MICHAEL I. BIRD,* FREDERICK J. LONGSTAFFE, and WILLIAM S. FYFE. Department of Geology, University of Western Ontario, London, Ontario N6A 5B7, ...
Geochimica et Cosrnochimica Acta Vol. 57, pp. 3083-3091 Copyright © 1993 Pergamon Press Ltd. Printed in U.S.A.
0016-7037/93/$6.00 + .00
Oxygen-isotope fractionation in titanium-oxide minerals at low temperature MICHAEL I. BIRD,* FREDERICK J. LONGSTAFFE,and WILLIAM S. FYFE Department of Geology, University of Western Ontario, London, Ontario N6A 5B7, Canada (Received May 12, 1992; accepted in revised form February 11, 1993)
Abstract--Oxygen-isotope results for natural and synthetic Ti-oxide minerals are presented. Rutile-water oxygen-isotope fractionation factors ( a ) of 1.0061 _+ 0.0006 ( l a ) at 22°C and 1.0030 + 0.001 (ltr) at 50°C have been obtained for synthesis experiments under a variety of conditions. These results are in close agreement with recent, semiempirical predictions and with values determined using naturally occurring rutile. Nevertheless, the true fractionation factors may lie toward the upper bounds of the error limits, given the possibilities that ( i ) isotopic equilibrium may not have been achieved in some experiments and that (2) hydration water may not have been removed completely from some samples prior to analysis. Anatase-water oxygen-isotope fractionation factors of 1.0087 _+0.0015 ( 1a) at 22°C and 1.0049 -+ 0.0005 ( 1a) at 50°C were obtained from synthesis experiments in which sulphate was not added to the starting solutions. Higher values of a for anatase-water vs. rutile-water are in accord with predictions based on theoretical and empirical considerations. However, fractionation factors increased dramatically to maxima of 1.0143 at 22°C and 1.0081 at 50°C for anatase-water syntheses in which sulphate ions were added to the starting solutions. The large increase in the 6180 values ofanatase produced in such solutions may result from donation of, or exchange with, high-180 sulphate oxygen during formation of Ti-O-OH-SO4 complexes. The range of fractionation factors obtained from the syntheses is mirrored by the range obtained for natural samples. This similarity suggests that the isotopic composition of anatase is controlled not only by its temperature of formation and the isotopic composition of associated water, but also by the mechanism of formation. Moreover, in cases where rutile formed by inversion of less stable anatase, rutile appears to at least partially inherit its isotopic composition from the precursor polymorph. However, coprecipitated quartz and rutile (formed directly, that is, in the absence of an anatase inversion) has potential as an oxygen-isotope geothermometer for low-temperature environments. INTRODUCTION
1979; AGRINIER, 1991 ). However, despite their ubiquitous occurrence in the low-temperature environment, there are few published oxygen-isotope analyses for rutile or anatase formed at temperatures less than 300°C. The increment method developed by SCH~3TZE(1980) to calculate oxygen-isotope fractionation in silicate minerals has recently been modified by ZHENG ( 1991 ) to include metal oxides. From such calculations, ZHENG ( 1991 ) has reported the following equation for the rutile-water oxygen-isotope fractionation between 0 and 1200°C:
TITANIUM-OXIDEMINERALSare widely distributed in a variety of geologic environments. Rutile a n d / o r anatase are common accessory minerals in eclogites (DESMONSand O'NE1L, 1978 ), hydrothermally altered rocks (PUTNIS and WILSON, 1978; YAU et al., 1987; RICHARDSet al., 1988; LEITCH and FEENAN, 1989), and diagenetically altered clastic sedimentary rocks (IXER et al., 1979; CORLETT and MCILREATH, 1974; MORAD, 1989). In the low-temperature environment, rutile and anatase are produced from weathering of primary Ti-bearing minerals such as ilmenite, biotite, and sphene. Ti-oxides are widely reported from bauxites (HARTMAN, 1959), laterites (ANAND and GILKES, 1984), and a variety of soils (SHERMAN, 1953; BAIN, 1976). Anatase is common in sedimentary kaolinite deposits (SAYIN and JACKSON, 1975; WEAVER, 1976) and silcretes (MILNES and HUaq'ON, 1974; SUMMERFIELD, 1983). High-grade Ti-oxide ores are formed by weathering of ilmenite-rich beach sands to rutile (GREY and REID, 1975) and perovskite-rich parent rocks to anatase (SOUBIES et al., 1990). The oxygen-isotope systematics of rutile have been the subject of several experimental and empirical studies at high temperatures (VOGEL and GARLICK, 1970; ADDY and CARLICK, 1974; DESMONS and O'NEIL, 1978; MATTHEWS et al.,
1031n
O~rutile.water
= 3.45 × 106/T 2 - 10.60 × 1 0 3 / T + 2.55.
(1)
The slope of this curve below ~ 2 0 0 ° C is very low in comparison to most other minerals formed at low temperatures. This behaviour suggests that oxygen-isotope values for rutile/ anatase and a coexisting mineral that exhibits strongly temperature-dependent fractionation could be paired, hence providing a geothermometer of considerable utility in weathering and diagenetic environments. A suitable mineral which commonly occurs with rutile/anatase is quartz. The temperature resolution of a quartz-rutile/anatase pair arising simply from analytical reproducibility of 6180 values could be as good as ~ _ 2 ° C at 25°C. The 6180 values for a suite of natural and synthesized anatase/rutile samples are reported in this study with a view to establishing their oxygen-isotope fractionation tendencies at low temperatures (
