Zhenling Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Zhenling Wang has a total of 34 co-authors. Here are the top co-authors by number of shared publications:
- Martin Head-Gordon
University of California
12 shared publications Xiaoye S. Li
Applied Mathematics and Computational Research Division
4 shared publicationsYang Liu
Applied Mathematics and Computational Research Division
4 shared publications- Abdulrahman Aldossary
University of Toronto
3 shared publications Kevin Ikeda
Department of Chemistry, University of California, Berkeley, California 94720, USA
3 shared publications- Tianyi Shi
Scalable Solvers Group, Lawrence Berkeley National Laboratory, Berkeley, CA 94720 USA
3 shared publications
Latest Publications
- 28 October 2026
- Published by AIP Publishing in The Journal of Chemical Physics
- Vol. 165 (16)
Third-order size-consistent Brillouin–Wigner perturbation theory
Low-order Rayleigh–Schrödinger perturbation theory with Møller–Plesset partitioning is size-consistent but not regular, and for this reason, the third-order variant, MP3, is not commonly used in computational quantum chemistry. Recently, it has become clear how to modify Brillouin–Wigner perturbation theory at second order to be both size-consistent and regular, by using a non-Møller–Plesset partitioning to define BWs2. In this work, a solution is presented for the problem of defining an iterative third-order generalization, iBWs3, that is, size-consistent, regular, and exact for the electron correlation of two electrons in two orbitals. iBWs3 can be seen as a scaled symmetrized version of coupled cluster theory with doubles including only the linear and the mosaic quadratic terms, thus with the same compute scaling. An implementation and a series of tests on restricted orbital dissociations and relative energies are reported.
- 28 October 2026
- Published by AIP Publishing in The Journal of Chemical Physics
- Vol. 165 (16)
Third-order size-consistent Brillouin–Wigner perturbation theory
Low-order Rayleigh–Schrödinger perturbation theory with Møller–Plesset partitioning is size-consistent but not regular, and for this reason, the third-order variant, MP3, is not commonly used in computational quantum chemistry. Recently, it has become clear how to modify Brillouin–Wigner perturbation theory at second order to be both size-consistent and regular, by using a non-Møller–Plesset partitioning to define BWs2. In this work, a solution is presented for the problem of defining an iterative third-order generalization, iBWs3, that is, size-consistent, regular, and exact for the electron correlation of two electrons in two orbitals. iBWs3 can be seen as a scaled symmetrized version of coupled cluster theory with doubles including only the linear and the mosaic quadratic terms, thus with the same compute scaling. An implementation and a series of tests on restricted orbital dissociations and relative energies are reported.
- 7 October 2026
- Published by Royal Society of Chemistry (RSC) in Physical Chemistry Chemical Physics
- Vol. 28 (38) , 23426-23441
Using energy decomposition analysis from correlated wavefunction theory to characterize a range of hydrogen, halogen, and tetrel bonds: the relative importance of electrostatics, charge-transfer and polarization
MP2 ALMO-EDA-II is used to characterize the nature of a range of non-covalent interactions, including hydrogen bonds, tetrel bonds, and halogen bonds. MP2 correlation contributes to each term, not just dispersion.
- 7 October 2026
- Published by Royal Society of Chemistry (RSC) in Physical Chemistry Chemical Physics
- Vol. 28 (38) , 23426-23441
Using energy decomposition analysis from correlated wavefunction theory to characterize a range of hydrogen, halogen, and tetrel bonds: the relative importance of electrostatics, charge-transfer and polarization
MP2 ALMO-EDA-II is used to characterize the nature of a range of non-covalent interactions, including hydrogen bonds, tetrel bonds, and halogen bonds. MP2 correlation contributes to each term, not just dispersion.
