Ri

MJa

b

c

h

�����

a

ARRAA

KAEHII

1

tcoiaRtca

0h

Journal of Hazardous Materials 227– 228 (2012) 126– 134

Contents lists available at SciVerse ScienceDirect

Journal of Hazardous Materials

j our na l ho me p age: www.elsev ier .com/ locate / jhazmat

emoval of iopromide and degradation characteristics in electron beamrradiation process

inhwan Kwona, Yeojoon Yoona, Eunha Choa, Youmi Junga, Byung-Cheol Leeb, Ki-Jung Paengc,oon-Wun Kanga,∗

Department of Environmental Engineering (YIEST), Yonsei University, 234 Maeji, Heungup, Wonju 220-710, Republic of KoreaQuantum Optics Laboratory, Korea Atomic Energy Research Institute, 1045, Daedeok-daero, Yuseong-gu, Daejeon 305-353, Republic of KoreaDepartment of Chemistry, Yonsei University, 234 Maeji, Heungup, Wonju 220-710, Republic of Korea

i g h l i g h t s

The second-order kinetic was fitted in overall removal tendency of iopromide.In the electron beam/H2O2 process, enhanced removal rate of iopromide was observed.The iopromide removal rate increased in the presence of OH• scavengers.The mineralization was mainly performed in the electron beam/H2O2 condition.The e−

aq mainly attacks the iodo-group, whereas the OH• reacts non-selectively.

r t i c l e i n f o

rticle history:eceived 30 January 2012eceived in revised form 4 May 2012ccepted 5 May 2012vailable online 14 May 2012

eywords:

a b s t r a c t

The aim of this study is to evaluate the removal efficiency of iopromide using electron beam (E-beam)irradiation technology, and its degradation characteristics with hydroxyl radical (OH•) and hydrated elec-tron (e−

aq). Studies are conducted with different initial concentrations of iopromide in pure water and inthe presence of hydrogen peroxide, bicarbonate ion, or sulfite ion. E-beam absorbed dose of 19.6 kGy wasrequired to achieve 90% degradation of 100 �M iopromide and the E-beam/H2O2 system increased theremoval efficiency by an amount of OH• generation. In the presence of OH• scavengers (10 mM sulfite

dvanced oxidation processlectron beamydrogen peroxide

odinated contrast mediaopromide

ion), the required dose for 90% removal of 100 �M iopromide was only 0.9 kGy. This greatly enhancedremoval was achieved in the presence of OH• scavengers, which was rather unexpected and unlike theresults obtained from most advanced oxidation process (AOP) experiments. The reasons for this enhance-ment can be explained by a kinetic study using the bimolecular rate constants of each reaction species.To explore the reaction scheme of iopromide with OH• or e−

aq and the percent of mineralization for thetwo reaction paths, the total organic carbon (TOC), released iodide, and intermediates were analyzed.

. Introduction

A substantial number of emerging contaminants (pharmaceu-icals and personal care products (PPCPs), endocrine disruptinghemicals (EDCs), surfactants, and other chemicals) are continu-usly being released into the environment, and have been detectedn various aquatic environments. According to recent research,bout 31 species of EDCs and PPCPs were detected in the Haniver of Korea [1], iopromide (IPM) was the major species among

he identified species. IPM is an iodinated X-ray contrast mediaompound (ICM) and is one of the widely used pharmaceuticalgents in Korean hospitals. It is used for imaging of organs or blood

∗ Corresponding author. Tel.: +82 33 760 2436; fax: +82 33 760 2571.E-mail address: jwk@yonsei.ac.kr (J.-W. Kang).

304-3894/$ – see front matter © 2012 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.jhazmat.2012.05.022

© 2012 Elsevier B.V. All rights reserved.

vessels during diagnostic tests, and it is designed to be resistantto human metabolism; therefore, the admitted agent be mostlyeliminated via urine or feces without transforming to anothermetabolite [2,3]. Steger-Hartmann et al. [4] reported that no toxiceffects of IPM were observed in both short-term and chronic tox-icity tests. However, the environmental safety of IPM cannot beguaranteed [2,5], since the side effects and toxic effects on renalcells by IPM have been consistently reported. Recently, IPM isknown to induce the vacuolization and apoptosis of tubular cells,and endothelial dysfunction leading to kidney injury [6]. Sincethe IPM is resistant in the conventional wastewater treatmentprocesses [7,8], the detected concentration of IPM in wastewater

effluent, surface water, and even in treated tap water is relativelyhigh in comparison to other PPCPs and EDCs [9,10]; therefore, moreeffective processes to remove the IPM in wastewater and drinkingwater treatment are urgently needed.

M. Kwon et al. / Journal of Hazardous Ma

Table 1Main reactions occurring in E-beam irradiation with rate constants for reactions ofhydrated electrons, hydroxyl radicals, and hydrogen atom in pure water (A) and IPMcontained solution (B).

Reaction number Reaction k (M−1 s−1) Refs.

(A.1) e−aq+H2O2 → OH•+OH− 1.1 × 1010 [19,27]

(A.2) e−aq+OH• → OH− 3.0 × 1010 [19,27]

(A.3) e−aq+O2 → O2

• 1.9 × 1010 [19,27](A.4) e−

aq+H → H• 2.3 × 1010 [19,27](A.5) e−

aq+O−•

2 → O2−2 1.3 × 1010 [19,27]

(A.6) e−aq+HO−

2 → OH•+2OH− 3.5 × 109 [19,27](A.7) OH•+OH•→ H2O2 5.5 × 109 [19,27](A.8) OH•+H2O2 → HO2

• 2.7 × 107 [19,27](A.9) OH•+OH− → H2O+O•− 1.3 × 1010 [19,27](A.10) OH•+O−•

2 → O2 + OH− 8.0 × 109 [19,27](A.11) OH•+HO

•2 → H2O + O2 6.0 × 109 [19,27]

(A.12) OH•+HO−2 → HO

•2 + OH− 7.5 × 109 [19,27]

(A.13) H•+H• → H2 7.8 × 109 [19,27](A.14) H•+O2 → HO

•2 2.1 × 1010 [19,27]

(A.15) H• + H2O2 → H2O + OH• 9.0 × 107 [19,27](A.16) H++O−•

2 → HO•2 4.5 × 1010 [19,27]

