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

Journal of Statistical Physics

The temperature zero limit

Joel Feldman, Horst Knörrer, Manfred Salmhofer B, Eugene Trubowitz B

Published Jan 1, 1998• 11 citations
Publisher

Abstract

We prove that, for a broad class of many--fermion models, the amplitudes of renormalized Feynman diagrams converge to their temperature zero values in the limit as the temperature tends to zero.
DOI
10.3929/ethz-a-004353120

Related papers

Finite-temperature renormalization-group study of a dissipative gauge theory with fermions

Hiroshi Takano, Masaru Onoda, Ikuo Ichinose +1 more

Nuclear Physics B•Mar 1, 1999•0 citations

Monte Carlo test of critical exponents and amplitudes in 3D Ising and phi^4 lattice models

J. Kaupužs

We have tested the leading correction-to-scaling exponent omega in O(n)-symmetric models on a three-dimensional lattice by analysing the recent Monte Carlo (MC) data. We have found that the effective critical exponent, estimated at finite sizes of the system L and L/2, decreases remarkably within the range of the simulated L values. This shows the incorrectness of some claims that omega has a very accurate value 0.845(10) at n=1. A selfconsistent infinite volume extrapolation yields row estimates omega=0.547, omega=0.573, and omega=0.625 at n=1, 2, and 3, respectively, in approximate agreement with the corresponding exact values 1/2, 5/9, and 3/5 predicted by our recently developed GFD (grouping of Feynman diagrams) theory. We have fitted the MC data for the susceptibility of 3D Ising model at criticality showing that the effective critical exponent eta tends to increase well above the usually accepted values around 0.036. We have fitted the data within [L;8L], including several terms in the asymptotic expansion with fixed exponents, to obtain the effective amplitudes depending on L. This method clearly demonstrates that the critical exponents of GFD theory are correct (the amplitudes converge to certain asymptotic values at L tending to infinity), whereas those of the perturbative renormalization group (RG) theory are incorrect (the amplitudes diverge). A modification of the standard Ising model by introducing suitable "improved" action (Hamiltonian) does not solve the problem in favour of the perturbative RG theory.

arXiv:cond-mat/0106271•Jun 14, 2001•0 citations

CONSTRUCTIVE MANY-BODY THEORY

Joel Feldman, JACQUES MAGNEN, Vincent Rivasseau +1 more

We review the rôle played by renormalization, particle-number symmetry breaking, Goldstone bosons and the renormalization group in a mathematically rigorous construction of many-Fermion models at temperature zero.

Reviews in Mathematical Physics•Jan 1, 1994•1 citation

Momentum occupation number bounds for interacting fermions

Matthew F. Lapa

We derive rigorous bounds on the average momentum occupation numbers $\langle n_{\mathbf{k}σ}\rangle$ in the Hubbard and Kondo models in the ground state and at non-zero temperature ($T>0$) in the grand canonical ensemble. For the Hubbard model with $T>0$ our bound proves that, when interaction strength $\ll k_B T\ll$ Fermi energy, $\langle n_{\mathbf{k}σ}\rangle$ is guaranteed to be close to its value in a low temperature free fermion system. For the Kondo model with any $T>0$ our bound proves that $\langle n_{\mathbf{k}σ}\rangle$ tends to its non-interacting value in the infinite volume limit. In the ground state case our bounds instead show that $\langle n_{\mathbf{k}σ}\rangle$ approaches its non-interacting value as $\mathbf{k}$ moves away from a certain surface in momentum space. For the Hubbard model at half-filling on a bipartite lattice, this surface coincides with the non-interacting Fermi surface. In the Supplemental Material we extend our results to some generalized versions of the Hubbard and Kondo models. Our proofs use the Fermi statistics of the particles in a fundamental way.

