Degradation of Amine-Based Water Treatment Polymers during

Feb 4, 2009 - AMELIA E. TAYLOR, §. CÉDRICK FAVERO, |. AND. CHING-HUA HUANG* , †. School of Civil and Environmental Engineering, Georgia...
0 downloads 0 Views 1MB Size
Download PDF
Environ. Sci. Technol. 2009, 43, 1360–1366

Degradation of Amine-Based Water Treatment Polymers during Chloramination as N-Nitrosodimethylamine (NDMA) Precursors SANG-HYUCK PARK,† SHUTING WEI,‡ BORIS MIZAIKOFF,‡ AMELIA E. TAYLOR,§ ´ DRICK FAVERO,| AND CE C H I N G - H U A H U A N G * ,† School of Civil and Environmental Engineering, Georgia Institute of Technology, Atlanta, Georgia 30332, School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia 30332, Polychemie Inc., a subsidiary of SNF, Pearlington, Mississippi 39572, and SNF Floerger, ´ 42163 Andrezieux Cedex, France

Received September 26, 2008. Revised manuscript received December 3, 2008. Accepted December 18, 2008.

Recent studies indicated that water treatment polymers such as poly(epichlorohydrin dimethylamine) (polyamine) and poly(diallyldimethylammonium chloride) (polyDADMAC) may form N-nitrosodimethylamine (NDMA) when in contact with chloramine water disinfectants. To minimize such potential risk and improve the polymer products, the mechanisms of how the polymers behave as NDMA precursors need to be elucidated. Direct chloramination of polymers and intermediate monomers in reagent water was conducted to probe the predominant mechanisms. The impact of polymer properties including polymer purity, polymer molecular weight and structure, residual dimethylamine (DMA), and other intermediate compounds involved in polymer synthesis, and reaction conditions such as pH, oxidant dose, and contact time on the NDMA formation potential (NDMA-FP) was investigated. Polymer degradation after reaction with chloramines was monitored at the molecular level using FT-IR and Raman spectroscopy. Overall, polyamines have greater NDMA-FP than polyDADMAC, and the NDMA formation from both polymers is strongly related to polymer degradation and DMA release during chloramination. Polyamines’ tertiary amine chain ends play a major role in their NDMAFP, while polyDADMACs’ NDMA-FP is related to degradation of the quaternary ammonium ring group.

Introduction N-Nitrosodimethylamine (NDMA) is a probable human carcinogen (1, 2) that has been identified as an emerging disinfection byproduct (DBP) since its occurrence in drinking water supplies in the U.S. and Canada was shown to be related * Corresponding author phone: 404-894-7694; fax: 404-385-7087; e-mail: [email protected]. † School of Civil and Environmental Engineering, Georgia Institute of Technology. ‡ School of Chemistry and Biochemistry, Georgia Institute of Technology. § Polychemie Inc. | SNF Floerger. 1360

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 5, 2009

to chlorine-based disinfection processes (3-10). Compared to the currently regulated DBPs (e.g., the maximum contaminant level (MCL) for total trihalomethanes ) 80 µg/L), NDMA is considerably more toxic based on its 0.7 ng/L level for 10-6 cancer risk (11). Currently, the California Department of Public Health has set a notification level of 10 ng/L for NDMA and a public health goal (PHG) of 3 ng/L (3). NDMA is now listed along with five other nitrosamines in the second unregulated contaminant monitoring regulation (UCMR2) (12) for drinking water treatment systems in order to determine future regulation. It is also listed among the latest drinking water contaminant candidate list 3 (CCL 3) issued by the U.S. EPA in 2008 (13). Previous research has investigated dimethylamine (DMA) and other simple amines as representative precursors to elucidate NDMA formation mechanism during chloramination. Results showed that NDMA is likely formed from the reaction of DMA with dichloramine to generate a chlorinated unsymmetrical dimethylhydrazine (Cl-UDMH) intermediate, followed by subsequent oxidation of Cl-UDMH by dissolved oxygen (5, 6, 14-16). Although DMA and DMA-based moieties are shown to be significant contributors to NDMA formation, the majority of NDMA precursors in wastewater and drinking water treatment plants are still not well identified and require more research (17-20). Among potential NDMA precursors, cationic amine-based polymers that are commonly used as coagulant and flocculant aids for water and wastewater treatment and for sludge dewatering (e.g., aminomethylated polyacrylamide (Mannich polymer), poly(epichlorohydrindimethylamine) (polyamine), and poly(diallyldimethylammonium chloride) (polyDADMAC)) were reported to form NDMA in water that contained chlorine or chloramines (21-27). The DMA-based moieties in the polymer structures were hypothesized to be responsible for NDMA formation potential (NDMA-FP) (19, 23, 24, 26, 28). Despite the reports, limited knowledge is available on how the polymers breakdown by chlorine oxidants and what the critical influencing factors are for their NDMA-FP. Furthermore, polymers typically contain residual starting materials, monomers, and oligomers because of polymerization procedures; whether polymers and the impurities associated with them contribute to NDMA-FP needs to be evaluated. In this work, studies were conducted to investigate NDMA formation from polyamines and polyDADMACs (Figure 1), two commonly used polymers in potable water treatment. Polyamine is among the polymers with quaternary ammonium ionic groups integrated all along the backbone of the polymer chain instead of in the pendent position. The synthesis of polyamine is well established in the literature (Supporting Information (SI) Figure S1). Polymerization usually takes place with two steps: first, at low temperature, DMA is added to epichlorohydrin, or vice versa, to make a DMA-epichlorohydrin adduct, and then the DMA-epichlorohydrin adduct undergoes condensation polymerization at high temperature to produce the polymer (29). PolyDADMAC was the first quaternary ammonium cationic polymer approved by the U.S. Food and Drug Administration for potable water treatment. As shown in SI Figure S2, typically, the DADMAC monomer is prepared from the reaction of allyl chloride and DMA. Then, the nonconjugated diene of DADMAC undergoes cyclopolymerization to form the pendent 5-membered ring structures of polyDADMAC (30). The objectives of this study were to assess how the structures of polyamines and polyDADMACs and reaction conditions affect their NDMA-FP during chloramination, and identify the predominant mechanisms of polymer degrada10.1021/es802732z CCC: $40.75