(A.17) HO2• + HO2

• → O2 + H2O2 8.3 × 105 [19,27](A.18) HO

•2+O−•

2 → O2 + H2O2 + OH− 9.7 × 107 [19,27](B.1) Iopromide + OH•→ products 3.30 ± 0.6 × 109 [38]

−

a(oOgsprcasdief

H

wt

rirrwItia�aooaepm

C C0

(B.2) Iopromide + eaq → products 3.25 ± 0.05 × 1010 [20]

Some studies on IPM removal using ozonation, photocatalyticctivity of TiO2, UV/H2O2, and other advanced oxidation processesAOPs) have been reported [11–14]. However, the removal schemef these processes relies on the oxidizing agents such as OH• and/or3, but not on the reducing scheme. It should be noted that the halo-enated compounds have high reactivity toward reducing agentsuch as e−

aq [15,16]. Since the electron beam (E-beam) irradiationrocess in water can induce not only oxidation but also reductioneaction, this technology can be advantageous if the contaminantan be removed via both an oxidation and reduction scheme. Inddition many researchers have proposed some combined systemsuch as the addition of a catalyst of reactive oxidants before the irra-iation of E-beam to improve its removal efficiency [17,18]. The

rradiation of E-beam in water results in the production of sev-ral radicals, ions, and molecules [19], and can be described as theollowing Eq. (1):

2O → [2.7]OH• + [2.6]e−aq + [0.6]H• + [0.7]H2O2 + [2.6]H3O+

+ [0.45]H2 (1)

here the numerical values in [] are the G values, which refer tohe yield of each species per 100 eV of absorbed dose.

Of the three radicals species (hydrated electrons (e−aq), hydroxyl

adicals (OH•) and hydrogen radicals (H•)) produced by E-beamrradiation, the species e−

aq and OH•, were regarded as importanteactive species in degrading IPM. The H• is also known to be aeactive reducing species, but its contribution in IPM degradationas neglected since it has a considerably lower reactivity toward

PM compared to e−aq and a lower G value (=0.6) H• as compared

o other species. Table 1 lists the major reactions that might occurn the E-beam irradiation process in pure water and IPM containedqueous solutions. Jeong et al. investigated the IPM removal with-irradiation [20], focusing on the degradation mechanisms in anttempt to elucidate the structures of major by-products. The aimf our study is also to investigate IPM removal by E-beam, butur study focuses on the detailed radical reaction schemes of e−

aqnd OH• with IPM, the matrix (or scavenger) effect of removalnhancement, and the role of e− and OH• in the partial or com-

aqlete mineralization of IPM. Overall, our utmost aim is to explore aethod to enhance IPM removal in the E-beam process.

terials 227– 228 (2012) 126– 134 127

2. Experimental methods

2.1. Solution preparation

All of the solutions were prepared using 18 M� deionized dis-tilled water. A reagent grade (pure material) of IPM (C18H24I3N3O8;MW 791.11) was purchased from U.S. Pharmacopeia. H2O2 solu-tions were prepared from Sigma–Aldrich® (30 wt% solution), andwere added before E-beam irradiation for the E-beam/H2O2 pro-cess. Quenching reagents of OH• were prepared with the sodiumsulfite (Na2SO3, 95.0%, SHINYO Pure Chemicals Co.) and sodiumbicarbonate (NaHCO3, 99.5%, Sigma–Aldrich®).

2.2. Electron-beam irradiation

All E-beam irradiation experiments were performed using anelectron accelerator made by EB TECH Co. (ELV-8, 2.5 MeV, 100 kW,Korea). The samples were prepared with 50 mL in a polyethylene-pack and were placed on the moving and irradiated so that theabsorbed dose would be from 0.5 to 50 kGy.

2.3. Analysis

The IPM concentration in solution was measured with a highperformance liquid chromatography (HPLC) system (Gilson Inc.,USA) equipped with a reverse phase column (Waters 5 �m ODS24.6 × 250 mm, C18) and a variable UV–vis wavelength detector(242 nm wavelength). An elution was performed with 10 mM phos-phoric acid/acetonitrile (80/20, v/v) at a flow rate of 1 mL/min.

The total organic carbon (TOC) contents of samples were mea-sured to investigate the degree of mineralization using a TOCanalyzer-TOC–VCPH/CPN (Shimadzu Co., Japan). To measure thereleased iodide ion, ion chromatography (IC) was employed usingan IC (DX-120, Dionex) equipped with a Dionex an Ion Pac AS16 anion column (4 mm × 250 mm, Dionex) and a conductivitydetector. To investigate some of the fragmented IPM breakdownproducts and degradation path, the samples were analyzed witha LC/IT-ESI-MS (LCQ Advantage, Thermo Finnigan) by direct injec-tion.

3. Result and discussion

3.1. Degradation of IPM by E-beam irradiation

Fig. 1A shows the degradation profile of IPM as a function ofabsorbed dose (from 0 to 25 kGy) for varied IPM initial concen-trations (10, 50, and 100 �M). In a 10 �M of initial IPM addedsolution, rapid degradation of IPM was observed by E-beam irradia-tion. However, the degradation rate was slower with the increase ofthe initial IPM concentration; therefore, higher doses were requiredfor complete removal of IPM (e.g. 3, 10, and >25 kGy required for95% removal of 10, 50, and 100 �M of initial IPM, respectively).The first-order kinetic is common for most of the E-beam irradi-ation for contaminate removal [21,22], in which the half-life ofthe contaminant should be constant throughout the reaction. Inthis study, IPM seems to be also removed by first-order kinetic(R2 = 0.85), but the second-order kinetic (R2 = 0.98) was better fittedin overall removal tendency, which reveals a rapid removal fol-lowed by a rather retarded degradation profile. The second-orderkinetic equation was represented at Eq. (2).

