SOME INSTRUMENTAL EFFECTS IN THE DETERMINATION OF STABLE CARBON ISOTOPE RATIOS BY GAS-CHROMATOGRAPHY ISOTOPE RATIO MASS-SPECTROMETRY

Some sources of instrumental errors in the determination of C-13/C-12 in organic compounds by gas chromatography-isotope ratio mass spectrometry (GC-IRMS) have been investigated. For mass 44 ion beam intensities in the range 1 . 10(-10) to 1 . 10(-8) A, mass-spectrometric pressure effects do not significantly affect data accuracy, thus obviating the necessity of matching sample and reference ion beam intensities for each compound in a complex sample. Data quality is influenced by: (a) the quality of the furnace-tube packing; and (b) the performance of the cryogenic trap. On analysis of large (70 ng) samples, precisions (1-sigma) of less-than-or-equal-to +/- 0.23 parts per thousand were obtained using 0.35-0.5-mm-grade CuO. However, there was a marked deterioration in precision (1-sigma less-than-or-equal-to +/- 0.62) when coarse CuO (0.35-1 mm grade) was tested. Use of fine CuO also permits analysis of smaller samples. Inefficient trapping, and release of water during overnight defrosting leads to erroneously low-delta-C-13-values, accuracies of -1.1 to -0.25 parts per thousand resulting from analysis of 70-ng samples. Accuracy error increases to less-than-or-equal-to 1.86 parts per thousand on analysis of 10-ng samples. Efficient water trapping combined with the improvements in furnace performance results in precisions generally much better than +/- 0.4 parts per thousand and accuracies better than +/- 0.65 parts per thousand for high-molecular-weight compounds (268-338 amu), and very accurate and precise results (better than +/- 0.2 parts per thousand) for lower-molecular-weight compounds (142-173 amu).