- 21 August 2026
- Published by American Chemical Society (ACS) in The Journal of Physical Chemistry Letters
- Vol. 17 (35) , 10309-10311
Correction to “Third-Order Perturbation Theory Made Regular: A Noniterative Correction to the Size-Consistent Second-Order Brillouin–Wigner Perturbation Theory”
- 21 August 2026
- Published by American Chemical Society (ACS) in The Journal of Physical Chemistry Letters
- Vol. 17 (35) , 10309-10311
Correction to “Third-Order Perturbation Theory Made Regular: A Noniterative Correction to the Size-Consistent Second-Order Brillouin–Wigner Perturbation Theory”
- Open Access
- 10 August 2026
- Published by American Chemical Society (ACS) in ChemRxiv
Third-order size-consistent Brillouin-Wigner perturbation theory
Low order Rayleigh-Schrödinger (RS) perturbation theory with Møller-Plesset partitioning is size-consistent but not regular, and for this reason, the third order variant, MP3, is not commonly used in computational quantum chemistry. Recently, it has become clear how to modify Brillouin-Wigner perturbation theory at second order to be both size-consistent and regular, by using a non-Møller-Plesset partitioning to define BWs2. In this work, a solution is presented for the problem of defining an iterative third-order generalization, iBWs3, that is size-consistent, regular, and exact for the electron correlation of 2 electrons in 2 orbitals. iBWs3 is closely related to coupled cluster theory with doubles, with the same compute scaling. An implementation and a series of tests on restricted orbital dissociations and relative energies are reported.
- 10 August 2026
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessThird-order size-consistent Brillouin-Wigner perturbation theory
Low order Rayleigh-Schrödinger (RS) perturbation theory with Møller-Plesset partitioning is size-consistent but not regular, and for this reason, the third order variant, MP3, is not commonly used in computational quantum chemistry. Recently, it has become clear how to modify Brillouin-Wigner perturbation theory at second order to be both size-consistent and regular, by using a non-Møller-Plesset partitioning to define BWs2. In this work, a solution is presented for the problem of defining an iterative third-order generalization, iBWs3, that is size-consistent, regular, and exact for the electron correlation of 2 electrons in 2 orbitals. iBWs3 is closely related to coupled cluster theory with doubles, with the same compute scaling. An implementation and a series of tests on restricted orbital dissociations and relative energies are reported.
- Open Access
- 10 August 2026
- Published by American Chemical Society (ACS) in ChemRxiv
Using energy decomposition analysis from correlated wavefunction theory to characterize a range of hydrogen, halogen, and tetrel bonds: The relative importance of electrostatics, charge-transfer and polarization.
Second-order Møller-Plesset theory (MP2) is the simplest wavefunction-based treatment of electron correlation and is a good compromise between accuracy and efficiency for evaluating the strength of many non-covalent intermolecular interactions. It has also been improved in accuracy and robustness by regularized methods, such as the size-consistent second order Brillouin-Wigner perturbation theory (BWs2) and κ-MP2. This work presents a non-technical summary of the recently developed second-generation absolutely localized molecular orbital (ALMO) energy decomposition analysis (EDA) method at the post-SCF level of MP2 and regularized MP2. The so-called frozen term is redefined to ensure that long range correlation contributions to electrostatics are correctly captured. The resulting EDA provides useful basis set limits to the correlation corrections for all physical contributions: donor-acceptor charge transfer, electrical polarization, dispersion, and Pauli repulsions and electrostatic interactions associated with frozen monomer orbitals. EDA calculations using MP2 are reported on several classes of systems: model systems to assess correlation effects on permanent and induced electrostatics, hydrogen bonds between water and model carbohydrate molecules, tetrel bonds ranging from 15 to 90 kJ/mol in strength, two halogen bonds-one of which is known to be solvent resistant, and a pair of “anti-electrostatic” halogen bonds. Several general conclusions emerge. First, to obtain physically reasonable conclusions, it is essential that the correlation contribution is not identified with dispersion alone: large corrections to the frozen term corresponding to changes in electrostatics are observed due to correlation. Second, the fingerprint of intermolecular interactions provided by the EDA reveals a range of interesting trends in the character of intermolecular interactions as a function of their type and strength. For example, the solvent-stable halogen bond is confirmed to be dominated by charge-transfer stabilization. On the other hand, the strongest tetrel bonds studied here are dominated by contributions from polarization, while more conventional hydrogen bonds between water and model carbohydrates represent a synergy between electrostatics, dispersion, and charge transfer. The metastability (or stability) of nominally antielectrostatic interactions depends strongly on the differing distance dependence of attractive and repulsive terms, and we demonstrated tuning based on changes in permanent electrostatics by increasing charge-separation through use of phenyl linkers.
