Methyl Tert-Butyl Ether (MTBE) Degradation by Ferrous lon-Activated Persulfate Oxidation: Feasibility and Kinetics Studies K. E. Che&*, O. M. Kao2, L. O. Wu3, R. Y. Surampalli4, S. H. Liang2 K’TRACT:

Thc objcctivc of this study was to cvaluatc thc fcasibility of Ierruua ¡on-acuvated persulfate oxidrnion lo remediate groundwaler :.rninated with meíhyl tert-butyl ether (MTBE). In this sntdy, batch nments we conducted :o evaluate the effeccs of various factors on (he ency of MTBE degradation including persulfate cor.cenn-ations, feffous oncentrations, and persulfate coupled with hydrogen pci-oxide Restilis .hat ferrous ion-activated persulfate oxidation was capable of rading MTBE efflcicntly. Persulfatc and fcrrous ion concentrations rated with MTBE degradation rates. However. excess additin of -us ion rtsultcd in decreased MTBE degrading rates mosi likely because omoet ron for sulfate free radicais betwcen fen-ous ion and MTBE. Two .on byproducts of MTBE degradation, tert-butyl formate asid tert-butyl ohol, were detected in the experimente; both were. however. subsequently racIed. Resulta of sulfate analysis show that proper addition of ferrous n could prevent unncce.ssary persulfate decomposition. Warer Environ. s., 81, 687 (2009). KEYWORDS: Methyl teri-hulyl eher (MTBE), persulfate. ferrous ion, Fenton’s reaction, in situ chemical oxidation (ISCO). doi: 10.21751106143008X370539 Introduction Oxygenates such as ethers asid alcohois usually are introduced into gasoline as additives to replace lead as an octane index cnhancer and to improve air quality. Arnong dic oxygen-containing compounds. meihyl tert-butyl ether (MTBE) is the most commonly used gasoline additive because of its low COSI and case of produc( ion asid blending (Fayolle eL al.. 2001; Kharoune el al., 2001). llowever, bccausc of the large amowit of MTBE used, contamin ation has occurred in groundwater, surface water, and air (Juhier asid Felding, 2003; Ayotte el al., 2005; Kolb and Püttmann. 2006; Toran et al., 2003; Schmidt et al., 2004; RoselI es aL, 2006; Guillard ct al., 2003; Lin el al., 2005). Methyl tert-butyl ether has buen demonstrated lo be an animal carcinogen. As a result, the U.S. Environmental Protection Agency (U.S. EPA) has temporarily classified MTBE as a possible human carcinogen (U.S. EPA, 1997).

Methyl tert-butyl ether is an organic chemical with high water solubility and low adsorption to soil. In addition, MTBE is not readily biodegradable because of the

tertiary butyl group and ether linkage on it. Thus, it usually migrates a longer distance than other gasoline components such as benzene, toluene. ethylbenzcne, and xylenes (BTEX) (US EPA, 1998; Braids, 2001; U.S. EPA, 2004). For this reason, more active measures may be required to control the movement of MTBE plumes at gasoline-contaminated sites. Chemical oxidation technology is an effective, potent groundwater remedial option capable of breaking down many contaminants in water. Hydrogen peroxide. Fenton’s reagent, permanganate, persulfate, and ozone are commori oxidants uscd lo remediase organic pollutante (Danim es al., 2002; Mitani es al., 2002; Burbano et aL, 2005). Recently. persulfate oxidation has been proposed for in sites chemical oxidation OSCO) proccsscs [Intenstate Technology & Regu!atory Council (rrCR), 2005J. The redox potential of persulfate is about 2.01 V. which is higher ¡han nhose of hydrogen pcroxide asid permariganase but lower ihan those of hydroxyl radicais (OH) and ozone (Huang el al., 2002; ITRC, 2005).

Persulfate can be activated thermally or chemically by initiators such as heat or transition mctals (cg., Fc2+) to produce more powerful sulfate free radicals (S04’) (l3lock eL al., 2004):

S2O8 + initiator —y 2S04 ‘(or SO4 + SO4 ) (1)

S04- + e — SO4 E° = 2.6V (2)

Sulfate free radicals are capable of degradinorganic compounds so carbon dioxide and water. For example, degradation of MTBE by sulfate free radicals can be illustrated as reaction 3:

30SO - + C5H120 +91120 — 5CO + 30H ± 30SO (3)

Generally, ferrous ion is a commonly used initiator for persulfale activation (Liang et aL, 2004a):

Fe24+S20[ —. Fe3+SO+SO. (4)

k = 2.0 X 1& M s (Travina et al., 1999)

k = bimolecular Tate constant.

SO4- +Fe2 —‘ Fe3+SO (5)

k 4.6 X M_’ s (Buxton el al., 1997)

From reactions 4 asid 5, overall reaction is obtaincd as reaction 6:

2Fe2 + S20r — 2Fe3 -f 2S0 (6)

k = 3.1 X M1 s (Buxton es al., 1997)

Chen et al.

Pcrsulfate can exisI longer than other oxidants (such as 03 and H202) resulting in a greater transpon distance iii (he subsurface. Ui addition, because persulfate reacis less ith natural organic matters (NOM), it appears to be more appropriatc for the remediation of contaminated aquifers conlaining high NOM (Brown et aL, 2003; Block et al., 2004; ITRC, 2005). Dose demand of persulfate would be less because of ita lower affinity for NOM and its longer retenhion lime in the subsurface. As a result, thc cost of using persulfatc oxidation may be lower compared lo other oxidants, making it a promising alternative in site remediation. Persulfate activated by ferrous ion or heat has been used to treat groundwater contaminants including chlorinatcd organics and petroleum hydrocarbons (Anipsitakis and Dionysiou, 2004; Liang el al., 2004a, 2004b: Huang et aL, 2005). Although degrading etiiciency is lower, polycyclic aromatic hydrocarbons, explosives, and pesticides could also be oxidized via persulfate (ITRC, 2005; Rivas, 2006). Recent studies have focused on kmetjcs and efñc iency of Lrichloroethylene (TCE) removal by persulfate under various environmental conditions (Liang et al.. 2004a, 2004h: Liang et al.. 2006; Liang eL al., 2007). However, few studies have cvalu ated dic feasibility of using pcrsulfate for tite degradation of petroleuni hydrocarbons such as MTBE and BTEX (Huang el al., 2002; Block el al., 2004; Huang et al., 2005; Liang et al., 2008). Huang el al. (2002) reported that cifective removal of MTBE and its degrading intermediates —including TBF, TBA, acetone, and mcthyl acetate—was achieved via heat-assisted persulfate oxidat ion. Block eL al. (2004) also showed that more than 90% of MTBE was decomposed by persulfare oxidation at 35°C. In addition, 80 lo 98% of MTBE was degraded by fcrrous iron-aclivated persulfate al a pH of about 2. Kinetics of MTBE removal by iron-activated persulfate has nol been studied. lii ibis study. batci expenimenis were conducted to evaluatc ihe various controlling factors—diffcrent persulfate and ferrous ion conccntrations, nature of solution (deionized water or grouridwater), pcrsulfate couple.d wjth hydrogen peroxide—on the efficiency of MTBE degradation. The main ohjectives of ibis study were to (1) evaluate the efficiency and kinetics of MTBE degradation via persulfate oxidation under different conditions; (2) evaluate the feasibility of using ferrous ion-activated persulfate oxidanon lo 688

