This official electronic version was created by scanning the best available paper or microfiche copy of the original report at a 300 dpi resolution. Original color illustrations appear as black and white images. For additional information or comments, contact: Library Without Walls Project Los Alamos National Laboratory Research Library Los Alamos, NM 87544 Phone: (505)667-4448 E-mail: [email protected]
THE EFFECT OF FLUORIDE AND ALUMINUM ON THEANION EXCHANGE OF PLUTONIUM FROM NITRIC ACID
by
S. Fredric Marsh
ABSTRACT
Anion exchange in nitric acid is a prominent aqueous process usedto recover and purify plutonium from impure nuclear materials. Thisprocess is sensitive to fluoride ion, which complexes plutonium in com-petition with the anionic nitrato complex that is strongly sorbed onthe anion exchange column. Fluoride interference traditionally has beencounteracted by adding a masking agent, such as aluminum, that formscompeting complexes with fluoride. The interfering effect of fluoride isknown to be a function not only of the fluoride-to-aluminum ratio butalko of the fluoride-to-plutonium ratio, yet no comprehensive study ofthis three-component system has previously been reported. This reportsummarizes a Los Alamos study of the effect of 25 fluoride-aluminum-plutonium combinations on the anion exchange sorption of plutonium.Five aluminum-to-plutonium ratios ranging from 0.10 to 10 were eachevaluated at five fluoride-to-aluminum ratios that ranged from O to 6.The fluoride-to-plutonium ratio has a greater influence on plutoniumsorption than does the fluoride-to-aluminum ratio. Aluminum was lesseifective as a masking agent than had been assumed, because measurablefluoride interference occurred at all levels of added aluminum.
INTRODUCTION
Anionexchange in nitric acid is the principal aque-ous process used to recover and purify plutonium atthe Los Alamos Plutonium Facility. A Los Alamosdevelopment effort directed at improving this processhas resulted in (1) the replacement of previously usedgel-type resin with macroporous resin,l (2) an im-proved feed treatment procedure that uses only hy-drogen peroxide,2 and (3) a systematic evaluation ofthe effect of plutonium concentration and solution flowrate on the effective capacity of the selected macro-porous resin.3
Aqueous feed solutions used in this process alwayscontain fluoride, which is required to dissolve refrac-tory plutonium oxide feed materials at an accept-able rate. Fluoride also is required to dissolve thelarge quantities of silicate materials often found inplutonium-contaminated scrap materials.
Fluoride accelerates the dissolution of difficult-to-dissolve compounds of plutonium because it formshighly stable complexes with plutonium. The anionexchange separation that follows, however, depends onthe anionic nitrato complex of plutonium that stronglysorbs on the resin. Hence, the plutonium fluoride corn-.plex that is an asset during dissolution is a liability
1
during the subsequent anion exchange separation be-cause it interferes with formation of the plutonium ni-trate complex.
Fluoride interferencecan be reduced by adding amasking agent known to form a competing, high-stability, soluble fluoride complex. Aluminum com-monly is added to plutonium feed solutions for thispurpose. The resulting solutiona contain plutonium,fluoride, and aluminum in relative amounts that varyover wide ranges. The many competing equilibria in-volved in such solutions make subsequent anion ex-change behavior of plutonium difficult to predict.
Interference by fluoride during the processing ofplutonium by anion exchange is known to be a functionof not only the fluoride-to-plutonium ratio but also thealuminum-to-fluoride ratio. Aluminum commonly hasbeen added to mask the interference of fluoride dur-ing most of the nearly thirty years that nitrate anionexchange has been used to purify plutonium. Yet weare unable to find a published, systematic study of theeffect of specific quantities of aluminum and fluorideon the plutonium anion exchange process. Our inves-tigation provides this previously unavailable informa-tion, which should facilitate the processing of fluoride-containing solutions of plutonium.
EXPERIMENTAL
Experimental Design
Plutonium. The plutonium concentration wasmaintained at a constant level of four grams per literthroughout this series of experiments. The oxidationstate of all plutonium used in this study was adjusted -to Pu(IV) with hydrogen peroxide,2 after which theabsence of Pu(VI) and Pu(III) was confirmed by spec-trophotometric analysis.
Aluminum. Five levelsof aluminum,relative toplutonium,wereobtainedby addingappropriatequan-tities of 2.2 Maluminumnitrate solution. The selectedaluminum-to-plutoniummole ratios were 0.10, 0.33,1.0, 3.0, and 10.
Fluoride. Fluoride was added, as hydrofluoricacid, to provide five fluoride-to-aluminumratios ateach level of aluminumto plutonium. The selectedmole ratios of fluorideto aluminumwere O, 1.5, 3.0,4.5, and 6.0.
Nitric Acid. Sufficient nitric acid and waterwere added to maintain the total acid concentration
at 7 M. (The small amount of hydrogen ion presentas hydrofluoric acid was included in the calculation oftotal acid concentration.)
Nitrate. Because aluminumwasaddedas the alu-minum nitrate salt, three moles of additionalnitratewere added with every mole of aluminum. This pro-vided additionalnitrate that contributed to a vary-ing total nitrate concentrationin excess of the nitratepresent as 7 M nitric acid.
Resin. Lewatit MP-500-FK (40- to 70-mesh)macroporous anion exchange resin (obtained fromMobay Chemical Company,Philadelphia,Pennsylva-nia) was used in this study. Individual10-milliliterportions of these solutionawere contacted with por-tions of air-dried, nitrate-form Lewatit resin. Aweighed amount of air-dried resin that correspondedto a 2-milliliter volume of nitrate-form re9in in waterwas taken for each contact. (The specified volume ofwet resin weighed 0.66 gram in air-dried form.)
Assay Technique
Dynamic batch contacts of each combination of so-lution and resin were achieved using a wrist-actionshaker. Each solution initially contained the selectedquantities of plutonium, aluminum, and fluoride. Af-ter each dynamic contact period, a measured portionof solution was removed for assay of plutonium usinggamma spectrometry, by assaying the 129-keV gammaray of 239Pu.
After the first dynamic contact period was com-pleted, the shaker was stopped only long enough toremove a solution aliquot for radioassay. The dynamiccontact was quickly resumed until the second contactperiod had been completed, after which another aque-ous portion was removed for assay. This procedurewas repeated until three aqueous portions had beenremoved for assay after dynamic contact perioda of10, 20, and 60 minutes.
The fraction of initial plutonium in each contactedsolution was calculated relative to the quantity ofZsgpu in ~ identical portion of uncontacted solution.From these data, plutonium distribution coefficientswere calculated for three contact periods at each of the25 fluoride-to-aluminum-to-plutonium ratios. Sequen-tial measurements for multiple contact perioda pro-vided important sorption kinetics data for the pluto-nium nitrato complex, whose sorption rate is knownto be particularly slow.
Distribution coefficients (Kd) were calculated asfollows:
In practice, Kd values were calculated by the equation:
Kd = [(At - At)/Al] /(VJVr),
whereAt is the activity of the actinide initially in the liquid,Al is the actinide activity in. the liquid after contact,V1 is the liquid volume, andV, is the volume of wet resin.
Because each fluoride-to-aluminum-to-plutoniumcombination was used for a series of three sequen-tial contact measurements, the liquid-to-resin ratiochanged as each aqueous portion was removed for as-say. We corrected for this changing ratio as well as forthe decrease in the total activity remaining in the vialafter the removal of each aqueous portion.
We elected to express Kd in terms of wet-resin vol-ume, contrary to the common practice of expressingKd in terms of dry-resin weight, for reasons explainedin detail in a previous report.1 Whichever conventionis used, however, the relative effect of fluoride and alu-minum remains the same.
RESULTS AND DISCUSSION
All solutions used in this study were prepared to ini-tially contain the appropriate amounts of fluoride, alu-minum, nitric acid, and resin. Plutonium was addedlast, immediately before the dynamic contact began.Under normal circumstances, the olive green color ofthe plutonium nitrate complex appears promptly whenPu(IV) is added to 7 M nitric acid. When plutoniumwas added to these solutions that contained varying ra-tios of fluoride and aluminum, however, a wide rangeof colors was observed. Some solutions showed theexpected green color, some showed a lighter shade ofgreen, and some showed no green. The intensity ofthe green color indicates the concentration of pluto-nium nitrate complex remaining in solution, and thus,the extent of fluoride interference in each case.
Because the maximum aluminum-to-plutonium ra-tio was 10 and the maximum fluoride-to-aluminumratio was 6, the highest fluoride-to-plutonium ratioin this study was 60. At this maximum fluoride-to-plutonium ratio, and in a few other solutions wherethe fluoride-to-plutonium ratio was high, the expected
green color of the plutonium nitrate complex was ei-ther absent or quite pale. Furthermore, in those so-lutions in which the fluoride-to-plutonium ratio was60 or 45, a straw-colored precipitate, assumed to bePuF4, was observed.
If plutonium were being precipitated from certainsolutions, in competition with removal from solu-tion by ion exchange processes, an assay of pluto-nium remaining in the liquid portion after the con-tact would seriously overestimate the quantity of plu-toni~ sorbed on the resin.
