- Email: [email protected]
Determination of uranium in minerals and rocks
Talonta. Vol. 23, pp 283-288
Pergamon
Press, 1976. Printed IIT Great Britam
DETERMINATION
OF URANIUM
IN MINERALS
AND ROCKS
J. KORKISCH and H. HiisNw Institute for Analytical Chemistry, Analysis of Nucbar Raw Materials Division, University of Vienna, W~hringerstra~ 38, A-1090 Vienna, Austria (Received
_^__.
i6 .iune i975. Accepreti 18 October IY/5j
Summary-A method is described for the determination of uranium in minerals and rocks by spectrophotometry and fluorimetry. After treatment of the sample with hydrochloric acid, uranium is separated from matrix elements by adsorption on a column of the strongly basic anion-exchange resin Dowex 1 X8 from an organic solvent system consisting of IBMK, tetrahydrofuran and 12M hydrochloric acid (1:&l v/v). Following removal of iron, molybdenum and co-adsorbed elements by washing first with the organic solvent system and then with 6M hydrochloric acid, the uranium is eluted with 1M hydrochloric acid. In the eluate, uranium is determined by means of the spectrophotometric arsenazo III method or fluorimetrically. The suitability of the method for the determination of both trace and larger amounts of uranium was tested by analysing numerous geochemical reference samples with uranium contents in the range lo-‘-104ppm. In practically all cases very good agreement of results was obtained.
air-dried and 3 g of it were slurried in a few ml of the IBMK-THF-HCl mixture (see below) and transferred to an ion-exchange column, with the same mixture used as a rinse. Standard uranium solutions. Aliquots of a stock solution containing 10 mg of uranium (as uranyl chloride) per ml of 12M hydrochloric acid were diluted with water to obtain standard solutions in 9M, 6M and 1M hydrochloric acid containing @l-lo4 ppm of uranium. I~~~-~~~-~C~ mixture. A mixture of IBMK (isobutyl methyl ketone), THF (tetr~ydrof~~) and 12M hydrochloric acid (1:8: 1 v/v) was prepared several hours before use (in order to prevent the formation of air-bubbles in the resin bed). This solution can be stored for at least a week without loss of effectiveness. Arsenazo III solution. Freshly prepared and filtered 0.1% aqueous solution. Other reagents. “Fluorhase” (200 mg) pellets consisting of 95% sodium fluoride and 5% lithium fluoride (Nordrhein-Chemie, 41 Duisburg 12, West Germany), zinc metal (dust or finely granulated zinc), oxalic acid 1.0, 6, 9 and 12M hydrochloric acid, 1M nitric acid, and reagent-grade methanol
Spectrophotometric and fluorimetric methods are employed extensively for the determination of uranium in a variety of geological samples, including materials of very low uranium contents such as basalt-
ic rocks, and of uranium-rich minerals, as for instance torbernite, carnotite, uraninite and pitchblende.’ For the successful analysis of both types of ~~iurn-~~~g materials the very sensitive spectrophotometric method of ur~ium detestation can be used, which is based on measurement of the absorbance of the green ur~i~(~~ar~n~o III complex in moderately concentrated hydrochloric acid medium.’ Interferences are caused by thorium, zirconium, titanium, rare earth elements, molybdenum, and iron, so it is necessary to separate uranium from these elements and also from other components such as the main constituents of minerals and rocks. For this purpose, methods based on liquid-liquid extraction and ion-exchange have found widespread application.’ In the present paper an application is described of a single separation step which is based on anionexchange in an organic solvent system. Following this quantitative separation, uranium can be determined free from interferences, by either the arsenazo III method or fluorimetrically.
EXPERIMENTAL
Reagents ion-exchanger. The strongly basic anion-exchanger Dowex 1 X8 (loo-Zoo mesh; chloride form) was used. To purify the exchanger (which contained, e.a., iron and zinc) 200 g of the resin were treated in succes~on with 2 htres of fM nitric acid, 2 htres of 6M hydr~hloric acid, 2 litres of 1M hydrochloric acid, 3 litres of distilled water and 1 litre of reagent grade methanol. This batch of resin was
Apparatus The
ion-exchange
~parations
of uranium
were per-
formed in columns of the type and dimensions described earlier.’ Determination of distribution coefficients
The distribution coefficients (&-values) of uranium and some of the more strongly adsorbed metal ions were determined by using the batch-equilibrium method, and the column technique was employed to measure the &-values of iron, molybdenum and other elements which are weakly retained by the resin.3 The usually high distribution coefficients of uranium were determined after equilibration (for 4 hr) of 1 g of the resin with 20 ml of the organic solventhydrochloric acid mixture tested (see Tabte 1) containing 20 mg of uranium, followed by filtration, etc3 In the determinations of the batch and column ~stribution coefficients of the other elements only 1 mg each of the meta ions was used (see Table 2).
283
284
J.
KORKISCH
Procedures D~sso~~~~o~ ~~su~~~es. To 1.0 g of the thoroughly homogenized sample, in a 2%ml beaker, 100 ml of concentrated hydrochloric acid are added, the beaker is covered with a watch-glass, and the mixture is heated on a sand-bath until its volume has been reduced to about 30 ml. Subsequently the solution is evaporated to dryness under an infrared lamp, another 100 ml of concentrated hydrochloric acid are added and the solution is evaporated to dryness again under the heat-lamp. To the residue 30 ml RF“l.l nnn l,.,,&.,Y.l.,rl,r;r nr;rl (IIS, nr* (LUUL,~) o&i-A lL‘lP l.a .LLII*Lu‘G ..&+.....1s :.. ,,ea&Xl L-,.+-A “, L’JuLw.,~II”IIrLIc.lll under the lamp for several minutes and then let stand (preferably overnight). Afterwards, insoluble material (mainly silica) is removed by filtering off on a dense filter, 6h;I hy~o~hlo~~ acid being used for washing the filter naner and residue free from -iron. The filtrate -6 evaporated- to dryness on a steam-bath and the residue is taken up in about 30 ml of the IBMK-THF-HCl mixture. If two phases are formed, more solvent mixture is added until a homogeneous solution is obtained* in which however, a portion of the solute (mainly silica) usually reaDDears as a suspension, owing to the- decrease of sblubihcy on addition of the IBMK-THE-HCl mixture. This insoluble material is filtered off on a dense filter, and washed with IBMK-THF-HCl solution. The filtrate (usually 5GlOO ml) is the sorption solution for the ion-exchange separation. Ion-exchange separation. The sorption solution is passed through the ion-exchange column containing 3 g of the resin (pretreated with 10 ml of the IBMK-THF-HCL mixture), at a flow-rate of about 08 mlimin. Iron and molvbdenum are removed by washing the column with 50*ml of the IBMK-THF-HCl mixture. and Conner. cobalt. etc. as well as organic solvents left in the resin bed are eluted with 100 ml of 6M hydrochloric acid. The adsorbed uranium is then eluted with 100 ml of 1M hydrochloric acid. The resin column may be used for the isolation of uranium from further samples provided that 30 ml of the IBMK-THF-HCl mixture are used to pretreat it following the eiution of uranium. Quantitative determination of uranium. The uranium eluate is evaporated to drvness on a steam-bath or under an infrared Iamp and the residue is dissolved in 9M hydrochloric acid and transferred to a 25-m] standard flask and diluted to volume with the acid. If this solution contains > 1 ppm of uranium (i.e., > 25 pg of uranium) it is necessary to dilute it (or an aliquot) with the same acid to a suitable volume in which this concentration is not exceeded A. Spectrophotometric method. Ten ml of the 9M hydrochloric acid solution of the uranium are transferred to a lOO-ml wide-neck Erlenmeyer flask, 0.3 g of oxalic acid and 1.10 g of zinc are added and the flask is covered loosely with a y.vYY”. QtnnntV. -l-m&lo . . “..b th_e reduction. the flask is shaken carefully until all the zinc has dissolved. Immediately afterwards 1.0 ml of the arsenazo III solution is
added and the absorbance is measured at 665 nm against a reagent blank prepared in the same way. A calibration curve for the range I-1Opg of uranium is constructed .according to the reduction procedure described above. Beer’s law is strictly obeyed up to 10 ng of uranium jabsorbance @400). Higher amounts give strong positive deviations from linearity. The absorbance remains constant for at least 100 min. If less than 0.2 pg of uranium is present in the 10 ml of test solution (<0.5 ppm in the original sample) this method is less reliable (because of the lo\u * If after addition of the organic solvent mixture a total volume of 100 ml should be reached without obtaining homogeneity, 1 ml of concentrated hydrochloric acid is added-this causes immediate disappearance of the two phases.
