INVESTIGATION OF A COMMERCIALLY AVAILABLE HIGH TEMPERATURE CARBON REDUCTION INSTRUMENT FOR THE ANALYSIS OF STABLE OXYGEN ISOTOPES IN SILICATES, OXIDES AND BIOGENIC SILICA

In the past decade, use of biogenic silica has been gaining popularity as a tool for paleoclimate investigations in lakes that lack significant carbonate deposits. Coupled with that is a continuing demand for δ¹⁸O analysis of silicates and oxides for petrologic applications. Fluorination techniques, though producing very accurate and precise δ¹⁸O data, are time-consuming and hazardous to generate a large number of analyses in a short amount of time, especially those required for paleoclimate research. We report here the use of a commercially available oxygen-nitrogen (O/N) analyzer capable of high temperature carbon reduction (HTCR) to produce accurate and precise stable oxygen isotope data for silicates, oxides, and biogenic silica.

Three quartz isotope standards (international standard NBS-28, ARQ, and NCSU) were used to test the accuracy and precision of oxygen isotope analyses. ARQ and NCSU are laboratory standards with δ¹⁸O values established by fluorination. One-mg samples of quartz are wrapped in tin capsules and dropped by an attached autosampler into a graphite crucible within the graphite-electrode furnace of the O/N analyzer. A temperature of ~2500°C is used to drive the reduction of SiO₂ in the presence of excess carbon to produce carbon monoxide (CO) in a continuous flow of He (~250ml/min), which carries the product directly to a stable isotope ratio mass spectrometer. Total trial time for extraction and analysis is 7 minutes.

Good precision and accuracy is observed in results from the quartz standards using HTCR. NBS-28 yields a HTCR precision of 9.6‰ ± 0.12‰ (n=4). ARQ yields HTCR δ¹⁸O of 18.73‰ ± 0.20‰ (n=7) vs. accepted δ¹⁸O of 19.0‰; and NCSU yields HTCR δ¹⁸O of 11.64‰ ± 0.16‰ (n=11) vs. accepted δ¹⁸O of 11.7‰. Oxygen yield in CO produced by the analyzer is calibrated with BaSO₄ and resulted in an average value of 102% quantitative conversion of quartz-oxygen to CO.

These results demonstrate that a HTCR-capable instrument, coupled to a mass spectrometer, can produce both accurate and precise δ¹⁸O data for quartz in very short trial times and without hazardous reagents. Use of such an instrument indicates sample throughput for biogenic silica is comparable to that of carbonates for use in paleoclimate investigations and allows for faster return time on data for petrologic studies.