Adsorption Plutonium Varied Oxidation States

7/29/2019 Adsorption Plutonium Varied Oxidation States

1/11

Cmrhrmica n Cosmahtmicn Acta Vol. 49, pp. 2297-230763 Pcqazx~on Pms Ltd.985. Rintcd in &A.

aoi6- 7037/85/53.00 + .a,

The adsorption of plutonium IV and V on goethiteARTHUR L. SANCHEZ'~~,AMESW. MURRAY and THOMAS H. SIBLEY'

I kh001fOceanography, University of Washington, Seattle. WA 98195 School of Fisheries, University of Washington, Seattle, WA 98 1953 Present address: Graduate School of Public Health, University of Pittsburgh, Pittsburgh, PA 1.5261(Received November 28, 1984; accepted n revisedform Jur y 23, 1985)

Abstract-The adsorption of Pu(IV) and Pu(V) on goethite (aFeGGH) from NaNO, solution shows distinctdifferences related to the different hyd~l~c character of these two oxidation states. Under similar solutionconditions, the adsorption edge ofthe more strongly hydrolyzable Pu(IV) occurs in the pH range 3 to 5while that for Pu(VI is at OH 5 o 7. he adsorption edge for Pu(V) shifts with time to lower pH values andthis appears to be due to-the reduction of PI&) to &(IV) in the presence of the goethite &face. Theseresults suggest that redox transformations may be an important aspect of Pu adsorption chemistry and theresulting scavenging of Pu from natural waters.Increasing ionic strength (from 0.1 M to 3 M NaCl or NaNO> and 0.03 M to 0.3 M NaSO,) did notinfluence Pu(IV) or I%(V) adsorption. In the presence of dissolved organic carbon (DOC), Pu(V) reductionto Pu(IV) occursed in solution. Pu(IV) adsorption on goethite decteased by 30% n the presenceof 240 ppmnatural DOC found in Soap Lake, Washington waters. ncreasing concentrations of carbonate ligands decmasedPu(IV) and Pu(V) adsorption on goethite, with an alkalinity of 1000 me@ totally inhibiting adsorption.The Pu-goethite adsorption system provides the data base for developing a thermodynamic model of Puinteraction with an oxide surface and with dissolved ligands, using the MINEQL computer program. Fromthe model calculations we determined equilibrium constants for the adsorption of Pu(IV) hydrolysis species.The model was then applied to Pu adsorption in carbonate media to see how the presence of Cd,- couldinlhtence the mobility of Pu. The decrease in adsorption appears to be due to fo~ation of a Pu-COJcomplex. Model calculations were used to predict what the adsorption curves would look like if Pu-C03complexes formed.

INTRODUCMONADSORPTIONONTO suspended particulate matter hasbeen suggested as the major removal mechanism forplutonium from natural waters. This removal processis thought to a&ct Pu in freshwater, estuarine, andmarine environments, regardless of whether its sourceis bomb fallout Pu, accidentally released Pu, or Puintroduced with low-level radioactive wastes (see e.g.BOWEN et l.,980). Many workers have argued thatthis observation is due to the interaction of the stronglyhydrolyzable Pu(IV) oxidation state with the surfacesof natural particulate matter. Thermodynamic calcu-lations, however, show that Pu(V) is the stable oxi-dation state in the pH range of 5 to 7 at a pE of 12and that Pu(VI) should be the most stable oxidationstate at higher pH values. Pu(IV) is predicted to bestable only below pH 5 (SANCHEZ,1983). Recent anal-yses support these calculations and suggest that thedominant oxidation state of Pu in oxygenated surfacewaters is the oxidized Pu(V) or Pu(V1) state (e.g., NEL-SON and LOVETT, 1978, 1981; NELSON and ORLAN-DINI, 1979; BONDIETTI and TRABALKA,1980). In suchenvironments, Pu removal to the sediments must beexplained by more than simple adsorption of the IVoxidation state.

School of Oceanography, University of Washington Con-tribution No. t&S.

There are few rigorous laboratory studies of the ad-sorption of actinides or of strongly hydrolyzable cationsin general. Some of the early work on Pu focused onits interactions with soils (e.g., PROUT, 1957; RHODES,1957). Distribution coefficient studies (e.g., DUURSMAand PARSI, 1974: GROMOV and SPITSYN, 1974: BON-DIETTI et al., 1976; SANCHEZet al., 1982) used waterand sediments from various locations to determine thepartitioning of Pu between sediment particles and wa-ter. While such studies are useful in developing pre-dictive models for Pu distribution in a particular en-vironment, the adsorbent and the solution phases aremuch too complex to provide a straightforward inter-pretation of the adsorption data. We choose to usesimpler, better-defined experimental systems to obtaindata that couid be readily interpreted using adsorption~uilib~um models. There have been few equivalentstudies. ROZZELL nd ANDELMAN 197 I) characterizedthe sorption behavior of Pu(IV) on various silica sur-faces. Ctay minerals have also been used in some ex-periments (e.g.. BONDIEIT~et al., 1976; DE REGGE dal., 1980; BILLON, 1982).In this paper, the results of experiments to determinethe effects of pH, ionic strength, alkalinity, and dis-solved organic carbon on the adsorption of Pu on goe-thite am presented. Two oxidation states of plutonium,PuflV) and Pu(V), were used in these experiments.The results of environmen~l stud& in Saanich Inietand Soap Lake (SANCHEZet al., in press) wem used asimportant criteria for determining the range of thevariables used in the laboratory experiments.

2297


2/11

298 Z 1.. Sanchez. J. W Murray and ^I li. Sible>METHODS

Ail experiments were carried out at room temperature (20i 2C) in borosilicate gIass vessels. Chemicals used were ACSanalytical reagent grade. Water was distilled. demineralizedand passed through 0.22 Frn Nuclepore filters prior to use inthe exmriments.

