role of humic acid and quinone model compounds in bromate reduction by zerovalent iron

Role of Humic Acid and QuinoneModel Compounds in BromateReduction by Zerovalent IronL I X I E A N D C H I I S H A N G *

Department of Civil Engineering, The Hong Kong Universityof Science and Technology, Clear Water Bay, Kowloon, HongKong, People’s Republic of China

Experiments were conducted to examine the role ofhumic acid and quinone model compounds in bromatereduction by Fe(0). The reactivity of Fe(0) toward bromatedeclined by a factor of 1.3-2.0 in the presence ofhumic acid. Evidence was obtained that the quickcomplexation of humic acid with iron species and itsadsorption passivated the iron surface and decreased therate of bromate reduction by Fe(0). On the other hand,in the long run, the reduced functional groups present inhumic acid were observed to regenerate Fe(II) and reducebromate abiotically. Compared with the case of humicacid only, the simultaneous presence of Fe(III) and humicacid significantly increased the bromate removal rate.Fe(III)/Fe(II) acted as a catalyst in the oxidation of humicacid by bromate. Anthraquinone-2,6-disulfonate (AQDS) andlawsone did not cause any significant effect on thebromate reduction rate by Fe(0). However, the redoxreactivity of lawsone in the presence of Fe(III) was evident,while AQDS did not show any under the tested conditions.The difference was attributable to the presence/absence of reducing functional groups in the modelcompounds. The electron spin resonance further demon-strated that the redox functional groups in humic acid aremost likely quinone-phenol moieties. Although thebromate reduction rate by regenerated Fe(II) is a fewtimes slower than that by Fe(0), the reactive Fe(II) can be,alternatively, reductively formed to maintain iron surfaceactivation and bromate reduction to prolong the lifetime ofthe zerovalent iron.

IntroductionBromate is a carcinogenic disinfection byproduct, resultingfrom ozonation of bromide-containing water. The USEPAhas established the maximum contaminant level (MCL) of10 µg/L for bromate (1). Surveys around the world have shownthat, in drinking water, the bromate concentration rangesfrom 0 to 127 µg/L, depending upon the bromide concen-tration in raw water and the disinfection process used (2, 3).

Reduced iron species, such as ferrous ions (Fe(II)) andzerovalent iron (Fe(0)), have been utilized as chemicalreductants for bromate removal (4-6). In our previous study,bromate removal by Fe(0) was described as a processcontrolled by surface-mediated reactions (6). Any factor thatalters the iron surface reactivity influences the rate of bromatereduction. Natural organic matter (NOM), ubiquitously

present in soils, surface water, and groundwater, has beenreported to form complexes with iron and iron oxide throughvarious types of carboxylate, phenolic, and carbonyl func-tional groups in NOM (7, 8). The adsorption of NOM isgenerally dominant over contaminants for available activesites on the iron surface (9, 10), thereby lowering thecontaminant removal efficiency. The accumulation of NOMat the iron-water interface may also adversely affectcontaminant reduction through the inhibition of metalcorrosion (11). On the other hand, NOM may act as anelectron-transfer mediator in the chemical reduction oforganic pollutants. It enhances the rate of chemical reductionof nitro aromatic or halogenated compounds by sulfide orferrous ions (12, 13) or facilitates the reduction of Fe(III) toFe(II) in the presence of microorganisms (14, 15). NOM hasbeen demonstrated to participate in the oxidation/reductionof iron as a factor controlling the iron speciation (16).Recently, oxidation of various types of NOM in the abioticreduction of Fe(III) to Fe(II) (17) or in the reduction ofhexachloroethane has been reported (18). The electron-transfer property of NOM has been proposed to be relatedto the formation of Fe(III)-NOM complexes (19) and thepresence of functional groups such as quinone, phenolic,and carboxylate moieties in NOM (7, 20, 21). In particular,quinone moieties in NOM have been speculated to play animportant role in electron transfer (12, 13). Tratnyek et al.(10) have found that quinone model compounds such asjuglone, lawsone, and anthraquinone-2,6-disulfonate (AQDS)can enhance the reduction rate of carbon tetrachloride byFe(0). Electron spin resonance (ESR) spectroscopy has beenused to measure the semiquinone radical concentration suchthat the quinone moieties in humic substances have beenconfirmed as electron-transfer mediators in biotic systems(21). Different redox capacities of NOM have been reported(19, 21), which are attributable to the different structuraland functional groups in NOM (17). Moreover, in additionto the quinone groups, complexed Fe(III) in NOM has beenhypothesized to participate in redox reactions as an electron-transfer mediator (19). Owing to the limited informationavailable in the literature, more studies are needed tounderstand the contributions of NOM and complexed metalions in the redox reactions and the mechanisms of thesereactions.

This study was, therefore, conducted to evaluate the rolesof humic acid (representing NOM) and quinone modelcompounds, such as lawsone and AQDS, in bromate reduc-tion by Fe(0) and to assess the redox reactivity of thesecompounds on the reduction of bromate and Fe(III). Therole of Fe(III)/Fe(II) in the reaction cycle was investigated.Fourier transform infrared (FTIR) and ESR spectroscopicanalyses were also conducted to substantiate the findingswith respect to the identification of the functional groups ofhumic acid that complex with the iron surface and the redoxfunctional groups that participate in the reduction of bromate.

