Answers are grounded in this paper and cite their sources.
A general model in which the nearest-neighbor exchange interactions J(→x,→x^') are of the form $J({{→}}{x},{{{→}}{x}}^{{'}})={J}₁+{J}₂{ε}({{→}}{x}){ε}({{{→}}{x}}^{{'}})+{J}₃[{ε}({{→}}{x})+{ε}({{{→}}{x}}^{{'}})]+{J}₄{μ}({{→}}{x},{{{→}}{x}}^{{'}})$ is considered. Here ${ε}({{→}}{x})$ is a random site variable and ${μ}({{→}}{x},{{{→}}{x}}^{{'}})$ is a random bond variable. It is argued that ${J}₄$ tends to produce the Edwards-Anderson-type spin glass, whereas ${J}₂$ produces a different type of phase which is like an alloy of up and down spins. That these two phases are distinct follows from the existence of a phase boundary in the ${J}₂{-}{J}₄$ plane when ${J}₁={J}₃=0$. A consistent, but qualitative, discussion of this model via the renormalization group is also given.
No takes yet. Share an insight, caveat, or question.
Harris et al. (1977) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: