LA-6577-PRProgress Report

C.3

t3G14 REPORT CC)&CTfON

REPRODUCTIONcoPy

UC-15

Issued: December 1976

Analytical Methods for Fissionable

Materials in the Nuclear Fuel Cycle

July 1, 1975— September 30, 1976

Compiled by

Glenn R. Waterbury

. –11

@

:

10s alamosscientific laboratory

of the University of CaliforniaLOS ALAMOS, NEW MEXICO 87545

/ \An Affirmative Action/Equal Opportunity Employer

g-

uNITED sTATES

ENERGY RESEARCH AND DEVELOPMENT ADMINISTRATION

CONTRACT W-740 S-ENG. 36

Previous reports in this series, unclassified, are LA-5064-SR, LA-5347-SR,LA-5696-SR, and LA-6045-SR.

This work was supported by the US Energy Research and DevelopmentAdministration, Division of Safeguards and Security.

Printed inthe United States of America. Available fromNational Technical Information Service

U.S. Department of Commcrcc528S Port Royal RoadSpringfic[d, VA 22161

Price: Printed Copy $3.5(3 Microfiche S3.1)(3

Th,. r.p,,,l S. SPWO.,MI a. mnaw.unt .rwu,k.wn,ordb, Ih.- (,nilcd State. (;... memt.t S.tth.r tbe I.n,t~S!. t..nor the [,.111+ S1.1,., En.rK\ Hesr”,(h ●nd 11...lqnnt.t Ad.minml..llon. nor . . . of!hcir.mplos (w,. .W. n. .r,hcircon.Ir.wt.r., .. bw.tr.tl. r.. . . Lh.ir ●mpl.. eex. m.ke. . . .W.rr. n!}: ?.pr..s.r ,mplied, .ar. . . . . . . . . . ICC.I Ii. billlt wrmpms!h, hl,r.r th. .cc. r..,. ..~pl. t...,..., .,d.l.”..r

. . . lmf., mm:, t,., .pp. r.lu.. prod.,+, or WK... dicclmed. orrc:c:,eql. th. ! it. . . . w..ld nti i. fri. ~e prI.. #eI%.wn~

ANALYTICAL METHODS FOR FISSIONABLEMATERIALS IN THE NUCLEAR FUEL CYCLE

July 1, 1975—September 30, 1976

Compiled by

Glenn R. Waterbury

ABSTRACT

Progress continued on development of dissolution techniques for difficult-to-dissolve nuclear materials, development of methods and automated in-struments for determinations of plutonium and uranium, preparation ofplutonium-containing materials for the Safeguards Analytical LaboratoryEvaluation (SALE) program, analysis of SALE uranium materials, andmeasurement of plutonium isotope half-lives. Gas-solid reactions at elevatedtemperatures using reactive gases such as chlorine continue to show promisefor separating uranium from refractory materials. An extensive study ofnonaqueous solvents for the dissolution of refractory materials is inprogress. An extraction-separation procedure, highly specific for microgram

~ =.amounts of uranium, has been developed, and its adaptation to the Los—.-=0 Alamos–Scientific Laboratory (LASL) automated spectrophotometer is being~m. evaluated. Development of an electrometric analysis method for plutonium;~$ . ..— .~= is nearing completion, and design of an automated instrument using the

~~m’eFi method has been started. Batches of plutonium oxide and mixed uranium-&—m

s= 0, plutonium, intended for issue as Secondary Reference and Calibration Test

j~:’” Materials, are being recharacterized for assay and isotopic contents. Them

8=m’ half-life of 2S9PUhas been determined by isotope-dilution mass-spectrometric“Scs’

zm;.measurement of 235Ugrow-in as a function of time.

-

—— ———— —— — —

1. INTRODUCTION

Adequate levels of safeguard control of uraniumand plutonium require confirmation by adequatesampling and analyses of inventories in the nuclearfuel cycle. Isotopic composition as well as the totalamounts present of these two actinides must bemeasured precisely in widely diverse types of nuclearmaterials, including highly pure products, reactorfuels having complex chemical compositions, and

many types of scrap materials. Dissolutions andanalyses of highly refractory alloys, oxides, andscrap materials containing uranium, plutonium, orboth are particularly difficult.

The objectives of this project are (1) the develop-ment of fast dissolution techniques and analyticalmethods for plutonium and uranium determination,with emphasis on difficult-to-dissolve materials, (2)the design and construction of automated apparatusfor plutonium and uranium determination, (3) the

1

preparation of well-characterized, plutonium-containing materials for use in the SafeguardsAnalytical Laboratory Evaluation (SALE) program,

(4) the preparation of well-characterized, highlypure plutonium metal standards for distribution bythe Nat ional Bureau of Standards (NBS), (5) par-ticipation in, and the preparation and characteriza-tion of the materials for, an inter-ERDA laboratoryprogram established to measure the half-lives of thelonger lived plutonium isotopes, and (6) thechemical characterization of special lots of nuclearmaterials as requested by the Energy Research andDevelopment Administration (ERDA).

II. DISSOLUTION OF FUEL CYCLEMATERIALS

A difficult and time-consuming operation inchemical analysis of nuclear fuel cycle materials istheir dissolution before assay or isotopic measure-ment. The rapid dissolution of fuel cycle materials or

solubilization of uranium and plutonium are beinginvestigated. Currently included in this investiga-tion are mineral acid reactions at high temperaturesin pressurized containers, reactions with reactivegases at high temperatures, and reactions with nona-queous acid and reactive organic solvents, and later,reactions with various media abetted by laser energyinput will be investigated.

A. Teflon-Container Metal-Shell Apparatus (S.F. Marsh, H. J. Kavanaugh, J. E. Rein)

A dissolution apparatus consisting of a Teflon con-tainer in a pressure-supporting metal shell permit-ted acid dissolutions at temperatures to 270”C andpressures to 320 atm (5 000 psi). The use of this ap-paratus is far-reaching, including uses for geologicalsamples, metals, and most other inorganicmaterials, and the nuclear industry from mining andmilling through nuclear fuel fabrication. The designof the LASL-developed apparatus was madeavailable to industry, and the Parr Instrument Com-pany now markets a slightly modified version.