Two oxidation states of the isotope Pu-238 (I,,~ = 88 yr).obtained from the Oak Ridge National Laboratory. were pre-pared for the experiments. The method recommended by FOTIand FREILING1964)or the preparation of Pu(IV) and Pu(V1)was used. Briefly, Pu(IV) was prepared by first evaporatingthe tracer solution to dryness with hydroxylamine-HCI to re-duce all Pu to the III state. The residue was then dissolved inI M HN03 with a few crystals of NaNO, added to oxidizePu(III) to the IV state. P&VI) was meoared bv evaooratinethe tracer solution to dryness with concentrated HNO,. anddissolving the residue with 0.05 M HCIOI containing 0.005N K&r207. After adding Pu(VI) to our adsorption media. itwas readily reduced to Pu(V).The techniques of s&vent extraction in TTA-xylene, lan-thanum fluoride coprecipitation, adsorption on silica gel. andCaCOr coprecipitation were used to characterize the Pu oxi-dation states in our tracer spike solutions and in our adsorptionexperiments (SANCHEZ,983). An organic phase containing0.5 M Z-thenoyIt~~uor~~tone (TTA) in xyiene is seIectivefor ionic PufiV) at pH 0.3 or lower. while the Pu(lI1). (Vf or(VI) states are not extracted at this low pH. Lanthanum flu-oride coprecipitates Pu(III), (IV) and possibly (V) but notPu(VI). Pu(V1) adsorbs more strongiy than Pu(V) on silicagel whife freshly precipitated CaC@ adsorbs Pu(V) morestrongly than Pu(VI). Details of these various techniques forcharacterizing the oxidation states are described in FOTI ndFREILING, 964; ONDIE'ITInd REYNOLDS. 1976; AG-NUZ~ON nd LACHAPEUE, 1948: NELSON nd LOVES. 1978,and NELSON nd OR~N~INI, 1979.Under the solution con-ditions of the adsorption experiments the oxidation statesPu(IV) and Pu(V) were confirmed to be present by a com-bination of these tests (SANCHEZ, 983).imilar characteristicsof these oxidation states were obtained by CHOPP~N ndMORSE in press).B. Adsorption

The synthetic goethite (aFeOOH) used in these ex~~rne~tswas prepared using the method of ATKINSON et i.1967).The surface characteristics of (YFeOOH prepared with thisprocedure have been studied extensively (ATKINSONet al..1967; BALISTRIERI, 977; BALISTRIERInd MURRAY, 1981.1982).The adsorption experiments included the following steps:(I) preparation of the solid suspension in a borosilicate glassvessel; (2) addition of the Pu tracer: (3) immediate adjustmentof pH with either 0.1 N HCI or 0. I N NaOH; (4) equilibrationof the spiked suspensions after adjusting the total volume to20.0 ml; (5) measurement of pH prior to samphng: (6) deter-mination of Pu partitioning between dissolved (co.45 pm)and adsorbed phases by filtering a subsample from each vesseland counting both dissolved and particulate fractions for Puactivity; and (7) oxidation state measurements For selectedsamples. Variations of this general procedure were used totest the effects of pH, ianic strength, carbonate alkalinity, anddissolved organic carbon on the adsorption behavior of bothPu oxidation states on aFeOOH. A time series of subsampleswere collectedrom each experimenti vessel until equilibriumdis~bution of Pu between the sofuble phase and the solidphase was attained. The goethite concentration was 28.5 m*i-.A blank electrolyte solution identical to the experimental me-dium but containing no cYFeDOH was run in parallel with

the adsorption soluuons to correct for adsorpuon on the vessels. No attempts were made to shield these experiments fromlight. It appears that the rates of adsorption may be differentunder light and dark conditions (l&?lEY-RENNlrr:. andMORSE. n press)1 +!$f%r~/prf on PU udsorpr~on. Each point cn the ad-sorption versr~ pH plots represents one experimental vessel.The background electrolyte was 0.1 M NaNO1. .~dsorpttonexperiments were run using both Puf IV) and Pu(\) initialspike solutions.In a separate set of expenments w&r Pu(V), samples werecollected for solvent extraction into TTA-xylem to determineif Pu(Vt had been reduced to the TTA-extractahie PufIVlstate (SANCHEZ, 1983). Two series of experiments for eachoxidation state (IV and V) were performed using 1 10 IM and I X IO- M r3sPu.2. @@cly/-tonic strength on Prc adsorption. Three different

electrolytes at the following concent~tions were used m theseexperiments: 0.1 M, 0.5 M. 1 O M and 3.0 M NaNOj: 0.5 Mand 3.0 M NaCI: and 0.03 M. 0.15 M and 0.30 M Na2S04.The range in ionic strengths for NaCl and NaNO, and thesutfate concentrations were chosen to approximate values ob-served for Soap Lake surface and monimolimnion waters.Soap Lake is an alkaline, meromictic lake in Eastern Wash-ington state, where we have measured relatively high concen-trations of Pu compared to other natural waters (SANCHEZ,1983). All experiments were at pH 7.0 + 0.2,2&S m* nFeOOHI-, and Pu(lV) and (V) coneent~tions of I X IOUr M.3. Eficcr qfcarbonafe aikaliniry on Pu adsorprrot7. Appro-priate volumes from a 1.0 M NaHCOs stock solution wereadded to distilled, deionized water to prepare adsorption mediawith the following total alkalinities: 10, 30, 100. 200, 400,550.700. and loo0 meq/L. No attempt was made to rn~n~ina constant ionic strength for these adsorption media and thesolutions were open to the atmosphere. The pH 8.6 ? 0. I andtotal Pu concentration was IO- M L3SP~(IVor V). Blankswere run along with the experiments to determine if PulVlwas reduced to Pu(IV) in carbonate media and to correct foradsorption on the vessels. Filtered (0.22 am) water from I?m and 24 m in Soap Lake were also used to determine Puadsorption under natural Soap Lake conditions. The anoxicsample from 24 m was left to equilibrate with air until all H2Swas oxidized prior to use in the experiments. It must be poiniedout that these samples from Soap Lake may also contain po-tential Pu ligands. such as dissolved organic carbon. in additionto the carbonate species.4. .!$&I $dissolvrd organic curbon (LXX_). Two sets oiexperiments were performed using the naturaiIy-~ccu~ngDOC concentrations in Soap Lake surface and monimolim-nion waters. Water from 12 m and 24 m were acidified bydropwise addition of concentrated HCI until all the CO> wasremoved (ptt c; 4). DOC measurements for these acidifiedsamples were done on a Dohrman DC-80 Total Organic Car-bon Analyzer system.In the lirst experiment. J8Pu(V) tracer was added to a setofsamples from 12 mat a concentration of I X IO- M, andthe pH readjusted to pH 4 and 6 using 0.1 N NaOH solution.Filtered Soap Lake water (not degassed) from i 2 m was alsoused as an experimental medium, with its pH constant at 9.8.An additional experimental vessel containing humic acid ex-tract from Washington continental shelf sediments (DCK= LOO pm, pH 9.6) was also included in this series. One mlsubsamples were analysed to determine if Pu(V) was reducedto Pu(IV) in these media using solvent extraction in TTA/xylene (SANCHEZ.1983).A second set of experiments was conducted to determmcthe adsorption of *38P~(IV) I X IO- M) at various concen-trations of DDC. Varying proportions of the acidified (pH3.8) surface and bottom samples from Soap Lake were com-bined to obtain a range of DOC values. The pH was thenadjusted to pH 7. I + 0. I. The pH was left at 7 because buffersfor other pt-1 values could complex Pu and confuw our results.


