Localized methods for more efficient quantum chemical calculations

Jeff Robert Keyes

Published Jan 1, 2025• 0 citations

Abstract

Localized methods for more efficient quantum chemical calculations, December 12, 2025, https://doi.org/10.14418/8qbns2ce, Keyes, Jeff Robert (creator), Jiménez-Hoyos, Carlos Alberto, dissertations, Doctor of Philosophy, Chemistry, This thesis will discuss two projects aimed at improving the computational cost of electronic structure methods using the local approximation. These are methods that take advantage of the local nature of the Hamiltonian in real space to provide low cost approximations to solutions of the Schrodinger equation. The first project is a modification of local correlation methods where a set of Hartree-Fock orbitals are produced to be compact in real space, then correlations are restricted to orbitals that are nearby, typically using perturbation theory or coupled cluster. These orbitals are usually constructed using a projection after variation approach where the canonical orbitals are produced, then localized through unitary transformation and orthogonalization tails are projected out. We instead consider a variation after projection approach, where we construct a set of localized constraints on the orbitals, then optimize them in those constraints. We will show that this approach produces local orbitals that are closer to the canonical orbitals and are very well-suited to local correlation techniques. The second project is a method for calculating the correlation energy of strongly correlated systems. The main idea is to construct local orbitals, then divide them into groups called clusters typically chosen by proximity. The orbitals in each cluster are then used to construct a local Fock space with which we use as a basis of matrix product states. First as a single product state called the cluster mean field(cMF) wavefunction, which can then be improved with the cluster versions of perturbation theory and coupled cluster. We will show that this method scales substantially better than other available methods and provides a high quality result that is systematically improvable. , d2026-keyes, Wesleyan University Library, Wesleyan University, Middletown, CT, USA, Quantum chemistry, Electronic structure theory, Local correlation, Matrix product states, Perturbation theory, Coupled cluster theory, Cluster mean field, Text, English, In Copyright, born digital, University Archives

DOI: 10.14418/8qbns2ce

Related papers

Modeling Electron Correlation In Quantum Chemistry

Eric Neuscamman

Accurately predicting the effects of correlations caused by strong interactions between electrons represents one of the key unsolved challenges in quantum chemistry. This thesis details theoretical developments addressing two aspects of this challenge. First, advances in Canonical Transformation (CT) theory address the task of describing weak interactions between the chemically active valence electrons and the chemically inert core electrons when the valence electrons are strongly correlated. Second, new methodologies for Jastrow factor wavefunctions are explored in an attempt to describe strong interactions amongst the valence electrons. In its original form, CT theory suffered from poorly understood numerical instabilities in strongly interacting systems and poor accuracy relative to coupled cluster theory in weakly interacting systems. Here the numeric issues are identified as consequences of CT theory's central approximation and shown to be similar to the intruder states of many-body perturbation theory. It is demonstrated that these instabilities can be avoided by placing particular restrictions on the theory's degrees of freedom. In weakly interacting systems, we show that using a quadratic, rather than linear, commutator approximation improves accuracy, and that the accuracy in any system can be improved by making use of the system's 3-body reduced density matrix. These improvements are demonstrated in a variety of systems, including N2 , H2 O, NiO, and free base porphin. Jastrow factor wavefunctions have traditionally been handled using quan- tum Monte Carlo techniques such as variational Monte Carlo (VMC). Here we explore how VMC can be made to accommodate the unconventionally large numbers of variational parameters present in many-body Jastrow factors and show that the stochastic reconfiguration algorithm can be thus improved by employing the method of conjugate gradients to solve its central linear equation. In addition to extending the reach of VMC, we have also developed nonstochastic methods for working with Jastrow factor wavefunctions that do not rely on random sampling. These methods, similar in design to those used with the coupled cluster wavefunction, produce energies very close to those of VMC in applications to the Heisenberg, spinless Hubbard, and full Hubbard model Hamiltonians.

Aug 31, 20110 citations

Coupled clusters tailored by matrix product state wave functions

Andrej Antalík

The central problem in the modern electronic structure theory is the calculation of cor- relation energy, possibly by an approach that would account for both static and dynamic correlation in an efficient, balanced and accurate way. In this thesis, I present a collection of methods that combine the effective treatment of dynamic correlation by the coupled cluster theory with density matrix renormalization group, a well-established technique for calculations of strongly correlated systems. The connection between them is achieved via the tailored coupled clusters (TCC) ansatz, which conveniently does not impose any ad- ditional computational costs. After the successful initial assessment, we developed more efficient implementations of these methods by employing the local approaches based on pair natural orbitals. This way, we extended the range of possible applications to larger systems with thousands of basis functions. To assess the accuracy of TCC as well as its local counterparts, we performed a variety of benchmark calculations ranging from small, yet challenging systems such as the nitrogen molecule or tetramethyleneethane diradical, to larger molecules like oxo-Mn(Salen) or Fe(II)-porphyrin model. 1