arXiv:2106.13818•Jun 25, 2021•0 citations

Renormalization group study of interacting fermions at finite temperatures

Adrian A. Drǎgulescu, Adrian Neagu

•Jan 1, 1995•0 citations

Renormalization of many body fermion models with singular fermi surfaces

Daniel Brox

The physical system modelled is a periodic lattice that provides a background for interacting fermions. An effective model of the interacting fermions at zero temperature may be studied using standard techniques of many-body quantum mechanics. All the physical information of such a system is encoded in its Schwinger's functions that tell us how states evolve in time and what the matrix elements of all operators are. It is natural to do calculations on the interacting system by treating it as a perturbation of the corresponding noninteracting system, as the eigenstates and energy levels of the latter are known. All Schwinger functions may be conveniently cast in the path integral formalism, and combined into a generating functional. The Schwinger functions are shown to be sums of Feynman diagrams, where the electron propagators blow-up at the Fermi surface. Power counting is described and used to reveal subdiagram divergences. Renormalization via the renormalized Gallavotti-Nicolo Tree Expansion is implemented in an attempt to circumvent these subdiagram divergences and obtain finite values for the Schwinger functions. It is shown that if the Fermi surface has at worst quadratic singularities, the perturbation coefficients of the renormalizing counterterms are finite and remain so as the infrared cutoff is removed.

•Jan 1, 2009•0 citations
We have tested the leading correction-to-scaling exponent omega in O(n)-symmetric models on a three-dimensional lattice by analysing the recent Monte Carlo (MC) data. We have found that the effective critical exponent, estimated at finite sizes of the system L and L/2, decreases remarkably within the range of the simulated L values. This shows the incorrectness of some claims that omega has a very accurate value 0.845(10) at n=1. A selfconsistent infinite volume extrapolation yields row estimates omega=0.547, omega=0.573, and omega=0.625 at n=1, 2, and 3, respectively, in approximate agreement with the corresponding exact values 1/2, 5/9, and 3/5 predicted by our recently developed GFD (grouping of Feynman diagrams) theory. We have fitted the MC data for the susceptibility of 3D Ising model at criticality showing that the effective critical exponent eta tends to increase well above the usually accepted values around 0.036. We have fitted the data within [L;8L], including several terms in the asymptotic expansion with fixed exponents, to obtain the effective amplitudes depending on L. This method clearly demonstrates that the critical exponents of GFD theory are correct (the amplitudes converge to certain asymptotic values at L tending to infinity), whereas those of the perturbative renormalization group (RG) theory are incorrect (the amplitudes diverge). A modification of the standard Ising model by introducing suitable "improved" action (Hamiltonian) does not solve the problem in favour of the perturbative RG theory.
We derive rigorous bounds on the average momentum occupation numbers ⟨nkσ⟩\langle n_{\mathbf{k}σ}\rangle⟨nkσ​⟩ in the Hubbard and Kondo models in the ground state and at non-zero temperature (T>0T>0T>0) in the grand canonical ensemble. For the Hubbard model with T>0T>0T>0 our bound proves that, when interaction strength ≪kBT≪\ll k_B T\ll≪kB​T≪ Fermi energy, ⟨nkσ⟩\langle n_{\mathbf{k}σ}\rangle⟨nkσ​⟩ is guaranteed to be close to its value in a low temperature free fermion system. For the Kondo model with any T>0T>0T>0 our bound proves that ⟨nkσ⟩\langle n_{\mathbf{k}σ}\rangle⟨nkσ​⟩ tends to its non-interacting value in the infinite volume limit. In the ground state case our bounds instead show that ⟨nkσ⟩\langle n_{\mathbf{k}σ}\rangle⟨nkσ​⟩ approaches its non-interacting value as k\mathbf{k}k moves away from a certain surface in momentum space. For the Hubbard model at half-filling on a bipartite lattice, this surface coincides with the non-interacting Fermi surface. In the Supplemental Material we extend our results to some generalized versions of the Hubbard and Kondo models. Our proofs use the Fermi statistics of the particles in a fundamental way.
The physical system modelled is a periodic lattice that provides a background for interacting fermions. An effective model of the interacting fermions at zero temperature may be studied using standard techniques of many-body quantum mechanics. All the physical information of such a system is encoded in its Schwinger's functions that tell us how states evolve in time and what the matrix elements of all operators are. It is natural to do calculations on the interacting system by treating it as a perturbation of the corresponding noninteracting system, as the eigenstates and energy levels of the latter are known. All Schwinger functions may be conveniently cast in the path integral formalism, and combined into a generating functional. The Schwinger functions are shown to be sums of Feynman diagrams, where the electron propagators blow-up at the Fermi surface. Power counting is described and used to reveal subdiagram divergences. Renormalization via the renormalized Gallavotti-Nicolo Tree Expansion is implemented in an attempt to circumvent these subdiagram divergences and obtain finite values for the Schwinger functions. It is shown that if the Fermi surface has at worst quadratic singularities, the perturbation coefficients of the renormalizing counterterms are finite and remain so as the infrared cutoff is removed.