The principal contaminants were iron, calcium, sodium, and silica. Under neutron activation t,hese elements undergo act'ivation reactions, in the manner shown in Table 111. Silver was present, only as a trace constituent, but owing to its sensitivity to neutron activation, it may be evident on a direct gamma spectrum. If the y spectrum is plotted after a decay period of 30 days, the main yemitting nuclides will be ZnB5and Fe59. These spectra are shown in Figure 1. The 1.1.-m.e.v. gamma peak occurs on both spectra. For a full spectrum of a head sample, containing approximately 5% iron, this iron cont'amination can be estimated and corrected for; a typical spectrum of a head ore, after 80-days decay, is shown in Figure 2. Whew a reactor is used as a source of neutrons it must be remembered that both fast and slow neutrons are present. Hence it is possible for radioact'ive zinc to be produced by other than t'he process described. The following have been reported in the literat'ure ( 4 ) :
Ga69( n ,p ) Zneg Ge72(n, d ) Zne9 Zn70 (n, an) Z d 9 .
mbarns 0 20 mbarn
U , 24
u,
Both gallium and germanium were below 20 p.p.m. in these materials. This, together with the low cross-section for the reactions producing zinc-69, makes any contribution to the zinc activity negligible. The contribution from the zinc-70 can likewise be ignored. In the case of determining zinc in the lead concentrates,. the difficulty of selfshielding is very apparent. For an accurate determination in this type of material, therefore, it is clear that a standard addition must be made to the sample material in order to assess a true value for self-shielding and absorption effects. I n the direct gamma spectrometric determinations, it is clear that the contamination in both the lead and zinc tail samples is too high for compensation: the chemical separations are, therefore, essential. In cases where a spectrometer is not
available it is sometimes possible to operate the counting equipment to count only those energies above a preselected value. This was attempted by setting the lower limit a t 1.0 m.e.v. Where no contamination due to higher energies was present, as in the case of the zinc concentrate samples, this gave a fair result. but in all other samples too much contamination was present. LITERATURE CITED
( 1 ) Bankes, T. E., Tupper, R., White. E. M. A., Wormall, A., Intern. J . A p p l , Radiation Isotopes 4 , 221-6 (1959). ( 2 ) Bowen, H. J. M., Ibid., 4 , 214-20 (1959). ( 3 ) Hevesy, G. yon, Kgl. Danske Videnskab. Selskab Mat.-fys. M e d d . 14, 5 i1936). (4)Paul, E. B., Clarke, R. L., Can. J . Phys. 31, 267 (19,53). ( 5 ) Pierce, T. B., Perk, P. F., Analyst 87, 367-73 (1962). RECEIVED for review Xovemher 18, 1963. Accepted January 31, 1964. IVork described in this paper forms part of a research program of Warren Spring Laboratory, D.S.I.R., Stevenage, and is published by permission of the Director.
Determination of the Zi nc-68-Zi nc -64 Ratio in Rocks and Minerals by Neutron Activation Analysis R. H. FlLBY Mineralogical-Geological Museum, University of Oslo, Oslo, Norway
b A neutron activation method for the microdetermination of the Zn68/Zn64 ratio in rocks and minerals is described and discussed. Samples are irradiated in a thermal neutron flux of approximately 1 0l2 neutrons per sq. cm. per second and dissolved, inactive zinc carrier is added, and the zinc is separated in a radiochemically pure state by anion exchange and ferric hydroxide scavenging. The ZnBgm/Zns5 activity ratio (relative to a standard zinc oxide sample) gives a measure of the Z d 8 / Z d 4 ratio. Precision ranges from 0.3% (relative standard debiation) for samples with high zinc contents to approximately 1 % for samples with low zinc contents. The sensitivity is estimated as 0.4 p.p.m. of zinc. The method is rapid, specific, and not subiect to serious contamination problems. Results for a sphalerite, granite G-1, and diabase W-1 indicate that there is no significant difference in the Zd8/Znfi4ratios relative to the standard zinc.