 2009 American Chemical Society

Published on Web 02/04/2009

FIGURE 1. Structure of polymers and intermediate compounds examined in this study. I: The chain ends of polyDADMAC are capped by the initiator radical (e.g., persulfate) used in the polymerization process. tion and NDMA formation in order to develop strategies to minimize NDMA formation and improve these polymer products. To obtain mechanistic insight, polymers (typically 10 mg/L) were directly exposed to preformed monochloramine (1-10 mg/L) in reagent water for predetermined time periods (0-24 h). The effects of polymers’ properties including polymer purification, polymer molecular weight (MW), and polymer intermediate compounds (structures shown in Figure 1), and reaction conditions including solution pH, chloramine dose, and contact time were systematically evaluated. Polymers were analyzed by FT-infrared (FT-IR) and Raman spectroscopy to identify structural change by chloramination. Based on the results, recommendations were made to minimize NDMA formation from these two polymers.

Materials and Methods Chemicals and Reagents. Sources of NDMA, DMA, deuterated NDMA (NDMA-d6), deuterated DMA (DMA-d6) and other chemicals, and preparation and quantification of fresh monochloramine stock solutions were described in SI Text S1. Polyamines and polyDADMACs (5.0-49.9% wt. aqueous solutions) containing a range of molecular weights, active contents, and Brookfield viscosities were obtained from SNF Polychemie (Pearlington, MS) and used as received. Fresh stock solutions (500 mg/L as active ingredient) of polymer or intermediate compound were made every time for each experiment. Polymer intermediate compounds DMA-epichlorohydrin-DMA (60% wt. solution), allyldimethylamine (ADMA) (98+ %), and DADMAC monomer chloride salt were obtained from SNF Polychemie. Preparation of DMAepichlorohydrin-DMA and ADMA is described in SI Text S2. Purified polymer samples were obtained by dialysis to remove low MW impurities such DMA and oligomers, with the details described in SI Text S3. Experimental Procedures. Ten mg/L as active ingredient of fresh polymer or intermediate compound solutions were chloraminatedwith10mgasCl2/Lofpreformedmonochloramine at 23 °C for 24 h. Solution pH was controlled at pH 5-9 using 10 mM acetate (pH 5), phosphate (pH 6, 7, and 8), or borate (pH 9) buffer. Afterward, the reactions were quenched by 30 mg/L ascorbic acid before NDMA and DMA analysis. The NDMA concentration measured at the end of the reaction was referred to as the “NDMA formation potential (NDMAFP)” of the polymers. These reaction conditions differ from those in the previous studies by Mitch and co-workers (18, 19), in which the authors utilized a much higher monochloramine dosage (140 mg/L) and longer reaction time (10 days) in NDMA formation potential tests as a mean to quantify the total concentration of NDMA precursors in different water samples. To evaluate the effect of oxidant dosage and reaction kinetics in this study, 4-13 mg/L of preformed monochloramine and reaction time from 0 to 144 h were used. Experiments

