JANUARY, 1967 THE ANALYST Vol. 92, No. 1090 Activation Analysis A Review* BY R. F. COLEMAN ( A tomic Weafions Research Establishment, Aldevmastoqt, Bevks.) AND T. B. PIERCE (Atomic Energy Research Establishment, Havwell, Bevks.) SUMMARY OF COXTESTS Introduction Nuclear interactions Irradiation facilities Nuclear reactors Accelerator neutron sources Isotopic neutron sources Charged particles y-Photons Radiochemical separations Rapid separations Group separations Automatic separation systems Substoicheiometry y-Spectroscopy Other counting methods Prompt radiation techniques Applications High purity materials Biology Forensic science Geochemistry On-line analysis Surface analysis Measurement of induced activity ACTIVATION analysis is one of the most rapidly developing analytical techniques at the present time.I t is difficult to obtain a comprehensive and accurate picture of the state of the art because the papers are widely scattered throughout the literature of many different scientific disciplines. In this review an attempt has been made to indicate the diverse fields from which developments in technique have come; to evaluate their significance for the analyst and show how they have been used in practice and, for some, their potential in the near future. In order that this review will be of use to the newcomer to this subject, as well as the specialist, a section briefly outlining the theoretical basis for the subject is included. The paper has concentrated on work published in the last 3 years, and earlier material is included only if especially significant.Rather than merely list all the papers on a topic, a few key references have been selected, and, wherever possible, they are those that have improved the technique or evaluated carefully the effect of interferences, precision and accuracy, or compared activation analysis with other methods. It is felt that the reviewers have a duty to comment on the potential of new developments and, further, if first-hand knowledge is available to recommend a particular approach to a problem. This is necessarily subjective, but it is hoped it will guide the less experienced workers in this field. For details see Summaries in advertisement pages. * Reprints of this paper will be available shortly. 12 COLEMAN Simple nuclear interact ions by the following equations- AND PIERCE : ACTIVATION ANALYSIS NUCLEAR INTERACTIONS [AwaZyst, Vol. 92 used as a basis for activation analysis may be represented .. - . (1) A + a - + A + a . . . . - + A * + a .. . . .. * - (2) .. - - (3) - + B + b . . . . where A and B are the target or residual nuclei, and a and b are the bombarding or emitted particles or photons. Processes (1) and (2) are both known as scattering. In reaction (1) the kinetic energy of reactants and products is the same, whereas in reaction (2) the total kinetic energy of the products is less than that of the reactants, the difference being used to raise the nucleus A to an excited state. Process (1) is known as elastic scattering, process (2) as inelastic scattering. Reaction (3) represents the transmutation of one element to another and one or more particles or y-photons being emitted.The residual nucleus B may be formed in ground or excited states, the latter usually decaying very rapidly to the ground state by the emission of y-radiation, and the ground state of nucleus B itself may be unstable and may decay to the ground or excited states of yet another nucleus, for example, by emission of cc or P-radiation. A further form of interaction that may occur between a charged particle and a target nucleus, coulomb excitation, has not hitherto been applied to activation analysis to any significant extent and will not be considered further here. An arbitrary distinction is sometimes drawn between prompt and delayed radiation. Prompt decay refers to processes that occur rapidly after nuclear interaction, while delayed radiation is measurable over a much longer period of time so that the sample can usually be removed from the place of irradiation before counting.The Q-value or energy balance of a reaction as calculated from the masses of reactants and products shows whether the process is endo-ergic or exo-ergic. Theoretically, an exo-ergic reaction can occur with a zero energy particle assuming no coulomb barrier restriction, while for endo-ergic reactions with an energy balance of -Q, an amount of energy equal to Q must be supplied by the kinetic energy of the incoming particle for the reaction to occur. As reaction between a stationary target nucleus and a bombarding particle will impart momentum to the compound nucleus, the reaction threshold will be greater than the Q value.The rate of a nuclear reaction is given by RR = fan where f is the particle flux; u is the reaction cross section; and n is the number of target atoms. A counter observing prompt radiation emitted from a target during irradiation will yield a count-rate, C,, which is proportional to the rate of reaction, i.e., where K , is a constant that depends upon the efficiency and geometry of the detector and the proportion of the decay emitting the radiation to be determined. When the product is a radioactive nuclide, this will have a rate of decay R, = AN, where N and A are the number of atoms and the decay constant of the nuclide, respectively. The net rate of growth of a radioactive nuclide will be- .. . .- * (4) C, = K,RR = K,fun .. dN L- dt = RR - RD =fun - AN, and thus the disintegration rate D(t) after irradiation for time t will be given by- .. . . * - (5) D(t) = fan (1 - e-0*693t/rt) . . . . . . * - (6) D(t) = fun (1 - ,-at) .. As the half-life, T,, of a radionuclide and the decay constant are related by the expression h = 0*693/T,, equation (5) becomes- Thus for a sequence of nuclear reactions in which a compound nucleus is de-excited by prompt decay to form a radionuclide, the disintegration rates yielding prompt and delayed radiation will only be equal a t saturation. After the completion of the irradiation, the decay rate of the radionuclide will then decay exponentially with time.January, 19671 COLEMAN AND PIERCE : ACTIVATION ANALYSIS 3 Prompt counts can be accumulated indefinitely by extending the duration of the irradi- ation, (i.e., from equation (4) the total counts registered in time t, C,, will be C, = Klfont) but in practice a useful lower limit will be imposed by counts from other sources that interfere with the determination. Radionuclides of a sufficiently long half-life can be removed from the place of irradiation and chemically separated from other active elements before counting.IRRADIATION FACILITIES NUCLEAR REACTORS- The nuclear reactor is the most prolific source of neutrons for activation analysis, although only a limited number of reactors have been designed specifically for this purpose. The neutron flux available for the irradiation of samples ranges from loll to 1014 neutrons per second per cm2, with most research reactors having a maximum flux of about 10I2 to The method of loading samples varies according to the installation, but generally tw7o methods are required: (a), a pneumatic tube for rapidly transferring samples from the laboratory to the reactor core for periods of seconds to about an hour; and ( b ) , a facility for long irradiations of up to about a week, which can operate more slowly and must be made of radiation-resistant materials.The reactor is a source of neutrons of widely varying energy from thermal neutrons of about 0.02 eV up to fast neutrons of greater than 10 MeV. Usually, no special effort is made to select the energy of neutrons for irradiation of samples; it is, however, essential to have a knowledge of the energy spectrum because of possible interferences discussed below. Furthermore, there is often a rapid change in flux over short distances in the irradiation position, and significant errors may result if this is ignored.I t is essential that the methods of transporting samples are reproducible, and often there is a variation in flux throughout the sample carrier requiring carefully packed samples. Activation analysis is not usually an absolute method, although Girardi, Guzzi and Paulyl have shown that the nuclear constants involved are usually sufficiently well known to allow many elements to be determined absolutely with an accuracy of better than 10 per cent. The more usual method is to irradiate the sample simultaneously with known amounts of a pure material. If the sample and standard are then counted under similar conditions the unknown quantity of the element in the sample is readily calculated.Care must be taken in the selec- tion of materials for standards; it is not unusual to find pure chemicals containing traces of elements of high neutron cross section that can seriously affect the validity of the standard. If a great many samples are to be analysed for many elements then the simultaneous irradiation of standards each time would be very time consuming. The comparator method requires that only one standard is irradiated each time to monitor the neutron flux, and the activity of the samples is compared with standards separately irradiated at a known neutron flux. A single flux monitor cannot, of course, detect changes in neutron spectrum, and it is essential that this remains reasonably constant over long periods of time.A study of the errors that can occur with this method, particularly in reactors with a high fast-neutron flux, has been made by Girardi, Guzzi and Pauly.2 There are many compilations of sensitivity for thermal neutrons; possibly the most comprehensive and also indicating all of the interfering reactions is the handbook by K ~ c h . ~ The presence of fast neutrons in a reactor neutron spectrum demands that careful con- sideration be given to interfering reactions; it is possible to produce nuclides with atomic number 1 or 2 less than the matrix elements by (n,p) and (n,a) reactions. Durham, Navalkar and Ricci4 discuss these interferences and also indicate certain examples in which fast neutrons can be utilised to advantage.Borg, Segal, Kienle and Campbell5 showed that by surrounding samples with boron and cadmium it was possible to improve the detection limit of manganese in tissue by a factor of 7 because of the greater suppression of sodium-24 activity compared with manganese-56. Yule6 experimentally determined the sensitivity of 28 elements for reactor fast neutrons. He showed that for 5 elements the sensitivity is greater than for thermal neutrons