In late fiscal year 1975, Parr-supplied Teflon con-

tainers experienced an unusually high number offailures, both at the New Brunswick Laboratory

(NBL) and at the Los Alamos Scientific Laboratory(LASL). Teflon containers fabricated at LASL weretested under conditions causing failures with Parrcontainers and shells. The LASL-fabricated con-tainers did not fail; in fact, the safety rupturediaphragm distorted only slightly, indicating low-to-moderate internal pressures.

We notified Parr Instrument Company of our testresults, along with speculation that the Teflon lotused to fabricate the Parr containers might havecontained reactive organic components or was un-usually porous. Porosity permits excessive transpira-tion of acid vapors at the operating temperature andconsequent corrosion and failure of the safety rup-ture diaphragm. Specifications and supplier of theTeflon in use at LASL were supplied to Parr. Todate, Parr has not responded.

B. Gas-Solid Reactions (D. D. Jackson, R. M.Hollen, J. E. Rein)

The investigation of gas-solid reactions with theobjective of converting refractory plutonium anduranium materials to more readily soluble com-pounds was continued, with emphasis on reactinguranium-containing materials with chlorine gas. Aspreviously reported,’ react ion of chlorine gas withuranium oxides (U02, UOa, and U~O~) and UC1 at1000”C for 12 h in a quartz-tube furnace producedcomplete volatilization of the uranium. The tech-nique has been applied successfully to scrapmaterial samples as well.

iNBL-supplied scrap materials known to containzirconium, probably as uranium-zirconium cermetfuel, were reacted with chlorine gas at temperaturesto 1000° C in the quartz-tube furnace. For a l-g sam-ple, 8-h reaction time, and 0.2 atm chlorine pressure,about 40 wtYOof the total sample and >95 wtYOofthe uranium volatilized. The quartz tube wasspecially fabricated to provide cent roll able at-mosphere and effective recovery of the volatilizedsample fraction by dissolution with HNOS.

Typical samples were subjected to this treatmentand the HN08 solutions were analyzed for uraniumusing the LASL automated spectrophotometer

prototype. For some samples, the organic phase wascloudy, contributing to uncertainty in the results.This cloudiness was probably caused by hydrolytic

2

precipitation of zirconium that dispersed into theorganic phase. To eliminate this effect, a two-stepfurnace treatment was investigated. The sample wasreacted with chlorine at 500” C for 4 h, the conden-sate was dissolved with HNOS, the residue was reac-ted at 1000° C for 8 h, and the condensate formedfrom this reaction was dissolved with HNO,. Thefirst reaction volatilized 11 wtY. of the sample andno significant amount of uranium. The second reac-tion increased the quantity of sample volatilized to40 wt% (based on initial sample weight) and gave 95wt% volatilization of the uranium. Significantly, theorganic phase remained clear during the uranium

extraction. Because the sublimation point of ZrC14is331 “C, the first reaction undoubtedly volatilized amajority of the zirconium.

A furnace designed to operate with the quartzreaction apparatus at temperatures to 1300” C is be-ing installed for continued investigation usingchlorine and other reactive gases. This furnace willbe in a glovebox so that applicability to plutoniummaterials also can be studied.

C. Reactions with Nonaqueous Solvents (W. D.Span, M. J. McLeod, S. F. Marsh, F. R. Roensch,J. E. Rein)

This year, an investigation was begun of nona-queous acid media and organic solvents that provideincreased acidity, greater oxidation-reductionpotentials, and unique high reactivities. Testmaterials were refractory components often presentin nuclear fuel cycle materials, including fused ox-ides of UOZ, ThOz, ZrOz, AlzO~, and SiOz, andfirebrick to simulate slag and similar components inscrap materials.

A 100-mg sample of a sieved 100-200 mesh testsample was reacted with a measured volume of thenonaqueous acid or organic solvent in a Teflon orborosilicate glass container fitted with a water-cooled cold finger. The mixture was heated for aselected time, usually 16 h, cooled, filtered, and theresidue weighed. Loss of weight was taken as ameasure of dissolution.

Nonaqueous acid media studied have been HI,HBr, HNO,, H,SO,, and Br, in glacial acetic acidcontaining a slight excess of acetic anhydride. Therewas no significant dissolution of the test materials,

except that UOZ completely dissolved in HN03 inless than 1 h at 50”C, giving a yellow solution ofUO~+ upon dilution with water. Because these mediaare considered to be oxidizing, and because all of themetals in the test series, with the exception ofuranium in UOZ, are in their highest oxidation state,the insolubility is not surprising. Over 50% dissolu-tion of an unground piece of zirconium-uranium car-bide in HNO,-glacial acetic acid indicated thepotential applicability of such oxidizing systems todissolution of carbide and nitride fuels.

Some refractory oxides are dissolved by reactionswith reducing agents, generally as high-temperaturefusions. Very strong reducing media are produced bydissolving alkali metals in low-molecular-weightamines, selected amides, and ethers. Such mediastudied to date include sodium and potassiummetals in ethylene diamineformamide, N, N-dimethylformamide, tetrahydrofuran, p-dioxane,triethylamine, diethylamine, isobutylamine, andacetonitrile. The generally low dissolution of testmaterials has been attributed to the low solubilities,and hence concentrations, of sodium and potassiumin these solvents. The solubilities of sodium andpotassium are higher in liquid ammonia,methylamine, and ethyiamine. Handling techniquesmust be developed for these low-boiling solventsbefore they can be evaluated. Other reducing media,including hydrazine hydrochloride in formamideand dimethylformamide, and sodium and potassiumnaphthalates in tetrahydrofuran, are beingevaluated. To be studied are hydroxylamine,dimethylphosphite, triphenylphosphite, andhydrazine in amine solvents.

Free radicals formed in the high-temperaturesealed-tube dissolution technique using mineralacids provide much of the reaction. Nonaqueous sol-

vents can generate free radicals without pressure-sealing to promote high temperature. Initial at-tempts using sulfuryl chloride alone, and mixed witht-butylperoxybenzoate as an initiator, have givenlimited success. Solvents generating more energeticfree radicals will require development of safe han-

dling techniques.In addition to the solvent types discussed above,

several organic solvents known to dissolve inorganicsalts are being tested, among them formamide, N, N-dimethylformamide, hexamethylphosphoroamide,

acetonitrile, dimethylsulfoxide, and dibutyl-phosphate. Dissolution of the refractory testmaterials was less than 20?4.except for dibutylphos-phate, which dissolved nearly all of UOZ and 15-607.of the others.

