Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

The Journal of Chemical Physics

Spin contamination in quantum Monte Carlo wave functions

Chien-Jung Huang, Claudia Filippi, Cyrus J. Umrigar

Published Jun 1, 1998• 58 citations
Publisher

Abstract

The wave function usually employed in quantum Monte Carlo (QMC) electronic structure calculations is the product of a Jastrow factor and a sum of products of up-spin and down-spin determinants. Typically, a different Jastrow factor is used for parallel- and antiparallel-spin electrons in order to satisfy the cusp conditions and thereby ensure that the local energy at electron–electron coincidence points is finite. However, when the Jastrow factor is not completely symmetric under interchange of the spatial coordinates of the electrons, the resulting wave function may not be an eigenstate of the spin operator Ŝ2. For the Li and Be atoms, we evaluate the spin contamination in a variety of wave functions with progressively higher body-order correlations in the Jastrow factor. The spin contamination is found to be small for all the wave functions (10−3–10−5), the smaller values being obtained for the more accurate wave functions. On the other hand, if we eliminate the spin contamination by employing a symmetric Jastrow (and sacrificing the parallel-spin cusp condition), the resulting wave functions typically have about 20%–40% larger root mean square fluctuations in the local energy. A wave function which both satisfies the cusp conditions and has definite spin can be constructed, but for systems with many electrons, its computational cost is higher than for the commonly used QMC wave functions.

DOI
10.1063/1.476330

Related papers

Spin contamination in quantum Monte Carlo wave functions

Chang Jiang Huang, Claudia Filippi, Cyrus J. Umrigar

Abstract The Pauli exclusion principle requires that a complete spacespin wavefunction be antisymmetric to the exchange of electrons. This requirement can be met by single-determinant and multi determinant wavefunctions obtained in SCF and CI calculations, which may then be considered acceptable as trial functions for specifying nodal surfaces and for importance sampling in determining energies. If multiplied by a Jastrow correlation function symmetrical to exchange of electrons, the function remains acceptable. However, in QMC, the Jastrow functions most often used have different coefficients for pairs of electrons of the same and opposite spins. These are not symmetric to all exchanges, and in combination with the determinantal function they give a trial function which is not antisymmetric to exchange of electrons and is not an eigenfunction of the spin operator 82 The problem is discussed in detail in this paper. The spin contamination oS2 of a wavefunction may be readily determined by Metropolis integrations as in variational QMC. The calculations reported for Li and Be show the spin contamination to be very small for unsymmetrical Jastrow terms with the usual coefficients of ¼ and ½ for parallel and antiparallel electron pairs. For higher order terms in the Jastrow function, the contamination could be reduced further. \Vith use of symmetric Jastrow factors, the electron-electron cusp conditions are not properly satisfied, so that variance in local energies is higher and computed energies are less accurate, but other properties such as electron densities may be more accurate. An alternative is the use of more complex and computationally expensive functions satisfying both spin and cusp conditions.

Oxford University Press eBooks•Jun 18, 2007•1 citation

Auxiliary-field-based trial wave functions in quantum Monte Carlo calculations

Chia-Chen Chang, Brenda M. Rubenstein, Miguel A. Morales

Quantum Monte Carlo (QMC) algorithms have long relied on Jastrow factors to incorporate dynamic correlation into trial wave functions. While Jastrow-type wave functions have been widely employed in real-space algorithms, they have seen limited use in second-quantized QMC methods, particularly in projection methods that involve a stochastic evolution of the wave function in imaginary time. Here we propose a scheme for generating Jastrow-type correlated trial wave functions for auxiliary-field QMC methods. The method is based on decoupling the two-body Jastrow into one-body projectors coupled to auxiliary fields, which then operate on a single determinant to produce a multideterminant trial wave function. We demonstrate that intelligent sampling of the most significant determinants in this expansion can produce compact trial wave functions that reduce errors in the calculated energies. Our technique may be readily generalized to accommodate a wide range of two-body Jastrow factors and applied to a variety of model and chemical systems.

arXiv:1604.00345•Physical review. B./Physical review. B•Dec 19, 2016•22 citations

Quantum Monte Carlo study of the three-dimensional spin-polarized homogeneous electron gas

G. G. Spink, Richard J. Needs, N. D. Drummond

We have studied the spin-polarized three-dimensional homogeneous electron gas using the diffusion quantum Monte Carlo method, with trial wave functions including backflow and three-body correlations in the Jastrow factor, and we have used twist averaging to reduce finite-size effects. Calculations of the pair-correlation function, including the on-top pair density, as well as the structure factor and the total energy, are reported for systems of 118 electrons in the density range ${r}_{\mathrm{s}}=0.5$--20 a.u., and for spin polarizations of 0, 0.34, 0.66, and 1. We consider the spin resolution of the pair-correlation function and structure factor, and the energy of spin polarization. We show that a control variate method can reduce the variance when twist averaging, and we have achieved higher accuracy and lower noise than earlier quantum Monte Carlo studies.