and H. HCJBNER absorbance) than the fluorime~ic procedure described below. B. Fluorimetric method. A suitable volume (e.g., 0.1 ml) of the uranium eluate or of the 25 ml of the 9M hydru chloric acid solution containing the uranium is evaporated in a small platinum dish“ and after addition of a “Fluorbase” pellet, a melt is prepared under strictly controlled conditions.’ The fluorescence intensity of the cold flux is measured and compared with the intensity of fluxes of known uranium concentrations.
RESULTS AND DISCUSSION The dissolution procedure described, in which only hydr~hloric acid is used, proved to be highly satis-
factory for all of the samples analysed although it is in principle based on leaching only. Investigation of the completeness of dissolution showed that the portions insoluble in hydrochloric acid never retained more than about 0.1% of the total uranium content. During filtration of the IBMK-THF-HCl mixture to remove the insolubles, some of the silica passes :..+, L.* F-..,A .L‘“L ..^I *-^ llll” tr., 111efin..,+, 11111arr; “UL +L:” LLUS._.,.” was l”U,,U L” L.+.WL...,,,LeI‘G‘e with the subsequent anion-exchange separation of uranium. The silica passes right through the resin bed and hence does not block the column. Investigations of the effect of varying the concentrations of IBMK, hydr~hloric acid and of water-miscible organic solvents on the adsorption of uranium and iron on Dowex 1 gave the results presented in Table 1. At high acid concentrations in the presence of IBMK, acetone or tetrahydrofuran, the &-values and elution volumes for iron are considerably lower than those for the other systems shown in Tabie i. That is not surprising because these media contain three CIESE-active components,G6 namely hydrochloric acid and IBMK mixed with either acetone or tetrahydrofuran so that ~on(II1) chloride shows practically no tendency to be adsorbed on the anion-exch~ger. This is also true for moly~enum although to a lesser extent in most of the media; like iron, molybdenum is also readily extracted by IBMK or other ketones or ethers from hydrochloric acid solutions3 Investigations using molybdenum dissolved in the systems in which elution volumes for iron of 2 or 3 ml were measured (Table 1) have shown that for molybdenum similarly small elution volumes, which are necessary to effect complete removal of molybdenum, were obtained with only three media i.e., in the 1:8:1, 4: 5: 1 and 8: 1: 1 v/v IBMK-THF-HCl mixtures. In all other mixtures the elution volumes for molybdenum were found to exceed 50 ml. As will later be shown in Table 3, the virtually complete removal (separation) of molybdenum and iron is essential for the accurate spectrophotometric determination of uranium. From Table 1 it is also seen that uranium (as the anionic chloride complex) is strongly adsorbed from virtually all of the systems investigated-the smallest but still sufficiently high distribution coefficients being found in the media containing acetone or tetrahydro-
Table
1. Distribution
coefficients
Composition IMBK, % c/v 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 * The elution 741 2: -1 ,I
of uranium and iron concentrations
of the IMBK-HCl
Organic solvent, % VIV 10 (Methanol) 40 (Methanol) 80 (Methanol) 10 (Methanol) 40 (Methanol) 80 (Methanol) 10 (Methanol) 40 (Methanol) 80 (Methanol) 10 (Ethanol) 40 (Ethanol) 80 (Ethanol) 10 (Ethanol) 40 (Ethanol) 80 (Ethanol) 10 (Ethanol) 40 (Ethanol) 80 (Ethanol) 10 (Propan-l-01) 40 (Propan-l-01) 80 (Propan-l-01) 10 (Propan-l-01) 40 (Propan-l-01) On /n.._-?._ I -1, 0” \rl “pa‘l-l-ul, 10 40 80 10 40 80 10 40 80 10 40 80 10 40 80 10 40 80 10 40 80 10 40 80 10 40 80 10 .^ 40
80 10 40 80 10 40 80 10 40 80 10 40 80
(Propan-l-01) (Propan-l-01) _ (Propan-l-01) (Acetic acid) (Acetic acid) (Acetic acid) (Acetic acid) (Acetic acid) (Acetic acid) (Acetic acid) (Acetic acid) (Acetic acid) (Propan-2-01) (Propan-2-01) (Propan-2-01) (Propan-2-01) (Propan-2-01) (Propan-2-01) (Propan-2-01) (Propan-2-01) (Propan-2-01) (Methyl glycol) (Methyl glycol) (Methyl glycol) (Methyl glycol) (Methyl glycol) (Methyl glycol) (Methyl glycol) .._ . (Methyl giycoi) (Methyl glycol) (Acetone) (Acetone) (Acetone) (Acetone) (Acetone) (Acetone) (Acetone) (Acetone) (Acetone) (Tetrahydrofuran) (Tetrahydrofuran) (Tetrahydrofuran)
10 -- iTc=trahvrlmfnrani ,------,-_--.I.I-.,
40 80 10 40 80 volume
(Tetrahydrofuran) (Tetrahvdrofuran) (Tetrahydrofuran) (Tetrahydrofuran) (Tetrahydrofuran) (ml) of iron (l-g column
on Dowex 1, X8 in IMBK-HCI of organic solvents
mixtures
Distribution
systems HCI, % 0 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 lo I” 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10
Uranium(V1)
(12M) (12M) (12M) (6M) (6M) (6M) (1M) (1M) (1M) (12M) (12M) (12M) (6M) (6M) (6M) (1M) (1M) (1M) (12M) (12M) (12M) (6M) (6M) ,l*I\ (“WI, (1M) (1M) (1M) (12M) (12M) (12M) (6M) (6M) (6M) (1M) (1M) (1M) (12M) (12M) (12M) (6M) (6M) (6M) (1M) (1M) (1M) (12M) (12M) (12M) (6M) (6M) (6M) (1M) (1M) (1M) (12M) (12M) (12M) (6M) (6M) (6M) (1M) (1M) (1M) (12M) (12M) (12M) (6.M) (6M) (6M) (1M) (1M)
(1M)
of Dowex
1) is shown 285
1.25 1.74 20.3 two-phase 6.65 4.80 two-phase 650 460 625 3.50 14.3 two-phase 2.20 3.00 two-phase two-phase 330 610 3.60 57.0 two-phase 4.80 ILn IV” two-phase two-phase 1.23 830 9.80 ti two-phase 20.0 > two-phase two-phase > 580 3.40 85.0 two-phase 3.70 60.0 two-phase two-phase 6.80 two-phase 3.50 14.0 two-phase 4.80 22.0 two-phase 1.60 10.3 500
x lOa x lo3 x lo3 mixture x lo3 x lo3 mixture
varying
coefficients
Iron(III)* 1 (3) 3 (>50) >50 7 (ti50) >50 >50 >50
x 10s x loa mixture x 10s x lo3 mixture mixture
1 (3) 1 (>50) >50 1 (ZSO) >50
>50 1 (3) 1 (3) 50
x 10’ x 103 mixture x lo3 .,x ,n3 I”
>:o
mixture mixture x lo3
>50
x lo3 10s mixture x lo3 lo5 mixture mixture 10s x lo3 x lo3 mixture x lo3 x lo3 mixture mixture x lo3 mixture x lo3 x lo3 mixture x lo3 x 103 mixture x 10s x lo3
zz two-phase two-phase 480 two-phase two-phase 1.05 x 380 550 550
1 (>50)
1 (3) 2 (>50) 9 50 2 (>50) + 50
5 50 1 (2) 1 (5) l(150) 1 (50) 30 (B50)
$50 1 (>50) 3 (>50) 2 (250) 10 (+ 50) 2 30 1 1 1
(b 50) (9 50) (2) (2) (3)
mixture mixture 1 (>50) mixture mixture lo3
twn_nhn.e . . . ., y*Lyy’
mivt,,re II.I,,...I”
two-phase 310 two-phase two-phase 100
mixture
in brackets,
containing
1 1 1 1
(>50) (2) (2) (3)
1 (5) mixture mixture 1 (20)
286
J. KORKISCH and H. HLJBNER Table
2. Distribution
Metal ion
UO,(IU
Distribution coefficient 550 I* 2* 125 50 75 110 170 200 50 1
Fe(III) MO(W) Cu(lI) Co(H) Pb(II) Mn(II) Cd(H) Zn(II) Ni(II) Cr(lII) *The
coefficients of metal ions in IBMK-THF-HCI (1 g of Dowex 1 X8; 1-mg load)
elution
Metal ion
(1:X: 1)
Distribution coefficient
Be(H) Mg(II) Ca(I1) Na(I) K(I) Al(II1) Ce(III) V(V) Ti(IV) Zr(IV) Th(IV)
volume on a l-g column of the resin is equivalent to 3 ml.