remediate MTBE-contaminatecj groundwater; (3) assess (he feasih ility of MTBE removal by pcrsulfatc couplcd with hydrogen peroxide; and (4) determine dic formation and reniova] of byproducis, tert-butyl forrnate (TBF) and tert-butyl alcohol (TBA), during MTBE degradation. Materjais and Methods Chemicais. In ibis study. MTBE (99.97%, TEDIA Company Inc., Fairfleld. Ohio) and sodium persulfate (Na2S,0g, > 99%; Riedel-de Haen, Germany) were selected as dic target compound and dic oxidani, respectively. Ferrous sulfate (FeSO4. 7H20.

99.5%; Riedel-de Haen, Germany) was used lo activate persulfate. Batch Kinetic Experiments. Volatile onganic analysis viuls (40 tnL) werc used as batch reactors. Ten mgjl. (0.114 mM) of MTBE was prepared in a 2-L bottle previously and tIten ferrous ion was introduced into tIte botile followed by tIte addition of pers ulfate. Each 40-ml vial wa.s tIten fihled with dic prepared solution via a dispenser and then placed at 25 ± 1°C. Theexriments of persulfate coupled with hydrogen peroxide were conducted by the same procedure with simultaneous addition of hydrogen peroxide and pcrsulfate. Control experimentx without the addition of pers ulfate or hydrogen peroxide also were conducted. No disturbance was applicd to tIte batch rcactors lo simulate nearly static conditions in aquifers. During tIte experinients, sainples were collected afler various reaction times. Typically, most of the expenimcnts were stopped when MTBE, TBA, and TBF wcre completely removed. TIte efficicncy of MTBE rernoval was obtained by calculating tIte pcrcentage of MTHE consumption al tIte end of tite experimenis. AlI batch kinctics experirneuta were conducted with duplicates. Two aqueous media including deionizcd watcr (typically used) and groundwatcr were applied Lo evaluate tIte efficiency of persulfate oxidation on MTBE degradation. TIte initial pH values of deionized watcr and groundwater were 5.5 and 7.0, respcctively. Table 1 shows tite paramclers of each experimen of MTBE oxidation. Analytical Methods. Organics including MTBE, TBF, and TBA samples were pretreated by a headspace aulosampler (G1888 Network Headspace Sampler, Agilent Technologies. Santa Clara, California), and then analyzed using gas chromatography (6890N, Agilent Technologies) cquipped with (lame ionization detector

Resul Per: TBE MTBE (mM) NaaSzOa (mM) Fe2 (mM)

MTBE/S2082/Fe2 (molar ratio)

Water Environment Research, Volume 81, Number 7 Jul

Table 1—Parameters of Factors TriaIs 1. Persulfate

methyl tert-butyl

ether (MTBE)

oxidation experiments.

H202lPersuifate (molar ratlo)

0.114 1.13 3.57

5.69 1/10/31 0

11.39 1/50/31

34.16 1/100/31

56.93 1/300/31

11.39 0 1/500/31 1/100/0

2 Ferrous iOn

0.114 11.39 1.13

0.36

1.79

3.57 11.43 0.18

1/100/3.1 1/100/15.5

1/100/31 1/100/100 1/10/1.55

0

3. Persulfate with 11202

0.114 11.39 3.57 1/100/31

1.18 mM o! H202 only 0.01/1 (H202 = 0.11 mM)

4. In situ grou9dwater

0.114 11.39 1.79 1/100/15.5 (H02 = 1.18 mM) 0

Chen et al.

.pillary o1umn (HP-5, 5% phenyl rnethyl siloxane, 30 m X :s X 0.25 m, Agilent Technologies). Oven temperature - ‘tarted at 50°C for 1 minute, and then increased 5°C per minute tinal temperature of 80°C md hcld for 1 minute. A split ratio • 5 v.as used, and the injector and detector temperatures were maintained at 250°C. Method detection limits of MTBE, TBF, TBA were 0.01. 0.04, and 0.05 mg/L, respectively. Sulfate was performed by an ion chromatography (ICS-1500. ‘nex Corp.. Sunnyvale, California) equipped with an analytical .arnn (lonPac AS 14, 4 X 250 mm, Dionex). A dissolved oxygen :ter (ORION Model 1810A+, Thermo Fisher Scientific Inc., .I:ham, Massachusetts) was used to measure dissolved oxygen. dation-reduction potennal (ORP) and the potential of hydrogen H were measured by an OR.P meter (ORION Model 250A+, rmo Fisher Scientiflc) and a pH meter (TES 1380, Taiwan), pectiveIy. Results and Discussjon Persulfate Activation by Ferrous Ion. Tn a preliminary test, \ITBE oxidaflon by persulfate with and without the addition of rrous ion was conducted to determine ihe effectiveness of ferrous on on persulfate activation. ‘lhe initial MTBE concentrations in the -:actors wcre approximately 0.114 mM, and ferrous ion of 3.57 mM -vas applied. Results show that MTBE was removed completely hin 78 hours in the presence of ferrous ion; 7% uf M’I’BE was .legraded in ¡he system with persulfate alone (data not shown). No \ITBE degradation was observed when only ferrous ion md MTBE ‘sted in ¡he reactors. Results reveal that the oxidative power of persulfate was enhanced by ferrous ion because of production of ulfate free radicals (Liang et al., 2004a).