III. ANALYTICAL METHODS ANDAUTOMATED INSTRUMENTS FOR THEDETERMINATION OF PLUTONIUM ANDURANIUM

Past efforts have produced an automated instru-ment for the determination of milligram amounts of

both uranium and plutonium using a highly specificspectrophotometric method. Development of a moresensitive method for microgram amounts of uranium(and possibly plutonium) is nearing completion, andits adaptation to the automated spectrophotometerappears promising. An electrometric method for thedetermination of plutonium is being developed toserve as the basis for a high-precision automatedanalyzer.

A. Automated Spectrophotometer (D. D.Jackson, R. M. Hollen, J. E. Rein)

A report detailing all aspects of the instrumentfabricated for NBL has been issued.’ The main bodyof the report offers (1) information on chemistryaspects of determination of plutonium and uranium,including discussions of tolerances for diverse metal

ions, inorganic and organic anions, and acidity, (2) adescription of the instrument, component by compo-nent, and (3) data showing accuracy and precisioncapability. A series of appendixes gives (1)preoperational instrument adjustments and test, (2)operational procedures for the determination ofuranium and plutonium, (3) details of mechanicalcomponents, including engineering drawings, (4)details of electrical control circuitry, (5) microcom-puter readout system hardware, and (6) software forthe readout system.

As one of the final investigations for theautomated spectrophotometer, its precision wascompared to the precision of the uranium deter-mination procedure done manually by an experien-

ced analyst. The operations were very similar to

those described for the original method,’ except that2-nitropropane was used as the organic extractant,and the volumes of the salting-completing solutionand the organic extractant were changed to thosevolumes used for the automated instrument. Theanalyst prepared a uranium calibration solution us-ing a LASL-produced highly pure uranium metal,Four weight aliquots at uranium levels of 1,6, and 10mg were processed on each of seven alternate days togive a total of 84 measurements. The absorbancemeasurements were made using l-cm cells and aHeath EU-707 recording spectrophotometer. Resultswere computed using both the two-wavelengthcalculational mode (as used in the automated in-strument) and the more precise three-wavelengthcalculational mode. Second-order polynomials werefitted and regression lines were computed for the 84values. The averages of the average deviations were1.8, 0.42, and 0.55 rel% at the l-mg, 6-mg, and 10-mg levels for the three-wavelength calculation. Thecorresponding values for the two-wavelengthcalculation were 4.0, 0.48, and 0.50 rel%. The preci-sion of 0.9, 0.30, and 0.27 rel% that were obtainedwith the automated instrument, in which the two-wavelength measurement system is used, aresuperior to those obtained by manual operations bythe experienced analyst, especially at the l-mguranium level.

B. High-Sensitivity SpectrophotometricUranium Determination and Application toAutomated Spectrophotometer (S. F. Marsh, M.R. Ortiz, F. R. Roensch, R. M. Hollen, D. D.Jackson, J. E. Rein)

The procedure used in the automated spec-trophotometer has a lower limit of 1 mg of uranium.A major application of the instrument is the analysisof scrap samples, many of which are characterizedby low uranium contents and high contents of ex-traneous elements. Because the automated spec-trophotometer procedure is restricted to a maximumsample volume of 0.5 ml, it is necessary to evaporatelarger aliquots of these samples with low uraniumcontents. Here, the large quantities of extraneoussalts can interfere with uranium extraction and ab-sorbance measurement. The development of a high-sensitivity spectrophotometric method compatible

4

with the automated instrument is therefore the ob:jective of this task. The major compatibility require-ment is that all operations including the final absor-bance measurement are done in a glass sample tubewith no transfer of the sample.

1. Chemical Separation and MeasurementProcedure. Of various procedures evaluated, one of

the most promising is based on the extraction of theuranium -benzoyltrifluoroacetone complex intobutyl acetate.’ Initial experiments with this

procedure showed that the degree of extraction of thecomplex was below the 99~0 level that is consideredminimum to produce highly reliable measurements.Other solvents evaluated were sec-butyl acetate,benzyl acetate, toluene, butyl propionate, isobutylpropionate, isobutyl isobutyrate, benzyl-n-butyrate,pentyl hexanoate, isopentyl benzoate, and isopentylhexonoate. Of these, butyl propionate gave thehighest uranium extraction ( >99~0, based on ‘“Utracer experiments), as well as the highest absor-bance for the extracted complex. The molar ab-sorptivity was 14000.

The extraction of the uranium complex isrelatively insensitive to pH change over the pHrange 6 to 8. However, many other elements also ex-tract under this condition. The extraction of ex-traneous elements can be suppressed by adding achelon as a masking agent, ideally without suppress-ing the extraction of the uranium complex. Chelonsinvestigated were ethylenediaminetetraacetic acid(EDTA), diethylenetriaminepentaacetic acid(DTPA), hydroxyethylethy lenediaminetriaceticacid (HEDTA), and cyclohexanediaminetetraaceticacid (CDTA). As expected, CDTA was the most ef-fective. It forms particularly stable complexes withmost cations; however, the U03+ -CDTA complex isat least 100 times weaker than complexes formedwith the other chelons. The effect of CDTA onuranyl ion was further minimized by adding CDTAas equimolar Mg-CDTA. The stability constant ofMg-CDTA ia -105 higher than for UO\+-CDTA.Thus, CDTA remains tightly bound to M&+ in thepresence of UO~+ ions. Most potentially interferingcations, however, form CDTA complexes whosestability constants are >10° higher than for Mg-CDTA. In samples containing such cations, CDTAreleases from the Mg-CDTA complex to form thestronger complexes.

The initial anion interference studies, from achloride system, demonstrated that the followinganions did not interfere at a 1000:1 mole ratio touranium SO:-, SO:-, F-, NO;, NO;, CIOZ, and l-.Phosphate did not interfere at a 1:1 mole ratio, butdid at a 10:1 ratio.

