Treatment ofPerchlorate-ContaminatedGroundwater Using Highly Selective,Regenerable Ion-ExchangeTechnologiesB A O H U A G U , * , †

G I L B E R T M . B R O W N , ‡ A N DC H E N - C H O U C H I A N G §

Environmental and Chemical Sciences Divisions, Oak RidgeNational Laboratory, Oak Ridge, Tennessee 37831, and CalgonCarbon Corporation, Pittsburgh, Pennsylvania 15205

Treatment of perchlorate-contaminated water using highlyselective, regenerable ion-exchange and perchlorate-destruction technologies was demonstrated at a field sitein California. Four treatment and four regenerationcycles were carried out, and no significant deteriorationof resin performance was noted in 2 years. The bifunctionalresin (Purolite A-530E) treated about 37 000 empty bedvolumes (BVs) of groundwater before a significantbreakthrough of perchlorate occurred at an average flowrate of 150 gpm (or 1 BV/min) and a feed perchlorateconcentration of about 860 µg/L. Sorbed perchlorate (∼20kg) was quantitatively recovered by eluting with as littleas 1 BV of the FeCl3-HCl regenerant solution. The elutedClO4

- was highly concentrated in the third quarter ofthe first BV of the regenerant solution with a concentrationup to 100 000 mg/L. This concentrated effluent greatlyfacilitated subsequent perchlorate destruction or recoveryby precipitation as KClO4 salts. High perchlorate destructionefficiency (92-97%) was observed by reduction with FeCl2in a thermoreactor, which enabled recycling of the FeCl3-HCl regenerant solution, thereby minimizing the need todispose of secondary wastes containing ClO4

-. This studydemonstrates that a combination of novel selective,regenerable ion-exchange and perchlorate-destruction and/or recovery technologies could potentially lead toenhanced treatment efficiency and minimized secondarywaste production.

IntroductionPerchlorate (ClO4

-) has emerged as one of the most wide-spread contaminants found in sediments, groundwater, andsurface water (1-5), and cost-effective remediation tech-nologies are needed to remove trace quantities of ClO4

- fromcontaminated media. Among various treatment options, ion-exchange technology has long been used for water treatmentbecause of its simplicity, high capacity, and capability ofoperating at a relatively high flow rate with a small treatment

unit. Currently, the two most commonly used ion-exchangetechnologies for perchlorate removal in contaminated waterare (i) selective but non-regenerable strong-base anionexchange and (ii) nonselective or low-selective anion ex-change resins with sodium chloride (NaCl) brine regenera-tion. In the first case, the spent resin cannot be regeneratedby desorption from the resin using a conventional brinesolution (6-8), so the resin is discarded after it reaches itssorption capacity. The resin bed must be replaced, and thechange-out time depends on the feed ClO4

- concentrationand water quality. Additionally, the spent resin containingperchlorate has to be properly disposed of as hazardouswaste. In the second case, the resin can be regenerated byflushing with concentrated NaCl brine solution. However,because of its relatively low selectivity, the resin preferentiallyremoves other common anions such as sulfate and nitratein water. Because these anions exist in contaminatedgroundwater or surface water usually at orders of magnitudehigher concentrations than ClO4

-, they occupy most of ionexchange sites on the resin (>99%), resulting in an extremelylow sorption efficiency or capacity for ClO4

-. Therefore, theresin has to be regenerated frequently, producing largevolumes of secondary brine wastes containing ClO4

-. Thesefactors contribute to relatively high capital and operatingcosts for conventional ion-exchange technologies, which arediscussed in numerous reports and publications (2, 6, 8, 9).Alternatively, the bioremediation technique has been suc-cessfully demonstrated to remove ClO4

- from contaminatedwater and is cost-effective, especially for treatment ofcontaminated groundwater with relatively high ClO4

- con-centrations but poor water quality (e.g., with high organics,co-contaminants, and suspended solids) (2, 3, 10, 11).However, the technique could be ineffective or costly fortreatment of large plumes with a low ClO4

- (e.g., at tens orhundreds of µg/L concentration levels) because a highlyreducing environment has to be maintained for the bio-degradation of ClO4

-. The major challenge here is to sustainenough biomass to create continuous reducing conditionsat a high flow rate without adding substantial amounts ofelectron donors and/or acceptors (as a food source formicrobes). Additionally, many groundwater constituents suchas dissolved oxygen, nitrate, Fe(III), and Mn(IV) ions areknown to be preferred electron acceptors for biologicalreduction and have to be reduced before the degradation ofClO4

- occurs. Groundwater with extreme pH conditions(either acidic or alkaline) also has to be preconditioned tocreate a favorable environment for microbes to be functional(11). Furthermore, for drinking-water treatment, posttreat-ment is usually required to remove added nutrients, otheringredients, and/or potential pathogens (2, 3).