Experimental SectionChemicals. All chemical stock solutions were prepared usingdoubly distilled deionized (DDDI) water and stored at 4 °C.Sodium bromate was obtained from Nacalai Tesque. Ferroussulfate and ferric nitrate were from Riedel-deHaen. Lawsoneand AQDS (reagent grade) were from Acros Organics andSigma, respectively. Irons purchased from Connelly-GPM(catalog no. 1004) were sieved to obtain iron grains withsizes of 425-1000 µm. The iron grains were pretreated withsonication and acid washing prior to the experiments in thesame manner as described in ref 6. Humic acid (representing

* Corresponding author phone: (852)2358 7885; fax: (852)23581534; e-mail: [email protected].

Environ. Sci. Technol. 2005, 39, 1092-1100

1092 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 4, 2005 10.1021/es049027z CCC: $30.25 2005 American Chemical SocietyPublished on Web 01/15/2005

NOM) was obtained from Aldrich, which has been applieddirectly or after pretreatment in numbers of studies on therole of humic substances as an electron-transfer mediator(15, 22, 23). In this study, the humic acid stock solution wasprepared by dissolving 1 g of humic acid powder in 1 L ofDDDI water and filtering the solution through a 0.45 µmfilter paper (Advantec MFS) without further purification. Themajor impurities of concern are Fe and Cu, and theirconcentrations under the test conditions were less than 200and 1.2 µg/L, respectively, according to the report from thesupplier.

Experimental Procedures. Batch experiments were per-formed in 43-mL lined, capped glass bottles (Wheaton) atroom temperature (20 ( 1 °C) in the dark under abioticconditions (filtered solutions and autoclaved bottles). Thetests were initiated by turning on an end-over-end rotatorat 47 rpm immediately after the additions of reactants intothe bottles. The time period between the two events waswithin 5 s. The test solutions were not buffered against pHchange to prevent any potential interference. Samples weretaken out at regular intervals, and these were subjected tobromate, bromide, dissolved iron, and dissolved organiccarbon (DOC) measurements after filtration.

A series of batch experiments were conducted to evaluatethe inhibitive effects of humic acid on bromate reduction byFe(0) and Fe(II) and to study the redox reactivity of humicacid, lawsone, and AQDS on bromate reduction. In the firstcase, two different scenarios with respect to the mode ofaddition of humic acid were considered. First, the humicacid was added at different concentrations (0-35 mg/L asDOC) simultaneously with a fixed concentration of bromate(5 mg/L) into the reactors, each filled with 1 g of preweighedirons. In the second scenario, to study the effect of ac-cumulation of humic acid on bromate reduction by Fe(0) inthe long-term run, irons were presoaked in humic acidsolutions of various concentrations for 24 h and then thehumic acid-amended irons were vacuum freeze-dried andused for bromate reduction under conditions similar to thosein the first case, but without humic acid addition. In both thescenarios, no adjustment to the solution pH was made. Theexperiments on the inhibitive effect of humic acid on bromatereduction by Fe(II) were conducted in the same manner, butinstead of using Fe(0) powders, 2 mL of a freshly preparedferrous stock solution was pipetted into the bottle to get atarget concentration of 20 mg/L. The effects of lawsone andAQDS on bromate reduction by Fe(0) were evaluated in thesame manner as the first scenario described above, exceptthat the humic acid was replaced with either lawsone orAQDS at predetermined concentrations.

In the case of evaluating the redox reactivity of humicacid, lawsone, or AQDS on bromate reduction in the absenceof Fe(0), the batch experiments were conducted with solutionscontaining bromate (5 mg/L) and humic acid/lawsone/AQDS(45 mg/L each as DOC). The experiments were conductedwith or without the presence of Fe(III), and the initial solutionpH was adjusted between 3.2 and 8.7 by the addition ofH2SO4 or NaOH. The concentration of Fe(III) was maintainedat 20 mg/L. The reduction of bromate with variations in humicacid concentrations (0, 10, 30, and 45 mg/L) and 20 mg/LFe(III) at pH 3.2-3.9 was also tested. Fe(III) reduction byhumic acid alone was studied with solutions containinghumic acid (45 mg/L as DOC) and 20 mg/L Fe(III). The pHwas maintained at 3.9, and no bromate was added.

Analytical Methods. Bromate and bromide concentra-tions were measured with an ion chromatograph (Dionex500) using an AS-9HC analytical column. The total solubleiron concentration in the filtered samples was determinedwith an atomic absorption spectrometer (SpectrAA 220FS).Humic acid, lawsone, and AQDS were quantified as dissolvedorganic carbon by a TOC analyzer (TOC-5000A, Shimadzu).

FTIR spectra of humic acid and complexes were recordedby a Fourier transform infrared spectrometer (Bio Rad FTS6000) equipped with a DTGS detector. The spectral resolutionwas 4 cm-1. Purified humic acid powders were mixed withan aliquot amount of spectroscopic grade KBr. The mixturewas finely ground with an agata ball and mill. Pellets wereprepared using a pressing machine for further FTIR analysis.The reflectance Fourier transform spectrometer equippedwith a Bio-Rad UMA500 IR microscope was used for theanalysis of the iron surface adsorbed with or without NOMafter freeze-drying, with 128-time scanning.

