Review
Chemistry, Physical
Prachi Sharma, Jie J. Bao, Donald G. Truhlar, Laura Gagliardi
Summary: Kohn-Sham density functional theory is less accurate for strongly correlated systems compared to weakly correlated systems. The available functionals for spin densities do not accurately predict energies for strongly correlated systems when using multiconfigurational wave functions with spin symmetry. Multiconfiguration pair-density functional theory overcomes these limitations by using a functional of the total density and on-top pair density, allowing efficient calculation of energy for strongly correlated systems.
ANNUAL REVIEW OF PHYSICAL CHEMISTRY, VOL 72
(2021)
Article
Chemistry, Physical
Paul B. Calio, Donald G. Truhlar, Laura Gagliardi
Summary: We present the first implementation of multiconfiguration pair-density functional theory (MC-PDFT) ab initio molecular dynamics, which has similar or even better accuracy compared to complete active space second-order perturbation theory (CASPT2) but at a lower computational cost. Our implementation of MC-PDFT into the SHARC molecular dynamics program for ab initio, nonadiabatic molecular dynamics simulations is verified through the examination of the intersystem crossing dynamics of thioformaldehyde, showing excellent agreement with CASPT2 and experimental findings. Furthermore, MC-PDFT enables dynamics simulations with large active spaces that were computationally expensive with CASPT2.
JOURNAL OF CHEMICAL THEORY AND COMPUTATION
(2022)
Article
Chemistry, Physical
Abhishek Mitra, Matthew R. R. Hermes, Laura Gagliardi
Summary: We propose a quantum embedding method that combines multiconfiguration pair-density functional theory (MC-PDFT) with periodic density matrix embedding theory (pDMET) to calculate ground and excited states of extended systems. By applying this method, we accurately compute local excitations in oxygen mono- and divacancies on a magnesium oxide (100) surface, with deviations of only 0.05 eV compared to the nonembedded MC-PDFT approach. We also demonstrate the efficiency of our method by calculating local excitations in larger supercells for the monovacancy defect.
JOURNAL OF CHEMICAL THEORY AND COMPUTATION
(2023)
Article
Chemistry, Physical
Aleksandr O. Lykhin, Donald G. Truhlar, Laura Gagliardi
Summary: The dipole moment directly indicates molecular polarity, and the accuracy of computed dipole moments relies on the calculated electron density quality. The multiconfiguration pair-density functional theory (MC-PDFT) shows reasonable behavior in predicting dipole moments, especially for stretched geometries, and it significantly improves electron correlation compared to other methods. The recommended on-top density functionals and active space selection lead to more accurate predictions of dipole moments, with MC-PDFT and its hybrid extension (HMC-PDFT) showing comparable performance to other advanced methods.
JOURNAL OF CHEMICAL THEORY AND COMPUTATION
(2021)
Article
Chemistry, Physical
Prachi Sharma, Andrew J. Jenkins, Giovanni Scalmani, Michael J. Frisch, Donald G. Truhlar, Laura Gagliardi, Xiaosong Li
Summary: A new computational method called X2C-MC-PDFT is proposed to account for both electron correlation and relativistic effects in a computationally inexpensive way. Benchmark studies show that X2C-MC-PDFT significantly improves the results compared to previous methods.
JOURNAL OF CHEMICAL THEORY AND COMPUTATION
(2022)
Article
Chemistry, Physical
Dihua Wu, Chen Zhou, Jie J. Bao, Laura Gagliardi, Donald G. Truhlar
Summary: Zero-field splitting (ZFS) is a fundamental molecular property that is important for various applications. Predicting ZFS parameters accurately requires addressing the challenge of including external correlation. This paper presents a cost-effective method, MC-PDFT, which combines spin-orbit-inclusive multiconfiguration and multistate pair-density functional theory, to accurately calculate ZFS.
JOURNAL OF CHEMICAL THEORY AND COMPUTATION
(2022)
Article
Chemistry, Physical
Dayou Zhang, Matthew R. Hermes, Laura Gagliardi, Donald G. Truhlar
Summary: The paper introduces a new theory named MC-DCFT, which provides a new route to define density functionals for multiconfiguration wave functions by using the one-particle density matrix in the coordinate representation. The theory is demonstrated by calculating dissociation curves of four diatomic molecules and achieving comparable accuracy with other calculation methods by introducing two parameters in converted density functionals.
JOURNAL OF CHEMICAL THEORY AND COMPUTATION
(2021)
Article
Chemistry, Physical
Thais R. Scott, Meagan S. Oakley, Matthew R. Hermes, Andrew M. Sand, Roland Lindh, Donald G. Truhlar, Laura Gagliardi
Summary: Density fitting reduces computational cost by avoiding computation and manipulation of four-index electron repulsion integrals, allowing efficient optimization of large systems. The new approach increases computation speed with negligible loss in accuracy, while smaller grid sizes further reduce costs with minimal impact on optimized geometries and gradient values.
