1-s2.Method for the separation of uranium( IV) and (VI) oxidation states in natural waters 0-002196739600235X-main

7/28/2019 1-s2.Method for the separation of uranium( IV) and (VI) oxidation states in natural waters 0-002196739600235X-m

1/6

J OURNAL OFCHROMATOGRAPHY A

ELSEVIER Journal of Chromatography A, 743 (1996) 335-340

Short communicationMethod for the separation of uranium( IV) and (VI) oxidation states

in natural watersMartine Carol Duff*, Christopher Amrhein

Department of Soil and Envi ronmental Sciences, The Univ ersit y of Calif orni a at Riv erside, Ri verside, CA 92521, USA

Received 6 November 1995; accepted 15 March 1996

AbstractA low-pressure cation-exchange method has been developed for separating U(IV) and U(VI) in natural, high carbonate

waters. Uranium(IV)/(VI) solutions were prepared in 0.125 M H,C,O,-0.25 M HNO ,, equilibrated with a N, atmosphere,applied to a Dow ex AG 5OW -X8 cation-exchange resin, and eluted with 0.125 M H,C,O,-0.25 M HNO,. Uranium(IV)eluted in four lo-ml fractions w herea s U(V 1) eluted within 9 to 19 lo-ml fractions. Additional U(IV)/U (VI) separationswere done on solutions containing known concentrations of PO:-, Ca*+ , Nat, Cl-, SO:-, CO:-, and F-. Interferenceswith U(VI) elution were observed at 0.05 M Ca*+.Keywords: Water analysis; Uranium; Inorganic ions; Metal cations

1. IntroductionElevated levels of U are found in shallow ground

waters of agricultural drainage water evaporationponds in the San Joaquin Valley (SJV ), CA, USA.Drainage waters are periodically pumped into evapo-ration ponds where they are evapoconcentrated.Consequently, the ponds undergo wetting and dryingcycles w hich affe ct the solubilities of redo x sensitiveelements in the pond waters and sediments. Thesebasins support migrating and native waterfo wl,whic h are threatened by elevated levels of potentiallytoxic U [l].

In the aqueo us environm ent, U(V 1) (uranyl,UOt+) is the most soluble U oxidation state. Understrongly reducing conditions, U(V 1) can be reduce d

* Corresponding author. Present address: Chemical Science andTechnology Division, Los Alamos National Laboratory, LosAlamos, NM 87545, USA.

to U(IV) (uranous), which is thought to precipitate asan insoluble oxide (UO,(,,) [2]. Recent res earchsuggests that U(IV) may be more soluble thanpreviously thoug ht - particularly in high carbonatealkalinity and saline waters [3-61. S tudies w ith Th[an actinide with a chemistry similar to U(IV)] inhigh carbonate alkalinity waters show that Th(IV) isquite soluble due to the formation of Th(IV)-carbon-ate species [7]. Hum ic and fulvic acid s form solublecomplexes with U(IV) in addition to U(VI) [8].Uranium(V ) is generally considered electrochem ical-ly unstable and is unlikely to exist in appreciableconcentrations in natural water s.

In natural water s, disso lved ions such as Ca+ ,CO:-, and PO:- commonly interfere with thedetermination of U oxidation states. One methodapplied to seawater is based on the coprecipitation ofU(IV) with NdF, [9,10]. This method is subject tointerferences from dissolved Ca2+ (-0.01 M). Highpressure liquid column chromatography has been

0021-9673/96/$15.00 0 1996 Elsevier Science B.V. All rights reservedPII SO02 I -9673(96)00235-X