and is only slightly less for 15 others, and hence may be superior in some matrices. In the TRIGA reactor it is possible to obtain, safely and reproducibly, neutron pulses of about 30-millisecond duration with peak fluxes of 5 x 10l6 neutrons per second per cm2, as well as a continuous flux of 1013 neutrons per second per cm2.Lukens7 showed that there is an increase in sensitivity for elements with half-lives of less than 50 seconds by pulsed operation of the reactor.4 COLEMAN AND PIERCE ACTIVATION ANALYSIS [AnaZ$!yst, Vol. 92 The interaction of the nucleus with a bombarding particle is independent of the matrix, but in practice the matrix can exert an influence by perturbing the neutron flux seen by different parts of the sample. This self-shielding effect has been studied by many workers; most of this work has been summarised by Hogdahl,8 who has tested the accuracy of the calculations by experiment and has also suggested methods by which the effect can be calculated. If the cadmium ratio for any nuclide is greater than 50, the effect of absorption of epi- thermal and fast neutrons can be ignored.Kamenoto’sg simple equation, oW/A e 0-03, for a 10 per cent. correction owing to absorption, where CT is the cross section in barns, W the weight of sample in grams and A the atomic weight of the nuclide, can be used to test quickly the need to consider self-shielding effects in more detail. With the availability of very high flux reactors for activation, other second-order reactions can cause significant interferences. Ricci and Dyer10 have found 42 instances in which interferences can occur, and in 23 of these they were able to compute the interference from currently available nuclear data. So many elements in a wide variety of materials have been determined by thermal neutron activation it would be foolish to quote just a few examples.The bibliographies of Bock-Werthman,ll published at regular intervals by the A.E.D. Information Service, have element and matrix indexes and provide an excellent reference source for published papers of this type. ACCELERATOR NEUTRON SOURCES- So that the advantages of activation analysis can be applied to laboratories without access to a reactor and to extend further the range of the technique, alternative sources of neutrons have been studied. In many nuclear reactions a neutron is the product of bombard- ment, but for practical purposes only, three reactions provide a useful neutron flux with reasonably priced accelerators. (i) 3H + 2H -+ 4He + n + 17-6 MeV. (ii) 2H + 2H -+ 3He + n + 3.3MeV. (iii) 9Be + 2H -+ 1°B + n + 4.4 MeV.Reaction (i), usually referred to as the DT reaction, is the most prolific source of neutrons for accelerators capable of generating voltages of less than 500 kV. The neutron energy is approximately 14 MeV, and is sufficient to produce (n, an), (n,p) and (n,cc) reaction products with almost all elements. Reaction (ii), the DD reaction, lias only found limited application in activation analysis. The neutron yield is much lower than for the DT reaction and the neutron energy of 2.5 MeV is generally insufficient to produce (n, 2n) reactions and only a limited number of elements can react to produce (n,p) and (n,cc) products. The reaction has been used in preference to the DT reaction if the latter formed the same product from two elements and only one reaction was energetically possible with DD neutrons.Reaction (iii) produces a broad energy spectrum of neutrons up to about 6 MeV, and is the most prolific neutron source for accelerators rated at 1 MeV or above. All types of accelerators can be used as sources of thermal neutrons by surrounding the target area with a water - paraffin moderator. I t is not possible, however, to operate in a region of high thermal flux that is free from fast neutrons, and interference from fast neutron reactions must be carefully considered. A recent survey12 conveniently tabulates most of the data concerning accelerators com- mercially available for neutron production. Most laboratories have found that 150 to 200-keV accelerators that use the DT reaction for neutron production are the most useful at reasonable cost (LBO00 to LSOOO) .The small sealed-tube accelerators are particularly well suited to routine use in a factory for monitoring a specific product. The shielding of a neutron generator is often an expensive item, particularly if a reason- abl~7 large space is enclosed ; several installations are briefly described by Strain.13 Neutron tubes have been installed in a small hole in the ground and then covered with polythene chips. This provides a shield that only costs about LSOO and includes a fluidiser to allow the tube to be easily inserted and withdrawn. However, there must be a higher thermal neutron flux around the sample which may cause interference in certain determinations.January, 19671 COLEMAN AND PIERCE ACTIVATION ANALYSIS 5 The inability to manufacture tritium targets which will maintain a high yield of neutrons over a long period has slightly limited the use of the neutron generator.In some earlier accelerators the target life was controlled by the loss of tritium by over-heating and the deposition of films of carbon on the target surface. By suitable design these problems can be overcome, but the loss of tritium by exchange with deuterium from the beam and subse- quent diffusion of tritium out of the target is unavoidable. In most accelerators the neutron output from a titanium tritide target decreases by a factor of 2 after 60 to 100 minutes at about 800-pA beam current .13 Erbium tritide targets, although possessing greater thermal stability, have a shorter life.To avoid the necessity of replacing used targets, regeneration is sometimes used.14 The deuteron beam is temporarily switched to a tritium beam that replaces the lost tritium from the target. Some sealed tubes15 accelerate a mixture of deuterium and tritium, and a constant neutron output of 10lo neutron per second can be maintained for about 100 hours. Most manufacturers of accelerators supply a transfer system to allow the sample to be moved rapidly from the neutron target to the counter. For reliable analysis it is essential that the system is capable of accurately locating the sample in a position of high flux each time. Because of the limited region of high neutron flux many workers do not simultaneously irradiate sample and standard, which is the common practice with reactor irradiations.Each is irradiated separately, and the change in neutron flux is monitored and normalised by measuring the flux with boron trifluoride counters, fission chambers, cooling water activity or a plastic scintillator. IddingP has compared most of these methods and shown that the comparator method, involving the simultaneous irradiation of sample and standard, is the most accurate, but that it must inevitably lead to a decrease in sensitivity of the method. The difficulty of determining oxygen by other methods and the high specificity of fast- neutron activation for oxygen under certain conditions has prompted many groups to study this determination extensively. Coleman17 has shown that, by correcting for y-absorption by the sample, it is possible to use one primary standard for oxygen determination in a variety of materials, which does not rely on samples analysed by other techniques.Satisfactory agree- ment with the vacuum fusion method for a variety of metals was achieved. Anders and Bridenls improved the accuracy by rotating samples while irradiating and counting, and showed the need for neutron-absorption corrections if a large sample is irradiated. Mottlg further reduced the experimental error (1 standard deviation) to about 0.4 per cent. The apparent increase in oxygen content arising from nitrogen-16 recoils from the air can be avoided by using nitrogen in the pneumatic transfer system,20 or by rapid etching of the sample after irradiation and before counting, Fast-neutron activation is finding application in other analytical problems, particularly for light elements.Nitrogen has been measured in rubber by Walker and Eggebraaten,21 but in many materials copper-62 produced by the reaction, 63C~(n,2n)62C~, is a serious inter- ference having the same half-life and decay mode as nitrogen-13. Blackburn22 has determined fluorine in pure organo fluorine compounds with errors of about 1 per cent. by using the lgF(n,2n)lsF reaction. The rapid and sensitive reaction l9F(n,~)l6N is seriously interfered with by the formation of nitrogen-16 from oxygen if the sample is irradiated with 14-MeV neutrons. A n d e r ~ ~ ~ used neutrons from the deuteron bombardment of beryllium, and Steele24 used neutrons from the DD reaction; in both instances neutrons are sufficiently energetic for the fluorine reaction but are below the threshold level for oxygen.The 2sSi(n,p)2sAl reaction permits the trace determination of silicon in many matrices, the only likely interference being phosphorus. has used this reaction for the precise determination of silicon in small meteorites. Gorski26 has determined the copper content of ores at the rate of several hundred per day. Tables of calculated values for the sensitivity of fast-neutron activation have been published by Gillespie and Hill2’ and also by Coleman.28 In practice it is found that the limits of detection are usually about 100 times larger than calculated, particularly if the active species is measured by y-spectroscopy. ISOTOPIC NEUTRON SOURCES- Isotopic neutron sources have not been used extensively in activation analysis because of their low neutron flux, but nevertheless there are applications in which this is not a serious disadvantage. Many combinations of a-emitter and y-emitter with light elements produce6 COLEMAN AND PIERCE ACTIVATION ANALYSIS [Analysf, VOl.92 satisfactory neutron sources. In Table I, some typical sources supplied by the Radiochemical Centre, Amersham, are listed. TABLE I ISOTOPIC NEUTRON SOURCES output, Maximum neutron per second activity available, Cost for Source per C C maximum activity 124Sb - Be 5-2 x 10' 10 L110 zzsTh - Be 2 x 107 5 -$so0 2a1Ain - Be 2.5 x lo6 5 L400 ZlOPo - Be 2.5 x 10' 50 (80 A new type of neutron source has been developed by Amie129 in which an cc-emitter, usually americium-241 or thorium-228, is surrounded by oxygen-18 in a gaseous form such as carbon dioxide.As the gas can be readily separated from the a-source, the neutrons can be turned on and off, greatly simplifying shielding problems during transportation and