3/11

Pu adsorption on goethite 2299C. Radioactivity measurements

Liquid scintillation counting was used to measure the %Iactivity in ail samples. Ten ml of a commercially preparedcounting cocktail, 3a70B, (Research Products InternationalCorp., Elk Grove Village, Illinois) were added to 4 ml ofaqueous sample adjusted to an acidity of 0.5 N HCI. For or-ganic samples from ITA-xylene extraction. IOml scintillationmixture were added directly without acidification. NOquenching was observed in the acid phase or the TTA-xyieneorganic phase, as checked with internal 23*Pu tandards pre-pared in either 0.5 N HC1or ITA-xylene extractant. An in-ternal 23pPu tandard prepared in acid solution containinggoethite had showed similar counts as a standard in acid so-lution alone. A Packard TriCarb Model 3375 Liquid Scintii-lation Spectrometer was used for counting samples.RESULTS AND DISCUSlON

I. Kinetics of Pu(i V) and Pu(V) adsorptionTwo observations distinguish Pu(IV) from Pu(V)adsorption on cuFeOOH: (1) rapid kinetics for Pu(IV)

adsorption compared to Pu(V), and (2) a significantdifference in their initial adsorption edges under similarsolution conditions. Equilibrium adsorption for Pu(IV)on crFeOOH is attained within one hour (Fig. la) andthe adsorption edge occurs from pH 3 to 5. Undersimilar experimental conditions, Pu(V) does not attainequilibrium distribution even after 20 days, and theadsorption edge gradually shifts from pH 7 to lowerpH values (Figs. 1b and lc).

The observed shift in the adsorption edge for Pu(V)and its slow adsorption kinetics appear to be inconsis-tent with thermodynamic calculations which predictthat Pu(V) is stable under the conditions of these ex-periments (SANCHEZ,1983). The gradual increase inPu(V) adsorption with time, caused its adsorption edgeto approach that for Pu(IV), suggesting that Pu(V) maybe reduced to the lower oxidation state. This was testedexperimentally by investigating the solvent extractionbehavior into TTA/xylene of the adsorbed and solublePu fractions. Any significant removal of Pu into theorganic phase under the low pH extraction conditionsindicates the presence of the Pu(IV) oxidation state.

The amount of TTA-extractable Pu as a functionof time in the adsorbed and the soluble fractions ofgoethite experiments spiked initially with Pu(V) isshown in Fig 2. The initiai time points are for 1 hourafter adsorption. The data show that soluble Pu re-mained unextmcted (solution.^ .*._at IO- M. c) Adsorption of I%(V) on goethite as a functionof pH from 0.10 M NaNO, solution at lo-i0 M.


4/11

300 fl I Sanchez. J. W. Murray and T. I-1.Sthlei100 - pH 3.8

A- -Apw 4. 0I D. . . cl pH 5. 3

; riSOi 0-Q pH 6. aQ- 4 pl i 7. 5

Adsorned

d0 80 120 160 200TI ME , HR

FIG. 2. The percent oftotal Pu(V) added that becameTTA-extractable Pu(IVf as a function of time in the adsorbed andthe dissolved fractions. Results from goethite experiments ini-tially spiked with Pu(V). Initial data points are for 1hour afteradsorption started. Error for TTA extraction is -+5%.

sequently rapidly adsorbed. The data are not suiKcientto distinguish which of these two mechanisms is oc-curring. However, recent observations of KEENEY-KENNICUTT and MORSE 1985) indicate that Pu(V)adsorbs as this oxidation state on a variety of mineralsurfaces. Their results would support the tirst mecha-nism to account for our observations, particularly sincewe found that no reduction occurs when PufV) is notadsorbed.Three possibilities may be suggested to account forthe observed reduction of Pu(V) after adsorption: ( 1)thermodynamic data are wrong and Pu(IV) is the stableform under the pH conditions of our experiments (pH5 to 7). (2) trae amounts of reducing agents are presenton the goethite surface; or (3) the reduction of Pu(V)occurs via a radiolysis or disproportionation reaction.if PufIV) is really the stable oxidation state the impli-cation is that Pu(V) is reduced slowly by water whenit is immobilized by adsorption. Pu(V) in homogeneoussolutions is kinetically stable and is not reduced forindefinite time periods. It was not possible to determinethe presence of trace amounts of reducing agents inthe experiments. Given the small concentrations of Puused, however, it is conceivable that any trace constit-uent occurring as an impurity in the media may po-tentially act as a reducing agent. There is the moreunlikely possibility that as soon as the radioactivitybecomes concentrated on the solid surface. the pro-duction of reducing agents such as Hz&. OH radicalsand nitrite (MINER nd SEED,1967) vicehe radiolysismechanism becomes important. The concentrationsof these species would probably be too small to be de-tected because of the minimal radioactivity in the ex-periments. The final edibility is the reduction ofPu(V) via the disproportion&on reaction. whereby

Pu(lV) and PufVtt species are produced ((. oNw-h,1949; RABIDEAI J . 1957). KEENEY-KENNICI: I-T andMORSE (in press) invoked this mechanism to accountfor the reduction of PufVI to Pu(IV) in their goethiteadsorption experiments.. &l!& t$ /Ffl 011PrrCl 1 rxf hi(V) u~~si)r/?tt~)~~

The major difference between the adsorption ofPu(IV) and Pu(V) is in the position of their initial pHadsorption edges under similar solution conditions (I= 0.1 M NaN03). In Fig. la. the adsorption edge forPu(IV) occurs at the pH range 3 to 4. in contrast tothat for Pu(V) (Figs. 1b and Ic), which has an mitiaiadsorption edge between pH 6 to 7. The pH of theadsorption edge for Pu(V) is a lower limit because ofthe uncertain role of Pu(V) reduction to Pu(lV). Theone hour data are probably a close approximation.

The difference in the adsorption edge is possihljlinked to the tendency of these ions to hydrolyze insolution. BALISTRIERI f uf. f I98 1) found a linear cor-relation between the apparent adsorption equilibriumconstant for metal adsorption on goethite and the hy-drolysis constants of that specific metal ion. The largerthe hydrolysis constant. the stronger the adsorptionand the lower the pH of the adsorption edge. Pu(lVf(in the ionic form Pu4+) is the most strongly hydrolyzedand Pu(V) (as PuOt ion) the least hydrolyzed amongthe common oxidati~)n states of Pu in aqueous solu-tions. The observed difference in their initial adsorptionedges is therefore consistent with their different hydro-lytic character.