2/6

336 MC Duff, C. Amrhein 1 J. Chromaogr. A 743 (1996) 335-340

used to separate 23gU(IV) and 239U(VI) species withanion-exchange resins in relatively pure laboratorysolutions free of potential interfering ions [ll]. Theseparation of U oxidation states has also beenperform ed with cation-exchang e resins in solutionsfree of interfering ions [ 121. Other U(IV)/U( VI)separation methods include the use of various com-plexing ions such as arsenazo (I, II, and III) whic hform distinct co mplexes with U(IV) and U(V1)whic h can be calorimetrically identified [ 13-151.Arsenazo compounds are also used as extractants forU(IV) and U(V1). These methods are subject tointerferences from Th(IV), F-, Fe3+, phosphates,vanadates, and arsenates. Another separation methodfor U oxidation states involves the pH dependentchelation of different oxidation states of U by adiketone (dibenzoylmethane, DBM ) in an organicsolvent [16,17]. This method has not been tested foruse with natural waters. When we attempted to usethe DBM extraction method with solutions con-taining soluble PO:-, Ca* +, Na+ , Cl-, SO :-,HCO ,, CO:- and F- and known concentrations ofU(VI), w e found the DBM method unsatisfactory inthe presence of carbonates, PO:-, F- and SOi-(Table 1). These anions, which form stable, dis-solved U(VI)-sp ecies, interfered with the chelationmethod. The DB M extraction was severely limited inthe presence of dissolved CO:- and to a lesserextent, SO:-. This made the extraction procedureunsuitable for the pond waters high in CO:- andSOi- concentrations. Therefore, we developed anew method using low-pressure chromatographicseparation.

We separated U(IV) from U(V1) on a cation-exchange resin using 0.125 A4 H,C,O,-0.25 MHNO , to elute the U species. Eluent fractions werecollected and analyzed for U with inductively cou-pled plasma mass spectrometry (ICP-MS). Theseparation method was tested for its applicability toagricultural subsurface drainage waters in the SJV .Synthetic water s containing soluble PO:-, Ca2 + ,Na+, Cl-, SO:-, CO:-, F- and known concen-trations of U(IV) and U(V1) w ere tested w ith themeth od to determine potential interferences withdissolved ions.

2. Experimental2.1. M ater ials

Reagen t-grade oxalic acid and Fisher trace-metal-grade concentrated HN O, were used to make aneluent of 0.125 M H,C,O,-0.25 M HNO , fromdeionized, organic-free water . Uranium(IV) in theform of a UCl, solid (in a sealed glass ampou leunder N,(,,) was purchased from ROC/RIC(Orange, CA, USA). An atomic absorption standardof U(V1) (1000 mg 1-l) in 5% nitric acid w aspurchased from Aldrich (USA).2.2. Chromatographic apparatus

The columns consisted of Beckman and Dickson60-ml plastic, luer-lock syringes and polycarbon ate,luer-lock stop cock s. A small plug of silanized glass

Table 1Recovery of 3.0 pg U(W) from various solutions using 0.5 M dibenzoylmethaneTreatment Recovery (%)

TreatmentNo Added PO:- or F- 2 mg 1-l P-PO:- 1mg 1-l F-

Deionized water0.1 M NaCl0.1 M NaOH0.1 M Na,SO,0.1 M NaHCO,0.1 M Na,CO,a Mean of triplicate samples.

98.420.1 59.8kO.2 96.1+0.192.7kO.l 64.oeo.2 74.120.294.820.1 81.1ZO.l 88.7kO.l87.3kO.2 59.920.0 70.1 kO.268.420.5 80.0k0.3 87.220.328.0k0.7 72.0k0.4 72.8eO.3


3/6

M.C. Duff, C. Amrhein I J. Chromatogr. A 743 (1996) 335-340 33-l

wool was placed at the syringe tip of each column. ABio-Rad Dow ex AG 5OW -X8 H-saturated cation-exchange resin with a 200 to 400 mesh size, amedium pore size, and a cation-exchange capacity of1.7 mequiv. ml- (resin bed) was hydrated witheluent prior to packing in the columns. The hydratedresin was poured to a volume of 40 ml (which gave acolumn height of 6.5 cm and a width of 2.6 cm) andhad a 20-ml po re volume. The columns were washedwith 200 ml of eluent to stabilize the resin or untilthe outgoing eluent w as colorless.2.3. Glove bo x

