Analyst, December, 1973, Vol. 98, pp. 873-885 873 Simultaneous Determination of Actinide Nuclides in Environmental Materials by Solvent Extraction and Alpha Spectrometry BY B. L. HAMPSON AND D. TENNANT (Ministry of Agriculture, Fisheries and Food, Fisheries Radiobiological Laboratory, Hamilton Dock, Lowestoft, Suflolk) Actinide analysis based on the selection of groups, instead of separation of individual elements, has been applied to monitoring and control of the increasing variety and amounts of actinide nuclides in environmental materials contaminated by controlled discharges of liquid wastes. Multi-element acti- nide analysis is achieved by extracting the whole group, or part of it, in the tri-n-octylphosphine oxide - n-heptane - nitric acid - sodium nitrate system, stripping into ammonium carbonate solution and electrodeposition, followed by solid-state alpha spectrometry, with unusual actinide nuclides as yield tracers. This system gives efficient separation from virtually all common elements. Common operations for the several functions of sample dissolution, tracer exchange, solvent extraction and electrodeposition have been developed, which are suitable for the group from actinium to curium.Detailed procedures, with variations, for simultaneous measurement of these actinides are stated, in order to allow application to biological materials and radioactive effluents in two combinations, the first in valency states 111, IV and VI, and the second in valency states IV and VI with valency state I11 eliminated. They have been fully evaluated for plutonium-239 $us -240 with plutonium-236 as tracer, and americium-241 with americium-243 as tracer, and the scope is indicated for other members of the group.Up to 2 kg of edible seaweed or 10 kg of fish flesh can be handled, with detection limits (in terms of activity to double background) of 2 x 10-6 and 4 x lo-7pCig-1, respectively, for a 1-week counting time. Sensitivities for precision with 4 per cent. standard deviation are 4 x lo-* and 8 x 10-5 pCi 8-1, respectively, which corresponds to levels associated with fallout. MEASUREMENT of actinide nuclides in environmental materials is becoming increasingly important as nuclear power programmes build up. Comprehensive actinide analysis is needed for control measures associated with the discharge of these radionuclides in liquid effluents, and also for radio-ecological research. The artificial nuclides of most concern are 238Pu, 239Pu, 240Pu, 241Am and 242Cm, although others, for example, 237Np, may occur.Members of the thorium and uranium series, which are constituents of the natural background, are conveniently measured at the same time. The range of sensitivity required varies from derived working limits in marine foodstuffs (typically of the order of lo2 pCi g-l) down to levels associated with natural activities or fallout (as low as In hazard assessment, all actinide nuclides should be detected and measured simul- taneously. The aim with most current chemical methods is to separate individual actinides, although with some several are separated sequentially, e.g. , by fractional elution from solid1 or liquid2 ion exchangers, followed by total alpha counting.Yield tracers have previously been used only in exceptional instances, e.g., 236Pu, so as to assess the recovery3 of 239+240Pu; their universal adoption is essential. Alpha spectrometry has been used in two ways for collective measurement of actinides in environmental materials, namely, with the Mayneord- Hill large-volume gridded ionisation chamber,4 and co-precipitation with barium sulphate com- bined with solid-state alpha spectrometry.5,6 However, the sensitivity of these methods (detector limits of about 0.1 pCi per g of ash) is insufficient for the materials mentioned above, because only small samples can be handled (in the latter method calcium, aluminium, etc.interfere). The present method combines highly selective actinide group separation with sensitive alpha spectrometry, by using yield tracers throughout , and sensitivity is increased at least 103 times. pCi g-l). @ SAC; Crown Copyright Reserved.874 HAMPSON AND TENNANT : SIMULTANEOUS DETERMINATION OF [Analyst, VOl. 98 A review of actinide separation indicates that neutral rather than ionic extractants are best suited to actinide group selectivity. Separation of all actinides together in a single operation cannot be achieved by ion exchange as both cation and anion exchange are required, with widely varying conditions for different actinides. The actinides are particularly prone to form strong ion-association complexes with neutral organophosphorus extractants, with reaction mechanisms not directly dependent on hydrogen-ion concentration (a main cause of differentiation with ionic extractants), and broad ranges of conditions are available for separation of actinides of valency states 111, IV and VI, or of sub-groups from almost all other elements under common conditions.The actinide selectivity derives from combination with the strongly nucleophilic phosphoryl 0 atom whose basicity is dependent on substituents, and is most highly developed in the c8 to C,, straight-chain trialkylphosphine oxides in similar hydrocarbon diluents. Electrostatic-type binding occurs with the actinide ion acting as a “hard” Lewis acid, and simultaneously n-bonding due to the actinide inner transition d orbitals. With the comparable nucleophilic amines, of which the tertiary and quaternary amines are the most potent, the latter effect is prevented.Of the many neutral extractants tri-n-octylphosphine oxide (TOPO) gives the highest extraction coefficients for americium(II1) ,7 c~rium(III),~ plutonium(IV)s and uranium(V1) .8 Of the TOPO - nitric, hydrochloric and sulphuric acid systems the nitric acid systems are the best c h ~ i c e , ~ as they give the most efficient extraction and separation of actinides from other elements. Nitric acid systems are also the most satisfactory for the best extraction of most actinides from quaternary amines, although not always from tertiary amines. However, in all instances, the use of nitric acid systems is necessary for effective separation from iron(II1) and other matrix elements.TOPO gives higher extraction efficiency for actinides of valency states 111, IV and VI, which is much higher for tervalent actinides, than the optimum tertiary and quaternary amines, and better separation from iron(II1); in fact, it gives effective separation from all significant biological matrix elements. EXPERIMENTAL Because of the separate demands of spectral resolution and sensitivity, the actinides must be extracted together, with high distribution coefficients for effective volume reduction and high separation factors from matrix elements so as to provide thin sources from massive samples. Some of the separation factors from individual interfering elements need to be very high, e.g. 107 for iron, and the over-all factor lo6 or more.Actinide valencies vary with redox conditions in such a way that in order to deal with the group from actinium to curium under common conditions it is necessary to separate actinides in a t least the 111, IV and VI valency states together. The behaviour of the bivalent oxy-cations is often intermediate between that of the ter- and quadrivalent cations. Most of the lighter common elements of valency I, I1 and I11 will be present in the matrix, but lanthanides and elements of higher valency, e.g., titanium, zirconium and niobium, can generally be ignored. The central issue is to obtain adequate separation between tervalent actinides and tervalent iron and aluminium, which tend to overlap in a scheme in which the elements of high valency are separated from the remaining elements.Sample dissolution, chemical exchange between sample nuclides and yield tracer nuclides, pre-concentration of actinides from bulky samples and removal of silica, solvent extraction and stripping and electrodeposition must each be carried out by successive processes that are commonly valid for all actinides. Uncommon nuclides used as tracers include 236Pu, 243Am and others (see under Scope). Actinides may co-exist in several valency states including both simple and oxygenated ions, which hinder free interchange with tracers. Some, for example, plutonium(IV), which are intimately bound to the insoluble fraction of ash from biological material,l* may be in polymeric or refractory form. Complete dissolution and tracer exchange are ensured by first reducing all actinides to the lowest states that can exist in aqueous solution, followed by re-oxidation.Boiling hydrochloric acid complexes most actinides well and was chosen in preference to nitric acid or nitric - hydrofluoric acid mixture+ for leaching the main salts and reducing the ferri-manganese component of ash from biological material. For reduction ascorbic acid was preferred to magnesium and agents of comparable potential,12 as its action is prolonged and the products are eliminated in the subsequent oxidation. Direct conversion Extraction with TOPO meets the above requirements.December, 19731 ACTINIDE NUCLIDES IN ENVIRONMENTAL MATERIALS 875 of plutonium and neptunium into valency IV,13 and uranium into valency VI by the nitrite - nitrate systems,14 leaving americium and curium in the tervalent state, was satisfactory, and gave results that were similar to those obtained by preliminary oxidation of uranium, nep- tunium, plutonium and americium to valency VI with silver peroxide and persulphate.12 The nitrite - nitrate stabilisation was preceded by hydrogen peroxide treatment so as to prevent formation of nitroso compounds of iron(I1) ; it also brings plutonium and neptunium to valency IV.Pre-concentration of the actinides from large samples of ash from biological material disposes of the bulk salts, and so simplifies the removal of silica; if left, these salts would form insoluble fluorides that carry actinides, and would interfere during extraction by com- plexing action and precipitation, as would phosphates.Sensitivity is improved by selective separation of tervalent actinides from iron at this stage. After the initial equilibration, two quantitative co-precipitations (to combine the actinides from the soluble and siliceous frac- tions) were adopted in order to suit the two variants of the procedure, according to whether tervalent actinides were included or excluded. With the latter the pH was raised to about 2 to permit the formation of a small amount of iron(II1) phosphate precipitate, which carried plutonium(1V) quantitatively but was unsuitable for tervalent actinides, because their phosphates precipitate at higher pH than the corresponding iron compounds. With the former at pH 1.5, oxalate carried all actinides without phosphate precipitation, kept iron complexed in solution, and was subsequently destroyed by nitric acid. Complete elimination of silica ensures the release of bound actinides and prevents their removal by colloid formation during the subsequent extraction.The hydrofluoric acid digest is evaporated with nitric acid in order to minimise the level of fluoride and the residue dissolved in nitric acid - boric acid mixture so as to complex any remaining fluoride before