however, that under the conditions of analysis, i.e., in the presence of large amounts of chlorides as is the case after treatment of the geological samples with hydrochioric acid, some of these eiements (and aiso those, e.g., lead, for which the adsorption decreases with increase in the chloride concentration) are much less strongly adsorbed than their &-values would indicate (salt-effect). The washing with IBMK-THF-HCl mixture immediately after passage of the sorption solution through the resin bed removes residual iron and molybdenum* so that these will not interfere with either the spectrophotometric or the fluorimetric procedure for uranium assay. Other elements which, if not separated from uranium, may cause interferences in either or both of these methods, are eluted subsequently with 6M hydrochloric acid. Thus, this eluent quantitatively eliminates from the resin all of the adsorbed elements listed in Table 2 except uranium, cadmium and zinc (The distribution coefficients in 6M hydrochloric acid are 283, > 100 and > 100 for uranium, cadmium and zinc respectively.) At the same time the residual organic solvents are also removed from the resin bed. On subsequent elution of uranium with 1M hydrochloric acid (Kd - l), cadmium and * Complete removal of iron (and also molybdenum) is zinc (Kd,,, _ 103; K,,,, > 100) are not co-eluted with impossible if it has been adsorbed from 6M hydrochloric .* or wnen 1 an ,-1x_ TT,T- rrn1 “___*:__ ,._,..*:__ __.. it. acm ~DLV,~- I 21=‘-“~1 su~p~luu ~VIUIIUIILULLThe ion-exchange separation method described was taining more than 100 mg of iron/30 ml is passed through applied to the analysis of numerous geological specithe resin column.
furan. From the latter media the 1:s: 1 v/v mixture was selected as the most suitable for the separation of uranium from geological samples. In this system the distribution coeficient of uranium (55G) is high enough to guarantee the complete adsorption of even mg-amounts of uranium and the &-values and elution volumes of iron and molybdenum are very low (see also Table 2), so they can be separated from uranium most effectively. An additional advantage of this medium stems from the fact that the tendency for formation of two phases on dissolution of the chloride residue in IBMK-THF-HCl mixtures decreases with increasing concentration of tetrahydrofuran in the mixtures. Investigations of the adsorption behaviour of numerous elements on Dowex 1 from the selected IBMK-THF-HCl mixture gave the results presented in Table 2. Atomic-absorption spectrophotometry was used for the necessary analyses. From these distribution coefficients it is evident that during the sorption process not only uranium but also copper, cobalt, lead, manganese, cadmium and zinc will be adsorbed on the resin. Previous experience has shown
Table 3. Effect of iron and molybdenum on the spectrophotometric determination of uranium with arsenazo III (5.0 pg of uranium taken) Fe present, /%l
U found, fig 5.00 5.00 5.00 5.00 4.94 4.80 4.54 4.00 1.53
MO present, PS 0
5 50 75 loo 200 300 400 500
U found, &I 5.10 5.40 5.80 4.90 4.50 000 0.00 000 @OO
257
Uranium in minerals and rocks Table 4. Results of uranium determinations in reference samples from the Internation~ Austria7
Atomic Energy Agency, Vienna,
u,a, % Spectrophotometry A B
Sample S-l (Torbernite; Australia) S-2 (Torbernite; Spain) S-3 (Carnotite; U.S.A.) S-4 (Uraninite; Australia) S-12 (Pitchblende; Spain) S-13 (Pitchblende; Spain)
0.456 0.298 0.401 0,378 0016 0,035
A
0457 (3000) 0.300 (3000) 0402 (3000) 0.377 (3000) 0.016 (100) 0.035 (100)
Other laboratories II I
Fluorimetry B
0.456 0.300 0401 0.379 0.016 0.035
0455 (3000) @300 (3ooo) 0402 (3000) 0378 (3000) 0.016 (100) @035 (100)
0483 0313 0.415 0377 0014 0037
O-471 0.313 0.418 0.375 0.014 0,039
A = Results obtained for an unspikcd sample. B = Results obtained after deduction of a spike (shown in parentheses as pg of uranium added per g of sampIe) added before the dissolution of the sample. I = Arscnaro III method.’ II = Mean of all methods used for uranium determination.’