JOURNAL OF CHEMICAL PHYSICS
(2021)
Article
Chemistry, Multidisciplinary
Chen Zhou, Matthew R. Hermes, Dihua Wu, Jie J. Bao, Riddhish Pandharkar, Daniel S. King, Dayou Zhang, Thais R. Scott, Aleksandr O. Lykin, Laura Gagliardi, Donald G. Truhlar
Summary: This article provides an introduction to strongly correlated systems and multiconfiguration pair-density functional theory (MC-PDFT) which is used to quantitatively treat near-degeneracy correlation and dynamic correlation in these systems. The article also discusses recent developments and applications of MC-PDFT and related methods.
Article
Chemistry, Physical
Daniel S. King, Donald G. Truhlar, Laura Gagliardi
Summary: The selection of active orbitals for modeling strongly correlated electronic states is challenging and highly dependent on the specific states and molecules of interest. The authors propose a discrete variational selection (DVS) approach to tackle this problem by generating multiple trial wave functions from systematically constructed active spaces and choosing the best one variationally. They apply this approach to a database of small-to-medium-sized molecules and find that it performs well in selecting wave functions based on the energy from multiconfiguration pair-density functional theory.
JOURNAL OF CHEMICAL THEORY AND COMPUTATION
(2023)
Article
Biochemistry & Molecular Biology
Meagan S. Oakley, Laura Gagliardi, Donald G. Truhlar
Summary: The study utilized CAS and SP wave functions to simulate the bond dissociation of transition metal silicides and predict experimental results, with typical errors of 0.2 eV and 0.3 eV, respectively. The ground and low-lying excited states were explored to predict state ordering for each molecule, and potential energy curves were calculated for comparison to experimental results. English Summary: The study utilized CAS and SP wave functions to simulate the bond dissociation of transition metal silicides and predict experimental results, with typical errors of 0.2 eV and 0.3 eV, respectively.
Article
Chemistry, Physical
Andrew M. Sand, Justin T. Malme, Erik P. Hoy
Summary: Single-molecule electronics have attracted research interest due to their small size and unique properties. A new approach, NEGF-MCPDFT, combining MCPDFT with NEGF, allows for efficient inclusion of electron correlations in the electronic structure description. The method was used to study alkane and benzyne junctions, showing agreement with DFT-NEGF results for alkane junctions and capturing additional electron correlation effects for benzyne junctions.
JOURNAL OF CHEMICAL PHYSICS
(2021)
Article
Chemistry, Physical
Chen Zhou, Dihua Wu, Laura Gagliardi, Donald G. Truhlar
Summary: The study demonstrates how spin-orbit coupling can be included in MC-PDFT and applied to the calculation of magnetic g tensors for transition-metal molecules. The MC-PDFT methods are shown to be superior in both accuracy and efficiency compared to complete active space second-order perturbation theory.
JOURNAL OF CHEMICAL THEORY AND COMPUTATION
(2021)
Article
Chemistry, Physical
Dayou Zhang, Donald G. Truhlar
Summary: We present a hybrid multiconfiguration density coherence functional theory (HMC-DCFT) and optimize a density coherence functional by parametrization against a diverse dataset of 59 bond energies and 60 barrier heights. The results are compared to calculations using CASSCF, CASPT2, six Kohn-Sham and hybrid Kohn-Sham exchange-correlation functionals, and three on-top functionals for pair-density functional theory (PDFT) and hybrid PDFT. The new functional outperforms all the compared methods.
JOURNAL OF CHEMICAL THEORY AND COMPUTATION
(2023)
Article
Materials Science, Multidisciplinary
Xuecheng Shao, Wenhui Mi, Michele Pavanello
Summary: Nonlocal kinetic energy functionals with a density-dependent kernel are the most accurate functionals for orbital-free density functional theory simulations. The HC functional is the most accurate for bulk semiconductors, but faces numerical instability for nonbulk systems. A revision, revHC, has been proposed to achieve improved numerical stability, efficiency, and applicability, particularly for surfaces of semiconductor materials.
Article
Chemistry, Physical
Jiaxin Ning, Donald G. Truhlar
Summary: Using theoretical computations, we studied the photodissociation process of ortho-Fluorothiophenol (o-FTP). We found that the introduction of a fluorine atom in o-FTP leads to the formation of a hydrogen bond with the hydrogen atom of the SH group, which is not commonly observed in other pi sigma* processes. Theoretical calculations are important for studying these reactions as they provide a deeper understanding of the processes compared to current spectroscopic methods.