This work reports the first field-scale demonstration usinghighly selective ion-exchange technology for removing ClO4

-

from contaminated groundwater after four regenerationcycles using novel tetrachloroferrate (FeCl4

-) displacementtechniques (7). The new technology also enables either quan-titative destruction or recovery of eluted perchlorate for pos-sible reuse, thus overcoming problems of conventional throw-away ion-exchange and/or brine regeneration methodologies.The technology, including the bifunctional anion-exchangeresin (Purolite A-530E), the ferric chloride-hydrochloric acid(FeCl3-HCl) regeneration, and the perchlorate destruction/recovery, made it possible to recycle both the spent resinand the regenerant solution, leading to practically noproduction of secondary wastes containing ClO4

-. In brief,the bifunctional resin is composed of two quaternaryammonium groups: the first has a long alkyl chain for higher

* Corresponding author phone: (865) 574-7286; fax: (865) 576-8543; e-mail: gub1@ornl.gov.

† Environmental Sciences Division, Oak Ridge National Laboratory.‡ Chemical Sciences Division, Oak Ridge National Laboratory.§ Calgon Carbon Corporation.

Environ. Sci. Technol. 2007, 41, 6277-6282

10.1021/es0706910 CCC: $37.00 2007 American Chemical Society VOL. 41, NO. 17, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 6277Published on Web 07/28/2007

selectivity, and the second has a shorter alkyl chain forimproved reaction kinetics. Both laboratory and small-scalefield tests indicate that the resin is highly selective andefficient at removing ClO4

- from the contaminated water.For example, the resin was able to treat ∼100 000 bed volumes(BVs) of groundwater before a breakthrough of ClO4

- occurredunder continuous flow conditions (running at ∼2 BV/min atan influent ClO4

- concentration of ∼50 µg/L) (12). The newregeneration technique involves the use of FeCl4

- ions,formed in a solution of FeCl3 and HCl, to displace sorbedClO4

- on the spent resin bed (7); it is followed by a rinse withwater, in which the sorbed FeCl4

- dissociates by chemicalequilibrium and thus desorbs so that the resin bed isregenerated to its original state with chloride as counterionsby charge balance. This technique has been shown to behighly efficient, and nearly 100% recovery of the exchangesites can be achieved by rinsing with as little as 1 BV of theregenerant solution (this study) and <2 BVs in smalllaboratory column studies (13). Additionally, a new per-chlorate-destruction technology using ferrous chloride (FeCl2)was developed to reduce or degrade ClO4

- into Cl- and waterin spent regenerant solutions (14). While ClO4

- ions aredegraded, ferrous ions are oxidized to ferric ions so that theentire process renews the FeCl3-HCl regenerant solutionand allows it to be recycled.

The primary objectives of this field demonstration were,therefore, to evaluate the treatment efficiency and longevityof the bifunctional resin after regenerations and to determinethe efficacy and feasibility of the perchlorate destruction.Results of this study indicate that no significant deteriorationof the resin’s sorption capacity or performance after fourwater-treatment and resin-regeneration cycles over a periodof about 2 years. Regeneration has proven successful inrestoring the ion-exchange sites, and the eluted ClO4

- wasfound to be highly concentrated in about a quarter BV of theregenerant solution, enabling effective recovery or destruc-tion of perchlorate.

Experimental SectionThe treatment site is located at the Aerojet property in RanchoCordova, CA, where perchlorate was first detected ingroundwater in 1997. Since 1953 Aerojet has been involvedin manufacturing and testing rocket engines using liquidand solid propellants for military and commercial use (15).Some wastes from these manufacturing activities weredisposed of in surface impoundments, landfills, and openburn areas. As a result of these former disposal practices,groundwater has been contaminated at various sites withinor near the Aerojet property. Groundwater characteristics atthe demonstration site (EW 4220) include major competinganions such as nitrate and chloride at 10 mg/L, sulfate at 15mg/L on average, and a pH of 7.3. The influent ClO4

-

concentration was relatively constant at 830-890 µg/L ClO4-

during the water-treatment phase. This particular ground-water source with a relatively high ClO4

- concentration wasused to accelerate the breakthrough and thus maximize thenumber of water-treatment and regeneration cycles withinthe limited demonstration period.