mens (see later) and to test-solutions containing known amounts of uranium (1, 10 and 100 pg), molybdenum (5 mg), iron (100 mg) and cobalt (1 mg). Analyses of these mixtures showed that in the entire
uranium eluate not more than 25 pg of iron were present (determined by atomic~abso~tion spectrophotometry) even if as much as 100 mg of this element were present in the sorption solution. Molybdenum and cobalt were completely separated, no detectable amounts being found in the uranium eluate. The effect of iron and molybdenum on the spectrophotometric determination of uranium with arsenazo III was tested with various concentrations of these elements. From Table 3 it is seen that the tolerance
limit for iron is relatively high, up to 1.50pg of iron not interfering with the uranium determination. Thus, the method is much less sensitive to iron than the fluorimetric procedure in which much smaller quantities of iron cause an appreciable quenching of the uranium fluorescence.’ On the other hand molybdenum has prakally no influence on the fluor~et~~ method but interferes seriously when arsenazo III is used, giving a positive error when ~75 fig are present and a large negative error when > 100 ,~g are present in the solution measured (Table 3). In the presence of 200-500 pg ofmolybdenum no uranium was found. Interferences by iron and molybdenum are not to be expected to occur, however, after the ion-exchange
Table 5. Results of uranium determinations
Source’
Sam&
Rd. 93-h 2.76 324
2
3
4
5
in reference rock samples
MRG-I (Gabbro) SY-2 (Syenite rock) SY-3 (Syenite rock) NIM-G NIM-L NIM-N NIM-P NIM-S NIM-D
(Granite) (Lqavritr) (Norite) (Pyroxenite) (Syenite) (Dunk)
Granite GA Granite GH Basalte BR Biotite IMicsFe Pblogopite hfica-Mg Ihonk DR-N Serpentine ?JB-N Bauxite RX-N Dish&c DT-N USGS-G-2 (Granite) USGS-GSP-I (Granodiarite) USGS-AGV-I (And&e)
0.30 246 634
94.8 (100) 2.85 (5) 3.23 (5) 025 (Iq 250 (ZOO) 630 (500)
91.2 2.73 3.25 a21 250 6(M
93.1 (lo@ 2.63 (5) 336 (5) O-19 (1+3) 255 (200) 623 (500)
1580 15.00 038 O-30 060 0.18
16.00 (100) 15ciO(IDO) 037 (1.0) 030 (1.0) a55 (1.0) 018 (1.0)
Ifi, 1637 Q.35 0.26 0.56 @I6
16.17 16.05 0.35 026 0.50 0.13
4.75 1581 1.66 66.10 0.16 I.50
4.80 (5.0) 15.84(15) 169(1.0) 66W 1501
4.78 IS.7 f46 ii@31 010 1.38 lxx3 4% I.37
4.70 (501 1s.7 (15) 1.40
UfOif+ij I47 11.01
226 2.42 1.31
(100) (10.0) (1.0) (1.0) (I.01 (1.0)
1PO!
@Ha (50) DIO(l.Of 145 (l.Oj wn5 {l-o) 4.30 (5Q i-39 (1.0) 2.22 2.45 1.35
8 8 8
104~2 (mean] 367 (mean) 3-34 (mean) -
9 9
5~500 loo-loo0 IO-63 (14) 8-38 (13) 021-07 (05) 021-07 (05) Oc6-09 (06) 014-Z (05) 1.7: 2; 4.26; 4”86 10: 12; w2: 19.71 09: 1.9; 2.1: 2.65 35:4(t; 57; 94 I.2 1: 2
10 10 10 IO 10 IO II if II II 11 11
-..
l%t
(1,96) lI,SRl
* 1, IAEA, Vienna; 2, Canadian Standard Reference Materials Project; 3, National Institute for Metallurgy, Johannesburg; 4, Centre de Recberches Petrographiques et Gtochimiques, Vandoeuvre-l&-Nancy; 5, U.S. Geological Survey. t A, B - See footnotes to Table 4. 4 Figures in parentheses arc quoted from Flanagan” $ Determined fluorimetrically after separation by the THF-methyl &cot-HCl methoc14
288
J. KORKISCH and H. HUEZNER
Table 6. Results of uranium determinations
Uranium, “/:, Fluorimetry A
Spectrophotometry Sample DH DL BL-1 BL-2 BL-3 BL-4 A,&See
A
B
0.194 0+)042 0.0230 0.43 1 @923 0.171
0.176 (1500) 0.0042 (50) 0.0227 (250) 0,439 (5000) n.o,< (1” ,,I+) , “7I.J 0.170 (1500)
0.190 00&l2 0.0227 0.431 @922 0.171
Other laboratories (mean of all results)‘3
B
0.181 (1500) oQO40 (50) 0.0210 (250) 0.433 (5000) n.nlrr (1 n41, “7J”[I”
0.177 00039 0.0219 0.450
0.174 (1500)
0.173
::Q:6
footnotes to Table 4.
separation, because the amount of iron accompanying uranium in the eluate is always much lower than 150 pg and molybdenum is separated quantitatively. Furthermore, thorium, zirconium, titanium, rare-earth elements and phosphate, which also strongly interfere with the arsenazo III method,’ will have no effect on the determination of uranium because they are rnmnl~tc=ki ~““‘y’~‘~‘J
in Canadian radioactive ore standardsI
cpnaratcvl y’y”‘“.-..
hv vJ
the . .._
aninn-t=urhrrnoc= u__v__ _1.__ --_a-.
l&s:
100 mg of sodium
phosphate gave no interference: and up to 100 mg of sulphate did not interfere with the ion-exchange separation of uranium. The results of uranium determinations in numerous reference samples are shown in Tables 4-6. The spectrophotometric method gives somewhat higher results than fluorimetry when very iron-rich samples, e.g., Biotite Mica-Fe (total Fe as FezO, = 2575%) and Bauxite BX-N (total Fe as Fez03 = 23.21%) (see Table 5) are analysed, because the amount of iron accompanying uranium into the eluate causes quenching of the uranium fluorescence. Apart fi0iii these few instances in which the fluorimetric procedure gives lower results, the uranium results are generally in reasonably good agreement with the results obtained in other laboratories, exceptions being Sediment-2 and Phlogopite Mica-Mg (see Table 5). Because the method is based on separations on anion-exchange columns, which can be performed more or less automatically, it is possible to analyse numerous samples simultaneously, i.e., the procedure is very well suited for the routine determination of uranium over a concentration range of about five orders of magnitude.
Acknowledgements-This research was sponsored by the Fonds zur Fijrderung der wissenschaftlichen Forschung, Vienna, Austria. The generous support from this Fund is gratefully acknowledged. Acknowledgement is also made to the following institutions which have supplied us with sample material: International Atomic Energy Agency, Vienna, Austria; Canada Department of Energy, Mines and Resources, Mines Branch, Ottawa, Canada; Geological Survey of Canada, Ottawa, Canada; National Institute for Metaiiurgy, Johannesburg South Airica; Centre de Recherches PCtrographiques et Gtochimiques, VandoeuvreI&s-Nancy, France. REFERENCES 1. J.Korkisch and F. Hecht, Handbook of the Analytical Chemistry of Uranium; Quantitative Analysis, Vol. VI, b; fl. Springer-Verlag, Berlin, 1972. 2. W. Koch and J. Korkisch, Mikrochim. Acta, 1973, 157. Modern Methods for the Separation of 3. J. Korkisch, Rarer Metal Ions. Pergamon, Oxford, 1969. 4. J. Korkisch and I. Steffan, Mikrochim. Acta, 1972, 837. 5. J. Korkisch, Sepn. Sci., 1966, 1, 159. 6. Idem, Osterr. Chem. Ztg., 1966, 67, 309. ~_I._ ~_I II_._‘_ _.-.__ I-...... A...1...!--1 A... Htornlc Energy Agency, nnalyucal vua7. lmernauonal lity Control Services, Laboratory Seibersdorf, Vienna, Austria. 0. Suschny and L. Gbrski, Intern. At. 8. J. Heinonen, Energy Agency (Vienna), Rep. IAEA/RL/13, 7 June 1973. 9. Canadian Standard Reference Materials Project, Geological Survey of Canada, Ottawa, Canada. 10. B. G. Russell, R. G. Goudvis, G. Dome1 and J. Levin, Nat. Inst. Metall. Rex (Johannesburg) Rept. No. 1351, 1972. Revue du GAMS, 11. H. de la Roche and K. Govindaraju, Centre de Recherches PCtrographiques et Gtochimiques, Vandoeuvre-l&s-Nancy, 1971, 7, 314. Geochim. Cosmochim. Acta, 1973; 37, 12. F. J. Flanagan, 1189. and D. Dimitriadis, Talanta, 1973, 20, 13. J. Korkisch 1199.
