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

arXiv:2207.06297 · Physical review. B./Physical review. B

Instanton gas approach to the Hubbard model

Maksim Ulybyshev, Christopher Winterowd, Fakher F. Assaad, Savvas Zafeiropoulos

Published Jan 30, 2023• 3 citations
arXivHTMLPDF

Abstract

In this paper, we consider a path integral formulation of the Hubbard model based on a Hubbard-Stratonovich transformation that couples the auxiliary field to the local electronic density. This decoupling is known to have a saddle-point structure that shows a remarkable regularity: The field configuration at each saddle point can be understood in terms of a set of elementary field configurations localized in space and imaginary time which we coin instantons. The interaction between instantons is short ranged. Here, we formulate a classical partition function for the instanton gas that has predictive power. For a given set of physical parameters, we can predict the distribution of instantons and show that the instanton number is sharply defined in the thermodynamic limit, thereby defining a unique dominant saddle point. Decoupling in the charge channel conserves SU(2) spin symmetry for each field configurations. Hence, the instanton approach provides an SU(2) spin-symmetric approximation to the Hubbard model. It fails, however, to capture the magnetic transition inherent to the Hubbard model on the honeycomb lattice despite being able to describe local moment formation. In fact, the instanton itself corresponds to local moment formation and concomitant short-ranged antiferromagnetic correlations. This aspect is also seen in the single particle spectral function that shows clear signs of the upper and lower Hubbard bands. Our instanton approach bears remarkable similarities to local dynamical approaches, such as dynamical mean-field theory, in the sense that it has the unique property of allowing for local moment formation without breaking the SU(2) spin symmetry. In contrast to local approaches, it captures short-ranged magnetic fluctuations. Furthermore, it also offers possibilities for systematic improvements by taking into account fluctuations around the dominant saddle point. Finally, we show that the saddle point structure depends upon the choice of lattice geometry. For the square lattice at half filling, the saddle-point structure reflects the itinerant to localized nature of the magnetism as a function of the coupling strength. The implications of our results for Lefschetz thimble approaches to alleviate the sign problem are also discussed.

DOI
10.1103/physrevb.107.045143

Related papers

Beyond the instanton gas approach: dominant thimbles approximation for the Hubbard model

Maksim Ulybyshev, Fakher F. Assaad

To each complex saddle point of an action, one can attach a Lefschetz thimble on which the imaginary part of the action is constant. Cauchy theorem states that summation over a set of thimbles produces the exact result. This reorganization of the path integral, is an appealing starting point for various approximations: In the realm of auxiliary quantum Monte Carlo methods it provides a framework to alleviate the negative sign problem. Here, we suggest to constrain the integration to the \textit{dominant} thimbles: the thimbles attached to the saddle points with the largest statistical weight. For the Hubbard model, in a formulation where the the Hubbard Stratonovitch field couples to the charge, this provides a \textit{symmetry} consistent approximation to the physics of the Hubbard model: constraining the integration domain does not explicitly break a symmetry. We can test this approach for the Hubbard model at half-filling on a bipartite lattice. The paper builds on the previously developed instanton gas approach, where an exhaustive saddle point approximation was constructed. We present results, showing that the dominant thimbles approximation provides results that are in remarkable agreement with the exact results for various fermionic observables including spin and charge order parameters and single electron spectral functions. We discuss implications of our results for simulations away of half filling.

arXiv:2407.09452•Jul 12, 2024•0 citations

Hubbard model on the honeycomb lattice: From static and dynamical mean-field theories to lattice quantum Monte Carlo simulations

Marcin Raczkowski, Robert Peters, Phùng Thị Thu +4 more

We study the one-band Hubbard model on the honeycomb lattice using a combination of quantum Monte Carlo (QMC) simulations and static as well as dynamical mean-field theory (DMFT). This model is known to show a quantum phase transition between a Dirac semimetal and the antiferromagnetic insulator. The aim of this paper is to provide a detailed comparison between these approaches by computing static properties, notably ground-state energy, single-particle gap, double occupancy, and staggered magnetization, as well as dynamical quantities such as the single-particle spectral function. At the static mean-field level, local moments cannot be generated without breaking the SU(2) spin symmetry. The DMFT approximation accounts for temporal fluctuations and thus captures both the evolution of the double occupancy and the resulting local moment formation in the paramagnetic phase. As a consequence, the DMFT approximation is found to be very accurate in the Dirac semimetallic phase where local moment formation is present and the spin correlation length small. However, in the vicinity of the fermion quantum critical point, the spin correlation length diverges and the spontaneous SU(2) symmetry breaking leads to low-lying Goldstone modes in the magnetically ordered phase. The impact of these spin fluctuations on the single-particle spectral function---waterfall features and narrow spin-polaron bands---is only visible in the lattice QMC approach.

