE (TDsBSE) approach to silicon and CdSe nanocrystals up to size of ~3000 electrons is presented and discussed., Comment: 12 pages, 5 figures

We reformulate 1D boson-fermion duality in path-integral terms. The result is a 1D counterpart of the boson-fermion duality in the 2D Chern-Simons gauge theory. The theory is consistent and enables, using standard resummation techniques, to obtain the long-wavelength asymptotics of the collective mode in 1D boson systems at the Tonks-Girardeau regime. The collective mode has the dispersion of Bogoliubov phonons: ω(q) = q(U(q)/m)1/2, where is the bosons density and U(q) is a Fourier component of the two-body potential.

We develop a formalism for efficient iterative solutions of scattering problems involving the Maxwell equations. The methods, borrowed from recent advancements in chemical reaction dynamics, represent the scattering wavefunctions on two grids; one used for the initial wave and is one-dimensional; the other is a small three-dimensional grid padded with absorbing-potentials on which the scattered function is represented. The formalism is automatically suitable for scattering studies of transmission, reflection and scattering components of a wave. The simulations can be done with time-dependent wavepackets or direct iterative solution for the Green's function, but the results are rigorous time-independent (frequency-dependent) scattering amplitudes. Model time-dependent simulations involving up to a 100×100×100 grid for the inner wavefunction were numerically done on a simple PC.

A semiclassical cellular method is proposed. Signals generated by semiclassical techniques generally deteriorate over time as trajectories become chaotic. One approach to remedy this problem has been to have each trajectory weighted by an entire cell of nearby trajectories (Filinov transform). But even in this approach the exponential part of the propagator typically becomes large and positive over time. Here the cellularization (Filinov) parameter is subject to constraints which make it time dependent and trajectory dependent. It also depends on dimensionality, so it ends up as a matrix. Physically, the Filinov transform is done differently in different directions associated with the stability matrix for the phase—essentially a more confined integration in directions where the matrix diverges and a wider integration in other directions. This squelches the contribution from any part of a trajectory that becomes excessively chaotic. A trajectory-dependent cellurized frozen Gaussian is applied here within the Herman–Kluk semiclassical approach. It is tested by looking at a single-particle three-dimensional problem, He attached to a rigid immovable naphtalene, where it is shown to be more accurate than the original HK approach, without the divergence of the correlation function common in the usual cellular dynamics (HK) formulation, and is able to separate a low-lying excited state from the ground state.

The dynamics of a chain of vibrational bonds which develop a classical solitary compression wave is simulated. A converged fully correlated quantum mechanical calculation is compared with a time dependent mean field approach (TDSCF) and with a classical simulation. The dynamics were all generated from the same Hamiltonian. The TDSCF and classical calculations show a fully developed solitary wave with the expected dependence of group velocity on amplitude. The full quantum calculations show a solitary-like wave which propagates for a while but then degrades. The robustness of the compression wave depends on the initial preparation. Evidence of partial recurrence of the wave has also been observed.

A stochastic approach to time-dependent density functional theory (TDDFT) is developed for computing the absorption cross section and the random phase approximation (RPA) correlation energy. The core idea of the approach involves time-propagation of a small set of stochastic orbitals which are first projected on the occupied space and then propagated in time according to the time-dependent Kohn-Sham equations. The evolving electron density is exactly represented when the number of random orbitals is infinite, but even a small number (? 16) of such orbitals is enough to obtain meaningful results for absorption spectrum and the RPA correlation energy per electron. We implement the approach for silicon nanocrystals (NCs) using real-space grids and find that the overall scaling of the algorithm is sublinear with computational time and memory., 7 pages, 4 figures

We develop a formalism to calculate the quasiparticle energy within the GW many-body perturbation correction to the density functional theory. The occupied and virtual orbitals of the Kohn-Sham Hamiltonian are replaced by stochastic orbitals used to evaluate the Green function G, the polarization potential W, and, thereby, the GW self-energy. The stochastic GW (sGW) formalism relies on novel theoretical concepts such as stochastic time-dependent Hartree propagation, stochastic matrix compression, and spatial or temporal stochastic decoupling techniques. Beyond the theoretical interest, the formalism enables linear scaling GW calculations breaking the theoretical scaling limit for GW as well as circumventing the need for energy cutoff approximations. We illustrate the method for silicon nanocrystals of varying sizes with N_{e}3000 electrons.

