A comparison of the toluene distillation and vacuum/heat methods for extracting soil water for stable isotopic analysis

Hanford Loam, from Richland, Washington, was used as a test soil to determine the precision, accuracy and nature of two methods to extract soil water for stable isotopic analysis: azeotropic distillation using toluene, and simple heating under vacuum. The soil was oven dried, rehydrated with water of known stable isotopic compositions, and the introduced water was then extracted. Compared with the introduced water, initial aliquots of evolved water taken during a toluene extraction were as much as 30 ‰ more depleted in D and 2.7 ‰ more depleted in 18O, whereas final aliquots were as much as 40 ‰ more enriched in D and 14.3 ‰ more enriched in 18O. Initial aliquots collected during the vacuum/heat extraction were as much as 64 ‰ more depleted in D and 8.4 ‰ more depleted in 18O than was the introduced water, whereas the final aliquots were as much as 139 ‰ more enriched in D, and 20.8 ‰ more enriched in 18O. Neither method appears quantitative; however, the difference in stable isotopic composition between the first and last aliquots of water extracted by the toluene method is less than that from the vacuum/heat method. This is attributed to the smaller fractionation factors involved with the higher average temperatures of distillation of the toluene. The average stable isotopic compositions of the extracted water varied from that of the introduced water by up to 1.4 ‰ in δD and 4.2 ‰ in δ 18O with the toluene method, and by 11.0 ‰ in δD and 1.8 ‰ in δ 18O for the vacuum/heat method. The lack of accuracy of the extraction methods is thought to be due to isotopic fractionation associated with water being weakly bound (not released below 110°C) in the soil. The isotopic effect of this heat-labile water is larger at low water contents (3.6 and 5.2% water by weight) as the water bound in the soil is a commensurately larger fraction of the total. With larger soilwater contents the small volume of water bound with an associated fractionation is not enough to affect the remaining unbound introduced soil water. Pretreatment of the soil to equilibrate the heat-labile water to the test water produced good results for the toluene distillation but not the vacuum/heat extraction method. Vapors collected over the soils also show stable isotopic variations related to soilwater content. These vapors also appear to be in closer equilibrium with the free water, as extracted by the toluene method, than with the originally introduced water; thus, the soil vapors do not appear t