Subsequent experiments were conducted with ¡he addition of ferrous ion to accelerate MTBE oxidation. Effects of Persulfate Concentrations. Figure 1 presents ¡he resulta of MTBE degradation under different persulfate concent rations. As shown in Figure la, complete MTBE oxidation was accomplished with the molar ratios of MTBE/S2O82/Fe2 ranging from 1/50/31 to 1/500/31. However, no MTBE degradation was observed with a MTBE/S2O82/Fe2 molar ratio of 1/10/31 during 270 hours of reaction time. Theoretieafly, 15 mole of persulfate is capable of degrading one mole of MTBE based on (he stoichiometiy illustrated below: l5S2O + C5H20 + 9H20 —.. 5C02 + 30H + 30SO (7) Although (he MTBE/SaOg27Fe2 molar ratio of 1/10/31 was not sufficient for (he complete degradation of MTBE, sorne rcmoval was expected to occur according lo reaction 7. Nevertheless, no MTBE degradation was observed under this molar ratio. The competition of ferrous ion for sulfate free radicals (reaction 5) was responsible for (he ineffectivcness of MTBE removal because the ratio of fcrrous ion to persulfate iii this experinient was much higher than those of other four experimenta (Table 1) (Liang et aL, 2004a). The kinetics of contaminant removal using persulfate oxidation is commonly proposed as a pseudo-first-order reaction (Huang et al., 2002; Huang et al., 2005; Liang el al., 2007). Resulta revealed that the pseudo-first-order reaction rate constants (k) of MTBE/S2082/ Fe2 molar ratios of 1/50/31, 1/100/31, 1/300/31, md 1/500/31 were 1.2 X l02, 4.4 X 10, 8.4 X 10—2, md 20.8 x 10-2 h, respcctivcly. As shown in Figure ib, persulfate concentration has a linear correlation with MTBE degradation rate. l-ligher persulfate concentration caused more efficient MTBE degradation. However, (he MTBE reaction rates were Jess (han the rates in an MTBE

a 0.8 0.6 a) 0.4 0.2 o Tiae (h)

—fr— MTlWJPersuJlaleJFe(11) = 1110/31 —.--- MTBE/PersuffiutefFe(11) 1/50/31 TBE/PersuIfine/Fe(ll) = 1/100/31 ---- MTBEIPersuiteJFe(I1) = 1/300/31 -*- MTBE/Persul6ste)Fe(ll) =11500/31

Figure 1—Methyl tert-butyl ether (MTBE) degradation under different persulfate concentrations: (a) normalized MTBE concentrations versus reaction time ([MTBE] 0.114 mM; [Fe29 = 3.57 mM); (b) pseudo-first-order reaction rate constants (k) of MTBE oxidation ([MTBEJ 0.114 mM; [Fe29 = 3.57 mM). oxidation study conducted with thermal activation under similar conditions (Huang ct al., 2005). This rnay bc bcuuse of Llie undisturbed sysrem of this study, which resulted in lower reaction rete constants. Effccts of Ferrous ion Concentrations. Figure 2 shows (he resulis of MTBE degradarion under different rrous ion conccnt rations. Results revealed hat MTBE could be totally consumed with Ihe molar ratios of MTBE/S2Og2/Fe2 ranging from 1/100/ 3.1 ta 1/100/10(1 (Figure 2a). Additionally. no significant differe nces were observed in required MTBE oxidanon time with ¡he molar ratios of MTBE/S2O82/Fe2 between 1/100/3.1 md 1/100/ 15.5. This indicates that persulfate activation could be achieved even

with (he addition of a low concentration of ferrous ion. However, MTBE oxidation rete decreased significantly at MTBE/ S2O2 /Fc2 molar ratio of 1/100/100. This could because high concentrations of fermus ion caused competition for sulfate free radicals between ferrous ion md MTBE (reactions 3 md 5) (Liang el al., 2004a). According to available literature, ¡he concentration of ferrous ion betwccn 100 mg/L md 250 mg/L was suggested for persulfate activation (Block et al., 2004; ITRC, 2005). Removal efficiency of contaminants would drop because of (he rapid decomposition of persulfate if the conceatration of ferrous ion exceeded 750 mg/L. Thus, resuhs of this study were similar to other reports (Block et al., 2004; ITRC, 2005). Figure 2b presenta ¡he pseudo-first-order reaction rete constants (k) of MTBE oxidation under various fenous ion concentrations. The reaction rete constants with thc molar ratios of MTBE ,O82/Fe2 of 1/100/3.1, 1/100/

1.2

Tvre (h) —a-- ff BEfPersulfitejFe (II) • 1110131 —.— MffiEfPerujtteJFe (II) 1/10/1.5 (e) Figure 2—Methyl tert-butyl ether (MTBE) degradation under dilfereni ferrous ion concentratjons: (a) normaljzed MTBE concentratlons versus reaction time ([MTBE] 0.114 mM; [Na2S2O8] = 11.39 mM); (b) pseudo-fjrst-order reaction rate constants (k) of MTBE oxidation ([MTOE] 0.114 mM; [Na2S2O8] = 11.39 mM); (c) normailzed MTBE concentratjons versus reaction time under high and approprjate ferrous ion addition ([MTBE] 0.114 mM; [Na2S2O8] = 1.13 mM). 15.5, 1/100/31, and 1/100/100 were 2.3 X lCr2, 35 X 10—2, 4.4 X lCr2, and 1.7 X 10’ h, respectively. As shown in Figure 2b, ferrous ion concentration had a linear correlatjon with MTBE removal with ferrous ion concentratjons of 0.36 to 3.6 mM. The molar ratio of MTBE/S2Os2’/Fe2 of 1/100/loo resulted m the dala point outlier. As disctissed, the overdose of fcrrous ion resulted in a lower MTBE degmdation rate bccause of competition for sulfate free radicals between ferrous ion and MTBE. Results from the experiments indicated thai molar ratios of S2052/Fe2 between 1/0.031 and 1,i3l dcmonstratej good MTBE degradation efficiency. Resulis ¡it Figure la show thai 690 r —..— 1.18 mM H202 addition —a- 0.11 mM,additjJ —.—- without H202 —.-. Fenton’s reaction Figure 3—Methyl tert-butyl ether (MTBE) degradation uslng persulfate coupled wlth hydrogen peroxide ([MTBE] 0.114 mM; [Na2S2O8] = 11.39 mM; [Fe29 = 3.57 mM; Fenton’s reaction: [H202] : 1.18 mM + [Fe2’]: 3.57 mM).

