Iron Covalency Assumptions and Redox Equilibrium in Vitrification

Abstract

Iron Redox equilibrium is important to nuclear waste vitrification because it impacts melt foaming, crystallization, viscosity, heat transfer, and potentially decreases the durability of the glass product. The last 40 years has seen much debate concerning the correct stoichiometry of the iron Redox reactions in molten glass. The reason for this debate on something seemingly simple is that assumptions have been imposed on the reaction regarding the inclusion or omission of different types of oxygen anions in order to rationalize glass composition effects. One popular stoichiometry has been rationalized in the literature by the tacit assumption that iron in octahedral coordination is primarily ionically bonded to oxygen whereas iron in tetrahedral coordination is primarily covalently bonded. This paper challenges that assumption. In this study, Racah B Parameters, a measure of iron covalency, were compiled and sorted by coordination number for ferric iron in all of the silicate and oxide minerals available. This comparison found no correlation between iron coordination number and Fe(III)-O bond covalency. The most ionic ferric iron in the database was iron-doped quartz, where iron is in tetrahedral coordination. Conversely, the most covalent iron oxide in the database was hematite (Fe 2 O 3 ), which has iron in octahedral coordination. This indicates that it is unsafe to assume that tetrahedral iron is more covalent in glass than octahedral iron. A reaction stoichiometry is proposed that includes the entire iron coordination-sphere because this mole balance demonstrates that oxygen anions are traded between the iron coordination-sphere and the bulk glass during the Redox reaction. These traded oxygen anions account for the effects of glass composition and iron coordination number on the Redox equilibria without invoking covalency assumptions.