Tulanla, Vol. 32, No. 5, pp. 395-398. 1985 Printed in Great Britain 0039-9140/85 %3.00 + 0.00 Pergamon Press Ltd zyxwvut A METHOD FOR THE DETERMINATION OF VANADIUM AND IRON OXIDATION STATES IN NATURALLY OCCURRING OXIDES AND SILICATES zyxwvutsrqponmlkjihgfed RICHARD B. WANTY and MARTIN B. GOLDHABER U.S. Geological Survey, MS 916, Box 25046 Denver Federal Center, Denver, CO 80225, U.S.A. zyxwvutsrqponml (Received 21 September 1984. Accepted 14 December 1984) Summary-A valence-specific analytical method for determining V3+ m ore minerals has been developed that involves two steps: dissolution of a mineral sample without disturbing the V3+/Vt0’ratio, followed by determination of V3+ in the presence of V4+. The samples are dissolved in a mixture of hydrofluoric and sulphuric acids at 100” in Teflon-lined reaction vessels. Tervalent vanadium is then determined calorimetrically by formation of a V3+-thiocyanate complex in aqueous-acetone medium. Fe3+ is measured semi-quantitatively in the same solution. The method has been tested with two naturally occurring samples containing vanadium and iron. The results obtained were supported by those obtained by other methods, including electron spin resonance spectroscopy, the~ogravimet~c analysis, and MSssbauer spectroscopy. In natural systems, vanadium is present in minerals and waters in the + 3, +4 and + 5 oxidation states. This fact may be advantageously applied in the determination of the redox state of an entire system. For instance, the presence of tervalent vanadium is an indication of an extremely reducing environment,‘” whereas the quadrivalent and quinquevalent states predominate in systems of intermediate and high oxidation potential, respectively. Many analytical techniques have been developed for the deter- mination of total vanadium in rock samples,“’ but no techniques have been reported whereby one vanadium oxidation state in a mineral sample is measured in the presence of another. This paper presents an analytical procedure that has been devel- oped to detect tervalent vanadium in minerals in the presence of quadrivalent vanadium. Tervalent iron in the minerals is also measured. The method has been tested with samples of the mineral roscoelite, a vanadiferous mica, and of a vanadium oxide phase from a sandstone-hosted vanadium-uranium ore- body from the Colorado Plateaus. The results should aid in the inte~retation of the processes of ore genesis and contribute to our understanding of the importance of redox processes in ore genesis. EXPERlM ENTA L Sample collection and handling Vanadium(II1) is unstable in the presence of oxygen. Therefore, all sample handling must be done in an inert atmosphere. In this study, high-purity nitrogen containing <SO ppm of oxygen was used. Roscoelite samples were obtained from an inactive mine at Placerville, Colorado. Samples of vanadium oxide were collected from freshly exposed faces in the Tony M mine, near Hanksville, Utah. The samples were stored in Mason jars flushed with nitrogen immediately after sample collection. Once the samples had been collected and brought back to the laboratory, all sample handling (grinding, size and density separation) was done under nitrogen in a glove-bag. All reagents which were to contact the samples were purged with nitrogen for at least 30 min. The efficiency of this purging process was verified by Winkler titration.’ Heavy- mineral separations were performed with fresh bromoform which had been purged with nitrogen for ~2 hr, and the samples were then rinsed with nitrogen-purged acetone. All size and density fractions were dried in a vacuum desiccator and stored in nitrogen-purged desiccators. Although the roscoelite samples were not collected from freshly exposed mine fz&es, repeated analyses over the course of several months showed no change in the V3+/Vto1 ratio. Similar experiments on the vanadium oxide mineral showed that after several weeks significant oxidation had occurred. Typical results (vanadium oxide sample TM-5 from the Tony M mine) are shown in Fig. 1. The analysis at zero time corresponds to a fresh sample stored under nitrogen. After one week of exposure to air, the V3+/Vtot ratio was virtually unchanged from the initial value of 0.82. After 6 weeks, the ratio had dropped to a value of 0.46. A split of the sample was stored in a cloth bag to allow complete exposure to air. Analysis of this split after 22 months exposure to air gave a V3+/V“” ratio of 0.10, assumed to be the lowest achievable value. It is interesting to note that two of the four analyses reported by Weeks et af.8 on the vanadium oxide mineral montroseite yielded V3+/Vtot ratios of approximately 0.15, and a third showed no V3’. These values are close to our “infinite time” value of 0.10, suggesting that the earlier analysesa were performed on oxidized material. Reagents Ammonium thiocyanate solution (2.5OM) in acetone. Dis- solve 28.5 g of reagent-grade ammonium thiocyanate in 150 ml of acetone. Prepare fresh daily. Stock vanadium solution (0.05OM). Dissolve 0.585 g of reagent-grade ammonium metavanadate, NHYO,,- in 100 ml of 6M hydrochloric acid. On standing for about 2 months, this solution will turn bright blue as quin- quevalent vanadium is reduced to the quadrivalent form, with liberation of chlorine. Procedure Introduce the samples (appro~mately 0.1 g) into the reaction vessels. Add to each 5.0 ml of 5M sufphuric acid and 5 ml of concentrated hydrofluoric acid, and seat the 395