0 30 Tuw(h) 60 90

o so too 150 200 250 3(10 Tune (h) —a-- MTBEcuJJ (U) 1/100/3.1 -.— M EE/PefsLIJlae/Fe (It) 1(100/15.5 - MiTJEPersijIfa,,JEe (11)- 1/100/31 -a-- I.fl’BEI SLJIr.atc/Fe (U) = 1!I I00 (a)

‘ 0.02 y=0.0064x+0,0217 R2 0.9756 0.0I o 0 2 4 6 8 10 12 14 Fe(]I)(n)

MTBE was not degraded with a MTBE/S2Og2”,?e2 molar ratio of 1/10/31 (S2082jFe2’ molar rano 1/3.1). This likely was because of ___________________________________________ ¡he effect of the high rabo of ferrous ion thai was added to the system. which would reaci with sulfate radicals and decrease dic oxidation power. To tesi ihis hypothcsis, a lesi with MTBEJS2O82”/ Fe2 molar ratio of 1/10/1.55 (S2O827Fe2 molar ratio of ljO.155) was conducted for companson. Figure 2c presents tIte companson of MTBE degradation under high and appropriate ferrous ion addition. More iban 60% of MTBE was removed within 430 hours with S2O52/Fc2 molar ratio of 1,). 155, and complete degradation was observed after 912 hours (38 days) with a reaction rate constant of 0.5 X 10-2 h’. However, no MTBE degradation was observed with S2O82/Fe2 molar ratio of 1/3.1 during a 460-hour reaction time. Resulis of dic cxperiment suggested thai exceas addirion of ferrous ion was tIte major factor inhibiting MTBE degradation in tIte 0 200 400 600 800 1000 System. , Accordtng to reacnon 7, 1.14 mM of persulfaic thcorctically ¡a capable of dcgrading 0.076 mM of MTBE. However, 0.114 mM of MTBE was completely removed in tIte test with MTBE/S20827 Fe2 molar rabo of 1/10/1.55. The degrading efficiency of MTBE was approximately 1.5 times higher tItan tIte theoretica1 value. un fact, the oxidation of contaminarns in a persulfate system involves a complicated chain reaction including ¡nitiation, propagation, and temljna,jon reactions (House, 1961). It has been reponed that if the chain teiniination reactions are slower iban chain propagation processes, ihen even small amounts of sulfate radicais can result in significant contamjnant removal. For this reason, consumplion of persulfate for contajnjtant treatment will be lower than tIte stoichiometrjc quantities (ITRC, 2005). Although stoichiometric equations such as reaction 7 were commonly used to predici dic reaction of pcrsulfate with contaminants, results from ibis study suggesi thai it is necessary to construct a more detailed mechanism of dic reaction between persulfate and contanhinants (ITRC, 2005; Liang ci al., 2007), Resulta also showed thai MTBE could be degraded effcctively with appropriate ferrous ion addition even if luw concentraijon of persulfate was applied. Therefore, molar ratio of S2O82fFc2’ ¡it a persulfate oxidation system was dic most importani controlling factor to achicve effective MTBE removal. Removal by Pcrsulfate and Hydrogen Peroxide. Hydrogen peroxide was added into batch reactors to enhance dic efficiency of MTBE oxidation because it can activate persulfate (8100k et al., 2004). Figure 3 shows tIte results of MTI3E degradation by

Water Envlronment Research, Volume 81, 4imbor 7 Ju

•1•

- --:fate coupled with hydrogen peroxide. Tite addition of 1.18 ‘.1 Df hydrogen peroxide enhanced MTBE removal efficiency, sugh 0.11 mM of

hydrogen peroxide showed no significant ovcment. This indicates thai 0.11 mM of hydrogen peroxide not sufficient to improve persulfate activation. Because ferrous -: ‘.as present in tite system, formation of hydroxyl free radicais uld generale more sulfate free radicais Lo accclcrate MTBE radaüon according lo reaction 8 (House, 1961): OH + S2O —. SO; ÷ HSO ÷ 1/202 (8) In addition, formation of hydroxyl free radicais also would ttribute ro MTBE consumption. Block er al. (2004) reponed - synergistic attributes caused by hydrogen peroxide and pera te would enhance removal of contaminants. Furthennore, heat rerated from tite decomposiuon of hydrogen pemxide may also mally activate persulfate (Waldemer et al., 2007). Vhen 1.18 mM of hydrogen peroxide was added alone with rrous ion of 3.57 mg/L (Fentoia’s reaction), more titan 80% of ITBE was degraded efficiently within 15 minutes. However, :.irther MTBE removal was not observed because of complete con. umption of hydrogen peroxide in tite system. It has been reported :hat two stages of MTBE decomposition were observed iii Fenton’s action (Sun asid Pignatello, 1993; Xu et al., 2004). En the first tage, MTBE is degraded rapidly with large amount of hydroxyl ree radicais, which are generated from quick reaction of fcrrous ions and hydrogen peroxide. lxi tite second stage, a lower rate of \ITBE degradation occurred because ferric ions react much more slowly with hydrogen peroxide than fermus ions. Thus, if hydrogen peroxide was still in tite system, continuous but slow MTBE degradatioti should be observed because of the catalyst of ferric DflS. In Ihis study, however, MTBE was not further degraded until ihe end of tite experimeni. Results revealed thai hydrogen peroxide was no longer present in tite system. Low conceitiration uf hydrogen peroxide (1.18 mM) was added to enhance persulfate oxidation in this study. Thus. Fenton’s reaction was conducted with 1.18 mM of hydmgen peroxide lo compare ro the combined persulfate/peroxide system. En this study. although MTBE degradation vEa Fenton’s oxidation was much faster titan the combined persulfate/peroxide system, it could fbi remove MTBE completely. Thus. persulfate in the combined systern could be used to remove rhe residual MTBE. Thc resulta also indicate, huweer, that Fenton’s reaction was stallcd because of persullale. lii fact, using Fenton’s system with higher concentrations of hydrogen peroxide can improve tite efficiency of MTBE removal. However, in tite case of improving a persulfate system, application of Fenton’s reaction with proper hydrogen peroxide concentrations (cg., 1.18 mM) followed by persulfate oxidation is an appropriate ISCO combination lo readily and completely degrade contaminants. [)egradation of Oxidation Byproducts. Figure 4 shows Lhc variation ni MTBE and jis byproducts, TBF and TBA, chining the oxidation prccesses. Firsts TBF. then - TBA were produced while MTBE degraded. Huang et al. (2002) and Xu el al. (2004) proposed a similar degrading pathway of MTBE by hydroxyl or sulfate free radicais. ‘[he Lwo stuclies also reported thai accione asid methyl acetate were formed duning MTBE degradation process (Huang et al.. 2002; Xu et al., 2004). Table 2 summarizes the results of MTBE oxidation. As shown En Table 2, TBA was more persistein titan TBF m tite system. In addition, complete degradation of IBA was not observed with low concentration of oxidani (MTBE/ Og2/Fe2 = 1/50/31 asid 1/10/1.55) or high concentration of