repeating the treatments with ascorbic acid and nitrite in order to effect exchange and stabilisation of valency state. CONDITIONS FOR ACTINIDE EXTRACTION WITH THE TOPO - NITRIC ACID SYSTEM- Distribution coefficients for actinides and matrix elements in this system are influenced by several factors, nitric acid concentration, the nature of the cation, the total nitrate-ion concentration, the TOPO concentration and nature of the diluent being the most important.By choosing appropriate conditions, separation procedures can be devised to meet particular needs in respect of actinide selection and matrix element rejection, For the present purpose two selections have been made: (1) all actinides together with valency states adjusted to 111, IV and VI, and (2) actinides of valency states IV and VI with I11 eliminated. The latter combination is needed so as to discriminate between plutonium-238 and americium-241, whose spectra could not be resolved from a mixture. The conditions chosen for full actinide group selection were 0-4 M nitric acid - 4 M sodium nitrate for biological materials or 0-2 M nitric acid - 4 M sodium nitrate for effluents, with an organic phase of 0.5 M TOPO in n-heptane. For actinides of valency states IV and VI with I11 eliminated the phases were 2 M nitric acid and 0.1 M TOPO in n-heptane.The first system gives distribution coefficients of lo2 to lo3 for actinides of valency state I11 and much higher values for actinides of valency states IV and VI, while the latter gives distribution coefficients of about for americium(III), curium(III), etc., but more than lo2 for plutonium(1V) and plutonium(VI), and other actinides of higher valency. Matrix element distribution coefficients range from to Amounts of transition elements, iron, etc., are easily reduced to less than 1 pg so as to prevent inter- ference with electrodeposition.There is a large margin of separation after prior elimination of most of the iron, etc., from large biological samples (100 g to 10 kg) by preliminary co-precipitation, especially in the presence of oxalate. STRIPPING OF ACTINIDES FROM TOPO SOLUTION, AND SOURCE PREPARATION BY ELECTRO- The actinides must be stripped from the TOPO prior to electrodeposition, preferably into a solution that can serve directly as the electrolyte. The two stripping agents that best satisfy the dual criteria are ammonium oxalate and ammonium carbonate. The oxalate complexes of actinides of valency states 111, IV and VI are much stronger than the TOPO binding, and they can be stripped into a small volume for plating. Carbonate complexes are even stronger, and this ligand was adopted for stripping; it was also more suitable for DEPOSITION-876 HAMPSON AND TENNANT: SIMULTANEOUS DETERMINATION OF [AnabSt, VOl.98 electrodeposition as oxalate15 could be destroyed easily only if chloride was added to generate chlorine electrolytically. (The platinum anode was attacked by chlorine and deposited on the cathode, thus degrading the source.) Formate is another effective electrolyte for all the actinides, is easily destroyed without using chlorine, but is not a good stripping agent from TOP0 for all the valency states. Electrodeposition could follow slight acidification of the ammonium carbonate stripping solution with nitric acid, because all actinides then migrate to the cathode. However, ammonium formate was usually added as an additional complexing agent for electrodeposition so as to destroy any excess of nitric acid, which can otherwise degrade the source during prolonged electrolysis.INSTRUMENTATION- The alpha spectrometer was developed so as to have sufficient stability to maintain a constant resolution of 35 keV or better over yearly or longer periods. Drift was limited to h3.5 keV in the 5-MeV region over counting periods of 10 000 minutes. It consisted of silicon surface barrier detectors, a charge-sensitive transistorised pre-amplifier with field-eff ect transistor input stage, and Intertechnique 400 channel pulse-height analyser (SA 40B) or Northern Scientific 256 channel (NS 601). Source - detector distance was controlled by using an adjustable source mount holder based on a 1-mm pitch screw.The vacuum chamber (Fig. 1) was redesignedl6 for continuous operation over long periods. The instrument was maintained at 20 & 0.5 "C, and provided with smoothed stable power supplies. The system gave a resolution of 14 keV with 25-mm2 detectors, with an americium-241 source, and noise determined with a mercury pulser was less than 5 keV. The empirical relationship (Fig. 2) between counting efficiency and resolution was evalu- ated for detectors of various sizes (for example, 100-mm2 Ortec A series and 300-mm2 20th Century Electronics) with various source - detector spacings, as both are functions of the solid angle subtended. Detectors that gave the highest efficiency for a specified limiting resolution were chosen; the larger detectors are not necessarily the most suitable, particularly when low background is required.A C E 0 1 2 inches 0 1 2 3 4 5 cm A= Glass feed-through B=Signal feed-through C=lnsulator D = O-ring E= Spring-loaded brass finger for applying potential to sample F = Detector G =Copper face H =Acrylic sample holder J =Threaded support K=Toggle clamp Fig. 1. Section of detector chamberDecember, 19731 ACTINIDE NUCLIDES IN ENVIRONMENTAL MATERIALS 877 ANALYSIS OF SPECTRA- Alpha spectra from environmental materials containing several artificial and natural nuclides are complex, and individual peaks are not always completely resolved. Spectra from individual nuclides are typically band spectra, with about 95 