Grantee Research Project Results
Final Report: Formation and Stability of Ozonation By-Products in Drinking Water
EPA Grant Number: R826833Title: Formation and Stability of Ozonation By-Products in Drinking Water
Investigators: Weinberg, Howard S.
Institution: University of North Carolina at Chapel Hill
EPA Project Officer: Hahn, Intaek
Project Period: November 1, 1998 through October 31, 2001
Project Amount: $441,261
RFA: Drinking Water (1998) RFA Text | Recipients Lists
Research Category: Drinking Water , Water
Objective:
It generally is perceived that disinfection by-products (DBPs) produced by ozonation alone represent less of a health concern than that associated with the sole use of chlorine for disinfection. It is true that bioassays generally have shown that ozonation by-products, at least those isolated by existing methodologies, cause fewer mutations and other cellular effects than chlorine or chlorine dioxide DBPs. Also, it has been noted that the types of oxidation by-products produced by ozonation of natural waters are often the same as compounds produced by natural oxidation processes in streams, lakes, and reservoirs. The implication is that naturally occurring compounds will be safer than "unnatural" compounds such as trihalomethanes (THMs) and haloacetic acids (HAAs) produced by chlorination. There are, however, fallacies associated with each of these arguments. Short-term bioassays are not at a stage of development to use in relative risk assessments. Also, the preconcentration methods used to obtain extracts for bioassays may not efficiently trap polar by-products from ozonation of natural waters. Finally, natural waters often are mutagenic, so the argument that ozone produces "natural" organics is not encouraging.
When ozone is used to reduce DBP precursor levels, the natural organic matter (NOM) in water is converted to aldehydes, aldo and keto acids, and carboxylic acids. These ozonation by-products, however, represent only about 40 percent of the assimilable organic carbon produced during ozonation. It is not just ozonation by-products that are in question, however. Ozone will not be used in the United States as the terminal disinfectant. Rather, it always will be followed by another disinfectant, probably chlorine or chloramines, and such disinfectant combinations (e.g., from pre- and postdisinfection) can form secondary products. As analytical methodologies only exist for a subset of all by-products, it has not yet been possible to perform comprehensive occurrence studies in the field, and so there is very little information available on the fate of these by-products in distribution systems and their subsequent levels in consumers' drinking waters.
Given the relatively large ozonation by-product fraction that remains unidentified, the objective of this research project was to employ advanced analytical techniques aimed at their positive identification and their recovery and isolation from the aquatic matrix. These techniques then were used with the analysis of known by-products in a study aimed at obtaining occurrence data in drinking water.
Summary/Accomplishments (Outputs/Outcomes):
Peroxides
Preconcentration of the target analytes in ozonated aqueous solutions was achieved through the use of solid phase extraction (SPE). Once extracted, the peroxides are separated by reverse phase high performance liquid chromatography. As the peroxides elute from the analytical column, they are converted to peroxide radicals by horseradish peroxidase in a post-column reactor. These radicals then react with p-hydroxyphenylacetic acid to form a fluorescent dimer, the signal of which is increased by the addition of sodium hydroxide. Without SPE, the method detection limit (MDL) for hydrogen peroxide is an adequate 270 ng/L, but the MDL values for peracetic acid, 2-butanone peroxide, cumene hydroperoxide, and tert-butyl hydroperoxide are all above 50 µg/L. With the SPE pretreatment, these latter values were lowered to 1µg/L. Hydrogen peroxide is a significant by-product of ozonation but does not persist in distributed drinking water. Peracetic acid was identified among the several organic peroxide by-products generated during ozonation, but in the absence of standards, other chromatographic peaks remain unidentified.
Epoxides
Epoxides have long been expected to be at least transition by-products during ozonation of organic moieties, but because they are unstable, highly polar water-soluble species, their extraction from aqueous matrices is not straightforward. Instead, it has been necessary to devise a method that initially will "flag" the functional group rather than individual species and then subsequently elucidate the specific identity. These challenges have been met with the following general approach: (1) functional group-specific aqueous-phase derivatization to provide identifiable flags and improve extractability of analytes; (2) extraction of derivatives by selective SPE methods; (3) further derivatization of remaining active functional groups; (4) analysis by gas chromatography-mass spectrometry (GC-MS) in electron ionization (EI) mode targeting applied flags; and (5) closer structure elucidation by GC-MS in EI and chemical ionization modes.
Initial investigations were undertaken with 2,5-di- and pentafluoroaniline. The mass spectral fragmentation of these derivatives, however, did not produce adequate identifiable spectral flags for a satisfactorily wide variety of epoxide compounds. They did, however, display some usefulness for the flagging of terminal-carbon epoxides. A more broadly satisfactory derivatizing agent was found in 3,5-difluorobenzylamine (DFBA). The use of DFBA has proved successful in both enhancing the extractability of epoxides from water and their analytical detectability. Furthermore, because epoxides are known to be somewhat unstable, the process has the extra advantage of producing a stable product for analysis. The difluorobenzyl ion [C6H3F2-CH2]+ is a relatively stable fragment in most mass spectra with m/z 127, and this provides the necessary general flag for epoxides. A further useful ion commonly was observed to indicate terminal-carbon epoxides with the highly abundant fragment ion [C6H3F2-CH2-NH-CH2]+ with m/z 156. The identity of this fragment was confirmed by its movement to m/z 228 following subsequent vigorous derivatization with silylating agents. Among the by-products associated with ozonation of a high total organic carbon containing surface water, oxirane and phenyloxirane were positively identified by this method.