Tiire (it) Figure 4—Oxidation of methyl tert-butyl ether (MTBE) and its byproducts,

tert-butyl formate (TFB) and tert-butyl alcohol (TBA) (MTBE/S2082/Fe2 = 1/10013.1; [MTBE] 0.114 mM; LNa2S2O8] = 1 1.39 mM; [Fe2J = 0.36 mM). ferrous ion (MTBE/S2Og’/Fe2 1/100/31), This indicates that higher oxidant concentration and proper ferrous ion addition are necessary to achieve complete TBA removal. Moreover, resulis of ‘FRA degradation from MTBE/S2O82’7Fe2 molar ratio of 1/10/ 1.55 also suggest that persulfate was able to sustain tite oxidation poer in tite system more than 43 days because TBA concentration dropped from 4.0 mg/L. on day 43 (dala not sitown) Lo 0.6 mg/L on day 96 (Table 2) Continuous oxidarive ability of persulfate occurred in thc systcm afier a 1 .5-nionth oxidation period. Oxidation Efliciency In Groundwater. Uncontaminated groundwater spiked with MTBE was used fon funiher evaluation Lo determine tite efficiency of itt situ persulfare oxidation. The initial conditions of the groundwater were: pH = 6.99; dissolvcd oxygen = 3.0 mg/L; ORP 459 mV; and total organic carbon (TOC) = 2l. mgfL. As shown m Table 2, complete MTBE depletion ni groundwater was observed within a 11 6-hour persulfate reaction (98 hours in deionized water) with a firsi-order reaction rate constant of 3.1 X 10-2 h (3.5 X 10 h m deioiiized water). In addition, byproducts of MTBE degradation (TBF asid TBA) also could be degraded completely En groundwater within a similar oxidation time to thai m deionized waten (Table 2). Although TOC m gnoundwater was highcr titan tite concentraon of MTBE, little effect was observed on removal of MTBE and its byproducts ni tite tesis with groundwater addition. This likely is because the neaction between persulfale asid NOM is limited (ITRC, 2005). Tite slightly decrease in tite MTBE removal ratc also may result from tite presence of natural radical scavengers En groundwater (e.g., carbonate, bicarbonate, chlorides) (JTRC, 2005; Liang et al., 2007). Effects of Persulfate Oxidation on Water Quality Parameters. In Ihis study, two aqueous media including deionized wtuer (typically uscd) and groundwater weic applied Lo evaluate lite efficiency of persulfate oxidation on M’lHE degradation. The initial conditions of dcionizcd water were p11 5.5, dissolved oxygen = 3.5 mg/L, asid ORP = 459 mV (see previous section for groundwatcr). Overail, ORP increased dramatically (more iban 700 to 800 mV) because of production of sulfate free radicais. In addition, pH values in deionized waten dropped rapidly from 5.5 to less than 3.0 (2.3 ro 2.9) by the end of ah experiments except for those wiih tite MTBEIS2O82/Fe2 molar rano of 1/10/1.55 (pH = 3.4) (Table 2). Although groundwater showed a stronger buffering capacity duning the first 70 hours of reaction (data foL shown), tite final pH in groundwater also dropped to 2.8 itt tite reactors. ‘[he results indicate thut thc application of in airo pcrsulfate oxidation

Chen el al.

Table 2—Results of methyl tert-butyl ether (MTBE) oxidation by persulfate [TBF = tert-butyl formate; TBA = tert-butyl alcohol; ORP = oxidation-reduction potential].

1 Nol available,

has the potential lo lower the PH of groundwater. In a persulfate oxidation systcm, high concentiations of sulfate rnay be produced to form sulfuric acid. Although not as severe in aquifers because of the buffering capacity of aquifer sedimenis. reductinns in pl-I

should still be considered when persulfate oxidation is applicd in situ (ITRC, 2005). Liar.g ct al. (2007) have investigated TCE degradad on by persulfate oxidation under various PH values. The results suggested that maximum TCE degradation occuned al pH 7. The TCE degradation ratea were higher at pH 9 than at pH 4. Becausc pH values were nol controlled in this study, further studies need Lo be conducted tu determine the optimum pH for M’lBE degradation. Dissolved oxygen did not noticeably increase except for whcn hydrogen peroxide was added. Studies have reported that the hydrolysis of persulfate in alkaline, neutral, and dilute acid solutions could gcnerate oxygen (Kolthoff and Miller, 1951): s2o[ + 1120 —. 2HSO; + 1/202 (9) However, based on the resulta of pH measurement (pH < 3), conditions in the batch rcactors were unfavorable for oxygen generalion. Hydroxyl frcc radicais produced from the reaction of sulfate free radicais with water (reaction 10) or hydroxyl ion (reaction 11) potentially could generale oxygen (Al-Ananzeh eL aL, 2006). S0 -t-H20 — •OH+HSO; (10) k [H20j <2 X l0— s (Pennington and Haim, 1968) SOL. 0H —‘ II0.+SO (11) k = 7.3 X 10’ Ms’ (Chawla and Fessenden, 1975)

Nevenheless, reaction 10 has a low reaction rate constant whi]e reaction 11 requires alkaline conditions te increase Ihe production of hydroxyl fice radicais. Thus, oxygen generation via hydroxyl free radicais was not prcdominant in tlie sysLem. The presence of hydrogen peroxide in the system secms to be another cntical source of oxygen. Generally, hydrogen peroxide eould be produced from monopersulfuric acid (H2S05), which is fonned from the decorup osition of persulfate in strong acidic solutioris (> 0.5 M HCIO4) (Kolthoff ancl Miller, 1951): H,S205 + H20 — H,S05 + H2S04 H2S03 + H20 — H202 + H2S04 Although pH values in the reactors were less than 3, it did not meel conditions for predominant hydrogen peroxide production. This resulted in unfavorable effects on oxygen production. Based on dic aboye discussion, significant oxygen generation would not take place in dic systems. In addition, fleid applications also showed that dic incrcasc of pressure in injection wells was not observed, indicating a limited release of gases from persulfate (ITRC, 2005). Results of this study were in accordance with dic data obtained from the fleid demonstratjon, Production of Sulfate during Oxidation. Final sulfate conc entratiotis were analyzed in dic experiments because sulfate producnon is related to persulfate decomposition and consumption. Experimental resulta reveal thai no significant differenccs in sulfate production were observcd under various MTBE concentrations except for dic MTBE/S2Og27Fe2 molar ratio of 1/10131 (Table 2). However, ferrous ion caused a significant effect on final sulfate