JOURNAL OF PHYSICAL CHEMISTRY A
(2023)
Article
Chemistry, Physical
Aleksandr O. Lykhin, Moritz K. A. Baumgarten, Donald G. Truhlar, Laura Gagliardi
Summary: We develop response-function algorithms for dipole moments and transition dipole moments for compressed multistate pair-density functional theory (CMS-PDFT). We use the method of undetermined Lagrange multipliers to derive analytical expressions and validate them using numerical differentiation. We show that CMS-PDFT has good accuracy for ground-state and excited-state dipole moments, and can provide correct behavior for the dipole moment curves in the vicinity of conical intersections, which opens the door to molecular dynamic simulations in strong electric fields and the discovery of chemical reactions that can be controlled by an oriented external electric field upon photoexcitation of the reactants.
JOURNAL OF PHYSICAL CHEMISTRY A
(2023)
Article
Chemistry, Physical
Jiaxin Ning, Donald G. Truhlar
Summary: The geometries, spin states, and bonding characteristics of NdO2 and SmO22+ were studied and compared using spin-orbit-free wave functions. The ground spin states and structures were found to be different for the two molecules, with NdO2 favoring a linear ONdO triplet structure and SmO22+ favoring a linear SmOO2+ quintet structure. The bonding characteristics were investigated using state-averaged complete active-space self-consistent-field (SA-CASSCF) calculations, which showed electron transfer from Nd to O in NdO2 but not between Sm and O in SmO22+.
JOURNAL OF PHYSICAL CHEMISTRY A
(2023)
Article
Chemistry, Physical
Farideh Badichi Akher, Yinan Shu, Zoltan Varga, Suman Bhaumik, Donald G. Truhlar
Summary: This article presents a method of adding a new type of activation function to a neural network to enforce low-dimensional constraints. By using this method, the interaction potential can be forced to approach zero when subsystems are too far separated to interact, even without sufficient training data. Additionally, improved potential energy surfaces for the 14 lowest (3)A' states of O-3 are provided, and a more general method called parametrically managed diabatization by deep neural network (PM-DDNN) is introduced to add other low-dimensional or lower-level knowledge to machine-learned potentials.
JOURNAL OF PHYSICAL CHEMISTRY A
(2023)
Article
Chemistry, Physical
Maryam Mansoori Kermani, Donald G. Truhlar
Summary: We discovered that two different basis sets labeled as cc-pVxZ are widely used in calculations on Li compounds. However, these calculations produce different results, causing reproducibility issues. In this study, we examined the discrepancies between these supposedly equivalent basis sets for wave function and density functional calculations of Li clusters. We found deviations as large as 22 millihartrees among the basis sets. The noncanonical choices generally resulted in lower energies. Therefore, it is crucial to specify the version of the basis set used in calculations.
CHEMICAL PHYSICS LETTERS
(2023)
Article
Computer Science, Interdisciplinary Applications
Wenqi Chen, Jingjing Zheng, Junwei Lucas Bao, Donald G. Truhlar, Xuefei Xu
Summary: The MSTor program is a computer program that calculates partition functions and thermodynamic functions of complex gas-phase molecules with multiple torsions. It can use multi-structural approximation with torsional anharmonicity (MS-T) based on either a coupled torsional potential or an uncoupled torsional potential. The program also includes eight utility codes for generating input files and comparison results. The 2023 version of MSTor introduces automatic identification of torsional modes with redundant internal coordinates.
COMPUTER PHYSICS COMMUNICATIONS
(2023)
Article
Chemistry, Physical
Farideh Badichi Akher, Yinan Shu, Zoltan Varga, Donald G. Truhlar
Summary: Dynamics simulations of high-energy O(2)-O collisions are crucial for understanding thermal energy content and heat flux in flows around hypersonic vehicles. Recently, a new diabatization method called parametrically managed diabatization by deep neural network (PM-DDNN) has been developed to efficiently treat collisions involving many coupled electronic states. In this study, the PM-DDNN method was applied to the six lowest-energy potential energy surfaces in the O-3 (5) A′ manifold to perform simultaneous diabatization and fitting, and three dynamic methods were used for calculations. The results showed good agreement among the three methods, and electronically nonadiabatic, electronically inelastic, and dissociative cross sections were calculated for various initial conditions.
JOURNAL OF CHEMICAL THEORY AND COMPUTATION
(2023)
Article
Chemistry, Physical
Dayou Zhang, Donald G. Truhlar
Summary: We present a hybrid multiconfiguration density coherence functional theory (HMC-DCFT) and optimize a density coherence functional by parametrization against a diverse dataset of 59 bond energies and 60 barrier heights. The results are compared to calculations using CASSCF, CASPT2, six Kohn-Sham and hybrid Kohn-Sham exchange-correlation functionals, and three on-top functionals for pair-density functional theory (PDFT) and hybrid PDFT. The new functional outperforms all the compared methods.
