ELECTRON RELAXATION IN DEERITE

Résumé
Des résultats obtenus antérieurement par spectrométrie Mössbauer d'un minéral rare, la déérite, sont réinterprétés en introduisant un processus de relaxation electronique.

Abstract
Up to now iron sites in Deerite, a relatively rare mineral, have been investigated by two research groups, using NGR spectroscopy. Bancroft et al. /1/ publishing the first data, interpreted their spectra as a superposition of three doublets : one from Fe⁺⁺ in a tetrahedral site. Franck and Bunbury /2/ performed a temperature dependent analysis. They only admit octahedral sites ; in spite of an extremely high temperature behaviour of the third doublet. Susceptibility measurments /3/ could not be clearly interpreted, neither assuming only low spin iron or only high spin ions. Overmore a mixture of both is evidently in contradiction with the former NGR experiments. Not before 1977 (Fleet /4/) was the crystal structure of Deerite well understood. Cations occupy only octahedral sites, forming strips of 6 sites in the a-b plane and continuous band along the c-axis by the superposition of three of those strips. The sites fall in three nearly equivalent classes with bond lengths of 2.09 for M1, 2.15 for M4 and 2.15 Å for M7 sites. As the M1 site has 4 edges with neighbours, electron hopping, as proposed by Pollak et al. /5/ and Evans /6/ should be expected. Reconsidering, with the aid of the now known crystal structure of Deerite, the results of Franck and Bunbury with the assumption of an electron relaxation we obtain a coherent explanation with the temperature dependence. We use, instead of a fast electron hopping model, excluded by the shape of the spectra and their temperature behaviour, the general relaxation expression /7/. This formulation includes two experimental parameters : the activation energy and the transition rate. We arrived althus at very good fits, although the ferrous/ferric ration is somewhat too low, when not taking in account the exchange spectrum. In fact those should be considered as due mainly to ferrous ions giving a ferrous/ferric ratio in agreement with crystallographic data.