Search

Dissolved organic matter (DOM) represents the major pool of organic precursors for harmful disinfection byproducts, such as haloacetic acids (HAAs), formed during drinking water chlorination, but much of it remains molecularly uncharacterized. Knowledge of model precursors is thus a prerequisite for understanding the more complex whole water DOM. The utility of HAA formation potential data from model DOM precursors, however, is limited due to the lack of comparability to water samples. In this study, the formation kinetics of dichloroacetic acid (DCAA) and trichloroacetic acid (TCAA), the two predominant HAA species, were delineated upon chlorination of seventeen model DOM precursors and sixty-eight inland lake water samples collected from the Upper Midwest region of the United States. Of particular interest was the finding that the DCAA and TCAA formation rate constants could be grouped into four statistically distinct clusters reflecting the core structural features of model DOM precursors (i.e., non-β-diketone aliphatics, β-diketone aliphatics, non-β-diketone phenolics, and β-diketone phenolics). A comparative approach built upon hierarchical cluster analysis was developed to gain further insight into the chlorine reactivity patterns of HAA precursors in inland lake waters as defined by the relative proximity to four model precursor clusters. This work highlights the potential for implementing an integrated kinetic-clustering approach to constrain the chlorine reactivity of DOM in source waters.

This paper demonstrates that disinfection byproducts (DBP) concentration profiles in heated water were quite different from the DBP concentrations in the cold tap water. Chloroform concentrations in the heated water remained constant or even decreased slightly with increasing distribution system water age. The amount of dichloroacetic acid (DCAA) was much higher in the heated water than in the cold water; however, the maximum levels in heated water with different distribution system water ages did not differ substantially. The levels of trichloroacetic acid (TCAA) in the heated water were similar to the TCAA levels in the tap water, and a slight reduction was observed after the tap water was heated for 24 h. Regardless of water age, significant reductions of nonregulated DBPs were observed after the tap water was heated for 24 h. For tap water with lower water ages, there were significant increases in dichloroacetonitrile (DCAN), chloropicrin (CP), and 1,1-dichloropropane (1,1-DCP) after a short period of heating. Heating of the tap water with low pH led to a more significant increase of chloroform and a more significant short-term increase of DCAN. High pH accelerated the loss of the nonregulated DBPs in the heated water. The results indicated that as the chlorine doses increased, levels of chloroform and DCAA in the heated water increased significantly. However, for TCAA, the thermally induced increase in concentration was only notable for the chlorinated water with very high chlorine dose. Finally, heating may lead to higher DBP concentrations in chlorinated water with lower distribution system temperatures.

Haloacetic acids (HAAs) are a family of compounds whose environmental concentrations have been extensively studied, primarily in Europe. Depending on the compound, their sources are believed to be both natural and anthropogenic. To better understand possible sources and contribute to the knowledge of the global distribution of these compounds, especially between the Northern and Southern Hemispheres, samples of precipitation, soils, and conifer needles were collected from Canada, Malawi, Chile, and the U.K. Precipitation samples exhibited highest HAA concentrations in collections from Canada, and lowest in those from Malawi. Malawi samples contained measurable levels of monobromoacetic acid (MBA) (56 ng/ L) unlike those from most other locations (< 9 ng/L). Soil HAA concentration levels were highest in the U.K. (e.g., 7.3 ng/g average TCA) and lowest in Malawi (0.8 ng/g average TCA), with Chile having higher levels (4.8 ng/g average TCA) than Canada (3 ng/g average TCA). Malawi soils contained small amounts of MBA (2 ng/g), in common with the two most southern of the 11 Chilean sites. Analysis of soil cores (10-cm depth sliced at 1 cm) from sites in Malawi and Chile showed that trichloroacetic acid (TCA) generally declined with depth while mono- and dichloroacetic acid (MCA and DCA) showed no trend. MCA, DCA, and TCA concentrations in archived U.K. soil samples increased by factors of 2, 4, and 5-fold over 75 years while TFA showed no consistent trend. Monochloroacetic acid (MCA) was detected in pine needles collected from Malawi. U.K. needle samples had the highest concentrations of all chloroacetic acids (CAAs): MCA, 2-18 ng/g; dichloroacetic acid (DCA), 2-38 ng/g; and trichloroacetic acid (TCA), 28-190 ng/g. Conifer needles from Canada and Chile contained CAAs at levels ranging from < 2 to 16 ng/g wet wt. Trifluoroacetic acid concentrations generally declined with increasing elevation in the samples from the Rocky Mountains in western Canada. The results indicate that concentrations of HAAs are greatest in the industrialized Northern Hemisphere but there are significant amounts of these compounds in the less industrialized Southern Hemisphere.

