Arthur Ramirez

The specific heat of the linear chain magnetic compound CsNiF(,3) has been measured at temperatures between 4.2 and 20 K and in magnetic fields up to 1.0 T applied perpendicularly to the chain direction. For this purpose, a computer-controlled automated calorimeter was developed which enabled concurrent specific heat data acquisition and analysis.

Rounded peaks were observed in both the field and temperature dependence of the specific heat. Such peaks can not be explained in terms of either spin waves or Zeeman split impurity levels.

These peaks are in qualitative agreement, however, with a classical sine-Gordon theory involving linear waves, solitons, and a phase shift interaction between them. The agreement in the peak positions is improved by an ad hoc reduction of the soliton rest energy by 30%. This energy reduction is consistent with neutron scattering, NMR, and light absorption measurements of CsNiF(,3) using a soliton interpretation of the data. The discrepancy between predicted and measured peak heights, however, is larger than that found for the peak positions and can not be removed by simple energy scaling.

The discrepancies for both peak position and height may be ascribed to one or more of the approximations made in mapping the CsNiF(,3) Hamiltonian to the sine-Gordon Hamiltonian and several independent attempts have been made to understand the specific heat results in terms of these approximations. A review is presented of the theories relevant for a modified soliton interpretation of the specific heat measurements. It is found that there is no complete theory for the observed nonlinear phenomena at this time but that such a theory will most likely involve modified sine-Gordon solitons and remains a challenging and important problem.