- 10 August 2026
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessUsing energy decomposition analysis from correlated wavefunction theory to characterize a range of hydrogen, halogen, and tetrel bonds: The relative importance of electrostatics, charge-transfer and polarization.
Second-order Møller-Plesset theory (MP2) is the simplest wavefunction-based treatment of electron correlation and is a good compromise between accuracy and efficiency for evaluating the strength of many non-covalent intermolecular interactions. It has also been improved in accuracy and robustness by regularized methods, such as the size-consistent second order Brillouin-Wigner perturbation theory (BWs2) and κ-MP2. This work presents a non-technical summary of the recently developed second-generation absolutely localized molecular orbital (ALMO) energy decomposition analysis (EDA) method at the post-SCF level of MP2 and regularized MP2. The so-called frozen term is redefined to ensure that long range correlation contributions to electrostatics are correctly captured. The resulting EDA provides useful basis set limits to the correlation corrections for all physical contributions: donor-acceptor charge transfer, electrical polarization, dispersion, and Pauli repulsions and electrostatic interactions associated with frozen monomer orbitals. EDA calculations using MP2 are reported on several classes of systems: model systems to assess correlation effects on permanent and induced electrostatics, hydrogen bonds between water and model carbohydrate molecules, tetrel bonds ranging from 15 to 90 kJ/mol in strength, two halogen bonds-one of which is known to be solvent resistant, and a pair of “anti-electrostatic” halogen bonds. Several general conclusions emerge. First, to obtain physically reasonable conclusions, it is essential that the correlation contribution is not identified with dispersion alone: large corrections to the frozen term corresponding to changes in electrostatics are observed due to correlation. Second, the fingerprint of intermolecular interactions provided by the EDA reveals a range of interesting trends in the character of intermolecular interactions as a function of their type and strength. For example, the solvent-stable halogen bond is confirmed to be dominated by charge-transfer stabilization. On the other hand, the strongest tetrel bonds studied here are dominated by contributions from polarization, while more conventional hydrogen bonds between water and model carbohydrates represent a synergy between electrostatics, dispersion, and charge transfer. The metastability (or stability) of nominally antielectrostatic interactions depends strongly on the differing distance dependence of attractive and repulsive terms, and we demonstrated tuning based on changes in permanent electrostatics by increasing charge-separation through use of phenyl linkers.
- 24 April 2026
- Published by American Chemical Society (ACS) in Journal of Chemical Theory and Computation
- Vol. 22 (9) , 4439-4451
An Improved Size-Consistent Second-Order Brillouin–Wigner Perturbation Theory: Which Desirable Properties Are Compatible with Unconditional Size-Consistency and Optimized Chemical Accuracy?
Recently, it has been established that a size-consistent second-order Brillouin–Wigner perturbation (BW-s2) theory can be obtained by satisfying a trace condition with a repartitioning of the Hamiltonian between a one-body reference and a correlation-containing perturbation. BW-s2 has the twin advantages of being regular like the conventional second-order BW theory and being conditionally size-consistent like the conventional second-order Mo̷ller–Plesset theory. Conditional size-consistency means that spin-polarized reference orbitals must be used whenever the spin-restricted orbitals are unstable, whereas unconditional size-consistency means that size-consistency follows from the use of orbitals yielding the lowest BW-s2 energy. There is a wide range of possible forms for the partition, and here we present an exploration of the most desirable choices for regularization in the occupied subspace (controlled by a parameter α) and in the virtual subspace (controlled by a second parameter γ). Desirable parameter choices for BW-s2(α, γ) should exhibit (i) numerical stability, (ii) satisfactory accuracy for chemical energy differences, (iii) unconditional size-consistency, (iv) exactness for 2 electrons in 2 orbitals (2-in-2) at dissociation (which requires α + γ = 1), (v) flatness of the 2-in-2 PES approaching dissociation (which requires α + γ = 4), and (vi) particle-hole symmetry (α = γ). Tests on a range of bond-dissociation curves reveal stronger constraints on (α, γ) values that achieve unconditional size-consistency (α + γ > 1) than previously known. We combine this new information with minimizing errors against reference data across 14 data sets of chemical energy differences to recommend the nonempirical choice of BW-s2(α = 2, γ = 2). This choice appears to ensure unconditional size-consistency, as well as satisfactory accuracy on chemical problems, and enforces 2-in-2 flatness toward dissociation and particle-hole symmetry. Our analysis additionally reveals that BW-s2(α = 4, γ = 0), which was previously recommended on the basis of minimizing errors on chemical data, is also an alternative nonempirical choice that sacrifices 2-in-2 exactness and particle-hole symmetry.