showed that the 10 mg/L monochloramine dose was sufficient except for ADMA and the lowest MW polyamine because the residual monochloramine concentrations at the end of 24 h reaction time were >7.8 mg/L for low and high MW polyamines, polyDADMACs, and DADMAC, and 1.5 mg/L for DMA-epichlorohydrin-DMA. For all the experiments, matrix and reagent controls as well as duplicate or triplicate experiments were conducted for each condition. Analytical Methods. NDMA and DMA were analyzed using the methods described in SI Text S4. DADMAC monomer in solutions was analyzed by an adapted method (31) using an Agilent LC/MS system with electrospray ionization (ESI) in positive ion mode. The LC/MS conditions are described in SI Text S5. Polymer Structural Analysis by IR and Raman Spectroscopy. The polymer structures before and after monochloramination were analyzed by FT-IR and Raman spectroscopy. In order to meet the detection limits of IR and Raman, 1 g/L of polyamine and polyDADMAC solutions were each reacted with 200 mg/L of preformed monochloramine for 24 h at room temperature at pH 7.5. Afterward, the polymers were dried and analyzed by IR and Raman. IR measurements were performed at 2 cm-1 for 100 scans with a mercury cadmium telluride (MCT) detector. Raman measurements were performed at 4 cm-1 for 500 scans. The Raman laser power was 550 mW.

Results Impact of Polymer Purification. Because polymer solutions contain residual DMA as well as other impurities including monomers, oligomers, intermediate compounds, catalysts, etc., removal of these impurities allows for the differentiation of the relative contribution of polymer versus residual compounds to the overall NDMA-FP. Polymers purified by dialysis were tested for NDMA-FP at pH 7.5. Compared to unpurified samples, purification effectively reduced the initial residual DMA concentration (i.e., DMAi) to less than 1 µg/L in the 10 mg/L polymer solution, as well as the polymers’ NDMA-FP (Figure 2). However, despite of lower DMAi, polymers’ NDMA-FP was not eliminated by purification. A significant amount of DMA was released from the polymers as seen by the increase of DMA concentration after the reaction (i.e., DMAf) to levels almost comparable to those in the unpurified polymers, strongly indicating polymer degradation leading to DMA release. Separate experiments were conducted to measure NDMA formation from 50 and 200 µg/L of DMA under the same chloramination conditions used for the polymers (SI Table S1). Based on the observed NDMA yield (1.0-1.5% on molar basis), the initial DMA residues contributed negligibly to the purified polymers’ NDMA-FP, while could account for 14% and 34% of the unpurified polyamine’s and polyDADMAC’s NDMA-FP, respectively. VOL. 43, NO. 5, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1361

FIGURE 2. Effect of polymer purification (by dialysis) on NDMA formation. Ten mg/L as active ingredient of polyamines (PA) or polyDADMACs (PD) were reacted with 10 mg as Cl2/L of preformed monochloramine for 24 h at pH 7.5 and 23 °C. (A) NDMA concentration measured at the end of reaction. (B) DMA concentration measured before (DMAi) and after (DMAf) the reaction. Effect of Polymer Molecular Weight. Polyamines of three different MW ranges (lowest MW (MW < 50 000 g/mol), low MW (20 000 < MW < 80 000 g/mol), and high MW (60 000 < MW < 300 000 g/mol)) and PolyDADMACs of two different MW ranges (low MW (50 000 < MW < 200 000 g/mol) and high MW (300 000 < MW < 1 000 000 g/mol)) were evaluated. Comparison of different MW polymers allows for the assessment of the role of polymer chain ends versus polymer backbone in the overall NDMA-FP. At the same total mass and similar degrees of branching, polymers of lower MW contain shorter chain lengths and a greater number of chain ends compared to polymers of higher MW that contain longer chain lengths. As shown in Figure 3A, the highest NDMA formation was from the Lowest MW polyamine. The low MW and high MW polyamines generated comparable, but much lower NDMA than the lowest MW polyamine. PolyDADMACs formed lower NDMA than all polyamines and, in contrast, their NDMA-FP was not affected by variation in MW (low vs high). The two polymers also differed in DMA release: for polyamines, DMA concentration was increased in all cases after chloramination, whereas DMA did not increase much for either low or high MW polyDADMACs (Figure 3B). Polymer Intermediate Compounds (DMA-epi-DMA, ADMA, and DADMAC). To evaluate the effect of polymer structural moieties and other potential residues in the polymer solutions on the overall NDMA-FP, three model compounds representing intermediate compounds during 1362