A shop-built glove bo x was used to maintain a low0 2(gj atmosphere. It consisted of a clear, 115 cmX61cmX5 5 cm polycarbonate housing with portholes forattachable gloves. Small holes in the box permittedpurge gas (N2) in and out of the box and vacuumlines in. A portable balance was positioned in a boxhousing within the glove box. This allowed the UCl,solid to be weighed in the absence of the purge g asstream.2.4. Procedures for the cation-exchange metho d

All solutions were placed in the glove box. Theywere loosely covered, and purged with N, gas for atleast 20 min prior to use. The glove box w as thenpurged for 20 min to ensure removal of O,(,,. Theampoule containing the UCl, was opened and ap-prox. 0.1 g of UCl, was weighed out within theglove bo x and dissolved in 100 ml of deionizedwate r. This solution was diluted with eluent bytransferring approx. 1 ml of the concentrated UCl,solution into 50 ml of eluent. A U(V 1) sto ck solutionwas made in eluent by diluting the 1000 mg U(V1)l_ standard (Aldrich).

The columns were wash ed with at least 50 ml ofeluent to remov e dissolved 0,. Fifty ,ul of U[containing approx. 6.4 mg U(IV) 1-l and 5.8 mgU(V 1) 1-l in eluent] were carefully applied to thecolumn, to insure minimal disturbance to the resinbed, and twenty lo-ml elution fractions were col-lected.

To determine if there were interferences with ionscommonly found in the SJV subsurface drainagewaters and ground waters, several solutions con-

Table 2Recovery of approx. 2.3 pg U(IV)/O.2 pg U(W) from varioussolutions using cation exchange and an eluent of 0.125 MH&0,-0.25 M HNO,Treatment Recovery (%)

WV) uwDeionized water 99.9 99.90.05 M NaCl 99.9 94.90.03 mM PO:- 100 97.40.05 mM F- 100 1000.05 M Na,SO, 99.0 99.30.05 M CaCl, 99.5 79.50.01 M CaCl, _ 98.30.003 M CaCl, _ 99.50.05 M Na,COt 100 93.5a Mean of duplicate samples.b Sample carbonate alkalinity was neutralized with nitric acid priorto separation.

taining PO:-, Ca* , Na+ , Cl-, SO :-~, CO:- and F-(Table 2) were made. Known quantities of U(IV)and U(V1) were added to the synthetic solutions inthe glovebox and 150 ~1 of these solutions wereapplied to the column. Twenty-five lo-ml fractionswere collected.

2.5. Analysis of uraniumUranium was determined by ICP-MS (VG Plas-

maquad 2T , Fisons Instruments, USA). Bismuth wasused as the internal standard . The ICP-M S sensitivitywas greatly improved with an enhanced interface(detection limit of -9 ,ug 1-l U).

3. Results and discussion3.1. Cation-exchange procedure

The separation of approx. 0.32 ,ug U(IV) and 0.29,ug U(V1) by cation exchange is shown in Fig. 1.This figure represents the results of two columns andtwo replicates per column. The U(IV) elutes withinthe first four lo-ml fractions and the U(V 1) eluteswithin nine to nineteen lo-ml fractions. The U(IV)


4/6

M.C. Duff, C. Amrhein I J. Chromatogr. A 743 (1996) 335-34038

U25

20

15

10

5

0

1.

,

L

1 2 3 4 5 6 7 8 9 1011121314151617191920

Dowex 5OW-X8 Cation-Exchange Resin0.125 A4 H&O,

0.25 AUHNO,Column

WV)/ 01I42

Ten mL Elution FractionsFig. 1. The elution of 0.32 yg U(W) and 0.29 pg U(V1).

peak is sharp whereas the U(VI) peak is somewh atbroad.

