质谱与光谱技术高精度测试水中三氧同位素的对比

High-precision measurements of oxygen isotopic compositions in water are of great importance for understanding regional and global hydrological cycles and historical hydroclimate change. Several methods have been developed to determine the triple oxygen isotopic proxy (¹⁷O-excess) in water. However, these methods are not applicable because of poor precision or poisonous compound utilisation. Recently, a method based on CO₂-H₂O equilibration followed by an isotopic exchange between CO₂ and O₂ catalysed by hot platinum, was developed to measure water ¹⁷O-excess with high precision (less than 10 per meg, 1 per meg=10^-6). It has also been demonstrated that this method has the capability to obtain sufficiently precise water ¹⁷O-excess values, by performing more injections per sample and/or giving longer integration time per injection. Here, we describe the water ¹⁷O-excess measurement method using cavity ring-down spectroscopy (Picarro L2140-i) and mass spectrometry (MAT 253) devices, the latter of which comprises CO₂-H₂O equilibrium and CO₂-O₂ equilibration systems that produce pretreated O₂ for measurement. The excellent agreement between the oxygen isotopic compositions of international standard waters, derived from both methods, and the recommended or previously reported values, confirms the reliability and accuracy of the two methods. Each method has advantages and disadvantages in terms of efficiency, sensitivity to environmental disturbance, sample quantity demand, and so on. Repeated measurements of Greenland Ice Sheet Precipitation (GISP) were performed to check the reliability of the data normalisation method to the Vienna Standard Mean Ocean Water-Standard Light Antarctic Precipitation (VSMOW-SLAP) scale. For the Picarro method, δ¹⁸O_VSMOW-SLAP and δ¹⁷O_VSMOW-SLAP values equalled -24.80‰ ± 0.03‰ and -13.15‰ ± 0.02‰, respectively (1σ), while for the mass spectrometry method, δ¹⁸O_VSMOW-SLAP and δ¹⁷O_VSMOW-SLAP values equalled -24.82‰ ± 0.04‰ and -13.16‰ ± 0.03‰, respectively (1σ). The International Atomic Energy Agency (IAEA) reported a GISP δ¹⁸O_VSMOW-SLAP value of -24.76‰ ± 0.09‰ (1σ), and the average GISP δ¹⁷O_VSMOW-SLAP value from previous studies is -13.12‰ ± 0.06‰. ¹⁷O-excess values were calculated from oxygen isotope ratios, and we obtained an average value of 26 ± 8 and 21 ± 6 per meg (1σ) for the Picarro and mass spectrometry methods, respectively, both of which are close to the average of previously reported values (22 ± 11 per meg). Collectively, these results demonstrate that both methods can be used to obtain accurate water triple oxygen isotopes with high precision. Moreover, both methods reproduce the previously reported 26 per meg difference between USGS 45 and USGS 47 ¹⁷O-excess values. This demonstrates the ability of the methods to distinguish the differences in precipitation, atmospheric vapour, cave drip water, and ice core water samples. Triple oxygen isotope measurement methods were applied to determine summer and winter precipitation ¹⁷O-excess values from Nanjing, Xi'an, and Urumqi, which exhibit distinct spatiotemporal differences. Although the precipitation δ¹⁸O values in Nanjing and Xi'an were significantly greater in winter and lower in summer, seasonal changes in ¹⁷O-excess appear complex, and thus more data are required. The precipitation δ¹⁸O values from Urumqi show the opposite pattern, with higher ¹⁷O-excess values in winter. Additional measurements are required to confirm this pattern.