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 5, 2009

FIGURE 3. Effect of polymer molecular weight on NDMA formation from polyamines (PA) and polyDADMACs (PD). Reaction conditions the same as those in Figure 2. (A) NDMA concentration measured at the end of reaction. (B) DMA concentration measured before (DMAi) and after (DMAf) the reaction. For PA: lowest MW ) MW < 50 000 g/mol; low MW ) 20 000 < MW < 80 000 g/mol; high MW ) 60 000 < MW < 300 000 g/mol. For PD: low MW ) 50 000 < MW < 200 000 g/ mol; high MW ) 300 000 < MW < 1 000 000 g/mol. polymer synthesis were investigated. DMA-epichlorohydrinDMA (DMA-epi-DMA, Figure 1), obtained by condensing the DMA-epichlorohydrin adduct (SI Figure S1) with another DMA molecule, was investigated for polyamine. Allyldimethylamine (ADMA) and DADMAC monomer (Figure 1) were investigated for polyDADMAC. DMA-epi-DMA formed significantly more NDMA than a polyamine (Figure 4A), indicating that the chain ends of a polyamine play a major role in NDMA formation since DMAepi-DMA represents an extreme case of a short-chain polyamine. PolyDADMAC’s model compound ADMA formed the second highest amount of NDMA, higher than polyDADMAC and polyamine. This result further supports that the DMAbased tertiary amine chain ends are a significant source for NDMA formation. Notably, both DMA-epi-DMA and ADMA also released much more DMA than the polymers (Figure 4B). In contrast, the DADMAC monomer formed little NDMA and had much lower DMAi and DMAf. The stability of DADMACmonomerwasfurthertestedatelevatedmonochloramine concentration (20 mg/L) at pH 7.5. More than 90% of 50 µM (6.3 mg/L) of DADMAC remained after three days (SI Figure S3), indicating that DADMAC monomer is resistant to degradation during chloramination. Because ADMA showed high NDMA-FP, several polyDADMACs were synthesized with spiking extra ADMA to DADMAC prior to polymerization (SI Figure S2) to evaluate

FIGURE 4. NDMA formation potential of polyamines (PA), polyDADMACs (PD) and their intermediate compounds (DMAEpi-DMA, DADMAC, ADMA). Reaction conditions the same as those in Figure 2. (A) NDMA concentration measured at the end of reaction. (B) DMA concentration measured before (DMAi) and after (DMAf) the reaction. whether an increase of ADMA in the polymer synthesis would lead to higher NDMA-FP. Although NDMA-FP increased with increasing ADMA-spiking in the polyDADMAC synthesis, the NDMA-FP increase was rather small compared to the amount of ADMA spiked (SI Figure S4). Furthermore, DMAi and DMAf were not affected by increased ADMA-spiking. These results showed that the spiked ADMA was not present as free residues after polymerization and lost most of its NDMA formation and DMA release tendency. Kinetics of NDMA and DMA Formation from the Polymers. NDMA formation and DMA release from the reaction of polymers with monochloramine were monitored for up to 144 h. The residual monochloramine concentration at the end of 144 h was 5.8 mg/L in the polyamine solution and 7.2 mg/L in the polyDADMAC solution. As shown in SI Figures S5 and S6, NDMA formation in both polymers increased continuously during the 6 days of reaction time and beyond, although the formation rate became slower after 5 days. In contrast, the DMA concentration change was more complex. For polyamine, DMA concentration increased continuously in the first 24 h and decreased afterward except for a second maximum at about 72 h. NDMA formation occurred at a slower rate than the release of DMA during the initial reaction period (SI Figure S5). For polyDADMAC, DMA release was less, more slowly and with more fluctuation than that of polyamine (SI Figure S6). The fluctuation of DMA concentration over time is likely caused by the different rates of polymer oxidation to release DMA and DMA oxidation to yield NDMA and other products by chloramines.

FIGURE 5. Effect of pH on NDMA formation from polymers. Ten mg/L of polymers were reacted with 10 mg of Cl2/L of preformed monochloramine for 24 h at 23 °C. (A) NDMA and (B) DMA concentrations measured at the end of reaction. Effect of Oxidant Dose. Investigation of the impact of monochloramine dose (4-13 mg/L) showed that NDMA formation from both polymers increased with increasing monochloramine dose (SI Figure S7A), and that DMAf at the end of 24 h of reaction time increased when monochloramine dose was increased from 4 to 7 mg/L (SI Figure S7B). While higher oxidant dose is expected to cause greater polymer degradation to release more DMA, the released DMA could also be degraded by excess amount of oxidant. A combination of the above two phenomena likely resulted in the comparable DMAf values at the higher oxidant dosages (7-13 mg/L) (SI Figure S7B). Overall, both polyamine and polyDADMAC did not exert a high monochloramine demand based on these results and the residual monochloramine concentrations measured after the reactions. Monochloramine dose at 4 mg/L or higher could cause degradation of these two polymers after extended contact time. Effect of pH. The effect of pH on NDMA-FP was tested on purified polymer samples to avoid interference from the impurities. NDMA formation was maximized at pH near 8 for both polyamine and polyDADMAC, and decreased when pH was lower or higher than 8 (Figure 5A). The DMA concentration measured at the end of reaction (Figure 5B) had the similar trend (highest at pH 8) as NDMA formation throughout the pH range of 5-9 in the case of polyamine. The DMAf in polyDADMAC was comparable from pH 6 to 8, and lower at pH 5 or 9. Solution pH affects the (i) degradation of polymers and (ii) formation of NDMA during chloramination. For (ii), maximum NDMA formation near pH 8 from reaction of DMA with chloramines was reported previously (6). NDMA formation is most favorable when unprotonated DMA precursors react with dichloramine (15). Since unprotonated DMA is more abundant at higher pH, whereas the dichloramine species is VOL. 43, NO. 5, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1363

