THE UNIVERSITY OF MICHIGAN INDUSTRY PROGRAM OF THE COLLEGE OF ENGINEERING EFFECT OF V205 ON SPINEL FORMATION D. M. GR][MES J. W. ~JIPER This paper is to be presented at the 1955 Thanksgiving meeting of the American Physical Society to be held at the University of Chicago on November 25 and 26, 1955. October, 1955 IP-135

ACKNOWLEDGEMENT We wish to express our appreciation to the authors for their permission to distribute this prepaper under the Industry Program of the College of Engineering.

INTRODUCTION In the accompanying article, by the term ferrite is meant a compound of trivalent iron and some univalent or divalent metal and accompanying oxygen. The structure is the spinel structure. This material is ferrimagnetic inasmuch as it has a permeability much greater than air, and a nonzero spontaneous magnetic moment. Some of the materials are magnetically "hard" but most are not. It is suggested that the mechanism of ferrite formation here discussed might be related to the problems involving corrosion of metals, including stainless steel, when subjected to exhaust fumes containing vanadium.

EFFECT OF V205 ON SPINEL FORMATION By D. M. Grimes and J. W. Kuiper Department of Electrical Engineering and Engineering Research Institute University of Michigan It is well known that certain substances act as mineralizers or fluxes inasmuch as they promote solid-state reactions. We have found and reported elsewhere* that V205 has such an effect upon a mixed nickel zinc ferrite. Further, we found that V205 seemed to be quite unique in this action. We have now extended our investigations to include Fe203 plus V205 and the following compounds: NiO, ZnO, Li2CO3, CdO, MnO2, MnCO3, Fe2O3, CoC03, C0203, MgO, CuO, BaO and CaO. All materials used were commercially available CP materials. The mole fraction of materials added were.4925 Fe203 plus.0075 V205 plus.5000 RO or equivalent in the case of the higher valent forms and of the carbonates. The amount was also adjusted for the lithium and barium compounds. In each case a control with.5000 Fe203 plus.5000 RO was made. Each material was pressed into a toroid. with a green OD of 25.4 mm. The material was fired four hours at 950OC. We use as a criterion for reaction completeness the final OD of the sample containing no vanadium minus the final OD of the sample containing vanadium. The results are shown in the table. *"Effect of V205 on Nickel Zinc Ferrite Formation," D. M. Grimes, L. Thomassen, C. F. Jefferson, N. C. Kothary. J. Chem. Phys. To be published. 1

Comparison of Shrinkage Caused by V205 Addition Reaction Furthered Reaction Retarded Reaction Unaffected Material tOD (mm) Material AOD (mm) Material AOD (mm) NiO 2.2 CuO -0.6 CaO 0.2 ZnO 3.6 Co203 -0.5 BaO -o.o6 CdO 3.0 MnO2 -0.6 MnC03 3.1 Li2C03 2.4 Fe203 1.0 CoC03 2.0 MgO* o.6 *Fired at 11000C. The difference in strength of attraction to an external magnet, in the case of the inverted spinels, was roughly equivalent to the difference in OD with the exception of the MnC03 material, which was not attracted to a magnet. Previous magnetic measurements on nickel zinc ferrites showed that the difference in magnetic properties obtained using V205 could be repeated by increasing the firing temperature. It remains to be determined if this is also true for materials such as MgFe204 which normally require a higher firing temperature. The mechanism by which V205 produces these results is uncertain. However, two methods are suggested: (1) The melting temperature of V205 is 8950C.** It is possible that the vanadium liquifies and dissolves some neighboring oxides. These in turn precipitate out as the spinel. Experimentally it was found that the material at no time underwent appreciable plastic flow. (2) Perhaps the vanadium goes into the spinel lattice as V5+, thus necessitating cation vacancies on the B sublattices, and thereby promoting diffusion. Method (2) would explain the action of CuO by presuming that in the neighborhood of the V5+ the copper cation would go from the plus two to the plus one **Handbook of Physics and Chemistry, Chemical Rubber Publishing Company.

valance state, thereby not only removing the vacancies but also placing blocks of three cations which must diffuse together and would thus block other diffusion. Likewise Co3+ and Mn4+ would presumably go to Co2+ and Mn2+. It is possible that both mechanisms prevail. Neither Ca2+ or Ba2+ form an iron spinel. Acknowledgement The authors wish to acknowledge the aid of Professor L. Thomassen and Mr. C. F. Jefferson for their interest and stimulating conversations.