Final Report: Membrane Introduction Mass Spectrometry Studies of Halogenated Cyano Byproduct Formation in Drinking Water

EPA 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 Amount: $334,666
RFA: Drinking Water (1999) RFA Text |  Recipients Lists
Research Category: Drinking Water , Water

Objective:

Description and Objective of Research:

Cyanogen halides are candidate DBPs that could become regulated in future Disinfection By-Product Rules, however, there is insufficient information available about the factors and precursors that promote their formation in drinking water. In this project, studies were conducted to 1) determine the most important amino acid compounds that serve as sources of cyanogen halides, 2) characterize the stability and decomposition mechanism of cyanogen chloride (CNCl) under disinfection conditions, 3) characterize the kinetics and formation mechanism of CNCl in model solutions of chlorinated amino acids, 4) examine the relative reactivity of amino acids with chlorine, and 5) assess the CNCl formation potential significance of amino acids in a natural surface water.

Summary/Accomplishments (Outputs/Outcomes):

Summary of Findings:

The research conducted in this study provides a new fundamental understanding for the phenomenological observations of CNCl occurrence in drinking water. The findings are useful in assessing the likely impact of shifts by utilities from chlorination to chlorination/postchloramination disinfection processes. The research was also successful in establishing a major CNCl precursor in chlorinated surface water, glycine, and in identifying some critical characteristics of the as yet unidentified pool of CNCl precursors. Thorough studies of the formation mechanism of CNCl from glycine, given its importance, were completed, and these studies establish the disinfection conditions that could affect CNCl yields.

Stability of CNCl. To fully establish the sources and conditions that promote CNCl formation in drinking water, it was necessary to first consider the stability of CNCl in the presence of free chlorine and chloramines and develop sufficient kinetic parameters so that CNCl decay processes could be modeled. This mechanistic study was published in a paper (Na and Olson, Environ. Sci. Technol., 2004, 38:6037-6043). The kinetic models were utilized later in studies of the formation of CNCl from glycine. In the presence of hypochlorite ion, it was shown that CNCl is catalytically hydrolyzed to cyanate. CNCl was stable, on the other hand, in the presence of monochloramine. The rate of CNCl hydrolysis due to hypochlorite was described by the following rate law,

d[CNCl] / dt = -kOCl[OCl¯][CNCl]

in which the second-order rate constant, kOCl¯ = 121 / M s (at 25°C). Simulations using this kinetic model were performed to assess the implications of CNCl hydrolysis under a range of conditions relevant to drinking water treatment (varying chlorine doses, pH, and temperature). Based on this analysis we concluded that extensive hydrolysis of CNCl would be expected at neutral pH and warmer temperatures with modest free chlorine residuals, but that CNCl becomes considerably more stable at colder temperatures, (e.g., half-life = 107 min, with 0.5 mg Cl2/L, 15°C).

The CNCl stability differences in these two disinfectant systems have important implications for utilities that utilize combined Cl2/post-chloramination systems. By combining residual free chlorine with ammonia the stability of any CNCl formed during chlorination is enhanced relative to chlorination alone. Previous utility surveys by Krasner et al. have reported an association of CNCl occurrence in finished water and these types of combined Cl2/chloramine facilities. The findings in this study provide in part some mechanistic explanations for this phenomenological association.

Amino Acid Screening Studies. Chlorination experiments with 17 DNA-derived amino acids indicated that only glycine was a significant CNCl precursor, in which the yields of CNCl from all other amino acids were less than 0.6%. These results were discussed in a paper published in Environmental Science & Technology, and motivated the mechanistic study of CNCl formation from glycine.

Mechanistic Study of CNCl Formation from Chlorinated Glycine. The formation kinetics of CNCl formation upon the reaction of free chlorine with glycine was examined using a real-time analysis method known as membrane introduction mass spectrometry (MIMS). The findings appeared in a paper published in Environmental Science & Technology (Na and Olson, 2006). The proposed reaction mechanism was consistent with a rapid dichlorination of glycine followed by the rate-limiting decomposition of dichloroglycine, as illustrated in Scheme 1. The rate of dichloroglycine decomposition and the yield of CNCl were found to be pH dependent, and two competing dichloroglycine decomposition pathways shown in Scheme 1 were invoked to account for these observations.

At pH > 6, which represents most drinking water chlorination conditions, glycine-nitrogen was 100% converted to CNCl-N with excess chlorine (greater than 2:1 molar ratio of free chlorine:glycine). The competing dichloroglycine decomposition pathway becomes significant at pH < 6, however, resulting in the conversion of glycine to other intermediates and products. Spectroscopic evidence (via MIMS) for the formation of chloromethylimine by this low pH decomposition pathway was also obtained.