FIGURE 6. FT-IR (left) and Raman (right) spectra of (A) polyamine and (B) polyDADMAC before and after 24 h chloramination at pH 7.5. more abundant at lower pH, the opposite pH demands by these two reactants resulted in a maximum reaction yield at intermediate pH range. The effect of pH on polymer degradation is discussed in the Discussion section along with the proposed degradation pathways. Polymer Structural Analysis by IR and Raman Spectroscopy. IR and Raman spectroscopy were used to investigate polymer structural changes by chloramination. Although much higher polymers (1 g/L) and monochloramine (200 mg as Cl2/L) concentrations were necessary to meet the detection limits of IR and Raman, the reactions actually involved a lower oxidant to polymer ratio than that in the kinetic experiments and thus excessive polymer degradation that did not correlate with the kinetic experimental results was unlikely. For polyamines (Figure 6A), the new peak at 1722 cm-1 in the IR spectra indicates oxidation of the OH group to >CdO group. In the Raman spectra, the cleavage of the CH2sN bond and oxidation of the OH group are proposed based on the decrease in CsN stretch (1134 cm-1) and CsO stretch (1065 cm-1). The change of CH/CH2 rocking at 572/512 cm-1 shows the backbone change. Both IR and Raman (Figure 6B) showed structural change of polyDADMAC in the CsH and CsN region after chloramination. The peak intensity at 1478 cm-1 (CH3 deformation vibration), 1383 cm-1 (CsH deformation vibration in ring), 1268 cm-1 (NsCH3 stretching vibration), and 794 cm-1 (NsCH2 stretch vibration) in IR were decreased after chloramination which can be proposed as the leaving of N(CH3)2 group. The decrease at 1480 cm-1 (CH3 asymmetric deformation vibration) and 791 cm-1 (NsC stretch vibration) in Raman can also be considered as evidence of the leaving of N(CH3)2 group. Ring opening of polyDADMAC after chloramination can also be proposed with the decrease of peak intensity at 3027 cm-1 (cyclic CH2 stretching vibration) in IR and 2934 cm-1 (cyclic CH2 stretching vibration) and 573 cm-1 (CsC skeletal vibration) in Raman. 1364

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 5, 2009

Discussion ReactionsofpolyaminesandpolyDADMACswithmonochloramine result in polymer degradation and release of DMA as demonstrated by the increase in DMA concentration after chloramination (i.e., DMAf versus DMAi) in many cases (Figures 2-4) and particularly in the studies involving purified polymers (Figure 2). The measured DMAf reflects the sum of initial DMA, newly released DMA, and loss of DMA by reaction with chloramines after the selected time period. When the NDMA formed (after 24 h) was normalized to DMAi or DMAf on a molar basis, a much more consistent NDMA/DMA ratio was obtained with DMAf than with DMAi (SI Table S1), indicating that NDMA formation correlates much more strongly to DMAf. In other words, NDMA-FP depends more strongly on the released DMA by polymer degradation than on the initial residual DMA in the polymers. This dependence is also supported by the pH effect experiments; generally, at pH values where higher NDMA formation was observed, higher DMAf was also found (Figure 5). To assess the significance of DMA release in the mechanisms of polymer’s NDMA-FP, crude estimations of DMA release from the polymers are made using the NDMA yield of 1.5% (on molar basis, SI Table S1) from chloramination of DMA as a guide. For example, the purified polyamines and polyDADAMCs formed around 1100 and 200 ng/L of NDMA, respectively, after 24 h of reaction with 10 mg/L of monochloramine (Figure 2A). This corresponded to 44 and 8 µg/L of DMA release from the two polymers, respectively, if all of the NDMA was formed from chloramination of the DMA released by the polymers. The experiments detected DMAf of 14 and 4.5 µg/L for the purified polyamines and polyDADMACs, respectively (Figure 2B), about 32 and 56% of the estimated DMA release. Such levels of DMA detection at the end of the reaction is within the reasonable range considering that (i) only ∼60% of DMA remains in the solution after 24 h of chloramination (SI Table S1), (ii) the DMAf, although related, is not equal to the amount of DMA release, and (iii) chloramination of polymers are much more complex reactions than chlorami-