The ESR spectra were obtained with a spectrometer (JEOLJES-TE ESR) at room temperature. Solid samples, includingFe(III) salts, humic acid, lawsone, and AQDS powders, andFe(III)-humic acid complexes after vacuum freeze-drying,were packed in quartz ESR tubes (Wilmad Glass710-SQ-250M)and placed into the ESR cavity. The ESR instrument wasoperated with the following parameters: a microwavefrequency of 9.45 GHz, a microwave power of 1.25 mW, anda modulation frequency of 100 kHz.

Data Analysis. Pseudo-first-order reaction kinetics wasused to fit bromate reduction by Fe(0) since Fe(0) was usedin excess:

where [BrO3-] is the bromate concentration at time t and

kobs is the observed rate constant.

Results and DiscussionBromate Reduction in an Fe(0)-H2O-Humic Acid System.Bromate reduction by Fe(0) was first examined by simul-taneously adding humic acid and bromate. The results ofthe experiment and the calculated kobs are presented in Figure1a. As shown, an increase in the humic acid concentrationled to a slight decrease in the bromate reduction rate. Theinitial pH of the test solutions varied in the range of 6.5-8due to the addition of different quantities of humic acid asno buffering was provided and slight increases in pH duringthe experiments due to proton consumption were observed.However, such a range of pH shall have little effect on thebromate reduction rate (6). Therefore, the decrease in thebromate reduction rate should be primarily attributable tothe increase in the humic acid concentration and somewhatto the increase in pH. During the reaction period, the aqueoushumic acid concentrations decreased continuously withincreasing reaction time. Meanwhile, the amount of dissolvedirons released to the aqueous phase increased with increasingreaction time and increasing initial humic acid concentration(see the inset in Figure 1a). This result may be explained bythe adsorption of humic acid on the iron surface and thecomplexation of humic acid with ferric and ferrous ions (20),which facilitates the detachment of metal ions from thesurface to the solution (24), thereby increasing the dissolutionof iron oxides and irons.

Several hypotheses were taken into consideration to findout the mechanisms behind the aforementioned observation.The accumulation of adsorbed humic acid on the iron surfacemay depress the bromate reduction rate. The complexationbetween humic acid and the generated Fe(II) may inhibitbromate reduction by Fe(II) or occupy the active surfacesites and inhibit iron corrosion. On the other hand, excavationof active surface sites by dissolution of passive iron oxidesmay accelerate the bromate reduction rate. The redox-activemoieties in humic acid or in iron-humic acid complexesmay act as reductants to reduce bromate to bromide or serveas electron-transfer mediators between Fe(0) and bromate.Therefore, further investigation was made to evaluate these

-d[BrO3

-]

dt) kobs[BrO3

-] (1)

VOL. 39, NO. 4, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 1093

inhibitory and promotional roles of humic acid in thechemical reduction of bromate.

Inhibition by Humic Acid. Considering the quick bromatereduction rate compared with the rate for humic acidadsorption as illustrated in Figure 1a, it may be concludedthat the humic acid may not have sufficient time to exert itseffects when added simultaneously with bromate. However,in the long run, the amount of humic acid adsorbed on theiron surface may increase, while the bromate is continuously

reduced, and this may eventually affect the rate of bromatereduction. Therefore, the inhibition role of humic acid wasevaluated by presoaking Fe(0) in a humic acid solution for24 h and then using the presoaked Fe(0) in the bromatereduction experiments. Different initial humic acid concen-trations were used for this purpose, and the results andcalculated kobs are presented in Figure 1b. After presoaking,the disappearance of bromate as a function of time couldnot be described very well by pseudo-first-order reaction

FIGURE 1. Effect of humic acid on bromate reduction by Fe(0), [BrO3-]0 ) 5 mg/L: (a) added simultaneously; (b) humic acid pretreatment;

(c) by Fe(II).

1094 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 4, 2005

kinetics. On the other hand, the initial rates in Figure 1bseem to be the same, irrespective of humic acid loading.Only after a certain decrease in the initial bromate concen-tration, the reaction becomes retarded and the onset of the“slowing down” seems to depend on the initial humic acidloading. The similarity in the initial rates is hypothesized toresult from the existence of specific active surface sitesdesignated for bromate reduction, not for humic acidadsorption; with the reactions proceeding to exhaust thosespecific sites, the common active sites (available for bothbromate reduction and humic acid adsorption) becomelimited. More studies need to be conducted to evaluate thistwo-stage trend in the future. Compared to the curves shownin Figure 1a, the 24 h presoaking period apparently facilitatedthe adsorption of humic acid on the iron-water interface,and its accumulation occupied the iron surface as indicatedin the literature (10) to passivate the iron surface and thereforeprevent electron transfer from the inner Fe(0) to bromate.The adsorption of humic acid on the iron surface led to thelatter, large rate changes. On the other hand, the slightdecreases in the initial bromate reduction are attributableto the formation of passive iron oxide during hydration.Meanwhile, during bromate reduction, the dissolution of ironto the aqueous solution was less significant in quantity (seethe inset in Figure 1b), and no humic acid was found torelease back to the solution phase. In addition, during thepresoaking step, it was found that the release of dissolvediron increased from 9.4 to 34.2 mg/L with increasingconcentrations of the presoaked humic acid from 20 to 35mg/L. These findings suggest that the complexation of humicacid with Fe(II) or Fe(III) facilitates the release of dissolvediron into the aqueous phase only during the presoakingperiod.