M
spectrographic methods for the determination of the isotopic compositions of elements have been used almost exclusively because of the ASS
high precision of which the methods are capable. For very low concentrations of elements in geological materials, however, the separation of the required amount of the element for mass spectrographic analysis is extremely tedious and a large sample is commonly required. Contamination from reagents and apparatus can thus be a serious problem and for this reason very few determinations of the isotopic compositions of trace elements in geological materials or meteorites have been made. Very little is known as to whether isotopic fractionation occurs in geological processes-for example, metamorphism or magmatic differentiation. Rankama ( 2 1 ) has suggested that isotopic fractionation of elements with masses up to 80 might occur in nature, given the right conditions. Klemm ( 5 ) has reported a 5.5y0 increase in the Cu"/ Cum ratio during the diffusion of copper in the system a-Cu2S-a-Xg2S a t 450' C., which suggests that solid-state diffusion in geological processes might lead to isotopic fractionation. Neutron activation analysis has been mentioned as a precise and accurate method for the microdetermination of isotopic ratios for certain elements (9). The neutron activation method is
feasible if, on thermal neutron irradiation, two nonadjacent isotopes of an element give rise to radioactive nuclides which are capable of being measured independently. The requirements of a suitable neutron activation method may be stated as follows. The thermal neutron cross sections of the isotopes involved must be large enough to give sufficient activity for low concentrations of the element. The half lives of the products must be short enough to permit irradiation times of less than 1 month and yet long enough to allow the necessary chemical separations to be made after irradiation. The products should have modes of decay that allow selective measurement. This is most conveniently achieved by y-ray spectrometry if the products have distinctly different y-ray spectra. The products should not result from other nuclear reactions on other elements present in the sample. Particular attention should be paid to fission products of the Uz35(n,f) reaction and to (n,p) and (n,a) reactions that may occur with the fast neutron flux of the reactor. The main advantages of the neutron activation method compared n ith other methods of isotopic analyqi. are high VOL. 36, NO. 8, JULY 1964
e
1597
Table I.
Stable isotope Znec Zn6* Zn@ Zn70
Abundance, 7, 48.89 18.56 18.56 0.62
Thermal Neutron Capture Reaction? of Zinc Isotopes
Cross ( n , y) Mode section, reaction of barn product decay 0.44 ZnG5 E C , p * , y 0.1 ZnGgm IT, y 1.0 Zn6@ pZn71m p-, y 0.09 ~~
Zn70
0.62
Zn71
p-,
y
sensitivity, freedom from contamination, and simplicity. Because of the high sensitivity of most neutron activation methods, isotopic ratios can be determined, where feasible, for elements present at very low concentrations without chemical separation of the element before irradiation. As no separation before irradiation is necessary, the danger of ' contamination is greatly reduced. Determinations have recently been made of the Hg196/Hgzoz,PbZ04/Pbzo8, Ba131/Bai35 ( l a ) , and Th230/Th232 (13) ratios in meteorites relative to terrestrial material and it is expected that other geochemical applications will be reported. No measurements of the isotopic composition of zinc in rocks have been published, although determinations of zinc isotope ratios in zinc minerals have been reported (1). Fractionation may occur in nature and the problem is thus worth studying. Zinc is a suitable element for the determination of the Zn68/Zn64ratio by neutron activation, as the two isotopes, Zn68 (abundance 18.56%) and ZnB4(abundance 48.89%), give 13.8-hour Znfiga and 245-day Znfi5 on thermal neutron irradiation. The two isotopes have sufficiently different y-ray spectra to facilitate selective determination by y-ray spectrometry. The thermal neutron capture reactions of the zinc isotopes are listed in Table I. I n the method discussed in this paper, the samples were irradiated in a high thermal neutron flux and dissolved, inactive zinc carrier was added to the solution and the zinc was separated in a radiochemically pure state. Zn65 and Zn69mwere measured by y-ray spectrometry after Zn71mhad been allowed to decay. Znfigm was chosen in preference to Zn6gbecause of the longer half life, although the cross section of the Zn68 (n,y)Zn69mreaction is less than that of ZnM (n, y)Zn". Also, Zn69 is a pure 8-emitter, thus making accurate measurement more difficult. EXPERIMENTAL
Apparatus and Reagents.
A Su-
clear-Chicago single-channel y-ray spectrometer (No. 1820) and scintillation counter with a 2 X 2 inch KaI(T1) well-type crystal were used. Ion exchange columns, 15 X 0.i5 1598
ANALYTICAL CHEMISTRY
Beta energies, m.e.v.
Gamma energies, m.e.v.