Pergamon
Press, 1976. Printed IIT Great Britam
DETERMINATION
OF URANIUM
IN MINERALS
AND ROCKS
J. KORKISCH and H. HiisNw Institute for Analytical Chemistry, Analysis of Nucbar Raw Materials Division, University of Vienna, W~hringerstra~ 38, A-1090 Vienna, Austria (Received
_^__.
i6 .iune i975. Accepreti 18 October IY/5j
Summary-A method is described for the determination of uranium in minerals and rocks by spectrophotometry and fluorimetry. After treatment of the sample with hydrochloric acid, uranium is separated from matrix elements by adsorption on a column of the strongly basic anion-exchange resin Dowex 1 X8 from an organic solvent system consisting of IBMK, tetrahydrofuran and 12M hydrochloric acid (1:&l v/v). Following removal of iron, molybdenum and co-adsorbed elements by washing first with the organic solvent system and then with 6M hydrochloric acid, the uranium is eluted with 1M hydrochloric acid. In the eluate, uranium is determined by means of the spectrophotometric arsenazo III method or fluorimetrically. The suitability of the method for the determination of both trace and larger amounts of uranium was tested by analysing numerous geochemical reference samples with uranium contents in the range lo-‘-104ppm. In practically all cases very good agreement of results was obtained.
air-dried and 3 g of it were slurried in a few ml of the IBMK-THF-HCl mixture (see below) and transferred to an ion-exchange column, with the same mixture used as a rinse. Standard uranium solutions. Aliquots of a stock solution containing 10 mg of uranium (as uranyl chloride) per ml of 12M hydrochloric acid were diluted with water to obtain standard solutions in 9M, 6M and 1M hydrochloric acid containing @l-lo4 ppm of uranium. I~~~-~~~-~C~ mixture. A mixture of IBMK (isobutyl methyl ketone), THF (tetr~ydrof~~) and 12M hydrochloric acid (1:8: 1 v/v) was prepared several hours before use (in order to prevent the formation of air-bubbles in the resin bed). This solution can be stored for at least a week without loss of effectiveness. Arsenazo III solution. Freshly prepared and filtered 0.1% aqueous solution. Other reagents. “Fluorhase” (200 mg) pellets consisting of 95% sodium fluoride and 5% lithium fluoride (Nordrhein-Chemie, 41 Duisburg 12, West Germany), zinc metal (dust or finely granulated zinc), oxalic acid 1.0, 6, 9 and 12M hydrochloric acid, 1M nitric acid, and reagent-grade methanol
Spectrophotometric and fluorimetric methods are employed extensively for the determination of uranium in a variety of geological samples, including materials of very low uranium contents such as basalt-
ic rocks, and of uranium-rich minerals, as for instance torbernite, carnotite, uraninite and pitchblende.’ For the successful analysis of both types of ~~iurn-~~~g materials the very sensitive spectrophotometric method of ur~ium detestation can be used, which is based on measurement of the absorbance of the green ur~i~(~~ar~n~o III complex in moderately concentrated hydrochloric acid medium.’ Interferences are caused by thorium, zirconium, titanium, rare earth elements, molybdenum, and iron, so it is necessary to separate uranium from these elements and also from other components such as the main constituents of minerals and rocks. For this purpose, methods based on liquid-liquid extraction and ion-exchange have found widespread application.’ In the present paper an application is described of a single separation step which is based on anionexchange in an organic solvent system. Following this quantitative separation, uranium can be determined free from interferences, by either the arsenazo III method or fluorimetrically.
EXPERIMENTAL
Reagents ion-exchanger. The strongly basic anion-exchanger Dowex 1 X8 (loo-Zoo mesh; chloride form) was used. To purify the exchanger (which contained, e.a., iron and zinc) 200 g of the resin were treated in succes~on with 2 htres of fM nitric acid, 2 htres of 6M hydr~hloric acid, 2 litres of 1M hydrochloric acid, 3 litres of distilled water and 1 litre of reagent grade methanol. This batch of resin was
Apparatus The
ion-exchange
~parations
of uranium
were per-
formed in columns of the type and dimensions described earlier.’ Determination of distribution coefficients
The distribution coefficients (&-values) of uranium and some of the more strongly adsorbed metal ions were determined by using the batch-equilibrium method, and the column technique was employed to measure the &-values of iron, molybdenum and other elements which are weakly retained by the resin.3 The usually high distribution coefficients of uranium were determined after equilibration (for 4 hr) of 1 g of the resin with 20 ml of the organic solventhydrochloric acid mixture tested (see Tabte 1) containing 20 mg of uranium, followed by filtration, etc3 In the determinations of the batch and column ~stribution coefficients of the other elements only 1 mg each of the meta ions was used (see Table 2).
283
284
J.
KORKISCH
Procedures D~sso~~~~o~ ~~su~~~es. To 1.0 g of the thoroughly homogenized sample, in a 2%ml beaker, 100 ml of concentrated hydrochloric acid are added, the beaker is covered with a watch-glass, and the mixture is heated on a sand-bath until its volume has been reduced to about 30 ml. Subsequently the solution is evaporated to dryness under an infrared lamp, another 100 ml of concentrated hydrochloric acid are added and the solution is evaporated to dryness again under the heat-lamp. To the residue 30 ml RF“l.l nnn l,.,,&.,Y.l.,rl,r;r nr;rl (IIS, nr* (LUUL,~) o&i-A lL‘lP l.a .LLII*Lu‘G ..&+.....1s :.. ,,ea&Xl L-,.+-A “, L’JuLw.,~II”IIrLIc.lll under the lamp for several minutes and then let stand (preferably overnight). Afterwards, insoluble material (mainly silica) is removed by filtering off on a dense filter, 6h;I hy~o~hlo~~ acid being used for washing the filter naner and residue free from -iron. The filtrate -6 evaporated- to dryness on a steam-bath and the residue is taken up in about 30 ml of the IBMK-THF-HCl mixture. If two phases are formed, more solvent mixture is added until a homogeneous solution is obtained* in which however, a portion of the solute (mainly silica) usually reaDDears as a suspension, owing to the- decrease of sblubihcy on addition of the IBMK-THE-HCl mixture. This insoluble material is filtered off on a dense filter, and washed with IBMK-THF-HCl solution. The filtrate (usually 5GlOO ml) is the sorption solution for the ion-exchange separation. Ion-exchange separation. The sorption solution is passed through the ion-exchange column containing 3 g of the resin (pretreated with 10 ml of the IBMK-THF-HCL mixture), at a flow-rate of about 08 mlimin. Iron and molvbdenum are removed by washing the column with 50*ml of the IBMK-THF-HCl mixture. and Conner. cobalt. etc. as well as organic solvents left in the resin bed are eluted with 100 ml of 6M hydrochloric acid. The adsorbed uranium is then eluted with 100 ml of 1M hydrochloric acid. The resin column may be used for the isolation of uranium from further samples provided that 30 ml of the IBMK-THF-HCl mixture are used to pretreat it following the eiution of uranium. Quantitative determination of uranium. The uranium eluate is evaporated to drvness on a steam-bath or under an infrared Iamp and the residue is dissolved in 9M hydrochloric acid and transferred to a 25-m] standard flask and diluted to volume with the acid. If this solution contains > 1 ppm of uranium (i.e., > 25 pg of uranium) it is necessary to dilute it (or an aliquot) with the same acid to a suitable volume in which this concentration is not exceeded A. Spectrophotometric method. Ten ml of the 9M hydrochloric acid solution of the uranium are transferred to a lOO-ml wide-neck Erlenmeyer flask, 0.3 g of oxalic acid and 1.10 g of zinc are added and the flask is covered loosely with a y.vYY”. QtnnntV. -l-m&lo . . “..b th_e reduction. the flask is shaken carefully until all the zinc has dissolved. Immediately afterwards 1.0 ml of the arsenazo III solution is
added and the absorbance is measured at 665 nm against a reagent blank prepared in the same way. A calibration curve for the range I-1Opg of uranium is constructed .according to the reduction procedure described above. Beer’s law is strictly obeyed up to 10 ng of uranium jabsorbance @400). Higher amounts give strong positive deviations from linearity. The absorbance remains constant for at least 100 min. If less than 0.2 pg of uranium is present in the 10 ml of test solution (<0.5 ppm in the original sample) this method is less reliable (because of the lo\u * If after addition of the organic solvent mixture a total volume of 100 ml should be reached without obtaining homogeneity, 1 ml of concentrated hydrochloric acid is added-this causes immediate disappearance of the two phases.