Mar 30, 20210 citations

Tensor-Structured Techniques for Electronic Structure Calculations

Lin, Ian

Electronic structure calculations have been one of the most successful scientific fields in the past 50 years. Density functional theory (DFT) with its nice balance between accuracy and efficiency has now become the standard technique for many materials research. On the other hand, the size of systems that wavefunction-based methods can handle has also been improved thanks to the recent developments in both computation power and more advanced methodologies. However, these methods are still suffering from the curse of dimensionality. The cubic scaling of a typical DFT calculation restricts the handleable system size of DFT to only a few thousand electrons. To this end, tensor-structured techniques provide a route to constructing a reduced-order algorithm and hence are promising to improve the computational efficiency. In this dissertation work, we aim at exploring the applications of the tensor-structured techniques on different aspects of electronic structure calculations. In the first part of this work, we present a tensor-structured algorithm for efficient large-scale DFT calculations by constructing a Tucker tensor basis that is adapted to the Kohn–Sham Hamiltonian and localized in real-space. The proposed approach uses an additive separable approximation to the Kohn–Sham Hamiltonian and an L1 localization technique to generate the 1-D localized functions that constitute the Tucker tensor basis. Numerical results show that the resulting Tucker tensor basis exhibits exponential convergence in the ground-state energy with increasing Tucker rank. Further, the proposed tensor-structured algorithm demonstrated sub-quadratic scaling with system-size for both systems with and without a gap, and involving many thousands of atoms. This reduced-order scaling has also resulted in the proposed approach outperforming plane-wave DFT implementation for systems beyond 2000 electrons. In the second part of this work, we present an acceleration scheme for the proposed tensor-structured algorithm. In the proposed algorithm, the discrete Kohn-Sham problem is solved using the Chebyshev filtering subspace iteration method that relies on matrix-matrix multiplications of a sparse symmetric Hamiltonian matrix and a dense wavefunction matrix, expressed in the localized Tucker tensor basis. These matrix-matrix multiplication operations, which constitute the most computationally intensive step of the solution procedure, are GPU accelerated providing ~8-fold GPU-CPU speedup for these operations on the largest systems studied. The computational performance of the GPU accelerated code is presented using benchmark studies on aluminum nano-particles and silicon quantum dots with system sizes ranging up to ~7,000 atoms. Finally, we present a computation kernel using tensor-structured techniques to evaluate the one-electron and two-electron integrals, which are the central quantities to construct the Hamiltonian matrix of a full configuration interaction (FCI) calculation, for any molecular orbitals projected on a tensor-structured finite-element mesh. The proposed computation kernel is used to compute the integrals using Hartree-Fock molecular orbitals, which is subsequently used to perform an FCI calculation. In our numerical study, the FCI energy using integrals from the proposed computation kernel is confirmed to be consistent with a standard FCI calculation using the same Hartree-Fock orbitals. The proposed computation kernel provides a useful tool to investigate the behavior of Kohn-Sham orbitals as a basis for FCI calculations in the future, which could be a route to construct a reduced-order basis for an FCI calculation to reduce the computational cost and enable FCI calculation for the previously computationally inaccessible systems.

Jan 1, 20220 citations

Coupled Cluster Theory in Materials Science

Igor Ying Zhang, Andreas Grüneis

The workhorse method of computational materials science is undeniably density functional theory in the Kohn-Sham framework of approximate exchange and correlation energy functionals. However, the need for highly accurate predictions of ground and excited state properties in materials science motivates the further development and exploration of alternative as well as complementary techniques. Among these alternative approaches, quantum chemical wavefunction based theories and in particular coupled cluster theory hold the promise to fill a gap in the toolbox of computational materials scientists. Coupled cluster (CC) theory provides a compelling framework of approximate infinite-order perturbation theory in the form of an exponential of cluster operators describing the true quantum many-body effects of the electronic wave function at a computational cost that, despite being significantly more expensive than DFT, scales polynomially with system size. The hierarchy of size-extensive approximate methods established in the framework of CC theory achieves systematic improvability for many materials properties. This is in contrast to currently available density functionals that often suffer from uncontrolled approximations that limit the accuracy in the prediction of materials properties. In this tutorial-style review we will introduce basic concepts of coupled cluster theory and recent developments that increase its computational efficiency for calculations of molecules, solids and materials in general. We will touch upon the connection between coupled cluster theory and the random-phase approximation to bridge the gap between traditional quantum chemistry and many-body Green’s function theories that are widely-used in the field of solid state physics. We will discuss various approaches to improve the computational performance without compromising accuracy. These approaches include large-scale parallel design as well as techniques that reduce the prefactor of the computational complexity. A central part of this article discusses the convergence of calculated properties to the thermodynamic limit which is of significant importance for reliable predictions of materials properties and constitutes an additional challenge compared to calculations of large molecules. Furthermore we mention technical aspects of computer code implementations of periodic coupled cluster theories in different numerical frameworks of the one-electron orbital basis; the projector-augmented-wave formalism using a plane wave basis set and the numeric atom-centered-orbital with resolution-of-identity.

