Some aspects of the geochemistry of fluorine in metamorphic bocks

Abstract

The literature on the determination of fluoride is reviewed. Attention Is given to its analysis in silicate materials, especially rocks and minerals; developments in method are traced from the first work, published in 1816, to the present. Special emphasis is given for the period 1951 through 1959, Important methods and other pertinent work and findings being summarized chronologically by subject and without regard to systematic exposition of contradictions or their explanation: The summaries of some work are nevertheless detailed enough to make manifest some contradictions. Conflicting findings on the precision in the visual thorium titration and on the visual ratings of various Indicators for this titration are thought merely to reflect Individual skill and choice rather than superiority of method and (or) indicator. A strong case is thereby made for Instrumental methods of analysis. A method developed to determine fluoride in metamorphic rocks Is described. A sodium peroxide decomposition, followed by a precipitation of silica and alumina and a steam distillation of the filtrates, gave quantitative isolation of fluoride. Aliquot8 of distillate were titrated spectrephotometric ally in perchloric acid-sodium sulfanilate buffered solution in the presence of sodium 2-(p-sulfophenylazo)-1,8-dlbydroxy- naphthalene-3,6-dlsulfonate (SPADNS). The method was tested with synthetic rocks and soda feldspar, to which known amounts, ) of fluoride were added. Aliquots were titrated spectrephoto metrically to within one microgram of fluoride. Results with G-l and w-1 are compared with analyses reported by others. Replicate results for ten raetamorphic rocks indicate a maximum proportional deviation of 2.0 percent. However, further replications have shown the error is somewhat greater, although nevertheless favorable: The standard deviation for 122 replicate determinations is +0.0014. percent fluoride for rocks having up to 0.09 percent fluoride. The applicability of the perchloric acid-sodium sulfanilate buffer for the thorium titration is examined. It is felt it is superior to the monochloracetlc acid buffer commonly used. The results of an Interlaboratory standardization program conducted to check the validity of several methods and the comparability of results and to suggest possible superior methods of analysis are presented. Agreement is not favorable when all results are compared; possible matrix effects for pyrolltic and chemical methods are considered. Four of the eight procedures employed show favorable agreement, aspects of which are studied. Work, carried out on synthetic solutions containing the elements commonly analyzed in the "main portion" of silicate rock analysis, is given and discussed. Minor modifications, mainly in technique, of the generally accepted methods are detailed. An important improvement in the examination of the calcium oxide and magnesium pyrophosphate precipitates for impurities such as strontium in the former and calcium and manganese In the latter has been developed. Considerable attention Is given to the problem of determining "blanks" for this part of silicate rock analysis.