and H. HCJBNER absorbance) than the fluorime~ic procedure described below. B. Fluorimetric method. A suitable volume (e.g., 0.1 ml) of the uranium eluate or of the 25 ml of the 9M hydru chloric acid solution containing the uranium is evaporated in a small platinum dish“ and after addition of a “Fluorbase” pellet, a melt is prepared under strictly controlled conditions.’ The fluorescence intensity of the cold flux is measured and compared with the intensity of fluxes of known uranium concentrations.
RESULTS AND DISCUSSION The dissolution procedure described, in which only hydr~hloric acid is used, proved to be highly satis-
factory for all of the samples analysed although it is in principle based on leaching only. Investigation of the completeness of dissolution showed that the portions insoluble in hydrochloric acid never retained more than about 0.1% of the total uranium content. During filtration of the IBMK-THF-HCl mixture to remove the insolubles, some of the silica passes :..+, L.* F-..,A .L‘“L ..^I *-^ llll” tr., 111efin..,+, 11111arr; “UL +L:” LLUS._.,.” was l”U,,U L” L.+.WL...,,,LeI‘G‘e with the subsequent anion-exchange separation of uranium. The silica passes right through the resin bed and hence does not block the column. Investigations of the effect of varying the concentrations of IBMK, hydr~hloric acid and of water-miscible organic solvents on the adsorption of uranium and iron on Dowex 1 gave the results presented in Table 1. At high acid concentrations in the presence of IBMK, acetone or tetrahydrofuran, the &-values and elution volumes for iron are considerably lower than those for the other systems shown in Tabie i. That is not surprising because these media contain three CIESE-active components,G6 namely hydrochloric acid and IBMK mixed with either acetone or tetrahydrofuran so that ~on(II1) chloride shows practically no tendency to be adsorbed on the anion-exch~ger. This is also true for moly~enum although to a lesser extent in most of the media; like iron, molybdenum is also readily extracted by IBMK or other ketones or ethers from hydrochloric acid solutions3 Investigations using molybdenum dissolved in the systems in which elution volumes for iron of 2 or 3 ml were measured (Table 1) have shown that for molybdenum similarly small elution volumes, which are necessary to effect complete removal of molybdenum, were obtained with only three media i.e., in the 1:8:1, 4: 5: 1 and 8: 1: 1 v/v IBMK-THF-HCl mixtures. In all other mixtures the elution volumes for molybdenum were found to exceed 50 ml. As will later be shown in Table 3, the virtually complete removal (separation) of molybdenum and iron is essential for the accurate spectrophotometric determination of uranium. From Table 1 it is also seen that uranium (as the anionic chloride complex) is strongly adsorbed from virtually all of the systems investigated-the smallest but still sufficiently high distribution coefficients being found in the media containing acetone or tetrahydro-
Table
1. Distribution
coefficients
Composition IMBK, % c/v 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 80 50 10 * The elution 741 2: -1 ,I
of uranium and iron concentrations
of the IMBK-HCl
Organic solvent, % VIV 10 (Methanol) 40 (Methanol) 80 (Methanol) 10 (Methanol) 40 (Methanol) 80 (Methanol) 10 (Methanol) 40 (Methanol) 80 (Methanol) 10 (Ethanol) 40 (Ethanol) 80 (Ethanol) 10 (Ethanol) 40 (Ethanol) 80 (Ethanol) 10 (Ethanol) 40 (Ethanol) 80 (Ethanol) 10 (Propan-l-01) 40 (Propan-l-01) 80 (Propan-l-01) 10 (Propan-l-01) 40 (Propan-l-01) On /n.._-?._ I -1, 0” \rl “pa‘l-l-ul, 10 40 80 10 40 80 10 40 80 10 40 80 10 40 80 10 40 80 10 40 80 10 40 80 10 40 80 10 .^ 40
80 10 40 80 10 40 80 10 40 80 10 40 80
(Propan-l-01) (Propan-l-01) _ (Propan-l-01) (Acetic acid) (Acetic acid) (Acetic acid) (Acetic acid) (Acetic acid) (Acetic acid) (Acetic acid) (Acetic acid) (Acetic acid) (Propan-2-01) (Propan-2-01) (Propan-2-01) (Propan-2-01) (Propan-2-01) (Propan-2-01) (Propan-2-01) (Propan-2-01) (Propan-2-01) (Methyl glycol) (Methyl glycol) (Methyl glycol) (Methyl glycol) (Methyl glycol) (Methyl glycol) (Methyl glycol) .._ . (Methyl giycoi) (Methyl glycol) (Acetone) (Acetone) (Acetone) (Acetone) (Acetone) (Acetone) (Acetone) (Acetone) (Acetone) (Tetrahydrofuran) (Tetrahydrofuran) (Tetrahydrofuran)
10 -- iTc=trahvrlmfnrani ,------,-_--.I.I-.,
40 80 10 40 80 volume
(Tetrahydrofuran) (Tetrahvdrofuran) (Tetrahydrofuran) (Tetrahydrofuran) (Tetrahydrofuran) (ml) of iron (l-g column
on Dowex 1, X8 in IMBK-HCI of organic solvents
mixtures
Distribution
systems HCI, % 0 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 lo I” 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10
Uranium(V1)
(12M) (12M) (12M) (6M) (6M) (6M) (1M) (1M) (1M) (12M) (12M) (12M) (6M) (6M) (6M) (1M) (1M) (1M) (12M) (12M) (12M) (6M) (6M) ,l*I\ (“WI, (1M) (1M) (1M) (12M) (12M) (12M) (6M) (6M) (6M) (1M) (1M) (1M) (12M) (12M) (12M) (6M) (6M) (6M) (1M) (1M) (1M) (12M) (12M) (12M) (6M) (6M) (6M) (1M) (1M) (1M) (12M) (12M) (12M) (6M) (6M) (6M) (1M) (1M) (1M) (12M) (12M) (12M) (6.M) (6M) (6M) (1M) (1M)
(1M)
of Dowex
1) is shown 285
1.25 1.74 20.3 two-phase 6.65 4.80 two-phase 650 460 625 3.50 14.3 two-phase 2.20 3.00 two-phase two-phase 330 610 3.60 57.0 two-phase 4.80 ILn IV” two-phase two-phase 1.23 830 9.80 ti two-phase 20.0 > two-phase two-phase > 580 3.40 85.0 two-phase 3.70 60.0 two-phase two-phase 6.80 two-phase 3.50 14.0 two-phase 4.80 22.0 two-phase 1.60 10.3 500
x lOa x lo3 x lo3 mixture x lo3 x lo3 mixture
varying
coefficients
Iron(III)* 1 (3) 3 (>50) >50 7 (ti50) >50 >50 >50
x 10s x loa mixture x 10s x lo3 mixture mixture
1 (3) 1 (>50) >50 1 (ZSO) >50
>50 1 (3) 1 (3) 50
x 10’ x 103 mixture x lo3 .,x ,n3 I”
>:o
mixture mixture x lo3
>50
x lo3 10s mixture x lo3 lo5 mixture mixture 10s x lo3 x lo3 mixture x lo3 x lo3 mixture mixture x lo3 mixture x lo3 x lo3 mixture x lo3 x 103 mixture x 10s x lo3
zz two-phase two-phase 480 two-phase two-phase 1.05 x 380 550 550
1 (>50)
1 (3) 2 (>50) 9 50 2 (>50) + 50
5 50 1 (2) 1 (5) l(150) 1 (50) 30 (B50)
$50 1 (>50) 3 (>50) 2 (250) 10 (+ 50) 2 30 1 1 1
(b 50) (9 50) (2) (2) (3)
mixture mixture 1 (>50) mixture mixture lo3
twn_nhn.e . . . ., y*Lyy’
mivt,,re II.I,,...I”
two-phase 310 two-phase two-phase 100
mixture
in brackets,
containing
1 1 1 1
(>50) (2) (2) (3)
1 (5) mixture mixture 1 (20)
286
J. KORKISCH and H. HLJBNER Table
2. Distribution
Metal ion
UO,(IU
Distribution coefficient 550 I* 2* 125 50 75 110 170 200 50 1
Fe(III) MO(W) Cu(lI) Co(H) Pb(II) Mn(II) Cd(H) Zn(II) Ni(II) Cr(lII) *The
coefficients of metal ions in IBMK-THF-HCI (1 g of Dowex 1 X8; 1-mg load)
elution
Metal ion
(1:X: 1)
Distribution coefficient
Be(H) Mg(II) Ca(I1) Na(I) K(I) Al(II1) Ce(III) V(V) Ti(IV) Zr(IV) Th(IV)
volume on a l-g column of the resin is equivalent to 3 ml.