Since Fe(II) could be generated by iron corrosion duringthe oxidation of Fe(0) and the presence of humic acidpromoted such a process, we felt that it was necessary toinvestigate the effects of humic acid on bromate reductionby Fe(II) in the aqueous phase. Batch tests were, therefore,conducted with aqueous solutions containing 20 mg/LFe(II), 5 mg/L bromate, and two different humic acidconcentrations (10-30 mg/L). The results of the tests arepresented in Figure 1c. As illustrated in the figure, an increasein the humic acid concentration decreased the bromatereduction rate, and the solution pH remained relatively stableat 5.1 even though no pH buffering was provided. This furtherindicated that the quick complexation between humic acidand Fe(II) dominated the available active Fe(II) needed forbromate reduction. Although this set of tests was conductedin the aqueous phase, the complexation is also likely to occurat the iron-water interface and prevents the active Fe(II)from participating in bromate reduction on the surface.

Redox Role of Humic Acid. Bromate reduction by humicacid alone was evaluated at various pH levels to examine theredox role of humic acid. The results of the batch experimentsare presented in Figure 2a. As shown, bromate reduction byhumic acid alone was possible, but the reduction wasextremely slow. As a result, only a small drop in the bromateconcentration could be observed after 284 h. The reductionrate decreased with increasing pH from 3.2 to 8.7, and almostno bromate reduction was observed at pH 8.7 at the end ofthe reaction period. In another set of experiments, theoxidation role of humic acid on bromate reduction in thepresence of Fe(III) was evaluated at various pH values andhumic acid concentrations. Figure 2b describes the disap-pearance of bromate at different pH levels with the samehumic acid concentration. As shown in the figure, thebromate removal efficiency by humic acid and Fe(III)decreased with increasing pH. Nevertheless, a comparisonwith the case of humic acid alone shows that the additionof Fe(III) into solutions increased the bromate removal rate

significantly. Figure 2c shows the bromate disappearance ina humic acid-Fe(III) system at different humic acid con-centrations at pH 3.2-3.9. As the figure illustrates, thebromate reduction rate was affected by the concentration ofthe humic acid in this system. An increase in the humic acidconcentration increased the bromate removal rate. It isimportant to note that no bromate reduction was observedwhen the solution contained only Fe(III) at the pH level usedin this experiment. The concentration of aqueous bromide,a reduction product, increased with decreasing bromateconcentrations in all cases with a bromine recovery ofapproximately 90%. As such, the removal of bromate shouldbe primarily attributable to the chemical reduction process.

The bromate reduction rate decreased with increasingsolution pH in the solutions containing humic acid only andin the humic acid plus Fe(III) solutions. There are severalpossible reasons for this. First, Steelink (25) observed thathumic substances under basic conditions have higher andstable radical concentrations than those under acidic condi-tions. At higher pH, the quinone moieties have the tendencyto attract electrons to form stable free radicals even withoutthe presence of reducing groups (26). If the quinone andphenol groups are close to one another, electron-transferreactions are possible, and thereby, radicals can be formedand stabilized in basic solutions (27). Second, the morphologyand structure of humic substances are greatly influenced bypH (28). Humic acid looks like fibers or a bundle of fibers atacidic or neutral pH, while its shape becomes sheetlike atbasic pH. Finally, as proposed by Chen et al. (17), thehydrolysis of Fe(III) at basic pH forms hydroxyliron(III)species. Precipitated amorphous iron hydroxides may beanother reason for the pH dependence of the Fe(III) reductionby humic acid. In this study, therefore, the observed declinein the bromate reduction rate with both the increasing pHand the decreasing initial humic acid concentration isspeculated to be influenced by the changes in the formationof the Fe(III)-humic acid complexes, in addition to thereducing reactivity of humic acid itself. However, only a verylow concentration of Fe(II) was detected in the aqueousphase, which may be attributable to the quick reoxidationof the formed Fe(II) to Fe(III) during bromate reduction.Therefore, additional tests were conducted in a water matrixcontaining humic acid and Fe(III) only to verify the generationof Fe(II).

An increase in Fe(II) concentration (up to nearly 2 mg/L)after 100 h as a result of the reduction of Fe(III) by humicacid was observed at pH 3.9. Fe(III)-humic acid complexes(in precipitates) were also observed at the bottom of thebottles. These findings suggest that the humic acid containsreduced functional groups that complex with Fe(III) andconsequently lead to the formation of Fe(II). Visible lightirradiation yielded higher dissolved Fe(II) concentrations ascompared with the yields obtained in the dark, indicatingthe photocatalytic behavior of the reduction of Fe(III) toFe(II) by humic acid. The abiotic photoreduction of Fe(III)to Fe(II) has been observed in aqueous solutions containingtannic acid (29) and humic substances (26).