Half life 0 . 3 ( 1 . 5 % ) 1.12, 0.511 245 day None 0.44 13.8 hr. 0.90 None 52 min. 1.5 0.38, 0.49, 3 . 0 hr. n AI 2.4
6;ii,0 . 5 1
2 . 2 min
em., were prepared from Amberlite CG-400 (100- to 200-mesh) anion resin in chloride form and were pretreated with 1.2M HCI. Zinc oxide reference material was Johnson, Matthey and Co., Ltd., Specpure. Zinc carrier solution was prepared from Merck (analytical grade) zinc sulfate; concentration 5 mg. of zinc per ml. Holdback carrier solution contained 0.5 mg. each of copper, cobalt, gallium, manganese, and nickel per ml. in 1.2X HC1. Procedure. Weigh out 0.5 to 1.0 gram (for samples containing 1 to 100 p.p.m. of zinc) of the powdered rock sample onto a 5-cm. square piece of aluminum foil (washed with water and then with acetone) and carefully wrap the sample in the foil. Wrap each sample in a further piece of foil. For zinc minerals, weigh out 50mg. portions. Weigh out 10-mg. aliquots of the standard zinc oxide and enclose in aluminum foil envelopes, making sure that no contamination of the samples is possible. Pack the samples, with the standards distributed uniformly, in an irradiation can and irradiate in a thermal neutron flux of 2.5 X 10l2 neutrons per sq. em. per second for one week. After irradiation, allow the samples to cool for a t least 2 hours, carefully unwrap each aluminum envelope, and transfer the material to a platinum dish (or beaker for standards), taking samples with the lowest zinc contents first. It is extremely important to avoid contamination of the samples by the standards. Proceed with the chemical separations as follows. ROCKSAMPLES.Transfer the sample to a platinum dish and add 10 ml. of zinc carrier solution, 10 ml. of perchloric acid, 5 ml. of sulfuric acid, and 15 ml. of hydrofluoric acid. Evaporate the solution to approximately 5 ml. and then cool. Wash the sides of the dish with a little water, add 3 ml. of perchloric acid, and evaporate to so3fumes. Add 1 ml. of water and then evaporate to near dryness. When cool, add 5 ml. of concentrated hydrochloric acid and 10 ml. of holdback carrier solution and warm gently until the residue has dissolved. Make up the solution to 50 ml. with water. If any sample is undecomposed, collect and fuse with sodium carbonate in a platinum crucible. Dissolve the melt in a minimum of water, neutralize, and add concentrated hydrochloric acid to make 1.224. Combine with the main solution.
Pass the solution through the ion exchange column at a flow rate of 2 ml. per minute and wash the column with 25 ml. of 1.2M hydrochloric acid. Elute the zinc from the column with 50 ml. of 0.01M hydrochloric acid a t a flow rate of 1 ml. per minute and to the eluate add 1 ml. of a 1% solution of ferric chloride, 1 gram of ammonium chloride, and 5 ml. of ammonium hydroxide. Heat to boiling and then place the beaker on a steam bath for 5 minutes. Filter off the ferric hydroxide and discard. Add 1 gram of thioacetamide to the filtrate and heat to boiling. Boil gently for 10 minutes, filter off the zinc sulfide on a sintered glass crucible, and wash the precipitate with a 0.1% solution of ammonium chloride. Dissolve the precipitate on the filter with 5 ml. of 6M hydrochloric acid and boil the solution to remove the HzS. Cool the solution and dilute to 25 ml. with water and then pass the solution through an ion exchange column. Wash the column with 25 ml. of I.2M hydrochloric acid and elute the zinc with 50 ml. of 0.01M hydrochloric acid. Precipitate the zinc from the solution as sulfide as described above and collect the zinc sulfide in a glass counting vial. Dissolve the zinc sulfide in the glass with 631 hydrochloric acid. ZINC SULFIDE SAMPLES.Dissolve the sample with 25 ml. of 1.2M hydrochloric acid and boil off the H2S. Cool the solution, add 10 ml. of the holdback carrier solution, and then pass through the ion exchange column. Wash the column with 25 ml. of 1.2M hydrochloric acid and then elute the zinc with 50 ml. of 0.01M hydrochloric acid. Add 1 gram of ammonium chloride, 1 ml. of a 1% solution of ferric chloride, and 5 ml. of ammonium hydroxide, and then heat to boiling. Allow the ferric hydroxide to coagulate and then filter the solution. Wash the precipitate with a 1% solution of ammonium chloride and make up the filtrate with water to 100 ml. in a volumetric flask. Take aliquots of the solution for counting. STANDARDS.Dissolve the zinc oxide standard in 25 ml. of 1.2M hydrochloric acid, add 10 ml. of holdback carrier solution, and then pass through an ion exchange column a t a rate of 2 ml. per minute. Wash the column with 25 ml. of 1.2M hydrochloric acid and then elute the zinc with 50 ml. of 0.01M hydrochloric acid. Make up the eluate to a final volume of 100 ml. in a volumetric flask. Counting. Approximately 15 hours after irradiation Zn71m has decayed and counting can begin. Record y-ray spectra of all samples and standards after separation of the zinc. Check the radiochemical purity of each sample. Count each sample on the 0.44- and 1.12-m.e.v. y-ray photopeaks of Zn6ga and Zn65, respectively, and take an aliquot of the standard zinc rolution for counting, so that the count rates of sample and standard are approximately equal. Equalize the volumes in all counting vials with water, so that the counting geometry is the same for all samples. Count the samples and standards alternately 10 times on the Znegm
photopeak to eliminate variations in the counting system. Record a t least 40,000 counts for each measurement. Repeat the procedure for the Zn65 photopeak. After 1 week, when all Zn6gm has decayed, count the samples and standards on the photopeak position to obtain the background due to Zn65. Subtract the values, corrected for decay of Zn65, from the Zn6qmreadings. Conirilite the value of r for each sample and zlandard from r =
activity of Zn6gma t arbitrary time t' activity of Zn65 a t arbitrary time t ' .