storage of the source. Californium-252, which emits lo9 neutrons per second per mg, will shortly be produced at the rate of 1 g per year in the high flux reactor at Oak Ridge. It should then be possible to obtain very small sources with a high neutron output. De and Meinke30 have made a general evaluation of 1 to 5-C antimony - beryllium sources for activation analysis. Isotopic sources can be useful in the rapid assay of samples in which conventional methods are rather tedious ; for example, Grunewald31 has determined iodine in organic compounds, and Bakes and J e f f e r ~ ~ ~ have determined fluorine in fluorite ores ; for both, the accuracy of the measurement was within 1 per cent.The absorption of neutrons from isotopic sources by elements of high neutron cross section has been utilised by Strain and L y ~ n ~ ~ for analysing flowing streams and batch samples con- taining cadmium and boron. CHARGED PARTICLES- The activation applications of charged-particle techniques have been limited by the relatively short distances that charged particles can penetrate into matter and the high heat dissipation that can occur if large beam currents are used. Nevertheless, a variety of charged particle methods have been developed for light elements that cannot be conveniently deter- mined after irradiation by reactor neutrons; as few radioisotopes suitable for activation analysis can be produced from these elements, reaction products are limited to a few radio- nuclides, for example, carbon-1 1, nitrogen-13 and fluorine-18.The coulomb barrier will restrict the reaction of low energy charged particles to elements of low atomic number, thus simplifying methods for the determination of light elements, but as particle energies are increased and the thresholds of additional reactions are crossed, the problem of nuclear interferences becomes more acute. Also, the matrix effect caused by variations in stopping power of different targets complicates standardisation. The prepara- tion of synthetic standards for geochemical use has been found to be possible by Sippel and Gl0ver,3~ and Ricci and Hahn35 have advocated the use of an average cross section when cal- culating charged particle results.Particle energies of 1 MeV or less can be conveniently obtained with a Cockcroft - Walton voltage multiplier or by other types of high voltage rectifiers. These machines are relatively cheap and can yield high ion currents. Van de Graaff generators provide higher accelerating voltages (approximately 6 MeV for a single-stage machine) and for both Van de Graaff and Cockcroft - Walton accelerators, targets can be conveniently placed at the end of the flight-tube, with or without a thin window to isolate the irradiation chamber from the rest of the evacuated volume of the machine. Higher particle energies can be attained with multiple-stage Van de Graaff generators, cyclotrons or linear accelerators.Targets irradiated inside a cyclotron usually intercept the beam only at an edge, and therefore the use of an external beam is to be preferred when lower beam currents are accept able.January, 19671 COLEMAN AND PIERCE ACTIVATION ANALYSIS 7 Relatively few activation procedures have been published which are based on charged particle irradiation, and there has therefore been little incentive to develop an accelerator for activation analysis other than for the production of neutrons. However, recent interest in helium-3 has prompted Markowitz and M a h ~ n y ~ ~ to propose a 9.2-inch cyclotron specifically for helium-3 irradiation at 8 MeV. A 30-inch general-purpose machine has been designedwhich is suitable for activation analysis; the Isotopic particle sources have found little application to activation analysis because of their low intensity and limited flexibility, but their small size is an advantage when space and weight is restricted, and curium-242 has been evaluated as an a-source for analysing lunar surfaces.38 The low depth of penetration of charged particles into matter, which may be accompanied by high heat dissipation, limits not only the type and form of sample that can be irradiated, but also the method of containment.Silicon samples, 3 x 3 x 40 mm, have been encapsu- lated in tantalum foil39 for cyclotron irradiation, platinum40 has been used to contain targets to be irradiated with helium-3 ions, and hydraulically pressed without any container have been held in the accelerator vacuum for the measurement of prompt radiation.More recent activation work with charged particles has been concerned with the use of helium-3 as an irradiating particle, as the helium-3 nucleus, which has a low binding energy, reacts exo-ergically with many elements ; in particular, sensitive methods of analysis are available for a number of light elements.36 Oxygen has been determined in a number of matrices by the reaction, 160(3He,p)18F and 160(3He,n)18NezgF, and the sensitivity has been estimated at a few parts per lo9 by M a h ~ n e y . ~ ~ However, the low depth of penetration of helium-3 ions necessitates the very care- ful removal of surface contamination. The reaction 12C(3He,~)11C can be used as a basis for the determination of carbon, and the sensitivity of the method has been reported43 as being better than 1 part per lo9.Analysis of the decay curves permits carbon and oxygen to be determined simultaneously. By using auto-radiography after irradiation of samples with helium-3 ions, Holm44 studied the location of concentrations of oxygen and carbon. Beam in- homogeneities and surface contamination complicated the technique, but concentrations at grain boundaries were visible. Carbon and oxygen have also been determined after irradiation with a-particles by E ~ ~ g e l m a n n . ~ ~ The product, fluorine-18, of the reaction l60 + a has a sufficiently long half-life to permit easy chemical separation but interference is experienced from the reaction lgF(a,orn)l8F.Carbon determinations have been based on the reaction l2C(cc,ccn)l1C, and by using a 10 pA per cm2 beam of 44-MeV a-particles, the estimated sensitivities for carbon and oxygen are 0.01 and 0.001 p.p.m., respectively. Oxygen has been determined in high purity silicon by a-par ticle activation. 46 The reaction llB(p,n)llC, induced by 20-MeV protons has been used by Gill39 to detect as little as 0-003 p.p.m. of boron in silicon. Measurement of phosphorus-30 produced by the reaction 30Si(p,n)30P was used to standardise the reaction. There are several reactions that interfere with this determinati~n~~; even at low proton energies the reaction 14N(p,a)11 C occurs. The reaction between nitrogen and protons, i.e., 14N(p,a)11C, has been used for the determina- tion of nitrogen48 in graphite at the 0-01 p.p.m.level. Boron can also be determined by deuteron irradiation with the aid of the reactions 1°B(d,n)llC and 11B(d,2n)11C, but once again nitrogen interferes, this time by the reaction 14N (d,an) 11C.49 Charged particles emitted as a result of reactor neutron irradiation can sometimes be used to induce secondary nuclear reactions of analytical interest and permit relatively cheap reactor irradiations to be used. In particular, the sequence “i(n,oc)T, 160(t,n)18F has been applied to the determination of oxygen in beryllium,50 gallium a r ~ e n i d e , ~ ~ on surfaces52 and to the determination of lithiumb3 in the presence of alkali metals. Aumann and Born54 assessed the oxygen-18 content of samples by the l80(p,n)lsF reaction induced by recoil protons, and deuterium55 in deuterated organic compounds has been measured by the reaction 12C (d,n) 13N.Y-PHOTOKS- y-Photons provide an alternative penetrating radiation to neutrons for inducing nuclear reactions, and they have been used mainly for the determination of those elements that cannot be measured satisfactorily after irradiation with reactor neutrons. is likely to be less than $200,000.8 COLEMAN AND PIERCE : ACTIVATION ANALYSIS [Analyst, Vol. 92 A variety of particles may be emitted as a result of photo-disintegration of a nucleus but most y-ray activation techniques are based on the (y,n) reaction. Photoneutron thresholds usually exceed 5 MeV, but photons of much higher energy are normally used to obtain adequate sensitivity and therefore samples are usually irradiated with bremsstrahlung from a betatron or linear accelerator.The penetration of y-photons permits easy encapsulation of targets, and pneumatic transfer systems56 can be used to move samples in and out of the irradiation position. Deuterium and beryllium exhibit low neutron binding energies of 2.2 and 1.7 MeV, respectively, and y-rays from isotopes such as antimony-124 have been used by Gold~tein~~ and many earlier workers to induce the (y,n) reaction in beryllium. Er~gelmann~~ has described the use of y-ray activation for several important light elements such as carbon, nitrogen and oxygen. The sensitivity of the method was better than 1 pg for the three elements. Schweikert and Albert59 have considered the potential use .of y-photon activation for heavier elements. Twenty-four elements were irradiated with bremsstrahlung from a linear accelerator and the induced activity counted; sensitivities were found to be better than 1 pg for many of the elements. y-Activation techniques have been used to analyse ores60 and concentrates for light elements, and for copper, zinc and zirconium. Samples of 5 to 500g were irradiated and the induced activity detected by y-ray scintillation spectrometry; interferences were reduced by careful choice of photon energy. Mulvey, Cardarelli, Meyer, Cooper and BurrowsG1 determined iodine by y-activation by using a 22-MeV linear accelerator, and detecting the 0.386 and 0-650-MeV y-rays of iodine-126; 1 pg was measured with a coefficient of variation of 0.56 per cent.Large accelerators are required for most y-photon activations, and it may not be either desirable or feasible to allocate most of the machine time to analytical work. Under these circumstances, y-photon activation can be used as a referee method62 to supplement more conventional methods of analysis. MEASUREMENT OF INDUCED ACTIVITY The method chosen for the measurement of the activity of an irradiated sample is governed by the decay characteristics of the active species and the relative activity of the element to be determined to the total activity of the sample. There is an increasing tendency towards non-destructive methods, which require specific counting methods such as y-spectrometry. RADIOCHEMICAL SEPARATIONS- Although instrumental