however, that under the conditions of analysis, i.e., in the presence of large amounts of chlorides as is the case after treatment of the geological samples with hydrochioric acid, some of these eiements (and aiso those, e.g., lead, for which the adsorption decreases with increase in the chloride concentration) are much less strongly adsorbed than their &-values would indicate (salt-effect). The washing with IBMK-THF-HCl mixture immediately after passage of the sorption solution through the resin bed removes residual iron and molybdenum* so that these will not interfere with either the spectrophotometric or the fluorimetric procedure for uranium assay. Other elements which, if not separated from uranium, may cause interferences in either or both of these methods, are eluted subsequently with 6M hydrochloric acid. Thus, this eluent quantitatively eliminates from the resin all of the adsorbed elements listed in Table 2 except uranium, cadmium and zinc (The distribution coefficients in 6M hydrochloric acid are 283, > 100 and > 100 for uranium, cadmium and zinc respectively.) At the same time the residual organic solvents are also removed from the resin bed. On subsequent elution of uranium with 1M hydrochloric acid (Kd - l), cadmium and * Complete removal of iron (and also molybdenum) is zinc (Kd,,, _ 103; K,,,, > 100) are not co-eluted with impossible if it has been adsorbed from 6M hydrochloric .* or wnen 1 an ,-1x_ TT,T- rrn1 “___*:__ ,._,..*:__ __.. it. acm ~DLV,~- I 21=‘-“~1 su~p~luu ~VIUIIUIILULLThe ion-exchange separation method described was taining more than 100 mg of iron/30 ml is passed through applied to the analysis of numerous geological specithe resin column.
furan. From the latter media the 1:s: 1 v/v mixture was selected as the most suitable for the separation of uranium from geological samples. In this system the distribution coeficient of uranium (55G) is high enough to guarantee the complete adsorption of even mg-amounts of uranium and the &-values and elution volumes of iron and molybdenum are very low (see also Table 2), so they can be separated from uranium most effectively. An additional advantage of this medium stems from the fact that the tendency for formation of two phases on dissolution of the chloride residue in IBMK-THF-HCl mixtures decreases with increasing concentration of tetrahydrofuran in the mixtures. Investigations of the adsorption behaviour of numerous elements on Dowex 1 from the selected IBMK-THF-HCl mixture gave the results presented in Table 2. Atomic-absorption spectrophotometry was used for the necessary analyses. From these distribution coefficients it is evident that during the sorption process not only uranium but also copper, cobalt, lead, manganese, cadmium and zinc will be adsorbed on the resin. Previous experience has shown
Table 3. Effect of iron and molybdenum on the spectrophotometric determination of uranium with arsenazo III (5.0 pg of uranium taken) Fe present, /%l
U found, fig 5.00 5.00 5.00 5.00 4.94 4.80 4.54 4.00 1.53
MO present, PS 0
5 50 75 loo 200 300 400 500
U found, &I 5.10 5.40 5.80 4.90 4.50 000 0.00 000 @OO
257
Uranium in minerals and rocks Table 4. Results of uranium determinations in reference samples from the Internation~ Austria7
Atomic Energy Agency, Vienna,
u,a, % Spectrophotometry A B
Sample S-l (Torbernite; Australia) S-2 (Torbernite; Spain) S-3 (Carnotite; U.S.A.) S-4 (Uraninite; Australia) S-12 (Pitchblende; Spain) S-13 (Pitchblende; Spain)
0.456 0.298 0.401 0,378 0016 0,035
A
0457 (3000) 0.300 (3000) 0402 (3000) 0.377 (3000) 0.016 (100) 0.035 (100)
Other laboratories II I
Fluorimetry B
0.456 0.300 0401 0.379 0.016 0.035
0455 (3000) @300 (3ooo) 0402 (3000) 0378 (3000) 0.016 (100) @035 (100)
0483 0313 0.415 0377 0014 0037
O-471 0.313 0.418 0.375 0.014 0,039
A = Results obtained for an unspikcd sample. B = Results obtained after deduction of a spike (shown in parentheses as pg of uranium added per g of sampIe) added before the dissolution of the sample. I = Arscnaro III method.’ II = Mean of all methods used for uranium determination.’