Another observation is the shift in the adso~tionedges to slightly higher pH range with increased Puconcentration for each oxidation state (Figs. I a, b. andc). This is similar to the results of BALISTRIERIandMURRAY (1982) for Cu. Pb, Zn, and Cd adso~tionon goethite as a function of the metal concentrationand may be due to strong binding site limitatmn. Thissuggests that adsorption of Pu rather than tts precipi-tation on the goethite surface is the controilin~ mech-anism.We proceeded to model the adsorption vcrw pHdata. We first determined adsorption equilibrium con-stants to enable us to predict Pu adsorption in digerentenvironments. The same constants are used later toestimate the values of dissolved Pu-carbonate com-plexes.

In recent years, several electrostatic models havebeen developed for describing ion adsorption at theoxide/water interface (see WESTALL and HOI% 1980.for a review of these various models). Our adsorptionexperiments on goethite were designed so that the re-sults could be interpreted using the site-binding or triplelayer model. This model has already been successfullyapplied by BALJ STRIERI nd MURRAY (198 I. I %2) forthe adsorption of major ions and trace metals on goc-thite and this was the main reason it was used rn thisstudy. Other models exist that would also tit the data


5/11

Pu adsorption on goethite 2301

equally well. This model was applied to our Pu ad-sorption data using the MINEQL computer program(WE~TALLet al., 1976).

The details of the site-binding model, and the theo-retical and experimental work that went into derivingthe various parameters used to describe the goethitesurface, will not be discussed here. The reader is re-ferred to YATES et al. f 1974). DAVIS et al. f 1978) WE-STALL and HOHL (1980), and BALISTRIERI nd MUR-RAY ( 98 1) for details. Briefly, the site-binding modelassumes a structure of the electrical double layer whichcan be described mathematically by a set of adjustablephysical parameters. BALIS-~RIERInd MURRAY(198 1)have presented the experimental data used for defininggoethite interactions in simple electrolyte systems.Their model parameters for the surface area, the totalnumber of surface sites, the inner and outer layer ca-pacitances, the surface hydrolysis constants of goethiteand the surface complexation constants for Na+ andNO; are used here.Modelling the Pu adsorption data as a function ofpH involves defining the intrinsic surface complexationconstants for Pu interaction with the goethite surface.Four reactions involving the adsorption of Pu(IV) hy-drolysis species were required to describe the adsorptionof Pu(IV).

*Ku&&~SOH + Pu*+ + 2H20-X SO-Pu(OH):+ + 3H+SOH + Pu4+ + 3Hz0 l l&ih+-SO-Pu(OH); + 4H+SOH + Pu4+ + 4Hz0 SSO-Pu(OH)$ + 5H+

In these equations, SOH represents a surface site, andthe *PM is the intrinsic constant describing Pu as-sociation with-goethite. The values of the log +KrNTthat were required to fit the adsorption data were 2.50,-2.00 -5.90 and -12.0 for PUN+, Pu(OH):*,Pu(OH); and Pu(OH)! respectively. The incrementaladdition of these constants to produce a model best fitis shown in Fig. 3. In this way the effect of the additionof each reaction on the final model curve can be seen.The values are a unique set and are accurate to at least0.5 log units. The hydrolysis constants for Pu(IV) re-ported by METIVIER and GUIL~UMONT ( 1976) wereused in these calculations. (BAESand MESMER, 1976:CHOPPIN, 1983). Although these constants were ob-tained at slightly higher ionic strengths (I = 1.0 M),no corrections were made for ionic strength in ourmodel calculations. It should be clearly stated that thereare still some uncertainties in the hydrolysis constantsfor Pu(IV) and the values for *tiNT depend on thehydrolysis constants used in the calculations. Theseuncertainties are not large enough to invalidate ourgeneral conclusions.

The data for Pu(V) were not interpreted using MI-NEQI modelling because an equilib~um dist~butionwas not attained. It was uncertain whether Pu(V) re-mained stable in the experiments. In addition, the hy-

I d \ I

2 4 6 a 10PHFIG. 3. Adsorption of Pu as a function of pH on goethite.Four reactions for the adsorption of Pu(IV) hydrolysis speciesare required to model the data (circles). The innmental ad-dition of these reactions to the equilibrium program MINEQLto produce a model best fit is shown.

drolysis constants for Pu(V) are even less well knownthan those for Pu(IV).

Changes in ionic strength do not appear to exert amajor control on the adsorption of either Pu(IV) orPu(V). The adsorption of 238Pu(IV) and 238Pu(V) oncrFeOOH did not vary with increasing ionic strength(up to I = 3) at pH 7.0 f 0.2, regardless of the back-ground electrolyte. As summarized in Table 1, greaterthan 90% of the total activity for both oxidation statesis found on the goethite phase after ~uilib~um dis-tribution is attained. Sulfate ion did not influence ad-sorption of either Pu(IV) or Pu(V) up to 0.3 M SOi-although it is one of the known complexing agents forPu (CLEVELAND,1979). Enhancement of Pu adsorp-tion due to SOi- as seen for Cu and Zn adsorption ongoethite (BALISTRIERI nd MURRAY, 1982) could notbe identified in these experiments because we were inthe greater than 90% adsorption range.4. Eflecf of carbonafe alkalinify

The adsorption of 238Pu(IV) and 238Pu(V) onaFeQQH as a function of increasing alkalinity is plot-ted in Figs. 4a and 4b. The experiments were open tothe atmosphere. The pH values for the individual ves-sels were constant at about pH 8.60 + 0.10 during thecourse of the experiments. The lower curve in eachplot represents adsorption data after 1 hour of equili-bration; the upper curve is for the true equilibriumadsorption points (96 hours for Pu(IV) and -300 hoursfor PufV)). Adsorption of both oxidation states is un-affected by alkalinity values less than about 100 meq/


6/11

2302 A ! Sanchez. J. u. Murray and r. H. %hle\

0.30 n tla2Sog IV 95t3 93s i

L, where the amount adsorbed on &eOOH is similarto that obtained at the same pH in 0. I M NaNOX so-lution. The effect of alkalinity, however, becomes pro-gressively more important for alkalinity values greaterthan 100 meq/L, and adsorption is totally inhibited ata concentration of 1000 meq/L. The adsorption fromnatural filtered Soap Lake water at alkalinities of f44meq/L (pH = 9.8) and - 1200 meq/L (pH = 9.5) areshown in the same plots. At equilibrium, the amountsof Pu IV and V adsorbed onto goethite in these SoapLake media agree with the expected removal based onalkalinity alone. However, it must be pointed out thatthe filtered Soap Lake water contains potential Pu li-gands other than the carbonate species, such as dis-solved organic carbon, that may also influence Pu ad-sorption behavior. The effects of organic carbon willbe discussed later.