arXiv:1307.5794•Physical Review B•Aug 22, 2013•99 citations

Capturing spin fluctuations in CaCuO 2 : Ab initio quantum Monte Carlo calculations with multideterminant wave functions

Chun Yu Chow, William A. Wheeler, Lucas K. Wagner

We present an advanced ab initio quantum Monte Carlo (QMC) calculation of the ground state of undoped ${\mathrm{CaCuO}}_{2}$. We extend the traditional single-determinant Slater-Jastrow approach to include multideterminant wave functions, inhomogeneous Jastrow factors, and orbital optimization. Our results demonstrate not only an improvement in the variational bound of the ground state energy---2.3 eV per formula unit lower than previous state-of-the-art techniques---but also confirm the presence of spin fluctuations in multideterminant wave functions in a strongly correlated cuprate system, which is integral to understanding high-${T}_{c}$ superconductivity. This is a demonstration of capturing spin fluctuations in QMC wave functions on a cuprate, establishing the groundwork for studies on doped cuprates in the superconducting state, where spin fluctuations require more accurate characterization.

arXiv:2409.00225•Physical review. B./Physical review. B•Dec 18, 2024•1 citation

Quantum Monte Carlo studies of density functional theory

Maziar Nekovee, W. M. C. Foulkes, Richard J. Needs

Abstract It is difficult to imagine that density functional theory (DFT) could be useful in improving quantum Monte Carlo methods, but it is easy to imagine that QMC could be useful in improving density functional methods. Since QMC can produce accurate wavefunctions and since wavefunctions lead to all other properties one should be able, in principle, to use QMC to predict the exchangecorrelation energy functional Exc[n(r)], the key (unknown) quantity required in DFT. In practice, that is not so easy, but this paper shows how QMC can take steps in that direction by providing insights gained from comparisons of results from QMC and DFT. The authors considered the case of the strongly inhomogeneous electron gas using variational Q:tvIC to obtain wavefunctions 1/J>. and energies E[7/J>.] for the ground state of the Hamiltonian iI>- = T +AVee+V >. associated with the coupling constant A in the range 0 to 1. Optimization of a Slater-Jastrow many-body wavefunction was made by minimizing a penalty function given by a weighted sum of the variance in local energies and a sum of squares of electron density differences. The variational parameters in 7/J>. and v>. were repeatedly varied to reach convergence followed by additional calculations to obtain the local values of the exchange-correlation energies e c and the exchange-correlation holes n c from which Exe and nxc could be determined.

Oxford University Press eBooks•Jun 18, 2007•0 citations

Studies of Quantum Dots using Machine Learning

Even Marius Nordhagen

With the ability to solve the many-body Schrödinger equation accurately, in principle all physics and chemistry could be derived from first principles. However, exact wave functions of realistic and interesting systems are in general unavailable because they are non-deterministic polynomial-hard to compute (Troyer, 2005). This implies that we need to rely on approximations. The variational Monte Carlo (VMC) method is widely used for ground state studies, but requires a trial wave function ansatz which must trade off between efficiency and accuracy. The method also has many common features with machine learning algorithms, and as neural networks have shown impressive power as function approximators, the idea is to use a neural network as the trial wave function guess. For fermionic systems, like electronic structure systems, the wave function needs to obey Fermi-Dirac statistics. This is typically achieved using a Slater determinant. As a neural network hardly can model this feature, our approach is to replace the single-particle functions in the Slater determinant with restricted Boltzmann machines, getting the RBM ansatz. In addition, we add further correlations via so-called Jastrow factors (Drummond, 2004). The de facto standard trial wave function ansatz for electronic structure calculations is the Slater-Jastrow ansatz, which was implemented as a reference. Our primary focus is on closed-shell circular quantum dots, where we compute the ground state energy and electron density of two-dimensional systems with up to 90 electrons and three-dimensional systems with up to 70 electrons. The energy obtained by the RBM ansatz was reasonably close to experimental results, and it gradually became closer as we added more complex correlation factors. For our most complicated Jastrow factor, the energy was found to be lower than the energy provided by the Slater-Jastrow ansatz for small dots, but for larger dots it was slightly higher. However, the one-body density profile reveals that the RBM ansatz gives more distinctly located electrons compared to the Slater-Jastrow ansatz. This can be explained by the way the RBM ansatz models the correlations. From the two-body density profile, we also observe that the repulsive interactions get more significant as we add a Jastrow factor. Based on the electron densities and the energy distribution between kinetic and potential energy, it is certain that the various methods provide different electron configurations. For low-frequency dots, the electron density becomes more localized with an additional radial peak compared to high-frequency dots. This is reminiscent of what is known as Wigner localization (Ghosal, 2007). The computational time consumption was found to be favorable for the RBM ansatz for small systems and the Slater-Jastrow ansatz for large systems. This can be explained by the exploding number of variational parameters in the RBM ansatz as the system sizes increase. The RBM ansatz with Jastrow factors were notably more computationally intensive than the other ansätze, and evidently, there is no reason to use the simplest Jastrow factor when more complicated Jastrow factors are available.