per cent. of the emission from two or three major lines within about 40 keV, and more than 70 per cent.of it from the primary line. Two types of interference occur, that from partial overlapping of the main bands from two or more nuclides, and that from the low-energy continua of scattered radiation from higher-energy nuclides. There is also broadening due to incomplete instru- mental resolution. Of the several unscrambling methods applicable to gamma spectrometry,17 simultaneous matrix solution was best suited to alpha spectrometry, and is capable of correcting both types of interference simultaneously . r 50 - 40 - 10 20mm , I I 0 5 10 15 Counting efficiency, per cent. Fig. 2. Relationship between reso- lution and efficiency for detectors of varying diameter. Marked distance on point denotes source - detector spacing : A, 300-mm2 detector; B, 100-mma detec- tor; and C, 25-mm2 detector Instrumental resolution remained constant at 30 to 35 keV (full width at half maximum; F.W.H.M.) and long-term stability was controlled at several points in the range to &0.5 channel (7 keV per channel).This arrangement did not cause significant broadening of the nuclide emission bands, and enabled routine analysis to be carried out with once-for-all calibration. In order to analyse spectra by the simultaneous matrix method, characteristic channel groups (equivalent to energy bands) are selected for each nuclide. The spectra of pure sources of each nuclide are first recorded over the entire energy range covered by all the nuclides. By comparing these spectra, nuclide channel groups are chosen; they are spaced so as to maximise the counts from that nuclide, and rninimise the counts from others, taking into account the expected activity ratios (e.g., 10: 1).Interference must not be a major component of any peak. The counts are expressed as rates in each channel group, norrnalised to unity in the characteristic channel group of each particular nuclide, and represented as a matrix. For example, for nuclides A, B, C, and characteristic channel groups a, b, c : where K is the response of pure nuclide in each channel group, X is the unknown sample count-rate of each nuclide in the channel group and M is the measured count-rate of the sample in the channel group.878 HAMPSON AND TENNANT : SIMULTANEOUS DETERMINATION OF [Analyst, VOl. 98 The inverse matrix is then obtained: (Z) = (1;: : :::) x (5) kca kc, kcc where k is the nuclide factor in each channel group. The elements of the inverse matrix are then used as parameters in order to evaluate the digitally recorded spectra from the sample sources ; for example, xA = kAa MA + kAb M B + kAc MC The degree of interference depends on the proximity of the primary emission lines of the nuclides, and also on which side of these lines the secondary lines occur (i.e., whether on the same or opposite sides of the primary lines).The extent of interference indicates the practical performance of matrix analysis in particular instances of closely spaced nuclides. However, in general, interference is only a minor fraction of peaks spaced more than 70 keV apart at a peak height ratio of 10: 1, and is readily corrected (see below).But when the primary lines differ by less than 40 keV the spectra cannot be resolved. In these situations it is necessary to distinguish between the nuclides by chemical means. RESULTS Although the method is generally applicable to actinides the two procedures discussed below have been particularly tested for plutonium-238, plutonium-239 plus -240, and americ- ium-241, the most common actinides arising in controlled discharges. Some of the results that indicate their validity are given here. SPECTRAL INTERFERENCE : ASSESSMENT AND CONTROL OF ERRORS THAT ARISE- procedures A and B (see under Methods). (F.W.H.M.). The spectra shown in Fig. 3 (a), ( b ) and (c) are typical of those obtained by using Resolution in all instances is 30 to 35 keV An evaluation of the major interferences is shown in Table I for the particular I ( c ) 23"u 1 Alpha energy/NleV Fig.3. Alpha spectra of actinides: (a), separated from 10 ml of nuclear power station cooling pond effluent by procedure A; ( b ) , separated from 5 g of Porfihyra ash from Sellafield, Cumberland, by procedure A; and (c), separated from 75 g of Porphyra ash from Bude, Cornwall, by procedure BDecember, 19731 ACTINIDE NUCLIDES IN ENVIRONMENTAL MATERIALS 879 measurement of plutonium-239 and americium-241, with plutonium-236 and americium-243 as yield tracers, in materials that have curium-242 as the only other important alpha-emitter. Interference between pairs may be up to 36 per cent. for ten-fold activity ratios, but errors after matrix correction are about twenty times less because each correction has been esti- mated to be 3 2 to 5 per cent. The greatest neighbouring-pair interference is normally between plutonium-239 plus -240 and americium-243, for which diverse ratios can be avoided by judicious spiking.TABLE I SPECTRAL INTERFERENCE BETWEEN NEIGHBOURING PAIRS IN THE MEASUREMENT OF PLUTONIUM AND AMERICIUM NUCLIDES 230+240pu 243Am 241Am 23Spu 242Cm 238pu Energy band for 95 per cent. emission 5.099 to 5.224 to 5.433 to 5.717 to 6-066 to 5.452 to Energy gap between neighbouring 74 167 240 303 Energy band chosen/MeV 5-04 to 5-21 to 5.39 to 5.64 to 5.99 to 5-17 5-31 5.53 5-80 6.13 Width of energy band/keV 130 100 140 160 140 6.110 5.495 - - - - - with primary line underlined/MeV 5.150 5.266 5.476 5.763 primary or secondary lines/keV Energy gap between neighbouring 40 80 110 190 Emission appearing in neighbouring - 0-7 0.4 3.3* -0.1 Emission appearing