1 − 1 = kD (2)

where C is the residual concentration of IPM in �M, C0 is the initialconcentration of IPM in �M, D is the absorbed dose in kGy, andk is the second-order rate constant in �M−1 kGy−1. The inserted

128 M. Kwon et al. / Journal of Hazardous Materials 227– 228 (2012) 126– 134

F trations

fid0(b

D

10

lk(btitd

ig. 1. (A) Degradation of IPM by E-beam irradiation in varied IPM initial concenolution, (C) IPM removal (%) vs mineralization (%).

gure of Fig. 1A represents the fitted results with Eq. (2) for threeifferent initial concentrations of IPM (R2 values: 0.90, 0.99, and.96 at 10 �M, 50 �M, and 100 �M of IPM, respectively). From Eq.2), the dose required for the IPM to reach 90% removal, D0.9, cane calculated as given by Eq. (3):

0.9 = 9C0 · k

(3)

The calculated D0.9 for 10, 50, 100 �M of IPM were 2.3, 6.5,9.6 kGy, respectively, with the respective k values of 0.3977,.0279, and 0.0046 �M−1 kGy−1.

The observed kinetic characteristics can be explained as fol-ows. (1) The parent compound, IPM, has a high reactivity towardey reactive species, OH• and e−

aq generated by the E-beam processEqs. (B.1) and (B.2), Table 1). The key reactive species could alsoe reactive toward intermediate IPM reaction products, degrading

hem to smaller products and eventually leading to the mineral-zed end products. Fig. 1B demonstrates the absorption spectra ofhe treated IPM solution. As the peak of IPM (wavelength at 242 nm)ecreases, one of its notable by-product peaks (wavelength at

s (10, 50, and 100 �M of IPM), (B) the absorption UV spectra of the treated IPM

225 nm) appeared at 1 kGy and increased with the increase of doseto 10 kGy, but the peak almost disappeared with a further increaseof radiation dose to 25 kGy. These results imply that the key reactivespecies react not only with the parent compound, IPM, but also withits reaction intermediates for further breakdown. Actually, under alower irradiation dose, the IPM rapidly degrades under the inter-mediate free matrix conditions, but the IPM degradation rate couldbe lowered as the run proceeds due to the buildup of intermediateproducts during reaction. The supporting evidence for this inter-pretation is revealed in Fig. 1C, showing TOC removal (%) versusIPM removal (%). Until 80% of IPM removal, the TOC removal (%)was substantially low and linear with respect to IPM removal (%).However, a significant increase in TOC removal (%) was observedafter achieving significant removal of the parent compound, whichimplies that the key reactants would preferably react with theparent compound, and then the reactants will continuously react

with the intermediate products. (2) Another plausible reason forthe retarded reaction at higher doses, which is the cause of thesecond-order degradation kinetics of IPM, is the radical-radicalrecombination reactions among the primary radicals (e.g. Eq. (A.2)

M. Kwon et al. / Journal of Hazardous Materials 227– 228 (2012) 126– 134 129

F(

art

3

Hbr

Ebr3Ta(p

tibcriid

3

sou(

ig. 2. Effect of hydrogen peroxide (H2O2) on IPM ([IPM]0 = 100 �M) removalinserted Fig.; the IPM removal rate constant vs H2O2 doses).

nd (A.4), etc.), some of which could be competitive with the IPMemoval reaction. Mak et al. also observed similar phenomena inheir work with chlorinated compounds [23,24].

.2. E-beam/hydrogen peroxide (H2O2) process

In AOPs such as the UV/H2O2 process and O3/H2O2 process,2O2 is frequently added to generate additional OH• [25]. The E-eam/H2O2 process is also known to induce the synergy effect inemoving contaminants, generating additional OH• [26].

In order to observe any enhancement of IPM removal in the-beam/H2O2 process, H2O2 was initially added from 0 to 8 mMefore irradiation. In this set of H2O2 added cases, enhancedemoval rate of IPM was observed from an increase of 1.70 to.47 times with reference to the E-beam alone process (Fig. 2 andable 2). The reason for this enhancement can be attributed to thedditional OH• production due to the reaction of H2O2 and e−

aqEq. (A.1), Table 1) or H• (Eq. (A.15), Table 1) in the E-beam/H2O2rocess.

Maximum enhancement was observed at [H2O2]0 = 2 mM, buthe rate of enhancement decreased with the increase of H2O2. Thiss because the excess amount of H2O2 could scavenge OH• as giveny the reaction (Eq. (A.8), Table 1), which will lower the net OH•

oncentration [27]. Similar results of negative effects on substrateemoval by excessive H2O2 addition were reported in other H2O2nvolved AOPs [28,29]. In fact, an optimum level of H2O2 additions also required in the E-beam/H2O2 process, and a 2 mM of H2O2ose is suggested in removing IPM ([IPM]0 = 100 �M).

.3. Effect of OH• scavenger on IPM removal

In the E-beam process, both OH• and e−aq are important reactive

pecies for IPM removal. To investigate the relative contributionf each in IPM removal, an OH• quenching study was performedsing SO3

2− (sulfite). Since the sulfite slowly reacts with e−aq

ksulfite/e−aq

= < 1.5 × 106 M−1 s−1 [19]), whereas it quickly reacts

Fig. 3. Sulfite effect of on IPM ([IPM]0 = 100 �M) removal.

with OH• (ksulfite/OH• = 5.5 × 109 M−1 s−1 [19]), the sulfite was rec-ommended as a suitable OH• scavenger in the E-beam process [30].Recently, sulfate radical (SO4

−•) was received attention becauseof its high reactivity of organic pollutants including pharmaceuti-cals [31–33], but the possibility of generating the sulfate radical isremote in E-beam/sulfite condition. Fig. 3 shows the change of IPMremoval rates by E-beam irradiation in varied sulfite concentrations([sulfite]0 = 2, 10, and 20 mM). The IPM removal rate increased inthe sulfite added condition. IPM removal was only 43% without sul-fite addition, whereas, a considerably increased IPM removal wasobserved with increasing sulfite (82%, 93%, and 96% of IPM removalin 2, 10, and 20 mM sulfite concentration, respectively) under thesame irradiation dose of 1 kGy.