The treatment system consisted of a primary resin canisterfor water treatment and backup resin canisters for polishing.The primary resin canister had a capacity of 270 gal but wasfilled with only 150 gal of the bifunctional anion-exchangeresin Purolite A-530E. Eight lateral distributors (with 0.01-in.slots) were installed at both the top and bottom of the canisterfor even distribution of the water flow. Groundwater waspumped directly from the extraction well to the resin canister(from top to bottom) after passing through a series of bagfilters (5-µm cutoff) to remove suspended particles andsediments. The system was continuously operated at about150 gpm or 1 BV/min. The effluent ClO4

- concentrations

were monitored daily during treatment and analyzed inaccordance with U.S. Environmental Protection AgencyMethod 314.0 using an ion chromatograph equipped withDionex AS16 analytical and AG16 guard columns (7, 16, 17).The analytical detection limit was about 1 µg/L ClO4

-.When the breakthrough of ClO4

- was observed in theprimary resin canister, it was taken offline for regenerationusing a mixed solution of FeCl3 (1 M) and HCl (4 M) (7, 13).The spent resin canister was transported off-site to the CalgonCarbon Corporation’s facility at Pittsburgh, PA, for regenera-tion and perchlorate destruction processes. Because of therelatively high sorption capacity of the resin and the longevityof each treatment cycle, studies were focused on only theprimary resin canister to maximize the number of water-treatment and regeneration cycles within the limited dem-onstration period. A dedicated regeneration canister (2 ft indiameter and 8 ft in height) was used for the regeneration.Spent resin beads were transferred from the treatment vesselto the regeneration canister and drained. The regenerantsolution was then pumped through the resin bed at a flowrate of ∼0.6 gpm until about 6 BVs of the regenerant solutionhad passed through the resin bed. The effluent regenerantsolution was periodically sampled for analysis and collectedin tanks approximately every quarter-BV so a mass-balanceanalysis of ClO4

- could be performed. The elution profilesof perchlorate, nitrate, and sulfate were determined as afunction of BVs of the regenerant passed through the resinbed. The eluted ClO4

- was usually concentrated in the thirdquarter BV of the regenerant solution, which was thensubjected to the perchlorate-destruction process usingferrous ion as a reducing agent at an elevated temperature(190 °C) in a thermoreactor (9, 14). A pilot-scale, flow-throughthermodestruction unit was constructed and capable ofrunning at about 1-2 gph (9). Because of the relatively lowvolume of spent regenerant solution needing treatment, thiswas sufficient capacity for perchlorate destruction.

Results and DiscussionResin Performance after Regenerations. The performanceof the treatment system was evaluated after four water-treatment and four regeneration cycles in nearly 2 years offield operation. The results of the first and second treatmentand regeneration cycles have been documented (18) andshowed that the initial breakthrough of ClO4

- occurred atapproximately 38 000 BVs at an influent ClO4

- concentrationof ∼830-890 µg/L. Figure 1 shows subsequent breakthroughcurves and the cumulative perchlorate removal after the twoinitial treatment and regeneration cycles. The breakthroughof ClO4

- was essentially the same as that observed using thevirgin resin (the initial breakthrough). About 36 000-38 000BVs of groundwater were treated before the breakthrough ofClO4

- occurred. The total amount of ClO4- removed by the

resin bed was about 20.1 kg each run (Figure 1). In comparisonwith the results obtained from a small field column experi-ment (12), the amount of groundwater treated in this study(∼38 000 BVs) appeared lower than that observed in theprevious study (∼100 000 BVs). This observation was pri-marily a result of a greatly increased feed ClO4

- concentration,which was more than 17 times greater than that observed inprevious studies (∼50 µg/L) (12). The difference is becausethe amount of groundwater which can be treated (or theamount of ClO4

- removed) depends not only on the anion-exchange capacity of the resin, but also on the distributioncoefficient (Kd) of the resin and the influent ClO4