Many cations are known to form highly extrac-table chloride complexes. For this reason, the systemwas modified from a chloride to a nitrate medium,and an extensive cation interference study was madeof all elements commonly present in nuclear scrapmaterials. For pH control, a highly buffered solutionof hexamethylenetetramine and triethanolaminewas incorporated into the system in anticipation ofautomated operation that would accept l-ml sam-ples ranging from 1A4HNO, to 8M HNO, with no pHadjustment. The effect of each cation was deter-mined at these extremes of 1 and 8 meq of acid andat mole ratios to uranium of 10, 100, and 1000 (or un-til interference was apparent). Of 42 cations studied,there were no effects for 23 at 1000:1 ratios, nine at100:1 ratios, and seven at 10:1 ratios. A few experi-ments remain for completion of the interferencestudy.

Development and evaluation of this procedure formanual operation is nearing completion. At thattime it will be published as a journal article ortopical report.

2. Adaptation to the AutomatedSpectrophotometer. The above procedure,developed for manual operation, uses inversion mix-ing for the extraction and a double-beam spec-trophotometer with a reagent blank in the referencecell for absorbance measurements. In the automatedspectrophotometer, phase mixing in the tube is at-tained by magnetic stirring, and a Teflon-coveredmagnet rotating at 1200 rpm proved effective for thenew procedure. The uranium-BFTA complex extrac-ted completely in 6 rein, the phases separated in lessthan 1 min to produce a clear organic phase, andspattering of droplets was insignificant.

Because the absorbance peak of the extracteduranium -BTFA complex is broad and is notseparated from the reagent absorbance, measure-ment of the complex absorbance with a single-beaminstrument like the automated spectrophotometer isdifficult. Absorbance by the tube’s glass walls also is

appreciable at the 380-nm wavelength of the com-plex’s absorbance peak, and the measured absor-bance must be corrected for variations arising fromdifferent wall thicknesses.

To investigate the conditions for measurement ina single-beam mode, a photomultiplier tube detectorand a chamber to hold the glass sample tubes wereconstructed and attached to the exit face of amonochromator that has variable wavelength andbandwidth control. Output from the photomultiplierdetector was measured with a picoammeter.

Samples containing up to 42 pg of uranium wereextracted in the automated spectrophotometer for 10min to assure complete extraction. Absorbancewere measured at 384.0 nm for the peak and at threewavelengths of 400.0, 415.0, and 425.0 nm for thebackground correction using bandwidths up to 0.4nm, the maximum attainable with themonochromator. Although 400 nm is on thedescending slope of the UO;+ -BTFA complex peak,its use gave best precision. That the precision im-proved slightly with increasing bandwidth is at-tributed to the higher signal-to-noise ratio producedby increasing incident light intensity.

Measurement precision was established for theabove instrument conditions at four uranium levelsin the range of 7 to 42 Kg of uranium. The relativestandard deviations, computed for six replicatemeasurements at each level, ranged from 0.3% at 42yg to 0.9% at 7 pg. Without the background correc-tion at 400 nm, precision was about tenfold worse.

C. Investigation of Electrotitrimetry for anAutomated Plutonium Analyzer (D. D. Jackson,R. M. Hollen, F. R. Roensch, J. E. Rein)

The objective of this task is to develop an elec-trometric method for the determination ofplutonium, to serve as the basis for an automated in-strument that features (1) high specificity, (2) preci-sion of 0.1-0.2% relative standard deviation, and (3)measurement of low-milligram levels. A versatilecoulometric apparatus consisting of commercialcomponents has been assembled around a PrincetonApplied Research Corporation 173D PotentiostatGalvonostat, a 179D Digital Coulometer, and aHewlett-Packard 9821 Programmable Calculator.Interfaced and under control of the calculator are a

scanner, a digital multimeter, a digital-to-analogconverter, and a plotter. A schematic of this ap-paratus is shown in Fig: 1.

The calculator controls titrimetric conditions suchas controlled-potential and controlled-current, andmonitors the electrolysis measurements of interest,such as current, voltage, coulomb, and time. Theelectrolysis data can be processed on-line, decisionscan be made, and conditions can be adjusted, andthe data can be stored on a magnetic tape cassettefor later analysis. The digital-to-analog converterpermits the calculator to select any desired potentiallevel when operating in the controlled-potentialmode or the current level when operating in

POTENT IOSTAT - GA LVANOSTAT

I CO ULOMETER

ITJ ;D- TO-ASCANNER

CONVERTER

/l\

rb’

DVMPROGRAMMABLE

- calculatorI.—

PLOTTER I

Fig. 1.Schematic of versatile electroanalytical ap-paratus for the investigation of automatedplutonium determinations.

6

constant-current mode. During all phases of an elec-trolysis, the on-line digital plotter can record avariable of interest, usually a plot of log10current vstime or electrode potential vs time.

This apparatus has given excellent performanceand will serve as the basis for the automated instru-ment. For example, 5-mg amounts of iron in 0.5A4H, SO, electrolyte were titrated with a precision of0.05% relative standard deviation using controlled-potential coulometry and an electrolysis cell con-sisting of a platinum-gauze working electrode, asaturated-calomel reference electrode, a Vycor coun-ter electrode, and a glass stirrer driven by a syn-chronous motor. Iron was reduced to Fez’ at 0.25 Vand then titrated to Fe8+ at 0.60 V. (All electrode

potentials in this report are relative to the saturatedcalomel electrode. ) Time for the oxidative titrationwas 6 min to a current end point of 30 pA.

The general system under consideration for thedetermination of plutonium is an initial reduction toPu’+, an oxidation in a medium in which PU8+is notoxidized but potentially interfering diverse ions areoxidized, addition of a completing agent to lower thePU3+-PU4+potential, and oxidative titration of Pug+to PU4+ using controlled-potential coulometry.

Aspects of the Davies and Townsend poten-tiometric methode are being considered as the basisfor a controlled-potential coulometric method. In theDavies and Townsend method, reduction and oxida-tion are done with chemicals, as follows. In amedium of hydrochloric acid-aluminum chloride-sulfamic acid, excess cuprous chloride reducesplutonium to Pug+, a titration with potassiumbichromate to a potentiometric end point oxidizesexcess Cul+ and any Fez+ without oxidizing PU8+ toPU4+, sulfuric and phosphoric acids are added tolower the PUS+-PU4’ potential, and the Pus+ istitrated to PU4+ with potassium bichromate to asecond potentiometric end point.

Because iron is the most troublesome diverse ionin electrotitrimetric determinations of plutonium, itis being investigated along with plutonium to es-tablish procedure conditions.