arXiv:1908.04307•Physical review. B./Physical review. B•Mar 5, 2020•57 citations

Moment formation and pairing in the 2D Hubbard model

Zizhe Cai

The two-dimensional (2D) Hubbard model is solved in a new way. The zero-frequency components of the Hubbard interaction are separated out and solved to infinite order to avoid the traditional difficulties associated with perturbation theory in the strong-coupling limit. Specifically the zero-frequency spin interaction is decoupled by the introduction of Stratonovich-Hubbard fields, while the zero-frequency charge interaction is treated by "slave" Hartree-Fock. This approach leads to the formation of large fluctuating spin moments, and as a consequence the one-electron spectral density exhibits a pseudogap. A Metropolis Monte Carlo method is used to perform the configuration averaging over Stratonovich-Hubbard fields. Within single Monte Carlo configurations the non-zero frequency components of the Hubbard interaction are treated by a high-order perturbation method, which incorporates the exchange of dynamical particle-hole fluctuations. Since a single configuration has neither translational symmetry nor spin rotational symmetry, these fluctuations do not have a well-defined crystal momentum or total spin. When the dynamical particle-hole fluctuations are included in the calculation, localized electron pairing results at elevated temperatures. Superconductivity presumably onsets when these localized pairs develop coherence, but this is not studied in the present work. The nature of the pairing interaction and the resulting pair eigenfunctions is studied.

•Jan 1, 2015•0 citations

Mean-field approach to antiferromagnetic domains in the doped Hubbard model

Edwin Langmann A, Mats Wallin

We present a restricted path integral approach to the two-dimensional and three-dimensional repulsive Hubbard model. In this approach the partition function is approximated by restricting the summation over all states to a (small) subclass that is chosen so as to well represent the important states. This procedure generalizes mean-field theory and can be systematically improved by including more states or fluctuations. We analyze in detail the simplest of these approximations, which corresponds to summing over states with local antiferromagnetic (AF) order. If in the states considered the AF order changes sufficiently little in space and time, the path integral becomes a finite-dimensional integral for which the saddle point evaluation is exact. This leads to generalized mean-field equations allowing for the possibility of more than one relevant saddle point. In a big parameter regime (both in temperature and filling), we find that this integral has two relevant saddle points, one corresponding to finite AF order and the other without. These degenerate saddle points describe a phase of AF ordered fermions coexisting with free, metallic fermions. We argue that this mixed phase is a simple mean-field description of a variety of possible inhomogeneous states, appropriate on length scales where these states appear homogeneous. We sketch systematic refinements of this approximation, which can give more detailed descriptions of the system.

arXiv:cond-mat/9610030•Physical review. B, Condensed matter•Apr 15, 1997•11 citations

Magnetism in the two-dimensional Hubbard model (abstract)

J. E. Hirsch, Hsueh‐Ju Lin

The Hubbard model has been widely used to describe magnetism due to itinerant electrons. However, except for the one-dimensional case, no exact solution of the model exists. We have studied the two-dimensional Hubbard model on a square lattice using a Monte Carlo simulation technique. Simulations were performed on lattices of up to 8×8 in spatial size. We will discuss results for thermodynamic properties and correlation functions as a function of temperature and interaction strength. For the half-filled band case, our results indicate that the system exhibits long-range antiferromagnetic order in the ground state for any nonzero interaction, although substantially reduced from the classical Néel state due to both charge and spin fluctuations. For the non-half-filled band case, our results suggest that no long-range magnetic order exists. We also don’t find any indication of short-ranged ferromagnetic correlations in the parameter range studied. We have also studied the predictions of various approximate treatments for this model, particularly the static approximation and the coherent potential approximation (CPA). By comparison with the Monte Carlo results, we discuss the validity of these schemes. Finally, we discuss the effect of changing the band structure on the magnetic properties of the system.