JOURNAL OF CHEMICAL THEORY AND COMPUTATION
(2023)
Article
Chemistry, Physical
Maryam Mansoori Kermani, Hanwei Li, Alistar Ottochian, Orlando Crescenzi, Benjamin G. Janesko, Giovanni Scalmani, Michael J. Frisch, Ilaria Ciofini, Carlo Adamo, Donald G. Truhlar
Summary: Theoretical characterization of reactions of complex molecules requires accurate determination of the relative energies of intermediates and transition states. In this study, the DLPNO-CCSD(T) method was employed to provide benchmark values for Diels-Alder transition states leading to strained pentacyclic adducts. Various wave function and density functional methods were compared for their prediction of absolute and relative barrier heights. The results show that only a few density functionals can satisfactorily predict absolute barrier heights, while relative barrier heights are more accurate. The findings of this study can guide the selection of density functionals for future studies on crowded, strained transition states of large molecules.
JOURNAL OF PHYSICAL CHEMISTRY LETTERS
(2023)
Article
Chemistry, Multidisciplinary
Bo Long, Yu Xia, Yu-Qiong Zhang, Donald G. Truhlar
Summary: In this study, rate constants for the reaction of sulfur trioxide with two water molecules were calculated. It was found that the SO3 + H2O + H2O reaction mechanism is more important than the SO3 + (H2O)2 mechanism at altitudes between 5 and 50 km, and the rate ratio between the two mechanisms is greater than 20 at altitudes between 10 and 35 km.
JOURNAL OF THE AMERICAN CHEMICAL SOCIETY
(2023)
Article
Chemistry, Multidisciplinary
Ying Yang, Siriluk Kanchanakungwankul, Suman Bhaumik, Qing Ma, Sol Ahn, Donald G. Truhlar, Joseph T. Hupp
Summary: In this study, a Cu-doped ZIF-8 metal-organic framework was synthesized and activated by heating and dosing with O2, methane, and steam. The framework showed catalytic activity for the oxidation of methane to methanol and formaldehyde. Characterization and analysis techniques were used to understand the mechanism of the catalytic reaction.
JOURNAL OF THE AMERICAN CHEMICAL SOCIETY
(2023)
Article
Chemistry, Physical
Yinan Shu, Donald G. Truhlar
Summary: Decoherence refers to the phenomenon where the density matrix of a subsystem evolves into a statistical ensemble of states instead of a coherent combination of pure-state wave functions. In molecular processes involving changes in electronic states and nuclear coordinates, the interaction between the electronic and nuclear subsystems leads to decoherence of the electronic subsystem. This paper provides background information on decoherence and discusses its treatment in the coherent switching with decay of mixing algorithm and the trajectory surface hopping method for semiclassical calculations of electronically nonadiabatic processes.
JOURNAL OF CHEMICAL THEORY AND COMPUTATION
(2023)
Article
Computer Science, Interdisciplinary Applications
Yiwei Liu, Cheng Zhang, Zhonghua Liu, Donald G. Truhlar, Ying Wang, Xiao He
Summary: In this study, we optimized a singly hybrid functional called CF22D, which outperforms most existing non-doubly hybrid functionals in terms of accuracy in chemistry. By using a flexible functional form that combines a global hybrid meta-nonseparable gradient approximation with a damped dispersion term, we trained CF22D using a large database and performance-triggered iterative supervised training. The use of a combined database demonstrated CF22D's good performance across various chemical properties and systems.
NATURE COMPUTATIONAL SCIENCE
(2023)
Review
Chemistry, Physical
Indrani Choudhuri, Jingyun Ye, Donald G. Truhlar
Summary: Metal-organic frameworks (MOFs) have exceptional properties for various functions. Experimental characterization and examination of all potentially useful combinations of MOFs are practically impossible. Computational modeling has become an important tool for understanding the functionalization and mechanisms of MOFs.
CHEMICAL PHYSICS REVIEWS
(2023)
Article
Multidisciplinary Sciences
Zhihong Hu, Qimei Di, Bingzhi Liu, Yanbo Li, Yunrui He, Qingbo Zhu, Qiang Xu, Philippe Dagaut, Nils Hansen, S. Mani Sarathy, Lili Xing, Donald G. Truhlar, Zhandong Wang
Summary: Hydroperoxides play crucial roles in atmospheric oxidation, combustion, and catalytic reactions. This study developed a new method for synthesizing organic hydroperoxides and measured their photoionization cross-sections using mass spectrometry. The findings are important for improving chemical models and studying reaction kinetics.
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA
(2023)