Characterization of TOX Produced During Disinfection Processes

Project Team

Faculty PIs: David A. Reckhow, Patrick G. Hatcher

Graduate Students and Post-Docs: Guanghui Hua, and Junsung Kim, Sarah A.L. Caccamise and Rakesh Sachdeva

Journal Publications

Hua GH, Reckhow DA. Determination of Tocl, Tobr and Toi in Drinking Water by Pyrolysis and Off-Line Ion Chromatography. Analytical and Bioanalytical Chemistry 2006; 384: 495-504.

Hua GH, Reckhow DA, Kim J. Effect of Bromide and Iodide Ions on the Formation and Speciation of Disinfection Byproducts During Chlorination. Environmental Science & Technology 2006; 40: 3050-3056.

Hua, Guanghui and Reckhow David A. Characterization of Disinfection Byproduct Precursors Based on Hydrophobicity and Molecular Size. 2006 in review.

Hua, Guanghui and Reckhow David A. Comparison of TOX formation from chlorine and alternative disinfectants.  2006 in review.

Hua, Guanghui and Reckhow David A. Factors Affecting the Formation of Disinfection Byproducts during Chlorination and Chloramination.. 2006 in review.

Hua, Guanghui and Reckhow David A. Hydrophobicity and Molecular Size Distribution of Unknown Halogenated Disinfection Byproducts in Drinking Water.  In preparation.

Project Report

Reckhow, David A., Hua, Guanghui, Kim, Junsung, Hatcher, Patrick G., Caccamise, Sarah A. L., and Sachdeva, Rakesh. Characterization of TOX Produced During Disinfection Processes.  2007. Denver, CO, American Water Works Association Research Foundation.

Presentations:

Hua, G., and DA Reckhow, “Determination of TOCl, TOBr, and TOI in Drinking Water by Pyrolysis and Off-line Ion Chromatography”, Proceedings of the Water Quality Technology Conference.  2004, AWWA, Denver, CO.

Hua, G., and DA Reckhow, “Effect of Bromide and Iodide Ions on the Formation and Speciation of DBPs During Chlorination”, Proceedings of the Water Quality Technology Conference.  2004, AWWA, Denver, CO..

Hua, G.., DA Reckhow, “Characterization of Unknown TOX Precursors based on Hydrophobicity and Molecular Size”, Proceedings AWWA Annual Conference, June 12-16, 2005; San Francisco, CA.

Hua, G., and DA Reckhow, “Determination of TOCl, TOBr, and TOI in Drinking Water by Pyrolysis and Off-line Ion Chromatography”, oral presentation, Water Quality Technology Conference.  November 16, 2004, San Antonio, TX.

Hua, G., and DA Reckhow, “Effect of Bromide and Iodide Ions on the Formation and Speciation of DBPs During Chlorination”, oral presentation, Water Quality Technology Conference.  November 17, 2004, San Antonio, TX.

Hua, G.., DA Reckhow, “Characterization of Unknown TOX Precursors based on Hydrophobicity and Molecular Size”, oral presentation, AWWA Annual Conference, June 12-16, 2005; San Francisco, CA.

Sachdeva, R., PG Hatcher, S. Kim, AG Marshall, RP Rogers and DA Reckhow, „Characterization of Total Organic Halogen (TOX) Produced During Disinfection Processes by ESI-TOF MS and ESI-FTICR MS“ poster presentation, EMSI/North Central NOM Workshop, Columbus, OH, June 15-17, 2005

Hua, G.., DA Reckhow, “Comparison of total organic halogen formation from various disinfection processes”, oral presentation, Water Quality Technology Conference,  November 6-10, 2005 (TUE1); Quebec City, Que.

Hua, G.., DA Reckhow, “Factors affecting the formation and destruction of TOX in drinking water”, oral presentation, Water Quality Technology Conference,  November 6-10, 2005; Quebec City, Que.