Over the pH range of 6 to 8, and in the presence of excess chlorine, the rate of CNCl formation by the reaction of glycine and free chlorine was mathematically modeled with the following rate law:

d[CNCl]f / dt = k*2[Cl2-Gly]T,0 exp(-k*2t),

however, to model the actual CNCl concentration, the kinetics of CNCl hydrolysis must be considered. Since chlorine is nominally in excess relative to the naturally occurring concentrations of amino acids in chlorine disinfection processes, the actual observed concentrations of glycine-derived CNCl should be significantly less than the glycine content due CNCl hydrolysis.

The half life for CNCl formation at 25°C is approximately 4 min, and hence glycine conversion to CNCl during chlorination should be nearly complete in a typical chlorine contact chamber. The mechanism in Scheme I suggests that there are few ways to prevent the conversion, however, except to remove glycine before chlorination. In other words, glycine is rapidly chlorinated and the subsequent decomposition of dichlorinated glycine to form CNCl is independent of pH (above pH 6) and chlorine concentration.

If the mechanism of Scheme I is analogously extended to other aliphatic amino acids, the expected products would be nitriles, rather than CNCl. This prediction was in part consistent with the findings of the screening studies described above, in which the 16 other aliphatic amino acids were chlorinated, and no CNCl was detected.

CNCl Precursor Analysis in Chlorinated Huron River Water (HRW). Since glycine was identified as the most significant amino acid CNCl precursor in well-defined model solutions, experiments were designed to examine glycine’s significance as a CNCl precursor using natural river water samples. These samples were collected from the Huron River, which drains a predominantly agricultural basin. The approach was to determine the amino acid contents of the water samples, utilize chlorine doses that would not significantly hydrolyze the CNCl formed, and measure the yield of CNCl after chlorination. Calculations of the glycine-derived CNCl fraction were made assuming the yields determined in the previously described model study. These studies were summarized in a paper published in Environmental Science and Technology (Lee et al., 2006).

Estimates of the fraction of glycine-derived CNCl ranged from 42-45% of the measured CNCl after 60 min. The findings are unique, in that while other precursors are important, a single precursor, glycine, is responsible for such a significant fraction of CNCl. The kinetics of CNCl formation in the natural sample was also compared with the previous mechanistically determined kinetic models of CNCl formation from glycine. Both rapid and slow CNCl formation pathways were evident in the kinetic studies of its formation in Huron River water. The more rapidly formed CNCl fraction could be predominantly due to glycine as illustrated in Figure 1, while other unidentified precursors may contribute the more slowly formed CNCl. Glycine-derived CNCl, in other words would form nearly completely in a disinfection contact chamber, while unidentified precursors may contribute more importantly to longer time-scale formation of CNCl, say in the distribution system.

From literature suggestions, two types of CNCl precursors, proteinaceous compounds and humic matter, were hypothesized as possible sources of the non-glycine derived CNCl produced after chlorination of HRW. Although these components are complex, it was reasoned that they might be distinguished by applying a non-destructive separation technique, known as immobilized metal ion affinity chromatography (IMAC). This analytical procedure was originally developed to isolate and purify proteins in biochemistry applications. The technique relies on the use of metal-loaded resins with a high affinity to bind proteins. More recently these resins were used by geochemists to fractionate strongly binding proteins and weakly binding humic matter in natural water samples. By processing HRW samples with an IMAC resin system in our laboratory, it was possible to demonstrate that the unidentified CNCl precursor pool has a much weaker affinity for IMAC resins than glycine and hence these precursors are more likely to be humic in character than proteinaceous. The findings provide an important basis for directing future research efforts to establish sources of cyanogen halide in drinking water.

Figure 1. Cyanogen chloride formation kinetics in a Huron River water sample (data points) with 2.3 mg Cl as Cl2 L-1, pH 8.3. Dashed line is a calculated prediction of glycine-derived CNCl formed based on the glycine content of HRW.

Relative Reactivity of Amino Acids with Chlorine. While amino acids are present in much less concentration than the disinfectant in drinking water, the opposite condition, a stoichiometric excess of total amino acid, may be found once drinking water is consumed or used. Chlorine may be the limiting reagent, for example, in vivo in the human gastrointestinal tract or during food washing and preparation. Under these conditions, the possible formation of cyanogen chloride from glycine would depend on the competitive nature of glycine for chlorine among other amino acids. There was insufficient data in the literature, however, to evaluate glycine’s relative reactivity, however, so experiments were devised in this study to directly address the question. Using competitive kinetic principles and examining the titration of amino acids with chlorine (over ratios of free chlorine:amino acid varying from 0 to 2), it was determined that glycine and proline were the least reactive of 17 amino acids. Amino acids with thiol groups (methionine and cysteine) were the most reactive, while all other amino acids were similar with intermediate reactivity. At an equal ratio of chlorine:total amino acid, only approximately 26% of the amino acids were dichlorinated. Since dichlorination of glycine is necessary to produce CNCl and since glycine is one of the slowest to react with chlorine, it is unlikely that amino acid-derived CNCl can form under these conditions. The results of these relative reactivity experiments have been submitted as a paper to the journal of Environmental Science & Technology.

