Environ. Sci. Technol. 2007, 41, 8315–8320
The Impact of UV/H2O2 Advanced Oxidation on Molecular Size Distribution of Chromophoric Natural Organic Matter SIVA R. SARATHY AND MADJID MOHSENI* Department of Chemical and Biological Engineering, University of British Columbia, 2360 East Mall, Vancouver, BC V6T 1Z3 Canada
Received June 29, 2007. Revised manuscript received October 11, 2007. Accepted October 18, 2007.
The impact of hydroxyl radical (•OH) on the molecular weight distribution of natural organic matter (NOM) was investigated. •OH was generated via the photolysis of hydrogen peroxide (H2O2) by ultraviolet (UV) radiation of 254 nm, also known as UV/ H2O2 advanced oxidation (AO). Additionally, the impact of combined membrane and UV/H2O2 treatment on the molecular weight distribution of NOM was studied. High performance size exclusion chromatography (HPSEC) was used to determine the apparent molecular weight (AMW) distribution of chromophoric NOM (CNOM). Prior to AO, 33% of the CNOM in the water had AMW greater than 1400 Da. Meanwhile, lower AMW CNOM made up smaller amounts of the CNOM, with CNOM of AMW less than 450 Da making up 5% of the total. Under the AO conditions typically applied in drinking water treatment applications, NOM was not mineralized but was partially oxidized resulting in significant reduction in aromaticity. •OH preferentially reacted with higher AMW CNOM and the fragmentation of high AMW CNOM led to the formation of smaller AMW CNOM. Ultrafiltration removed all CNOM greater than 1400 Da AMW and a large portion of other high AMW fractions. In the absence of high AMW CNOM, •OH reacted more readily with all AMW fractions leading to a reduction in concentration of most AMW fractions. Whereas •OH reacted nonspecifically with all AMW fractions, the reaction rate between •OH and CNOM was observed to be dependent on molecular size.
Introduction Ultraviolet/hydrogen peroxide (UV/H2O2) advanced oxidation process (AOP) has been extensively researched for the removal, from drinking water, of organic contaminants such as N-nitrosodimethylamine (1, 2), methyl tert-butyl ether (3–6), herbicides and pesticides (7), and taste and odor compounds, such as 2-methylisoborneol and geosmin (8). Furthermore, there are currently a number of commercial UV/H2O2 applications for the treatment of drinking water originating from ground and surface reservoirs (9, 10). While much research has focused on finding target contaminants susceptible to UV/H2O2, the fate of natural organic matter (NOM) during typical UV/H2O2 treatment conditions has received little attention. * Corresponding author e-mail:
[email protected]. 10.1021/es071602m CCC: $37.00
Published on Web 11/13/2007
2007 American Chemical Society
NOM, a complex mixture of humic and fulvic acids, is ubiquitously present in surface waters and poses several challenges to drinking water treatment operations. With respect to water quality, NOM acts as a precursor for disinfection byproducts (DBPs) and has potential to increase the biological regrowth potential of water in distribution systems. During UV/H2O2 AOP, NOM screens light needed for the photolysis of H2O2 and scavenges hydroxyl radicals (•OH) needed for contaminant removal. Under strong advanced oxidation (AO) conditions (i.e., long irradiation time and/or high H2O2 concentration), studies have found that NOM becomes mineralized, indicated by a decrease in total organic carbon (TOC) (11–17). From a water quality standpoint, the removal of NOM is beneficial since this leads to reductions in the formation of DPBs and biological regrowth potential. However, the strong AO conditions required for mineralization are not economically feasible. Commercial UV/H2O2 systems, applied for the removal of trace organic contaminant in drinking water, operate at conditions sufficient for degradation of the target pollutants. Under these conditions NOM is not mineralized but rather partially oxidized. Studies have reported that this partial oxidation may lead to increases in the formation potential of DBPs (11, 16–18) and biodegradability of NOM (12, 16–18). But, these studies focused on UV/H2O2 as a treatment for the removal of NOM so there is insufficient data with respect to what happens to water quality when NOM is partially oxidized. Therefore, NOM plays a critical role during commercial UV/H2O2 applications since it not only impacts the effectiveness of UV/H2O2, but also undergoes transformations that may impact water quality. Proper understanding of how NOM is transformed under the conditions used in commercial applications and what products are formed are issues of key importance in the development of solutions to address potential problems. That is, how the physical and chemical characteristics of NOM are altered as a result of reaction with •OH requires attention. The aim of this study was to investigate how UV/H2O2 impacts NOM by observing the effects of UV/H2O2 on NOM’s molecular weight (MW) distribution. MW distribution of NOM is an important property in drinking water treatment since changes in MW distribution can lead to changes in DBP formation potential and biological regrowth potential (14–16, 19–24). In general, lower MW species are more biodegradable, whereas higher MW species are more humic and aromatic in nature and may be more reactive with chlorine. Furthermore, the impact of membrane separation as well as combined membrane and UV/H2O2 treatment on NOM was studied. The latter was employed to investigate how •OH affects NOM’s MW distribution when the initial MW distribution of NOM is altered by pretreatment. High performance size exclusion chromatography (HPSEC) has been demonstrated to be an effective technique for determining the apparent MW (AMW) distribution of NOM (25). During HPSEC, NOM is detected by absorbance of 260 nm UV, thus yielding the AMW distribution of only chromophoric NOM (CNOM), NOM that is able to absorb at 260 nm (i.e., chromophores). In this study, HPSEC was employed to investigate the impact of •OH generated by UV/H2O2 on the AMW distribution of CNOM.
Materials and Methods Waters. The original source of water used in all experiments was the Capilano Reservoir, which provides drinking water for the Greater Vancouver region, British Columbia, Canada. The damming of the Capilano River, which is fed by fall and VOL. 41, NO. 24, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
8315
TABLE 1. Characteristics of Capilano Source Water during 2005 and Waters Used in This Studya CW reservoir (2005)b parameter alkalinity as CaCO3 (mg/L) dissolved organic carbon (mg/L) total organic carbon (mg/L) pH turbidity (NTU) A254 (cm-1) a
N/A, not measured.
avg
CW used in experiments (March 2006)
range
avg
9
stdev
avg
stdev
avg
stdev
2.7
2.1–3.6
N/A
N/A
N/A
2.0
1.6–2.7
same as TOC
same as TOC
same as TOC
2.0
1.5–2.9
2.2
0.2
0.81
0.1
1.5
0.1
6.8 N/A 0.098
0.1
6.8 N/A 0.011
0.1
6.8 N/A 0.038
0.1
6.5 1 0.081 b
6.2–6.9 0.32–5.9 0.055–0.108
0.008
0.001
0.001
Source: The Greater Vancouver Water District Quality Control Annual Report 2005.
winter rain runoff and the spring snowmelt, forms the reservoir. Given its low TOC, turbidity, and alkalinity (Table 1) Capilano water (CW) was a surface water of very high quality and presently undergoes no coagulation/flocculation or filtration prior to chlorine disinfection. The water used for this particular study was obtained in March of 2006, and its characteristics are given in Table 1. The water with