At equilibrium, the adsorption versusalkalinity plotsfor both Pu(fV) and Pu(V) overlap (Figs. 4a and 4b).This suggests that either the adsorption behavior ofPu(lV) and Pu(V) under high carbonate concentrationsare similar or that there is a redox transformation ofone oxidation state to another. To test this hypothesis,solutions of both Pu(IV) and (V) in high carbonatemedia in the absence of solid were taken for solventextraction with TTA-xylene.

The PuffV) tracer in 1.0 M NaHC03, at pH 8.6 inthe absence of fuFeOOH, remained in the soluble(x0.45 pm) fraction but only 50% of the total activityin solution could be extracted into TTA after 300 hours{Table 2). This observation suggests that althoughPu(IV) remained filterable, only 50% was in the ionicform which can be extracted into TTA. In a less con-centrated carbonate medium (0.03 M NaHC03, pH8.6), increasing amounts of non-extractable Pu (co.45pm), presumably polymeric species of Pu(IV). were

found with time. This suggests that hydroiysls IS ~nt-portant in this medium and that carbonate complexmgcan not compete effectively wtth the formation of h\-drolytic Pu(iW species. BONDIETTI cv ui. i IQ?+) ob-served that polymeric Pu was tbrmed and absorbed tothe walls of the reaction vessel in 0. I M NaH( 0, SO-lution at pH 8.0. At the Pu concentration ( -- 111 M)which they used in their experiments. they concludedthat carbonate ions are unable to stabilize ;b Pu(fVfmonomeric complex.

Similar tests with the Pu( V1oxidation state showedthat it remained soluble and did not extract significantlyinto TTA in either the I .O M NaHCO? or the 0.03 MNaHCOl blank solutions (no goethite present) lTablr2). Pu(V) appeared to be stable in this oxidatron stateat these solution conditions. In the presence otaFeOOH. the amount of TTA-extractable Pu in thesoluble fraction increased with time. In one experiment( I .O M NaHC03, pH 8.6). to which Pu(V) was Initiallyadded. only 2 to 4% of the total Pu adsorbed and 939of the soluble fraction was extracted into T7.A after300 hours (Table 2). These observations indicate thatin the presence of the adsorptive surface, PutV) is re-duced to Pu(lV). This reduction was also observed inthe pH experiments but in that case. Pu(IV) remainedon the solid. Here, it appears that the complexation ofPu(IV) by carbonate ligands removes the adsorbed Pu.This provides further evidence that Pu(V) reductionoccurs in the presence of goethite. Thus. the similarityin the sorption plots for Pu(IV) and Pu(VI ~Y~.UI.Sn-creasing alkalinity is due to the reduction of Pu(V) toPu(IV) in the presence of the oxide surface followedby complexation with carbonate ions.

In the carbonate experiments. three possible factorsare likely to be influencing the Pu distribution: ( I) ad-sorption onto the goethite surface. (2) hydrolysis ofPu. and (3) carbonate complexing. We have attempteda prelimina~ inte~re~tion of our data using the site-binding model including all three interactions. in thepreceding section, an adsorption model that sufficientlydescribed the adsorption behavior of Pu(IV) as a func-tion of pH was developed. An extension ofthis modelto include carbonate complexing was then earned out.We will show here how the formation of such com-plexes could possibly influence the adsorption ofPu(IV). This model can be refmed as the thcrmody-namit data is improved.

The formation of dissolved Pu(lV)-CO3 complexesis poorly understood (CLEVELAND. 1979: KIM ~1 al,19831. MOSKV~N nd GELMAN (1958) attempted todetermine the formation constants for PutIV)-CO3complexes by measuring the solubility of Pu(OHh (am)at pH 1 I .5 in concentrated K&03 solutions. They as-sumed that the complex PuCO: was formed by thereaction:

pu4+ + co:- = puce:+and estimated the formation constant to vary from1O34o 1042.SCHWAB and FELMY (19831 reviewed this


7/11

PU adsorption on goethite 2303

oo-fz::= 0-E P (IY) * 10-111z-l% ptl a.6to.l

.*-

96 HR AM-

,b ,1,,1Il LlTOTAL ALKALINITY , neqJr

So4p Like (ZCm)f1200 mcqJ4COO

_ pH B.60.1

20- 1. HRO168 HR A

1 I I I1111TOTAL ALKALIN1lT , aeq/t

Soip Lakt (24~)

FIG.4. a) The effect of carbonate alkalinity on the adsorption of Pu(IV) on goethite. b) The effect ofcarbonate alkaiinity on the adsorption of I%(V) on goethite. Adsorption from natural Soap Lake water isalso shown. (Solid symbols).data set and proposed that the main species is probablya mixed hydroxy~ar~nate complex formed by thereaction:

Pu4+ + 4HrO + CO:- = Pu(OH)~CO:- + 4H+with log K = -4.1 + 1.0. From a coordination chem-istry point of view the species Pu(OH)&O:- seemsunlikely. A third possibility that has not been exploredis that a CO:- Iigand could replace one OH- to formPu(OH)&O; . The formation of the species PuCO: ,Pu(OH)&O$- and Pu(OH)rCO; were added to themodel and tested using MINEQL. In addition we as-sumed that the carbonate ions interact with the surfaceof goethite as described by BALISTRIERInd MURRAY( 1982):

SOH + CO:- + H+= SOH;CO:- + Hz0 log Kg; = 15.90

SOH + CO:- + 2H= SOHfHCO; + Hz0 log @$,, = 22.30.

We also assumed that the intrinsic adsorption constantsderived previously for Pu(IV)-goethite interaction arestill valid.

In our first calculations we assumed that aPu(IV)-CO3 complex forms and that these species donot adsorb. The formation of either PuCOf+ orPu(OH),C@,- with the formation constants citedabove would result in no adsorption under the con-ditions of our experiments (curve A). This model isclearly inconsistent with our experimental data (curve


8/11

PUIIV) 1.0 H. PH 8.6 TTIAeYtFaetlo*

PU( IV) 0.03 M, PH 8.6

93; i n TTA aft er 300hr.