Journal of Applied Physics•Apr 15, 1985•0 citations

A STRONG-COUPLING EXPANSION FOR THE HUBBARD MODEL

N. Dupuis, Stéphane Pairault

We reconsider the strong-coupling expansion for the Hubbard model recently introduced by Sarker and Pairault et al. By introducing slave particles that act as projection operators onto the empty, singly occupied and doubly occupied atomic states, the perturbation theory around the atomic limit distinguishes between processes that do conserve or do not conserve the total number of doubly occupied sites. This allows for a systematic t/U expansion that does not break down at low temperature (t being the intersite hopping amplitude and U the local Coulomb repulsion). The fermionic field becomes a two-component field, which reflects the presence of the two Hubbard bands. The single-particle propagator is naturally expressed as a function of a 2×2 matrix self-energy. Furthermore, by introducing a time- and space-fluctuating spin-quantization axis in the functional integral, we can expand around a "non-degenerate" ground-state where each singly occupied site has a well defined spin direction (which may fluctuate in time). This formalism is used to derive the effective action of charge carriers in the lower Hubbard band to first order in t/U. We recover the action of the t–J model in the spin-hole coherent-state path integral. We also compare our results with those previously obtained by studying fluctuations around the large-U Hartree–Fock saddle point.

arXiv:cond-mat/0008447•International Journal of Modern Physics B•Sep 30, 2000•6 citations
In this paper, we consider a path integral formulation of the Hubbard model based on a Hubbard-Stratonovich transformation that couples the auxiliary field to the local electronic density. This decoupling is known to have a saddle-point structure that shows a remarkable regularity: The field configuration at each saddle point can be understood in terms of a set of elementary field configurations localized in space and imaginary time which we coin instantons. The interaction between instantons is short ranged. Here, we formulate a classical partition function for the instanton gas that has predictive power. For a given set of physical parameters, we can predict the distribution of instantons and show that the instanton number is sharply defined in the thermodynamic limit, thereby defining a unique dominant saddle point. Decoupling in the charge channel conserves SU(2) spin symmetry for each field configurations. Hence, the instanton approach provides an SU(2) spin-symmetric approximation to the Hubbard model. It fails, however, to capture the magnetic transition inherent to the Hubbard model on the honeycomb lattice despite being able to describe local moment formation. In fact, the instanton itself corresponds to local moment formation and concomitant short-ranged antiferromagnetic correlations. This aspect is also seen in the single particle spectral function that shows clear signs of the upper and lower Hubbard bands. Our instanton approach bears remarkable similarities to local dynamical approaches, such as dynamical mean-field theory, in the sense that it has the unique property of allowing for local moment formation without breaking the SU(2) spin symmetry. In contrast to local approaches, it captures short-ranged magnetic fluctuations. Furthermore, it also offers possibilities for systematic improvements by taking into account fluctuations around the dominant saddle point. Finally, we show that the saddle point structure depends upon the choice of lattice geometry. For the square lattice at half filling, the saddle-point structure reflects the itinerant to localized nature of the magnetism as a function of the coupling strength. The implications of our results for Lefschetz thimble approaches to alleviate the sign problem are also discussed.
To each complex saddle point of an action, one can attach a Lefschetz thimble on which the imaginary part of the action is constant. Cauchy theorem states that summation over a set of thimbles produces the exact result. This reorganization of the path integral, is an appealing starting point for various approximations: In the realm of auxiliary quantum Monte Carlo methods it provides a framework to alleviate the negative sign problem. Here, we suggest to constrain the integration to the \textit{dominant} thimbles: the thimbles attached to the saddle points with the largest statistical weight. For the Hubbard model, in a formulation where the the Hubbard Stratonovitch field couples to the charge, this provides a \textit{symmetry} consistent approximation to the physics of the Hubbard model: constraining the integration domain does not explicitly break a symmetry. We can test this approach for the Hubbard model at half-filling on a bipartite lattice. The paper builds on the previously developed instanton gas approach, where an exhaustive saddle point approximation was constructed. We present results, showing that the dominant thimbles approximation provides results that are in remarkable agreement with the exact results for various fermionic observables including spin and charge order parameters and single electron spectral functions. We discuss implications of our results for simulations away of half filling.
We study the one-band Hubbard model on the honeycomb lattice using a combination of quantum Monte Carlo (QMC) simulations and static as well as dynamical mean-field theory (DMFT). This model is known to show a quantum phase transition between a Dirac semimetal and the antiferromagnetic insulator. The aim of this paper is to provide a detailed comparison between these approaches by computing static properties, notably ground-state energy, single-particle gap, double occupancy, and staggered magnetization, as well as dynamical quantities such as the single-particle spectral function. At the static mean-field level, local moments cannot be generated without breaking the SU(2) spin symmetry. The DMFT approximation accounts for temporal fluctuations and thus captures both the evolution of the double occupancy and the resulting local moment formation in the paramagnetic phase. As a consequence, the DMFT approximation is found to be very accurate in the Dirac semimetallic phase where local moment formation is present and the spin correlation length small. However, in the vicinity of the fermion quantum critical point, the spin correlation length diverges and the spontaneous SU(2) symmetry breaking leads to low-lying Goldstone modes in the magnetically ordered phase. The impact of these spin fluctuations on the single-particle spectral function---waterfall features and narrow spin-polaron bands---is only visible in the lattice QMC approach.