3.2. Cation exchange pro cedure with solutionscontaining potential interfering ions

In the presence of known concentrations of U(IV),U(VI), PO:-, Ca*+, Naf, Cl , SO:-, CO:- and F-,the elution of U(IV) occurs within the five lo-mlfractions whereas U(VI) elutes within eight totwenty-two lo-ml fractions with the exception of thesolution containing Ca2+ (Table 2). The greater peakbroadness is probably due to the presence of theinterfering dissolved ions. This factor decre ases th eseparation efficiency of the columns causing theU(IV) and U(VI) peaks to elute closer together.

There is some interference with the separation ofU(VI) in the presence of dissolved Ca*+ (Table 2).Calcium ion may compete with U(VI) for cation-exchange sites and further broaden the U(VI) peak,causing U(VI) to elute within the first eight totwenty-five lo-ml fractions as opposed to the firsteight to twenty-two lo-ml fractions. When sepa-rations are performed with lower concentrations ofCa*+ (0.01 M and 0.03 M), U(VI) elutes w ithineight to twenty-two IO-ml fractions and the recoveryof U(VI) is greatly improved.

3.3. The mechanism for the elution order of U(N)and U(W)

The elution order for this separation is not intui-tive. One would anticipate U(IV) to elute after U(VI)because the U(IV) ion has a greater valence andsmaller size than the U(V1) ion (which exists as theuo i ion). How ever, U(IV) ion generally formsmore stable inorganic complexes than U(VI) ion[ 181. Both U(IV) and U(VI) form several strong,highly soluble oxalate complexes (Table 3) but atlow pH, the UOz+ ion is the dominant U(VI) speciesin solution and it is not expecte d to form solublecomplexes with oxalate [19,20]. At high concen-trations of dissolved oxalate, U( IV)-oxalate com-plexes should dominate U(IV) solution speciation.

Table 3The U formation constantsReaction log KU4+ +H OtlUOH++H+U4+2&OttU(OH);++2H+

-1.0-2.0

U +3H,OtiU(OH); +3H+ -5.0U4+ +4H,O+&J(OH),.+4H+ -9.0U +5HzOtlU(OH),_ +5H+ -13.0UO;+H,OWJO,OH++H+ -5.2UO;++2H,O&JO,(OH),.+2H+ ~ 12.0UO;+ +3HzOttU0,(OH), +3H+ -20.0UO;+ +4H,OtlUO,(OH):- +4H+ -33.02UO; +H,0tl(U02),0H7+ +H+ -2.82UO;++2H,O++(UO,),(OH);++2H+ -5.63UO;++4H,Oti(UO,),(OH):++4H+ - 11. 93UO; +SH,Oti(UO,),(OH); +5H+ -15.53UO;++7H,O+UO,),(OH), +7H+ -31.04UO;+7H,O~(UO,),(OH);+7H+ -21.9uj+ +c,o:- +CJc,o:+ 8.6u~++2czo:ttu(c,o,),. 16.9U4++3Czo:~ttu(C,o,);- 22.7u4++4czo:~c+u(c,o,):- 27.7uo;++c~o~~+-+uo,c,o,. 4.6uo:+ +2c~o:tluo,(c,o,);~ 8.7uo;+ +3c~o:wJo,(c,o,)4~ 12.0U4+ + +NO,&JNO;+ -0.3UO;++NO, tlUO,NO; 0.1Stability constants for U-oxalate species from Zakharova andMoskvin [19]; Mikaye and Numberg [20].Stability constants for U-nitrate species from Arhland [21] andreferences therein.Stability constants for U(IV)-hydrolysis species from Lemire andTremaine [23].Stability constants for U(VI)-hydrolysis species from Grenthe etal. [24].