The redox potentials of the Fe’+ -Fe’+ and thePU8+-Pu4+ couples in the 1.111 AIC1,-1.1A4 HCl-0. 15Jkfsulfamic acid electrolyte of Davies and Town-send were measured as 0.41 V and 0.67 V for a poten-tial difference of 0.26 V. A greater potential dif-ference was desired to decrease the small quantity of

iron not oxidized at the first end point that then ox-idized with and biased the plutonium result. To es-tablish whether these half-cell potentials were suf-ficiently separated to permit the unbiased deter-mination of plutonium, the oxidation levels of Fez+and Pue+ were determined at various potentials. Foriron, 98.0?4 oxidized at 0.49 V, 98.8% at 0.50 V,99.2% at 0.51 V, and 99.5% at 0.52 V. Theplutonium oxidations were 0.20% at 0.53 V, 0.32% at0.54 V, and 0.58% at 0.55 V. Thus, at a potential of0.50-0.51 V, 99% of the iron oxidizes with no signifi-cant oxidation of plutonium. A correction for thesmall amount of iron that would not have been ox-idized at 0.50-0.51 V and that would then oxidizewith the plutonium could be computed, based on thecoulombs measured at 0.50-0.51 V. However, the ac-curacy of this correction will depend on the relativequantities of other diverse ions that oxidize at thefiist end point..

The precision for the determination of 5 mg of ironin the AICl~-HCl-sulfamic acid electrolyte, usingreduction and oxidation potentials of 0.25 V and0.60 V, was 0.46% relative standard deviation, con-siderably poorer than that obtained in 0.5A4 HZS04electrolyte. Also, electrolysis times were 2 minlonger, requiring 8 min for the oxidation.

Hydrochloric acid, lithium chloride, and mixturesof the two were evaluated as electrolytes. Advan-tages, compared to the HC1-AICl,-sulfamic acidelectrolyte, would be high purity and an ability toobtain high chloride concentration levels coupledwith low viscosity. A disadvantage appears to be theinability to complex fluoride, expected to be presentin many samples. This could be overcome by using afluoride complexant such as H,BO,.

The half-cell potentials of the Fe’+ -Fe’+ and Pug+-Pu’+ couples in various combinations ofhydrochloric acid and lithium chloride are given inTable I. These potentials were measured by sweep-ing the electrode potential rapidly over the region ofinterest while measuring the instantaneous current.Scans were made with increasing and decreasingelectrode potential, starting, respectively, with thereduced and oxidized species in the electrolyte. Theintersection of the two curves on a plot of electrodepotential vs instantaneous current yields the half-cell potential. These half-cell potentials were con-firmed by controlled-potential coulometric deter-minations.

According to Table I, the half-cell potentials forboth the Fez+-Fe’+ and the PUS+-Pu4+ couplesdecrease with increasing chloride concentration.This is expected because the chloride complexes ofthe higher oxidation states are stronger than thecomplexes of the reduced states. At a constant totalchloride level, the half-cell potential for irondecreases as the HC1/LiCl ratio increases, whereasthere is little effect on the half-cell potential forplutonium. Thus, the electrolyte preferred to obtainlargest differences in the iron and plutonium half-cell potentials is hydrochloric acid without lithiumchloride.

The half-cell potentials of over 0.6 V for the Pu’ +-PU4+ couple require greater than 0.8 V for completeoxidation of PU8+ to PU4+. At this high potential,hydrochloric acid undergoes electrolysis and thelikelihood increases that diverse ions will oxidize andinterfere. The effect of sulfate, selected as a com-plexant to lower the half-cell potential of the Pu*+-PU4+ couple, was determined for the hydrochloricacid-lithium chloride electrolyte combinations

(Table I). The decrease was only about 0.1 V for upto 2A4 H,SO,, which is not adequate (discussedlater). There was even less effect upon the half-cellpotential of the Fe’+ -Fe’+ couple, and the previouslyoxidized Fes+ remained stable,

TA13L131

HALF-CIH,L POT13NI’IALS 01: Fe” -FE8+AND PUS4-}’U’+ CO[lPLllSIN HYDROC1lLOR!C ACllj-L!’l’Hl[llK CllLOI{llJE-SUI.l~ UItlC

ACID llLI.?CrROLYTI?S”

Hnlf-(kll Potmtids Without 11.S0, Half-Cell Potentials After Addition of 11.,S0,

IW2ctrolytc

2.5MLiC1

1.5A4 LiC1-lM HC1

2.5A4 HCl

5 M LiCl

4 M LiC1-lM HC1

3 M LiC1-2A4 HCl

5MHC1

4.5M LiC1-3M HC1

7.5M HC1

———_

Pus+ -PU4 ‘

0.47

0.44

0.42

0.42

0.41

().3.9

0.37

0.35

0.31

0.68

0.67

O.oa

0.65

0.69

0.69

0.67

0.65

0.63

])iffercncc

0.21

0.23

0.26.

0.23

0.28

0.30

0.30

0.30

0.32

MolmityH,so,b

123.4

12

12

1

1

12

123.4

12

123.4

Fd’-Fd ‘

0.430.420.38

0.420.41

0.410.40

0.40

0.39

0.380.38

0.360.350.34

0.340.32

0..310.290.28

“Potentials vs SCE. Mcmured with 2M H, SO, in reference compmtmcnt; therefore, includes o liquid junction potential.

PU’+-I)U’+

0.530.510.50

0.53Q.5~

0.540.52

0.51

0.53

0.540.52

0.5(30.530.50

0.540.51

0.570.510.50

‘Sulfuric acid was added to the electrolyteto give the indicated molarity.

8

The half-cell potentials of the two couples inhydrochloric acid media were determined in detail toestablish the concentration level at which the dif-ference between the two half-cell potentials isgreatest. This level would be used as the electrolytein conju~ction with a complexant (other than sul-fate) for plutonium. As shown in Table II, the dif-ference of 0.32 V is essentially maximum and cons-tant over the range from about 5 to 7.4M HCI, andboth half-cell potentials decrease with increasinghydrochloric acid concentration. At a 0.32-V dif-ference, the Nernst equation predicts that 99.6% of

the iron can be oxidized with oxidation of less than0.1 % of the plutonium. This is considered satisfac-tory. The apparent advantage of using a highhydrochloric acid concentration to attain lower ox-idation potentials is more than offset by an increasedbackground current and an increased measurementimprecision caused by electrolysis of the hydro-chloric acid. Based on the above, 5.5M HC1 wasselected as the electrolyte.