- concen-tration. This relationship is commonly expressed as follows:

where Q is the amount of ClO4- sorbed by the resin, and C

is the ClO4- concentration. However, this relation is only

Q ) Kd × C

6278 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 17, 2007

valid within a limited concentration range by assuming Kd

is a constant. It cannot be extrapolated for estimating theamount of water treatment at different feed ClO4

- concen-trations because the sorption isotherms are usually nonlinearor the Kd is not a constant over a wide concentration range(12). The Kd value is also a function of the composition andconcentration of competing ions present in the groundwater(9). Therefore, although the feed ClO4

- concentration was 17times higher, the volume of groundwater treated was stillmore than one-third of that observed in the previous study.The relatively high feed ClO4

- concentration was used in thepresent study to accelerate the evaluation of resin perfor-mance after regenerations and to maximize the water-treatment and regeneration cycles within the limited dem-onstration time period. Our results (Figure 1) neverthelessdemonstrate and confirm the high sorption capacity andefficiency of the bifunctional resin for removing ClO4

- fromcontaminated groundwater. In addition, regeneration usingthe tetrachloroferrate displacement technique was successful,and we were able to recover most of the exchange sites inthe spent resin bed.

Regeneration and Elution Profiles of Perchlorate, Sul-fate, and Nitrate. Regeneration of the spent resin usingtetrachloroferrate displacement is illustrated in elutionprofiles of ClO4

- (Figure 2). In this process the effluentregenerant solution was collected every quarter BV to facilitatethe analysis of recovered ClO4

- and its mass balance. Theresults indicate that the sorbed ClO4

- was rapidly eluted andconcentrated in roughly the first BV of the regenerantsolution. The maximum ClO4

- concentration in the thirdquarter of the first BV reached as high as 100 000 mg/L (or∼1 M), representing a concentration factor of more than 5orders of magnitude in comparison with the ClO4

- concen-tration in groundwater. The total perchlorate recovered wasabout 20.1 kg (Figure 2), which is nearly identical to theamount removed during the treatment phase. The amountof perchlorate recovered in the third run was about 19.2 kg,which also compared favorably with the amount of per-chlorate loading (20.1 kg) before the regeneration (Figure 1).Therefore, regeneration using the tetrachloroferrate dis-placement yielded a recovery of 96-100% of the influentperchlorate. Given experimental error in the measurementsof the concentration and volume under field conditions, thesedata suggest a nearly complete regeneration and confirmour early laboratory findings that no significant deteriorationof the resin performance would occur after repeated loading

and regeneration (7 cycles) using the FeCl3-HCl regenerationtechnique (12). Although the total amount of ClO4

- recoveredmay vary depending on the initial loading, as describedearlier, the sorbed ClO4

- can be quantitatively recoveredregardless of the water quality. This statement is supportedby our laboratory experiments (7) as well as field experimentsat the Edwards Air Force Base, CA (18), at which thegroundwater contained much higher competing ion con-centrations such as sulfate (up to 180 mg/L) and chloride(up to 400 mg/L). These results are therefore in sharp contrastwith the extremely low efficiency of the regeneration usingconcentrated NaCl brines. Previous studies have shown thateven with a relatively nonselective resin and with counterflowof the brine, regeneration usually requires a large excess ofthe brine solution (6, 8, 19). With relatively selective anion-exchange resins, many BVs of 12% NaCl brine were able toremove only ∼6% of the loaded ClO4

- from the resin, andheating the ClO4

--laden resins during regeneration had onlylimited success (6).

The exceptionally high affinity of ClO4- for Type-I poly-

styrenic anion-exchange resins has been attributed toperchlorate’s low hydration energy and large ionic radius.The driving force for anion sorption is the net ion dehydrationand the electrostatic ion-pairing attraction energies (20, 21).The Type-I polystyrenic resins have a natural bias for sorbingpoorly hydrated anions such as ClO4

- because the polystyrenedivinylbenzene matrix is nonpolar or hydrophobic comparedwith matrices containing oxygen functional groups, such asthose containing polyacrylic esters and Type-II resin matrixes.It is thus not surprising that the affinity of singly chargedanions for Type-I polystyrenic resins is in the order ofperchlorate > nitrate > chloride (6, 8, 9). The correspondinghydration energy of these anions is in the order of perchlorate(∆G0 ) -205 kJ/mol), nitrate (∆G0 ) -314 kJ/mol), chloride(∆G0 ) -371 kJ/mol) (19, 20).