A possible interpretation could be that even though bothOH• and e−

aq are reactive species degrading IPM, the bimolec-ular reaction rate constant of IPM with e−

aq (kIPM/e−aq

= 3.25 ×1010 M−1 s−1, Table 1) is one order greater than the IPM withOH• (kIPM/OH• = 3.3 × 109 M−1 s−1, Table 1). According to a recentlyresearch by Jeong et al. [20], the IPM degradation efficiency withe−

aq is greater than with OH•. It should also be noted that thetwo reactive species, OH• and e−

aq, could rapidly react with eachother by radical-radical recombination reaction (kOH•/e−

aq= 3.0 ×

1010 M−1 s−1, Table 1) in the absence of each background scav-enger. In this sulfite added experimental condition, a greaterconcentration of e−

aq could be maintained, suppressing the radical-radical recombination reaction by scavenging OH•; therefore, thereaction opportunity of more reactive species, e−

aq, with IPMcould be increased. Similar results were reported by Cooper et al.in removing chloroform, which is also more reactive towarde−

aq than OH• (kchloroform/e−aq

= 1.1 × 1010 M−1 s−1, kchloroform/OH• =

5.4 × 107 M−1 s−1) [23,34,35].To confirm the enhanced IPM removal in the presence

of the OH• scavenger, other scavenging experiment wasperformed using bicarbonate. In a natural aquatic system,bicarbonate ion is always present and is known to trap OH•

(kbicarbonate/OH• = 1.0 × 107 M−1 s−1 [19]). As shown in Fig. 4,

an enhanced removal of IPM was also observed as the bicar-bonate concentration increases from 2 to 16 mM; even the

smaller enhancement was observed compared with exper-iments observed in the sulfite added case. This is becausethe second-order rate constant value of OH• with sulfite is

130 M. Kwon et al. / Journal of Hazardous Materials 227– 228 (2012) 126– 134

Table 2Removal rates of IPM in varied operating conditions ([IPM]0 = 100 �M).

Conditions Rate constant,k (�M−1 kGy−1)k (kGy−1)a

Rate of enhancement D0.9 (kGy) R2

IPM alone 0.0046 1.00 19.56 0.96IPM + 1 mM H2O2 0.0091 1.98 9.89 0.96IPM + 2 mM H2O2 0.0156 3.47 5.77 0.94IPM + 4 mM H2O2 0.0095 2.06 9.47 0.98IPM + 8 mM H2O2 0.0078 1.70 11.54 0.99IPM + 10 mM SO3

2− 2.7370a – 0.84 a 0.99a

IPM + 20 mM SO32− 3.2445a – 0.71 a 0.99a

IPM + 2 mM HCO3− 0.0070 1.52 12.85 0.99

IPM + 8 mM HCO − 0.0092 2.00 9.78 0.99

s([edtcittcccwmcEmo

Fo

3

IPM + 16 mM HCO3− 0.0120

a Calculated by first-order kinetic equation.

ignificantly greater than the value of OH• with bicarbonateksulfite/OH

• = 5.5 × 109 M−1 s−1 > kbicarbonate/OH• = 1.0 × 107 M−1 s−1

19]). Table 2 lists the second-order rate constant values, rate ofnhancement, and corresponding D0.9 values determined in allifferent matrix conditions. Under the E-beam/sulfite condition,he e−

aq, which is the single dominant major reacting species, onlyontributes to the dehalogenation of IPM not forming any variousntermediates; therefore, the removal tendency is suitable forhe first-order kinetic equation. In Table 2, we cannot comparehe rate constants between second-order and first-order rateonstants, but the D0.9 value (calculated by each rate constant)ould be substituted to compare the removal efficiency of eachondition. In all matrix conditions, the enhanced removals of IPMere observed. However, it should be noted that the enhancementechanism of the E-beam/H2O2 process must differ from the

ases of E-beam/sulfite and E-beam/bicarbonate. Compared to

-beam/sulfite and E-beam/bicarbonate, the E-beam/H2O2 processaintains a lower concentration of e−

aq, but a higher concentrationf OH•. This is undesirable in increasing IPM removal since e−

aq

ig. 4. Effect of alkalinity on IPM ([IPM]0 = 100 �M) removal (inserted Fig.; the sec-nd order kinetic plots of IPM for varied bicarbonate concentrations).

2.61 7.50 0.99

is more reactive toward IPM than OH•. However, some enhance-ments were observed in reality. Plausible interpretations evenwith the unfavorable distribution of the reactive species betweenOH• and e−

aq are: (1) the E-beam/H2O2 process can generatea greater amount of OH• than the E-beam alone, because thereaction between the H2O2 and e−

aq or H• produces additionalOH•, and (2) the H2O2 addition can suppress the radical-radicalrecombination between e−

aq and OH•, leading to the significantlygreater concentration of OH•, which contributed to the overall netreaction increase.

To confirm the previously discussed IPM removal mechanismand the IPM removal enhancement for the E-beam/sulfite conditionwith respect to the E-beam alone, a simple calculation approachwas conducted using an equation proposed as given by Eq. (4):

Enhanced removal = (�[IPM]/�kGy)|sulfite

(�[IPM]/�kGy)|alone

=(kIPM/OH• [OH•]∗IPM + kIPM/e−

aq[e−

aq]∗IPM)|sulfite

(kIPM/OH• [OH•]∗IPM + kIPM/e−aq

[e−aq]∗IPM)|alone

(4)

where �[IPM]/�kGy is the removed IPM per given absorbed dose,k is the second-order rate constant value of IPM with OH• or e−

aq,[OH•]IPM, [OH•]*

IPM, [e−aq]IPM, and [e−

aq]∗ are the concentration ofeach reacted species used for IPM removal in the E-beam aloneand E-beam/sulfite condition. As previously described, the moleconcentration of the key reacting species, OH• and e−

aq, would bechanged upon the addition of OH• scavengers.