On the basis of the arguments presented above, anelectron-transfer scheme is proposed to account for the fasterbromate reduction rate observed in the solution containinghumic acid and Fe(III); this scheme is illustrated in Figure3a. As the figure shows, the Fe(III)/Fe(II) couple acts as acatalyst for bromate reduction by humic acid. The reducedfunctional groups in the humic acid complex with Fe(III)and then transfer electrons to Fe(III) to form Fe(II). TheFe(II) thus formed donates the accepted electrons to bromate,the terminal electron acceptor. A similar role of the Fe(III)/Fe(II) redox couple as a catalyst has also been observed byMiles and Brezonik (29) in a study of oxygen depletion inhumic-containing water. The above reaction process is


expected to occur in the environment, since the reducingcapacity of Aldrich humic acid was reported to be lower thanthat of the soil and aquatic humic acid (18). In the processof bromate reduction by Fe(0), a more comprehensive processinvolving Fe(III) and humic acid may also take place inaddition to the direct reaction between Fe(0) and bromate.In such a process, the oxidized functional groups in the humicacid may be reduced by the bulk reductant, Fe(0), and then

donate electrons to electron acceptors. This electron-transferprocess has been proposed to account for the role of naturallyoccurring electron mediators in the reduction of 4-amino-azobenzen by Fe(0) (30). Figure 3b shows that two kinds ofelectron-transfer mediators are involved, the humic acid andthe Fe(III)/Fe(II) couple, after humic acid is adsorbed on theiron surface. The humic acid acts to mediate electrons fromFe(0) to the electron acceptors. Since Fe(III) has a higher

FIGURE 2. Bromate reduction in the solution containing (a) humic acid under various pH conditions, (b) humic acid and Fe(III) under variouspH conditions, and (c) humic acid and Fe(III) under various humic acid concentrations.


priority to accept the electrons as compared to BrO3-, Fe(II)

will be formed, which will thereafter mediate the electronsto the terminal electron acceptor, BrO3

-. On the other hand,when surface passivation occurs, the reactive Fe(II) can be,alternatively, reductively regenerated on the formed outeroxide layer by the reducing functional groups of humic acid.Additionally, direct electron transfer between Fe(III) andFe(0) at the inner Fe(0)-Fe(III) interface may also occur andcannot be ruled out.

Effects of Quinone Model Compounds on BromateReduction. Quinone moieties in humic acid have beenidentified recently as the electron-transfer mediator in redoxreactions (15, 21). Two quinone model compounds, lawsoneand AQDS, which have been commonly used to representqunione moieties in humic substances to study the redoxfunction of quinone moieties in humic acid (10, 18, 21), werechosen. Parts a and b of Figure 4 show the results of bromatereduction with Fe(0) when the bromate was simultaneouslyadded with lawsone and AQDS, respectively. Changes in theconcentrations of lawsone and AQDS yielded only small,

insignificant (within 2% of the experimental errors) changesin the bromate reduction rate. Meanwhile, the complexationand adsorption onto the iron surface caused the removal oflawsone or AQDS from the aqueous phase. The dissolvediron concentration increased with prolonged reaction timeand also with increasing lawsone or AQDS concentrations.

The reduction of bromate by lawsone and AQDS was alsostudied with or without the presence of Fe(III). These resultsare presented in parts a and b, respectively, of Figure 5.Without Fe(III), only lawsone at pH 4.1 demonstrated notablebromate reduction over a 168 h time span. There was nosignificant change in the bromate concentration in allremaining cases. When Fe(III) was added into a solutioncontaining bromate and lawsone, the bromate reduction ratewas enhanced significantly. However, the addition of AQDS,even in the presence of Fe(III), did not remove bromate overa reaction time of about 180 h. These findings are attributableto the differences in the chemical structures of lawsone andAQDS. Although both compounds contain quinone moieties,lawsone contains an additional phenolic group, which hasbeen identified as a reductant group in NOM (7), and maydonate electrons to bromate, the electron acceptor. Usually,quinones are reduced to semiquinone radicals, and thecontinuous electron transfer during such a reduction resultsin the formation of hydroquinone (31). As such, quinone isan oxidized form and does not have any reduction capabilityin the presence of other oxidants such as bromate or Fe(III).Only if other reductants such as HS- or Fe2+ reduce AQDSto AHQDS, it can mediate the accepted electrons to thecontaminants of interest to facilitate chemical reduction (12,13).

Spectroscopic Analysis. Information on the specificfunctional groups of the humic acids adsorbed on the ironsurface was provided by FTIR spectroscopy. Figure 6 showsthe spectra of the humic acid powders (line a), of the iron

FIGURE 3. Electron-transfer process (a) from the electron donor(humic acid) to the electron acceptor (bromate) via Fe(III) and (b)from the electron donor (Fe(0)) to the electron acceptor (bromate)via humic acid and Fe(III).

FIGURE 4. Bromate reduction by Fe(0) in the presence of (a) lawsoneand (b) AQDS.

FIGURE 5. Bromate reduction by lawsone and AQDS under variouspH conditions.


surface after humic acid adsorption (line b), and of the ironsurface without adsorbed humic acid (line c). On the basisof the band assignments obtained from the literature (32-34), the following major bands are observed: 3429 cm-1

(hydrogen bonding and O-H stretching), the doubletbetween 2920 and 2851 cm-1 (C-H stretch of aliphaticgroups), 1580 cm-1 (aromatic CdC), 1389 cm-1 (symmetricstretching of COO-, C-OH stretching of phenolic OH), and1099/1034 cm-1 (C-O stretching of carbohydrate or alcohol).After adsorption, the observations of the shifts at stretchingbands of OH, CdC, and COO- or phenolic OH, althoughsome of which may not be conclusive, are consistent withmany studies of NOM adsorption on surface minerals,especially the surfaces of iron oxides (24, 35, 36). Similarpeak shifts at the sites of OH, CdC, and COO- were alsoobserved in the spectra of the formed Fe(III)-humic acidcomplex (not shown). This limited FTIR evidence suggeststhat the aromatic COOH and hydroxyl or phenolic OH arethe main functional groups that complex with and adsorbonto the iron oxide surfaces. It has been reported in theliterature that the adsorption of these functional groups onthe metal surface can shift the electron densities to the centralmetal ions at the surface to facilitate the breakage of metal-oxygen lattice bonds, leading to the detachment of the centralmetal ion into the solution (24).