- 24 April 2026
- Published by American Chemical Society (ACS) in Journal of Chemical Theory and Computation
- Vol. 22 (9) , 4439-4451
An Improved Size-Consistent Second-Order Brillouin–Wigner Perturbation Theory: Which Desirable Properties Are Compatible with Unconditional Size-Consistency and Optimized Chemical Accuracy?
Recently, it has been established that a size-consistent second-order Brillouin–Wigner perturbation (BW-s2) theory can be obtained by satisfying a trace condition with a repartitioning of the Hamiltonian between a one-body reference and a correlation-containing perturbation. BW-s2 has the twin advantages of being regular like the conventional second-order BW theory and being conditionally size-consistent like the conventional second-order Mo̷ller–Plesset theory. Conditional size-consistency means that spin-polarized reference orbitals must be used whenever the spin-restricted orbitals are unstable, whereas unconditional size-consistency means that size-consistency follows from the use of orbitals yielding the lowest BW-s2 energy. There is a wide range of possible forms for the partition, and here we present an exploration of the most desirable choices for regularization in the occupied subspace (controlled by a parameter α) and in the virtual subspace (controlled by a second parameter γ). Desirable parameter choices for BW-s2(α, γ) should exhibit (i) numerical stability, (ii) satisfactory accuracy for chemical energy differences, (iii) unconditional size-consistency, (iv) exactness for 2 electrons in 2 orbitals (2-in-2) at dissociation (which requires α + γ = 1), (v) flatness of the 2-in-2 PES approaching dissociation (which requires α + γ = 4), and (vi) particle-hole symmetry (α = γ). Tests on a range of bond-dissociation curves reveal stronger constraints on (α, γ) values that achieve unconditional size-consistency (α + γ > 1) than previously known. We combine this new information with minimizing errors against reference data across 14 data sets of chemical energy differences to recommend the nonempirical choice of BW-s2(α = 2, γ = 2). This choice appears to ensure unconditional size-consistency, as well as satisfactory accuracy on chemical problems, and enforces 2-in-2 flatness toward dissociation and particle-hole symmetry. Our analysis additionally reveals that BW-s2(α = 4, γ = 0), which was previously recommended on the basis of minimizing errors on chemical data, is also an alternative nonempirical choice that sacrifices 2-in-2 exactness and particle-hole symmetry.
- 13 January 2026
- Published by American Chemical Society (ACS) in The Journal of Physical Chemistry Letters
- Vol. 17 (4) , 909-917
Third-Order Perturbation Theory Made Regular: A Noniterative Correction to the Size-Consistent Second-Order Brillouin–Wigner Perturbation Theory
While second- and third-order Møller-Plesset perturbation theories are a useful description of weak electron correlations in molecules, they fail catastrophically in the presence of strong correlations. A promising recent alternative to MP2 is the size-consistent second-order Brillouin-Wigner perturbation theory, BWs2, which is both regular and size-consistent. Here we develop and assess a noniterative third-order correction to BWs2, called BWs2.5. This method has identical computational costs to MP3 and is both regular and size-consistent. For the model problem of two electrons in two orbitals, it removes the leading error of BWs2 if 50% of the third-order contribution is included. BWs2.5 is stable across a variety of tests of bond breaking using spin-restricted orbitals, with dissociation limits above the spin-polarized solution as needed for size consistency. Numerical tests on a variety of chemical energy differences establish that BWs2.5 is notably more accurate than BWs2.