The efficacy of using the FeCl4--displacement technique

to regenerate spent resins loaded with ClO4- relies on the

fact that the FeCl4- is also a large, poorly hydrated anion and

known as one of the most strongly extracted anions fromHCl solutions by either liquid-liquid solvent extraction oranion exchange (7, 20-22). However, the FeCl4

- ion has amuch-desired chemical property: it dissociates in water ordilute acidic solution and forms positively charged Fe(III)species such as Fe3+, FeCl2+, and FeCl2

+, which are thenreadily eluted from the resin by charge repulsion. Therefore,the resin is regenerated to its original state with Cl- as thecounterion by charge balance. In practice the resin bed isready to be reused after regeneration by rinsing with about10-20 BVs of potable water. Because the rinsewater contains

FIGURE 1. Breakthrough curves (solid symbols) and mass removalof perchlorate (open symbols) after treatment and regenerationsusing the tetrachloroferrate displacement technique. The triangleand square symbols represent the third and fourth treatment andregeneration cycles, which compared favorably with the initialbreakthrough and perchlorate removal (circles) (18). The influentconcentration was 830-890 µg/L ClO4

-, and the flow rate was∼150 gpm.

FIGURE 2. Elution profiles (solid symbols) and the cumulativeamounts of ClO4

- recovery (open symbols) during the third (triangle)and fourth (square) regeneration of the spent resin bed after watertreatment. The regenerant solution consisted of 1 M FeCl3 and 4 MHCl. The flow rate was ∼0.6 gpm.

VOL. 41, NO. 17, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 6279

practically no detectable amounts of ClO4- ions, it can be

readily disposed of via a process drain after neutralizationand precipitation of residual Fe3+ in solution (e.g., by passingit through a carbonate gravel bag filter and, if necessary, asmall polishing resin canister).

The sorption affinity of perchlorate and its desorption byFeCl4

- are also illustrated in the elution profiles of nitrateand sulfate during the regeneration of the resin bed (Figure3). Both nitrate and sulfate are major competing ions sorbedby the resin during the water-treatment phase because theirconcentrations are orders of magnitude higher than that ofClO4

- in groundwater.However, during the regeneration these anions are

expected to be eluted off the resin bed more quickly thanClO4

- because the bifunctional resin shows relatively lowselectivities for nitrate and sulfate (23, 24). Indeed, resultsindicate that sulfate was eluted and concentrated in the firstquarter of the first BV of the regenerant (since the resin bedwas drained before regeneration). On the other hand, nitratewas eluted and concentrated in the second quarter BV orbetween the elution profiles of sulfate and perchlorate. Thiselution sequence (i.e., sulfate f nitrate f perchlorate) isconsistent with previous studies which showed that sulfateis the least strongly sorbed by Type-I polystyrenic anion-exchange resins, whereas perchlorate is the most stronglysorbed (24). This finding may also be partially explained bythe fact that sulfate has an extremely high hydration energy(∆G0 ) -1103 kJ/mol), as noted earlier. The elution profilesare analogous to chromatographic separation of ions basedon their retention time or selectivity in ion-exchangereactions, although the elution profiles of sulfate, nitrate,and perchlorate overlapped. The overlapping of the elutionprofiles could be partly attributable to a relatively highconcentration of the regenerant solution (1 M FeCl3 and 4M HCl) as well as a low sample frequency (every quarter BV)used during the regeneration. In fact, previous laboratorystudies have shown that the sulfate elution profile could becompletely separated from that of perchlorate if the regen-eration was performed at a slower flow rate and the effluentwas sampled more frequently (18). Separation of these anionsin the spent regenerant solution is of great interest becauseremoval of nitrate and sulfate is beneficial prior to thermalreduction or destruction of ClO4

- (described below) in thata smaller amount of reducing agent is needed for the process.

