You are here:
MEETING IN CHARLOTTE: SOLID-PHASE EXTRACTION OF 35 DBPS WITH ANALYSIS BY GC/ECD AND GC/MS
Citation:
CHINN, R., T. LEE, S. KRASNER, M. DALE, S. D. RICHARDSON, J. G. PRESSMAN, T. F. SPETH, R. J. MILTNER, AND J. E. SIMMONS. MEETING IN CHARLOTTE: SOLID-PHASE EXTRACTION OF 35 DBPS WITH ANALYSIS BY GC/ECD AND GC/MS. Presented at 2007 Water Quality Technology Conference, Charlotte, NC, November 04 - 08, 2007.
Impact/Purpose:
(1) Use a toxicity-based approach to prioritize and identify DBPs that show the greatest toxic response. (2) Comprehensively identify DBPs formed by different disinfectant regimes for the 'Four Lab Study'. (3) Determine the mechanisms of formation for potentially hazardous bromonitromethane DBPs.
Description:
An analytical method for 35 disinfection by-products (DBPs) was developed for a U.S. Environmental Protection Agency health effects study. A toxicological evaluation was conducted on drinking water that was scaled-up (using reverse osmosis) by concentrating the total organic carbon (TOC) from a treated surface water by ~130-fold, adjusting the bromide concentration to re-establish the natural TOC to bromide ratio, and subjecting the concentrate to chlorination, using a similar chlorine dose to TOC ratio as would be used in conventional treatment. This concentrated water presented analytical challenges, which were resolved by merging two methods, which provided excellent quality control data while increasing the efficiency of the analysis and offering confirmation data for 19 of the target analytes.
In previous studies, the analysis of a broad suite of DBPs was accomplished using two analytical methods, micro liquid-liquid extraction (MLLE) with gas chromatography/electron capture detection (GC/ECD), and solid-phase extraction (SPE) with GC/mass spectrometry (MS). Due to the complex sample matrix in this study, attempts to use conventional MLLE produced thick layers of emulsions that made the separation of a clean solvent layer very difficult. The analysis was labor-intensive and produced low analyte recoveries. However, sample preparation by SPE using a styrene divinylbenzene polymer cartridge proved successful. An automated SPE system was used for all sample preparations. Because GC/ECD offered some detection sensitivity advantages over GC/MS, the SPE extracts were split into two separate vials and both types of instruments were used for sample analyses over a six-month period.