- 13 January 2026
- Published by American Chemical Society (ACS) in The Journal of Physical Chemistry Letters
- Vol. 17 (4) , 909-917
Third-Order Perturbation Theory Made Regular: A Noniterative Correction to the Size-Consistent Second-Order Brillouin–Wigner Perturbation Theory
While second- and third-order Møller-Plesset perturbation theories are a useful description of weak electron correlations in molecules, they fail catastrophically in the presence of strong correlations. A promising recent alternative to MP2 is the size-consistent second-order Brillouin-Wigner perturbation theory, BWs2, which is both regular and size-consistent. Here we develop and assess a noniterative third-order correction to BWs2, called BWs2.5. This method has identical computational costs to MP3 and is both regular and size-consistent. For the model problem of two electrons in two orbitals, it removes the leading error of BWs2 if 50% of the third-order contribution is included. BWs2.5 is stable across a variety of tests of bond breaking using spin-restricted orbitals, with dissociation limits above the spin-polarized solution as needed for size consistency. Numerical tests on a variety of chemical energy differences establish that BWs2.5 is notably more accurate than BWs2.
- 27 October 2025
- Published by American Chemical Society (ACS) in Journal of Chemical Theory and Computation
- Vol. 21 (21) , 10910-10929
More Numerical Precision for Less Compute Cost: Optimizing a Local Correlation Algorithm for Second Order Møller–Plesset Theory and Comparing against Pair Natural Orbital Methods
Correlations associated with dynamical fluctuations of electrons in molecules can be spatially localized, which offers the promise of reducing the compute complexity of many-body quantum chemical methods such as second-order Møller-Plesset (MP2) theory. This work reports efforts to significantly improve a recent approach to local MP2 that is based on a single numerical threshold to control accuracy versus compute cost, and the use of localized orthogonal orbitals for both occupied and virtual spaces. The most important improvement is a novel embedding correction to the right-hand-side of the linear equations for the retained MP2 amplitudes, which includes the effect of integrals that are evaluated but discarded as below threshold. Together with a modified set of occupied orbitals that increases diagonal dominance in the occupied-occupied Fock operator, and an on-the-fly block Kapuy solver for below-threshold amplitudes, this scheme provides roughly an order of magnitude improvement in accuracy. Other algorithm optimizations include nonrobust local fitting, sharper occupied/virtual sparse maps, and on-the-fly selection of locally BLAS-2 and BLAS-3 evaluation of matrix-vector products in the conjugate gradient iterations. These advances have been shown to significantly reduce memory and improve compute efficiency, without affecting accuracy control by ϵ. Detailed comparisons against the domain-localized pair natural orbital, DLPNO-MP2, algorithm in the ORCA 6.0.1 package demonstrate significant improvements in accuracy for given time-to-solution. Reported benchmarks include ACONF20 conformational energies, L7, S12L, and ExL8 noncovalent interactions, C60 isomerization energies, and selected MME55 transition metal complex energetics.