Hua, G, and DA Reckhow, “Comparison of total organic halogen formation from chlorine and alternative disinfectants” oral presentation at the AWWA Annual Conference, June 12, 2006, San Antonio, TX

Hua, G, and DA Reckhow, “Hydrophobicity and molecular size distribution of unknown halogenated disinfection byproducts in drinking water” oral presentation at the Water Quality Technology Conference,  , November 6, 2006, Denver, CO

Project Summary

Total organic halide (TOX) is an analytically defined measurement that is often applied to environmental waters, especially drinking waters.  The intent of the TOX measurement is to provide an estimate of the total amount of organically-bound (i.e., by covalent C-X bonds) chlorine, bromine and iodine in a dilute water sample.  The standard method for measuring TOX does not allow for separate determination of the various halogen species.  Alternative TOX methods capable of differentiating between the halogens (i.e., TOCl vs TOBr vs TOI) are of interest to the water industry, because of the perceived toxicity of bromine and iodine containing disinfection byproducts.

Research Objectives

The purpose of this research was: (1) to determine the best TOX protocol for use with ion chromatography (IC) analysis for the purposes of discriminating between TOCl, TOBr and TOI. (2) to determine the nature and chemical characteristics of the unknown fraction of the total organic halogen (UTOX) produced during chlorination and alternative disinfection processes (i.e., chloramination, chlorine dioxide, ozone disinfection), and (3) to assess the impact of treatment on removal of UTOX precursors.

Approach

This work was conducted in several phases; and it built upon the latest fundamental advancements in NOM characterization.  First, a series of TOX methodology studies were undertaken.  This was needed to validate existing TOX methods before they could be reliably applied to the analysis of TOBr and TOI.  Next, a survey of selected North American utilities was conducted.  This involved the collection of waters of diverse quality and geographic location for laboratory treatment with 5 basic disinfection scenarios (chlorination, chloramination, both with and without preozonation, and chlorine dioxide).  Analysis of these samples for TOX species and known DBPs was undertaken to help the researchers better assess the full range of UTOX occurrence and the raw water characteristics that are associated with higher levels.  In addition, distribution system samples were fractionated according to hydrophobicity and molecular size, and then analyzed for UTOX.  The next task focused on factors influencing UTOX concentrations, especially engineering factors.  This task was designed to examine impacts of chemical conditions during disinfection on ultimate UTOX concentrations.  The final phase was directed to the application of advanced chemical techniques (borrowed from the humic substances research field) to the characterization of UTOX.  A set of innovative techniques were explored for assessing UTOX on a molecular level.  Most prominent among these was electrospray ionization fourier transform ion cyclotron resonance mass spectrometry (ESI-FT-ICR-MS).

Conclusions

Modeling and Occurrence

Using laboratory data, an empirical model was developed for Unknown TOX based on easily measured DBPs and various reaction conditions.  Test data were extracted from the Information Collection Rule (ICR) federal database allowing the calculation of known to unknown TOX quotients for hundreds of waters across the US.  Estimates were made of nationwide average ratios of known/unknown TOX.  Further examination of the data revealed extreme sites where with high humic waters that tend to form more identifiable byproducts, and low humic waters that tend to produce a higher relative amount of UTOX.

Based on the laboratory analysis of 6 contrasting waters, there is a broad range in HAA9/THM precursor ratios (0.8-1.7 on a weight basis) in North American NOM.  This implies a substantial diversity in NOM types, especially considering that both THMs and HAAs are general oxidation end products of a wide range of organic molecules.  The data also supported an earlier assertion that terrestrial and lignin-based precursors are both highly-absorbing and rich in HAA precursors as compared to THM precursors.  Based on this set of waters, there appeared to be little difference in the UTOX/TOX ratios, suggesting that UTOX may not be a sensitive indicator of the quality of the unknown byproducts. 

TOX Method Refinement

The pyrolytic analyzer using pure O2 and off-line IC combined with a standard TOX carbon (coconut based) achieved complete recovery of TOCl, TOBr and TOI from model compounds.  The O2/CO2 pyrolytic system and off-line IC showed incomplete recovery for some samples due to condensation of halide ions and the difficulty in flushing these halide ions.  The use of CO2 as an auxiliary gas resulted in interference in IC analysis, which made it necessary to purge dissolved CO2 before IC analysis.  There was no obvious difference between the two analyzers when used in microcoulometric detection mode.  The TOX method is moderately sensitive to nitrate rinse volume.  The monohaloacetic acids were partly washed out during sample preparation.  This problem was solved by modifying the nitrate rinsing procedure.  Complete recovery was achieved for all the selected compounds by this modified TOX protocol.  The two coconut based carbons (CPI-001 and CPI-002) gave nearly identical recoveries for natural water samples.  The bituminous coal based carbon (F-600) suffered from excessive inorganic iodide retention.