in neighbouring 0.01 0.06 0.02 0.02 - Estimated maximum individual 7.4 3.7 32.7 1 channel blocks/keV nuclide band below, per cent.nuclide band above, per cent. interference component for ten-fold activity ratio, per cent. - * Includes typical component of 238Pu impurity in 236Pu tracer. Accuracy of measurement of americium-241 and plutonium-238 by the joint use of procedures A and B depends on the decontamination from americium by procedure B. Current ratios of plutonium-238 to plutonium-239 PZus -240 in discharges are typically 0.1 to 0-2, while the ratio of americium-241 to plutonium-238 may vary from less than 1 to 100.Minimum decontamination factors needed to guarantee an error of less than 1 per cent. are 104 for americium-241 and 105 for plutonium-238 measurement, with the “difference” method inapplicable at an americium-241 to plutonium-238 ratio of less than 0.1. A decontamination factor of lo6 was achieved, and was established by adding known amounts of americium-241 prior to analysis. Curium decontamination is probably similar, and will prevent further ingrowth of plutonium-238 from curium-242. PRECISION OF MEASUREMENT BY PROCEDURES A AND B- Precision by the two procedures (Table 11) was estimated by using samples of the seaweed Porphyra from the Cumberland coast with various concentrations of radionuclides and in- soluble components. Counting statistics were 3 per cent.(standard deviation) for individual nuclide activities, equivalent to 4 per cent. (standard deviation) for nuclide and tracer together. The estimate of 6 to 8 per cent. over-all precision, including all errors for the more silty samples, gives confidence with less exacting biological materials, e.g., fish tissues. When plutonium alone is required procedure B should be used because of its precision and simpler operation. ACCURACY OF MEASUREMENT O F PLUTONIUM-239 PlUS -240 AND AMERICIUM-241 BY PROCEDURES A AND B- The critical processes in which bias of plutonium results might occur are dissolution and exchange. Different modifications of procedures A and B were compared and are referred to in Table I11 as methods Al, A2, B1, B2, C2 and D2. Two reducing agents were investigated, metallic magnesium, a very rapid reductant for plutonium,12 and ascorbic acid, which is effective with insoluble iron compounds and man-880 HAMPSON AND TENNANT : SIMULTANEOUS DETERMINATION OF [AndySi!, VOl.98 TABLE I1 PRECISION OF 239+240PU AND 241AM MEASUREMENTS IN POYphyYU TAKEN FROM THE CUMBERLAND COAST BY PROCEDURES A AND B 23s+240pu =*lArn Sample No. 1 1 2 3 3 4 5 6 7 8 Mean . . Mean .. CoefficieI;t of Mean/ variation, Procedure minations pCi g-1 per cent. pCi g-l per cent. Number I n t of of deter- Mean/ variation, A B A A B B B B B B A B 10 2 5 4 2 3 3 2 2 2 19 16 0.530 0.567 0- 194 0-899 0.960 0.260 0-181 0.104 0.040 0.022 * 23sPu corrections applied, based on determination by procedure B. ganese(1V) oxide. Both hydrochloric and nitric acids were tested; the former is the better reducing medium, although nitric acid is completely effective in conjunction with nitrite for interchange and stabilisation of plutonium in the quadrivalent ~ t a t e .1 ~ Combinations of these reagents, also various acid concentrations, were tried out on a range of Porphyra samples. Results in Table I11 are expressed in relation to the highest plutonium-239 plus -240 content, corresponding to the most complete release and exchange with tracers. The lowest recoveries occurred when 1 M acid was used, probably resulting from losses by hydrolysis. Otherwise the range, of about &6 per cent., showed that, provided the acid concentration is maintained at 2 M, both reducing agents are equally effective. Nitric acid, however, provided better conditions for subsequent re-oxidation. Sodium carbonate fusion applied to the insoluble matter from a silty ferruginous Porphyra sample after the primary dissolution - exchange stage, but before hydrofluoric acid treatment, released no additional plutonium-239 PLUS -240, indicating that total extraction had already been achieved.Tests on the co-precipitation step by equilibrating the acid-soluble fraction with the nuclides used for spiking showed the precipitation to be complete. The phosphate precipita- tion of procedure B was not in doubt because plutonium is quadrivalent and the pH of the solution is relatively high. The oxalate precipitation of procedure A for both plutonium and americium (Table IV) was found to be virtually complete in a single stage but a second precipitation was included as a safety precaution.An estimate of accuracy by comparing the sum of plutonium and americium present with that of total alpha-activity, making allowance for minor alpha-emitters, was made on Cumberland Porphyra (Table V) in which samples 9 to 12 are bulked from nine locations for average matrix composition. Total alpha-activity was measured with 200 mg of ash (less than 22 pm particle diameter) spread over a 200-cm2 source area, by using a Beckman proportional alpha counter with adequate discrimination against beta- and gamma-radiation. Measurements of uranium, thorium (and daughter products) polonium, neptunium and curium TABLE I11 EXCHANGE OF ENVIRONMENTAL PLUTONIUM IN Porphyra ASH WITH 236Pu TRACER Number of Method Solution conditions samples tested Bias * A1 Mgo - 1 M HNO, 6 0.817 B1 Mgo - 1 M HC1 6 0.774 B2 MgO - 2 M HC1 5 0.927 c 2 Ascorbic acid - 2 M HNO, 5 0.965 D2 Ascorbic acid - 2 M HC1 5 0-885 * Degree of incomplete exchange in relation to method A2 taken as 1.A2 Mgo - 2 M HNO, 5 1.000December, 19731 ACTINIDE NUCLIDES IN ENVIRONMENTAL MATERIALS 881 nuclides