Substitutions of Divalent Transition Metal Ions in Yttrium Iron Garnet

Abstract

Various amounts of the divalent ions of manganese, iron, cobalt, and nickel have been substituted for trivalent iron in yttrium iron garnet. The effect of substitution of Fe2+ or Mn2+ ions in dodecahedral sites has also been studied. Electrical balance was accomplished by the simultaneous substitution of tetravalent silicon or germanium. Under these conditions, it is concluded that the divalent ions prefer octahedral sites. The great affinity of silicon for the tetrahedral sites in the garnets results in the reduction of trivalent to divalent iron at high temperatures even in air atmosphere when appropriate amounts of finely divided silica are present. In Y3Fex2+Fe5−2x3+SixO12, the maximum attainable value of x (under the conditions of our experiments) appears to be 0.45. In attempting to put Fe2+ ions into dodecahedral sites, greatest success was attained when there was also some in the octahedral sites as in {Y2.9Fe0.12+}[Fe0.32+Fe1.73+](Fe2.63+Si0.4)O12. The 0.1 was the maximum amount that could be put into the dodecahedral sites. Although, it appears that the Co2+ ion is the easiest to substitute into yttrium iron garnet, it is perhaps the most difficult to understand, and only a few of the results on experiments with this ion are included. In all substitutions of divalent magnetic ions not involving the dodecahedral sites, the 0°K moments are smaller than those calculated on the basis of the simplest interaction model. Several possible explanations are explored; a plausible one is the random canting of some tetrahedral ion moments resulting from weaker interaction with the octahedral divalent ions. The possibility that some of the divalent ions enter tetrahedral sites is not entirely excluded especially in the case of the Co2+ ion. The divalent magnetic ions do not behave like the trivalent magnetic lanthanide ions when substituted in the dodecahedral sites. They appear not to contribute to the net spontaneous moment appreciably.