ESR spectroscopy can supply useful information on themolecular structure of humic substances (27) or about thestereochemistry and the type of coordination sites of metal-ligand bonding (37, 38). The ESR spectral parameters suchas g factors and line width, unaltered with various experi-mental and laboratory factors, can be compared and relatedto other chemical and physicochemical properties. Similarg values point out that the free radicals of different types ofNOM are of similar nature (27). In this study, the ESR spectraof humic acid and lawsone, which are shown in Figure 7a,were very similar with obtained g values of 1.996 and 1.995,respectively. The ESR signal was caused by the presence ofquinone hydrate from the reduction of quinone, includingsemiquinone and hydroquinone (27). On the other hand, noESR signal was observed in the AQDS samples as expectedsince AQDS contains quinone moieties only. It has beenreported in the literature that the radical concentration ofAQDS increased from below a detection limit to 7 × 1018

spins/g after microbial reduction and the ESR signal com-pletely disappeared when oxygen was introduced (21). TheESR spectra of the humic acid and lawsone in this studywere in agreement with the literature that the ESR spectraof humic acid and fulvic acid are qualitatively described by

a sharp and narrow resonance at g ≈ 2 (39), which is consistentwith semiquinone radical units possibly conjugated toaromatic rings (40). On the basis of the chemical structureof lawsone, the similar ESR signals for lawsone and humicacid should be attributable to the same nature of radicals,which would be quinone-phenol groups. Steelink and Tollin(41) proposed that the quinone-phenol groups in hydro-anthraquinones, tannies, lignins, solid humic, and fulvic acidsaccount for their similar ESR properties. No ESR signal wasobserved in a solution containing glycine (42) and in solidhydroquinone and 2,3-dihydroxybenzoic acid used as thereferences of phenolic and/or carboxylate groups in our study.It should be noted that although iron porphyrin has beenstudied in other studies as an electron mediator (43, 44), theESR analyses in the current study showed the presence ofquinone-phenol groups only. Iron porphyrin is commonlyrecognized as a functional group in hemoproteins (45, 46),but it was reported to be a trace component in nationalorganic matter with concentrations of sub-µg/L in naturalwater (47).

The ESR spectra of Fe(III) and Fe(III)-humic acidcomplexes were also recorded (Figure 7b). A broad resonancesignal centered near g ) 2, which is an envelope of severalresonances from Fe(III), and a little shoulder at g ) 4 wereobtained in the Fe(III)-humic acid complex. Usually, themagnetic parameters of the Fe(III)-NOM complexes havebeen observed to exist at a g value of 2 or at a g value of 4(38). Since the ESR spectra of Fe(III)-humic acid complexesmainly center at a g value of 2, this indicated that mostFe(III) is bound to phenolic and/or carboxylic groups atoctahedral sites with little or no axial distortion from thecubic symmetry ligand field, where Fe(III) has been reported

FIGURE 6. FTIR spectra of humic acid (a) before and (b) afteradsorption on the iron surface. (c) FTIR spectrum of the iron surfacewithout adsorption.

FIGURE 7. ESR spectra of (a) humic acid and lawsone and (b) ferricand humic acid-Fe(III) complexes.


to be easily reduced (48). The little shoulder at g ) 4 indicatesthat there is a small amount of Fe(III) coordinating with humicacid in the tetrahedral or octahedral sites in a low-symmetryligand field, where Fe(III) is hard to be reduced (48).

Environmental Implications. In this study, batch testsshow that humic acid has relatively little inhibitory effect onthe rates of bromate reduction by Fe(0), wherein the ratesare only retarded by a factor of about 1.3-2.0, not by ordersof magnitude, due to the adsorption of humic acid. Theadsorption of humic acid also inhibits iron corrosion, therebyprolonging the lifetime of the zerovalent iron. On the otherhand, the adsorbed humic acid can transfer electrons fromthe inner Fe(0) to Fe(III) to reduce bromate in solution. TheFe(III)-humic acid complexes formed on the outer oxidelayer or in solution can regenerate reactive Fe(II) to reducebromate; however, at a much slower rate (kobs of ap-proximately 0.01 min-1, based on the data and conditions inFigure 1c). Thus, it is expected that when the bromatereduction rate by Fe(0) is reduced to a certain extent due toboth humic acid and iron oxide coverage, there will be aturning point at which such bromate reduction by regener-ated Fe(II) becomes prevalent. If the quantity of the Fe(III)-humic acid complexes is sufficient, the amount of Fe(II)regenerated may be high enough to maintain iron surfaceactivation and reduction of bromate at a typical concentrationof 0-127 µg/L in drinking water to prolong the lifetime ofthe zerovalent iron. In the remediation of halogenated ornitro aromatic compound contaminated groundwater usingFe(0), the electron-transfer function of humic acid shall bemore significant due to the refractory nature of thesepollutants.

AcknowledgmentsThis study was supported in part by the Hong Kong ResearchGrants Council under Competitive Earmarked Research GrantHKUST6106/03E.

Literature Cited(1) USEPA. National Primary Drinking Water Regulations: Disin-

fectants and Disinfection Byproducts. Final Rule. Fed. Regist.1998, 63 (241), 69390.

