I s 0topic Determination of Ca I c ium-48 by Proton Activation MAX PEISACH and RENE PRETORIUS Southern Universities Nuclear Institute, Faure, Activation with 4.75-m.e.v. protons of calcium samples containing enriched calcium-48 as tracer, leads to the formation of 44-hour scandium-48 and 4-hour scandium-43 and 44. The ratio of these activities gives a measure of the isotopic concentration of calcium-48. Other calcium isotopes do not interfere. The relative standard deviation over the concentration range from 0.18 (natural) to 4 atom % Ca48is 2.7%.
R
adiotracers are being used for calcium tracing in vivo, but the reluctance of the medical profession to use radioactive material on healthy subjects prompted the use of stable calcium isotopes for tracing purposes. However, the utilizaticn of stable isotopes depended, among other considerations, on the availability of a suitable method of analysis. The most sensitive method of isotopic analysis is by mass spectrometry. Nuclear methods of analysis, though not as sensitive, can serve a useful purpose in this field. Natural calcium consists of six stable isotopes, of which calcium40 makes up almost 97y0 of the natural element. Accordingly any one of the five heavier isotopes could serve as a tracer in biological studies. In this paper the determination of the isotopic
Table 1.
C.P., South Africa
concentration of calcium-48 used as a tracer is described. Nuclear methods have been developed for determining calcium48 by neutron activation either through measuring the high energy gamma-rays emitted by calcium49 (2), or by allowing the calcium49 to decay and measuring the daughter scandium49 after chemical separation (6). These methods, though useful for calcium48 analysis, have the disadvantage that the total calcium content has to be determined separately. It would therefore be advantageous to devise a direct method which would lead to the isotopic ratio of tracer to total calcium directly. In order to obtain reasonable yields from the proton activation of calcium, it would be advisable to irradiate at energies somewhat above the coulomb barrier, 4.1 m.e.v. At such energies (p,n) reactions are most likely to occur. Table I lists the &-values of (p,n) reactions on calcium isotopes and some nuclear characteristics of the products obtained. From Table I it follows that a (p,n) reaction is impossible with calcium40 or calcium-42 for incident proton energies between 4 and 5 m.e.v. If i t is assumed that the (p,n)reaction cross sections for the remaining four isotopes are comparable, the main activities that can be expected in a proton irradiated calcium target are the
Nuclear Properties of Target Nuclides
Product Target nuclide Ca40 Ca42
Natural abundance
(%I
96.97 0.64
&-value(1) (p,n)(m.e.v.)
- 14.680 -6.700
Nuclide Sc" Sc41m
Half-life 0.18 sec. 62 see.
Maina decay gamma-rays (m.e.v.) ( p + ) 3.75 ( p + ) 1.52, 1 . 2 3 0.44 ( p + ) 0.37 0.271 (p+) 1.16
3 . 9 hr. 2.4 d 4.0hr. sc44 -2.165 Sc46m 20 see. 0.0033 Ca46 84 d 1.12, 0 . 8 9 Sc46 -0.515 Sc48 44 hr. 1.31,1.04,0.99 0.18 Ca48 F 16 10-19 sec. 99.759 0 0.037 0 17 -3:544 F17 66 sec. (8+) F 18 -2.450 110 m. (0 +) 0.204 018 Ne18 -4.031 18 sec. ( P +) F19 100 -0.083.(3) Wlmm 5 msec. 0.10 0.0123 Talrn w'm stable +0.242 (S) W181rn 14 pec. 0.15, 0.14 TalB1 99.988 W181 130d 0 ( p + ) refers to positron decay and hence the appearance of 0.51 m.e.v. gamma-rays.