MTBE/S2082jFe2 Trials (molar ratio)

(12) (13) 692

Water Envlronment Research, Volume 81, Number 7

time of time of time

complete complete Final

MTBE TBF

degradation degradatiori

complete Final

dissolved

Final

(hours) (hours) degradation (hours)

Final pH

ORP

(mV)

oxygen (mg/L)

S042 (mgíL)

x 102

(h)

(rio dogradatlon)

1/50/31 260.0 141.3

> 338.0

1/100/31 77.5 72.5 170.0 2.6

778

3.5

1.197

1.2

1/300/31 51.6 51.6 97.8 2,5

839

3.5

1.085

4.4

1/500/31 20.3 29.3 55.8

2.3 2.5

844 893

4.5 3.9

,614 8.4

2

1/100/3.1 93.8 125.5

149.0 1,291

20.8

1/100/15.5 98.0 110.0

173.5 2.9

825

3.3

219 2.3

1/100/31 77.5 72.5 170.0 2.4

770

5.6

751 3.5

1/100/100 241.0 40.5 > 2.5

839

3.5

1.085

4.4

1/10/1.55 912.0 864.0

> 2,304.0

2.5 3.4

812 477

3.6 4.6

2,807 171

1.7

3

HO2 1.18 mM > 268.0 (e 80% removal)

49.0 > 268.0

2.6

467

5.3

— 0.5 61.2

1/100/31 775 72.5 170.0

without H202 2.5

839

3.5

1,085

4.4

1/100/31 81.0 166.0

> 169.0

with I-fO 0.11 mM

2.3

741

5.2

1,256

3.4

1/100/31 47.0 47.0 167.5

wit*i H202 1.18 mM

596

5.6

1,550

166

4

1/100/15.5 115.8 115.8

178.5

(in situ groundwater)

2.8

701

38 731 3.1

Chen et al.

‘entraticns. Experiments with highcr fcrrous ion concentratiotis ‘ulied in higher sulfate production (Figure 5). Table 2 shows althougb 2710 mg/L of persulfate degraded MTBE completely Jer different ferrous ion concentrations, sulfate production reased with higher ferrous concentrations. This indicates that

-.ru1fate decomposition rates would increase when higher - nentration of ferrous ion was added. Thus, proper addition of -rous ion could prevent unnecessary persulfate decomposiiion. Conclusions In this study, MTBE ocidation using ferrous ion-activated :.:rulfate oxidation was investigared. Results from the experirnents • -ow that persulfate oxidation was capable of degrading MTBE - :-ic.ently. Persulfate concentration correlated with MTBE degra: iion rate. Higher persulfate concentration caused more efficient \ ITBE degradation. No signifkant dissolved oxygen increase was ‘served in most expenmenis bccause of unfavorable conditions 2.3 Lo 2.9). Ferrous ion concentration presents a near-linear effect on MTBE rnova1 (between 0.36 and 3.57 mM of ferrous iron addition). tiowever, when excess ferrous ion was applied (11.43 mM). grading efficiency of .MTBE decreased because of competition :r sulfate free radicais between ferrous ion and MTBE. Results rom the experilllenis showed that MTBE eould be degraded ffective1y under appropriate femus ion addition (molar ratios of S’O2’/Fc2’ between 1/0.031 and 1/0.31) even if low concentra¡ ion of persulfate was applied (1.13 mM). Thus, the molar ratio of S’082/Fe2 in a persulfate oxidation system is the most important :nnrrolling factor to achieve effective MTBE removal. Moreover, the results of sulfatc analysis revcaled that persulfate decomposition would be increased if higher concentration of ferrous ion was applied. Thus, appropriate addition of ferrous ion should be controlled to prevent unnecessary persulfate decomposition. The results of TBA dcgradaiion show that low concentration of persulfate (approximately 1.13 mM) was able to remain in the %ystem for more than 43 days. This indicates that persulfare is a stable oxidant in water. In addition, the degrading efficiency of MTBE was higher than that predicted fmm the stoichiometric cquation. Tl’is may result from the complicated chain reactions of sulfate [roe radicais. The combined system with the addition of persulfate and peroxide sirnultaneously is able to enhanee (he cfficiency of MTT3E degradation effectively. Tlie first-order reaction rate of the combined system was approximately four times higher than that of the experimeni without hydrogen peroxide addition. Resulis also indicate that application of Fenton’s reaction followed by persulfatc oxidation has the potential to be developed jato a more effective ISCO system. Results of this study could aid in designing systems to remediate MTBE-contaminatcd sies. Acknowledgments This study was funded by the ApolLo Technology Corp., Ltd., Taiwan. The authors also thank Mr. Y.Y. Li of the Department of Marine F.nvironmental Engineering, National Kaohsiung Marine University, Taiwan, for his assistarlce throughout tuis project. The views or opinions expressed iii tIsis article are thosc of ttie writers and should not be construed as opinions of the U.S. Environmental Protection Agency. Submitred for publicarion Augus: 5, 2008; revised manuscripi submitred Ocrober ¡6, 2008; accepted for publicalion Januaiy 5, 2009.

Fe(11) mM Figure 5—Sulfate production under various ferrous ion concentrations ([MTBE] 0.114 mM; [Na2S2O8] = 11.39 mM) (MTBE = methyl tert-butyl ether). Tite deadline ro submir Discussions of ihis paper is Ocrober 15, 2009.