Carbonyl-Containing Compounds
A significant improvement and expansion of existing analytical methods has been achieved for the analysis of carbonyl by-products of drinking water ozonation. SPE has enabled improved cleanup and selectivity of analytes and allows for extracts to be dried thoroughly prior to secondary derivatization. Silylation has enabled the successful GC and mass spectral detection of carbonyls featuring secondary functional groups such as alcohols and carboxylic acids. The use of phosphate buffer has reduced significantly the background contamination of mass spectral chromatographs.
Ion-trap MS with electron impact ionization has provided the opportunity to detect and identify a wide range of anticipated and unanticipated carbonyl compounds at concentrations as low as 100 ng/L. Structural confirmation is achieved through secondary ionization of target fragments (MS/MS).
Many oxime-derivatized carbonyl compounds are suitable for direct GC analysis. Those with secondary functional groups, including alcohols and acids, pose further complications however. The hydrogen bonding and low volatility of these compounds render them unsuited to chromatography under typical environmental analytical conditions. Under these circumstances, the precise identification of the compounds is difficult, if not impossible. Such difficulties largely have been overcome by silylation of these functional groups. Bis-(trimethysilyl) trifluoroacetamide (BSTFA) derivatization provides suitable stable characteristic ion fragmentation. Other by-products of the derivatization are volatile and are typically eluted during the GC-MS solvent delay. An important consideration of BSTFA derivatization is that trimethylsilyl ethers are highly susceptible to hydrolysis by water. Therefore, it is crucial to dry thoroughly the SPE cartridges before elution and to keep all extracted solutions under anhydrous conditions.
The MDLs for the extraction of 20 mL samples currently are around 10 µg/L in scan mode and 1 µg/L in selected ion monitoring mode for most of the identified by-products. Among the new ozonation by-products positively identified in this category are cyanoformaldehyde, 6-hydroxy-2-hexanone, hydroxyacetaldehyde, and hydroxyacetone. Numerous other by-products with carbon chain lengths varying from 3 to 15 were tentatively identified, but standards to absolutely confirm their identity were unavailable.
Postdisinfection of Ozonation By-Products
Major ozone by-products (i.e., carbonyl-containing) were reacted with both chlorine and preformed chloramine to identify additional components of finished water under such a treatment scenario. The by-products were extracted from the aqueous samples by sequential derivatization with: (1) pentafluorobenzylhydroxamine to form oximes that could be extracted into hexane; (2) silylation of extracted hydroxyl groups with N-methyl-N-(tert-butyldimethylsilyl) trifluoracetamide; and (3) esterification of acid functionalities with diazomethane. In addition, low molecular weight carboxylic acids were analyzed by ion chromatography. The stability of both the parent carbonyl-containing compound and resulting by-products were investigated so that predictions of persistence, or lack thereof, could be provided.
Significant formation of haloacetaldehydes and haloacetamides resulted from subsequent chlorination and chloramination of ozone by-products, and all of these persisted in finished drinking waters. Chlorination of carbonyl-containing ozone by-products is part substitution (e.g., formation of chloral hydrate from acetaldehyde) and part oxidation (e.g., formation of oxalic acid from glyoxal), but under treatment plant conditions it is always an incomplete reaction, leaving a mixture of by-products in drinking water.
Conclusions:
The 1986 Safe Drinking Water Act amendments set in motion the recently completed negotiations over new federal rules for DBPs. Although the main aim of these negotiations was to set new maximum contaminant levels for certain halogenated by-products, there was almost unanimous agreement on the need for more research on DBPs and their fate in drinking water. The approach undertaken in this study successfully extracted and flagged by-products for identification of general classes of compounds. The classes described are carbonyls (including those with acidic and hydroxyl functional groups) and epoxides. Once flagged and assigned to these categories, the specific compounds were identified further by closer inspection of their EI and MS/MS mass spectra. Where possible, compound assignments were confirmed by comparison of mass spectra and retention times with those of reference standards.
In addition to the identification of a number of new classes of ozone by-products (e.g., multifunctional carbonyl-containing compounds, epoxides, peroxides, and haloacetamides), this research project has shown that although the precursors of the traditional chlorination by-products such as THMs and HAAs have been changed to reduce their formation, ozone by-products are precursors of other halogenated by-products when chlorine or chloramines are used as terminal disinfectants during drinking water treatment. Some of these products include haloacetaldehydes and haloacetamides for which health effects have not yet been studied. Ozonation cannot, therefore, be considered a magic bullet technology for improvement of finished drinking water quality without a full evaluation of the mechanisms of reactions with NOM.
A comprehensive survey of by-products in ozonated drinking water has afforded significant gains in the identification of by-products formed from ozonation processes. This information will allow a more complete understanding of the role ozonation may play in the removal of the precursors of the chlorine DBPs.
Journal Articles on this Report : 1 Displayed | Download in RIS Format
Other project views: | All 18 publications | 1 publications in selected types | All 1 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Weinberg H. Disinfection byproducts in drinking water: the analytical challenge. Analytical Chemistry 1999;71(23):801A-808A. |
R826833 (Final) |
|
Supplemental Keywords:
disinfection, exposure, organics, analytical methods, ozonation, alternative disinfection methods, chloramines, chlorine-based disinfection, community water system, disinfection by-products, DBPs,, RFA, Scientific Discipline, Water, Environmental Chemistry, Chemistry, Analytical Chemistry, Drinking Water, alternative disinfection methods, public water systems, water quality parameters, exposure and effects, disinfection byproducts (DPBs), stability, exposure, community water system, treatment, chlorine-based disinfection, chloramines, DBP risk management, water quality, drinking water contaminants, water treatment, formation, drinking water systemRelevant Websites:
None.
Progress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.