In the E-beam alone process, the concentration of each react-ing species used for IPM removal, [OH•]IPM and [e−

aq]IPM, can becalculated as the equation given by Eqs. (5) and (6):

[OH•]IPM|alone = G([OH•]) × kIPM/OH• [IPM]

kIPM/OH[IPM] + ke−aq/OH• G([e−

aq])(5)

[e−aq]

IPM|alone = G([e−

aq]) ×kIPM/e−

aq[IPM]

kIPM/e−aq

[IPM] + kOH•/e−aq

G([OH•])(6)

which is the fractional distribution of each species reacted with IPMconsidering reactions (A.2), (B.1) and (B.2) as the key reactions ofthe all reactions in Table 1. The G value of each species (G([OH•]) and(G([e−

aq])) is the mole concentration of the reacting species initiallygenerated upon E-beam irradiation, which can be calculated fromG values given in Eq. (1). Under the E-beam sulfite condition (Fig. 3),the initial slope (below 1 kGy) was calculated to know the removaltendency of IPM; therefore, the initial concentrations of OH• and

e−

aq calculated by the G value at 0.5 kGy were G([OH•]) = 0.141 mMand (G([e−

aq])) = 0.136 mM, respectively.For the case of E-beam/sulfite radiation, the added sul-

fite will preferentially react with OH• owing to the high

M. Kwon et al. / Journal of Hazardous Materials 227– 228 (2012) 126– 134 131

F majoi nging

r

r

fIwE

[

[

w

[

It(batso(c

scavenger (Fig. 5B) and H2O2 (Fig. 5C). In Fig. 5B, compared tothe reactions in Fig. 5A, the reaction opportunities of IPM with e−

aqwill be increased, whereas the reactions of IPM with OH• and e−

aqwith OH• decrease in the presence of sulfite of bicarbonate. Fig. 5C

Table 3Experimental vs model predictions of enhanced IPM removal (%) in E-beam/sulfiteprocess ([IPM]0 = 100 �M).

Conditions Enhancedremoval (%)(experiment)

Enhancedremoval (%)(model)

ig. 5. Reaction scheme of iopromide with key reacting species (A), and change ofncreased reaction opportunities; dotted line represent decreased reactions in a cha

eactivity (kSO2−3

//OH• = 5.5 × 109 M−1 s−1 [19]), and the radical-

adical recombination reaction (kOH•/e−aq

) would be decreased. In

act, e−aq will be the major reacting species used for IPM removal.

n the sulfite added case of E-beam irradiation, the reacted speciesith IPM, [OH•]*

IPM|sulfite and [e−aq] ∗ IPM|sulfite, can be expressed by

qs. (7) and (8):

OH•]∗IPM|sulfite = G([OH•])

× kIPM/OH• [IPM]

kIPM/OH• [IPM] + ke−aq/OH• G([e−

aq]) + kSO2−3

/OH• [SO2−3 ]

(7)

e−aq]

IPM|sulfite = G([e−

aq]) ×kIPM/e−

aq[IPM]

kIPM/e−aq

− [IPM] + kOH•/e−aq

− [OH•]∗e−aq

(8)

here [OH•]∗e−aq

is given by Eq. (9):

OH•]∗e−

aq|sulfite = G([OH•])

×ke−

aq/OH• G([e−aq])

kIPM/OH• [IPM] + ke−aq/OH• G([e−

aq]) + kSO2−3

/OH• [SO2−3 ]

(9)

Comparing the concentrations of the reacted species withPM in E-beam alone and in E-beam/sulfite, the OH• concen-ration in E-beam/sulfite is smaller than that in E-beam alone[OH•]IPM|alone � [OH•]*

IPM|sulfite) but the e−aq concentration in E-

eam/sulfite will be considerably greater than that in E-beamlone ([e−

aq] ∗ IPM|sulfite � [e−aq] ∗ IPM|alone. Since IPM is more reac-

ive with e−aq, the IPM removal enhancement can be observed in the

ulfite added solution. Plugging Eqs. (5)–(8) with known second-rder rate constant values of IPM with reacting species into Eq.4), the enhancement removal (%) can be calculated. In Table 3, thealculated values were compared with the experimental results

r reaction path in addition of sulfite (B) and H2O2 (C) (solid line represents key or water matrix conditions).

determined with the initial data points observed under 1 kGy ofirradiation dose (Fig. 3), which shows a good agreement betweenthe two.

Even though the above modeling approach is useful in explain-ing the enhanced removal of a compound such as IPM, it shouldbe noted that this approach is limited to a simple matrix andearly reaction phase. For the case of E-beam/H2O2 process, rad-ical reactions associated with the key reacting species, OH• ande−

aq, are significantly complicated and interconnected to each otherwith chain reactions. In fact, the above approach is inapplicableto the E-beam/H2O2 process due to the difficulty of prediction ofthe key reacting species in the H2O2 spiked water matrix condi-tion. Even though the lack of precise model equation for everywater matrix conditions, the reaction schemes of IPM with the keyreacting species in the three different conditions (E-beam alone,E-beam/sulfite, E-beam/H2O2) can be (Fig. 5). The solid lines rep-resent the major or increased reactions occurring in a reference(Fig. 5A) or a changing matrix condition, and the dotted lines arethe decreased reaction opportunities in the presence of the OH•

IPM alone 100% 100%IPM + 2 mM SO3

2− 194% 170%IPM + 10 mM SO3

2− 277% 266%IPM + 20 mM SO3

2− 293% 301%

132 M. Kwon et al. / Journal of Hazardous Materials 227– 228 (2012) 126– 134

F

sHwb((i(ss(oesi

3

teAwcd

etaef1trdIbdaacdpaibg

ig. 6. Mineralization of IPM in various water matrix conditions ([IPM]0 = 100 �M).

uggests why the enhanced IPM removal could be observed in the2O2 added condition, even though the reaction opportunity of IPMith e−

aq was decreased. This is because the reaction opportunitiesetween IPM and OH• could be greatly increased as shown in path1) and (2) by adding H2O2. Considering that the reaction path (1)reaction between IPM and OH•) is dominant in E-beam/H2O2, thenitially added H2O2 will scavenge e−

aq with a rapid reaction ratepath (3)), producing additional OH•. The reaction path (2) repre-ents the reaction between IPM with the secondarily formed OH•. Ithould be noted that H2O2 itself can act as an OH• scavenger (path4)), but it will eventually transform back to OH• by the reactionf e−

aq and H2O2. The excess amount of H2O2, which could scav-nge the OH•, is greater for the lower dose; therefore, the initiallope could not reflect the overall removal rate and the model isnapplicable to the E-beam/H2O2 condition.

.4. Mineralization

The mineralization extent was investigated for each differentest condition (E-beam alone, E-beam/H2O2, E-beam/sulfite) toxplore the major reacting species involved in IPM mineralization.s discussed previously in Section 3.1 with Fig. 1C, low TOC removalas observed even with the high E-beam irradiation dose. Fig. 6

ompares the %TOC removal of IPM as function of the absorbedose for the tested conditions.