The ratio, R , where R
= Tstandard
gives
a measure of the Zn68/Zn64 ratio of the sample relative to that of the standard. Assuming isotopic abundances of 18.56% for Zn@ and 48.89% for Zn64 ( I C ) , calculate the Zn@ and Zn'j4 values for the sample. RESULTS
The method was used to determine the Zn68/Zn64 ratio of a sample of zinc oxide (British Drug Houses, AnalaR). a sphalerite (ZnS) of sedimentary origin from Yorkshire, England, and a sample of zinc sulfate (Merck, analytical grade). The ZnMjZnB4ratios were also determined for two standard rocks, granite G-1 and diabase W-ll containing 40 and 82 p.p.m. of zinc, respectively (E). I t can be seen from Table I1 that there is no difference in the Zn68/Zn64 ratios for the samples and the standard zinc oxide, within the limits of experimental error, and no significant difference between the isotopic ratios for G-1 and W-1, and the standard. DISCUSSION
Table II.
Difference Weight ZnS8m/Zn6j irradiated, relative to Zn6*/Zn64 from Saniple mg. standard ratio standard Std. dev. ZnO standard 40 1.000 0.3796 ... ... ZnO (B.D.H.) 40 1,003 0.3807 0.0011 0.0012 ZnS (Yorks) 100 1.004 0.3811 0.0015 0.0011 ZnS047H,0
(Nerck)
100 1.003 320 1.012 w- 1 187 0.9905 a Accepted values for half lives of tively (6). G- 1
Z r ~ ~ ~ ( n , p 12.8 ) C u hours, ~~ u
< 10 pb. (6)
Z n 6 ' ( n , p ) C ~61 ~ ~hours, u = 1 3 mb. (6)
Pvlellish (8) has reported that for the irradiation of zinc in the Harwell BEPO reactor a C d 4activity equivalent to 700 p.p.m. of copper was produced by the reaction. Possible (%,a)reactions are
< 20 pb. 6 hours u < 20 pb.
Zn66(n,a)Si6380 years u
(6)
Zn68(n,a)Ni652
(6)
The standards must therefore be purified after irradiat,ion to remove copper and nickel activities. I n the samples, Zn6j and Zn6gm may be produced by interfering reactions. Zn'j5, for example, may be produced by the secondary reaction
The ZnG5contribution from this reaction relative to that produced from natural zinc may be calculated from the equation (4) Zn6j activity from Cu63 Zn65 activity from Zn6'
Half lifea Zn69m,
zn,.
hours 13.7
days 238
13.8
. ..
13.8
. . .