•Jan 1, 2019•0 citations
The wave function usually employed in quantum Monte Carlo (QMC) electronic structure calculations is the product of a Jastrow factor and a sum of products of up-spin and down-spin determinants. Typically, a different Jastrow factor is used for parallel- and antiparallel-spin electrons in order to satisfy the cusp conditions and thereby ensure that the local energy at electron–electron coincidence points is finite. However, when the Jastrow factor is not completely symmetric under interchange of the spatial coordinates of the electrons, the resulting wave function may not be an eigenstate of the spin operator Ŝ2. For the Li and Be atoms, we evaluate the spin contamination in a variety of wave functions with progressively higher body-order correlations in the Jastrow factor. The spin contamination is found to be small for all the wave functions (10−3–10−5), the smaller values being obtained for the more accurate wave functions. On the other hand, if we eliminate the spin contamination by employing a symmetric Jastrow (and sacrificing the parallel-spin cusp condition), the resulting wave functions typically have about 20%–40% larger root mean square fluctuations in the local energy. A wave function which both satisfies the cusp conditions and has definite spin can be constructed, but for systems with many electrons, its computational cost is higher than for the commonly used QMC wave functions.
Abstract The Pauli exclusion principle requires that a complete spacespin wavefunction be antisymmetric to the exchange of electrons. This requirement can be met by single-determinant and multi determinant wavefunctions obtained in SCF and CI calculations, which may then be considered acceptable as trial functions for specifying nodal surfaces and for importance sampling in determining energies. If multiplied by a Jastrow correlation function symmetrical to exchange of electrons, the function remains acceptable. However, in QMC, the Jastrow functions most often used have different coefficients for pairs of electrons of the same and opposite spins. These are not symmetric to all exchanges, and in combination with the determinantal function they give a trial function which is not antisymmetric to exchange of electrons and is not an eigenfunction of the spin operator 82 The problem is discussed in detail in this paper. The spin contamination oS2 of a wavefunction may be readily determined by Metropolis integrations as in variational QMC. The calculations reported for Li and Be show the spin contamination to be very small for unsymmetrical Jastrow terms with the usual coefficients of ¼ and ½ for parallel and antiparallel electron pairs. For higher order terms in the Jastrow function, the contamination could be reduced further. \Vith use of symmetric Jastrow factors, the electron-electron cusp conditions are not properly satisfied, so that variance in local energies is higher and computed energies are less accurate, but other properties such as electron densities may be more accurate. An alternative is the use of more complex and computationally expensive functions satisfying both spin and cusp conditions.
Quantum Monte Carlo (QMC) algorithms have long relied on Jastrow factors to incorporate dynamic correlation into trial wave functions. While Jastrow-type wave functions have been widely employed in real-space algorithms, they have seen limited use in second-quantized QMC methods, particularly in projection methods that involve a stochastic evolution of the wave function in imaginary time. Here we propose a scheme for generating Jastrow-type correlated trial wave functions for auxiliary-field QMC methods. The method is based on decoupling the two-body Jastrow into one-body projectors coupled to auxiliary fields, which then operate on a single determinant to produce a multideterminant trial wave function. We demonstrate that intelligent sampling of the most significant determinants in this expansion can produce compact trial wave functions that reduce errors in the calculated energies. Our technique may be readily generalized to accommodate a wide range of two-body Jastrow factors and applied to a variety of model and chemical systems.
We have studied the spin-polarized three-dimensional homogeneous electron gas using the diffusion quantum Monte Carlo method, with trial wave functions including backflow and three-body correlations in the Jastrow factor, and we have used twist averaging to reduce finite-size effects. Calculations of the pair-correlation function, including the on-top pair density, as well as the structure factor and the total energy, are reported for systems of 118 electrons in the density range rs=0.5{r}_{\mathrm{s}}=0.5rs​=0.5--20 a.u., and for spin polarizations of 0, 0.34, 0.66, and 1. We consider the spin resolution of the pair-correlation function and structure factor, and the energy of spin polarization. We show that a control variate method can reduce the variance when twist averaging, and we have achieved higher accuracy and lower noise than earlier quantum Monte Carlo studies.