The spectra of metal-containing molecules and crystals often contain “interference dips,” i.e., sharp decreases in absorbance at the energy of forbidden or weakly allowed intraconfigurational transitions [1–14]. The loss of absorbance is naively unexpected; the spectrum should consist of the superposition of the absorbances resulting from the transitions to the two different excited electronic states. The spectroscopic dips are routinely interpreted using the theory developed by Fano that was derived for atomic absorption in the presence of a slowly varying background continuum [15]. In molecular and crystal spectroscopy, however, the “background” contains an underlying dispersion part that varies strongly with energy [16–18]. It results from a dipole allowed electronic transition with a bandwidth determined by progressions in normal vibrational modes that are displaced between the excited and ground electronic states. A new treatment of the loss of absorbance that has physical relevance to molecules or to metal ions with strong phonon coupling to host lattices is needed. In this Letter we derive a simple analytical expression for the intensity decrease and quantitatively interpret the phenomenon. The dips are a form of destructive quantum interference. The important molecular properties that govern the interference are the relative energies of the states, the vibrational frequencies in the electronic states, and the coupling properties. Our analytical derivation uses the simplifying assumption that only one vibrational eigenstate of the “forbidden” electronic state is involved. This assumption is accurate for intraconfigurational transitions in metal compounds, where the spin multiplicity changes but the orbital configuration does not [19]. When this type of transition is well separated from more intense transitions, it is observed as a single line (i.e., little or no vibrational progression.) Our analytical expression for the interference dips is compared to the calculation using the full Hamiltonian. The total electron-nuclear Hamiltonian for the coupled excited states is described in terms of the usual diabatic Hamiltonians [20]: H p 2

The shifted-contour auxiliary-field Monte Carlo (SCAFMC) approach has been recently developed by Rom, Charutz and Neuhauser as an extension of the auxiliary-field Monte Carlo (AFMC) method. AFMC replaces the difficult fully interacting electrons problem by an ensemble of simpler problems where the electrons interact with a fluctuating electric field but not with each other. SCAFMC is based on shifting the auxiliary-field contour of integration to pass through the (imaginary) stationary point, leading to numerical stability at long propagation times. The new approach converges to the full CI energy in electronic structure calculations (both ground and low-lying excited states). Here we expand the application of SCAFMC from atomic to molecular problems. First, we calculate ground-state energies of a highly correlated transition-metal system (Cr2) with a moderate (12 orbitals) active space size, and demonstrate that SCAFMC is able to extract the energies accurately. In addition, we use SCAFMC to calculate a ...

Several years ago we introduced a new approach for signal extraction, Filter-Diagonalization, which is capable of handling signals containing millions of points or more. The method was recently extended to handle multidimensional signals, as appear in NMR, by us and separately by Mandelshtam and Taylor. We further extend the original formulation by showing that the method is capable of handling degenerate eigenvalues and eigenstates through the use of a simultaneous diagonalization algorithm. We apply the method to simple two-dimensional NMR signals. Comparisons to a standard FFT treatment show that FDG recovers the salient spectral features using a quarter of the signal required by FFT.

A simple method that incorporates coupling between bath and system and between individual bath modes is proposed. In the new locally propagating gaussian (LPG) approximation, the bath is modeled by Gaussians at each system site that propagate in time. Each site couples to others by the overlap of their Gaussians. In a test case with weak but non-negligible coupling, LPG is nearly as rapid as TDSCF but more closely follows exact calculations. LPG can be extended to larger simulations. Strongly coupled system coordinates will be treated with a grid-type calculation and weakly coupled bath modes will be handled by LPG.

A procedure to control harmonic generation is proposed using initial state control. We show that an initial molecular state can be prepared before shining a strong laser pulse to maximize the output harmonic generation. We demonstrate the method by maximizing emission for the sixth harmonic on a given initial IR pulse in a model H 2 + system using five initial vibrational states.

The filter-diagonalization method is used to find the broad and even overlapping resonances of a 1D Hamiltonian used before as a test model for new resonance theories and computational methods. Is is found that the use of several complex-scaled cross-correlation probability amplitudes from short-time propagation enables the calculation of broad overlapping resonances, which can not be resolved from the amplitude of a single complex-scaled autocorrelation calculation.

A numerical scheme is suggested for solving the scalar wave equation for optical devices characterized by a z-dependent refractive index. The homogeneous wave equation is converted into a single-column inhomogeneous linear system by applying imaginary (absorbing) boundary potentials at the device boundaries. The imposed absorbing boundary conditions enable discrete representation of the device on a compact grid. Our scheme is based on applying an efficient sparse preconditioner to the linear system, which enables its global solution (i.e., simultaneously for all z values) by fast iterative methods, such as the quasi-minimal residual algorithm. A single solution of the inhomogeneous equation with the imaginary boundary operators allows the calculation of mode-specific as well as region-specific light-intensity coupling probabilities for initial mode of interest. Numerical examples illustrate the usefulness of the suggested scheme to the optimization of optical devices.