5/6

Using the comp uter equilibrium speciation pro-gram FITEQL [22] and the U formation constants inTable 3, the solution speciation distribution forU(IV) and U(V1) species in the eluent was calcu-lated at 0.38 M ionic strength with the Daviesequation (Table 4). The speciation calculations pre-dict that in 0.125 M H,C,O,-0.25 M HN O,, theU(C,O,)!- species dominates U(IV) speciationwhereas the free UOi+ ion, and to some extent, theUO,C,O,. species are the predominant U(V1)species in solution. This is found to be applicable forthe U concentration range of 10m5 to lo- lo M [forboth U(IV) and U(VI)].

u jJgL-l)WI)

4 I ,n J Column

An explanation for the elution order may be (1)the relative low affinity of the U(C,O ,):- species,which does not adsorb to the resin and (2) the strongadsorption of U(V1) (as UO:) to the resin. Assum-ing that U-oxalate complex ation reactions in aque-ous solutions are rapid proc esses, it is likely thatU(IV) would elute prior to U(V1). The interferenceby high concentrations of dissolved Ca*+ with theelution of U(V1) may be explained by competitionbetween Ca*+ with UOz+ for exchange sites on theresin.

1

0 5 10 15 20 25Ten mL Elution FractionsFig. 2. The elution of 0.19 pg U(W) in a reducing natural watersample.

3.4. Applying the metho d to natural, high-salinityand high-carbonate waters

The method has been applied to a natural water incontact with reducing sediments. The redox potentialof the waterfrom the sediments was -0.0% At thispotential, U(V 1) is thermodyn amically stable. Thesample has the following dissolved constituents0.030 M Na+, 0.004 M Ca*+, 0.008 M Mg*+, 0.002M K+, 0.026 M SO;-, 0.007 M Cl- and a carbonatealkalinity of 0.02 5 M.

A 0.5-ml sample is filtered and diluted with 0.5 mlof 0.25 M H,C,O,-0.55 M HN O, that has beenequilibrated with a N, atmosphere. The additionalHN O, neutralizes the samp les carbonate alkalinityand helps low er the sample pH. The sample is mixed(by vortexing) one min in a N,(,, head space . A50-~1 aliquot is applied to the exchange resin and 25lo-ml elution fractions are collected as describedpreviously (Fig. 2). The elution of U(V 1) [-0.19 pugU(VI)] occurs between the 9 and 22 IO-ml elutionfractions. Although U(IV) was predicted to be thestable U oxidation state, there was little U(IV) in thesample (co.1 ng). In addition, a small d etectablequantity of U is observ ed within the fifth and sixthlo-ml elution fractions. This may be due to the

Table 4The calculated equilibrium percent distribution obtained for the various U(W) and U(V1) species in the eluent of 0.125 M H,C,O,-0.25 MHNO,U(W) solution species:U(C,OX U(C,Ox U(C,O,),. uc,o:+ UNO;%98.2 I .I co.1 co.1 co.1

M.C. Duff, C. Amrhein I J. Chromatogr.A 743 (1996) 335-340 339

U(V1) solution species:uo:+%82.2

uo,c,o,. uo,tc,o,):~ UO,(C,O,):- UO,NO_:

12.9 1.9 0.5 2.9


6/6

340 M.C. Duff, C. Amrhein I J. Chromatogr A 743 (1996) 335-340

presence of small quantities of U(V ) in the naturalsample. No peaks are ever observed between the fifthand eighth lo-ml elution fraction s when knownquantities of U(IV) and U(VI) are separated withthis method as in Section 3.1 and Section 3.2 above.

In conclusion, the cation-exchan ge proce dure isless subject to interferences than the DB M chelationmethod. Because of considerable interferences withdissolved ions, the DBM chelation method is notrecommended for U(IV) and U(VI) separations inhighly-alkaline, saline wate rs. The cation-exchan gemethod is subject to interferences from high con-centrations of Ca+, w hich broaden the U(VI) peak.Low er concentrations of Ca*+ have little effect onthe recovery of U(V1). T he detection of U wasgreatly aided by the enhanced interface, whichenables ICP-MS detection below 1 O pg U I-.

AcknowledgmentsThis work w as supported by a grant from the

University of California Salinity/Drainage Progra m(No. 92-l 1). We would like to extend great ap-preciation to Dr. Paula J. Bosserman for her sug-gestions and assistance with this research.

