Variational Active Space Selection with MC-PDFT

Daniel S. King, Donald G. Truhlar, Laura Gagliardi

Published Jun 19, 2023• 0 citations

Abstract

The selection of a good set of active orbitals for modeling strongly correlated quantum states is difficult to automate as it is highly dependent on the state and molecule of interest. As such, although many approaches have been proposed with some success, no single approach has worked in all cases. Here, we propose an improved framework for automated selection in which (i) multiple wave functions based on different active-spaces are generated, and (ii) the resulting wave functions are chosen between by some means of facile evaluation. Using this framework, we propose a method in which (i) we construct different active space orbitals through diagonalization of a parameterized operator, and (ii) we choose the state with the lowest tPBE absolute energy from multiconfigurational pair density functional theory (MC-PDFT) averaged over the targeted states of interest. We test the method using density matrix renormalization group (DMRG) wave functions with 40 active orbitals and bond index of 700, with no further orbital optimization following the active space selection. We find that with only four values of the parameter (i.e., with only four trial wave functions), we can obtain a mean unsigned error of only 0.19 eV for 199 vertical excitation energies in the QUESTDB database. Furthermore, the tPBE absolute energy proves robust in selecting between active spaces of very different sizes, over and above the wave function (CASCI/DMRG-CASCI) absolute energy. We believe this new framework is promising for the application of active space methods to chemical problems in a high-throughput fashion.

DOI: 10.26434/chemrxiv-2023-9b3g4

Related papers

Active Space Selection Based on Natural Orbital Occupation Numbers from n-Electron Valence Perturbation Theory

Abhishek Khedkar, Michael Roemelt

Efficient and robust approximations to the full configuration interaction (full-CI) method such as the density matrix renormalization group (DMRG) and the full-CI quantum Monte Carlo (FCIQMC) algorithm allow for multiconfigurational self-consistent field (MC-SCF) calculations of molecules with many strongly correlated electrons. This opens up the possibility to treat large and complex systems that were previously untractable, but at the same time it calls for an efficient and reliable active space selection as the choice of how many electrons and orbitals enter the active space is critical for any multireference calculation. In this work we propose an Active Space Selection based on 1st order perturbation theory (ASS1ST) that follows a "bottom-up" strategy and utilizes a set of quasi-natural orbitals together with sensible thresholds for their occupation numbers. The required quasi-natural orbitals are generated by diagonalizing the virtual and internal part of the one-electron reduced density matrix that is obtained from strongly contracted n-electron valence perturbation theory (SC-NEVPT) on top of a minimal active space calculation. Self-consistent results can be obtained when the proposed selection scheme is applied iteratively. Initial applications on four chemically relevant benchmark systems indicate the capabilities of ASS1ST. Eventually, the strengths and limitations are critically discussed.

Journal of Chemical Theory and ComputationMay 6, 201965 citations

Machine Learning-Assisted Selection of Active Spaces for Strongly Correlated Transition Metal Systems

Pavlo Golub, Andrej Antalík, Libor Veis +1 more

Active space quantum chemical methods could provide very accurate description of strongly correlated electronic systems, which is of tremendous value for natural sciences. The proper choice of the active space is crucial but a nontrivial task. In this article, we present a neural network-based approach for automatic selection of active spaces, focused on transition metal systems. The training set has been formed from artificial systems composed of one transition metal and various ligands, on which we have performed the density matrix renormalization group and calculated the single-site entropy. On the selected set of systems, ranging from small benchmark molecules up to larger challenging systems involving two metallic centers, we demonstrate that our machine learning models could predict the active space orbitals with reasonable accuracy. We also tested the transferability on out-of-the-model systems, including bimetallic complexes and complexes with ligands, which were not involved in the training set. Also, we tested the correctness of the automatically selected active spaces on a Fe(II)-porphyrin model, where we studied the lowest states at the DMRG level and compared the energy difference between spin states or the energy difference between conformations of ferrocene with recent studies.