Heller’s expression for the absorption cross-section in the weak field limit is extended to cases where the total Hamiltonian contains a strong time-dependent component, supplemented by a weak field. A very similar expression to the original case then results when the (t,t) formalism is used; one only needs to construct a correlation function for the system without the weak field, and use it to extract the absorption probability for any value of the weak-field frequency (or pulse shape). In addition, a numerical approach for extracting Floquet states without full-matrix diagonalization is demonstrated, by filtering (or filter-diagonalization) a single wave function (or the correlation function) propagated under the (t,t) Hamiltonian.

A converged Monte Carlo approach for electronic-structure calculations of ground and low-lying excited-states is developed. The method, Shifted-Contour Auxiliary-Field Monte Carlo (SCAFMC), is an extension of the Auxiliary-Field Monte Carlo (AFMC) approach, and is based on a shift of the auxiliary-field contour of integration to pass through the (imaginary) stationary point. It eliminates sign difficulties for systems with a strong contribution from the mean field, such as electronic-structure applications. The method is exemplified for ground and excited-states calculations of Ne.

We present a new and more efficient implementation of a hybrid approach to computing the solution of scattering problems affected by resonances. In the computationally expensive part of the calculation, wave packet propagation is used to obtain the time-dependent wave function Ψ(t) up to some time τ at which direct scattering is over. This part is made efficient by using a recently introduced modification for the absorbing boundary conditions evolution operator which allows the use of real operator algebra if the initial wave function is chosen real. In the second part of the calculation, filter diagonalization is used to efficiently obtain the energies, widths, and expansion coefficients of resonances needed to describe the long time behavior of the scattering wave function. This part is made efficient by using a recently introduced algorithm which avoids the storage of energy-dependent basis functions. We demonstrate the application of the method to a two-dimensional reactive scattering problem.

We formulate the Kohn-Sham density functional theory (KS-DFT) as a statistical theory in which the electron density is deter-mined from an average of correlated stochastic densities in a trace formula. The key idea is that it is sufficient to converge the total energy per electron to within a predefined statistical error in order to obtain reliable estimates of the electronic band structure, the forces on nuclei, the density and its moments, etc. The fluctuations in the total energy per electron are guaranteed to decay to zero as the system size increases. This facilitates "self-averaging" which leads to the first ever report of sublinear scaling KS-DFT electronic structure. The approach sidesteps calculation of the density matrix and thus is insensitive to its evasive sparseness, as demonstrated here for silicon nanocrystals. The formalism is not only appealing in terms of its promise to far push the limits of application of KS-DFT, but also represents a cognitive change in the way we think of electronic structure calculations as this stochastic theory seamlessly converges to the thermodynamic limit., 4 pages, 4 figures

A time‐independent approach to calculate scattering matrix elements using the complex coordinate method is examined. This approach is based on a combination of the expressions derived by Tannor and Weeks [J. Chem. Phys. 98, 3884 (1993)] and by Kouri, Huang, Zhu, and Hoffman [J. Chem. Phys. 100, 3662 (1994)], with an analytic continuation of the Hamiltonian, while keeping the initial and final wave packets unscaled. The procedure is examined using a one dimensional Eckart barrier representing the H+H2 reaction, and a comparison between two complex scaling schemes and an optical potential one shows good convergence of the method. In addition, a one‐dimensional electron scattering from a barrier is calculated, showing an advantage here of the complex‐scaling approach over the optical potentials method when very light particles are involved in the dynamics. The complex‐scaling version enables the use of iterative techniques, hence is a promising tool for calculating dynamics in large systems of light particles.

We investigate the usefulness of a hybrid method for scattering with resonances. Wave packet propagation is used to obtain the time‐dependent wave function Ψ(t) up to some time T at which direct scattering is over. Next, Ψ(t) is extrapolated beyond T employing resonance eigenvalues and eigenfunctions obtained in a Lanczos procedure, using Ψ(T) as starting vector to achieve faster convergence. The method is tested on one two‐dimensional (2D) and one four‐dimensional (4D) reactive scattering problem, affected by resonances of widths 0.1–5 meV. Compared to long time wave packet propagation, the hybrid method allows large reductions in the number of Hamiltonian operations NH required for obtaining converged reaction probabilities: A reduction factor of 24 was achieved for the 2D problem, and a factor of 6 for the 4D problem.

