Partition function zeros for the Ising model on complete graphs and on annealed scale-free networks

M. Krasnytska, B. Berche, Y. Holovatch, Ralph Kenna

    Research output: Contribution to journalArticlepeer-review

    22 Citations (Scopus)
    63 Downloads (Pure)

    Abstract

    We analyze the partition function of the Ising model on graphs of twodifferent types: complete graphs, wherein all nodes are mutually linked and annealedscale-free networks for which the degree distribution decays as P(k) ∼ k−λ. We areinterested in zeros of the partition function in the cases of complex temperature orcomplex external field (Fisher and Lee-Yang zeros respectively). For the model on anannealed scale-free network, we find an integral representation for the partition functionwhich, in the case λ > 5, reproduces the zeros for the Ising model on a completegraph. For 3 <λ <5 we derive the λ-dependent angle at which the Fisher zerosimpact onto the real temperature axis. This, in turn, gives access to the λ-dependentuniversal values of the critical exponents and critical amplitudes ratios. Our analysisof the Lee-Yang zeros reveals a difference in their behaviour for the Ising model on acomplete graph and on an annealed scale-free network when 3 <λ <5. Whereas inthe former case the zeros are purely imaginary, they have a non zero real part in lattercase, so that the celebrated Lee-Yang circle theorem is violated.
    Original languageEnglish
    Article number135001
    JournalJournal of Physics A: Mathematical and Theoretical
    Volume49
    Issue number13
    Early online date17 Feb 2016
    DOIs
    Publication statusPublished - 1 Apr 2016

    Bibliographical note

    Due to publisher policy, the full text will not be available on the repository until 17th February 2017. The full text can, however, be found online here :http://arxiv.org/abs/1510.00534

    Fingerprint

    Dive into the research topics of 'Partition function zeros for the Ising model on complete graphs and on annealed scale-free networks'. Together they form a unique fingerprint.

    Cite this