Exploring intra-crystalline B-isotope variations in mixed-layer illite-smectite

The isotopic composition of boron in illite-smectite (I-S) can be important for monitoring fluid/rock interactions in sedimentary basins. Boron substitutes for Si during reaction of smectite to illite and can preserve information about paleofluid B-isotopic composition. Boron is enriched in oilfield brines, therefore the isotopic composition of those brines may be recorded during illitization and represent a monitor of hydrocarbon maturity and migration. We re-examined previously published experimental results on B-isotope fractionation between I-S and water. By separating B from two crystallographic sites of I-S (tetrahedral and interlayer), we found differences in the dB that might be used as a single-mineral geothermometer. Boron incorporation in I-S follows a non-linear kinetic pathway. Maximum interlayer-B incorporation occurs during R1-ordering. R3-ordering approaches equilibrium with expulsion of interlayer-B leaving only tetrahedral layer-B. The important discovery is that tetrahedral layer dB does not change between R1 and R3 ordering. Boron substitutes in an equilibrium ratio early in the crystallographic reordering of I-S. Natural I-S samples were tested from Gulf Coast mudstones, increasingly illitized with burial depth. Diagenetic reaction kinetics differ from hydrothermal experiments, but still reveal large dB differences (up to 40‰) between the interlayer and tetrahedral layer. Interlayer dB decreases with increasing temperature and illitization. We propose that interlayer dB values represent metastable equilibrium, whereas tetrahedral layer B represents a temperature-dependent equilibrium. If this is true, then the B-isotope geochemistry of I-S can be used to determine paleotemperatures and monitor the influence of hydrocarbons on pore fluids associated with diagenetic I-S. WILLIAMS AND HERVIG: INTRA-CRYSTALLINE BORON ISOTOPES IN ILLITE-SMECTITE 1565 changes in the bulk mineral B-isotope ratios as equilibrium was approached between the minerals and fluids (Williams et al. 2001a). The samples showed increasing degrees of illitization with time at temperatures of 300 and 350 ∞C, similar to results of Whitney and Northrup (1988) who studied O-isotope variations during illitization. These samples were used to test a method of separating interlayer B from tetrahedral layer B in an attempt to compare the isotopic compositions of each site. This allows us to examine more closely the mechanism of B incorporation into I-S. The method was then applied to natural I-S from the Gulf Coast sedimentary basin, with promising preliminary results showing a correlation between the intracrystalline B-isotope differences and temperature. EXPERIMENTAL METHODS The quantity of B incorporated in the bulk I-S, and its isotopic composition, were measured using secondary ion mass spectrometry (SIMS). The instrument (Cameca IMS 3f) used a 12.5 V O primary beam to bombard the silicate sample held at +4500 V in a vacuum of <10 Torr. The primary beam was defocused to 30–50 mm resulting in a total depth of analysis of ~3 mm. This method samples many <2 mm clay particles during a single analysis. Analyses on more than three (up to 30) spots within a single sample were averaged to get acceptable statistical errors. The instrumental conditions for B-isotope analyses require closing down the entrance slits to the mass spectrometer to separate the BH from B. This method effectively separates the hydride species (Williams et al. 2001a) at a mass resolution of ~2500 (m/Dm). Silicate glass standards were used to construct a calibration curve for B content of silicates by measuring the ratio of B to Si. The less-abundant mass of Si was used so that both masses are measured using an electron multiplier, thus avoiding discrepancies inherent in switching between a faraday cup and electron multiplier. The results using this calibration (Hervig 1996) are reproducible with errors of <10% at the level of a few hundred ppm B. Errors increase to 20% as concentrations decrease to 1 ppm. Clay minerals require special preparation for isotope ratio analysis because of the different adsorption and exchange sites of the layered silicate (Fig. 1). Surface-adsorbed B must first be removed as it can represent 10–20% of the total-B composition (Spivack et al. 1987). Hingston (1964) showed that by washing clay minerals in mannitol (a polyhydric alcohol that complexes B), adsorbed B is effectively removed from clay-mineral surfaces. We added 50 mL of a 1.82% mannitol solution to 200 mg of clay, ultrasonified to disperse the clay minerals, and then shook the suspension for 2 or more hours to remove adsorbed B. The samples were then washed in triplicate using amberlite resin exchanged de-ionized water. A small drop of the clay/water slurry was dried onto a B-free glass slide and gold-coated for analysis by SIMS. Figure 2 shows B-isotope ratio analyses of the starting material SWy-1 smectite before and after washing in mannitol as described above. There is a change of –15‰ in the bulk analysis after washing off the adsorbed B, showing how important it is to remove B contaminants that may interfere with interpretations. A standard was mounted on each glass slide containing the samples to be analyzed. The standard we chose was IMt1 illite. Analyses of this standard have been made by two separate laboratories using positive and negative thermal ionization mass spectrometry (TIMS) and by ICP atomic emission spectroscopy. The TIMS analyses agreed within 0.5‰ giving an average dB of –9.1 ± 0.3 and the B-content is 240 ± 10 ppm (Williams 2000). The standard was checked before and between analyses of each sample to correct for any instrumental drift. The mineralogical changes that occurred as smectite reacted to illite were monitored by X-ray diffraction (XRD) of oriented clay mounts. The analyses were performed on a FIGURE 1. Schematic diagram showing the crystallographic positions of mixed-layer I-S that may contain B. Boron in the silicate framework goes only into tetrahedral layers. The molecular form of B in the interlayer is unknown. FIGURE 2. Results of multiple SIMS analyses of the B-isotope composition of reactant SWy-1 smectite, before and after washing in mannitol to remove adsorbed B. WILLIAMS AND HERVIG: INTRA-CRYSTALLINE BORON ISOTOPES IN ILLITE-SMECTITE 1566 Siemens D-5000 diffractometer using CuKa radiation. The procedure used was that described in detail in Moore and Reynolds (1997).