A statistical analysis of IRMS and CRDS methods in isotopic ratios of 2H/1H and 18O/16O in water

Abstract

Quantitative information about the variation in natural isotopic abundances in water is of great importance in a variety of fields. Due to the wide range of applications and types of samples, it is necessary that isotopic analyses have precision, accuracy and reproducibility. The present study compares the techniques of cavity ring-down spectroscopy (CRDS) and isotope-ratio mass spectrometry (IRMS) for the determination of the isotopic ratios of 2H/1H and 18O/16O in water in the two secondary standards, denoted PB3 and PB4, and in a certified material, GISP, Greenland Ice Sheet Precipitation, used as a quality tester of such measurements. The traditional method for measuring isotopic ratios is IRMS. Because of the nature of the molecule, the samples are not introduced directly into the mass spectrometer. Instead, the water is chemically converted to CO2 and H2. The other technique, CRDS, is a system of laser absorption that has great potential for the detection of atomic and molecular species with high sensitivity by measuring the light absorption ratio as a function of time, confined within an optical cavity of high finesse. In this technique, the water sample is converted into steam without undergoing conversion processes. Parametric (test T) and nonparametric (Wilcoxon) statistical tests were performed to compare the results obtained in the system, and CRDS and IRMS are from the same population. The values of the isotopic abundances of the two secondary standards [PB3, δD = − 1.9 ± 0.4 (‰) and δ18O = − 2.19 ± 0.24 (‰) and PB4, δ2H = − 71.4 ± 0.4 (‰) and δ18O = − 10.08 ± 0.19 (‰)] were determined with accuracy. For the certified standard GISP, values of δ2H = − 189.3 ± 0.5 (‰) and δ18O = − 24.69 ± 0.20 (‰) were obtained. Both techniques have factors that interfere with the accuracy of the measurements and require corrections. Comparing the results revealed that there was a greater accuracy for measurements with CRDS and greater precision for IRMS. However, the results are within the tolerance range of 0.2‰ for δ18O and 2.0‰ for δ2H in isotope hydrology.