Ca43 Ca4'
956
0.145 2.06
-3.003 -4.431
ANALYTICAL CHEMISTRY
sc43
Sc44m
scandium isotopes 43, 44, and 48. The yield of scandium46 is expected to be very small because of the low natural abundance of calcium46 (0.0033%) and the relatively short half-life of scandium46m and relatively long one of scandium-46. The activity, A , produced in w gram of a target element irradiated for a time t in a constant flux of $ particles per cm.2-second by a reaction with cross section u is given by
where N is the Avogadro number, 6.025 X loz3, iM the atomic weight of the target element, a the fractional abundance of the target nuclide producing the required activity, and x the decay constant of the radioactive product. Proton activation of calcium samples containing enriched calcium-48 would produce readily distinguishable scandium48 activity and some other identifiable scandium isotope whose activity could be used as a measure of natural calcium within the target. The ratio of these activities, R, would be given by
where the subscripts 1 and n refer, respectively, to the scandium isotopes obtained from the label, calcium-48, and the natural calcium isotopes. In Equation 2 it is assumed that the enrichment of calcium48 is .not large enough to change the value of M appreciably. Because a, is constant and the cross sections are determined by the energy of the irradiation beam, then, for a fixed irradiation duration, it is clear that the isotopic concentration of calcium-48 would be proportional to the activity ratio, R R = kai (3) Charged particle irradiation is limited to thin deposits because of the very short range of charged particles in matter. At the same time, the entire energy of the beam is deposited within a very short distance in the target, thus resulting in large temperature rises. Indeed, the thermal properties of a material frequently determine whether
A
A 1;
0.23 haup
E 1;
8.60 b u r s
C
-
66.37 hours D 1: 147.83 hours t;
-u-
b 16' I
l
0
Figure 2.
IF Figure 1. Rotating sample holder cooled by circulating water A. 8. C. D. E. F.
G. H. I. J.
K. 1. M. N:
Motor and gear box Nylon bearings W a t e r inlet W a t e r chambers Vacuum seal Base plate Rotating cooled target holder Cooling water stream guide Target clamps Target 1 2 flats for mounting target W a t e r and vacuum seals Target holder support W a t e r outlet
irradiation with charged particles is practical or not. For the analysis of calcium by proton irradiation, few forms of the target material are suitable. The more common calcium compounds usually precipitated in analyses, such as the carbonate and oxalate, decompose on heating. Possible interference from the nuclear reactions of other elements also make it desirable that only simple compounds such as the oxide and fluoride which melt a t 2580" and 1360" C., respectively, be considered. Proton activation of oxygen will result in the formation o€ 66-second fluorine-17 and 110-minute fluorine-18 by (p,n)reactions on the stable isotopes of oxygen, while fluorine will only form 18-second neon-19. Comparing the properties of these radio-nuclides with the ( p , n ) products of calcium in
l
I
1
I
0.5
I
I
I
,
I
l
l
l
1.0 1.5 GAMMA RAY ENEffiY(MPV)
l
l
l
l
l
l
l
l
l
/
2.5
2.0
Gamma spectra from proton irradiated natural calcium fluoride
Table I it is clear that fluoride would be preferable to oxide. The backing material for the target should be inert to activation and capable of withstanding relatively high temperatures. Tantalum meets these requirements, because the only activity that could be measured with the sample is tungsten-181 (see Table I). However, even this nuclide is not expected to be present in any appreciable concentration, because the coulomb barrier for protons is about 10.5 m.e.v.
protons in a beam current of about 2 Pa. After irradiation, samples were analysed either by gamma-ray spectrometry or by gross gamma-ray counting with a 3- X 3-inch XaI(T1) scintillation detector. Counting usually started about 15 minutes after irradiation, thereby allowing sufficient time for very short-lived activities to decay, and continued periodically until sufficient data had been accumulated to allow the major components to be separately determined. RESULTS AND DISCUSSION
EXPERIMENTAL
Preparation of Samples and Standards. Calcium, separated from biological material as oxalate or carbonate, was redissolved and converted to calcium fluoride by evaporation with HF in a platinum crucible. The calcium fluoride was transferred to a tungsten boat and distilled in vacuum onto tantalum discs over an area compatible with the cross sectional area of the irradiation beam. Standards were similarly prepared from calcium carbonate enriched to 97.98% in calcium-48, obtained from the Oak Ridge Xational Laboratory, U.S.X., which was diluted with natural calcium to the required isotopic concentration. Irradiation and Measurement. Sample discs were mounted on a rotating holder cooled by circulating coolant (see Figure 1) inside a vacuum chamber which fitted onto the beam tube of the 5.5 m.e.v. Van de Graaff accelerator a t the Southern Universities Kuclear Institute. The irradiation beam was defocussed until its cross section exceeded that of the calcium fluoride deposit. Targets were irradiated up to 2 hours with 4.75 m.e.v.
.