0.3807 0.0011 0.0015 13.9 . . . 0.3842 0.0046 0.0053 13.7 249 0.3761 -0.0035 0.0057 13.7 255 and ZnGSare 13.8 hours and 245 days, respec-
mentally. Samples of Specpure zinc oxide of different weights were irradiated simultaneously and the relative Zn6gm/ ZnG5ratios obtained are shown in Table I11 Examination of the results in Table I11 shows that samples of different weights gave the same Zn69m/Zn65ratios, within the limits of experimental error. In the standards nuclides other than ZnG5and Zn6g/6gmmay be produced by (n,2n), (n,p),or (n,a)reactions with the fast neutron flux of the reactor. Any 38-minute Zn63 produced by the reaction will have decayed before counting begins and thus this reaction may be ignored. Possible (n,p) reactions are
OF METHOD
I t follows from the equation relating the induced activity, D,to the thermal neutron flux, f, the thermal neutron capture cross section, u, and the number of target atoms, N , D = ujN(1 - e - x L ) that the Zn6gm/Zn65activity ratio is directly proportional to the Zn'/Zn64 ratio. Flux variations which may occur between the sample and standard should not affect the activity ratio, provided that the energy distribution of the neutrons does not change, causing a The change in the u C ~ / U ~ + ratio. thermal neutron cross section for zinc is 1.1 barns and thus significant selfshielding in the standards is unlikely to occur. Differential absorption of resonance neutrons by the zinc isotopes can cause differences in the ratios of nuclides produced from samples of different sizes. Examination of the cross-section data for the zinc isotopes (3) indicates that this effect can be ignored. However, as the neutron energy distribution of the irradiation position mas not known, the influence of sample size was investigated experi-
Determinations of Zn68/Zn64Ratios
Table 111.
Effect of Size of Sample on ZnM/Zn64 Ratio
Zn69m(Znes ratio
Weight of ZnO, mg. 140 (standard) 41 6 32 0 10 0 0 1
relative to standard 1 000 0 997 1 001 1 004 1 014
Std. dev. 0 004 0 004 0 003 0 004 0 013
where Nc" and N z n = number of copper and zinc atoms in sample f = neutron flux, neutrons sq. cm./ second t = time of irradiation, seconds u = cross section for C U ~sq. ~ ,cm. A = decay constant for Zn65
A 10-day irradiation for a sample with a Cu/Zn ratio of 3.5 X lo4 would give a Zn65activity from the interfering reaction equal to approximately 0.1% of the total ZnG5activity. This interference is therefore unimportant, except for copper-rich samples. Zn65 may also be formed by the fast neutron reaction Zn66(n,2n)Zn65,but as the variation of Zn66 in nature will be small and the cross section of the reaction with reactor neutrons very low, the interference will be negligible. Zn69m may be formed by the fast neutron reactions Ga6g(n1p)Znagm Ge72(n,CY) Zn69m The cross sections of these reactions with reactor neutrons are low and interference will be negligible except for samples of very high gallium or germanium contents, a n unlikely occurrence. If interference from gallium or germanium is expected, the samples can be irradiated in a pure thermal neutron flux. Neither Zn65nor Zn6gmis produced by thermal neutron fission of U235. VOL. 36, NO. 8, JULY 1964
1599
Decomposition of the sample is important,, .as anions present in the final solution may affect the anion exchange behavior of zinc or other elements. The presence of fluoride should not affect the adsorption characteristics of zinc, as the adsorption behaviors of zinc in hydrochloric acid and hydrochloric acid-hydrofluoric acid media are similar (10). However, the adsorption behavior of other elements may be affect,ed by fluoride; therefore fluoride was removed by evaporation with sulfuric-perchloric acids. Perchlorate was removed by evaporation t’o SO3 fumes. The chemical separation of zinc from other elements in complex silicat’e samples presents no great difficulty, as zinc can be separated by anion exchange in hydrochloric acid media. Examination of the adsorption behavior of the elements with a strongly basic anion resin ( 7 ) shows that if the adsorption step is carried out in 1.2M hydrochloric acid and zinc eluted with 0.01X hydrochloric acid, only P b ( I I ) , U(VI), and possibly In(II1) accompany the zinc. These elements can be removed after the anion exchange separation by precipitation with ammonium hydroxide, using ferric hydroxide as a carrier. The chemical yield of zinc, for the separation process ranged from 85 to 98%. Gamma-ray spectra of the zinc separated from G-1, W-1, and the zinc sulfide sample were identical to the standard zinc spectrum, indicating that the separation produced zinc in a radiochemically pure st’ate. This was further confirmed by the half-life data for ZnG5 and Zn69mobtained from the decay curves (Table 11).