D). We then tested the possibility that the formationconstants are too Iarge. We varied the formation con-stants until the % adsorbed was about 50% at analkalinity of 300 meq I-. Then we varied the alkalinityto generate the curve. Because the formation of eachof the three species considered involves only oneCO%-, the effect of all three species is the same (curveB). The log formation constants for PuCO:,Pu(OH),CO~- and Pu(OH)&O; would have to be+27.2, -7.1, and + 1.40 respectively. The formationconstants for P&O:+ and Pu(OH),CO? would beweaker than those previously estimated. Regardless ofwhether these are reasonable constants the formationof these species would not result in a steep enoughdependency of % adsorbed on alkalinity. Adding com-plexes with two or more carbonate ligands (e.g.,Pu(CO& would make the curve steeper.We also tested what the distribution would look likeif the dissolved Pu(OH),COj- species could adsorb.!SCHINDLER 98 I ) reviewed some of the experimentalevidence for the formation of ternary surface com-plexes (i.e., complexes that participate in a surfacereaction) and suggested that these complexes shouldnot be ignored in model calculations which considerthe effects of dissolved hgands. We assumed that theformation constant for Pu(OH),CO:- is accurate andthat this species adsorbs on the surface of goethite as-cording to:Pu4+ + SOH f CO:- + 4H20

= SOH;Pu(OH)X@- + 3H.The curve C in Fig. 5 corresponds to this example withan intrinsic adsorption constant of log K!$HJQ-= 15.6. The alkalinity dependency of this curve is amuch closer representation of the experimental data.Though there are some uncertainties in the adsorp-tion models one point is clear and that is that increasing

Murray and I. H. Sthle!

n I- 38Pan a ICI M20. 28.5 m2 [tFe OOH/i r \\\, -pti 8.6 +o.r /x >,,oi__-_-__*_____,_T-crl _,J___n\,__,__,_--___.$-.10 400 1000

TOTAL ALKALINITY, msqllFIG. 5. Adsorption equilibrium model fits to adsorptionversus alkalinity data (curve D) of Pu(IV) on goethite. Thecurves represent the following conditions. a) Formation con-stant for Pu(OI-I),COj- is tog K = -4.1 or for PuCO: is logK = 34. b) Formation constant for Pu(OH),CO? is log K

= -7.2 or for Pu(OH)&O; is log K = t 1.40 or for PuCO:is log K = +27.2. c) Formation constant for Pu(OH)XO:- islog K = -4.1 and the adsorption equilibrium constant forPu(OH),C@- is log K = 15.6.alkalinity above 100 meq I- can decrease the adsorp-tion of Pu(IV) and Pu(V). This appears to be due tothe formation of Pu-CO3 complexes. Model profilessuggest that the best ftt to the data will be when Pu-CO3 compiexes form and then form a ternary complexwith the goethite surface.5. Effect of dissolved organic carbon(DOC) on Pu adsorption

The naturally-occurring LXX in Soap Lake waterreduced Pu(V) to the IV oxidation state, as indicated

FIG. . he kinetics of Pu(V) reduction to TTA extractablePu(IV) in Soap Lake water samples containing the naturally-occurring dissolved organic carbon fraction in the lake water.Humic acid extract is used as a reference organic material.


9/11

Pu adsorption on goethite 2305by its increased extraction into TTA/xylene with time(Fig. 6). The rate of reduction increased with increasingpH, and was fastest in the nondegassed Soap Lakesample at pH 9.8. A humic acid extract from Wash-ington continental shelf sediments was also found toreduce Pu(V) (Fig. 6). These results are similar to thoseobtained by NASH et al. (1981) for Pu(VI) reductionby humic acids in NaHC03 solution. These taboratoryobservations agree with the suggestion of BONDIETTIet al. (1976) that Pu(V) should be unstable to reductionto Pu(IV) in the presence of organic substances. Fur-thermore, the results suggest that Pu in Soap Lake sur-face waters should be in the IV oxidation state, inagreement with the observation of WAHLGRENet al.(1978) that Pu(IV) is the dominant oxidation state ineutrophic waters.Since DOC was found to reduce Pu(V), the effectof DOC on Pu adsorption was studied using only theIV oxidation state. Two observations were noted inthese experiments: (1) a much slower approach toequilibrium adsorption compared to the pH experi-ments, and (2) a slight decrease in adsorption with in-creasing DOC concentration (Fig. 7). These are similarto results of NELZX?N t al. ( 1980, 198 1), who foundthat the distribution coefficient for *Pu between sed-iments and water decreased as the DOC concentrationincreased.A simple mathematical model similar to the one wedeveloped for the Pu-goethite-carbonate system wasalso applied to the adsorption data. The results of ourcalculations are published elsewhere (SIBLEY et al..1984). Essentially, we found that a model assumingtwo different binding sites on DOC as suggested byALBERTSet al. (1980) and neglecting DOC-goethiteinteractions reasonably fit the adsorption data.

One major difference between these adsorption dataand those obtained at increasing carbonate ion con-centrations is that even at the high DOC concentrations

maa0 PU(lV) * lo-1ZS.Sr2.F.00H/1

pn 7.1to.110

FIG. 7. Theeffect of dissolved Organic carbon from SoapLake waters on the adsorption of Pu(IV) on goethite. Thealkalinity was zero in these experiments.

for Soap Lake monimolimnion waters (DOC = 220ppm), a substantial fraction (70%) of Pu(IV) still ad-sorbs on goethite. In this same water with greater than1 M total carbonate con~ntmtion, no adsorption ongoethite was obtained. These results suggest that car-bonate anions inhibit Pu(IV) adsorption on goethitemore strongly than DOC at the concentration rangestested for these ligands. SIMPSON t al. ( 1980) reacheda similar conclusion for Mono Lake.

CONCLUSIONS1. Under similar solution conditions, the adsorption

edge for Pu(IV) occurs at a significantly lower pH range(pH 3 to 5) than for Pu(V) (initially at pH 5 to 7). Thisresult is consistent with the different hydrolytic char-acter of these two oxidation states in aqueous solutions.

2. For Pu(V), a gradual shift in the adsorption edgeto lower pH ranges occurs with increasing time of ad-sorption. Solvent extraction tests with TTA/xylene in-dicate that the adsorbed Pu(V) has been reduced tothe IV oxidation state. Although thermodynamic cal-culations predict that Pu(V) is stable under the solutionconditions of the experiments, it appears that it is un-stable to reduction to Pu(IV) in the presence of anadsorptive surface.