Gamma- Ray Spectrometry Gamma-ray spectra recorded during the decay of an irradiated sample of natural calcium fluoride are shown in Figure 2 . Immediately after irradiation, the presence of scandium44 was proved by the pronounced peak from positron annihilation radiation of 0.51 m.e.v. and the photopeak from the 1.16 m.e.v. gamma-ray, the apparent photopeak a t 1.67 m.e.v. is a sum peak (Sum peak 1 in Figure 2 ) . After 8.6 hours the presence of the longer-lived scandium48 could already be distinguished by the appearance of photopeaks from gamma-rays of 0.99 and 1.04 m.e.1'. (unresolved) and 1.31 m.e.v. Sum peak 1 was still evident, and similar but smaller sum peaks (2 and 3) could be distinguished a t about 2 and 2.3 m.e.v. After about 46 hours the characteristics of scandium-44 had almost disappeared. The shape of the spectrum thereafter remained virtually unaltered, but for the disappearance of sum peaks as a result of the decreased count rate. The identity of each radionuclide VOL. 38, NO. 8, JULY 1966
957
was confirmed by the rate of decay of the respective gamma-rays. At lower proton energies scandium-43 was shown to be the only positron cmitter formed (..$), but under the conditions of the irradiation described above, scandium43 was not ol)scrved for the following reasons: the calcium isotope from which it would be produced has an abundance of 0.145% wliidi is low compared to that of calcium-44, 2.06%. Apparently the cross section of the reaction Ca48(p,n)Sc43 is too small to compensate for the lower concentration; it emits only a 0.37-m.e.v. gamma-ray for about 217c of its disintegrat'ions, so that the intensity of this gamma-ray would be too small to be distinguished from the high Compton continuum caused by higher energy gainnia-rays emitted by other nuclides; it has a half-life so close to that of scandium44 that its positron annihilation radiat,ion cannot be resolved froin that of scandium44 by decay measurements alone. Because of the above last-mentioned reason, it was convenient to consider the mixture of scandium-43 and scandium-44 as a single radioactive species. The half-life was found to be 4 hours whether measured by the decay of the positron annihilation radiat'ion or the decay of the 1.16 n1.e.v. gainma-ray. The Effect of Irradiation Duration. The intensity of the 4-hour activity was taken as a measure of the total natural calcillm contcnt of the sample. The time-delicndent factor (1 - e-'~')/ (1 - e-',,() in Equation 2 could then he considered as relating to two iadionuclides. The vaiiation of this timedependent factor with the irradiation duration is shown in Figure 3 from nhivh it can be scen that the yicld of scantliuin48 relative to that of the 4-hour activity (scandium43 and scantliuin-44) would change slowly with tinic. An irradiation of 20 hours would approximately double the act,ivity ratio obtained after 2 hours, but despite the increase in sensitivity this would offer, the increased irradiation cost would not be warranted. Accuracy and Precision. The nctivity ratio of scandium-48 to the 4hour activity a t t,he end of the irratlia-
Table 11. (1)
Known Ca4* concn. (atom %)
Activity ratio
( x 100)
0.11
-
10.6
-
I -
20.4-
%
0.1
r
tion changcs linearly with calciuni-48 content, SO t h a t one standard sample and a sample of natural compocition are sufficient to fix the calibration line, provided targets of approximately the same thickness are irradiated for a fixed duration. If the variation in thickness is such as to cause an appreciable variation of the proton energy within the target from sample to sample, the change in cross section of the (p,n) reaction at the lower proton energies would affect the activity ratios differently, thereby giving erroneous results. A variation in the length of irradiation \vould result in a change in the activity ratio from sample to sample as ivell, but this could be corrected for by using the appropriate values from Figure 3. Table I1 shows a typical set of results from an irradiation of 2 hours a t a proton energy of 4.75 m.e.v. The relative stantlard deviation was 2.7Oj, while the mean value of the activity ratio in this caw was 0.252 per atom 7@ caloium-48. The relative standard dcvintion is a measure of the precision of the method, but the absolute value of the activity ratio changes with the length of irradiation and may further vary if the irradiation flu\: undergoes a1)prcviable change during irradiation. The precision was thus sufficient to measure a 107c increase over the natural c+al(.iuni-48isotopic concentration. The mean error, -0.004 atom %, \vas suf
A Typical Set of Results Activity ratio
per unit
atom
Yo
4.75 0.257 4.83 0.252 9.32 0.249 0.262 21.57 0.254 26.38 0.242 33.15 &lean activity ratio 0.252 f 0.007 per iiiiit atom yo Relative st,andard deviation *2.7% Mean error -0.004 atom 70 0.185 0.192 0.375 0.825 1.040 1.370
