Determining the ~(17)O/~(16)O ratio of water using a water-CO_2 equilibration method: Application to glacial-interglacial changes in ~(17)O-excess from the Dome Fuji ice core, Antarctica

摘要

Recentw studies show that oxygen three isotope measurement (~(16)O, ~(17)O, and ~(18)O) of water provides additional information for investigating the hydrological cycle and paleoclimate. For determining the ~(18)O/~(16)O value of water, a conventional CO_2-water equilibration method involves measurement of the ratios of CO_2 isotopologues which were equilibrated with water. However, this long-established technique was not intended to measure the ~(17)O/~(16)O ratio, primarily because the historic ion correction scheme does not allow for possible deviations from a fixed (and mass-dependent) relationship between ~(17)O/~(16)O and ~(18)O/~(16)O isotope ratios. Here, we propose an improved method for obtaining the ~(17)O/~(16)O isotope ratio of fresh water by the equilibration method and measurement of the 45/44 CO_2 ion abundance ratio. Equations which we formulated for ~(17)O/~(16)O measurement have two features: first, instead of absolute isotope ratio (R), all equations are formulated in δ values, measured by isotope ratio mass spectrometry. Second, we include two " assigned" δ values of water standards in the equations, because the δ~(18)O are commonly measured against two working standards to normalize the span of the δ scale. This approach clarifies that the contribution from ~(17)O (~(12)C~(16)O~(17)O~+) to the molecular ion current at mass-to-charge ratio m/z 45 signal depends not on the absolute ~(13)C/~(12)C ratio, but on the relative δ~(13)C differences between the working standards and the sample. The pH value of water affects δ~(17)O estimation because δ~(13)C of CO_2 was changed in the water-CO_2 system. We reevaluated this effect using a set of equations, which explicitly includes CO_2 partial pressure effect on pH value. Our new estimation of pH effect is significantly smaller than previously reported value, but it does not alter the main conclusions in the previous study. The method was verified by δ~(17)O measurements of an international standard reference water (GISP) provided by the IAEA. We applied the method to investigate ~(17)O-excess of the ice core drilled at the Dome Fuji station, Antarctica. A total of 1320 samples from a 130m section around Marine Isotope Stage 9.3 (～330,000 years before present) were measured. The error of a measurement for δ~(17)O is 0.175‰ and that of ~(17)O-excess is 184per meg. Although these analytical uncertainties hampered accurate estimation of the changes in ~(17)O-excess, the averaged data indicate that ~(17)O-excess around MIS 9.3 was higher than during the subsequent glacial period. This approach can be applied only to fresh water samples, and additional improvements will be needed to measure samples which contains significant amount of carbonate minerals.