on occasional samples of Porphyra from the area indicate that the sum of such activities corresponds approximately to the mean “unaccounted” activities in Table V. Thus errors caused by bias are not serious. TABLE IV COMPLETENESS OF PRECIPITATION OF PLUTONIUM AND AMERICIUM AS OXALATES FROM SOLUTIONS OF SEAWEED ASH AT pH 1.5 Recovery as percentage of total nuclide recovered each with 200 mg of oxalic acid 23sPu z43Am First 95.7 98.1 Second 2-92 1-34 Third 1.34 0.54 Consecutive co-precipitations, r- For effluents, curium-242 was also measured as a major nuclide.Its recovery was calculated from the americium-243 spike, as a curium tracer was not then available. The mean results are shown in Table VI and involved variants of acidic media and redox agents in the dissolution - exchange operation, but showed only small differences, the conditions giving the highest activity values being those given in the procedure. Small differences between methods imply that the dissolution - exchange stage is efficient. Two other variants were tested. In one, with the redox cycle omitted, results varied widely; in the other 1 0 ~ nitric acid - 0.1 M hydrofluoric acid mixture was used for the pre-exchange leaching of actinides from insoluble matter, as recommended by other workers.18 Results were invariably low, indicating that actinides are not completely available under these conditions : they must be leached and equilibrated before the introduction of hydrofluoric acid.Total alpha measure- ments on effluents were made from thin evaporated sources with a zinc sulphide alpha scintillation counter gated not to register beta- and gamma-radiation and calibrated with a similar standardised source. The comparison of the sum of plutonium-239 PZNS -240, americium-241 and curium-242 with total alpha-activity (Table VI) shows only a small discrepancy accountable to uranium, neptunium and other nuclides noted in the spectra, and indicated accuracy to within a 10 per cent.limit. TABLE V MEASUREMENT OF PLUTONIUM , AMERICIUM AND TOTAL ALPHA-ACTIVITY IN Porphyra Sample No. 9 10 11 12 13 14 0.75 1-16 0.86 0.74 0.60 1.07 z3spu - 2 4 1 ~ ~ / pCi 8-l 0-72 1.77 0.87 0.81 0.32 0.46 Mean .. * . 238+239+240pu - 24lAm/ pCi 8-1 1.47 2.93 1-73 1-55 0.92 1.53 1.69 Total alpha- activitylpci g-’ 1.37 4.11 1.96 1.40 1.93 1.82 2.10 SCOPE- The method is applicable to all the actinides except the quinquivalent oxy-cations, and, by maintaining all in one of the other states, most of them up to curium have been measured. Convenient nuclides for spiking include actinium-225 , thorium-229 , uranium-233, neptun- i~m-235,~~ plutonium-236 ,19 americium-243 and curium-244. On occasions when energies conflict alternatives can sometimes be used; otherwise differential extraction procedures are readily developed.Successful application to natural waters had been made after suitable pre-concentration. SENSITIVITY- Attainable sensitivity depends on sample size, extent of chemical recovery and counting efficiency. Amounts of seaweed samples up to 2 kg, and of fish flesh up to 10 kg, have been handled. Chemical recoveries are virtually 100 per cent. at the precipitation stages and882 HAMPSON AND TENNANT: SIMULTANEOUS DETERMINATION OF [Analyst, VOl. 98 75 to 90 per cent. at the solvent-extraction and electrodeposition stages, with more than 50 per cent. overall. In this work 300-mm2 detectors with 25 to 30 keV (F.W.H.M.) nominal alpha energy resolution were used. With the 10 mm diameter sources, discrimination factors were chosen equal to 12 per cent.of 47r geometry (Table I). With counting periods up to 1 week, ultimate sensitivity relative to a standard deviation on counting of 4 per cent. is 4 x pCi g-l for the 10-kg samples. Nominal detec- tion limits, in terms of the activity to double background, are 2 x and pCi g-l, respectively. Sensitivity for measurements of plutonium-239 and americium-241 to 4 per cent. (stan- dard deviation) precision when using 100-g samples of foodstuffs and counting for 100 minutes is about 0.8 pCi g-l. Measurements on marine foodstuffs, fish flesh or Porphyra, can thus be made at concentrations of about 1 per cent. of the derived working limits for consumption by the general public.20 Actual concentrations in individual Cumberland Porphyra samples are currently 0.001 to 5 pCi g-l, and about 0.005 pCi g1 in fish flesh from the vicinity of the Windscale outfall.Up to l-kg samples are needed for 4 per cent. precision in 24-hour counts. The fallout concentration in seaweed estimated at 2 to 5 x 10-4pCig-1 can be measured with 4 per cent. (standard deviation) precision with a 2-kg sample in 1 to 2 weeks’ counting time, although in fish, about In conclusion, actinide analysis by solvent extraction and solid-state alpha spectrometry is sufficiently selective and accurate for control of actinides in foodstuffs, and for surveillance of their spread in aquatic organisms. TABLE VI MEASUREMENT OF MAJOR INDIVIDUAL ALPHA-EMITTERS IN EFFLUENTS, AND COMPARISON (MEAN OF THREE DETERMINATIONS) pCi g-l for the 2 kg and 8 x pCi g1 can be detected with lower precision.O F THEIR SUM WITH TOTAL ALPHA-ACTIVITY MEASUREMENTS Effluent No. 1 2 3 4 5 6 Mean . . 