The total absence of green color in certain high-fluoride solutions confirmed that the fluoride level wassufficient to prevent formation of the plutonium ni-trate complex. In the three solutions in which precip-itate formation precluded measurement of the quan-tity of plutonium sorbed on the resin, we can confi-dently conclude that fluoride interference was severe,even though we are unable to provide quantitativevalues.
Experimental Data
The experimental data are presented in two for-mats. In Figs. 1 through 5, the ratio of aluminum-to-plutonium is held constant to demonstrate the ef-fect of varying fluoride-to-aluminum ratios. In Figs. 6through 10, the fluoride-to-aluminum ratio is held con-stant to demonstrate the effect of varying aluminum-to-plutonium ratios.
Figure 1 shows that even small amounts offluoride measurably suppress the distribution coef6-cient of Pu(IV) onto the resin. The interference offluoride increases for a given fluoride-to-aluminumratio as the fluoride-to-plutonium ratio increases,as shown in Figs. 1 through 5. (The fluoride-to-plutonium ratio is the product of the fluoride-to-aluminum ratio and the aluminum-to-plutonium ratioin Figs. 1 through 5.)
Figure 6 shows the effect of increasing levels of alu-minum nitrate when fluoride is absent. The increas-ing distribution coefficient of Pu(IV) at the higheraluminum-to-plutonium ratios is attributed to the in-creasing concentration of nitrate ion present as alu-minum nitrate. In the maximum case of A1/Pu = 10,the concentration of additional nitrate ion is 0.5 M.
Figures 7 through 10 present the same data shownin Figs. 1 through 5 in a different format. These fig-ures show the effect of increasing total amounts of flu-oride when the fluoride-to-aluminum ratio is held con-stant. The increasing interference of -fluoride at thehigher fluoride-to-plutonium ratios is apparent. (The
3
A1/F’u-O.10 AliRJ -1
1“~ low I 1 I I I I I
D–-–F/N .0
— WA “ 1.5‘*— FIN = 3.0—— FIN = 4.6‘-— FIAI = 6.0 ----
<
----
“;”>-”
/
-./--—--—../
//y’,”,/
Im‘ --- FIAI = O— FJAI = 1.6‘. — FIN = 3.o——- flAl = 4.6‘--— FIN = &O
---- -1-Z / /~”Mg ./”g 100
u-Illo0 42=-
./” ———-----/
/ --—--—---/-----//
/
,~10 20 30 40 60
CONTACT PER10D,5~in CONTACT PERIOD, min
Fig. 3. Sorptionof Pu(lV) on LewatitMP-500-FKanionexchange reain from,7 M nitric acid that contains varyingratios of fluoride to aluminum for A1/Pu = 1, as a functionof dynamic contact time.
Fig. 1. Sorption of Pu(IV) on Lewatit MP-500-FK anionexchange resin from 7 M nitric acid that contains varyingratios of fluoride to aluminum for A1/Pu = 0.1, as a func-tion of dynamiccontact time.
A1/Pu =0.3Slow
r
m
‘ --- F/N = O— F/Al = 1.6‘* — FIN = 3.0—— FIN = 4.5‘--- F/N = 6.0
1-
~’” ~/./
{<,/
/
A1/Fu= 3
‘“m ‘ ‘r----FIAJ = 0.0 --- —-.—
,/ 0“1- .
’21 /“//
.-.—.— .—. -./.0
/
-— —————-//
t -1-l—-_—__—--—--—-
110 20 30 40 60 60
CONTACT PERIOD, minCONTACT PERIOD, min
Fig. 2. Sorption of Pu(IV) on Lewatit MP-500-FK anion
exchange resin from 7 M nitric acid that contains varyingratios of fluoride ta aluminum for A1/Pu = 0.33, ss a fnnc-tion of dynamic contact time.
4
Fig. 4. Sorption of Pu(IV) on Lewatit MP-500-FK anion
exchange reain from 7 M nitric acid that contains varyingratios of fluoride to aluminum for A1/Pu = 3, ss a functionof dynamic contact time.
I -.—.—.—.—.—~.. ICONTACT PERIOD, min
Fig. 5. Sorption ,of Pu(IV) on Lewatit MP-500-FK anionexchange resin from 7 M nitric acid that contains varyingratios of fluoride to aluminum for A1/Pu = 10, as a functionof dynamic contact time.
fluoride-to-plutonium ratio may be calculated as pre-viously explained for Figs. 1 through 5.) Figures 9 ahd10 show the almost total interference of fluoride in so-lutions in which .41/Pu = 3. Precipitation preventedmeasurement of distribution coefficients at the highestlevels of fluoride in solutions for which A1/Pu = 10, asshown in Figs. 9 iind 10.