Incorporation of Bromine and Iodine

Molar yields of trihalomethanes (THMs) and haloacetic acids (HAAs) increased as the initial bromide ion concentration increased.  No significant change in total organic halogen (TOX) concentration was found for varying bromide concentrations.  However, TOX concentrations decreased substantially with increasing initial iodide ion concentrations.  At higher levels of bromide, there was a decreasing level of unknown TOX and unknown total organic chlorine (UTOCl) but an increasing level of unknown total organic bromine (UTOBr).  Bromine substitution into THMs and dihaloacetonitriles (DHAN) was shown to be more complete than into trihaloacetic acids (THAA) and dihaloacetic acids (DHAA).  Generally, the order of bromine substitution was DHAN> DHAA& THM >THAA. The extent of iodine substitution was much lower than that of bromine substitution when comparing identical initial concentrations since a substantial amount of iodide was oxidized to iodate by chlorine.  Increasing chlorine dose was able to control the formation of iodinated organic compounds.  The Cl2/TOC or Cl2/I ratio is critical to the formation of iodinated disinfection byproduct (I-DBPs) and iodate (IO3-).  While high levels of I-DBPs may be rare in practice, when they do occur, increasing chlorine dose is a feasible method to control the formation of these compounds.  Low dose chlorination may result in a substantial amount of I-DBPs for iodide containing water. 

TOX from Alternative Disinfectants

Preformed chloramines produced from 7% to 18% of the TOX produced by free chlorine.  It should be noted that these experiments were designed to isolate the impacts of pure monochloramines, whereas full-scale systems always involve some exposure (sometimes transient) to free chlorine.  Chlorine dioxide showed a slightly lower and narrower range (5%-8%).  Preozonation usually resulted in less UTOX upon subsequent free chlorination.  However, when followed by chloramination, the impacts of preozonation were of little benefit and more often detrimental.  One water showed net increases in UTOX as a result of pre-ozonation regardless of the final disinfectant.

Characterization of UTOX Precursors

In tests using chlorination, hydrophobic and high MW (e.g. >3kDa) precursors produced more trihalomethanes (THMs), trihaloacetic acids (THAA) and unknown total organic halogen (UTOX) than corresponding transphilic, hydrophilic and low MW (e.g. <3kDa) precursors.  However, the formation of THMs and THAA was similar among different fractions for a water with low humic content.  Hydrophilic and low MW fractions (<0.5k) gave the highest dihaloacetic acid yields.  No significant difference was found for dihaloacetic acid and UTOX formation among different fractions during chloramination.  It appears that chloramination DBP precursors are more evenly distributed among NOM fractions.  High pH favors the formation of THMs and HAAs over UTOX.  Increasing pH also led to lower TOX formation for hydrophobic and high MW fractions, but had little impact on TOX yields from hydrophilic and low MW fractions.  Bromine and iodine were found to be more reactive with hydrophilic and low MW fractions as measured by THM and HAA formation than their corresponding hydrophobic and high MW fractions.  However, hydrophobic and high MW fractions produced more UTOX when reacting with bromine and iodine.  A conceptual model was proposed involving aliphatic and aromatic precursors.  This was used to bridge the empirical observations on halide incorporation with precursors of varying characteristics

Characterization of UTOX Compounds

Chemical and physical property based measurements (i.e., resin adsorption and membrane separation) indicate that most UTOX is in the mid-size range (0.5-10 kDa), but it can have a wide spectrum of partitioning properties or hydrophobicities.  These sizes suggest that the bulk of the UTOX resembles halogenated fulvic acid molecules with little fragmentation, but substantial modification in the form of greater densities of hydrophilic groups (carboxylic acids) may occur.  This material also seems to have a special affinity for XAD-4 over XAD-8 (i.e., transphilic) suggesting an intermediate polarity and aromaticity.