2 3 9 + 2 4 O P ~ - pCi ml-l per cent. 1.46 4.2 0.217 12.2 0.825 13-4 0.605 20.7 0.893 12.8 0.361 12.9 12-7 Coefficient Mean/ of variation, 238pu - 241Am f - - ~ - - - - - ? Coefficient Mean/ of variation, pCi ml-l per cent. 0.855 1.2 0.1 13 6.7 0.531 12.9 0.406 21.7 0.605 9.9 0.242 8.2 10.1 249Cm, single Sum of determina- 23*+239+240Pu, tions/ 2r1Am and 2a2Cm/ pCi ml-1 pCi ml-l 4-12 6.43 0.443 0.773 2-10 3.46 1.66 2-67 2.5 1 4.0 1 0.870 1-47 3.14 Total alpha-activity/ pCi ml-1 6.17 1.26 4.20 2.73 5.42 1.40 3-46 METHODS Two types of procedure for group selection of actinides of valency states 111, IV and VI, and for sub-group selection of actinides of valency states IV and VI with I11 eliminated, have been developed for application to biological materials and effluents.Most of the details of the four routines are similar, and that for actinides with valencies 111, IV and VI in biological materials is stated in full. The points of difference in the other procedure, and in the sub-routines then follow. REAGENTS- Plutonium-236 solution in 2 M hydrochloric acid. Americium-243 solution in 2 M hydrochloric acid. Tri-n-octylphosphzine oxide. n-Heptane. PROCEDURE A FOR THE SEPARATION OF PLUTONIUM, AMERICIUM AND OTHER ACTINIDES TOGETHER AS A GROUP FROM BIOLOGICAL MATERIALS- Sample dissolution and exchange of actinides with tracers-Weigh 5 g* of carbon-free ash * For amounts of sample larger than 5 g (up to 100 g) the amounts of reagents used should be varied in proportion.December, 19731 ACTINIDE NUCLIDES IN ENVIRONMENTAL MATERIALS 883 into a centrifuge tube and spike it with standardised solutions of suitable nuclides, pluton- ium-236, americium-243 and others as required (see under Scope).Boil the mixture for 5 minutes with 40ml of 4111 hydrochloric acid, cool and dilute to 80ml. Add 500mg of ascorbic acid and leave the solution for 5 minutes in the cold, then heat it. After cooling, add 1 ml of 30 per cent. hydrogen peroxide solution, and then dissolve 10 g of sodium nitrate and 1 g of sodium nitrite in the solution. Allow it to stand for 5 minutes, then heat it. Adjust the pH of the solution to 1.5 with ammonia solution, then add 4 ml of 5 per cent.oxalic acid solution, and leave the mixture to stand until a precipitate forms. Collect the precipitate, together with acid-insoluble matter, by spinning it in a centrifuge. Repeat the precipitation with a further 4 ml of oxalic acid solution. Combine the precipitates, and wash them with 50 ml of 2 per cent. ammonium oxalate solution. Transfer the combined precipitates into a 50-ml capped PTFE centrifuge tube with 20 ml of 16 M nitric acid. Add 20 ml of hydrofluoric acid (70 per cent. m/m if a pure grade is available) and digest the mixture at 100 "C so as to dissolve all siliceous matter. Transfer the solution, in portions, to a platinum crucible and evaporate it under radiant heat. Complete the transfer with nitric acid so as to ensure dissolution of all fluorides, etc.Evaporate just to dryness two or three times with mixed nitric and hydrofluoric acids, and then two or three times with nitric acid. Dissolve the dry residue in 20 ml of 0.4 M nitric acid that is half-saturated with boric acid. (A larger volume may be needed if much acid-insoluble matter was originally present. This volume should be kept small, but without approaching the solubility limit). Treat a 20-ml volume of the solution with 50 mg of ascorbic acid, or larger volumes with 100 mg, for 5 minutes. Make the solution 4 M in sodium nitrate. Treat it for 5 minutes with 150 mg of sodium nitrite for a 20-ml volume, or 300 mg for larger volumes. Centrifuge or filter the solution if it is not absolutely clear. Extraction with TOPO and stri$$ing-Decant 10 ml of freshly prepared 0.5 M solution of TOPO in n-heptane into a separating funnel, and purify it by washing it in turn with 5 per cent.sodium carbonate solution, 2 M nitric acid (three times) and 0.2 M nitric acid - 4 M sodium nitrate solution. Shake the aqueous phase prepared from the sample with the purified TOPO solution for 5 minutes. After separating the phases, scrub the TOPO solution three times with 10 ml of 0.2 M nitric acid - 4 M sodium nitrate solution for 5 minutes each time. Strip the actinides by shaking the TOPO solution three times, each for 5 minutes, with 3ml of 10 per cent. ammonium carbonate solution. Wash the combined aqueous phases once for 1 minute with 10 ml of n-heptane and transfer them to a plating cell.ELECTRODEPOSITION- Prepare a disposable plating cell, which can be made from the cut-off top and screw-cap of a polythene bottle (e.g., hydrofluoric acid type). Degrease a polished stainless-steel disc, 27 mm in diameter and 22 s.w.g. (0.711 mm) thick, in trichloroethylene, and mount it as the cathode, with a polythene gasket to expose a central 8 mm diameter area for a 10-mm detector, or 1Omm diameter area for a 20-mm detector. Complete the cell with a loop of 0-040-inch (1-mm) thick platinum wire to serve as the anode, inserted through a polythene foil cover. Acidify the stripping solution in the cell to pH below 2 with 16 M nitric acid, by using wide-range indicator paper in order to ensure that the excess of acid does not exceed one drop.Add 1 ml of 40 per cent. ammonium formate solution and electrolyse the solution at 3.5 mA mm-2 for 8 to 16 hours. The pH of the final bulk solution should be above 10. On dismantling the cell rinse the disc rapidly with methanol and dry it. A good source will be almost invisible. MEASUREMENT AND ANALYSIS OF THE ALPHA SPECTRA- Set the spectrometer to the required alpha energy range by using a multi-nuclide source. Make regular checks for spectral resolution, linearity and drift so as to ensure that all sources are counted on exactly the same channel settings. Calibrate with an absolutely standardised source that is similar to the sample sources (obtainable from the Radiochemical Centre Ltd., Amersham). Count a set of individual sources of all alpha nuclides in the samples and tracers,884 HAMPSON AND TENNANT: SIMULTANEOUS DETERMINATION