methods of activation analysis have received considerable atten- tion over recent years, the high separation factors that can be achieved by chemical techniques still result in chemical methods being used in many activation procedures, particularly when a small yield of activity of one element is sought in the presence of a highly active matrix or the maximum sensitivity is required.Chemical separations are usually devised to yield the elements to be determined, either in a radiochemically pure form that can be counted by simple counting equipment, or into several groups which can then be further examined by y-ray spectrometry. Inactive carriers are usually present during solution of the sample to enable a correction to be made for losses occurring during chemical processing, but treat- ment of the sample before irradiation has been carried out in a limited number of activation schemes when losses are expected to be Surface etching after irradiation can remove impurities from suitable samples.The radiochemistry of a great many elements is summarised in a series of monographs issued by the U.S. National Academy of Sciences, Nuclear Science Series NAS-NS, and many references are quoted. In addition, several laboratories publish their own separation procedures. Precipitation is still widely used for scavenging, for separating one or more nuclides from several others, preparing sources suitable for counting and for chemical-yield determina- tions. Precipitation from homogeneous solution can improve the purity of the precipitate, and S a l ~ t s k y ~ ~ has reviewed this and other forms of precipitation.The speed and versatility of solvent extraction makes the technique suitable for radio- chemical separations ; the book by Morrison and F r e i ~ e r ~ ~ is a valuable source of information, and Green66 has reviewed the uses of liquid ion exchangers.January, 19671 COLEMAN AND PIERCE ACTIVATION ANALYSIS 9 A monograph has been published on the use of high molecular weight amines in solvent e ~ t r a c t i o n , ~ ~ and a recent review by Morrison contains references to extraction procedures for many elements.68 Differential migration techniques have been extensively applied to radiochemical separa- tions, and in particular to ion-exchange chromatography on columns of organic ion exchangers ; the book by Samuelson is a useful guide for devising ion-exchange separation^.^^ Inorganic ion exchangers have been applied to radiochemical separations less frequently, but Girardi, Merlini, Pauly and Pietra have reported their use in an extensive chromatographic separation scheme.70 Reversed phase partition chromatography, in which the column material is an inert support retaining an organic extractant, has been developed more recently, but several workers have applied it to the radiochemical separation of a sample after irradiation.Towell, Volfousky and Winchester have determined rare-earth abundances in the standard granite G-1 and the standard diabase W-1 by neutron activation after separation of the elements on columns of di(2-ethylhexyl) hydrogen p h ~ s p h a t e .~ ~ Paper chromatography has also been used in chemical separation schemes for neutron-activation analysis, and in certain cases the separation has been carried out before irradiati0n7~ although limitations are imposed by impurities in the paper and radiation damage of the chromatogram if this is irradiated. Distillation can yield fast and clean separations of some elements and De V O ~ ' ~ has reviewed the technique in a short monograph. A simple apparatus has been described for vacuum distillation which permits the rapid distillation of small amounts of material.74 Other separation techniques applied to radiochemistry include electro-analytical methods, for example, controlled electro-potential deposition,75 amalgam exchange76 and the ring-oven technique.77 RAPID SEPARATIONS- Separation of short-lived radionuclides must be effected rapidly after the completion of irradiation, and special techniques have been devised for this purpose, Kusaka and Meinke78 have reviewed rapid radiochemical separation techniques, and Rengan and Meinke79 have described a rapid separation of rare earths by elution from columns of cation exchanger with a-hydroxyisobutyric acid.SamsahlsO has also carried out rapid separation with ion exchangers, but has used automatic equipment that permits many elements to be separated into several groups. Ruch, De Voe and Meinkes1 have demonstrated the speed of amalgam- exchange separations by separating indium from several elements in 11 minutes with a yield of higher than 95 per cent.The method was found to give more satisfactory results than the conventional bromide extraction. Kemp and Smaless2 have used solvent extraction for the rapid separation of 3-76-minute vanadium from rocks and meteorites after neutron irradiation. The vanadium was extracted into chloroform as the cupferron complex in the presence of EDTA, and the activity was measured 13 to 14 minutes after the sample left the reactor. Bakers3 has separated the 2-1-minute oxygen-15 isotope formed by the (y,n) reaction on oxygen-16 by inert gas fusion in a graphite furnace at 2000" C, and by subsequent gas phase and solution chemistry. In this way oxygen has been determined in steels and other metals down to 1 p.p.m. after irradiation with 40-MeV bremsstrahlung. Rapid radiochemical separations have recently been carried out on thin-layers of ion- exchange resins4 by using centrifugal acceleration to reduce development time, and Wolfs5 has described apparatus for rapid radiochemical separation by co-precipitation with a carrier.GROUP SEPARATIONS- A chemical separation procedure may be devised to divide the constituents of a sample into several groups which are then examined by y-ray spectrometry. This technique has been applied to a variety of samples. Samsahl, Brune and Westers6 have described a scheme for the determination of 30 trace elements in cancerous and non-cancerous human tissue after chemical separation into 16 or 18 groups. The separation scheme was based on distillation and the use of organic and inorganic ion exchangers.For the determination of 62 elements in high purity beryllium, aluminium and iron, Ross87 has combined non-destructive and chemical techniques. Forty-nine of the elements were determined after separation into 6 groups by distillation, precipitation and solvent extraction. Coulomb and Schiltzss have used paper chromatography to provide a preliminary separation of elements derived from geochemical materials. After separation, the y-spectra10 COLEMAN AND PIERCE ACTIVATION ANALYSIS [Analyst, Vol. 92 of different bands of the chromatogram were measured and analysed by computer techniques. Aubouin and Laverlocheresg have published a comprehensive ion-exchange separation for 30 elements and have determined impurities in several matrices. Many elements are separated individually, but the scheme provides a number of small groups which can be easily assayed by y-ray spectrometry.AUTOMATIC SEPARATION SYSTEMS- Several workers have operated chemical separations automatically to reduce the effort required when many samples have to be processed. Girardi, Merlini, Pauly and Pietra70 have devised an automatic system of chromatographic separation that is based on the use of two units: ( a ) , a pump and programming unit; and ( b ) , a fraction collector and motor-driven stopcock. By varying the number and arrangement of these units, constituents of a sample may be chromatographed through one or more columns and eluted with different solvents. Group separations may be effected by sorption of the constituents of the sample on to several columns connected in series. Samsahlso has devised a pumping unit which auto- matically introduces solution to the solvent stream at points between columns to achieve the conditions necessary for selective sorption.The technique has been applied to the group separation of trace elements in biological materiaLg0 Comar and Le Poecgl have reported an automatic method for the separation of iodine in biological fluids as part of a neutron-activation procedure. SUBSTOICHEIOMETRY- Activation procedures involving radiochemical separation usually require measurement of both the radioactivity and chemical yield of a radionuclide, but if a known or carefully controlled amount of the element can be sampled, only a measurement of activity is required. Sampling may be carried out by completely converting a known amount of reagent to a compound with the radionuclide and determining the activity of this compound ; the amount of reagent added must be insufficient to combine with all of the radionuclide present in the system.This technique, known as substoicheiometry, has been described in detail by RdiiEka and Star$,92 and procedures for many elements have been given. If the reagent reacts with the radioactive element to be determined in preference to other elements in the system, a substoicheiometric step offers some selectivity which may be exploited to simplify and increase the speed of the separation procedure. Thus a substoicheio- metric finish has been used for a determination of silverg3 based on measurement of the 2.3- minute silver-108 nuclide.RdiiEka and Williamsg4 have described a system for the continuous operation of sub- stoicheiometric isotope dilution analysis, but the method can also be extended to activation analysis. +SPECTROSCOPY- The most widely used measuring technique in activation analysis is now y-spectrometry ; this generally involves absorption of the y-rays in a sodium iodide crystal and then processing the pulses with a multi-channel analyser. Other crystal detectors, such as caesium iodide and calcium iodide, have been developed, but because of extra cost and, for the latter, technical difficulties in making large crystals, they have not found wide application. For background information the book edited by Crouthamelg5 is useful; it also contains a compilation of y-spectra that assists in identifying unknown species.More recently, Heathg6 has published an excellent two-volume report, which contains detailed information on his system and also over 200 spectra in graphical and digital form. These data can then be transferred to paper or magnetic tape and any selected spectrum read into the analyser, so allowing a direct visual comparison with the spectrum of interest. Other compilations have been produced specifically for activation analysis. Girardi, Guzzi and