References[I] J.E. Chilcott, D.W. Westcot, A.L. Toto, and CA. Enos. 1990.

Water quality in evaporation basins used for the disposal ofagricultural subsurface drainage water in the San JoaquinValley, California. Central Valley Reg. Wat. Qual. ControlBd. Rep., Dec., Oakland, CA.

[2] D. Langmuir, Geochim. Cosmochim. Acta, 42 (1978) 547-569.

[3] J. Bruno, I. Casas and I. Puigdomnech, Geochim. Cosmo-chim. Acta, 55 (1991) 647-658.

[4] D.W. Wester and J.C. Sullivan, Inorg. Chem., 19 (1980)2838-2840.

[5] L. Ciavatta, D. Ferri, I. Cirenthe, F. Salvatore and K. Spahiu,J. Inorg. Chem., 22 (1983) 2088-2092.

I61[ 71PI

[ 181

r191

WVI I

[ 221

1231~241

A. Giblin and E. Appleyard, Appl. Geochem., 2 (1987)285-295.E. Osthols, J. Bruno and I. Grenthe, Geochim. Cosmochim.Acta, 58 (1994) 613-623.G.R. Choppin and B. Allard, in A.J. Freeman and C. Keller(Editors), Handbook on the Physics and Chemistry of theActinides, Elsevier, Amsterdam, 1985, pp. 407-429.R.F. Anderson, Nucl. Instrum. Methods Phys. Res., 223(1984) 213-217.R.F. Anderson, M.O. Fleisher and A.P. LeHuray, Geochim.Cosmochim. Acta, 53 (1989) 2215-2224.I. Al Mahamid and J.M. Paulus, Radiochim. Acta, 48 (1989)39-42.Z.K. Karalova, N.P. Shibaeva and Z.I. Pyzhova, J. Anal.Chem. USSR, 23 (1968) 381-386.I. Kuznetsov and S.B. Savvin, Radiokhim., 2 (1960) 682-686.S.B. Savvin, Dovlady., 127 (1959) 1231-1234.J.S. Fritz and M. Johnson-Richard, Anal. Chim. Acta, 20(1959) 164-170.A. Saito and G.R. Choppin, Anal. Chem., 55 (1983) 2454-2457.G.R. Choppin, G.R. Choppin, J.D. Navratil and W.W. Schulz(Editors), Proc. Int. Symp. on Actinide/Lathanide Sepa-rations, World Science Publications, Phildelphia, PA, 1984,pp. 176-193.J.J. Katz, L.R. Morss and G.T. Seaborg, in J.J. Katz, G.T.Seaborg, and L.R. Morss (Editors), The Chemistry ofActinide Elements, Chapman and Hall, London, 1986, pp.1121-I 195.F.A. Zakharova and AI. Moskvin, Russ. J. Inorg. Chem., 29(1967) 241 I.C. Miyake and H.W. Nurnberg, J. Inorg. Nucl. Chem., 5( 1960) 592.S. Ahrland, in J.J. Katz, G.T. Seaborg, and L.R. Morss(Editors), The Chemistry of the Actinide Elements, Chapmanand Hall, London, 1986, pp. 1480-1546.A. Herbelin and J.C. Westall, A Computer Program for theDetermination of Chemical Equilibrium Constants fromExperimental Data, Version 3.1, Department of Chemistry,Oregon State University, Corvallis, OR, 1994.R.J. Lemire and P.R. Tremaine, J. Chem. Eng. Data, 25( 1980) 36 I-370.1. Grenthe, J. Fuger, R. Konings, R.J. Lemire, A.B. Muller,C. Nguyen-Trung and J. Wanner, The Chemical Thermo-dynamics of Uranium, Elsevier, New York, 1992.

Documents

1-s2.Method for the separation of uranium( IV) and (VI) oxidation states in natural waters 0-002196739600235X-main