Journal of Chemical Theory and ComputationSep 27, 202137 citations

Automated Selection of Active Orbital Spaces

Christopher Johannes Stein, Markus Reiher

One of the key challenges of quantum-chemical multi-configuration methods is the necessity to manually select orbitals for the active space. This selection requires both expertise and experience and can therefore impose severe limitations on the applicability of this most general class of ab initio methods. A poor choice of the active orbital space may yield even qualitatively wrong results. This is obviously a severe problem, especially for wave function methods that are designed to be systematically improvable. Here, we show how the iterative nature of the density matrix renormalization group combined with its capability to include up to about 100 orbitals in the active space can be exploited for a systematic assessment and selection of active orbitals. These benefits allow us to implement an automated approach for active orbital space selection, which can turn multi-configuration models into black box approaches.

arXiv:1602.03835Journal of Chemical Theory and ComputationMar 9, 2016326 citations

TNASS: Tensor Network Active Space Selection with the Entanglement Feature

Angus Mingare, Isabelle Heuzé, Peter V. Coveney

The quality of multi-scale modelling techniques in molecular electronic structure calculations, such as embedding and subspace methods, relies upon the chosen active space. The automation of active space selection is vital for ensuring the accuracy, reproducibility, and scalability in such calculations. In this work, we introduce Tensor Network Active Space Selection using the Entanglement Feature. Through the isolation of strongly correlated electrons, this method provides a scalable foundation for embedding methods in multi-scale modelling. By representing the purities of all possible orbital partitions as a Matrix Product State, our method isolates regions of strong electron correlation without requiring manual preselection of target atoms or the calculation of expensive high-order density matrices. The results demonstrate that this approach leads to lower ground state energies and more accurate dipole moments than other fully automated selection schemes such as those based solely on single-orbital entropy or the selection of spatial orbitals around the HOMO/LUMO gap.

physics.chem-phCross-listedarXiv:2608.03645Aug 5, 20260 citations

The density matrix renormalization group self-consistent field method: Orbital optimization with the density matrix renormalization group method in the active space

Dominika Zgid, Marcel Nooijen

We present the density matrix renormalization group self-consistent field (DMRG-SCF) approach that is analogous to the complete active space self-consisted field (CASSCF) method but instead of using for the description of the active space the full configuration interaction (FCI) method, the DMRG-SCF uses the density matrix renormalization group (DMRG) method. The DMRG-SCF approach, similarly to CASSCF, properly describes the multiconfigurational character of the wave function but avoids the exponential scaling of the FCI method and replaces it with a polynomial scaling. Hence, calculations for a larger number of orbitals and electrons in the active space are possible since the DMRG method provides an efficient tool to automatically select from the full Hilbert space the many-body contracted basis states that are the most important for the description of the wave function.

The Journal of Chemical PhysicsApr 11, 2008202 citations

A Perturbative Density Matrix Renormalization Group Algorithm for Large Active Spaces

Sheng Guo, Zhendong Li, Garnet Kin‐Lic Chan

We describe a low cost alternative to the standard variational DMRG (density matrix renormalization group) algorithm that is analogous to the combination of the selected configuration interaction plus perturbation theory (SCI+PT). We denote the resulting method p-DMRG (perturbative DMRG) to distinguish it from the standard variational DMRG. p-DMRG is expected to be useful for systems with very large active spaces, for which variational DMRG becomes too expensive. Similar to SCI+PT, in p-DMRG, a zeroth-order wave function is first obtained by a standard DMRG calculation but with a small bond dimension. Then, the residual correlation is recovered by a second-order perturbative treatment. We discuss the choice of partitioning for perturbation theory, which is crucial for its accuracy and robustness. To circumvent the problem of a large bond dimension in the first-order wave function, we use a sum of matrix product states to expand the first-order wave function, yielding substantial savings in computational cost and memory. We also propose extrapolation schemes to reduce the errors in the zeroth- and first-order wave functions. Numerical results for Cr 2 with a (28e, 76o) active space and 1,3-butadiene with an (22e, 82o) active space reveal that p-DMRG provides ground state energies of a similar quality to variational DMRG with very large bond dimensions but at a significantly lower computational cost. This suggests that p-DMRG will be an efficient tool for benchmark studies in the future.

arXiv:1803.07150Journal of Chemical Theory and ComputationJun 21, 201839 citations