Perchlorate Destruction, Recovery, and Waste Minimi-zation. Because the eluted perchlorate was so concentrated

in the third quarter of the first BV of the regenerant solution(Figure 2), three options are available for subsequentperchlorate destruction, recovery, and/or disposal. First, thissmall volume or the first BV of the regenerant solution cansimply be neutralized and disposed of as hazardous waste.Assuming that the resin bed is able to treat ∼100 000 BVs ofcontaminated water at influent ClO4

- concentrations of 50µg/L or less (12), this translates only about 0.001% ofsecondary wastes produced in comparison with the amountof water treated. Note that although about 2-6 BVs of theregenerant solution are typically used to ensure a completeregeneration, any solution in excess of the first BV of thespent regenerant can be reused in subsequent regenerationcycles; the presence of small quantities of ClO4

- in theregenerant solution has no significant impact on subsequentregeneration efficiencies (12, 18).

Second, perchlorate in the regenerant solution is destroyedby reduction with ferrous chloride at an elevated temperaturein a thermoreactor (14). While perchlorate is degraded, theferrous ion (Fe2+) is oxidized (to Fe3+), and this process renewsthe FeCl3-HCl regenerant solution because Fe3+ ions aredepleted as a result of the sorption of FeCl4

- ions during theregeneration. In other words, this process minimizes the needto dispose of the secondary regenerant waste by allowing itto be recycled. This option was evaluated in this study atCalgon Carbon Corporation. Ferrous chloride was used asa reducing agent in accordance with the following chemicalstoichiometry (14):

However, an excess amount of FeCl2 is usually needed dueto the presence of other oxidizing agents (such as nitrate anddissolved oxygen) in the spent regenerant solutions (18). Theresults (Table 1) indicate that perchlorate in spent regenerantsolution can be effectively destroyed at an operating tem-perature of 190 °C and a flow rate of ∼24-36 gpd (or a

FIGURE 3. Elution profiles of sulfate and nitrate during the regeneration of the spent resin bed. The regenerant solution consisted of 1M FeCl3 and 4 M HCl, and the flow rate was ∼0.6 gpm. The elution profile of ClO4

- (shown in Figure 2) was included for comparisons.

TABLE 1. Perchlorate Destruction Performance

temperature(°C)

flow rate(gph)

inlet ClO4-

(mg/L)outlet ClO4

-

(mg/L)

ClO4- destructionefficiency

(%)

190 1 7290 175 97.6190 1.5 7040 553 92.1

ClO4- + 8Fe2+ + 8H+ f Cl- + 8Fe3+ + 4H2O

6280 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 17, 2007

residence time of 40 min to 1 h). At an influent ClO4-

concentration of 7000-7300 mg/L (diluted due to the addi-tion of FeCl2), the effluent ClO4

- concentration was found torange from 175-550 mg/L, representing a destructionefficiency of approximately 92-98%. These results clearlydemonstrate that perchlorate can be rapidly and effectivelydestroyed in spent regenerant solution. Similar observationshave been made in a small-scale field demonstration at theEdwards Air Force Base in California (18). As noted earlier,although perchlorate was not completely destroyed in thiscase, the presence of residual amounts of ClO4

- should notaffect the regeneration efficiency when the treated solutionis reused for regenerating the resin bed. The residual ClO4

-

concentration was very low in comparison with the peakClO4

- concentration in spent regenerant solution (up to100 000 mg/L).

An operating temperature of about 190 °C was chosen forobtaining optimum perchlorate-destruction efficiency whilepreventing potential precipitation of mixed ferric and ferrous-oxide solids. Detailed temperature-dependent reaction ki-netics between ferrous ions and ClO4

- have been reportedelsewhere (14). The reaction was found to be slow at relativelylow temperatures, despite its favorable thermodynamics,because of the high activation energy (∼120 kJ/mol) requiredfor the degradation of ClO4

- to occur in aqueous solutions.The reaction followed a pseudo-first-order-rate law in thepresence of excess Fe(II), and the rate increased nearly 3orders of magnitude when the temperature was increasedfrom 110 to 195 °C. We also note that nitrate in the regenerantsolution was completely degraded whereas sulfate was notdegraded at all under the same experimental conditions.