SENSITIVITY A N D PRECISION
Standard deviations, coml)ut,ed from replicate determinations, are given in Table 11. As can be seen from t,he values of the standard deviations, precision is high and for the zinc-rich samples (sphalerite and zinc sulfate) the relative standard deviation is approximately 0.3Oj,. Differences of less than 0,5y0 in the Zn68/Zn64,ratio can thus be readily detected by the method. The lower count rates for the rock samples, G-1 and W-1, and the longer counting times involved are reasons for the lower precision for these samples. Differences of more than 1% in the isotopic ratio should be easily mea+ ured, however. Considerable improvement in precision could be obtained by using higher neutron fluxes. It, was experimentally ascertained that for a &day irradiation in a flux of 2 x lo1*neutrons per sq. em. per second, 1 gram of G-1 (40 p.p.m. of zinc) gave 20,000 c.p.m. for Zn69mand 5500 c.p.m. for Zn“, 33 hours after irradiation, with the counting conditions described. Thus for a 5-gram sample in a flux of 5 X 10l2 neutrons per sq. em. per second it should be possible to determine the Zn68/’Zn64rat,io for samples with approximately 0.4 11.p.m. of zinc with satisfactory precision. CONCLUSION
The neutron activation method described for the microdetermination of the Zn68t‘Zn64 ratio in rocks and mineralj can be used for samples containing as little as 0.4 p.p.m. of zinc. The method, although very sensitive. is free from contamination difficultieq.
Studies are now being made on a number of rocks of different geological origins to determine if there is any fractionation of the zinc isotopes during geological processes. Meteorites are also being examined for differences relative to terrestrial materials. LITERATURE CITED
V.,Wickman, F. E., Geochim. Cosmochim. Acta 11, 162 (1957). ( 2 ) Fleischer, >I., Stevens, R. E., Ibid., 26, 525 (1962). (3) Hughes, L>. J., Schwartz, R. B., L‘. S. At. Energy Comm., Rept. BNL325, 2nd ed. (1958). (4) Kant, A , , Cali, J. P., Thompson, H. D., AXAL.CHEM.28, 1867 (1956). (5) Klemm, A., 2. Physik. Chem. 193,29 (1943). (6) Koch, R. C., “Activation Analysis Handbook,” Academic Press, New Tork and London, 1960. ( 7 ) Kraus, K. A , , Selson, F., Proc. Intern. ( 1 ) Blix, R., Ubisch, H .
Conf. Peaceful Vses .4tomic Energy 7,
113 (1956).
(8) Mellish, C. E., Payne, J. A , , Otlet,
R. L., “Radioisotopes in Scientific Research,” Vol. 1, p. 35, Pergamon Press, London, 1958. ( 9 ) Merz, E., Herr, iV.j2nd Intern. Conf. Peaceful lyses Atomic Energy, Geneva 1958, A/Conf. 15/P/984. (10) h’elson, F., Rush, 11. XI., Kraus, K. A , , J . A m . Chem. SOC.82, 339 (1960). (11) Rankama, K., “Isotope Geology,” Pergamon Press, London, 1956. (12) Reed, G . W.,Proc. Radioactivation Analysis Symp. 1-ienna 1959, pp. 26, 48, Butterworths, London, 1960. (13) Rona, E., Trans. A m . Geophys. Union 38, 754 (1957). (14) Strominger, D., Hollander, J. M., Seaborg, G. T., Rev. M o d . Phys. 30, 585 (1958). RECEIVEDfor review January 13, 1964 Accepted Marrh 20, 1964.
Radiochemical Determination of Radium-228 and Thorium-228 in Biological and Mineral Samples HENRY G. PETROW, ARTHUR COVER, WALTER SCHIESSLE and ELIZABETH PARSONS Institute o f lndusfrial Medicine, New York University, Tuxedo, N . Y.
b A general radiochemical procedure for the determination of radium-228 and thorium-228 in biological and mineral specimens has been developed. The technique is sensitive and as little as 1 kpc. of radium-228 and 0.02 ppc, of thorium-228 can b e detected. Samples containing as much as 50 grams of ash can b e treated without adverse effects from either phosphate ion or calcium ion. Interference from other radioactive substances normally present in the environment is negligible. 1600 *
ANALYTICAL CHEMISTRY
A
AN OUTGROWTH of research directed toward the estimation of radium-228 and thorium-228 in the tissues of radium dial painters, this laboratory is now engaged in the determination of these nuclides in a large variety of environmental samples. The distribution of radium-226 in nature has been studied extensively. The concentration of radium-228, a beta emitter of low energy, has received far less attention, presumably becau5e of analytical difficulties. Some attempt
s
has been made to estimate the concentration of radium-228, via measurement of its granddaughter, 1.9-year thorium228, an alpha emitter ( 2 , 6). This is an unsafe practice. The relatively long half-lives of both nuclides, combined with their vast chemical differences, make translocation, relative to each other, a strong possibility. I n fact, thorium-228 cannot and should not be used to estimate radium-228 concentrations. Petrow and Allen ( 3 ) describe a method for the determination of radium-