Paulyg7 have listed all the reactions normally used in thermal neutron activation which produce y-emitting nuclides. The y-spectrum is illustrated and additional information is provided to allow the activity to be rapidly determined for any weight of material from a combination of graphs of decay growth during irradiation and sensitivity.Aude and Laverlochereg8 have produced experiment ally a similar compilation for fast -neutron products. Before buying equipment for y-spectrometry it is important to compare carefully the manufacturer's specifications with the experimental requirements. Crouch and Heathg9 haveJanuary, 19671 COLEMAN AND PIERCE ACTIVATION ANALYSIS 11 specified the standards required and the means of testing the equipment for accurate spectro- metry by using computer methods of data analysis. The performance of analysers available early in 1965 is conveniently grouped together in a report,loO but new and improved equipment is constantly coming on to the market. The quality-control methods of CovelllOl can be recommended for establishing and maintaining spectrometer stability.If many samples are to be processed it is likely that an automatic sample changer will be required ; most commercial equipment is unsuitable for y-spectrometry inside a large shield, but changers that are suitable have been developed in several 1ab0ratories.l~~ 9 1 ° 3 With the advent of computer methods of analysis of y-spectra, the method of transferring data from the analyser to the computer has been studied. Euler, Phelps and Covelllo4 have described a fast punched-card read-out system, which collects data from a 256-channel analyser in 12 seconds and provides for the insertion of seven record and information words. The automated system of Wainerdil05 transfers the data in less than 1 second.Although he has not applied it to y-spectrometry, McNaughtlo6 has accumulated counting data on magnetic tape and then successfully transferred it into a computer via telephone wires. As nuclear physics experimentation increases in complexity and vast amounts of data are produced there has been a tendency towards the use of small computers for data collection instead of fixed-wire analysers. Cohan1°7 has used the N.B.S. system for activation analysis, and SpinradlOs gives a general comparison of the merits of computers and analysers for multi- parameter analysis. The use of small computers will increase in activation analysis, par- ticularly when there is a need for on-line analysis, because the computer can be used for data processing as well as for a store.However, the conventional analyser is likely to provide the most economical answer for most routine operations for some time. The control of drift in a spectrometer is of particular importance if accurate measurements are required from complex spectra by computer methods. Many systems have been described which monitor the position of peaks in the spectrum derived from a y-emitter, &-emitter, a light source or a pulse generator. Comparatively few, however, control both the gain and threshold settings of the analyser, and in the authors’ opinion both are liable to drift. Dudley and Scarpatellilo9 have a method that limits changes in gain and threshold to less than 0-1 per cent. and does not add unwanted peaks to the spectrum. The change in the system can also be corrected for by computer methods.Schonfeld’sllO programme allows for automatic data processing as well as for a store. However, the conventional analyser is likely to provide changes in gain and threshold and has been used most satisfactorily in the author’s (R.F.C.) laboratory. Both of these last methods correct for all sources of drift, including the effect of high count-rates on the analyser and photomultiplier. For quantitative measurement of nuclides it is usual to standarise the equipment by count- ing sources prepared from a known weight of irradiated material. Poor standards are a frequent source of error in activation analysis. The material used for standards should be free from impurities that could give rise to interfering activities, and the irradiation conditions, counting geometry and total amount of scattering material near the source should be as similar as possible for sample and standard.For the evaluation of simple spectra containing a few well resolved peaks the method of Covelllll is widely used and is quite satisfactory. Yule112 has applied convolution techniques to y-spectra and then used a computer method to locate and determine the area of the photopeak for simple spectra. The direct determination of mixtures by complement subtraction, as proposed by Lee,113 is possible with the accessories available with most analysers. This technique was further improved by Anders and Beamerl14 by using a small computer to normalise the spectra for decay, neutron flux, etc., before stripping. To resolve and evaluate complex mixtures, a more sophisticated approach with larger computers is necessary. Zerby115 attempted to calculate pulse-height distributions from a knowledge of basic interactions occurring in a detector by using Monte Carlo methods. Heath116 was more successful, and empirically calculated the response function for any detector and experimental geometry.Many spectra of single y-ray emitters are measured to charac- terise the system, and then, by suitable machine programming, any unknown spectrum can be generated from an empirical analytical function representing the pulse-height distribution and a knowledge of the decay scheme. The resolution of the complex spectrum is then achieved by least squares analysis. Salmon,117 Trombka1lS and others have experimentally built up a library of standards and then used the least squares method.Perhaps it is surprising that although in activation analysis it is usual to measure several spectra over a period of[Analyst, Vol. 92 time, few analysts have used this extra parameter, the decay constant, in the analysis of spectra. Nicholson, Schlosser and B r a ~ e r ~ ~ ~ have developed such a programme, and more recently Schonfeld120 has applied this approach to activation analysis and demonstrated the improvement in precision. Computer methods of analysis of y-spectra have provided the key to rapid and economical routine analysis by activation methods. Automatic sample handling and irradiation facilities have been combined with computer methods of data analysis by Wainerdi, Menon and Fitel05 to provide a system capable of handling a great many of certain types of sample at low cost.Other workers have attempted to increase the scope of instrumental methods of analysis through coincidence and anti-coincidence counting. Ljunggren121 has used y - y coincidences for the measurement of several nuclides emitting y-rays in cascade. Borg et aL5 demonstrated the specificity of triple coincidence methods, two y-rays and a high energy /&particle, by determining manganese in blood directly without first removing the sodium-24 activity. The original method of Pierson,122 which compensates for the Compton continuum by subtracting the spectrum produced by an anthracene crystal from the sodium iodide spectrum, has been greatly improved by De Soete and H 0 ~ t e .l ~ ~ This reduction in the Compton con- tribution, by about 85 per cent. in the latter case, simplifies the task of interpreting and evaluating peak heights. Other methods of suppressing Compton pulses have been tried by surrounding the sodium iodide crystal with a large plastic ~ c i n t i l l a t o r , ~ ~ ~ a large sodium iodide crystal125 or a caesium iodide crystal.126 The most sophisticated and expensive device has been developed by Perkins,12' and consists of two sodium iodide crystals surrounded by a 12-inch diameter sodium iodide detector. An extremely low background is achieved, and by measuring y - y coincidences with a multi-parameter analyser, very complex mixtures can be readily determined. The use of sum coincidence y-spectrometry has been demonstrated; Adams and Hoste128 showed an improvement in sensitivity of a factor of 32 compared with the Compton com- pensated technique in the determination of antimony in lead.Wahlgren, Wing and H i n e ~ l ~ ~ have applied the method generally, and compare the sensitivity of a fast sum coincidence system with simple y-spectrometry. A relatively new and rapidly developing area of y-spectrometry involves the use of lithium- drifted germanium. Currently, detectors are commercially available in sizes up to about 2.5 cm in diameter and 1 cm thick. Larger volumes have been made at Chalk River by Malm, Tavendale and by using an axial drift technique; as these larger sizes become generally available they should find wide application in activation analysis.The resolution of such a detector is 4.8 KeV for 1.33-MeV cobalt-60 y-ray, and gives complete separation of many photopeaks unresolved by sodium iodide. The necessity of operating germanium semi- conductors in a vacuum and at 77" K is overcome by suitable design of equipment, and access to the detector is then quite easy. The application of these new detectors to activation- analysis problems has been demonstrated by Girardi, Guzzi and Pauly131 in analysing hafnium in zirconium oxide and by Prussin, Harris and Hollander132 in the non-destructive analysis of aluminium. OTHER COUNTING METHODS- P-Counting with Geiger or proportional counters is still frequently used for measuring the activity after a radiochemical separation that ensures purity of the source.Because of the high efficiency (up to 100 per cent.), such counters have advantages in achieving the maximum sensitivity possible by activation methods. Cerenkov counters have found some application for the measurement of high energy p-particles. L ~ k e n s l ~ ~ counted P-particles from nitrogen-16 (maximum energy 10 MeV) and demonstrated the high efficiency of rejection of weaker ,&particles, even from chlorine-38 (maximum energy 4.8 MeV). The Cerenkov counters were superior to sodium iodide crystals for the measurement of oxygen in some materials. Fission counters have been used by N i l ~ s o n l ~ ~ for the accurate assay of small samples of uranium; 100 to 140-mg samples can be determined with a precision of 0.7 per cent. (two standard deviations).Auto-radiography can be used for the quantitative measurement of boron135 by counting recoil tracks after irradiating the sample and emulsion in a reactor. The method is useful in defining the position of impurities in a sample; for example, the study of