FIGURE 7. Proposed degradation pathways for (A) polyamine’s tertiary amine chain end, and (B) polyDADMAC’s quaternary ammonium group during chloramination. nation of DMA. The above comparison suggests that polymer breakdown to release DMA accounts for a major mechanism of polymers’ NDMA-FP during chloramination. IR and Raman analyses show that chloramination causes structural changes of polymers including the cleavage of CH2sN bond in polyamine and the leaving of sN(CH3)2 group and ring opening in polyDADMAC that are consistent with DMA release (Figure 6). Polyamine has tertiary amine (sN(CH3)2) chain ends and quaternary ammonium groups (sN+(CH3)2s) on the backbone, whereas polyDADMAC has quaternary ammonium groups (sN+(CH3)2s) in ring structures along the polymer chain. Tertiary amines are more susceptible than quaternary ammonium compounds to electrophilic attack by chloramines (32) because of the availability of the lone-pair electrons and less steric hindrance at the N atom. Releasing DMA from tertiary amine groups is also easier than from quaternary ammonium groups because it requires breaking only one CsN bond rather than two. These rationales plus the results that the model intermediate DMA-epi-DMA releases a large amount of DMA and yields very high NDMA formation (Figure 4), and that lower MW polyamines generate more DMA and NDMA than higher MW polyamines (Figure 3) strongly suggest that the tertiary amine chain ends of polyamines are primarily responsible for their NDMA-FP. DMA-epi-DMA is an extreme case of a short-chain polyamine that contains only tertiary amine chain ends. On the same mass basis and similar degrees of branching, polyamines of lower MW have shorter chain length and thus a greater number of total tertiary amine chain ends than polyamines of higher MW. Therefore, we proposed that polyamine degrades via an attack of monochloramine on the tertiary amine chain end’s N atom, leading to an imine intermediate, which subsequently breaks down by hydrolysis to yield DMA (32) (Figure 7A). The released DMA then reacts with dichloramine to yield NDMA (15). This proposed mechanism is also consistent with the pH trend of polyamine degradation (indicated by DMAf value) by chloramination (Figure 5B). While unprotonated tertiary amine chain ends of polyamine are more reactive to oxidant than protonated ones, Cl+ transfer from monochloramine to polyamine’s tertiary amine chain end is facilitated by the presence of proton (33). Because of the opposite pH demands

by the above two factors, polyamine degradation is most significant at intermediate pH range (Figure 5B). For polyDADMAC, although the model intermediate ADMA yields significant NDMA formation and DMA release (Figure 4), the presence of an appreciable amount of free ADMA in polyDADMACs to contribute significantly to NDMAFP is unlikely based on the results of polyDADMACs synthesized with ADMA-spiked DADMAC monomers (SI Figure S4). Thus, polyDADMAC’s NDMA-FP is related to degradation of its quaternary ammonium group during chloramination. A plausible degradation pathway is that polyDADMAC’s quaternary ammonium group first undergoes a Hofmann elimination by the attack of a nucleophile (e.g., OH- or H2O) at the β-H leading to breaking of the ammonium group from the R-C (34), followed by attack of monochloramine on the resulted tertiary amine via the steps as described for polyamine to yield DMA (32) (Figure 7B). The same mechanism can also explain the results that DADMAC monomer shows little NDMA formation or DMA release during chloramination. The DADMAC monomer is resistant to Hofmann elimination due to the double bond linkage at the β-C (Figure 1). Although Hofmann elimination is favorable at higher pH, the attack of monochloramine on the resulting tertiary amine group is susceptible to the similar pH effect as that in polyamine, thus polyDADMAC degradation is greater at intermediate pH region (Figure 5B). Since the properties of polyDADMAC’s quaternary ammonium rings are not influenced by polymer MW, the result that polyDADMAC’s NDMA-FP is not affected by polymer MW is consistent with the above mechanism. Two additional factors may also contribute modestly to the overall greater NDMA-FP and DMA release of polyamines than polyDADMACs. First, polyamine has a higher mass fraction of DMA moiety per repeating unit in the polymer structure than polyDADMAC (32 vs 27%) (SI Figure S8). Because the experiments exposed the same mass (10 mg/L) of polymers to monochloramine, polyamine has higher DMA precursor content per mass basis than polyDADMAC. Second, the chains of polyamine and polyDADMAC have bond and torsional angles restricting their chain flexibility. The local stiffness of a polymer chain is usually expressed by the Flory’s characteristic ratio (C∞), which is higher in the polymer containing bulkier side VOL. 43, NO. 5, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1365