Complete (>99.9%) oxidation of Pu’+ in 5.5MHC1 would require a potential of almost 0.87 Vwhich is not feasible because hydrochloric acid ox-idizes and more diverse ions may interfere.Therefore, another complexant, phosphate ion, wasinvestigated to decrease the PUS+-Pu4+ half-cellpotential. The measured potentials as a function of

phosphate molarity in 5.5M HCI (Table III) showthat in lM phosphate, the Pus+ -Pu4+ half-cell poten-tial is lowered to almost 0.50 V, and greater than99.9% of the PU8+is calculated to be oxidized to PU4+at 0.68 V. At this potential, the oxidation ofhydrochloric acid is negligible. The Fe’+ -Fea+ half-cell potential was little affected by phosphate,decreasing from 0.36 Vat 0.56M phosphate to 0.34 Vat 1.33M phosphate. Therefore, the previously ox-idized iron remains oxidized upon addition of

phosphate.Two sources of phosphate, HaPO. and NaH,PO,,

were investigated. Neither contained measurablequantities of impurities that oxidized or reduced in5.5M HC1 over the 0.2-0.7 V potential range.Phosphoric acid, which can be added as a more con-centrated solution than NaHzP04, was investigatedfirst. After several months of successful use, unex-plained erratic potentials and increased electrolysistimes and background currents began occurring. Useof NaHzP04 for long periods of time had none ofthese adverse effects, and all subsequent data wereobtained with it. .

With 5.5A4 HC1 established as the electrolyte, andwith NaHzP04 added to lower the Pus+ -Pu’+ couple,detailed experiments were conducted to establish“optimum” potentials for the preliminary oxidationof iron (and presumably other diverse ions) and the

TABLE II

MORE DETAILED MEASUREMENTS OF THE HALF-CELLPOTENTIALS OF Fe2+-Fe3+AND PUS+-Pu’+ COUPLES IN HYDROCHLORIC ACIDa

HC1 corm Half-Cell Potentialsb

(M) Jj’&+-Feg+ PU9+-PU’+ Difference

2.5 0.446 * 0.0026 0.705 ● 0.0080 0.259 * 0.00844.7 0.397 * 0.0015 0.701 + o 0.304 + 0.00155.5 0.372 * 0.0021 0.686 i 0.0013 0.314 * 0.00255.8 0.371 + 0.0017 0.686 + 0.0015 0.315 * 0.00236.6 0.343 i 0.0021 0.662 ● 0.0010 0.319 A 0.00237.0 0.332 * 0.0015 0.649 * 0.0006 0.317 & 0.00167.4 0.323 * 0.0010 0.640 * 0.0017 0.317 + 0.0020

“Potentials w SCE.

The associated uncertainty is the standard deviation for a single measurement, calculatedfrom n = 3 measurements.

Pu~+-Pu’+IN 5.5M

TABLE III

HALF-CELL POTENTIALSHC1-PHOSPHATE MEDIA

Phosphate Potential(M) (V vs SCE)

0.56 0.540.83 0.521.08 0.501.33 0.48

titrimetric oxidation of plutonium. Variablesevaluated included the preliminary oxidation poten-tial, the end-point current, and the effect of time forboth the preliminary and titrimetric oxidations.

In general, experimental conditions were asfollows: equal-mole quantities of 1 mg iron and 5 mgplutonium, 10-ml volume of 5.5A4 HC1, nitrogen flowthrough the cell for 10 min before and then duringelectrolysis, preliminary reduction at 0.25 V to anend-point current of 50 KA, preliminary oxidation ata selected potential to a selected end point (currentand/or time), addition of 2.5 ml 5A4 NaHZPO,, andoxidative titration at 0.68 V to a selected end point(current and/or time).

To attain high precision and accurate coulometrictitration, proper background correction is impor-tant. In these experiments, background correctioncoulombs are established at both potentials by elec-trolysis of the electrolyte alone for times equal tothose used for the iron and plutonium oxidations.

The per cent oxidations of iron and plutonium atpreliminary potentials of 0.54, 0.55, and 0.56 V, andthe oxidative titration at 0.68 V are given in TableIV. The end point current for all titrations was 50PA.

Table IV gives average results for four to eightmeasurements. Measurement precision weretypically 0.08% relative standard deviation for theinitial iron oxidation and O.15% relative standarddeviation for the plutonium oxidation. For both ox-idations, logl~ current decreased linearly with time,in agreement with theory.

Mixtures of 1 mg iron and 5 mg plutonium wereelectrolyzed using the same conditions as described

TABLE IV

PER CENT OXIDATIONS OF IRONAND PLUTONIUM AT THE PRELIMINARY

AND OXIDATIVE TITRATION POTENTIALS

Preliminary Per CentPotential Oxidized

(v) Fe Pu

0.54 99.89 <0.010.55 99.93 0.0120.56 99.97 0.037

Per CentOxidizedat 0.68 V

Fe Pu

0.11 100.0.07 99.9&0.03 99.96,

above. Electrolysis times and measurement preci-sion were the same as for the separate iron elec-trolysis at the three (0.54, 0.55, and 0.56 V)preliminary potentials and the plutonium elec-trolysis at 0.68 V. Also, the sum of coulombs for thetwo electrolysis did not change. There was a 0.3%increase in the coulombs at the preliminary iron ox-idation, with a corresponding decrease at theplutonium oxidative titration. This was experimen-tally established to be a consequence of the longerelectrolysis time required at the preliminary oxida-tion to reach the 50-PA end point current for iron andplutonium mixtures compared to plutonium alone.During this time, more plutonium oxidized. This ef-

fect will be a function of the iron-to-plutonium ratioor of any other oxidized diverse-ion-to-plutoniumratio.

Decreasing this effect will lead to better accuracyand precision for the plutonium measurement.Procedural changes being investigated are a higherpotential for the preliminary oxidation andmodifications in end point establishment. Followingthis investigation, effects of diverse ions charac-teristic of nuclear fuel cycle materials will beestablished.