mens (see later) and to test-solutions containing known amounts of uranium (1, 10 and 100 pg), molybdenum (5 mg), iron (100 mg) and cobalt (1 mg). Analyses of these mixtures showed that in the entire
uranium eluate not more than 25 pg of iron were present (determined by atomic~abso~tion spectrophotometry) even if as much as 100 mg of this element were present in the sorption solution. Molybdenum and cobalt were completely separated, no detectable amounts being found in the uranium eluate. The effect of iron and molybdenum on the spectrophotometric determination of uranium with arsenazo III was tested with various concentrations of these elements. From Table 3 it is seen that the tolerance
limit for iron is relatively high, up to 1.50pg of iron not interfering with the uranium determination. Thus, the method is much less sensitive to iron than the fluorimetric procedure in which much smaller quantities of iron cause an appreciable quenching of the uranium fluorescence.’ On the other hand molybdenum has prakally no influence on the fluor~et~~ method but interferes seriously when arsenazo III is used, giving a positive error when ~75 fig are present and a large negative error when > 100 ,~g are present in the solution measured (Table 3). In the presence of 200-500 pg ofmolybdenum no uranium was found. Interferences by iron and molybdenum are not to be expected to occur, however, after the ion-exchange
Table 5. Results of uranium determinations
Source’
Sam&
Rd. 93-h 2.76 324
2
3
4
5
in reference rock samples
MRG-I (Gabbro) SY-2 (Syenite rock) SY-3 (Syenite rock) NIM-G NIM-L NIM-N NIM-P NIM-S NIM-D
(Granite) (Lqavritr) (Norite) (Pyroxenite) (Syenite) (Dunk)
Granite GA Granite GH Basalte BR Biotite IMicsFe Pblogopite hfica-Mg Ihonk DR-N Serpentine ?JB-N Bauxite RX-N Dish&c DT-N USGS-G-2 (Granite) USGS-GSP-I (Granodiarite) USGS-AGV-I (And&e)
0.30 246 634
94.8 (100) 2.85 (5) 3.23 (5) 025 (Iq 250 (ZOO) 630 (500)
91.2 2.73 3.25 a21 250 6(M
93.1 (lo@ 2.63 (5) 336 (5) O-19 (1+3) 255 (200) 623 (500)
1580 15.00 038 O-30 060 0.18
16.00 (100) 15ciO(IDO) 037 (1.0) 030 (1.0) a55 (1.0) 018 (1.0)
Ifi, 1637 Q.35 0.26 0.56 @I6
16.17 16.05 0.35 026 0.50 0.13
4.75 1581 1.66 66.10 0.16 I.50
4.80 (5.0) 15.84(15) 169(1.0) 66W 1501
4.78 IS.7 f46 ii@31 010 1.38 lxx3 4% I.37
4.70 (501 1s.7 (15) 1.40
UfOif+ij I47 11.01
226 2.42 1.31
(100) (10.0) (1.0) (1.0) (I.01 (1.0)
1PO!
@Ha (50) DIO(l.Of 145 (l.Oj wn5 {l-o) 4.30 (5Q i-39 (1.0) 2.22 2.45 1.35
8 8 8
104~2 (mean] 367 (mean) 3-34 (mean) -
9 9
5~500 loo-loo0 IO-63 (14) 8-38 (13) 021-07 (05) 021-07 (05) Oc6-09 (06) 014-Z (05) 1.7: 2; 4.26; 4”86 10: 12; w2: 19.71 09: 1.9; 2.1: 2.65 35:4(t; 57; 94 I.2 1: 2
10 10 10 IO 10 IO II if II II 11 11
-..
l%t
(1,96) lI,SRl
* 1, IAEA, Vienna; 2, Canadian Standard Reference Materials Project; 3, National Institute for Metallurgy, Johannesburg; 4, Centre de Recberches Petrographiques et Gtochimiques, Vandoeuvre-l&-Nancy; 5, U.S. Geological Survey. t A, B - See footnotes to Table 4. 4 Figures in parentheses arc quoted from Flanagan” $ Determined fluorimetrically after separation by the THF-methyl &cot-HCl methoc14
288
J. KORKISCH and H. HUEZNER
Table 6. Results of uranium determinations
Uranium, “/:, Fluorimetry A
Spectrophotometry Sample DH DL BL-1 BL-2 BL-3 BL-4 A,&See
A
B
0.194 0+)042 0.0230 0.43 1 @923 0.171
0.176 (1500) 0.0042 (50) 0.0227 (250) 0,439 (5000) n.o,< (1” ,,I+) , “7I.J 0.170 (1500)
0.190 00&l2 0.0227 0.431 @922 0.171
Other laboratories (mean of all results)‘3
B
0.181 (1500) oQO40 (50) 0.0210 (250) 0.433 (5000) n.nlrr (1 n41, “7J”[I”
0.177 00039 0.0219 0.450
0.174 (1500)
0.173
::Q:6
footnotes to Table 4.
separation, because the amount of iron accompanying uranium in the eluate is always much lower than 150 pg and molybdenum is separated quantitatively. Furthermore, thorium, zirconium, titanium, rare-earth elements and phosphate, which also strongly interfere with the arsenazo III method,’ will have no effect on the determination of uranium because they are rnmnl~tc=ki ~““‘y’~‘~‘J
in Canadian radioactive ore standardsI
cpnaratcvl y’y”‘“.-..
hv vJ
the . .._
aninn-t=urhrrnoc= u__v__ _1.__ --_a-.
l&s:
100 mg of sodium
phosphate gave no interference: and up to 100 mg of sulphate did not interfere with the ion-exchange separation of uranium. The results of uranium determinations in numerous reference samples are shown in Tables 4-6. The spectrophotometric method gives somewhat higher results than fluorimetry when very iron-rich samples, e.g., Biotite Mica-Fe (total Fe as FezO, = 2575%) and Bauxite BX-N (total Fe as Fez03 = 23.21%) (see Table 5) are analysed, because the amount of iron accompanying uranium into the eluate causes quenching of the uranium fluorescence. Apart fi0iii these few instances in which the fluorimetric procedure gives lower results, the uranium results are generally in reasonably good agreement with the results obtained in other laboratories, exceptions being Sediment-2 and Phlogopite Mica-Mg (see Table 5). Because the method is based on separations on anion-exchange columns, which can be performed more or less automatically, it is possible to analyse numerous samples simultaneously, i.e., the procedure is very well suited for the routine determination of uranium over a concentration range of about five orders of magnitude.
Acknowledgements-This research was sponsored by the Fonds zur Fijrderung der wissenschaftlichen Forschung, Vienna, Austria. The generous support from this Fund is gratefully acknowledged. Acknowledgement is also made to the following institutions which have supplied us with sample material: International Atomic Energy Agency, Vienna, Austria; Canada Department of Energy, Mines and Resources, Mines Branch, Ottawa, Canada; Geological Survey of Canada, Ottawa, Canada; National Institute for Metaiiurgy, Johannesburg South Airica; Centre de Recherches PCtrographiques et Gtochimiques, VandoeuvreI&s-Nancy, France. REFERENCES 1. J.Korkisch and F. Hecht, Handbook of the Analytical Chemistry of Uranium; Quantitative Analysis, Vol. VI, b; fl. Springer-Verlag, Berlin, 1972. 2. W. Koch and J. Korkisch, Mikrochim. Acta, 1973, 157. Modern Methods for the Separation of 3. J. Korkisch, Rarer Metal Ions. Pergamon, Oxford, 1969. 4. J. Korkisch and I. Steffan, Mikrochim. Acta, 1972, 837. 5. J. Korkisch, Sepn. Sci., 1966, 1, 159. 6. Idem, Osterr. Chem. Ztg., 1966, 67, 309. ~_I._ ~_I II_._‘_ _.-.__ I-...... A...1...!--1 A... Htornlc Energy Agency, nnalyucal vua7. lmernauonal lity Control Services, Laboratory Seibersdorf, Vienna, Austria. 0. Suschny and L. Gbrski, Intern. At. 8. J. Heinonen, Energy Agency (Vienna), Rep. IAEA/RL/13, 7 June 1973. 9. Canadian Standard Reference Materials Project, Geological Survey of Canada, Ottawa, Canada. 10. B. G. Russell, R. G. Goudvis, G. Dome1 and J. Levin, Nat. Inst. Metall. Rex (Johannesburg) Rept. No. 1351, 1972. Revue du GAMS, 11. H. de la Roche and K. Govindaraju, Centre de Recherches PCtrographiques et Gtochimiques, Vandoeuvre-l&s-Nancy, 1971, 7, 314. Geochim. Cosmochim. Acta, 1973; 37, 12. F. J. Flanagan, 1189. and D. Dimitriadis, Talanta, 1973, 20, 13. J. Korkisch 1199.