In the third option, perchlorate in the regenerant solutioncan be recovered as pure perchlorate salts such as potassiumperchlorate (KClO4) because of relatively low solubility ofthis species in water (Ksp ) 1.05 × 10-2 at 20 °C). At aconcentration of 100 000 mg/L ClO4

-, theoretical calculationssuggest that about 98% of ClO4

- could be precipitated asKClO4 solids, assuming that the final K+ concentration iskept at 0.5 M or higher. This option was tested in laboratorystudies, and we found that about 13.1 g of KClO4 (or ∼9.4 gof ClO4

-) was recovered by mixing 100 mL of the spentregenerant solution with 30 mL of saturated potassiumchloride (KCl) solution. This represents a recovery of 94% ofperchlorate from the spent regenerant solution. This amountof recovery was slightly lower than that predicted bytheoretical calculations, partly because of an increasedvolume (due to the addition of KCl) and a relatively low K+

concentration. This treatment option could be attractivebecause of its simplicity, greatly reduced volume of hazardousmaterial, and elimination of the destruction unit. Additionally,one highly significant result of perchlorate recovery is thatperchlorate recovered during environmental remediationcould be turned from a liability into the reutilization of avaluable material. For example, the recovered KClO4 couldpotentially be reused in pyrotechnics or munitions. Thisrecovery technology also has been demonstrated successfullyin recovering trace quantities of perchlorate in water andsediments for isotopic analysis and environmental forensicsstudies (25, 26).

AcknowledgmentsThe technical and field support operations provided by Y. K.Ku and H. Yan at Oak Ridge National Laboratory (ORNL) andpersonnel at Calgon Carbon Corporation are gratefullyacknowledged. This research was supported in part by Aerojetand the Environmental Security Technology CertificationProgram (ESTCP) of the U.S. Department of Defense. ORNLis managed by UT-Battelle, LLC, under contract DE-AC05-00OR22725 with the U.S. Department of Energy.

Literature Cited(1) Jackson, W. A.; Anandam, S. K.; Anderson, T.; Lehman, T.;

Rainwater, K.; Rajagopalan, S.; Ridley, M.; Tock, R. Perchlorateoccurrence in the Texas southern high plains aquifer system.Ground Water Monit. Rem. 2005, 25, 137-149.

(2) Urbansky, E. T. Perchlorate chemistry: Implications for analysisand remediation. Bioremed. J. 1998, 2, 81-95.

(3) Urbansky, E. T. Issues in managing the risks associated withperchlorate in drinking water. J. Environ. Manage. 1999, 56,79-95.

(4) Damian, P.; Pontius, F. W. From rockets to remediation: theperchlorate problem. Environ. Prot. 1999, 24-31.

(5) Aziz, C.; Borch, R.; Nicholson, P.; Cox, E. Alternative causes ofwide-spread, low concentration perchlorate impacts to ground-water. In Perchlorate Environmental Occurrences, Interactions,and Treatment; Gu, B., Coates, J. D., Eds.; Springer: New York,2006; pp 71-91.

(6) Batista, J. R.; McGarvey, F. X.; Vieira, A. R. The removal ofperchlorate from waters using ion exchange resins. In Perchloratein the Environment; Urbansky, E. T., Ed.; Kluwer/Plenum: NewYork, 2000; pp 135-145.

(7) Gu, B.; Brown, G. M.; Maya, L.; Lance, M. J.; Moyer, B. A.Regeneration of perchlorate (ClO4

-)-loaded anion exchangeresins by novel tetrachloroferrate (FeCl4

-) displacement tech-nique. Environ. Sci. Technol. 2001, 35, 3363-3368.

(8) Tripp, A. R.; Clifford, D. A. The treatability of perchlorate ingroundwater using ion exchange technology. In Perchlorate inthe Environment; Urbansky, E. T., Ed.; Kluwer/Plenum: NewYork, 2000; pp 123-134.

(9) Gu, B.; Brown, G. M. Recent advances in ion-exchange forperchlorate treatment, recovery and destruction. In PerchlorateEnvironmental Occurrences, Interactions, and Treatment; Gu,B., Coates, J. D., Eds.; Springer: New York, 2006; pp 209-251.

(10) Logan, B. E. Evaluation of biological reactors to degradeperchlorate to levels suitable for drinking water. In Perchloratein the Environment; Urbansky, E. T., Ed.; Kluwer/Plenum: NewYork, 2000; pp 189-197.

(11) Hatzinger, P. B.; Diebold, J.; Yates, C. A.; Cramer, R. J. Fielddemonstration of in situ perchlorate bioremediation in ground-water. In Perchlorate Environmental Occurrences, Interactions,and Treatment; Gu, B., Coates, J. D., Eds.; Springer: New York,2006; pp 311-341.