A recurring problem of controlled potential

coulometry using solid electrodes is instability of theworking electrode, resulting in increased backgroundcurrents and electrolysis times. The following treat-ment for the platinum gauze electrode restores itslow background current and reproducible perfor-mance. Immerse the electrode in hot 16A4 HNOS for

10

1 h, rinse thoroughly with water, immerse in 12MHCI for >2 h, and cycle through the reduction andtwo oxidations described above using 5.51kf HC1 un-til the background current stabilizes. Usually, onlytwo cycles are required. The electrode is stored in5.5A4 HC1.

Design of the first automated instrument usingelectrometric titration is under way. In addition toestablishing the basic electroanalytical system (seeFig. 1), mechanical components are being tested.These include reagent dispensers and cell-rinsingsystems.

IV. PREPARATION OF PLUTONIUM-CONTAINING MATERIALS FOR THE SALEPROGRAM (S. F. Marsh, J. E. Rein, G. R.Waterbury)

The major objectives of the Safeguards AnalyticalLaboratory Evaluation (SALE) program are (1) toupgrade the capabilities of US and foreignlaboratories to analyze for assay and isotopicuranium and plutonium contents of nuclear fuel cy-cle materials, and (2) to provide uranium andplutonium Reference and Calibration Test Materials(RCTMS).

In 1973, vials containing a PU02 powder lot and a

(U, PU)O, powder lot were supplied to AlliedChemical Corporation, Idaho National EngineeringLaboratory, for distribution as Secondary RCTMS(SRCTMS). When these two lots were characterizedfor their assay and isotopic values by LASL andchecked by other ERDA contractor laboratories,guidelines had not been formulated for such charac-terizations. Since then, NRC guides have beenprepared for the preparation and characterization of

Working Calibration and Test Materials (WCTMS).It is proposed that the two lots be recharacterized byat least the guidelines for WCTMS by at least twolaboratories. A plan is being formulated with NBL,the laboratory responsible for the SALE program, ef-fective October 1, 1976.

A SALE program meeting, June 29-30, 1976, atNBL, was attended by more than 60 people frommost of the US and foreign laboratories participatingin the program. Major recommendations were to (1)expand the program to provide greater internationalcoverage, and (2) maintain the voluntary aspects of

the program, rather than use any generated informa-tion for regulatory application, to ensure the objec-tive of technology upgrading. Five talks were presen-ted by LASL attendees (see the Publications andTalks section).

A report’ has been prepared that providesrelationships for computing the decreasingplutonium content and changing isotopic distribu-tion of plutonium materials for the radioactive decayof 2SSPU,2aePu,240Pu,and 242Puto uranium daughtersand of 241Puto “lAm. This computation is importantto the use of plutonium reference materials forcalibrating destructive and nondestructive methodsfor assay and isotopic measurements, as well as toaccount ability inventory calculations. Correctionsfor the 24’Pu decay are usually made; however, theeffects of the radioactive decay of the otherplutonium isotopes can exceed that of 24’Pu decay,

depending on the plutonium isotopic distribution ofthe material.

V. ANALYSES OF SALE URANIUMMATERIALS (A. D. Hues, W. H. Ashley, R. M.Abernathy, and J. E. Rein)

Analyses of SALE bimonthly samples of uraniumnitrate and uranium oxide were continued to deter-mine uranium content and isotopic distribution.There were no significant biases, and the measure-ment relative standard deviations were smaller thanthe averages for the reporting laboratories. A LASL-modified NBL-Davies Gray methodi is used for theuranium determinations, and thermal ionizationmass spectrometry is used for measuring isotopicdistributions.

VI. PLUTONIUM ISOTOPE HALF-LIFEMEASUREMENTS (S. F. Marsh, M. R. Ortiz, J.W. Dahlby, D. C. Croley, S. Kosiewicz, R. M.Abernathy, G. L. Tietjen, R. K. Zeigler, J. E.Rein, G. R. Waterbury)

As a part of an inter-ERDA laboratory program tomeasure accurately the half-lives of longer livedplutonium isotopes, LASL is (1) preparing, exten-sively characterizing, and distributing high-puritybatches of plutonium metal from specially provided,

11

enriched isotope materials, and (2) determining thehalf-lives using the technique of isotope dilutionmass spectrometry for measuring the produceddaughter isotope. The LASL characterizationmeasurements include assay, isotopic distribution,metal impurities including other transuranics, andnonmetal impurities.

The plutonium isotopes in the program are 2“Pu,ZWpu,andMOPU, and perhaps 241Pu. An initial batch

of 2SoPu metal (99.26Y0 ‘WPU isotopic abundance),designated UDPU.l, was prepmed, characterized,

sent to the Mound Laboratory for calorimetricmeasurement, and then returned. Accuratelyweighed portions of the metal were sealed inborosilicate glass tubes under a dry argon cover gasand distributed to laboratories.

A. Plutonium-239

Plutonium oxide of >99% 2SePuwas converted tohigh-purity metal at LASL for distribution to allparticipating laboratories. The chemical purity ofthis 2S’PUmetal was established by measuring totalplutonium assay using potentiometric titration, aswell as by measuring 44 impurity elements at LASLand the Rockwell International Rocky Flats Plantusing emission spectrography, thermal ionizationmass spectrometry, spark source mass spectrometry,radiochemistry, spectrophotometry, inert gas fusion,and combustion chemical analysis techniques.Plutonium isotope distribution was measured by thesame two laboratories using thermal ionization massspectrometry. The individual portions of plutoniummetal used in the program were weighed on the samebalance used to weigh NBS SRM-949 plutoniummetal,

The experimental program’ formulated at LASLto attain at least 1% relative accuracy in measuringthe half-life of 2gePu-1 involves the following opera-tions. The grown-in and impurity uranium wasseparated from six samples by ion exchange, ameasured quantity of ‘8U was added as the internal

standard, and the growth rates of daughter ‘SW arebeing measured at selected time periods by massspectrometric determinations of the 2s5UpsW ratio.The 2S’PU half-life will be calculated from the fun-damental decay equation

– (dnldt) 239 + (dn/dt) 235 = NA ,

in which

– (dn/dt) 239 = loss of ‘“PU atoms

per unit time, and

+ (dn/dt) 235 = growth of the ‘SW

daughter per unit time (Ref. 1).

