Routine high-precision analysis of triple water-isotope ratios using cavity ring-down spectroscopy

Abstract

Rationale: Water isotope analysis for δ(2) H and δ(18) O values via laser spectroscopy is routine for many laboratories. While recent work has added the δ(17) O value to the high-precision suite, it does not follow that researchers will routinely obtain high precision (17) O excess (Δ(17) O). We demonstrate the routine acquisition of high-precision δ(2) H, δ(17) O, δ(18) O, d, and Δ(17) O values using a commercially available laser spectroscopy instrument.

Methods: We use a Picarro L2140-i cavity ring-down spectroscopy analyzer with discrete liquid injections into an A0211 vaporization module by a Leap Technologies LC PAL autosampler. The instrument is run in two modes: (1) as recommended by the manufacturer (default mode) and (2) after modifying select default settings and using alternative data types (advanced mode). Reference waters analyzed over the course of 15 months while running unknown samples are used to assess system performance.

Results: The default mode provides precision for δ(2) H, δ(17) O, δ(18) O, d, and Δ(17) O values that may be sufficient for many applications. When using the advanced mode, we reach a higher level of precision for δ(2) H, δ(17) O, δ(18) O, d, and Δ(17) O values (0.4 mUr, 0.04 mUr, 0.07 mUr, 0.5 mUr, and 8 μUr, respectively, where mUr = 0.001 = ‰, and μUr = 10(-6) ) in a shorter amount of time and with fewer syringe actuations than in the default mode. The improved performance results from an increase in the total integration time for each injected water pulse.

Conclusions: Our recommended approach for routine δ(2) H, δ(17) O, δ(18) O, d and Δ(17) O measurements with the Picarro L2140-i is to make use of conditioning vials, use fewer injections (5 per vial) with greater pulse duration (520 seconds (s) per injection) and use only the first 120 s for δ(2) H measurements and all 520 s for δ(17) O and δ(18) O measurements. Although the sample throughput is 10 unknowns per day, our optimal approach reduces the number of syringe actuations, the effect of memory, and the total analysis time, while improving precision relative to the default approach. Copyright © 2016 John Wiley & Sons, Ltd.