Effect of Reaction Conditions on UTOX Formation

In the early stages of chlorination reactions, UTOX compounds are formed preferentially at neutral and alkaline pHs.  At acidic pHs the measureable byproducts and the unknown TOX seem to form at about the same rate.  In pure chloramination processes the reverse occurs.  Neutral to high pHs result in rapid formation of measureable byproducts (principally DHAAs), whereas the formation of UTOX is much slower.  DHAAs form very quickly in both chlorination and chloramination scenarios.  Unknown TOX also forms quickly by chlorination, but quite slowly by chloramination.  In systems where chlorine is limiting (i.e., low doses and no residual), most of the TOX is locked up in UTOX compounds with very little THAAs.  As higher doses are applied, the UTOX gives rise to high levels of THAAs.  This occurs in both free and combined chlorine systems.

Corrosion control chemicals can impact DBP formation through pathways that probably involve general or specific base catalysis.  For example, use of sodium hydroxide (up to pH 11) and silicate (at pH ~9.5) caused loss of TOCl, TOBr, and especially TOI.  At the same time, it resulted in increases in chlorinated THMs, probably via hydrolysis of meta-stable trichloromethyl intermediates.  The same phenomenon was observed for the DHAAs (chlorinated and brominated), but not the THAAs

Characterization of Individual UTOX Molecules

Our analysis of the four water samples by electrospray time of flight mass spectrometry shows spectral features characteristic of DOM samples.  The mass spectral data are very complex and we also notice the increase in the complexity of data upon chlorination.  Since the DOM samples have spectral intensity every 2 Da apart, analysis of the isotope patterns due to presence of halogen atoms in the molecules is not feasible at the resolution of the TOF MS instrument.

The ultra-high resolution mass spectral analysis of the samples on the 9.4 T FTICR mass spectrometer at the National High Magnetic Field Laboratory at Florida State University yielded a wealth of information.  All 13 samples were analyzed in the negative ion mode only to make the most use of our instrument time.  The high mass resolving power of the FTICR MS (approximately 400,000 at m/z 400) was able to separate these complex DOM mixtures.  In each ESI FTICR mass spectrum, there were an average of 10,000 to 14,000 different components resolved in an individual sample.  These resolving powers were achieved on the whole sample extract, with no prior separation technique employed (besides the initial C18 isolation/extraction). 

The majority of the high intensity ions found in the ESI FT-ICR mass spectra are C, H and O based compounds.  The halogenated DBPs that are C,H, O and halogen based compounds are believed to be the best molecular matches, based on known DBPs formed from other studies.  These molecular formula matches are still preliminary, and the isotopic abundances still need to be verified in future work.

Originally, it was thought that the halogenated DBPs would occur at negative mass defects, but from the analysis performed so far, there are several halogenated DBPs that occur at positive mass defects.  The mass defect is also based on the other elements and their number, and this can counteract the negative mass defect effect of the halogens.  This complicates the analysis, because not only the negative mass defect ions can be analyzed to look at the DBPs, the whole data set needs to be considered.

Kendrick plots are useful in comparing data sets.  These plots can also be helpful in the determination of homologous series in the data, which can aid in the data interpretation for the higher molecular weight components.  Kendrick plots of the different disinfection processes show that there are a large number of components that do not lie within a homologous series related by a CO2 group.  This means that there is a large number of components that do not contain only C, H and O, but have some other elements (chlorine?).  The Cl2 treated water has a large number of components, and many have Cl included in their proposed molecular formulas.  The other 4 disinfection processes show less deviation from the CO2 relationship, and have less chlorinated matches.  This shows that the other disinfection processes create less halogenated DBPs, and C,H, O based components are formed from the oxidation of the DOM.

            The two-dimensional van Krevelen plot of the chlorinated species in the Winnipeg chlorinated water illustrates that the chlorinated DBPs originate with DOM.  The three-dimensional van Krevelen plot of the species from m/z 400 to 450 show the typical DOM cluster at H/C ratio of 1 and O/C ratio of 0.5.  Plotting the Cl/C ratio on the z-axis shows layers within the data relating to the number of chlorines present.  These layers all lie above the DOM cluster, illustrating that these chlorine substituted components are a result of the reaction of DOM with chlorine.