segregation in metals by Makin.136 12 COLEMrlN AND PIERCE : ACTIVATION ANALYSISJanuary, 19671 COLEMAN AND PIERCE ACTIVATION ANALYSIS 13 For a limited number of elements, neutron counting has provided a specific method of analysis. By using the delayed neutrons from fission, two groups of workers, Amie1137 and Dyer, Emery and Leddic~ttel~~ have developed a rapid, routine method for uranium. In the author’s (R.F.C.) laboratory, more than 100 urine and rock samples are analysed weekly, each measurement requiring less than 3 minutes.Thorium, oxygen and lithium can also be determined, but with lower sensitivity, by modifying the conditions of irradiation. Amie1139 has further extended the value of neutron counting by using the photoneutrons emitted when high energy y-rays bombard deuterium and beryllium for the analysis of sodium, manganese, sulphur and calcium. PROMPT RADIATION TECHNIQUES- Analytical methods based on the measurement of prompt radiation have received rela- tively little attention. Prompt radiation must be measured with the sample in sitzi, usually with the irradiation in progress, and chemical separation of the constituents of a sample between irradiation and counting cannot be used to isolate the elements to be determined from all others, or to etch and remove surface contaminants.Moreover, interference is likely from any contamination that might build up on the sample, for example, carbon during accelerator irradiation. Therefore, if pre-separation of the constituents of the sample is to be avoided, prompt radiation techniques must rely heavily on instrumentation to distinguish the radiation to be measured from others emitted by the samples and from machine back- ground. However, prompt techniques permit the radiation emitted during the decay of excited nuclear states to be counted, so avoiding the necessity of choosing a nuclear reaction producing a radioactive isotope as a basis for the determination. Thus, a 0-87-MeV y-ray is emitted as a result of the irradiation of oxygen-16 with low energy deuterons, corresponding to the de-excitation of oxygen-17 produced by the (d,p) reaction, although both oxygen-16 and oxygen-17 are stable.Analytical methods have been proposed that are based on the measurement of particles of y-radiation. Energies of charged particles can be found with a magnetic spectrometer or with a semiconductor detector, but if thick targets are used, broadening of the particle lines caused by energy loss in the sample complicates spectral analysis. However, Patterson, Turkevich and Fran~grotel~~ have examined the possibility of using Rutherford scattering and (cc,p) reactions to analyse geochemical surfaces. Particle scattering can also be used to investi- gate thin films on s ~ r f a c e s , ~ ~ ~ , ~ ~ ~ particularly when a heavy element is present on alighter matrix, and layers less than 1-,u thick can be measured.Prompt neutrons have also been measured143; for example, deuterium in surface layers of Zircaloy-2 has been determined by measuring the neutrons emitted during the reaction D(d,n)3He. Standardisation was effected by comparing the neutron yield from the target with that from a deuterium gas target cell. Capture y-rays emitted during neutron irradiation are often complex, but Greenwood and Reed144 have discussed possible activation applications and compiled y - ~ p e c t r a . ~ ~ ~ The high cross-section of the reaction lOB(n,~x)~Li for thermal neutrons feeding the 0.478-MeV level in 7Li has been used by Isenhour and Morrison146 as the basis of a method for determining boron. Measurement of prompt y-rays emitted during irradiation of a sample with neutrons from a neutron generator is complicated by the problem of shielding the y-ray detector, and ring geometry, often used in nuclear physics experiments, requires a large sample which is not always convenient for analytical work.However, nitrogen has been determined in bulk material by the measurement of capture radiation with a polonium - beryllium source,147 and carbon has been determined in coal by the inelastic scattering of 14.8-MeV neutrons. Prompt y-rays emitted during irradiation of a sample with charged particles provide analytical methods for the determination of many light e l e m e n t ~ , ~ ~ ~ 1 ~ ~ ~ but information is only obtained from a thin section of the sample near to the surface.If a carefully collimated particle beam is used, localised areas of the surface of the sample can be examined. APPLICATIONS I t is not possible to review critically all of the fields to which activation analysis has made However, it was considered desirable to attempt to select a a substantial contribution.14 COLEMAN AND PIERCE : ACTIVATION ANALYSIS [Afialyst, Vol. 92 few reviews and papers to illustrate the wide application of activation analysis, and to provide a starting point for a more extensive literature search. HIGH PURITY MATERIALS- The high sensitivity of activation analysis for many elements has resulted in the tech- nique being applied to the determination of trace constituents of pure materials. Again, freedom from reagent blank during chemical processing enables the high sensitivity of the technique to be exploited, even when chemical processing of the sample is necessary between measurement and counting.When the matrix is relatively insensitive to activation by the radiation with which the sample is being irradiated, trace elements may sometimes be determined non-destructively. Lightowler~l~~ has determined small amounts of manganese, sodium and copper in natural diamonds by y-ray spectroscopy after neutron activation, and Petit and Engelmannlsl have described non-destructive methods of beryllium analysis. However, it is more often necessary to take advantage of the high separation factors that can be achieved by chemically separating the activity to be measured from all others before measurement.Thus, Gebauhr and Martin152 have described a scheme for the determination of thirty elements present at trace levels in high purity silicon. Auto-radiography was used to localise the impurities. A field of trace analysis to which activation analysis has been extensively applied has been the determination of trace-level impurities in semiconductor materials, and Cali153 has described the methods used in detail. Laboratories without direct access to suitable irradiation facilities may find value in using activation analysis as a referee method. Tellurium154 has been determined in tellurium- doped gallium arsenide by neutron-activation analysis to check the validity of spectrophoto- metric measurements. Extensive reference is made in bibliographies to the individual determination of many elements at trace levels in a variety of materials.BIOLOGY- The book by Bowen and Gibbons155 provides an excellent starting point for obtaining details of biological applications of activation analysis. Most of the papers up to 1961 are reviewed, and also the reported values of elemental concentration in organisms and mammalian tissue are tabulated. Practical details of methods of analysis for twenty elements in biological tissue have been published by Bowen and Cawse.156 Data on the elemental composition of blood as determined by activation analysis and many other methods have been collected by Bowen,15' and indicate that large discrepancies occur for many elements. A biological reference materiall58 has been prepared to permit intercomparison of analyses from different laboratories, and ultimately this should lead to an improvement in the reliability of analytical measurements.The proceeding~15~ of the biological symposium held at Saclay in 1964 demonstrate the wide variety of biological applications of activation analysis. FORENSIC SCIENCE- There has been considerable interest in the application of activation analysis to forensic science problems in the last few years, and useful reviews have been written by Guinn,lG0 and by Smith and Lenihan.lG1 Smith162 has demonstrated the value of activation analysis in the investigation of arsenic poisoning. Ruch, Guinn and Pinker1G3 have detected excess of antimony and barium on the hands of people who have recently fired a gun, and this method has already found some use in criminal investigations.Much of the work in forensic science laboratories is aimed at establishing if two samples have a common origin. As has been indicated by several papers, e.g., those by Ruch, Buchanan, Pinker, Guinn and BellancalG4 and Bate, Emery, Leddicotte, Lyon and activation analysis can be applied to this problem and so determine the trace-element content of many materials, such as hair, glass, rubber and plastics, that are of interest to forensic scientists. It is assumed that the variation in trace-element content of samples of different origin will appear significantly different, and only closely match in two samples of common origin. This is a dangerous assumption that is only justified if many samples have been analysed, indicating that a large variation in concentration does exist.Coleman166 has investigated the trace- element content of a great many hair samples and has used a statistical method to assess quantitatively the value of this kind of evidence for identification purposes.January, 19671 COLEMAN AND PIERCE ACTIVATION ANALYSIS 15 GEOCHEMISTRY- Activation analysis has found application to the determination of major, minor and trace constituents in geochemical samples, but methods of trace-element analysis have found widest application, as the efficient separation of small amounts of activity from a complex and highly active matrix by chemical techniques can be effected without introducing a reagent blank, and correction can be made for losses that may occur during chemical processing.MappeP7 has discussed in detail the application of radioactivation to geochemistry and has given many examples. The scope of activation analysis for geochemical determinations can be gauged from a summary of results recently obtained for the levels of different elements in the standard diabase W-1 and the standard granite G-1 in which reference is made to the determination of more than forty elements168 by neutron-activation analysis. Several authors have compared results found by activation analysis with those obtained by other analytical techniques. Thus Ball and F i l b ~ l ~ ~ obtained good agreement for the zinc content of some geochemical standards measured by neutron activation and X-ray fluorescence techniques, and Smales, Hughes, Mapper, McInnes