3. The triple layer model for the adsorption ofPu(IV) on goethite in NaN03 predicts that four hy-drolytic species of Pu(IV), Pu(OH)+, Pu(OH)$+,Pu(OH)f and Pu(OH)!: adsorb on the goethite surface.The log intrinsic constants for the interaction of thesespecies with the oxide surface that reasonably fit theadsorption data are 2.50, -2.0, -5.9 and -12.0, re-spectively. Since Pu(V) was unstable in the presenceof goethite, the data were not used in the triple layermodel calculations.

4. An increase in ionic strength from 0.1 M to 3.0M NaN03 or NaCl did not affect the adsorption of Puon goethite. Sulfate (0.03 M to 0.3 M) also did notdecrease Pu removal on goethite although it is one ofthe known inorganic complexers for Pu.

5. Pu(V) was reduced to Pu(IV) in solution in thepresence of natural dissolved organic carbon (DOC)from Soap Lake. At 240 ppm QOC con~ntmtion,Pu(IV) adsorption on goethite was reduced by 30%.

6. Increasing concentrations of carbonate ions de-creased Pu(IV) and Pu(V) adsorption on goethite. Ata concent~tion of 1 M NaHCOs (pH 8.6) Pu adsorp-tion was inhibited completely, The decrease in ad-sorption appears due to formation of a Pu-CO3 com-plex. however, the value of the formation constant ispoorly known and it is also uncertain whether or notthe complex adsorbs.Acknowledgements-This manuscript benefited greatly fromdiscussions and comments from P. Santschi, R. Anderson, L.Fiaiistrieri, B. ~oneyman, R. McDuff and L. Miller. Dr.W. R. Schell helped with the initial aspects of the senior au-thors study. L. Miller and J. 1. Murray helped collect the


10/11

7306 -\ i >anchez. J. H. Murray and I. H. S~hle\Soap Lake samples. T. Rasp typed the manuscript. This re-search was funded by NSF OCE 82-14072Editorial handling: E. R. Sholkovitz

REFERENCESALBERTS . J., LUTKENHOFFD., GEIGERR. A. and GANT D.(1980) Stabilitv constants determined for comnlexes of 233U.2'7P~.'237Npnd TC with natural stream organic materialIn Savannah River Ecology Lahoratory, iJniver,sit,v o/Georgia. Annual Report 1980ATKINSONR. J.. POSNERA. M. and QUIRK J. P. (1967) Ad-sorption of potential-determining ions at the ferric oxide-aqueous electrolyte interface. J fhy.s Chem. 71, 550-558.BAESS. F. and MESMERR. E. (1976) The Hvdrotvsts ofCat-ions. Wiley-Interscience. 489 pp.BALISTRIERI . S. ( 1977) The basic surface characteristtcs ofgoethite (aFeOOH). MS. thesis, University of Washington,Seattle.BALISTRIERI . S. and MURRAY J. W. (1981) The surfacechemistry of goethite (aFeOOH) in major ion seawater.Amer. J. Sri. 281, 788-806.BALISTRIERI . S. and MURRAY . W. (1982) The adsorptionofCu. Pb, Zn, and Cd on goethite from major ion seawater.Geochim. Cosmochim. Acta 46, I253- 1265.BALISTRIERI . S., BREWER . G. and MURRAY . W. (1981).Scavenging residence times of trace metals and surfacechemistry of sinking particles in the deep ocean. Deep-SeaRes. 28A, 101-121.BILLONA. (1982) Fixation delements transuraniens a differ-ents degr&s doxydation sur les argiles. In Proc. Symp. onMigration in the Terrestrial Environment of Long-lived Ra-dionuclides from the Nuclear Fuel Cycle. Knoxville, TN.27-31 July 1981. IAEA, Vienna.

B~NDIETTI E. A. and REYNOLDS . A. (1976) Field and lab-oratory observations on plutonium oxidation states. InProceedings, Actinide-Sediment Reactions Working Meet-ing, Seattle, IO-1 I Feb. 1976. (ed. L. L. AM=). BNWL-2117.B~NDIETTI E. A. and TRABALKA . R. (1980) Evidence forplutonium (V) in an alkaline, freshwater pond. Radiochem.Radioanal. Letters 42, I69- 176.BONDIETTI E. A., REYNOLDS . A. and SHANKSM. H. (1976)Interactions of plutonium with complexing substances insoils and natural waters. In Transuranium nuclides in the

environment, Proc. Symp.. San Francisco, 17-2 I November1975. IAEA, Vienna.B~WEN V. T., NOSHKINV. E.. LIVINGSTON . D. and VOL-

CHOK H. L. (1980) Fallout radionuclides in the PacificOcean: vertical and horizontal distributions, largely fromGEOSECS stations. Earth Plan. Ser. Lert. 49, 4 I l-434.CHOPPING. R. (1983) Aspects of plutonium solution chem-istry. In Plutonium Chemistry (eds. W. T. CARNALLandG. R. CHOPPIN), p. 2 13-230. ACS Symposium Series 2 16.ACS.CHOPPIN G. R. and MORSE . W. (in press) Laboratory studiesof plutonium in marine systems. In S.vmposium on Envi-ronmental Research for Actrnide ElementsCLEVELAND. M. (1979) Critical review of plutonium equi-

libria of environmental concern. In Chemical Modeling tnAqueous Systems (ed. E. A. JENNE). Amer. Chem. Sot..Washington, DC.CONNICK R. E. (1949) The mechanism of disproportionationof plutonium (V). In The Transuranium Elements (eds.G. T. SEABORG. . J. KATZ and W. MANNING).McGraw-Hill, New York.DAVIS J. A., JAMB R. 0. and LECKIEJ. 0. (1978) Surfaceionization and complexation at the oxide/water interface.1. Computation of electrical double layer properties in sim-ple electrolytes. J. CON.Interf: Sci. 63,480-499.