We present an advanced ab initio quantum Monte Carlo (QMC) calculation of the ground state of undoped CaCuO2{\mathrm{CaCuO}}_{2}CaCuO2​. We extend the traditional single-determinant Slater-Jastrow approach to include multideterminant wave functions, inhomogeneous Jastrow factors, and orbital optimization. Our results demonstrate not only an improvement in the variational bound of the ground state energy---2.3 eV per formula unit lower than previous state-of-the-art techniques---but also confirm the presence of spin fluctuations in multideterminant wave functions in a strongly correlated cuprate system, which is integral to understanding high-Tc{T}_{c}Tc​ superconductivity. This is a demonstration of capturing spin fluctuations in QMC wave functions on a cuprate, establishing the groundwork for studies on doped cuprates in the superconducting state, where spin fluctuations require more accurate characterization.
Abstract It is difficult to imagine that density functional theory (DFT) could be useful in improving quantum Monte Carlo methods, but it is easy to imagine that QMC could be useful in improving density functional methods. Since QMC can produce accurate wavefunctions and since wavefunctions lead to all other properties one should be able, in principle, to use QMC to predict the exchangecorrelation energy functional Exc[n(r)], the key (unknown) quantity required in DFT. In practice, that is not so easy, but this paper shows how QMC can take steps in that direction by providing insights gained from comparisons of results from QMC and DFT. The authors considered the case of the strongly inhomogeneous electron gas using variational Q:tvIC to obtain wavefunctions 1/J>. and energies E[7/J>.] for the ground state of the Hamiltonian iI>- = T +AVee+V >. associated with the coupling constant A in the range 0 to 1. Optimization of a Slater-Jastrow many-body wavefunction was made by minimizing a penalty function given by a weighted sum of the variance in local energies and a sum of squares of electron density differences. The variational parameters in 7/J>. and v>. were repeatedly varied to reach convergence followed by additional calculations to obtain the local values of the exchange-correlation energies e c and the exchange-correlation holes n c from which Exe and nxc could be determined.
With the ability to solve the many-body Schrödinger equation accurately, in principle all physics and chemistry could be derived from first principles. However, exact wave functions of realistic and interesting systems are in general unavailable because they are non-deterministic polynomial-hard to compute (Troyer, 2005). This implies that we need to rely on approximations. The variational Monte Carlo (VMC) method is widely used for ground state studies, but requires a trial wave function ansatz which must trade off between efficiency and accuracy. The method also has many common features with machine learning algorithms, and as neural networks have shown impressive power as function approximators, the idea is to use a neural network as the trial wave function guess. For fermionic systems, like electronic structure systems, the wave function needs to obey Fermi-Dirac statistics. This is typically achieved using a Slater determinant. As a neural network hardly can model this feature, our approach is to replace the single-particle functions in the Slater determinant with restricted Boltzmann machines, getting the RBM ansatz. In addition, we add further correlations via so-called Jastrow factors (Drummond, 2004). The de facto standard trial wave function ansatz for electronic structure calculations is the Slater-Jastrow ansatz, which was implemented as a reference. Our primary focus is on closed-shell circular quantum dots, where we compute the ground state energy and electron density of two-dimensional systems with up to 90 electrons and three-dimensional systems with up to 70 electrons. The energy obtained by the RBM ansatz was reasonably close to experimental results, and it gradually became closer as we added more complex correlation factors. For our most complicated Jastrow factor, the energy was found to be lower than the energy provided by the Slater-Jastrow ansatz for small dots, but for larger dots it was slightly higher. However, the one-body density profile reveals that the RBM ansatz gives more distinctly located electrons compared to the Slater-Jastrow ansatz. This can be explained by the way the RBM ansatz models the correlations. From the two-body density profile, we also observe that the repulsive interactions get more significant as we add a Jastrow factor. Based on the electron densities and the energy distribution between kinetic and potential energy, it is certain that the various methods provide different electron configurations. For low-frequency dots, the electron density becomes more localized with an additional radial peak compared to high-frequency dots. This is reminiscent of what is known as Wigner localization (Ghosal, 2007). The computational time consumption was found to be favorable for the RBM ansatz for small systems and the Slater-Jastrow ansatz for large systems. This can be explained by the exploding number of variational parameters in the RBM ansatz as the system sizes increase. The RBM ansatz with Jastrow factors were notably more computationally intensive than the other ansätze, and evidently, there is no reason to use the simplest Jastrow factor when more complicated Jastrow factors are available.