Abstract
A series of experiments were conducted to ascertain the accuracy of D/H determination by applying a TCD-gas chromatography and to improve the methods of sample preparation for the determination. The sensitivity, repeatability, and reproducibility of D/H determination by the TCD-gas chromatography was significantly improved when low D/H hydrogen (LDH), which is supplied from a hydrogen generator fed by water of low deuterium content, was used as the carrier gas . The calibration curve was fitted very well by a straight line in the range of D/H from 100 to 1000 ppm and the least significant difference at confidence level of 0.05 was less than 3 ppm and 8 ppm in the cases of more than 40 and 5 replications, respectively. The reproducibility of D/H determination was also satisfactory, especially in the cases of low D/H samples near natural abundance. When commercial high-quality hydrogen was used as the carrier gas, the linearity of the calibration curve was satisfactory, but both repeatability and reproducibility were somewhat inferior compared with those obtained by using LDH. In the direct distillation of soil samples to prepare water sample for the analysis, the efficiency of vacuum distillation apparatuses, which were proposed previously, was not satisfactory. The efficiency was improved significantly by using a larger distillation flask of more than 30 mm in inner diameter and the accuracy of determination was not influenced by the alteration. A tension lysimeter system was devised and tested to extract soil solution in the field experiments. It was concluded that the system was suitable when the soil water suction was less than about 0.3 bar. The isotope effect together with the extraction was not so significant, but some correction was recommended to compensate the reduction of D/H caused by the isotope effect. In field experiments, tracer solution is diluted rapidly by soil water. The determinations, therefore, have to be performed in the low range of D/H from natural abundance to about 300 ppm and, moreover, high reproducibility is required to evaluate the time series of D/H change in soils. The reduction of background by using LDH as the carrier gas and frequent calibration (two standard samples for a lot of determinations) were recommended to maintain a satisfactory accuracy and reproducibility.
