Loop expansion around the Bethe solution for the random magnetic field Ising ferromagnets at zero temperature

Maria Chiara Angelini, Carlo Lucibello, Giorgio Parisi, Federico Ricci‐Tersenghi, Tommaso RizzoProceedings of the National Academy of Sciences 2020 · DOI

TL;DR

This paper introduces a novel "loop expansion" method around the Bethe solution to study the complex behavior of the Random Magnetic Field Ising Model (RFIM) at zero temperature. This new approach offers a more appropriate framework for understanding strongly disordered systems, particularly those governed by a zero-temperature renormalization group fixed point, and reveals new theoretical insights that are absent in traditional methods.

What Problem Does This Paper Address?

The critical behavior of the Random Magnetic Field Ising Model (RFIM) has been a long-standing puzzle in statistical mechanics. While a ferromagnetic transition is expected in the presence of a random magnetic field, its critical properties differ from ordinary ferromagnets. Traditional methods, like the $\epsilon$-expansion around the upper critical dimension (D=6), have faced challenges in accurately describing these strongly disordered systems, especially at low temperatures and finite connectivity. A key issue is the potential breakdown of "Dimensional Reduction" (DR), a conjecture that relates the critical properties of the RFIM in D dimensions to a pure Ising model in D-2 dimensions. This breakdown is particularly evident at lower dimensions (D<6), where the validity of DR is questioned. This paper aims to provide a more robust theoretical framework to understand the RFIM's critical behavior at zero temperature, specifically addressing the limitations of existing methods and the potential failure of DR.

Methods

The authors employ a new "loop expansion" method that starts from the Bethe solution, a sophisticated mean-field theory applicable to random regular graphs (Bethe lattices). This approach is contrasted with the standard $\epsilon$-expansion, which is based on the Landau-Ginzburg (LG) effective Hamiltonian. The key differences lie in:

  • Starting Point: The Bethe solution is derived from a finite coordination number, unlike the infinite coordination number used in the standard LG approach.
  • Fluctuation Treatment: The Bethe solution inherently captures fluctuations, which are crucial for understanding strongly disordered systems at low temperatures, whereas the LG approach often neglects them at the leading order.
  • Effective Theory: The new loop expansion naturally leads to an effective theory with cubic vertices, in addition to the quartic vertices found in the standard LG approach.

The authors then compute one-loop corrections arising from these cubic vertices and compare them to the standard $\epsilon$-expansion results. They also utilize numerical simulations on Bethe lattices to validate their analytical findings.

Datasets

Not discussed in the paper. The study is theoretical and relies on analytical calculations and numerical simulations on model systems (Bethe lattices).

Key Insights and Ideas

  • Bethe Solution as a Better Starting Point: The Bethe solution, derived from finite coordination number models, provides a more appropriate starting point for understanding the RFIM at zero temperature compared to the standard Landau-Ginzburg approach, especially for strongly disordered systems.
  • Cubic Vertices Emerge Naturally: The loop expansion around the Bethe solution naturally introduces cubic vertices into the effective theory. These vertices are absent in the standard $\epsilon$-expansion and are crucial for describing phenomena like avalanches.
  • Avalanches Explained: The presence of cubic vertices is shown to be essential for correctly describing avalanches, a collective zero-temperature phenomenon in the RFIM that cannot be explained by the standard $\phi^4$ theory. The authors recover the correct mean-field exponent for avalanche size distribution.
  • Dimensional Reduction Holds (at 1-loop): Despite the introduction of new terms from cubic vertices, these terms are found to be subdominant compared to the standard supersymmetric terms at the one-loop level. This implies that Dimensional Reduction remains valid at this order of the loop expansion, even in dimensions close to the upper critical dimension (D=6-$\epsilon$).
  • Beyond Mean-Field Behavior: The Bethe lattice approach, by incorporating fluctuations, reveals a richer picture of critical behavior, particularly the dominance of rare fluctuations and the emergence of phenomena like avalanches, which are not captured by simpler mean-field theories.

Key Results

  • The one-loop corrections due to cubic vertices introduce new terms that are absent in the standard $\epsilon$-expansion.
  • These new terms are subdominant to the standard supersymmetric terms at the one-loop order, preserving Dimensional Reduction in the $6-\epsilon$ expansion.
  • The new framework naturally explains the critical exponent of avalanche size distribution on Bethe lattices, a feat not achievable with the standard $\phi^4$ theory.
  • Numerical simulations on Bethe lattices are used to support the analytical findings, particularly regarding correlation functions and the decay of the mean coupling.

Limitations and Future Work

  • Subdominance of New Terms: While new terms are found, their subdominant nature at one loop means that the most significant implications for Dimensional Reduction are not immediately apparent at this order. Further investigation at higher loop orders might be necessary to fully assess their impact.
  • Analytical Ansatz Completeness: The analytical Ansatz for the distribution of couplings and fields, while successful, has some undetermined coefficients and might lack certain terms from the J≠0 part of the expansion, limiting exact calculations for some quantities.
  • Future Work: The authors suggest extending the analysis to higher loop orders to fully understand the role of cubic vertices and their potential impact on Dimensional Reduction. They also plan to compute one-loop corrections for three-point functions to investigate the scaling dimension of avalanches and their fractal dimension. The method's applicability to other complex systems like spin glasses and the Anderson transition is also highlighted as a future direction.

Summary updated

Abstract

We apply to the random-field Ising model at zero temperature ([Formula: see text]) the perturbative loop expansion around the Bethe solution. A comparison with the standard ϵ expansion is made, highlighting the key differences that make the expansion around the Bethe solution much more appropriate to correctly describe strongly disordered systems, especially those controlled by a [Formula: see text] renormalization group (RG) fixed point. The latter loop expansion produces an effective theory with cubic vertices. We compute the one-loop corrections due to cubic vertices, finding additional terms that are absent in the ϵ expansion. However, these additional terms are subdominant with respect to the standard, supersymmetric ones; therefore, dimensional reduction is still valid at this order of the loop expansion.
BibTeX
@inproceedings{angelini2020loop,
  author = {Maria Chiara Angelini and Carlo Lucibello and Giorgio Parisi and Federico Ricci‐Tersenghi and Tommaso Rizzo},
  title = {{Loop expansion around the Bethe solution for the random magnetic field Ising ferromagnets at zero temperature}},
  booktitle = {Proceedings of the National Academy of Sciences},
  year = {2020},
  doi = {10.1073/pnas.1909872117},
}

Loading related content...

Figure