Dependence of Critical Properties of Heisenberg Magnets upon Spin and Lattice

Abstract

High‐temperature expansion methods have recently been used to predict the form of the divergence of the zero‐field static susceptibility χ at the critical temperature T_c. These studies have suggested that χ∼A (T−T_c)^−γ, with γ=$\frac{4}{3}$ independent of both lattice structure and spin quantum number S. Here we argue that the proposal γ=$\frac{4}{3}$ for all S is unjustified; we find a slow but nevertheless clear variation of γ with S. We further point out that there exists at least one physically interesting lattice—a normal cubic spinel with nearest‐neighbor ferromagnetic B—B interactions—for which the theoretical evidence indicates that if the power‐law form of divergence is correct, γ may differ from$\frac{4}{3}$ by as much as 50%.