Influence of nondiagonal dynamic susceptibility on the EPR signal of Heisenberg magnets

Abstract

In EPR with linearly polarised microwaves if just the diagonal contributions chi ++( omega ) and chi --( omega ) of the dynamic susceptibility are considered then a discrepancy occurs between the experimentally observed signal and the predictions of theory. By means of the Mori formalism the authors have calculated the non-diagonal contributions chi +-( omega ) and chi -+( omega ) for insulators with dominating isotropic exchange. For rather small linewidth these terms result in a new kind of dynamic shift of the resonance field, whereas for broad lines they give rise to a drastic anomaly of lineshape. All these effects could be verified experimentally in both one- and three-dimensional spin lattices of axial symmetry in the following ways: the dynamic shift expected for narrow lines could be observed in TMMC; lineshape anomalies were studied in the very broad line of CsNiF3; and in MnF2 the critical increase of linewidth near TN allowed variation between the two extreme cases.

Anisotropic exchange inLiCuVO4probed by ESR

“…Here, we are dealing with an insulator, where the asymmetry arises from the influence of non-diagonal elements of the dynamic susceptibility: This effect is often observed in systems with interactions of low symmetry and sufficiently broad resonance lines (H res ≈ ∆H) [32].…”

Spin Dynamics in the Paramagnetic Regime: NMR and EPR in Two-Dimensional Magnets

“…A critical examination of the underlying theory, however, attributes this shift to special apparatus conditions (linear polarization of the microwave) rather than to low-d spin dynamics [32]. So, due to other competing shift mechanisms, which may be even larger than the interesting one, and due to experimental uncertainly of the g-factor, it would be very difficult to probe the diffusive dynamics by analyzing the dynamic shift.…”

Electron paramagnetic resonance in weakly anisotropic Heisenberg magnets with a symmetric anisotropy

“…When H and u are not parallel, the spin precession around z is no longer circular but elliptic, and an asymmetry shows up. Such a result has already been obtained ͑at infinite temperature͒ by BBSW, 11 but without a rigourous treatment. In particular, they assumed the S z motion decoupled from S x and S y , which is only true when Hʈu.…”