References A1-Ananzch, N.; Bergendahi, .1. A.; Thornpson. R.W. (2006) Kinetic MMeI for he degradation of MTBE by Fentons Oxidation. Enviro,,. Chem., 3 (1), 40—47, Anipsitakis, O. P.; Dionysiou, O. 0. (2004) Transitron Metal/UV-based Advanced Oxidadon Technologies for Water Decontarnination. AppI. Cara!. B: Envi ron.. 54 (3). 155—163. Ayoue, J. O.; Argue. D. M.; Mcgarry, F. J. (2005). Mctli>t Tert-butyl Eiher Occurrence and Related Factors in Public and Private Wells in Southeast New Hampshire. Envrion. Sci. Technol., 39 (1), 9—16. Block, P. A.; Brown, R. A.; Robinson, D. (2004) Novel Activation Technologies Por Sodium Peraulfate In Sim Chemical Oxidazion. Pmceedings of ¿he 4rh International Conference on ¿he Remediarion of Chlorinatedand Recalcirrwu Compour.ds-, Monterrey, California, May 24—27; Battellc: Columbus, Ohio. Bra,ds. O. C. (2001) MTBE — Panacea or Problern. Environ. Forensics, 2 (3), 189—196. Brown, R. A.; Robinson. D.; Skladany, 0. (2003) Response to Naturally Occumng Organic Material: Permanganate versus Persulfatc. Proceedi ngs of rhe Srli Annual Cwiferene un Covaminared Son; Obent, Belgium, Germany. May 12—16; Forschungszentnim Karlsruhe GmbH, Eggenstein-Leopvldshafen, Germany. Burbano, A. A.; Dionysiou, 0. D.; Suidan, M. T.; Richarson. T. L. (2005) Oxidation Kinetics and Effeet of pH un the Degradation of MTBE with Fenton Reagent. Wa:cr Res., 39(1), 107—118. Buton, O. y.; Malone, T. N.; Sahnon, O. A. (1997) Reaction of SO4’ with Fe2’, Mn24 and Cuz” in Aqueous Solution. J. Che,,,. Sae., Faraday Trans.. 93 (16), 2893—2897. Chawla, O. P.; Fessenden, R. W. (1975) Electron Spin Resonance and Pulseradiolysis Studies of Sorne Reactions of SO4. 1,2, j Phys. Che,,,., 79 (24), 2693—2700. Damm, 3. H.; Hardacre. C.; Kalin, R. M.; Walsh. K. P. (2002) Kineücs of the Oxidation of Meihyl Tert-butyl Ether (MTBE) by Potassium Permang anare. Warer Rei, 36 (14), 3638—3646. Fayolle, F.; Vandecasteele, 1. P.; Monot. F. (2001) Microbial Degradation ¡md Fate in the Environmeat of Meshyl Tert-butyl Ether ¡md Relatad Fuel Oxygenatcs. Appl. MicrobioL BiotechnoL, 56 (3—4), 339—349. GuiUard, C.; Charton, N.; Pichat, P. (2003) Degradazion Mechanism of t-Butyl Methyl Ether (MTBE) in Atniospheric Dropleis. Chemosphere, 53 (5), 469—477. House. D. A. (1961) Kinetics ¡md Mechanism of Oxidations by Pemaydisulfate. Che’m. Rey, 62 (3), ¡85—203.

Chen et al.

Huang, K. C.; Couttenye, R. A.; Hoag, G. E. (2002) Kinetics of Heata ssistcd Persulfate Oxidation of MeIhyl Tcrt-Butyl Ether (MTBE). Chemospnere, 49 (4). 413-420. Huang. K. C: 7iian, Z.; Hoag. G. E.; Dahmani, A.; Block, P. A. (2005) Degradation of Volatile Organic Conipounds with Thermally Activated Persulfate Oxidation. Chemosphei, 61(4), 551—560. Intersiate Tcclmology & Rcguiato,y Council (ITRC) (2005) Technical and Reguiarory Guidance foi- in Sizu Chemical Oxuialto , 2nd ed.; ¡ntcrstate Technology & Regulatoiy Council: Washington, D.C. Juhier, R.; Felding, 0. (2003) Monitorio8 Methyl Teruary Butyl Ether (MTBE)

and Ocher Organic Micmpollutants in Groundwater: Results fmn, the Danish National Monitonng Prograni. Wager Air Sol! PoIl., 149 (1—4), 145- 161. Kharounc, H.; Pausa, A.; Lebcault, J. M. (2001) Aerobic Biodegradation of an Oxygenates Mil(ture: ETBE, MTBE and TAME ¡o an Uptlow Fixedb ed Reactor. Water Res., 35 (7), 1665—1674. Kolb, A.; Plttmann. W. (2006) Comparison of MTBE Concentrations in Groundwater of Urban and Nonurbaji Aseas in Gerniany. Water Res.. 40 (19), 3551—3558. Kolthoff. 1. M; Miller, 1 K. (1951) fle Chemistry of Persulfate. 1. Thc Kinetics and Mechanism of dic Decornposjtjon of dic Persulfate Ion in Aqueous Mcdium. J. Am. Che,n. Soç., 73 (7), 3055—3059. Liang, C.; Bruell, C. J.; Marley, M. C.; Speny, K. L (2004a) Persulfate Oxidation br Iii Situ Remediatum of TCE 1—Activated by Ferrous ion with asid without a Persulfate-thiosulfate Rcdox Couple. Chemosphere. 55 (9), 1213—1223. Liang, C.; Brueli. C. J.; Marley, M. C.: Sperry, K. L. (2004b) Persulfate O.’cidatinn fnr In Sirti Remediation of TCE II — Activated by Chclatad Ferrous Ion. Chemosphere, 55 (9), 1225—1233. Liang. C.; Huang, C. F.; Chen, Y. 3. (2008) Potential for Aclivated Persulfate Degradation of BTEX Contamination. Water Res.. 42 (15), 4091—4 100. Liang. C.; Wang, Z. S.; BrucIl. Ci. (2007) Influence of pH on Persulfate Oxidatiosi of TCE al Ambieni Temperatures. Chemosphere, 66 (1). 106—213. Liang, C.; Wang, Z. S.; Mohanty. N. (2006) lnfluencet of Carbonate and Chioride lons oit Persulfate Oxidation of Trichloroethylene at 20’C. Sci. Total Envjron., 370 (2—3), 27I..77 Lin, C. W.; Chiang, S. 13.; Lu. S. 3. (2005) Investigation of MTBE and Aromani Compound Concentrations al a Gas Service Staüon. Envimn. Monii. Assess., 105 (1—3), 327—339. Mitani, M. M.; KcIIer, A. A.: Bunton, C. A.; Rinker, R. G.; Sandail, O. C. (2002) ICinetica and Products of Reactiona of MTBE with Ozone asid

Ozone/Hydrogen Peroxide in Water. J. Hazard. Mates, B89 (2—3). 197—212. Pcnnington, D. E.; Haiiii, A. (1968) Stoichiomeoy and Mechanisrn’of the Chromium (U) Peroxydisulfate Reaction. J. Am, Chem. Soc.. 90 (14), 3700—3704. Rivas, F. J. (2006) Polycvclic Aromatic Hydrncaihons Sorbed on Soils: A Short Review of Chemical Oxidation Based Treatmcnts. J. h’azwd. Mates, 138 (2), 234—251. Ruad, M.: Lacorte, S.; Barceid. D. (2006) Analysis, Occurrence and Fate of MTBE in thc Aquauc Environmeni over dic Pasi Decade. Trac—Trend. Anal. Chem., 25 (10). 1016—1029. Schmjdt. T. C.: Haderlein. S. B.; Forsier, R. (2004) Occurrence and Fatc Modeling of MTBE and BTEX Compounds iii a Swiss Lake Usad as Drinidng Water Supply. Wares Res., 38 (6), 1520—1529. Sun, Y.; Pignatello, 3. 3. (1993) Photochemical Reactions ¡nvolved in dic Total Mineralization of 2,4-D by Fe3/H2O,/LJV. Environ. Se!. Technol., 27 (2). 304—310.