- 27 October 2025
- Published by American Chemical Society (ACS) in Journal of Chemical Theory and Computation
- Vol. 21 (21) , 10910-10929
More Numerical Precision for Less Compute Cost: Optimizing a Local Correlation Algorithm for Second Order Møller–Plesset Theory and Comparing against Pair Natural Orbital Methods
Correlations associated with dynamical fluctuations of electrons in molecules can be spatially localized, which offers the promise of reducing the compute complexity of many-body quantum chemical methods such as second-order Møller-Plesset (MP2) theory. This work reports efforts to significantly improve a recent approach to local MP2 that is based on a single numerical threshold to control accuracy versus compute cost, and the use of localized orthogonal orbitals for both occupied and virtual spaces. The most important improvement is a novel embedding correction to the right-hand-side of the linear equations for the retained MP2 amplitudes, which includes the effect of integrals that are evaluated but discarded as below threshold. Together with a modified set of occupied orbitals that increases diagonal dominance in the occupied-occupied Fock operator, and an on-the-fly block Kapuy solver for below-threshold amplitudes, this scheme provides roughly an order of magnitude improvement in accuracy. Other algorithm optimizations include nonrobust local fitting, sharper occupied/virtual sparse maps, and on-the-fly selection of locally BLAS-2 and BLAS-3 evaluation of matrix-vector products in the conjugate gradient iterations. These advances have been shown to significantly reduce memory and improve compute efficiency, without affecting accuracy control by ϵ. Detailed comparisons against the domain-localized pair natural orbital, DLPNO-MP2, algorithm in the ORCA 6.0.1 package demonstrate significant improvements in accuracy for given time-to-solution. Reported benchmarks include ACONF20 conformational energies, L7, S12L, and ExL8 noncovalent interactions, C60 isomerization energies, and selected MME55 transition metal complex energetics.
- Open Access
- 24 August 2025
- Published by American Chemical Society (ACS) in ChemRxiv
More numerical precision for less compute cost: Optimizing a local correlation algorithm for second order Møller-Plesset theory and comparing against pair natural orbital methods
Correlations associated with dynamical fluctuations of electrons in molecules can be spatially localized, which offers the promise of reducing the compute complexity of many-body quantum chemical methods such as second-order Møller-Plesset (MP2) theory. This work reports efforts to significantly improve a recent approach to local MP2 that is based on a single numerical threshold to control accuracy versus compute cost, and the use of localized orthogonal orbitals for both occupied and virtual spaces. The most important improvement is a novel embedding correction to the right-hand-side of the linear equations for the retained MP2 amplitudes, which includes the effect of integrals that are evaluated but discarded as below threshold. Together with a modified set of occupied orbitals that increases diagonal dominance in the occupied-occupied Fock operator, and an on-the-fly block Kapuy solver for below-threshold amplitudes, this scheme provides roughly an order of magnitude improvement in accuracy. Other algorithm optimizations include non-robust local fitting, sharper occupied/virtual sparse maps, and on-the-fly selection of locally BLAS-2 and BLAS-3 evaluation of matrix-vector products in the conjugate gradient iterations. These advances have been shown to significantly reduce memory and improve compute efficiency, without affecting accuracy control by ε. Detailed comparisons against the domain-localized pair natural orbital, DLPNO-MP2, algorithm in the ORCA 6.0.1 package demonstrate significant improvements in accuracy for given time-to-solution. Reported benchmarks include ACONF20 conformational energies, L7, S12L, and ExL8 non-covalent interactions, C60 isomerization energies, and selected MME55 transition metal complex energetics.
- 24 August 2025
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessMore numerical precision for less compute cost: Optimizing a local correlation algorithm for second order Møller-Plesset theory and comparing against pair natural orbital methods
Correlations associated with dynamical fluctuations of electrons in molecules can be spatially localized, which offers the promise of reducing the compute complexity of many-body quantum chemical methods such as second-order Møller-Plesset (MP2) theory. This work reports efforts to significantly improve a recent approach to local MP2 that is based on a single numerical threshold to control accuracy versus compute cost, and the use of localized orthogonal orbitals for both occupied and virtual spaces. The most important improvement is a novel embedding correction to the right-hand-side of the linear equations for the retained MP2 amplitudes, which includes the effect of integrals that are evaluated but discarded as below threshold. Together with a modified set of occupied orbitals that increases diagonal dominance in the occupied-occupied Fock operator, and an on-the-fly block Kapuy solver for below-threshold amplitudes, this scheme provides roughly an order of magnitude improvement in accuracy. Other algorithm optimizations include non-robust local fitting, sharper occupied/virtual sparse maps, and on-the-fly selection of locally BLAS-2 and BLAS-3 evaluation of matrix-vector products in the conjugate gradient iterations. These advances have been shown to significantly reduce memory and improve compute efficiency, without affecting accuracy control by ε. Detailed comparisons against the domain-localized pair natural orbital, DLPNO-MP2, algorithm in the ORCA 6.0.1 package demonstrate significant improvements in accuracy for given time-to-solution. Reported benchmarks include ACONF20 conformational energies, L7, S12L, and ExL8 non-covalent interactions, C60 isomerization energies, and selected MME55 transition metal complex energetics.
