Abstract
Bulk magnetic properties of the related antiferromagnets KNiF3 and K2NiF4 are analyzed by use of both high- and low-temperature theoretical methods. A crystal-field theory for the Ni2+ ion is outlined, using the strong-field coupling scheme, and is used to construct relevant spin Hamiltonians for the two systems. KNiF3 is found to be a simple-cubic antiferromagnet for which all except nearest-neighbor interactions are negligible. Writing an exchange interaction JSi·Sj between nearest-neighbor spins Si and Sj, the experimental observations are quantitatively fitted for J=89±4 °K with a suggestion of a small temperature dependence of J. Reasons are put forward to suggest that K2NiF4 is, to a very good approximation, a two-dimensional quadratic-layer antiferromagnet with a magnetic state which, over a large temperature region, shows a long-range order in two dimensions but not in the third. A good theoretical interpretation of most of the published experimental data can be obtained using such a picture. K2NiF4 is made up of NiF2 magnetic planes, which are essentially the same as those found in KNiF3, but with these magnetic layers separated in the third dimension by two nonmagnetic KF layers. We find an intraplane exchange J125 °K and an interplane exchange J which is at least an order of magnitude smaller and probably ferromagnetic in sign. A number of major differences between typical two- and three-dimensional magnetic structures are discussed, and some experiments are outlined for which the use of two- dimensional systems like K2NiF4 could provide new and interesting information concerning magnetic cooperative phenomena. Finally, the possible origins of the difference in exchange values J between KNiF3 and K2NiF4 are discussed, and the relevance of our findings in connection with the supposedly anomalous properties of another series containing two-dimensional antiferromagnets, namely, the calcium manganites, is assessed.