2001 Progress Report: Membrane Introduction Mass Spectrometry Studies of Halogenated Cyano Byproduct Formation in Drinking WaterEPA Grant Number: R828231
Title: Membrane Introduction Mass Spectrometry Studies of Halogenated Cyano Byproduct Formation in Drinking Water
Investigators: Olson, Terese M.
Institution: University of Michigan
EPA Project Officer: Nolt-Helms, Cynthia
Project Period: August 1, 2000 through August 1, 2003 (Extended to August 31, 2004)
Project Period Covered by this Report: August 1, 2000 through August 1, 2001
Project Amount: $334,666
RFA: Drinking Water (1999) RFA Text | Recipients Lists
Research Category: Drinking Water , Water
Cyanogen halides currently are tentative candidates that could fall under regulation as a disinfection byproduct (DBP) in Stage 2 of the Disinfection/DBP Rule. This research project seeks to: (1) determine the most important amino acid precursor compounds as sources of halogenated cyanogen DBPs; (2) characterize the kinetics and formation mechanism of chlorinated and brominated cyanogen compounds; and (3) model their formation in a natural water sample containing a significant fraction non-humic organic matter.
Initial project activity has focused on screening studies to determine whether certain amino acid precursor compounds are more significant as sources of cyanogen chloride (CNCl), and if so, to develop a smaller set of model compounds with which to further study reaction mechanisms. Because the question of importance is potentially a function of three factors-reactivity, yield, and natural abundance, the screening approach during this initial period has been to consider these factors separately and from a relative perspective, particularly in terms of the former two.
To assess the relative reactivity of amino acids with free chlorine, 17 amino acids were considered using a titration-type approach. Subset mixtures of the amino acids were chlorinated for a fixed period at varying Cl2:total amino acid molar ratios, and residual amino acid concentrations were determined using a fluorescent derivatization method and HPLC chromatographic separation. Three subset mixtures of about six or seven amino acids each were selected to eliminate interferences with products of the reaction that were observed when all amino acids were titrated simultaneously. An example of one of these subset titrations is illustrated in Figure 1.
The results of these titration experiments demonstrate that the largest differences in reactivity with chlorine are between those amino acids that contain sulfhydryl groups, methionine and cysteine, and those that do not. Both methionine and cysteine are titrated first and are significantly more reactive than amino acids that do not have sulfhydryl groups. Despite the reactivity of these two substrates, however, the reaction pathway for the methionine and cysteine is likely to involve oxidation of the sulfhydryl sulfur, rather than reaction at the amine nitrogen. Positive identification and the determination of significant yields of methionine sulfoxide upon the chlorination of methionine during these tests is supporting evidence that the initial reactivity of these compounds does not involve steps that lead to cyanogen chloride formation. Limited data in the literature also suggest that methionine and cysteine are typically less abundant amino acids in natural environments.
Among the remaining 15 amino acids considered, reactivity with chlorine could be characterized in terms of three groups. Basic amino acids (histidine, arginine, and lysine) were somewhat more competitive than a middle group (Ile, Val, Ala, Glu, Asp, Phe, Tyr, Ser, Leu, and Thr), which in turn were somewhat more competitive than glycine and proline. Given the relatively small differences in reactivity between these 15 substrates, however, our results suggest that other factors (yields and abundance) are probably more important in discriminating between the CNCl formation potential of amino acids.
Screening studies to compare the relative yields of CNCl currently are in progress. A continuous flow well-mixed reactor system (CSTR) with online detection of CNCl using membrane introduction mass spectrospectrometry (MIMS) is being used to compare the steady-state concentrations of CNCl. These experiments are being conducted over a range of Cl2:amino acid concentration ratios with the amino acid in excess to minimize effects of CNCl hydrolysis, because the latter reaction is catalyzed by the presence of free chlorine. In this way, relative yields alone can be compared, without the added complexity of CNCl decay. A schematic diagram of the system being used is provided in Figure 2.
Upon completion of these yield test experiments, a manuscript summarizing our screening results will be submitted for journal publication. While limited information of this type is available for a few amino acids, the assessment will be the most comprehensive and systematic to date. The information will help to assess such questions as to whether it is more important to assay for total or individual amino acids to understand CNCl formation potential of a water sample.
Figure 1. Example of subset titration of 7 amino acids after 10 hours at pH 7 as a function of Cl2:Total Amino Acid ratio.
Figure 2. Continuous Flow Reactor with in-line OCl-, CNCl, pH and pI detection.
During the next reporting period we hope to complete the screening test phase of the project and a kinetic study of the CNCl formation mechanism for the amino acid system giving the highest CNCl yields. An additional student with independent funding will join the project and conduct parallel studies of CNBr formation. Chlorination studies of Colorado River Water will be initiated towards the end of the next reporting period. These studies were postponed until the end of the screening studies to take better advantage of their findings.