Several interesting features can be observed in this set ofxperiments: (1) In E-beam alone without any scavenger addi-ion ([IPM]0 = 100 �M), the %TOC removal was approximately 40%t a 50 kGy absorbed dose. For the case of E-beam/H2O2, a rathernhanced TOC removal was observed. The %TOC removal increasedrom 40% in the blank solution to 50 and 60% in the 0.05 and.47 mM H2O2 added solution at a 50 kGy absorbed dose, respec-ively. In this case of the H2O2 added condition, it should beemembered that the increased IPM removal observed and theegree of mineralization also increased along with the enhanced

PM removal in the E-beam/H2O2 process. (2) For the case in E-eam/sulfite, it is interesting to observe that the %TOC removalecreased compared to the E-beam alone, even though we observed

significantly enhanced IPM removal with the increase of sulfiteddition (Fig. 3). The %TOC removal was only 3% or less even for theomplete removal of IPM with E-beam/sulfite at a low irradiationose (5 kGy). (3) The above mineralization data could support ourrevious interpretation on the major reacting species between e−

aq•

nd OH for both of the runs, E-beam/sulfite and E-beam/H2O2,

n which total enhanced IPM degradation was observed. The E-eam/sulfite condition is efficient in the removal of IPM since areater amount of e−

aq, which is considerably more reactive toward

Fig. 7. Deiodination of iopromide (A) according to electron beam irradiation in dif-ferent conditions and (B) according to the iopromide degradation ([IPM]0 = 100 �M).

IPM, could be maintained in this sulfite added matrix condition.However, the lower mineralized results in this process are due tothe key reacting species, e−

aq, which is limited only to attack theiodo-group of IPM and its intermediates, but not to the carboncontaining intermediates. OH• is known to be non-selective whenreacting with various compounds having diverse functional groups,and in fact, the further reaction of OH• with IPM intermediates willeventually lead to the complete mineralization. The enhanced TOCremoval in this E-beam/H2O2 process indirectly supports the pre-vious mechanistic interpretation of the IPM removal enhancementdue to the additional OH• production in the E-beam/H2O2 process.

3.5. Deiodination of IPM

To confirm the reaction schemes of the reacting species (OH•

and e−aq) with IPM in detail, the samples were analyzed by IC for the

released iodide (I−) and MS for the key intermediates. In general,the halogen substituents have an electron-withdrawing character-istic; therefore, the IPM carrying iodides could be resistant to anoxidation process [36], and the role of OH• would be minimal,

whereas, since an electron affinity to halogen atoms is high [37],the IPM will favorably react with e−

aq.Fig. 7A shows the %iodide release from degraded IPM for

the three tested processes, E-beam alone, E-beam/H2O2, and

M. Kwon et al. / Journal of Hazardous Materials 227– 228 (2012) 126– 134 133

FE

Ecrdeikwbi

Table 4Major deiodinated IPM peaks based on the mass to charge ratio (m/z) value appearedin the E-beam alone and E-beam/sulfite process.

Adsorbed dose E-beam alone E-beam/sulfite (=2 mM Na2SO3)a

0 kGy 791 791

1 kGy 791 591*682*b 554***

664* 427***554**

3 kGy 554** 427***444*** 553***

664* 591*

ig. 8. MS spectrum of IPM reaction products. (A) E-beam alone (No scavenger); (B)-beam/sulfite (=2 mM Na2SO3).

-beam/sulfite. For three case conditions, the released iodide con-entration rapidly reaches maximum level at around 5 kGy ofadiation dose, and stays almost constant, followed by a slightecrease, with an increase of the radiation dose. One possiblexplanation for the decrease of iodide after reaching the max-mum is the reaction between OH• and I− (OH• + I− → HOI−,

= 1.1 × 1010 M−1 s−1 [19]). The maximum %yield of iodide release

as highest (93%) in E-beam/sulfite and lowest (30%) in the E-

eam/H2O2. The lowest iodide release in the E-beam/H2O2 processs understandable since the deiodination reaction by e−

aq could be

a m/z number in E-beam/sulfite was adjusted excluding m/z value of Na (=23).b The number of released iodide.

reduced because H2O2 scavenges e−aq. In the E-beam/sulfite process,

the released iodide reached almost close to the level of completedeiodination, in which a sufficient concentration of e−

aq must havebeen maintained. This result also supports the previously describedreaction scheme of IPM by OH• and e−

aq for the tested processes. Toinvestigate the portion of deiodination reaction of the total IPMremoval, the released % iodide from the removed IPM was replot-ted with respect to % IPM removal (Fig. 7B). In the E-beam/sulfite(=2 mM) process, the IPM removal (%) was almost equivalent to thereleased iodide (%), which implies that IPM removal is mainly gov-erned by the deiodination reaction path by e−

aq. These results alsoconcurred with the lower mineralization rate (<3% TOC removal)observed in this condition (Fig. 6). In the E-beam/H2O2 process,the slope of IPM removal (%) versus released iodide (%) is aboutthree, implying that the deiodination reaction path by e−

aq con-tributes about 1/3 of the total IPM removal. Unlike the previoustwo processes, in the case of E-beam alone, the portion of deiodina-tion reaction changes as the reaction proceeds. The slope betweenIPM removal (%) and released iodide (%) was about two until 40% ofIPM removal, but then approached one. This slope change impliesthat at the initial reaction period, both e−

aq (by deiodination path)and OH• (by oxidation reaction path) equally contribute to the IPMremoval. However, as the reaction proceeds further, the role of e−

aqbecomes dominant over OH• because the attack of e−

aq to the iodo-group of fragmented intermediates could become easier after thebreakdown of the parent IPM into smaller fragments by OH•.

