There are variations of reported isotope enrichment factors of chlorinated organic contaminants in evaporation processes. Trichloroethene （TCE） and tetrachloroethylene （PCE） were chosen to study carbon and chlorine isotope effects during evaporation at different temperatures. Equilibrium vapor-liquid carbon and chlorine isotope effects experiments were also conducted. In the equilibrium liquid-vapor system, the 13C was enriched but 37Cl was depleted in the vapor phase, being consistent with previous results. For evaporation average carbon isotope enrichment factor εc were ＋0.28‰± 0.01‰ for TCE and ＋0.56‰±0.09‰ for PCE at temperature from 20 to 26 ℃. Meanwhile, average chlorine isotope enrichment factor εCl were -1.33‰±0.21‰ for TCE and -1.00‰±0.00‰ for PCE. The results indicate that during evaporation the equilibrium isotope effect attenuates the magnitude of carbon isotope fractionation whereas enhances the chlorine isotope effect. Isotope fractionation during evaporation is determined by both equilibrium and kinetic factors. Chlorine isotope fractionation is influenced by the evaporation rate which is linked to temperature. When using stable isotope to inves- tigate the behavior of chlorinated organic contaminants in groundwater with slow biodegradation rate, the isotope fractionation resulted from evaporation should be taken into consideration. Furthermore, the environment conditions such as temperature are also factors to be considered.
An online method using continuous flow isotope ratio mass spectrometry （CF-IRMS） interfaced with a Gasbench Ⅱ was presented to determine chlorine stable isotope composition. Silver chloride （AgCl） was quantitatively derived from chloride by using silver nitrate （AgNO3）, and then was reacted with iodomethane （CH3Ⅰ） to produce methyl chloride （CH3Cl）. A GasBench Ⅱ equipped with a PoraPlot Q column was used to separate CH3Cl from any other gas species. Finally, chlorine stable isotope analysis was carried out on CH3Cl introduced to the IRMS in a helium stream via an active open split. The minimum amount of Cl used in this method is of the order of 1.4 μmol. Inter-laboratory and inter-technique comparisons show that the total uncertainty incorporating both the precision and accuracy of this method is better than 0.007%. Furthermore, ten seawaters sampled from different locations have a narrow δ37Cl value range from -0.008% to 0.010%, with a mean value of （0.000±0.006）%. This supports the assumption that any seawater can be representative of standard mean ocean chloride （SMOC） and used as an international reference material.