In this equation, N is the number of 2gePuatoms andA is the 2S’PUdecay constant. The formation rate ofZ95U is measured bY isotope dilution mass spec-

trometry using an accurately measured initial addi-tion of ‘SSU as the internal standard.

To attain the stated accuracy goals, the variousoperations must be accomplished within an ac-curacy of 0.01 relOAfor weighing each of the six 2S0Pu-1 samples, 0.05 for the 2SoPucontent of the six solu-tions, 0.1 for the quantity of ‘W added, and 0.1 forthe ‘5UPS8U ratio measurements.i The ‘W spikesolution concentration was established by isotopedilution mass spectrometry using four separateuranium standard solutions, two prepared fromNBS SRM-960 (natural uranium metal) and twoprepared from high-purity 93% ‘W metal. Mixturesof each calibration solution and the 288U wereanalyzed in triplicate. The quantity of 2SEUadded tofour of the dissolved ‘g’Pu-l metal solutions wasselected to equal the quantity of 2*W grown-in dur-ing 1.5 yr. The quantity of 2WUadded to the remain-ing two solutions will provide equal amounts of ‘Wand 298Uafter 4.5 yr. Mass spectrometric measure-

ment of isotope ratios is most accurate and precise at1:1 ratios. Weighed portions of the calibrated ‘W

solution were added to each of the six 2S9PUsolutionsto produce solutions having 2WU~S’Pu ratios knownto *0.022Y0 relative standard deviation.

Four of the six ‘g’Pu-l solutions were measured inquadruplicate at five different times over 20 months.Plutonium half-life values were calculated for eachaverage quadruplicate measurement. The remainingtwo ‘g’Pu-l solutions were designed to provide moreaccurate half-life measurements after 4.5 yr of ‘Wdaughter growth. The first scheduled measurementof these solutions, after 1.5 yr, has yielded “PU half-life values in good agreement with the values ob-tained from the other four solutions.

The half-life data for the first four solutionscurrently are undergoing detailed statistical

12

analysis. The preliminary 2goPu half-life value is24150 yr with an overall relative standard deviati~nof about o.lh~o. The stated goal of the mass spec-trometric measurement technique has thus beenachieved and is expected to be further improvedwhen the 4.5-yr measurements are complete. Thisvalue is l% lower than the currently accepted valueof 24400 yr. An error this large represents a signifi-cant bias in all nondestructive analytical techniquesthat use the 280Puhalf-life. A l% bias in the 2SWPUcontent of the many tons of material assayed non-destructively represents a large amount of fissilematerial. More accurate half-life values will con-tribute significantly to the safeguarding of fissilematerial,

B. Plutonium-238

Plutonium metal of -90% 298Pu isotopic abun-dance has been prepared at LASL, weighed usingspecially developed, thermal-isolation weighingtechniques, and sent to the Mound Laboratory forweight verification and calorimetric measurements.The 2’8Pu metal was handled only in argon-atmosphere gloveboxes and sealed in an inner-welded tantalum capsule and two outer-welded stai-nless steel capsules for shipment. Nonetheless, the2’8Pu metal has apparently reacted with oxygen andperhaps nitrogen to cause the metal to be unusable.The metal will be repurified to remove the oxygenand nitrogen, and then will be resampled.

PUBLICATIONS AND TALKS

1. D. D. Jackson, D. J. Hodgkins, R. M. Hollen,and J. E. Rein, “Automated Spectrophotometer forPlutonium and Uranium Determination, ” LosAlamos Scientific Laboratory report LA-6091(February 1976).

2. D. D. Jackson, S. F. Marsh, J. E. Rein, and G.R. Waterbury, “Recent Developments in the Dis-solution and Automated Analysis of Plutonium andUranium for Safeguards Measurements, ” Inter-national Atomic Energy Symposium, Safeguardingof Nuclear Materials, Vienna, Austria, October 20-24, 1975.

3. S. F. Marsh, W. D. Span, R. M. Abernathy,and J. E. Rein, “Uranium Daughter Growth MustNot Be Neglected When Adjusting PlutoniumMaterials for Assay and Isotopic Contents, ” LosAlamos Scientific Laboratory report LA-6444(November 1976).

4. Talks presented at SALE Program Meeting,New Brunswick Laboratory, June 29-30, 1976.

a. J. E. Rein, “Preparation and Characterizationof Plutonium Materials. ”

b. G. R. Waterbury, “Uranium Methodology-LASL Experience with the NBL Titration. ”

c. G. R. Waterbury, “Plutonium Methods Used atLASL. “

d. J. E. Rein, “Automated Spectrophotometer forUranium and Plutonium. ”

e. J. E. Rein, “Isotope Measurements at LASL. ”

REFERENCES

1. G. R, Waterbury, Comp., “Analytical Methodsfor Fissionable Materials in the Nuclear Fuel Cycle,June 1974 to June 1975, ” Los Alamos ScientificLaboratory report LA-6045-SR (October 1975).

2. D, D. Jackson, D. J. Hodgkins, R. M. Hollen,and J. E. Rein, “Automated Spectrophotometer forPlutonium and Uranium Determination, ” LosAlamos Scientific Laboratory report LA-6091(February 1976).

3. W. J. Maeck, M. E. Kussy, G. L. Booman, andJ. E. Rein, “Spectrophotometric ExtractionMethods Specific for Uranium, ” Anal. Chem. 31,1130 (1959).

4. T. Shigematsu, M. Tabushi, and M. Matsui,“The Solvent Extraction and SpectrophotometricDetermination of Uranium Using Ben-zoyltrifluoroacetone as the Chelating Agent, ” Bull.Chem. Sot. Jpn. 37, 133-136 (1964).

15. W. Davies and M. Townsend, “AnalyticalMethod for the Titrimetric Determination ofPlutonium Using Cuprous Chloride as Reluctant, ”United Kingdom Atomic Energy Authority reportTRG-2463 (1974).

13

6. S. F. Marsh, W. D. Span, R. M. Abernathy, Materials for Assay and Isotopic Contents, ” Los

and J. E. Rein, “Uranium Daughter Growth Must Alamos Scientific Laboratory report LA-6444

Not Be Neglected When Adjusting Plutonium (November 1976).

* U.S. GOVERNMENT PRINTING OFFICE 1976–777-018/42

14