- 7 July 2025
- Published by American Chemical Society (ACS) in Journal of Chemical Theory and Computation
- Vol. 21 (14) , 6887-6904
Regularized Second-Order Møller–Plesset Theory: Linear Scaling Implementation and Assessment on Large-Molecule Problems
Formally fifth-order scaling with molecular size, second-order Møller-Plesset (MP2) theory is widely used in its own right, and as a component of double hybrid density functional theory. Nevertheless, MP2 is well-known to have limited accuracy for large intermolecular interactions, and the MP2 energy is not bounded from below. Regularized MP2 methods aim to address these issues without changing the formal fifth-order scaling and are therefore of interest as more accurate and robust alternatives. In this work, we report a linear scaling algorithm with shared and distributed memory parallel implementations, for the evaluation of two regularized MP2 methods: κ-MP2 and the size-consistent second-order Brillouin-Wigner perturbation theory (BW-s2). The algorithm uses a single numerical threshold to control accuracy, which is demonstrated to yield a roughly 10-fold reduction in numerical error for each factor of 10 reduction in the threshold. The scaling of time-to-solution is assessed as a function of molecule size (up to 1800 atoms and 25000 atomic orbitals), the choice of threshold, and the number of processors (up to 2000 cores), for a range of systems including linear alkanes and three-dimensional complexes. The results show an earlier crossover against the canonical implementation for BW-s2 than for either MP2 or κ-MP2. The new implementation is used to evaluate the accuracy of the κ-MP2 and BW-s2 methods for some of the largest test sets of intermolecular interactions for which coupled cluster benchmark data are available, including IONPI19, L7, S12L, and C60ISO. The results show that κ-MP2 and BW-s2 perform far better than MP2. Additional tests of κ-MP2 and BW-s2 are reported for the case of short-range amino acid interactions, where MP2 performs exceptionally well.
- 7 July 2025
- Published by American Chemical Society (ACS) in Journal of Chemical Theory and Computation
- Vol. 21 (14) , 6887-6904
Regularized Second-Order Møller–Plesset Theory: Linear Scaling Implementation and Assessment on Large-Molecule Problems
Formally fifth-order scaling with molecular size, second-order Møller-Plesset (MP2) theory is widely used in its own right, and as a component of double hybrid density functional theory. Nevertheless, MP2 is well-known to have limited accuracy for large intermolecular interactions, and the MP2 energy is not bounded from below. Regularized MP2 methods aim to address these issues without changing the formal fifth-order scaling and are therefore of interest as more accurate and robust alternatives. In this work, we report a linear scaling algorithm with shared and distributed memory parallel implementations, for the evaluation of two regularized MP2 methods: κ-MP2 and the size-consistent second-order Brillouin-Wigner perturbation theory (BW-s2). The algorithm uses a single numerical threshold to control accuracy, which is demonstrated to yield a roughly 10-fold reduction in numerical error for each factor of 10 reduction in the threshold. The scaling of time-to-solution is assessed as a function of molecule size (up to 1800 atoms and 25000 atomic orbitals), the choice of threshold, and the number of processors (up to 2000 cores), for a range of systems including linear alkanes and three-dimensional complexes. The results show an earlier crossover against the canonical implementation for BW-s2 than for either MP2 or κ-MP2. The new implementation is used to evaluate the accuracy of the κ-MP2 and BW-s2 methods for some of the largest test sets of intermolecular interactions for which coupled cluster benchmark data are available, including IONPI19, L7, S12L, and C60ISO. The results show that κ-MP2 and BW-s2 perform far better than MP2. Additional tests of κ-MP2 and BW-s2 are reported for the case of short-range amino acid interactions, where MP2 performs exceptionally well.