and Webster170 have compared results obtained for the rubidium and caesium content of stony meteorites by neutron- activation analysis and by an isotopic dilution method of mass spectrometry.For the latter, even at rubidium and caesium levels lower than 1 p.p.m., differences in results obtained by the two techniques exceeded 20 only for one sample in every twelve. Fast-neutron activation has been used to obtain figures for the silicon171 and oxygen172 contents of stony meteorites by using the reactions 28Si(n,p)28A1 and l60(n,p)l6N, respec- tively. Measurements were non-destructive, and the precision for the silicon determinations was usually found to be better than +3 per cent. The relatively small size of some neutron generators enables them to be used in portable instruments for geochemical analysis ; Caldwell, Mills, Allen, Bell and Heath173 have described the use of a pulsed fast-neutron source for the remote analysis of lunar and planetary surfaces.y-Rays produced by three different processes are distinguished by their time of emission with respect to the neutron pulse; y-rays from inelastic scattering are emitted during the pulse, capture radiation is emitted between pulses after the neutrons have been slowed down and y- rays from isotopes, produced, for example, by (n,p), (n,2n) and (n,cc) reactions, build up over a number of pulses. In addition, information about the composition of the target can be obtained from the rate at which the capture radiation, and hence the thermal neutrons, die away.ON-LINE ANALYSIS- Many analytical techniques are being adapted for process control, and activation methods can make a useful contribution in this field; Taylor's on activation analysis devotes a chapter to on-line analysis. A typical example has been the extensive study by Martin and co-worker~l7~ into coal analysis. Under laboratory conditions, carbon, oxygen, aluminium and silicon can be determined with an accuracy of 1 to 3 per cent. by using an accelerator neutron source and measuring y-rays produced after inelastic scattering and from the decay of radionuclides. Further work is continuing in several laboratories to apply such techniques on a plant scale. The continuous measurement of perborate in detergents has been examined by Ljunggren and Christell.l76 Neutron absorption by boron-10 and also the measurement of y-rays from the reaction loB(n,acy)7Li can form the basis of the analytical method.A nuclear method of monitoring the addition of fluorine to drinking water by using a 14-MeV neutron source and counting the activation product, oxygen-19, in a special cell downstream from the neutron source has been demonstrated by Norgolwalla and Jervis.17' On-line analysis would be of value in ore sorting plants; Ramdohr's on copper minerals is an example of the use of activation analysis. SURFACE ANALYSIS- Several nuclear methods are available for examining the surface of samples and these frequently involve the use of charged-particle irradiation. Charged-particle beams can be collimated relatively easily, and the depth of penetration of particles into the sample can be adjusted by careful control of particle energy; particle energy spectra can be obtained with a magnetic spectrometer or with a semiconductor detector assembly.16 COLEMAN AND PIERCE : ACTIVATION ANALYSIS [Aaalyst, Vol.92 Elastic scattering techniques, which have been described by Rubin,141 are capable of a high sensitivity, and can be used to examine layers of much less than 1 p in thickness, although if thick targets are used the technique is most satisfactorily applied to the deter- mination of a heavy element on the surface of a lighter matrix. Pliesach and Poole142 have used a semiconductor detector to measure the scattered particles. When there is a sharp resonance in the excitation function of a reaction of interest, the variation of yield with energy can provide useful information about the composition and thickness of surface films.Thus, Amsel and Samuel179 have used sharp resonances in the reaction l8O(p,cc)l5N to follow the growth of oxide films during anodic oxidation. The use of auto-radiography after helium-3 irradiation to locate concentrations of light elements in surfaces has been mentioned. In addition to conventional radiographic techniques, charged-particle tracks in a variety of materials have been used to provide information about surfaces, in some cases after enlarging the damage trails by chemical attack. For example, cc-particle tracks in plastic sheets have been used to study the inhomogeneous distribution of boron in the surfaces of metallurgical samples, and uranium has been deter- mined in very small samples down to levels of Fleischer, Price and Walkerlso have discussed analytical applications of track counting in some detail.L e ~ n h a r d t ~ ~ has used tritons, emitted as a result of the reaction “i(n,ct)T, to produce fluorine-18 from surface oxygen by the reaction l60(tJn)l8F, and was thus able to use a nuclear reactor as a radiation source. The amount of oxygen on the surface of foils was calculated from the annihilation radiation of the fluorine-18 after separation, and was found to vary from 1.4 pg of oxygen per cm2 for platinum to 7-5 pg of oxygen per cm2 for iron. p.p.m. by fission track counting. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. REFERENCES Girardi, F., Guzzi, G., and Pauly, J., Analyt. Chew., 1964, 36, 1588. KO&, R. C.,’ “Activation Analysis Handbook,” Academic Press, New York and London, 1960. Durham, R. W., Navalkar, M. P., and Ricci, E., “Proceedings of the International Conference Borg, D. C., Segal, R. E., Kienle, P., and Campbell, L., I n t . J . Appl. Radiat. Isotopes, 1961, 11, 10. Yule, H. P., Lukens, H. R., and Guinn, V. P., Nucl. Instvum. Meth., 1965, 33, 277. Lukens, H. R., Yule, H. P., and Guinn, V. P., Ibid., 1965, 33, 273. Hogdahl, 0. T., “Radiochemical Methods of ,4nalysis,” International Atomic Energy Agency, Kamenoto, Y., I n t . J . Appl. Radiat. Isotopes, 1964, 15, 447. Ricci, E., and Dyer, F. F., Nucleonics, 1964, 22, 6, 45.Bock-Werthmann, W., Atomkernenergie-Docuwentation, Beim Gmelin Institut, AED-C-14-01, 1961 ; Nucleonics, 1965, 23, 4, 60J. Strain, J. E., in Elion, H. R., and Stewart, D. C., Editovs, “Progress in Nuclear Energy,” Pergamon Hollister, H., Nucleonics, 1964, 22, 6, 68. Bounden, J. E., Lomer, P. D., and Wood, J . D. L. H., Nucl. Instvum. Meth., 1965, 33, 283. Iddings, F. A,, Analytica Chiw. Acta, 1964, 31, 206. Coleman, R. F., Analyst, 1962, 87, 590. Anders, 0. U., and Briden, D. W., Analyt. Chem., 1963, 36, 287. Mott, TV. E., and Orange, J. E., in Guinn, J. P., Editov, “Proceedings of the 1965 International Conference on Modern Trends in Activation Analysis,” Texas A and M IJniversity, 1966, p. 115. Anders, 0. U., and Briden, D. W., Analyt. Chem., 1965, 37, 530.Walker, L. J,, and Eggebraaten, V. L., in Guinn, J. P., Editor, 09. cit., p. 169. Blackburn, R., Analyt. Chem., 1964, 36, 669. Anders, 0. U., Ibid., 1960, 32, 1368. Steele, E., and Meinke, W. W., “Proceeding of the International Conference on Modern Trends in Activation Analysis,’’ Texas A and M University, 1961, p. 161. Vogt, J. R., Ehmann, W. D., McEllistream, M. T., I n t . J . Appl. Radiat. Isoto$es, 1965, 16, 573. Gorski, L., Kusch, W., and Wojtkowska, J., Talanta, 1964, 11, 1135. Gillespie, A. S., and Hill, W. W., Nucleonics, 1961, 19, 11, 170. Coleman, R. F., Aaalyst, 1961, 86, 39. Amiel, S., “Radiochemical Methods of Analysis,’’ International Atomic Energy Agency, Vienna, De, A. K., and Meinke, W. W., Analyt. Chew., 1958, 30, 1474. Grunewald, R., Ziemer, P.L., and Christian, J. E., Ibid., 1964, 36, 1138. Bakes, J. M., and Jeffery, P. G., Analyst, 1966, 91, 217. Strain, J . E., and Lyon, W. S., “Radiochemical Methods of Analysis,” International Atomic Sippel, R. F., and Glover, K. D., Nucl. Instruvn. Meth., 1960, 9, 37. , Ibid., 1965, 37, 1085. - -__ on Modern Trends in Activation Analysis,” Texas A and M University, 1961, p. 67. Vienna, 1965, Volume 1, p. 23. AED-C-14-02, 1963; AED-C-1P03, 1964. Press, 1965, Series IX, Volume 4, Part 3, p. 139. 1965, Volume 1, p. 258. Energy Agency, Vienna, 1965, Volume 1, p. 245.January, 19671 COLEMAN AND PIERCE ACTIVATION ANALYSIS 17 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70.71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. Ricci, E., and Hahn, R. L., Analyt. Chem., 1965, 37, 742. Markowitz, S. S., and Mahoney, J . D., Ibid., 1962, 34, 329. Fleischer, A. h., Trans. Amev. Nucl. Soc., 1965, 8, 84. Patterson, J . H., Turkevitch, A. L., and Franzgrote, E., J . Geophys. Res., 1965, 70, 1311. Gill, R. A., U.K. Atomic Enevgy Authovity Keport, A.E.R.E. C/R 2758, Harwell, 1958. Demildt, A. C., Analyt. Chem., 1963, 35, 1228. Pierce, T. B., Peck, P. F., and Henry, W. M., Analyst, 1965, 90, 339. Mahoney, J. D., U.S. Atomic Energy Commission Report, UCRL 11780, 196!. Markowitz, S . S., and Mahoney, J . D., “Radiochemical Methods of Analysis, International Atomic Holm, D., U.S. Atomic Energy Commission Report, L-4-DC 7319, 1962.Engelmann, C., “Radiochemical Methods of Analysis,” International Atomic Energy Agency, Saito, K., Nozaki, T., Tanaka, S., Furukawa, M., and Cheng, H., I n t . J . Appl. Kadiat. Isotopes, Rommel, H., Kernenevgie, 1962, 5, 859. Nozaki, T., and Yazawa, K., Bull. Chem. SOG. Japa,, 1964, 37, 1891. Rommel, H., Analytica Chim. Acta, 1966, 34, 427. Osmond, R. G., and Smales, A. A., Ibid., 1954, 10, 117. Bailey, R. F., and Ross, D. A., Analyt. Chem., 1963, 37, 791. Leonhardt, W., Analytica Chim. Acta, 1965, 32, 355. Smith, G. W., Santelli, D. J., and Fiess, H., Ibid., 1965, 33, 1. Aumann, D. C., and Born, H. J., I n t . J . Appl. Radiat. Isotopes, 1965, 16, 727. Fabbri, E., Lazzarini, E., and Sanquist, V., Ibid., 1964, 15, 437. Engelmann, C., Commissariat a 1’Energie Atomique, Report 2559, 1964.Goldstein, G., Analyt. Chem., 1963, 35, 1620. Engelmann, C., “Radiochemical Methods of Analysis,” International Atomic Energy Agency, Vienna, 1965, Volume 1, p. 341. Schweikert, E., and Albert, P., “Radiochemical Methods of Analysis,” International Atomic Energy Agency, Vienna, 1965, Volume 1, p. 323. Berzin, A. K., Vitozhents, G. C., Sulin, V. V., and Shornikov, S. I., “Radiochemical Methods of Analysis,” International Atomic Energy Agency, Vienna, 1965, Volume 1, p. 361. Mulvey, P. F., Cardarelli, J . A., Meyer, R. A,, Cooper, R., and Burrows, B. A., “Radioisotope Sample Measurement Techniques in Medicine and Biology,” International Atomic Energy Agency, Vienna, 1965, p. 249. Energy Agency, Vienna, 1965, Volume 1, p.429. Vienna, 1965, Volume 1, p. 405. 