arXiv:2004.06424Frontiers in MaterialsJun 5, 2019138 citations

Direct density matrix renormalization group approaches for strong correlation effects in quantum chemistry

Philipp Bertone Snajberk

An implementation of the single-site density matrix renormalization group (DMRG) approach for the quantum-chemical application to calculate approximate as well as exact electronic ground-state energies and wave functions of atoms and molecules is presented.The algorithm is formulated in terms of the variational and sizeconsistent class of wave functions, matrix product states, and their operator analogues, matrix product operators.A procedure to construct quantum-chemical Hamiltonian operators with long-range Coulomb interactions as matrix product operators is developed.The many-electron Hamiltonian operators are represented as sums and products of one-electron operators.For the efficient matrix product operator construction of quantum-chemical Hamiltonians, different basis spaces including discrete position and momentum space as well as molecular and atomic orbitals are investigated.Discrete position space yields the best matrix product operator size scaling, but introduces artificial long-range couplings in the kinetic energy term of the Hamiltonian which are a consequence of the DMRG algorithm in terms of matrix product states.The huge amount of discrete space points needed for an accurate description poses a problem because of the electron-electron interaction.Therefore, a momentum space or Fourier series expansion of the electron-electron interaction is shown which makes it possible to compress its matrix product operator representation.DMRG applied to quantum-chemical Hamiltonians in matrix product operator form can serve as means to directly (without mean-field treatment in advance) investigate quantum-chemical problems and find the analytically exact ground-state solution of small atoms and molecules.The next basis space analyzed in detail is the atomic orbital basis.The matrix product operator construction and the direct application of the DMRG method in this basis is examined.A physical compression scheme based on a singular value decomposition of the electron-electron interaction in the atomic orbital basis is demonstrated.Furthermore, DMRG calculations in the atomic orbital basis for certain systems where mean-field theory yields qualitatively wrong results and where higher-order approximative correction methods like truncated configuration interaction and coupled cluster cannot fully capture strong electron correlation effects are compared to results obtained in the molecular orbital basis.The comparison shows that atomic orbitals describe the physical situation well and thus lead to a tremendous reduction of the matrix product state dimension and the computational effort.i

Jan 1, 20171 citation

State-of-the-art local correlation methods enable affordable gold standard quantum chemistry for up to hundreds of atoms

Péter R. Nagy

We review the current capabilities of local electron correlation methods up to the coupled cluster model with single, double, and perturbative triple excitations [CCSD(T)], which is a gold standard in quantum chemistry. The main computational aspects of the local method types are assessed from the perspective of applications, but the focus is kept on how to achieve chemical accuracy (i.e., <1 kcal/mol uncertainty), as well as on the broad scope of chemical problems made accessible. The performance of state-of-the-art methods are also compared, including the most employed DLPNO and, in particular, our local natural orbital (LNO) CCSD(T) approach. The high accuracy and efficiency of the LNO method makes chemically accurate CCSD(T) computations accessible for molecules of hundreds of atoms with resources affordable to a broad computational community (days on a single CPU and 10–100 GB of memory). Recent developments in LNO-CCSD(T) enable systematic convergence and robust error estimates even for systems of complicated electronic structure or larger size (up to 1000 atoms). The predictive power of current local CCSD(T) methods, usually at about 1–2 order of magnitude higher cost than hybrid density functional theory (DFT), has become outstanding on the palette of computational chemistry applicable for molecules of practical interest. We also review more than 50 LNO-based and other advanced local-CCSD(T) applications for realistic, large systems across molecular interactions as well as main group, transition metal, bio-, and surface chemistry. The examples show that properly executed local-CCSD(T) can contribute to binding, reaction equilibrium, rate constants, etc. which are able to match measurements within the error estimates. These applications demonstrate that modern, open-access, and broadly affordable local methods, such as LNO-CCSD(T), already enable predictive computations and atomistic insight for complicated, real-life molecular processes in realistic environments.

Chemical ScienceJan 1, 202471 citations