958
ANALYTICAL CHEMISTRY
m
10
IR~POIATIQN1lME (MW4
Figure 3. The variation of the timedependent factor (I - e - V ) with irradiation time (1 A, and XI refer respectively to the 4 and 44-hour activities
(11)
hIeasurecl
Ca48C O I I ~ I L (atom 7;) 0.188 0.191 0.369 0.854 1.045 1.313
Okrved difference (11) - (1)
+0.003 -0.001 -0.006 + 0 , 029 0.005 -0.057
+
ficiently close to zero to indicate no bias. The main source of error affecting the accuracy of a determination would be contamination from other calcium salts, which would not be immediately obvious during activity measurements, and it would thus require stringent precautions to be taken to prevent such cont.amination. Contaniination from other sources would have very little effect. If any contamination occurs before the evaporation of the calcium fluoride, the contaminant is unlikely to follow calcium through the chemical stages. Should the contamination ocrur after the target is prepared, it would still be possible t o discriminate against it by half-life and energy measurements. It should be noted that the elements likely to be activated by proton irradiation at 4.75 m.e.v. would be those lighter than calcium, and the activities they would produce would not interfere in the analyses. The coulomb barrier for elements heavier than calcium would tend to decrease the activation cross sections thus still further discriminating against the contaminant. Advantages and Limitations. The method offers a simple procedure for determining isotopic concentrations of calcium-48, and, in tjhe sense t h a t the analyzed sample is still available after analysis, is nondestructive. The procedure, being h s e d on the determination of activity ratios, does not require yields, weights of targets, or irradiation fluxes to be known. If samples are irradiated successively or singly, it is necessary that the irradiation flus remains vonstant for all the samples. However, when the samples and stantlards are irradiated together, as is possible by the use of the rotating target holder shoivn in Figure 1, flux variations would only result in a calibration line with a different slope. It is essential, that the standard and sample should be counted under identical conditions arid that the irradiation duration be known. The method is applicable to samples of natural isotopic concentration of calcium-48 (0.18%) and samples enriched with cal(-ium-48. When the isotopic concentration of calcium48 is about 47c, the count-rates of the scandium48 and the 4-hour activities are about equal. At higher enrichments the yield of scandium-48 becomes so great that the error involved in memuring the shorter-lived activity lowers the precision of the method. Accordingly the method is best suited for low isotopic concentrations of calcium-48. It would thus be advisable to dilute very highly concentrated samples with a known amount of natural calcium. The method is not suitable for microamounts of calcium. Approximately a milligram of calcium fluoride is the least
needed to prepare the target. However, this amount is readily obtainable from most biological systems where the method would find ready application. ACKNOWLEDGMENT
The assistance of Robbie Verbruggen especially with the design and construeis tion Of the rotating gratefully acknowledged.
LITERATURE CITED
(1) Everling, Fa,Koenig, A*,Mattauch, J. H. E., Wapstra, ,,A. H.. “Nuclear Data Tables, Part I, National Academy of Sciences, Washington, 1961.
(2) Junod, E., Laverlochere,J., “Proceed.
3rd Intern. Colloquium on Biology,” Saclay, 1963. (3) Koenig, L. A., Mattauch, J. H. E., Wapstra, A. H., “Nuclear Data Tables Part 11,”National Academy of Sciences, Washington, 1961.
Radiochemical Separations
(4) Peisach,
M., Pretorius, R., unpublished data, Southern Universities Nuclear Institute, Faure, C. P. (1965). ( 5 ) Strelow, F. w. E., Staerk, H.9 ANAL. CHEM.35, 1154 (1963).
RECEIVEDfor review March 9, 1966. Accepted April 18, 1966. R. P. thanks Dr. H. P. Malan, his Promotor a t the University of Stellenbosch, for permission to publish his results which will form part of his doctorate thesis to be submitted to the University of Stellenbosch.
by Isotopic Ion Exchange
FOUAD TERA and G. H. MORRISON Department of Chemistry, Cornell University, Ithaca, N. Y .
b A , general approach has been developed for the rapid separation of radioactive species, which has particular application to matrix removal in activation analysis. Ion exchange and isotopic exchange operate simultaneously to effect multitrace element separation from large amounts of a radioactive matrix element. The technique has been applied to the activation analysis of trace elements in barium nitrate and sodium nitrate samples.