(2) Mari, A.; Takaka, A. Occurrence and control of bromate inaqueous media. J. Health Sci. 1999, 45, 344-355.

(3) Krasner, S. W.; Glaze, W. H.; Weinberg, H. S.; Daniel, P. A.; Najm,I. Formation and control of bromate during ozonatioin of waterscontaining bromide. J.sAm. Water Works Assoc. 1993, 85, 73-81.

(4) Gordon, G.; Gauw, R. D.; Emmert, G. L.; Walters, B. D.; Bubnis,B. Chemical reduction methods for bromate ion removal. J.sAm. Water Works Assoc. 2002, 94, 91-98.

(5) Siddiqui, M.; Amy, G.; Ozekin, K.; Zhai, W. Y.; Westerhoff, P.Alternative strategies for removing bromate. J.sAm. Water WorksAssoc. 1994, 86, 81-96.

(6) Xie, L.; Shang, C. Bromate reduction by zero valent iron in batchtests. Submitted for publication to J.sAm. Water Works Assoc.

(7) Stumm, W. Aquatic Surface Chemistry, Chemical Processes atthe Particle-Water Interface; John Wiley & Sons: New York, 1987.

(8) Gordon, E.; Brown, J. R. Sorption of trace elements on mineralsurfaces: modern perspectives from spectroscopic studies, andcomments on sorption in the marine environment. Int. Geol.Rev. 2001, 43, 963-1073.

(9) Klausen, J.; Vikesland, P. J.; Kohn, T.; Burris, D. R.; Ball, W. P.;Roberts, A. L. Longevity of granular iron in groundwatertreatment processes: solution composition effects on reductionof organohalides and nitroaromatic compounds. Environ. Sci.Technol. 2003, 37, 1208-1218.

(10) Tratnyek, P. G.; Scherer, M. M.; Deng, B. L.; Hu, S. D. Effects ofnatural organic matter, anthropogenic surfactants, and modelquinones on the reduction of contaminants by zero-valent iron.Water Res. 2001, 35, 4435-4443.

(11) Bouayed, M.; Rabaa, H.; Srhiri, A.; Saillard, J.-Y.; Ben Bachir, A.;Le Beuze A. Experimental and theoretical study of organiccorrosion inhibitors on iron in acidic medium. Corros. Sci. 1999,41, 501-517.

(12) Curtis, G. P.; Reinhard, M. Reductive dehalogenation ofhexachloroethane, carbon tetrachloride, and bromoform byanthrahydroquinone disulfonate and humic Acid. Environ. Sci.Technol. 1994, 28, 2393-2401.

(13) Dunnivant, F. M.; Schwarzenbach, R. P. Reduction of substitutednitrobenzenes in aqueous solutions containing natural organicmatter. Environ. Sci. Technol. 1992, 26, 2133-2141.

(14) Royer, R. A.; Burgos, W. D.; Fisher, A. S.; Unz, R. F.; Dempsey,B. A. Enhancement of biological reduction of hematite byelectron shuttling and Fe(II) complexation. Environ. Sci. Technol.2002, 36, 1939-1946.

(15) Lovley, D. R.; Coates, J. D.; Blunt-Harris, E. L.; Phillips, E. J. P.;Woodward, J. C. Humic substances as electron acceptors formicrobial respiration. Nature 1996, 382, 445-448.

(16) Voelker, B. M.; Morel, F. M. M.; Sulzberger, B. Iron redox cyclingin surface waters: effects of humic substances and light. Environ.Sci. Technol. 1997, 31, 1004-1011.

(17) Chen, J.; Gu, B.; Royer, R. A.; Burgos, W. D. The role of naturalorganic matter in chemical and microbial reduction of ferriciron. Sci. Total Environ. 2003, 307, 167-178.

(18) Kappler, A.; Haderlein, S. B. Natural organic matter as reductantfor chlorinated aliphatic pollutants. Environ. Sci. Technol. 2003,37, 2714-2719.

(19) Struyk, Z.; Sposito, G. Redox properties of standard humic acids.Geoderma 2001, 102, 329-346.

(20) Tipping, E. Cation binding by humic substances; CambridgeUniversity Press: New York, 2002.

(21) Scott, D. T.; Mcknight, D. M.; Blunt-Harris, E. L.; Kolesar, S. E.;Lovley, D. R. Quinone moieties act as electron acceptors in thereduction of humic substances by humic-reducing microor-ganisms. Environ. Sci. Technol. 1998, 32, 2984-2989.

(22) Benz, M.; Schink, B.; Brune, A. Humic acid Reduction bypropionibacterium freudenreichii and other fermenting bac-teria. Appl. Environ. Microbiol. 1998, 64, 4507-4512.

(23) O’Loughlin, E. J.; Burris, D. R.; Delcomyn, C. A. Reductionvedechlorination of trichloroethene mediated by humic-metalcomplexes. Environ. Sci. Technol. 1999, 33, 1145-1147.

(24) Stumm, W. Chemistry of the Solid-Water Interface: Processes atthe Mineral-Water and Particle-Water Interface in NaturalSystems; John Wiley & Sons: New York, 1992.

(25) Steelink, C. Electron paramagnetic resonance studies of humicacid and related model compounds. Coal Science; Advances inChemistry Series, No. 55; American Chemical Society: Wash-ington, DC, 1966; pp 80-90.