Toran. 1..: I.ipka. C.; Baehr, A.; Reilly, T.; Baker, R. (2003) Scasonal asid Daily Variations ¡si Concentrations of Mcthyl Teniary-bwvl Ether (MTBE) al Cranberry Lake, New Jersey. Warer Res., 37 (15), 3756- 3766. Travina, O. A.; Kozlov, Y. N.; Purmal’. A. P.; Rodko, 1. Y. (1999) Synergism of the Action of dic Sulfite Oxidation lnitiators, Iron and Peroxydisulfate Toas. Russ. J. Phys. Chem.. 73 (8), 1215—1219. U.S. Environmental Protection Agcncy (1997) Drinlcing Wa:er Advisory: Conswner Acceptability Advice ami Health Effects Analysis on Meihyl Terriary.butyl E:her (MTBE); EPA-822-F-97-009; U.S. Environmental Protection Agency: Washington, D.C. U.S. F.nvironmental Protection Agency (1998) MTBE Fact Sheer 02. Remediatjon of MTBE Consaminaged Sol! asid Groundwazer; EPA 510- F-97-015; IJ.S. Environmental Protcction Agency: Washington, D.C. U.S. Environmental Protection Agency (2004) Technologies for Trearing MTBE asid Other Fuel Oxygenares; EPA 542-R.04-009; U.S. F.nvironmental Protection Agcncy: Washington, D.C. Waldemer. R. H.; Tratnyck, P. G.; Johnson. R. L; Nurmi, J. T. (2007) Oxidation of Chiorinated Ethenes by Heat-activated Persu’fate: Kinetics asid Products. Envi ron. Sci. TechnoL, 41(3), 1010—1015. Xu, X. R.; Zhao, Y. Y.; Li. X. Y.: Gu, 3. D. (2004) Chemical Oiidative Degradation of Methyl Teri-butyl Ether in Aqueous Solution by Fenton’s Reagent. Chemosphere, 55 (1), 73—79.

694

Water Envlronment Research, Volume 81, Number 7

Oxygenates such as ethers and alcohols usually are introduced into gasoline as additives to replace lead as an octane index enhancer and to improve air quality. Among the oxygen containing compounds, meihyl tert-butyl ether (MTBE) is the most commonly used gasoline additive because of its low cost and ease of production and blending (Fayolle et al.,2001), (Kharoune et al., 2001). However, because of the large amount of MTBE used, contamination has occurred in groundwater, surface water, and air (Juhler and Felding, 2003; Ayotte et al., 2005; Kolb and Püttmann. 2006; Toran et al., 2003; Schmidt et al., 2004; RoselI es aL, 2006; Guillard ct al., 2003; Lin el al., 2005). Methyl tert-butyl ether has been demonstrated to be an animal carcinogen. As a result, the U.S. Environmental Protection Agency (U.S. EPA) has temporarily classified MTBE as a possible human carcinogen (U.S. EPA, 1997).

Methyl tert-butyl ether is an organic chemical with high water solubility and low adsorption to soil. In addition, MTBE is not readily biodegradable because of the tertiary butyl group and ether linkage on it. Thus, it usually

migrates a longer distance than other gasoline components such as benzene, toluene. ethylbenzcne, and xylenes (BTEX) (US EPA, 1998; Braids, 2001; U.S. EPA, 2004). For this reason, more active measures may be required to control the movement of MTBE plumes at gasoline-contaminated sites. Chemical oxidation technology is an effective, potent groundwater remedial option capable of breaking down many contaminants in water. Hydrogen peroxide. Fenton’s reagent, permanganate, persulfate, and ozone are commori oxidants uscd lo remediase organic pollutante (Danim es al., 2002; Mitani es al., 2002; Burbano et aL, 2005). Recently. persulfate oxidation has been proposed for in sites chemical oxidation OSCO) proccsscs [Intenstate Technology & Regu!atory Council (rrCR), 2005J. The redox potential of persulfate is about 2.01 V. which is higher ¡han nhose of hydrogen pcroxide asid permariganase but lower ihan those of hydroxyl radicais (OH) and ozone (Huang el al., 2002; ITRC, 2005).

Persulfate can be activated thermally or chemically by initiators such as heat or transition mctals (cg., Fc2+) to produce more powerful sulfate free radicals (S04’) (l3lock eL al., 2004):

S2O8 + initiator —y 2S04 ‘(or SO4 + SO4 ) (1)

S04- + e — SO4 E° = 2.6V (2)

Sulfate free radicals are capable of degradinorganic compounds so carbon dioxide and water. For example, degradation of MTBE by sulfate free radicals can be illustrated as reaction 3:

30SO - + C5H120 +91120 — 5CO + 30H ± 30SO (3)

Generally, ferrous ion is a commonly used initiator for persulfale activation (Liang et aL, 2004a):

Fe24+S20[ —. Fe3+SO+SO. (4)

k = 2.0 X 1& M s (Travina et al., 1999)

k = bimolecular Tate constant.

SO4- +Fe2 —‘ Fe3+SO (5)

k 4.6 X M_’ s (Buxton el al., 1997)

From reactions 4 asid 5, overall reaction is obtaincd as reaction 6:

2Fe2 + S20r — 2Fe3 -f 2S0 (6)

k = 3.1 X M1 s (Buxton es al., 1997)

‘ Dcpartmern of Civil Enginecring, National Chi Nan University, 1 University Rd., Nantou 545, Taiwan; e-mail: kfchen@ncnu.edu.tw. 2 Institute of Envisonmental Engineering, Natinnal Sun Yat-Sen tiniversity, Kaohsiung. Taiwan. CPC Corp.. Tainan, Taiwan. 41J.S. Environnental Prtaection Agericy, Kansas City, Kansas.