To confirm the deiodination reaction scheme of IPM, irradi-ated samples taken from the E-beam alone and the E-beam/sulfite(2 mM) process were analyzed by a mass spectrometer (MS). Tocapture the initial deiodination reaction path with the MS, the IPMcontained solutions were irradiated at relatively lower doses of 1and 3 kGy. Fig. 8 shows and compares the MS spectrum of two dif-ferent samples, one without sulfite (Fig. 8A) and the other withsulfite (Fig. 8B). It should be noted that the mass to charge ratio(m/z) values of the same species are greater in Fig. 8B than thevalue showed in Fig. 8A due to the contribution of the m/z valueof Na (m/z = 23) to the fragment under the MS analysis of the sam-ple containing Na2SO3 (=2 mM). In both runs, the MS spectrums ofthe deiodinated compounds corresponding to one to three iodidereleased fragments of IPM were all observed. However, compar-ing the two spectrums of the samples irradiated at the lowestdose, 1 kGy, the completely deiodinated peaks (m/z = 552 and 449)were detected in the E-beam/sulfite condition rather than in E-beam alone, although iodide contained fragments still appear inthe E-beam alone process. Table 4 lists the m/z values of the deio-dinated reaction products, and also compares the peaks detectedin the result of the deiodination favorable reaction for the two test

conditions. The sulfite added condition is significantly favorableto the deiodination reaction since the number of peaks corre-sponding to the three iodide released fragments is more dominant

1 ous M

tanEya

4

trboietbtlmEcmtpOrtc

A

REe

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

34 M. Kwon et al. / Journal of Hazard

han the peaks observed in the sulfite absence case. This resultgain suggests that the e−

aq involved deiodination reaction is sig-ificantly favored in the E-beam/sulfite process compared with-beam alone. This MS analysis was limited to the qualitative anal-sis only observing deiodination scheme; therefore, more detailednalysis employing quantitative analysis needs to follow.

. Conclusion

From this study, it appears that the E-beam process can effec-ively remove IPM. The E-beam absorbed dose of 20 kGy wasequired to achieve the 90% removal of 100 �M IPM, and the E-eam/H2O2 system increased the removal efficiency by an amountf OH• generation. On the contrary, the OH• quenching test alsoncreased the removal efficiency by the inducement of many more−aq reactions, which was unexpected for most advanced oxida-ion processes (AOPs). In this regard, the E-beam process maye proposed as a strong candidate among AOPs for removinghe reactive pharmaceuticals simultaneously present with a highevel OH• scavenger such as alkalinity, which is the case for the

unicipal wastewater and livestock wastewater. Furthermore, the-beam/H2O2 process could be designed according to the waterondition. The mineralization was improved when the OH• was aajor reactant, but it was decreased when e−

aq was a major reac-ant. This is the reason why each reactant has a different reactionreference. The e−

aq mainly attacks the iodo-group, whereas theH• non-selectively reacts with the side chains of the benzene

ing rather than the iodo-group. These results will provide bet-er estimates of the removal efficiencies according to the targetompound.

cknowledgments

This work was supported by BK21 Program and Nuclearesearch & Development Program of the Korea Science andngineering Foundation (KOSEF) grant funded by the Korean gov-rnment (MEST).

eferences

[1] Y. Yoon, J. Ryu, J. Oh, B.G. Choi, S.A. Snyder, Occurrence of endocrine disrupt-ing compounds, pharmaceuticals, and personal care products in the Han River(Seoul, South Korea), Sci. Total Environ. 408 (2010) 636–643.

[2] S. Perez, D. Barcelo, Fate and occurrence of X-ray contrast media in the envi-ronment, Anal. Bioanal. Chem. 387 (2007) 1235–1246.

[3] T.A. Ternes, J.L. Kormos, M. Schulz, H.P.E. Kohler, Biotransformation of selectediodinated X-ray contrast media and characterization of microbial transforma-tion pathways, Environ. Sci. Technol. 44 (2010) 4998–5007.

[4] T. Steger-Hartmann, R. Lange, H. Schweinfurth, Environmental risk assessmentfor the widely used iodinated X-ray contrast agent iopromide (Ultravist), Eco-toxicol. Environ. Saf. 42 (1999) 274–281.

[5] A. Peer, Z. Averbukh, S. Berman, D. Modai, M. Averbukh, J. Weissgarten, Contrastmedia augmented apoptosis of cultured renal mesangial, tubular, epithelial,endothelial, and hepatic cells, Invest. Radiol. 38 (2003) 177–182.

[6] Z.J. Yang, Y.M. Zhao, Z.W. Tao, Z.H. Xu, Z.X. Tao, B. Chen, L.S. Wang, C.J. Li, L.L.Chen, Q.Z. Jia, E.Z. Jia, T.B. Zhu, Toxic effects of a high dose of non-ionic iodinatedcontrast media on renal glomerular and aortic endothelial cells in aged rats invivo, Toxicol. Lett. 202 (2011) 253–260.

[7] N.J.D. Graham, T.W. Chan, W. Chu, Degradation of iopromide by combined UVirradiation and peroxydisulfate, J. Hazard. Mater. 181 (2010) 508–513.

[8] W. Kalsch, Biodegradation of the iodinated X-ray contrast media diatrizoateand iopromide, Sci. Total Environ. 225 (1999) 143–153.

[9] W. Seitz, J.Q. Jiang, W.H. Weber, B.J. Lloyd, M. Maier, D. Maier, Removal of iod-inated X-ray contrast media during drinking water treatment, Environ. Chem.3 (2006) 35–39.

10] J.Q. Jiang, W. Seitz, W.H. Weber, B.J. Lloyd, M. Maier, D. Maier, W. Schulz, Mon-itoring of iodinated X-ray contrast media in surface water, Chemosphere 64(2006) 1318–1324.

11] B. Ning, N.J.D. Graham, Ozone degradation of iodinated pharmaceutical com-pounds, J. Environ. Eng. – ASCE 134 (2008) 944–953.

[

[

aterials 227– 228 (2012) 126– 134

12] T.E. Doll, F.H. Frimmel, Kinetic study of photocatalytic degradation of carba-mazepine, clofibric acid, iomeprol and iopromide assisted by different TiO2materials—determination of intermediates and reaction pathways, Water Res.38 (2004) 955–964.

13] M. Sprehe, S.U. Geissen, A. Vogelpohl, Photochemical oxidation of iodized X-ray contrast media (XRC) in hospital wastewater, Water Sci. Technol. 44 (2001)317–323.