R
separations are essential for much of the work with radionuclides in nuclear chemical characterizations as well as in many applications of nuclear techniques such as activation analysis. The goal of these separations is to purify the desired radioisotopes for measurement of activity. Many excellent radiochemical procedures exist for the isolation of specific radionuclides (S), but they are generally time-consuming and any one method is usually very limited in scope. Isotopic ion exchange can be used for the rapid separation of interfering radioactive matrices in activation analysis permitting multielement trace determinations, as well as for the separation of radionuclides with similar radiations which cannot be easily resolved by physical means. Ion exchange and isotopic exchange are simultaneously operating to effect the separation. When the major component AI* (interfering matrix) of the radioactive mixture has a high selectivity for the cation exchange resin employed, the concentration of this component is adjusted so that the distribution coefficients of the other trace radioactive components are greatly suppressed. Upon passing the mixture through a column previously saturated with the same cation in the stable form ADIOCHEMICAL
(R-M), the major radioactive component M * will be retained on the column by isotopic exchange while the other radioactive components are easily eluted. The eluate may then be analyzed by gamma spectrometry or other suitable counting methods. Alternatively, if a major radioactive component N * has a low selectivity for the resin, the radioactive mixture is passed through a column previously saturated with any stable cation of high ion exchange selectivity (R-M). Under these conditions the other trace radioactive elements are held by ion exchange as well as small amounts of N * . By passing a solution containing stable N , the residual amounts of N * will be completely removed from the column by isotopic exchange. The radioisotopes retained on the column may then be determined by gamma spectrometry or other suitable counting methods. When two radioactive trace components are to be separated, one of them can be converted to a major component by the addition of its stable cation and either of the above two approaches applied depending upon its relative position in the ion exchange selectivity scale. The factors governing isotopic ion exchange have been investigated and the technique has been applied to matrix removal in activation analysis. Barium and sodium matrices were chosen as representative of larger groups of elements to which the approach is applicable. EXPERIMENTAL
Reagents and Tracers. All reagents used were analytical reagent grade. Dowex 50-X8, 100-200 mesh (J. T. Baker Co.) cation exchange resin was employed for column operations and was purified by passing excess 12M and 6-V HC1 and washing with distilled deionized water until a negative
acid test was obtained. Part of the resin was then transformed to the desired metal form by passing a large excess of the metal salt solution, then washing with distilled deionized water until a negative test for the metal cation was obtained. Radioactive tracers S a z 2 Rb*6 C ~ 1 3SF, ~ , Ba133 Y91 Ce141,CrL1, Mn644: Fe59, C060, Zn65,’Ag1;0, and Cdlo9 were obtained from Oak Ridge Xational Laboratory or Isoserve, Inc. Tracer Studies. APPARATUS. Eluate fractions were obtained using a Research Specialties Co. automatic fraction collector. Radioactive counting equipment consisted of a SaI(T1) well-type scintillation crystal and Baird-Atomic scaler. BATCHPROCEDURE. Approximately 1 ml. of water-soaked resin was measured exactly and transferred t o a 125ml. Erlenmeyer flask and evaporated to dryness a t -80’ C. Twenty-five milliliters of 0.05N HC1 containing a known amount of a matrix cation and a known radioactive tracer, either M* or t * , were then added. The flasks were agitated with a mechanical shaker for a period of 12 to 20 hours a t room temperature. The resin was always in the matrix form. Appropriate aliquots were then taken and the activity was determined by scintillation counting. The volume distribution coefficient for ion exchange of the trace elements is given by Dv = [ t * I T / [ t * ] , , where t* represents the radioactive trace element and the subscripts r and o refer to the resin and liquid phases, respectively. When the radioactive tracer is M*, a radioactive isotope of the matrix element JI, its isotopic distribution is given by Dv* = [ X * ] , / [ J I * ] , = [.Ml,/[Ll1IO. C o L u m PROCEDURE. Separations * were performed with the resin in the Ba+2, C O + ~and , Ca+2 forms. In all separations the column diameter was 6 mm. and the resin height 4 cm. Disposable Pasteur pipets with glasswool plugs were employed as columns. All solutions were 0.05W in HC1. To determine the breakthrough and elution behavior of matrix and trace radioactive elements, a stock solution VOL. 38, NO. 8, JULY 1966
0
959