The two-dimensional (2D) Hubbard model is solved in a new way. The zero-frequency components of the Hubbard interaction are separated out and solved to infinite order to avoid the traditional difficulties associated with perturbation theory in the strong-coupling limit. Specifically the zero-frequency spin interaction is decoupled by the introduction of Stratonovich-Hubbard fields, while the zero-frequency charge interaction is treated by "slave" Hartree-Fock. This approach leads to the formation of large fluctuating spin moments, and as a consequence the one-electron spectral density exhibits a pseudogap. A Metropolis Monte Carlo method is used to perform the configuration averaging over Stratonovich-Hubbard fields. Within single Monte Carlo configurations the non-zero frequency components of the Hubbard interaction are treated by a high-order perturbation method, which incorporates the exchange of dynamical particle-hole fluctuations. Since a single configuration has neither translational symmetry nor spin rotational symmetry, these fluctuations do not have a well-defined crystal momentum or total spin. When the dynamical particle-hole fluctuations are included in the calculation, localized electron pairing results at elevated temperatures. Superconductivity presumably onsets when these localized pairs develop coherence, but this is not studied in the present work. The nature of the pairing interaction and the resulting pair eigenfunctions is studied.
We present a restricted path integral approach to the two-dimensional and three-dimensional repulsive Hubbard model. In this approach the partition function is approximated by restricting the summation over all states to a (small) subclass that is chosen so as to well represent the important states. This procedure generalizes mean-field theory and can be systematically improved by including more states or fluctuations. We analyze in detail the simplest of these approximations, which corresponds to summing over states with local antiferromagnetic (AF) order. If in the states considered the AF order changes sufficiently little in space and time, the path integral becomes a finite-dimensional integral for which the saddle point evaluation is exact. This leads to generalized mean-field equations allowing for the possibility of more than one relevant saddle point. In a big parameter regime (both in temperature and filling), we find that this integral has two relevant saddle points, one corresponding to finite AF order and the other without. These degenerate saddle points describe a phase of AF ordered fermions coexisting with free, metallic fermions. We argue that this mixed phase is a simple mean-field description of a variety of possible inhomogeneous states, appropriate on length scales where these states appear homogeneous. We sketch systematic refinements of this approximation, which can give more detailed descriptions of the system.
The Hubbard model has been widely used to describe magnetism due to itinerant electrons. However, except for the one-dimensional case, no exact solution of the model exists. We have studied the two-dimensional Hubbard model on a square lattice using a Monte Carlo simulation technique. Simulations were performed on lattices of up to 8×8 in spatial size. We will discuss results for thermodynamic properties and correlation functions as a function of temperature and interaction strength. For the half-filled band case, our results indicate that the system exhibits long-range antiferromagnetic order in the ground state for any nonzero interaction, although substantially reduced from the classical Néel state due to both charge and spin fluctuations. For the non-half-filled band case, our results suggest that no long-range magnetic order exists. We also don’t find any indication of short-ranged ferromagnetic correlations in the parameter range studied. We have also studied the predictions of various approximate treatments for this model, particularly the static approximation and the coherent potential approximation (CPA). By comparison with the Monte Carlo results, we discuss the validity of these schemes. Finally, we discuss the effect of changing the band structure on the magnetic properties of the system.
We reconsider the strong-coupling expansion for the Hubbard model recently introduced by Sarker and Pairault et al. By introducing slave particles that act as projection operators onto the empty, singly occupied and doubly occupied atomic states, the perturbation theory around the atomic limit distinguishes between processes that do conserve or do not conserve the total number of doubly occupied sites. This allows for a systematic t/U expansion that does not break down at low temperature (t being the intersite hopping amplitude and U the local Coulomb repulsion). The fermionic field becomes a two-component field, which reflects the presence of the two Hubbard bands. The single-particle propagator is naturally expressed as a function of a 2×2 matrix self-energy. Furthermore, by introducing a time- and space-fluctuating spin-quantization axis in the functional integral, we can expand around a "non-degenerate" ground-state where each singly occupied site has a well defined spin direction (which may fluctuate in time). This formalism is used to derive the effective action of charge carriers in the lower Hubbard band to first order in t/U. We recover the action of the t–J model in the spin-hole coherent-state path integral. We also compare our results with those previously obtained by studying fluctuations around the large-U Hartree–Fock saddle point.