OF [Analyst, Vol.98 by using the actual batches of the latter to correct automatically for impurities. Prepare and invert a matrix based on chosen channel groupings so as to correct for spectral inter- ference arising from all the nuclides present. For close pairs check that acceptable error limits will not be exceeded. Count the sample sources to acceptable statistics, allowing for count-rates of required and tracer nuclides, and measure background between batches. Calculate the results by detecting background effect and calculating the activity due to each nuclide, by using the inverse matrix and the absolute calibration. Correct for yields on the basis of measured tracer recoveries.For samples containing both plutonium-238 and americium-241, measure the pluton- ium-238 (and plutonium-239 plus -240) by procedure B, spiking with plutonium-236. Obtain the americium-241 (and also plutonium-239 plus -240) by procedure A, spiking with americ- ium-243 and plutonium-236. Correction for plutonium-238 has to be made from its measure- ment by procedure B because it cannot be resolved from americium-241. In a batch of samples in which the 238Pu to 239+240P~ ratio is found to be constant, this ratio may be inserted into the matrix and the americium441 then measured by procedure A. PROCEDURE B FOR THE SEPARATION OF PLUTONIUM AND OTHER QUADRI- AND SEXAVALENT ACTINIDES, WITH ELIMINATION OF AMERICIUM AND OTHER TERVALENT ACTINIDES , FROM BIOLOGICAL MATERIALS- Sample dissolution and exchange of actinides with tracers-Carry out the spiking, leaching and chemical exchange stages as above, by using samples of any size up to 1OOg of ash.Neutralise the resulting solution with ammonia solution until an amount of precipitate of 100 to 200 mg forms on top of any acid-insoluble matter. Allow to stand so as to enable the precipitate to coagulate and collect it by spinning the mixture in a centrifuge. Wash the precipitate with 50 ml of 2 per cent. ammonium nitrate solution made slightly alkaline with ammonia solution. Dissolve the dry residue con- tained in a platinum crucible in 10 ml of 2 M nitric acid half-saturated with boric acid. A larger volume may be needed if much acid-insoluble matter was present originally.Carry out the reduction and re-oxidation as described above, by using the amounts of reagents given for solution volumes of less than and more than 20 ml, respectively. Extraction into TOP0 and stripping-Prepare the 0.1 M TOPO phase as above, but use only 2~ nitric acid for the final wash. Shake the aqueous phase prepared from the sample with the purified TOPO solution for 5 minutes. Separate the phases and scrub the TOPO solution four times with 10 ml of 2 M nitric acid, for 5 minutes each time. Strip the actinides as described above. APPLICATION OF PROCEDURES A AND B TO EFFLUENTS- Make effluent samples 4~ in nitric acid and store them in polythene bottles. Transfer 10 ml of suspension to a 50-ml PTFE tube, and spike with standardised solutions of yield tracer nuclides, Dilute the mixture to 20 ml and add 100 mg of ascorbic acid.Allow the mixture to stand for 5 minutes, then heat it. Add 500 mg of sodium nitrite, allow to stand for 5 minutes and again heat. Digest and complete the removal of silica, dissolution and valency adjustment as described above, but making the dissolution media only 0-2 M in nitric acid half-saturated with boric acid for procedure A, or 2 M nitric acid - boric acid for procedure B. The remaining operations for tracer exchange and extraction are as for biological materials described above. Carry out the dissolution and removal of silica as above. Add 10 ml of fuming nitric acid and 15 ml of hydrofluoric acid. REFERENCES 1. 2. 3. 4. 5. 6. 7. Magmo, R. J., Kauffman, P. E., and Schleien, B., Hlth Phys., 1967, 13, 1335. Butler, F. E., Ibid., 1968, 15, 19. Pillai, K. C., Smith, R. C., and Folsom, T. R., Nature, Lond., 1964, 203, 576. Hill, C. R., Hlth Phys., 1962, 8, 17. Gomm, P. J., and Eakins, J. D., Analyst, 1968, 93, 228. Sill, G. W., Hlth Phys., 1969, 17, 89. Gureev, E. S., Dedov, V. B., Karpacheva, S. M., Lebedev, I. A., Shvetsov, I. K., Yakovlev, G. N., Ryzhov, M. N., and Trukchlayev, P. S., Proc. 3rd Int. Conf. Peaceful Uses Atom. Enevgy, Geneva, 1964, p. 348.December, 19731 ACTINIDE NUCLIDES IN ENVIRONMENTAL MATERIALS 885 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Coleman, G. K., “The Radiochemistry of Plutonium,” Rep. Nat. Acad. Sci., Wash., NA S-NS-3058, White, J . C., and Ross, W. J., “Separations by Solvent Extraction with Tri-n-octylphosphine Toribara, T. Y . , Predmore, C., and Hargreve, R. A., Talanta, 1963, 10, 209. Ballada, J ., “Determination Analytique du Plutonium dans l’Environment,” Rapp. CEA R3220, Katz, J . J . , and Seaborg, G. J., “The Chemistry of the Actinide Elements,” Methuen, London, Weaver, B., and Horner, 13. E., J . Chem. Engng Data, 1960, 5, 260. Dukes, E. K., J . Amer. Ckem. Soc., 1960, 82, 9. Brooks, R. 0. R., Rep. U.K. Atom. Energy Auth., AM 60, 1960. Perry, K . E. G., in “Proceedings of the Symposium on Radioisotope Sample Measurement Tech- O’Kelley, G. D., “Detection and Measurement of Nuclear Radiation,” Rep. Nut. Acad. Sci., Black, R. M., and Drummond, J. L., Rep. U.K. Atom. Energy Auth., TRG 1072 (D), 1965. Jenkins, J . L., and Wain, A. J., Ibid., A E R E R5790, 1968. Preston, A., and Jefferies, D. F., Hlth Phys., 1969, 16, 33. 1965. oxide,” Ibid., NAS-NS-3102, 1961. 1967. 1957. niques in Medicine and Biology, Vienna, 1965,” I.A.E.A., Vienna, 1966, pp. 687-698. Wash., NAS-NS-3105, 1962. Received October 18th, 1971 Amended April 9th, 1973 Accepted June 26th, 1973