(12) Gu, B.; Brown, G. M.; Alexandratos, S. D.; Ober, R.; Dale, J. A.;Plant, S. Efficient treatment of perchlorate (ClO4

-)-contaminatedgroundwater by bifunctional anion exchange resins. In Per-chlorate in the Environment; Urbansky, E. T., Ed.; Kluwer/Plenum: New York, 2000; Ch. 16, pp 165-176.

(13) Gu, B.; Ku, Y.; Brown, G. M. Treatment of perchlorate-contaminated water using highly-selective, regenerable ion-exchange technology: a pilot-scale demonstration. Fed. Facil.Environ. J. 2003, 14, 75-94.

(14) Gu, B.; Dong, W.; Brown, G. M.; Cole, D. R. Complete degradationof perchlorate in ferric chloride and hydrochloric acid undercontrolled temperature and pressure. Environ. Sci. Technol.2003, 37, 2291-2295.

(15) U.S. Environmental Protection Agency. Record of decision forthe Western groundwater operable unit OU-3. U.S. EPA Region9: San Francisco, CA, 2001; http://www.epa.gov/superfund/new/awards/rods/aerojet.pdf.

(16) Jackson, P.; Laikhtman, M.; Rohrer, J. S. Determination of tracelevel perchlorate in drinking water and ground water by ionchromatography. J. Chromatogr. 1999, 850, 131-135.

(17) Urbansky, E. T.; Gu, B.; Magnuson, M. L.; Brown, G. M.; Kelty,C. A. Survey of bottled waters for perchlorate by electrosprayionization mass spectrometry (ESI-MS) and ion chromatography(IC). J. Sci. Food Agric. 2000, 80, 1798-1804.

(18) Gu, B.; Brown, G. M. Field demonstration using highly selective,regenerable ion exchange and perchlorate destruction tech-nologies for water treatment. In Perchlorate EnvironmentalOccurrences, Interactions, and Treatment; Gu, B., Coates, J. D.,Eds.; Springer: New York, 2006; pp 253-278.

(19) Brown, G. M.; Bonnesen, P. V.; Moyer, B. A.; Gu, B.; Alexandratos,S. D.; Patel, V.; Ober, R. The design of selective resins for theremoval of pertechnetate and perchlorate from groundwater.In Perchlorate in the Environment; Urbansky, E. T., Ed.; Kluwer/Plenum: New York, 2000; Ch. 15, pp 155-164.

(20) Moyer, B. A.; Bonnesen, P. V. Physical factors in anionseparations. In Supramolecular Chemistry of Anions; Bianchi,A., Bowman-James, K., Garcia-Espana, E., Eds.; VCH: New York,1997; Ch. 1.

(21) Diamond, R. M.; Whitney, D. C. In Ion Exchange, Vol. 1; Marinsky,J. A., Ed.; Marcel Dekker: New York, 1966; pp 277-351.

VOL. 41, NO. 17, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 6281

(22) Marcus, Y.; Kertes, A. S. Ion Exchange and Solvent Extractionof Metal Complexes; Wiley-Interscience: New York, 1969.

(23) Bonnesen, P. V.; Brown, G. M.; Bavoux, L. B.; Presley, D. J.;Moyer, B. A.; Alexandratos, S. D.; Patel, V.; Ober, R. Developmentof bifunctional anion exchange resins with improved selectivityand sorptive kinetics for pertechnetate. 1. Batch-equilibriumexperiments. Environ. Sci. Technol. 2000, 34, 3761-3766.

(24) Gu, B.; Ku, Y.; Brown, G. Sorption and desorption of perchlorateand U(VI) by strong-base anion-exchange resins. Environ. Sci.Technol. 2005, 39, 901-907.

(25) Bao, W.; Gu, B. Natural perchlorate has its unique oxygen isotopesignature. Environ. Sci. Technol. 2004, 38, 5073-5077.

(26) Bohlke, J. K.; Sturchio, N. C.; Gu, B.; Horita, J.; Brown, G. M.;Jackson, W. A.; Batista, J. R. Perchlorate isotope forensics. Anal.Chem. 2005, 77, 7838-7842.

Received for review March 20, 2007. Revised manuscriptreceived June 15, 2007. Accepted June 19, 2007.

ES0706910

6282 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 17, 2007