groups (30, 35). The chain of polyDADMAC is stiffer than that of polyamine due to the ring structures. The greater flexibility of polyamine chain permits easier movement in aqueous solution and better contact with monochloramine oxidant, leading to greater chain degradation and higher DMA release and NDMA formation. This study exposed polymers directly to chloramines for short to long reaction times in order to probe the mechanisms of NDMA formation. At optimum dosages and typical operation conditions at water treatment plants, most amine polymers are removed from the solution during coagulation and flocculation; therefore, very little polymers encounter chloramines. Nevertheless, the mechanisms elucidated in this study offer important guidance for future strategies to minimize the potential of NDMA formation from water treatment polymers, particularly for polyamine because of its considerably higher NDMA-FP than polyDADMAC. Minimizing the number of tertiary amine chain ends by producing polyamines of higher MW and less branching will reduce NDMA-FP. Appropriate capping of the tertiary amine chain ends may be another approach to protect their breakdown by chlorine oxidants to release DMA. Reducing the amount of residual oligomers and DMA by purification can also reduce the overall NDMA-FP of the polymers. NDMA-FP tests and assessing the tendency of free amine release of amine-based polymers should be included when developing new polymers for water treatment, domestic and/or food applications to produce safer products.

Acknowledgments This study was supported by a research grant sponsored by SNF Floerger. We gratefully acknowledge Lokesh Padhye for laboratory assistance and Dennis Marroni at SNF and Dr. Mustafa Aral at Georgia Tech for providing valuable comments for this study.

Supporting Information Available Text S1-S5, Schemes S1 and S2, Table S1, and Figures S1-S7. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) IARC. Some N-Nitroso Compounds. IARC Monographs on the Evaluation of Carcinogenic Risk of Chemicals to Humans, 17; International Agency for Research on Cancer: Lyon, France, 1987. (2) IARC. Overall Evaluations of Carcinogenicity. IARC Monographs on the Evaluation of Carcinogenic Risk of Chemicals to Humans, Supplement 7; International Agency for Research on Cancer: Lyon, France, 1987. (3) California Department of Public Health. A Brief History of NDMA Findings in Drinking Water Supplies. http://ww2.cdph.ca.gov/ certlic/drinkingwater/Pages/NDMAhistory.aspx. (4) Ontario Ministry of The Environment. Drinking Water Surveillance Program, 1996-1997, Executive Summary report. http:// www.ene.gov.on.ca/programs/3554e01.htm. (5) Choi, J. H.; Valentine, R. L. Formation of N-nitrosodimethylamine (NDMA) from reaction of monochloramine: a new disinfection by-product. Water Res. 2002, 36 (4), 817–824. (6) Mitch, W. A.; Sedlak, D. L. Formation of N-nitrosodimethylamine (NDMA) from dimethylamine during chlorination. Environ. Sci. Technol. 2002, 36 (4), 588–595. (7) Barrett, S. E.; Hwang, C. J.; Guo, Y. C.; Andrews, S. A.; Valentine, R. L. Occurrence of NDMA in Drinking Water: A North American Survey, 2001-2002. Proc. Annu. Conf. Am. Water Works Assoc. 2003, . (8) Krasner, S. W.; Weinberg, H. S.; Richardson, S. D.; Pastor, S. J.; Chinn, R.; Sclimenti, M. J.; Onstad, G. D.; Thruston, A. D. Occurrence of a new generation of disinfection byproducts. Environ. Sci. Technol. 2006, 40, 7175–7185. (9) Richardson, S. D.; Plewa, M. J.; Wagner, E. D.; Schoeny, R.; DeMarini, D. M. Occurrence, genotoxicity, and carcinogenicity of regulated and emerging disinfection by-products in drinking water: A review and roadmap for research. Mutat. Res. 2007, 636, 178–242. (10) Richardson, S. D. Environmental mass spectrometry: Emerging contaminants and current issues. Anal. Chem. 2008, 80, 4373– 4402.

