Skip Navigation


The Brown Group

Characterization and Quantification of Volatile Disinfection By-products Present in Drinking Water and Air by Capillary Membrane Sampling-gas chromatography-mass spectrometry

Previous research has shown that coupling a supported capillary membrane sampling device (SCMS) to a GC-MS is possible though preliminary results yielded higher detection limits than the electron capture detector method (Emmert et al., 2004; Emmert et al., 2007). This is to be expected, because an electron capture detector is more selective for halogen containing compounds. The development of the CMS allowed for on-line monitoring and easier sample analysis than the SCMS, since it had a “tube-within-a-tube” design. The CMS device is constructed of two stainless steel tee unions and a length Tefzel® tubing with silicone membrane tubing inserted through it. Drinking water or air flows through the CMS device, thus VOCs will permeate through the silicone membrane and into an inert gas stream (helium). The VOC laden gas stream is then flowed through a desiccator vessel to remove excess moisture and to concentrator that contains an adsorbent trap. The VOCs are trapped at ambient temperature then desorbed at 250º C and flowed to a GC-MS through a heated transfer line for analysis.

Modeling the Permeation Process of Volatile Compounds through Cylindrical Membranes

Further optimization of the CMS device for use in GC methods would benefit from knowing an approximate point where Fick’s second law (non-steady state) gives way to Fick’s first law (steady-state). Another issue that needs attention is eliminating the memory effect associated with membrane materials. All parameters should be taken into account (e.g. membrane thickness, inner/outer flow-rates, temperature, ect.), yet this is a lengthy process. Fundamental experiments will involve varying one parameter while several others are held constant to determine boundary layer conditions. In previous research, a “time-lag” method for diffusion was used for calculating permeation fluxes; however, a mathematical model could provide a good comparison to these results.
The initial goals of this research are to optimize the CMS device for both liquid/gas and a liquid/liquid extraction VOCs by comparing experimental values obtained from a CMS-GC to mathematical models. Ultimately, a better understanding of the overall diffusion process for this system can be obtained.

Emmert, G. L., Cao, G., Geme, G., Joshi, N., and Rahman, M. (2004). Methods for Real-Time Measurement of THMs and HAAs in Distribution Systems. Denver, Col.: AwwaRF and AWWA.

Emmert, G. L., Brown, M.A., Simone, P.S., and Geme. G. (2007). Methods for Real-Time Measurement of THMs and HAAs in Distribution Systems II. Denver, Colo.: AwwaRF and AWWA.