(26) Steelink, C.; Fitzpatrick, J. D. The origin of the paramagneticspecies in lignin solutions. Autoreduction of 2,6-dimethoxy-benzoquinone and related quinines to radicals anions in alkalinesolutions. J. Org. Chem. 1972, 37, 762-767.

(27) Senesi, N.; Steelink, C. Application of ESR spectroscopy to thestudy of humic substances. In Humic Substances II: In Searchof Structure; Hayes, M. H. B., MacCarthy, P., Malcolm, R. L.,Swift, R. S., Eds.; John Wiley & Sons: New York, 1989.

(28) Chen, Y.; Schnitzer, M. Sizes and shapes of humic substancesby electron microscopy. In Humic Substances II: In Search ofStructure; Hayes, M. H. B., MacCarthy, P., Malcolm, R. L., Swift,R. S., Eds.; John Wiley & Sons: New York, 1989.

(29) Miles, C. J.; Brezonik, P. L. Oxygen consumption in humic-colored waters by a photochemical ferrous-ferric catalytic cycle.Environ. Sci. Technol. 1981, 15, 1089-1095.

(30) Weber, E. J. Iron-mediated reductive transformations: inves-tigation of reaction mechanism. Environ. Sci. Technol. 1996,30, 716-719.

(31) Larson, R. A.; Weber, E. J. Reaction Mechanisms in EnvironmentalOrganic Chemistry; Lewis Publishers: Boca Raton, FL, 1994.

(32) Chen, J.; Gu. B.; LeBoeuf, E. J.; Pan, H.; Dai, S. Spectroscopiccharacterization of the structural and functional properties ofnatural organic matter fractions. Chemosphere 2002, 48, 59-68.

(33) Tanaka, T.; Nagao, S.; Ogawa, H. Attenuated total reflectionfourier transform infrared (ATR-FTIR) spectroscopy of functionalgroups of humic acid dissolving in aqueous solution. Anal. Sci.2001, 17 (Suppl.), i1081-i1083.

(34) Gu, B.; Schmitt, J.; Chen, Z.; Liang, L.; McCarthy, J. F. Adsorptionand desorption of natural organic matter on iron oxide:mechanisms and models. Environ. Sci. Technol. 1994, 28, 38-46.

(35) Murphy, E. M.; Zachara, J. M.; Smith, S. C. Influence of mineral-bound humic substances on the sorption of hydrophobic organiccompounds. Environ. Sci. Technol. 1990, 24, 1507-1516.

(36) Sposito, G. The Surface Chemistry of Soils; Oxford UniversityPress: New York, 1984.


(37) Senesi, N. Metal-humic substances complexes in the environ-ment. Molecular and mechanistic aspects by multiple spec-troscopic approach. In Biogeochemistry of Trace Metals. Adriano,D. C. Ed.; Lewis Publishers: Boca Raton, FL, 1992.

(38) EI-Sonbati, A. Z.; EI-Bindary, A. A. Stereochemistry of newnitrogen containing Aldehydes. V. Novel synthesis and spec-troscopic studies of some quinoline Schiff bases complexes.Pol. J. Chem. 2000, 74, 621-630.

(39) Olk, D. C.; Brunetti, G.; Senesi, N. Decrease in humification oforganic matter with intensified lowland rice cropping: A wetchemical and spectroscopic investigation. Soil Sci. Soc. Am. J.2000, 64, 1337-1347.

(40) Senesi, N. Application of electron spin resonance (ESR) spec-troscopy in soil chemistry. Adv. Soil Sci. 1990, 14, 77-130.

(41) Steelink, C.; Tollin, G. Biological polymers related to catechol:electron paramagnetic resonance and infrared studies ofmelanin, tannin, lignin, humic acid and hydroxyquinones.Biochim. Biophys. Acta 1966, 112, 377-379.

(42) Perlinger, J. A.; Kalluri, V. M.; Venkatapathy, R.; Angst, W.Addition of hydrogen sulfide to juglone. Environ. Sci. Technol.2002, 36, 2663-2669.

(43) Wade, R.; Castro, C. E. Oxidation of iron(II) porphyrins by alkylhalides. J. Am. Chem. Soc. 1973, 95, 226-230.

(44) Perlinger, J. A.; Buschmann, J.; Angst, W.; Schwarzenbach, R. P.Iron porphyrin and mercaptojuglone mediated reduction ofpolyhalogenated methanes and ethanes in homogeneousaqueous solution. Environ. Sci. Technol. 1998, 32, 2431-2437.

(45) Sheldon, R. A. Oxidation catalysis by metalloporphyrins: ahistoric perspective. In Metalloporphyrins in catalytic oxidations;Sheldon, R. A. Ed.; Marcel Dekker: New York, 1994.

(46) Lever, A. B. P.; Gray, H. B. Iron porphyrins Part III; Addison-Wesley: Reading, MA, 1983.

(47) Thurman, E. M. Organic geochemistry of natural waters; Nijhoff-Junk Publishers: Boston, MA, 1985.

(48) Senesi, N.; Griffith, S. M.; Schnitzer, M.; Townsend, M. G. Bindingof Fe3+ by humic materials. Geochim. Cosmochim. Acta 1977,41, 969-976.

Received for review June 28, 2004. Revised manuscript re-ceived November 11, 2004. Accepted November 24, 2004.

ES049027Z


role of humic acid and quinone model compounds in bromate reduction by zerovalent iron

Documents