Conclusions:

Conclusions:

Several fundamental conclusions regarding the specific conditions that lead to cyanogen halide formation and occurrence in drinking water system were established in this research. The most important conclusions are as follows:

  1. Cyanogen chloride is relatively unstable in the presence of free chlorine, but stable in the presence of monochloramine. The decay mechanism was found to be due to hypochlorite ion –catalyzed hydrolysis. Disinfection strategies such as post-chloramination after chlorination, therefore, may serve to stabilize CNCl relative to chlorination alone.
  2. Chlorination studies of seventeen DNA-derived amino acids demonstrate that glycine is the only amino acid to significantly form cyanogen chloride. With excess chlorine, dichlorination of the glycine-N rapidly occurs, which is then more slowly stoichiometrically converted to CNCl.
  3. While the predicted yield of CNCl from glycine with excess chlorine is 100%, observed yields as they would occur during disinfection are less and depend also on the extent of CNCl hydrolysis. With the quantitative kinetic models for both processes that were determined in this project, it is possible to predict the maximum glycine-derived CNCl concentration that might be expected.
  4. Chlorination studies with natural river water samples established the significance of glycine rather uniquely. Based on the concurrent measurements of natural glycine present and the amount of CNCl formed, it was estimated that up to ~45% of the CNCl in chlorinated river water may be glycine-derived.
  5. Based on non-destructive natural organic matter separation techniques, the remaining non-glycine-derived CNCl in the river water was more likely to come from humic matter, rather than protein- or peptide-like precursors.
  6. Although amino acids might be in high concentration relative to chlorine when drinking water is used or consumed, it is unlikely that much glycine-derived CNCl could form under these conditions, due to the uncompetitive nature of glycine kinetically, and the mechanistic requirement that glycine first be dichlorinated to form CNCl.


Journal Articles on this Report : 4 Displayed | Download in RIS Format

Other project views: All 14 publications 4 publications in selected types All 4 journal articles
Type Citation Project Document Sources
Journal Article Lee JH, Na C, Ramirez RL, Olson TM. Cyanogen chloride precursor analysis in chlorinated river water. Environmental Science & Technology 2006;40(5):1478-1484. R828231 (Final)
  • Abstract from PubMed
  • Full-text: ACS-Full Text HTML
    Exit
  • Abstract: ACS-Abstract
    Exit
  • Other: ACS-Full Text PDF
    Exit
  • Journal Article Na C, Olson TM. Stability of cyanogen chloride in the presence of free chlorine and monochloramine. Environmental Science & Technology 2004;38(22):6037-6043. R828231 (2003)
    R828231 (Final)
  • Abstract from PubMed
  • Full-text: ACS-Full Text HTML
    Exit
  • Abstract: ACS-Abstract
    Exit
  • Other: ACS-Full Text PDF
    Exit
  • Journal Article Na C, Olson TM. Mechanism and kinetics of cyanogen chloride formation from the chlorination of glycine. Environment Science & Technology 2006;40(5):1469-1477. R828231 (Final)
  • Full-text: ACS-Full Text HTML
    Exit
  • Abstract: ACS-Abstract
    Exit
  • Other: ACS-Full Text PDF
    Exit
  • Journal Article Na C, Olson TM. Relative reactivity of amino acids with chlorine in mixtures. Environment Science & Technology 2007;41(9):3220-3225. R828231 (Final)
  • Full-text: ACS-Full Text HTML
    Exit
  • Abstract: ACS-Abstract
    Exit
  • Other: ACS-Full Text PDF
    Exit
  • Supplemental Keywords:

    RFA, Scientific Discipline, Water, Environmental Chemistry, Health Risk Assessment, Environmental Microbiology, Drinking Water, alternative disinfection methods, halogenated disinfection by-products, halogenated cyano byproduct formation, monitoring, mass spectrometry studies, exposure and effects, disinfection byproducts (DPBs), exposure, bromate formation, brominated DPBs, chlorine-based disinfection, drinking water distribution system, treatment, DBP risk management, water quality, drinking water contaminants, drinking water treatment

    Progress and Final Reports:

    Original Abstract
  • 2001 Progress Report
  • 2002 Progress Report
  • 2003 Progress Report
  • 2004