CONCLUSIOIN”S
1. All levels of fluoride tested measurably suppress the distribution coefficient of plutonium, no mat-ter how much aluminum is added. Although alu-minum counteracts the interference of fluoride to a
large extent, some interference always occurs. Alu-minum therefore is less effective for masking fluoridethan had been supposed.
2. Aluminum is much more effective as a mask-ing agent at a given fluoridet~aluminum ratio whenthe fluoride-to-plutonium ratio is low, than when thefluoride-to-plutonium ratio is high. This indicates thatthe anion exchange sorption of plutonium is more af-fected by the fluoride-t~plutonium ratio than by thefluoride-to-aluminum ratio.
3. On the basis of this study, we recommend thatthe quantity of aluminum added be appr oximatelyequ”molar with the amount of fluoride present. Evenat this ratio, however, some interference is to be ex-pected, particularly when fluoride is high relative toplutonium.
4. Excessive aluminum should be avoided, as itscompetition for fluoride converts soluble fluoro com-plexes of elements such as silicon and tungsten to in-soluble oxides, which can coat ion exchange resin andplug proceas filters and valves.
5. We had hoped that this study would resultin a general equation that would define the level ofaluminum required to mask any specified amount offluoride. Unfortunately, the complex interactions ofthis chemical system do not lend themselves to such+rnplistic interpret&.ion. It appears, instead, thateach specific fluoride-to-plutonium ratio may requirea unique amount of aluminum to effectively mask flu-oride, without causing the precipitation problems de-scribed in the preceding paragraph.
We therefore are generating digitized, visible/near-infrared spectra for all 25 solutions examined in thisstudy. These data will be processed by modern chemo-metric techniques to produce an initial data base thateventually will allow us to use on-line apectrophoto-metric measurements as a basis for controlling andoperating the anion exchange process at near-optimume%iciency.
Fig. 7. Sorption of Pu(IV) on Lewatit MP-500-FK anion
exchange r=in from 7 M nitric acid that contains varyingratios of aluminum to plutonium for F/Al = 1.5, as a func-tion of dynamic contact time.
6
------------
I -.-”,.
,- .-.—. —0-, ,, ./
,,/
./
./ --——— —-
J’/ 0
/“/’/--—--—--—-.--
,./~
,~CONTACT PERIOD, min
Fig. 8. Sorption of Pu(IV) on Lewatit MP-500-FK anionexchange reain from 7 M nitric acid that contains varyingratios of aluminum to plutonium for F/Al = 3, sa a functionof dynamic contact time.
Fig. 9. Sorption of Pu(IV) on Lewatit MP-500-FK anionexchange resin from 7 M nitric acid that contains varyingratios of aluminuq to plutonium for F/Al = 4.5, ss a func-tion of dynamic contact time.
‘OOOIEEi!iEi!i-------
------.@”-,.-
/’/’ .= .-. — “—
./‘“.//’
-—————//
/’ /
IL10 20 30 40 60 s“
CONTACT PERIOD, min
REFERENCES
1.
2.
3.
S. F. Marsh, “Improved Recovery and Purifica-tion of Plutonium at Los Alamos Using Macro-porous Anion Exchange Resin,” Los Alamos Na-tional Laboratory report LA-10906 (May 1987).
S. F. Marsh and T. D. Gallegos, “Chemical Treat-ment of Plutonium with Hydrogen Peroxide Be-fore Nitrate Anion Exchange Processing,” LosAlamos National Laboratory report LA-10907 (May1987).
S. F. Marsh and T. D. Gallegos, “The Infiu-ence of Plutonium Concentration and Solution FlowRate on the Effective Capacity of Macroporous An-ion Exchange Resin,” Los Alamos National Labo-ratory report LA-1099O (July 1987).
Fig. 10. Sorption of Pu(IV) on Lewatit MP-500-FK anion
exchange main from 7 M nitric acid that contains varyingratios of aluminum to plutonium for F/Al = 6, as a functionof dynamic contact time.
7
Page R.wgeNTIS
PriIx Code
IX31.025026.0500514375076100101.12s126.150
A02A03A04A05A06A07
Printedinthe United Statesof AmericaAvailable from
National Tcchnicd lnforma!ion serviceUS Department of Commerce
S28S Porf Royal RoadSpringfield,VA 22161
Microfiche (AOI)
Page RangeNTIS
Pricecode Page RmgeNTIS
PriCCcode
Is 1.17s[76.200201.225226.2S0251-275.776.300
A08A09A 10AllA12A13
301.32S326.350351.375376.400401.42S426.450
A 14A 15A 16A 17A18A19
Page Range
45 I .47s476.500SO1.525S26.550551.57s576.60060[ .Up”