The concentrations and input/output fluxes of trichloroacetic acid (TCA) were measured in all relevant media for one year at a 0.86 km2 upland conifer plantation and moorland catchment in SW Scotland (n > 380 separate samples analyzed). Annual wet precipitation to the catchment was 2.5 and 0.4 m for rain and cloud, respectively. TCA input to the catchment for the year was 2100 g, predominantly in rainwater (86%), with additional input via cloudwater (13%) and gas plus particle dry deposition (1%). There were no seasonal trends in TCA deposition, and cloudwater concentration was not enhanced over rainwater. TCA in precipitation exceeded concentrations estimated using currently accepted routes of gas-phase oxidation from anthropogenic chlorinated hydrocarbon precursors, in agreement with previous studies. Export of TCA from the catchment in streamwater totalled 1970 g for the year of study. The TCA concentration in streamwater at outflow (median 1.2 microg L(-1)) was significantly greater than that before the stream had passed through the conifer plantation. To well-within measurement uncertainties, the catchment is currently at steady-state with respect to TCA input/output. The catchment reservoir of TCA was dominated by soils (approximately 90%), with the remainder distributed in forest litter (approximately 9%), forest branchwood and stemwood (approximately 0.7%), forest foliage (approximately 0.5%), and moorland foliage (approximately 0.1%). Although TCA is clearly taken up into foliage, which consequently may be important for the vegetation, this was a relatively minor process for TCA at the catchment scale. If it is assumed, on the basis of laboratory extraction experiments, that only approximately 20% of "whole soil" TCA measured in this work was water extractable, then total mass of TCA in the catchment is reduced from approximately 13 to approximately 3.5 kg. Comparing the latter value with the annual flux yields an average steady-state residence time for TCA in the catchment of approximately 1-2 y, if all TCA is involved in catchment turnover. Considering that other evidence indicates the lifetime of TCA in soil and biota is considerably shorter than this (weeks rather than years), the magnitude of the TCA reservoir is suggested to be strong evidence for net natural TCA production in soils and/or that the majority of TCA in the reservoir is not involved with external fluxes.

The first long-term concurrent measurements of trichloroacetic acid (TCA) in rainwater, in cloudwater, and in air (both gas and particle phase) are reported. Measurements were made weekly between June 1998 and April 2000 at a rural forested upland site in SE Scotland. Rainwater TCA concentration did not differ significantly between two elevations (602 and 275 m asl), with precipitation-weighted mean values of 0.77 and 0.70 microg L(-1), respectively (n > 75). The precipitation-weighted mean concentration of TCA in cloudwater at the highest elevation was 0.92 microg L(-1), yielding an average cloudwater enrichment factor of 1.2, considerably lower than for other inorganic ions measured. Rainwater and cloudwater TCA concentrations did not vary systematically with season. Since wet precipitation depth also did not vary systematically with season, the wet deposition fluxes of TCA were likewise invariant (annual fluxes at the highest elevation of 880 and 130 microg m(-2), respectively, for rain and cloud interception to spruce forest). Weekly integrated concentrations of TCA in air (gas and particle) were very low (median 25 pg m(-3), range < LOD-110 pg m(-3)). The estimated upper limit for annual dry deposition of TCA at this site was approximately 20 microg m(-2), assuming a deposition velocity of 2 cm s(-1). Concentrations of TCA in air correlated reasonably strongly with concentrations in rainwater, with a partition ratio approximately equal to the Henry's law coefficient. On average, only about 23% of TCA measured in Edinburgh air was associated with the particle phase. These measurements are consistent with the observed high scavenging ratio of TCA (ratio of concentration in air to concentration in rainwater). Overall, these data confirm that the atmosphere is an important source of TCA to the environment and that precipitation is the dominant transfer mechanism. In line with previous work, the atmospheric deposition flux is greater than expected from the current understanding of atmospheric production of TCA from anthropogenic precursors. It is suggested that aqueous-phase processes could lead to greater atmospheric conversion of chlorinated solvent precursors to TCA than is currently accepted.

This study reports an example of visible-light photocatalyst based on TiO2 modified by ruthenium-complex sensitizers and noble metal deposits. The photodegradation of trichloroacetate (TCA) and carbon tetrachloride was used as a probe reaction for evaluating the visible light activity of the photocatalyst under the illumination of lambda > 420 nm. Photodeposition of platinum nanoparticles on dye-sensitized TiO2 (Pt/TiO2/Ru(II)L3) drastically enhanced the degradation rate of TCA and CCl4. The visible light reactivity of Pt/TiO2/Ru(II)L3 was optimal with [Ru(II)L3] = 10 microM, [TiO2] = 0.5 g/L, and Pt loading of about 0.2 wt %. Although no electron donors to regenerate the oxidized Ru-sensitizers were added in the aqueous suspension, the photoreductive dechlorination of perchlorinated compounds proceeded far beyond the stoichiometric limit of the initial sensitizer concentration. Water acted as an electron donor to regenerate the sensitizer with a concurrent production of dioxygen. On the other hand, Pt/TiO2/Ru(II)L3 was completely inactive in the presence of dissolved oxygen and the in-situ generated dioxygen gradually decelerated the dechlorination rate. Conduction band electrons transferred to O2 in preference to CCl4 and TCA on Pt deposits. Other noble metals (Ag, Au, and Pd) deposited on TiO2 showed a better oxygen-tolerance but less visible-light reactivity than PtTiO2/Ru(II)L3. Effects of metal loading on the visible light activity and its implications for the efficientvisible-light photocatalyst development are discussed.