We investigate the performance of a scattering algorithm which uses purely real algebra for the major part of the wave function calculation, while incorporating automatically the appropriate boundary conditions. The algorithm falls in the category of time‐independent wave packet methods ([R. Kosloff, J. Phys. Chem. 92, 2087 (1988)], and, more specifically for scattering [Y. Huang, W. Zhu, D. J. Kouri, and D. K. Hoffman, Chem. Phys. Lett. 206, 96 (1993)]), and combines two previous approaches: A method [V. A. Mandelshtam and H. S. Taylor, J. Chem. Phys. 103, 2903 (1995)] in which the action of the absorbing potentials is implicitly inserted in a polynomial expansion of the Green’s function, and a real initial wave function approach, in which zero initial momenta are avoided. Compared to the conventional, multiple time‐step Chebyshev method, the new algorithm required three times less Hamiltonian evaluations for a model problem involving direct scattering. The new method also showed faster convergence for a problem involving resonances. Both methods showed convergence problems in the vicinity of a very narrow resonance.

In this paper, we apply the generalized filter-diagonalization (GFD) technique for extracting normal modes through correlation functions generated by molecular dynamics simulation. We test our method on a 256 argon particles system. Normal modes are calculated by the power spectrum of the velocity auto-correlation function, generalized filter-diagonalization, and direct diagonalization of the force matrix. The argon cluster provides a rigorous model for the generalized filter-diagonalization technique since the average separation between adjacent modes is less than 0.1 cm −1 . Excellent agreement was achieved even for normal modes that are separated by as little as ∼ 0.01 cm −1 .

This article has a few goals. First, a new highly accurate and anomaly free time‐independent approach to reactive scattering is presented, based on the use of very‐short‐range imaginary potentials. The range of the imaginary potentials is extremely short—in successful one‐dimensional simulations they cover only two grid points. The savings are incurred by limiting the role of the imaginary potentials to shifting the eigenvalues of the Hamiltonian away from the real axis, thereby avoiding anomalies; the imaginary potentials are not required to impose outgoing boundary conditions. Another goal is a rigorous derivation of a flux‐amplitude (FA) expression, whereby (for any scattering approach, whether using negative imaginary potentials or not) reactive S‐matrix amplitudes are extracted from the wavefunction at a single surface of any desired ‘‘internal’’ coordinate system—there is no need to project the wave function to the asymptotic products coordinates before determining its flux. With the FA expression, expensive volume integrals in state‐to‐state scattering are reduced to simple surface integrals. The FA expression also leads to a rigorous derivation of various alternate expressions for the scattering matrix which are useful whenever negative imaginary potentials are utilized. Finally, a new expression is presented for estimating the errors in absorbing potentials single‐column calculations due to imperfections in the absorbing potentials.

In a previous paper we developed a method, Filter‐Diagonalization, for extracting eigenvalues and eigenstates of a given operator at any desired energy range. In essence, the method eliminates correlation between distant eigenstates through a short‐time filter while correlations between closely lying states are eliminated by diagonalization. Here we extend Filter‐Diagonalization. When used to extract eigenvalues for a given operator H, we show that all eigenvalue information is directly extracted from a short segment of the correlation functionC(t)=(ψ(0)‖e −iHt ‖ψ(0)), or alternately from a small number of residues (ψ(0)‖R n (H)‖ψ(0)), where ψ(0) is a random initial function and R n (H) is any desired polynomial expansion in H. The implications of this feature are twofold. First, in contrast to the previous version the wave packet needs only to be propagated once (to prepare C(t)), and eigenstates at all desired energy windows can then be extracted with negligible extra computation time (and negligible storage requirements). In a simulation presented here, accurate eigenvalues are extracted using propagation times which are only a 0.0041 fraction of the ‘‘natural’’ time, i.e., the time by which the relative phase of the two closest eigenstates reaches 2π. The second and more important feature is that the method is automatically suitable for extracting eigenvalues (or normal modes) using a short‐time segment of any signal C(t) which is a sum of (unknown) Fourier components (C(t)=∑ nd ne −ie nt ) regardless of its origin. In addition to its use for determining eigenvalues of known operators, the method may also be utilized to extract normal modes from classical‐dynamics simulations, eigenstates from real‐time Quantum Monte‐Carlo studies, frequencies from experimental optical or electrical signals, or be utilized in any other circumstance where a correlation function or general signal is only known for short times (or expensive to generate at long times).