1366

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 5, 2009

(11) United States Environmental Protection Agency. Integrated Risk Information System (IRIS). http://cfpub.epa.gov/ncea/iris/ index.cfm?fuseaction)iris.showSubstanceList&list_type)alpha (12) U.S. Environmental Protection Agency. Unregulated Contaminant Monitoring Regulation (UCMR) for Public Water Systems Revisions. http://www.epa.gov/fedrgstr/EPA-WATER/2007/ January/Day-04/w22123.htm (13) United States Environmental Protection Agency Contaminant Candidate List 3. www.epa.gov/OGWDW/ccl/ccl3.html. (14) Schreiber, I. M.; Mitch, W. A. Influence of the order of reagent addition on NDMA formation during chloramination. Environ. Sci. Technol. 2005, 39 (10), 3811–3818. (15) Schreiber, I. M.; Mitch, W. A. Nitrosamine formation pathway revisited: The importance of chloramine speciation and dissolved oxygen. Environ. Sci. Technol. 2006, 40 (19), 6007–6014. (16) Mitch, W. A.; Schreiber, I. M. Degradation of tertiary alkylamines during chlorination/chloramination: Implications for formation of aldehydes, nitriles, halonitroalkanes, and nitrosamines. Environ. Sci. Technol. 2008, 42 (13), 4811–4817. (17) Gerecke, A. C.; Sedlak, D. L. Precursors of N-mitrosodimethylamine in natural waters. Environ. Sci. Technol. 2003, 37 (7), 1331–1336. (18) Mitch, W. A.; Gerecke, A. C.; Sedlak, D. L. A N-nitrosodimethylamine (NDMA) precursor analysis for chlorination of water and wastewater. Water Res. 2003, 37 (15), 3733–3741. (19) Mitch, W. A.; Sedlak, D. L. Characterization and fate of N-nitrosodimethylamine precursors in municipal wastewater treatment plants. Environ. Sci. Technol. 2004, 38 (5), 1445–1454. (20) Chen, Z.; Valentine, R. L. Formation of N-nitrosodimethylamine (NDMA) from humic substances in natural water. Environ. Sci. Technol. 2007, 41 (17), 6059–6065. (21) Kimoto, W. I.; Dooley, C. J.; Carre, J.; Fiddler, W. Role of strong ion exchange resins in nitrosamine formation in water. Water Res. 1980, 14 (7), 869–76. (22) Child, P. ; Kaa, G. ; Benitz, D.; Fowlie, P.; Hong-You, R. Reaction between chlorine and a dimethylamine containing polyelectrolyte leading to the formation of N-nitrosodimethylamine. http://www.pollutechgroup.com/papers/p11.htm. (23) Najm, I.; Trussell, R. R. NDMA formation in water and wastewater. J.- Am. Water Works Assoc. 2001, 93 (2), 92–99. (24) Wilczak, A.; Assadi-Rad, A.; Lai, H. H.; Hoover, L. L.; Smith, J. F.; Berger, R.; Rodigari, F.; Beland, J. W.; Lazzelle, L. J.; Kinicannon, E. G.; Baker, H.; Heaney, C. T. Formation of NDMA in chloraminated water coagulated with DADMAC cationic polymer. J.- Am. Water Works Assoc. 2003, 95 (9), 94–106. (25) Kohut, K. D.; Andrews, S. A. Polyelectrolyte age and Nnitrosodimethylamine formation in drinking water treatment. Water Qua.l Res. J. Can. 2003, 38, 719–735. (26) Najm, I.; Teefy, S.; Barrett, S. E. NDMA formation from water treatment polymers. In Proceedings of Annual Conference of the American Water Works Association; Orlando, FL, 2004. (27) Charrois, J. W. A.; Hrudey, S. E. Breakpoint chlorination and freechlorine contact time: Implications for drinking water N-nitrosodimethylamine concentrations. Water Res. 2007, 41, 674–682. (28) Kohut, K. D.; Andrews, S. A. Polyelectrolytes as amine precursors for N-nitrosodimethylamine in drinking water. Proc. Annu. Conf. Am. Water Works Assoc. 2003, 405–418. (29) Schwoyer, W. L. K. Polyelectrolytes for Water and Wastewater Treatment; CRC Press, Inc.: Boca Raton, FL, 1981; pp 26-46. (30) Shalaby, S. W.; McCormick, C. L.; Butler, G. B. Water-Soluble polymers: Synthesis, solution properties, and application. In ACS Symposium Series 467; American Chemical Society: Washington, DC, 1989; p 26-56. (31) Jin, F.; Hu, H. Y.; Yang, M.; Jin, X. H.; He, W. J.; Han, H. D. Determination of diallyldimethylammonium chloride in drinking water by reversed-phase ion-pair chromatography-electrospray ionization mass spectrometry. J. Chromatogr., A 2006, 1101 (1-2), 222–225. (32) Armesto, X. L.; Canle, L. M.; Garcia, M. V.; Santaballa, J. A. Aqueous chemistry of N-halo-compounds. Chem. Soc. Rev. 1998, 27, 453–460. (33) Snyder, M. P.; Margerum, D. W. Kinetics of chlorine transfer from chloramine to amines, amino acids and peptides. Inorg. Chem. 1982, 21, 2545–2550. (34) Boyd, M. Organic Chemistry, 5th ed.; Allyn and Bacon, Inc.: Newton, MA, 1987. (35) Rubinstein, M.; Colby, R. H. Polymer Physics; Oxford University Press: New York, 2003; pp 49-96.

ES802732Z