14] B. Ning, N.J.D. Graham, P.D. Lickiss, A comparison of ultrasound-based advancedoxidation processes for the removal of X-ray contrast media, Water Sci. Technol.60 (2009) 2383–2390.

15] H. Mohan, K.-D. Asmus, Radical cations from alkyl iodides in aqueous solution,J. Chem. Soc. Pak. 53 (1987) 1795–1800.

16] D. Behar, P. Neta, Intramolecular electron transfer and dehalogenation of anionradicals. 4. Haloacetophenones and related compounds, J. Am. Chem. Soc. 103(1981) 2280–2283.

17] N.K.V. Leitner, S.O.B. Boukari, F. Pellizzari, Influence of persulfate ions on theremoval of phenol in aqueous solution using electron beam irradiation, J. Haz-ard. Mater. 185 (2011) 844–851.

18] J.C. Gonzalez-Juarez, J. Jimenez-Becerril, Gamma radiation-induced catalyticdegradation of 4-chlorophenol using SiO2, TiO2, and Al2O3, Radiat. Phys. Chem.75 (2006) 768–772.

19] G.V. Buxton, C.L. Greenstock, W.P. Helman, A.B. Ross, Critical review of rateconstants for reactions of hydrated electrons, hydrogen atoms and hydroxylradicals (•OH/•O−) in aqueous solution, J. Phys. Chem. Ref. Data 17 (1988)513–886.

20] J. Jeong, J. Jung, W.J. Cooper, W.H. Song, Degradation mechanisms and kineticstudies for the treatment of X-ray contrast media compounds by advancedoxidation/reduction processes, Water Res. 44 (2010) 4391–4398.

21] M.G. Nickelsen, W.J. Cooper, K.J. Lin, C.N. Kurucz, T.D. Waite, High-energyelectron-beam generation of oxidants for the treatment of benzene and toluenein the presence of radical scavengers, Water Res. 28 (1994) 1227–1237.

22] Y.J. Yoon, Y.J. Jung, B.S. Han, J.W. Kang, Performance of electron beam irradiationfor treatment of groundwater contaminated with acetone, Water Sci. Technol.59 (2009) 2475–2483.

23] F.T. Mak, S.R. Zele, W.J. Cooper, C.N. Kurucz, T.D. Waite, M.G. Nickelsen, Kineticmodeling of carbon tetrachloride, chloroform and methylene chloride removalfrom aqueous solution using the electron beam process, Water Res. 31 (1997)219–228.

24] S.M. Pimblott, J.A. Laverne, Cooperative effects of scavengers on the scavengedyield of the hydrated electron, J. Phys. Chem. 96 (1992) 8904–8909.

25] W.H. Glaze, J.W. Kang, D.H. Chapin, The chemistry of water treatment processesinvolving ozone, hydrogen peroxide and ultraviolet radiation, Ozone Sci. Eng.9 (1987) 1987.

26] M.G. Nickelsen, W.J. Cooper, C.N. Kurucz, T.D. Waite, Removal of benzene andselected alkyl-substituted benzenes from aqueous-solution utilizing continu-ous high-energy electron-irradiation, Environ. Sci. Technol. 26 (1992) 144–152.

27] M.G.N.S.R. Zele, W.J. Cooper, C.N. Kurucz, T.D. Waite, Modeling kinetics of ben-zene, phenol, and toluene removal in aqueous solution using the high-energyelectron beam process, in: W.J Cooper, R.D. Curry, K.E. O’Shea (Eds.), Environ-mental Applications of Ionizing Radiation, John Wiley & Sons, Inc., New York,1998, pp. 395–415.

28] W.H. Glaze, J.W. Kang, Advanced oxidation processes. Test of a kinetic modelfor the oxidation of organic compounds with ozone and hydrogen peroxide ina semibatch reactor, Ind. Eng. Chem. Res. 28 (1989) 1580–1587.

29] M.I. Stefan, A.R. Hoy, J.R. Bolton, Kinetics and mechanism of the degradationand mineralization of acetone in dilute aqueous solution sensitized by the UVphotolysis of hydrogen peroxide, Environ. Sci. Technol. 30 (1996) 2382–2390.

30] H.S. Shin, Y.R. Kim, A.V. Ponomarev, Influence of sulfite on radiolytic conversionof nitrate and nitrite in dilute aqueous solutions, Mendeleev Commun. (2001)21–23.

31] Y.H. Guan, J. Ma, X.C. Li, J.Y. Fang, L.W. Chen, Influence of pH on the formationof sulfate and hydroxyl radicals in the UV/peroxymonosulfate system, Environ.Sci. Technol. 45 (2011) 9308–9314.

32] P. Nfodzo, H. Choi, Sulfate radicals destroy pharmaceuticals and personal careproducts, Environ. Eng. Sci. 28 (2011) 605–609.

33] B. Roshani, N.K.V. Leitner, Effect of persulfate on the oxidation of benzotriazoleand humic acid by e-beam irradiation, J. Hazard. Mater. 190 (2011) 403–408.

34] W.J. Cooper, M.G. Nickelsen, S.P. Mezyk, G. Leslie, P.M. Tornatore, W. Hardison,P.A. Hajali, MTBE and priority contaminant treatment with high energy electronbeam injection, Radiat. Phys. Chem. 65 (2002) 451–460.

35] J.A. Laverne, S.M. Pimblott, Yields of hydroxyl radical and hydrated electronscavenging reactions in aqueous-solutions of biological interest, Radiat. Res.135 (1993) 16–23.

36] H. Loonen, F. Lindgren, B. Hansen, W. Karcher, J. Niemela, K. Hiromatsu, M.Takatsuki, W. Peijnenburg, E. Rorije, J. Struijs, Prediction of biodegradabilityfrom chemical structure: modeling of ready biodegradation test data, Environ.Toxicol. Chem. 18 (1999) 1763–1768.

37] H. Mohan, P.N. Moorthy, Pulse radiolysis investigations on the oxidation ofiodobenzene in aqueous solutions, Radiat. Phys. Chem. 33 (1989) 211–217.

38] M.M. Huber, S. Canonica, G.Y. Park, U. Von Gunten, Oxidation of pharmaceuti-cals during ozonation and advanced oxidation processes, Environ. Sci. Technol.37 (2003) 1016–1024.