1963, 14, 357. Baker, C. A., Proc. SOG. Analyt. Chem., 1965, 2, 177. Kosta, L., and Cook, G. B., Talanta, 1965, 12, 977. Salutsky, M. L., in Kolthoff, I. M., and Elving, P. J., Editors, “Treatise on Analytical Chemistry,” Morrison, G. H., and Freiser, H., “Solvent Extraction in Analytical Chemistry,” John Wiley Green, H., Talanta, 1964, 11, 1561. Moore, F. L., “Liquid - Liquid Extraction with High Molecular Weight Amines,” U.S. Academy Morrison, G. H., Analyt. Chem., 1964, 36, 93R. Samuelson, O., “Ion Exchange Separations in Analytical Chemistry,” John Wiley and Sons, Girardi, F., Merlini, M., Pauly, J., and Pietra, R., “Radiochemical Methods of Analysis,” Inter- Towell, D. G., Volfouslq, R., and Winchester, J . W., Geochim.Cosmochim. Acta, 1965, 29, 569. Rakovic, M., Chemicke‘ Listy, 1964, 58, 1141. De Voe, J. R., “Application of Distillation Techniques to Radiochemical Separations, ’’ U S . De Voe, J . R., and Meinke, W. W., Analyt. Chem., 1963, 35, 2. Mark, H. B., jun., and Berland, F. J., Ibid., 1964, 36, 2062. De Voe, J . R., Nass, H. W., and Meinke, W. W., Ibid., 1961, 33, 1713. Hilton, D. A., and Reed, D., Analyst, 1964, 89, 599. Kusaka, Y., and Meinke, W. W., “Rapid Radiochemical Separations,” U.S. Academy of Sciences, Nuclear Science Series, NAS-NS 3104, 1961. Rengan, K., and Meinke, W. W., Analyt. Chem., 1964, 36, 157. Samsahl, K., Aktiebolaget Atomenergi Repovt, AE 159, 1964. Ruch, R. R., De Voe, J. R., and Meinke, W. W., Talanta, 1962, 9, 33. Kemp, D. M., and Smales, A.A., Analytica Chim. Acta, 1960, 23, 397. Baker, C. A., PYOG. Soc. Analyt. Chem., 1965, 2, 177. Ajuria-Garza, S., U.S. Atomic Energy Commission Report, UCRL 16015, 1965. Wolf, G. K., Kerntechnik, 1965, 7, 122. Samsahl, K., Brune, D., and Wester, P. O., Int. J . Appl. Radiat. Isotopes, 1965, 16, 273. Ross, W. J., Analyt. Chem., 1964, 36, 1114. Coulomb, R., and Schiltz, J. C., “Radiochemical Methods of Analysis,” International Atomic Aubouin, G., and Laverlochere, J., Commissaviat a L’Energie Atomique, Repovt CEA 2359, 1963. Samsahl, K., Aktiebolaget Atomenevgi Report, AE 215, 1966. Interscience Publishers, New York, 1959, Volume 1, Part I, p. 733. and Sons Inc., New York, 1962. of Sciences, Nuclear Science Series, NAS-NS 3101, 1960. New York, 1963. national Atomic Energy Agency, Vienna, 1965, Volume 2, p.1. Academy of Sciences, Nuclear Science Series, NAS-NS 3108, 1962. Energy Agency, Vienna, 1965, Volume 2, p. 177.18 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. COLEMAN AND PIERCE ACTIVATION ANALYSIS [A.naZyst, Vol. 92 Comar, D., and Le Poec, C., “Radiochemical Methods of Analysis,” International Atomic Energy RbZiCka, J., and Star$, J., Atom. Energy Rev., 1964, 2, 3. Pierce, T. B., and Peck, P. F., Analyst, 1963, 88, 603. RbiiCka, J., and Williams, M., Talanta, 1965, 12, 967. Crouthamel, C.E., Editor, “Applied Gamma-ray Spectroscopy,” Pergamon Press, Oxford, 1960. Heath, R. L., U S . Atomic Energy Commission Report, IDO-16880, 1964. Girardi, F., Guzzi, G., and Pauly, J., Euratom Report, EUR 1898e, 1965. Aude, G., and Laverlochere, J., “Spectres Gamma de Radioelements Formes par Irradiation sous Crouch, D. F., and Heath, R. L., U.S. Atomic Energy Commission Report, ID0 16923, 1963. Nucleonics, 1965, 23, 4, 60H. Covell, D. F., Nucl. Instrum. Meth., 1963, 22, 101. Robin, G., Atompraxis, 1965, 6, 325. Zinner, W. H., U.S. Atomic Energy Commission Report, H.W. 80798, 1964. Euler, B. A., Phelps, P. L., and Covell, D. F., Nucl. Instrum. Meth., 1965, 34, 266. Wainerdi, R. E., Menon, M. P., and Fite, L. E., U S . Atomic Energy Commission Report, TEES- McNaught, R.A., Nucl. Instrum. Meth., 1964, 31, 106. Cohan, M. D., in Guinn, J . P., Editor, op. cit., p. 157. Spinrad, R. J., Nucleonics, 1963, 21, 12, 46. Dudley, R. A., and Scarpatelli, R., Nucl. Instrum. Meth., 1964, 25, 297. Schonfeld, E., Kibbey, A. H., and Davis, W., U.S. Atomic Energy Commission Report, ORNL 3744, Covell, D. F., Analyt. Chem., 1959, 31, 1785. Yule, H. P., Ibid., 1966, 38, 103. Lee, W., Ibid., 1959, 31, 800. Anders, 0. U., and Beamer, W. H., Ibid., 1961, 33, 226. Zerby, C. D., and Moran, H. S., US. Atomic Energy Commission Report, ORNL 3169, 1962. Heath, R. L., U.S. Atomic Energy Commission Report, ID0 16784, 1962. Salmon, L., Nucl. Instrum. Meth., 1961, 14, 193. Trombka, J., Report JPL-TR-32-373, Jet Propulsion Laboratory, California Institute of Tech- nology, Pasadena, Calif., 1962.Nicholson, W. L., Schlosser, J. E., and Brauer, F. I?., U.S. Academy of Sciences Series NAS-NS 3107 1963, p. 254. Schonfeld, E., in Guinn, J. P., Editor, op. cit., p. 279. Ljunggren, K., “Radioisotopes in the Physical Sciences and Industry,” International Atomic Energy Agency, Vienna, 1962, Volume 2, p. 199. Pierson, D. H., Natuve, 1954, 173, 990. De Soete, D., and Hoste, J., Radiochim. Acta, 1965, 4, 35. Busuoli, G., Melandri, C., Rimondi, O., and Righini, B., Nucl. Instrum. iweth., 1963, 22, 324. Trail, C. C., and Raboy, S., Rev. Scient. Instrum., 1959, 30, 425. Lam, C. F., and Wainerdi, R. E., in Guinn, J . P., Editor, op. cit., p. 189. Perkins, R. W., and Robertson, D. E., in Guinn, J . P., Editor, op. cit., p. 48.Adams, F., and Hoste, J., Nucleonics, 1964, 22, 3, 55. Wahlgren, M. A., Wing, J . , and Hines, J., in Guinn, J. P., Editor, 09. cit., p. 134. Malm, H. L., Tavendale, A. J., and Fowler, I. L., Can. J . Phys., 1965, 43, 1173. Girardi, F., Guzzi, G., and Pauly, J., Radiochim. Acta, 1965, 4, 109. Prussin, S. G., Harris, J. A., and Hollander, J. M., Analyt. Chem., 1965, 37, 1127. Lukens, H. R., and Lasch, J . E., I n t . J . AppZ. Radiat. Isotopes, 1964, 15, 759. Nilsson, G., Nucl. Instrum. Meth., 1961, 13, 183. Edwards, L. C., I n t . J . Appl. Radiat. Isotopes, 1956, 1, 184. Makin, S. M., “Radioisotopes in the Physical Sciences and Industry,” International Atomic Amiel, S., Analyt. Chem., 1962, 34, 1683. Dyer, F. F., Emery, J. F., and Leddicotte, G. W., U.S. Atomic Energy Cowmission Report, ORNL Amiel, S., “ Radiochemical Methods of Analysis,” International Atomic Energy Agency, Vienna, Patterson, J .H., Turkevich, A. L., and Franzgrote, E., J . Geophys. Res., 1965, 70, 1311. Rubin, S., in Kolthoff, I. M., and Elving, P. J., Editors, “Treatise on Analytical Chemistry,” Interscience Publishers, a division of John Wiley & Sons Inc., New York and London, 1963, Volume 4, Part 1, p. 2075. Agency, Vienna, 1965, Volume 2, p. 15. Neutrons de 14 MeV,” Presses Universitaires de France, 1963. 2671-4, 1965. 1965. Energy Agency, Vienna, 1962, Volume 1, p. 217. 3342, 1962. 1965, Volume 2, p. 101. Pliesach, M., and Poole, D. O., in Guinn, J. P., Editor, op. cit., p. 206. Butler, J. R., “Radiochemical Methods of Analysis,” International Atomic Energy Agency, Greenwood, R.C., and Reed, J ., “Proceedings of the International Conference on Modern Trends Greenwood, R. C., U.S. Atomic Energy Commission Report, ARF 1193-26, 1963. Isenhour, T. L., and Morrison, G. H., Analyt. Chem., 1966, 38, 167. Nelligan, W. B., and Tittman, J., “Proceedings of the 6th Scintillation Counter Symposium,” Vienna, 1965, Volume 1, p. 391. in Activation Analysis,” Texas A and M University, 1961, p. 166. I.R.E. Trans. Nucl. Sci., 1958, p. 187.January, 19671 COLEMAN AND PIERCE 19 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 169. 168. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. Pierce, T. B., and Peck, P. F., in Shallis, P. W., Editor, “Proceedings of the SAC Conference Nottingham 1965,” W. Heffer and Sons Ltd., Cambridge, 1965, p. 159. Pierce, T. B., “Practical Aspects of Activation Analysis with Charged Particles,” European Atomic Energy Community Report EUR 2957 d.f.t. p. 53. Lightowlers, E. C., Analyt. Chew., 1963, 35, 1284. Petit, J ., and Engelmann, C., “Production and Use of Short-Lived Radioisotopes from Reactors,” Gebauhr, W,, and Martin, J., 2. analyt. Chevn., 1964, 200, 266. Cali, J. P., Editov, “Trace Analysis of Semi Conductor Materials,” Pergamon Press, Oxford, 1964. Lloyd, K. W., iuz Shallis, P. W., Editor, op. cit., p. 180. Bowen, H. J. M., and Gibbons, D., “Radioactivation Analysis,” Oxford University Press, 1963, Bowen, H. J. M., and Cawse, P. A., U.K. Atomic Energy Authority Report, AERE-R 4309, Harwell, Bowen, H. J. M., U . K . Atowzic Energy Authovity Report, AERE-R 4196, Harwell, 1963. Bowen, H. J. M., and Cawse, P. A., in Lenihan, J. M. A., and Thomson, S. J., Editovs, “Activation Analysis Principles and Applications,” Academic Press, London and New York, 1965, p. 149. Comar, D., Editor, “L‘analyse par Radioactivation et ses Applications aux Sciences Biologiques,” Presse Universitaires de France, Paris, 1964. Guinn, V. P., in Curry, A. S., Editor, “Methods of Forensic Science,” Interscience Publishers, a division of John Wiley & Sons Inc., New York and London, 1964, p. 47. Smith, H., and Lenihan, J. M. A., in Curry, A. S., Editor, 09. cit., p. 69. Smith, H., J. Forens. Sci. Soc., 1964, 4, 192. Ruch, R. R., Guinn, V. P., and Pinker, R. H., Nucl. Sci. Engng, 1964, 20, 381. Ruch, R. R., Buchanan, J . D., Pinker, R. H., Guinn, V. P., and Bellanca, S. C., J . Forens. Sci., Bate, L. C., Emery, J. F., Leddicotte, G. W., Lyon, W. S., and Pro, M. J., I n t . J . APPZ. Radiat. Coleman, R. F., J . Forens. Sci. Soc., 1966, 6, 19. Mapper, D., in Smales, A. A., and Wager, L. R., Editovs, “Methods in Geochemistry,” Interscience Publishers Inc., New York and London, 1960. Fleischer, M., Geochiw. Cosvnochiutz. Acta, 1965, 29, 1263. Ball, T. K., and Filby, R. H., Ibid., 1965, 29, 737. Smales, A. A., Hughes, T. C., Mapper, D., McInnes, C. A. J., and Webster, R. K., Ibid., 1964, Vogt, J. R., and Ehmann, W. D., Ibid., 1965, 29, 373. _ _ _ _ _ , Radiochivn. Acta, 1964, 4, 24. Caldkell, R. L., Mills, W. R., Allen, L. S., Bell, P. R., and Heath, R. L., Science, 1966, 152, 457. Taylor, D., “Neutron Irradiation and Activation Analysis,” G. Newnes, London, 1964, p. 119. Martin, T. C., Mathur, S. C., and Morgan, I. L., I n t . J . A p p l . Radiat. Isotopes, 1964, 15, 331. Ljunggren, I<., and Christell, R., Atomnpraxis, 1964, 10, 259. Norgolwalla, S., and Jervis, R. E., Trans. Autzev. Nucl. Soc., 1965, 8, 86. Ramdohr, H. F., Kerntechnik, 1963, 5, 204. Almsel, G., and Samuel, D., Physics Chew. Solids, 1962, 23, 1707. Fleischer, R. L., Price, P. B., and Walker, R. M., Science, 1965, 149, 383. International Atomic Energy Agency, Vienna, 1963, Volume 2, p. 29. p. 166. 1963. 1964, 9, 119. Isotofles, 1963, 14, 549. 28, 209. Received July 20th, 1966