Auxiliary‐field Monte Carlo (AFMC) is an exact approach for calculating the ground state of a system of fermions (or bosons) interacting by pair‐potentials. The method uses the Hubbard–Stratonovich transformation to replace the exact imaginary‐time propagator by an average over an ensemble of propagators for independent particles in the presence of a varying external field, so that the calculation of the exact energy is reduced to multiple independent calculations, each of which costs essentially the same as one Hartree–Fock iteration. Here we consider the application of AFMC to calculate molecular structure, and present preliminary simulations on He and Be. We develop two simple methods to partially alleviate a ‘‘sign‐problem’’ in AFMC through restriction of the length of the imaginary‐time propagation, by either a simultaneous propagation of several initial states followed by subspace‐diagonalization or by incorporation of information from all propagated time steps. The first method is tested and found to yield significant improvement in accuracy. For the present simulations, the single‐particle orbitals are expanded in a given set of primitive orbitals. The resulting spectral‐AFMC method yields, for sufficiently converged ensembles, the full‐CI energy associated with a given basis. The developments reported here, and in particular the demonstration of subspace‐diagonalization, have however general validity independent of whether a basis set or a grid representation is used for the single‐particle orbitals (in the first case a full‐CI result is obtained in the given basis, while a converged grid representation would yield the exact result).

In a previous paper [J. Chem. Phys. 93, 2611 (1990)] a new method, filter diagonalization, was introduced for extracting highly excited rovibrational states from an arbitrary Hamiltonian, in any desired energy range. In the method, an arbitrary initial wave packet is propagated for a short time and during the propagation a ‘‘short time filter’’ of the wave packet is accumulated at various energies in any desired ‘‘window,’’ yielding a small set of functions which span the eigenfunctions of the Hamiltonian in the desired range. A small Hamiltonian matrix is then evaluated in the filtered‐functions basis, to yield the eigenvalues in the desired range. The combination of the time‐dependent (TD) propagation with the small matrix diagonalization eliminates the uncertainty‐relation limitation associated with a pure TD approach and the large‐matrix diagonalization necessary in a purely time‐independent approach. In this paper we give the first demonstration of the power of filter diagonalization for a molecular Hamiltonian exhibiting accidental near degeneracies, thereby supplying a stringent test of our approach. A two‐dimensional model of LiCN (J=0) is used. Good agreement is established with previous results for high‐energy states. For a further check of consistency, we perform a large‐scale direct diagonalization of the Hamiltonian, and verify a very‐high accuracy even for nearly degenerate levels. Extraction of these levels by a purely TD approach would have necessitated ≊700‐fold increase in propagation time.

We develop near-field (NF), a very efficient finite-difference time-dependent (FDTD) approach for simulating electromagnetic systems in the near-field regime. NF is essentially a time-dependent version of the quasistatic frequency-dependent Poisson algorithm. We assume that the electric field is longitudinal, and hence propagates only a set of time-dependent polarizations and currents. For near-field scales, the time step (dt) is much larger than in the usual Maxwell FDTD approach, as it is not related to the velocity of light; rather, it is determined by the rate of damping and plasma oscillations in the material, so dt = 2.5 a.u. was well converged in our simulations. The propagation in time is done via a leapfrog algorithm much like Yee's method, and only a single spatial convolution is needed per time step. In conjunction, we also develop a new and very accurate 8 and 9 Drude-oscillators fit to the permittivity of gold and silver, desired here because we use a large time step. We show that NF agrees with Mie-theory in the limit of small spheres and that it also accurately describes the evolution of the spectral shape as a function of the separation between two gold or silver spheres. The NF algorithm is especially efficient for systems with small scale dynamics and makes it very simple to introduce additional effects such as embedding.

We propose a new approach to imaging astrometry, the uniaxial crystal interferometer (UCI). It uses the extremely sensitive dependence of the polarization phase change on the incident angle in uniaxial crystals. This is further combined with very sensitive polarization measurements. The polarization-coding is used for a fine-scale angle determination, and is simultaneously combined with a crude measurement through a standard telescope. Further, achromatic anamorphic prisms will amplify five to tenfold the resolution which is estimated to reach the scale ratio of large interferometers. Because the fine angular information is superimposed on the light at an early stage the optical system tolerances are relieved to the level of a standard low-weight optical imaging system. We also suggest solutions to make the system achromatic. Overall the system may cover sky segments of the order of fractions of a degree square and reach a resolution of tens of microas, even for faint stars.

Time‐dependent scattering is extended to systems possessing narrow resonances. At short times the wave function is integrated directly, and at late times the wave function is expanded in terms of the slowly decaying (complex) resonance eigenfunctions of the Hamiltonian, Ψ(X,t)≂Σn ane−ient−Γ nt/2Φn(X). The slowly decaying eigenfunctions are easily found via a short‐time filterization approach adapted from bound‐state studies, in which a random wave packet is filtered at various energies and the resulting vectors are then diagonalized. The method is exemplified for collinear reactions of H+H2, where it halves the propagation time.