DE REC~GE.. BONNEA.. Hti\ s I>. and HEREMANSK. t i/)tiOtExperimental mvestigatton of the behavior of long-livedradioisotopes in a natural water clay system. in \lurr,~c,Radioeco/ogy Proc~ 3rd NE.4 Seminar. Tokyo. I-5 Octokr1979. OECDDIJURSM* E. K. and PAKSIP. (1974) Distributmn coefficientofplutonium between sediment and seawater. In l(~~l~rrrc~\011hcInrcwational Lahorarori~ ot \!arrne Radr(wtrvrr\,IV?4 Report.\ IAEA. Vienna.FOTI S. C. and FREILIN~;E. C. (1964) The determination oithe oxidation states of tracer uranium, neptunium and plu-tonium in aqueous media. Talanta II, 385-392.GROMOVV. V. and SPITSY~:V. I. (1974) Sorption ofPu.Ru and *9Tc by bottom sediments of the Pacific Ocean.Sovie; Radiochem 16, 15?-161.KENNEY-KENNICUTTW. 1.. and MORSE J. W. ( 1985) Theredox chemistry of Pu(V) 0: interaction with mineral sur-faces in dilute solutions and seawater. Geochrm (o.rmcr-chum. .-lcta (in press).KIM J. I.. LIERSEC. H. and BAUMGARTNER. (1983) Com-plexation of the plutonium (IV) ion in carbonate-bicar-bonate solutions In Pluionium Chemistry (eds. W. T. CAR-NALLand G. R. CHOPPIN). ACS Symposium Series 2 16.ACS. 317-334.MAGNUSSONL. B. and LACHAPELLE . J. (1948) Ihe firstisolation of element 93 in pure compounds and a deter-mination of the half-life of Q1Nu27.. Am. Chem Sot 70.3534.METIVIERH. and GUILLAUMONTR. (1976) Hydrolyses andcomplexing of tetravalent plutonium. J. Inor,g.Nucl. Chem..Supplement 1976 179-183.MINER F. J. and SEED J. R. (1967) Radiation chemistry ofplutonium nitrate solutions. Chem. Rev. 67. 299-3 15.MOSKVINA. 1. and GELMANA. D. (1958) Determination ofthe composition and instability constants of oxalate andcarbonate complexes of plutonium (IV). Russ .I InorgChem. 3(4), 198-216.

NASH K., SHERMAN .. FIEDMANA. M. and SULLIVAN . C.( 198 I) Redox behavior, complexing, and adsorption ofhexavalent actinides by humic acid and selected clays. EmSci. Techn. 15, 834-837.NELSOND. M. and LOVETTM. B. (1978) Oxidation state ofplutonium in the Irish Sea. Nature 236, 599-601.NEL_WN D. M. and LQVEIT M. B. (1981) Measurements ofthe oxidation state and concentration of plutonium in in-terstitial waters of the Irish Sea. In Impacts oJRadionuclide

Releases into the Marine Environment. Pmt. Svmp.,Otaniemi. 30 June-4 July 1975. IAEA. Vienna.NELSOND. M. and ORLANDINIK. A. (1979) Identificationof Pu (V) in natural waters. In RadiologicaL and Envrron-mental Research Division Annual Report, ANL-79-65, PartIII, pp. 57-59.NELSON D. M.. KAR~UNEN J. O., ORLANDINIK. A. andLARSENR. P. (1980) Influence of dissolved organic carbonon the sorption of plutonium to natural sediments. In Ra-diological and Environmental Research Division AnnualReport, ANL-80-I 15, Part 111. p. 19-25.NELSON D. M.. KARTTUNEN . 0. and MEHLHOFFP. (1981)Influence of colloidal dissolved organic carbon (DOC) onthe sorption of plutonium in natural sediments. In Radio-

logical and Environmental Research Division .4nnual Report, ANL-81-I 15, Part III, pp. 48-52.PROUT W. E. (I 957) Adsorption of radioactive wastes by Sa-vannah River Plant soil. Soil Science 84, I3- 17.RABIDEAUS. W. (1957) The kinetics of dispropottionationof plutonium (V). J. Amer. Chem. Sot. 79,6350-6353.RHODESD. W. (1957) Adsorption of plutonium by soil. SoiLSri. 84, 465-47 I.ROZZELLT. C. and ANDELMAN . B. (1971) Plutomum mthe water environment. Il. Sorption of aqueous plutoniumon silica surfaces In Nonequilibrium Systems in Natural


11/11

Pu adsorption on goethite 2307Wuter Chemistry w i. . D. HEM). Adv. Chem. Ser. # 106,ACS.SANCHEZA. L. (1983) Chemi cal speciat i on and adsorpt ionbehavior f plut onium n natural w aters. h.D. dissertation,University of Washington, Seattle, WA.SANCHEZ . L., MURRAYJ. W., SCHELLW. R. and MILLERL. (in press) FaIlout plutonium in two oxic-anoxic envi-ronments. Li mnol . Oceanogr.SANCHEZA. L.. SCHELLW. R, and SIBLEYT. H. (1982) Dis-tribution coefficients for plutonium and americium on par-ticutates in aquatic environments In Prod Svmp. Envir on-mentatMigrat ionof ong-Li ved Radi onuchdes,Knoxville,TN, 27-31 July 1981. IAEA, Vienna.SCNINDLERP. W. (198 ) Surface complexes at oxide-waterinterfaces. Ch. 1. n Adsor~ion of norganics t Soii d-Li quidI nt erf kces eds. M . A. ANDERSON nd A. J. RUBIN). AnnA&r Sci. Publ., Ann Arbor, Michigan.SCHWABA. P. and FELMV A. R. (1983) Crit ical hermodv-namic data evaluat ion nd recalculat ion ir Pu and Np.Abstracts 185th Amer. Chem. Sot. National Meeting, Seat-tie, March 1983. p. 63.SIBLEYT. H., CLAYTON . R., WURTZ E. A., SANCHEZ . L.and ALBERTS . J. (1984) Effects of dissolved organic com-

pounds on the adsorption of transumnic elements. In~ornp~e~t~ on oj Trace h& tal s n Nat ural W aters eds.C. J. M. KRAMER nd J. C. DUIRKER). Martinus Niihoff/Dr. W. Junk Publishers, The Hague, Netherlands. _SIMPSON H. J.. TRIER R. M., OLSEN C. R., HAMMONDD. E.. EGE A., MILLER .and MEL&K J. M. ( 1980) FaIloutPlutonium in an alkaline, saline lake. Science 207, 1071-1073.WAHLGRENM. A., NELSOND. M., ORLANDINIK. A. andLARSENR. P. (1978) Plutonium in fresh water. In Rudio-logicai nd Environmental Research Di vi sionAnnual Re-port ,ANL-78-65, Part III. pp. 64-68.WESTALL . C. and HOHL H. (1980) A comparison of elec-trostatic models for the oxide/solution interface. Adv. CON .I nt er. Sci. 12, 265-294.WEBTALL, C., ZACHARY . L. and MORELF. M. M. ( 1976)MINEQL, a computer program for the calculation ofchemical equilibrium composition of aqueous systems.

Water Qual. Lab., Tech. Note No. 18, Dept. of Civil Eng.,Mass. Inst. of Tech., Cambridge.YA~XS . E., LEVINES. and HEALYT. W. (1974) Site bindingmode1 of the eIectrica1 double layer at the oxide/water in-terface. J. Chem. Sot . Puraday Trans. l & JO , 807- 18 8.

Documents

Adsorption Plutonium Varied Oxidation States