OCTOBER, 1971 THE ANALYST Vol. 96, No. 1147 Liquid Scintillation Counting as an Analytical Tool .A Review* BY J. A. B. GIBSON AND A. E. LALLY (Health Physics and Medical Division, A .E.R.E., Harwell, Didcot, Berkshire) SUMMARY OF CONTENTS Introduction Basic concepts Instrumentation Solvents, scintillators and additives Techniques Sample preparation for homogeneous systems Sample preparation for heterogeneous systems Sample preparation for Cerenkov counting Standardisation Data handling Conclusions Appendix Isotopes measured in a liquid scintillation counter INTRODUCTION LIQUID scintillation counting is commonly used for counting a wide range of p- and cc-emitting radioisotopes in many chemical forms. The technique provides for a good detection efficiency (up to about 100 per cent,), and normally involves a minimum of chemical preparation.The relatively high backgrounds obtained in this method compared with other detection systems, e.g., proportional counters, limit the sensitivity for some isotopes, particularly a-emitters. A second, more serious, limitation is the variable efficiency caused by a reduction in the light output (quenching) in the presence of certain chemical impurities. Other impurities may introduce chemilurninescence, which gives an unknown and variable background. These effects are most important for low-energy 8-emitters such as 3H and 14C. Reduction of quenching and chemilurninescence have been widely discussed throughout the literature, and many methods of measuring the efficiency have been devised. There already exist several reviews dating from the proceedings of a symposium in 1958, edited by Bell and Hayes,l and culminating in a second symposium (edited by Branksome2) in 1970.The proceedings of the latter will provide much up-to-date information for the experienced user of the technique. This more limited review cannot compete with the depth and coverage of such a symposium. We therefore aim to provide a critical introduction to the methods and materials used in liquid scintillation counting. The review is divided into four general sections, the basic concepts of the technique, instrumentation, the choice of scintillators and additives, and the techniques used in sample preparation and standardisation and analysis of the subsequent data. The information is finally summarised as a table in the Appendix, which gives an indication of the range of isotopes that can be counted with a liquid scintillation counter.The techniques that will not be discussed include the use of liquid scintillators for neutron measurements and as anti-coincidence shields. These and other techniques are discussed in detail by B i r k ~ . ~ BASIC CONCEPTS The basic organic liquid scintillator consists of a solution of one or more fluorescent aromatic solutes dissolved in an aromatic solvent (usually toluene or dioxan). Other com- pounds (additives) can be added to this basic solution so as to incorporate the various radioactive samples into the scintillator. * Reprints of this paper will be available shortly. For details see summaries in advertisement pages.0 SAC and the authors.682 GIBSON AND LALLY: LIQUID SCINTILLATION COUNTING [Artalyst, Vol. 96 #I * Fast coincidence < 41 P 1 . \ Photomultiplier Liquid scint iI lator 'Photomultiplier J. Pulse amplifier f Single channel analyser $ Gate* 4 = * 1 Solute - - - electrons ----- ,Photocathode , I * Pulseadder 4 \ I . J I I J. Pulse amplifier Pulse amplifier Y * Single Single $. I Gate* Gate* channel analyser channel a na I yser electrons ___f \ Dynode \ \ \ j'%k Electrons $. Scaler- readout channel 1 4 $. Scaler- readout Scaler- readout channel 2 channel 3 i * Gates only open when pulses are observed from both photomultiplier tubes., Fig. 1. Schematic diagram of the scintillation process and electronics Alpha and beta radiations from radioisotopes in the scintillator deposit energy in the solvent.This energy is transferred through the solvent until it reaches a solute molecule which converts the energy to a light quantum (Fig. 1). Light quanta from several solute molecules are detected by the photomultiplier tubes (usually two), which convert about 25 per cent. of the light quanta into electrons. The over-all quantum efficiency from the original particle to the production of electrons is about 0.2 per cent., and the deposition of 6 keV of electron energy (approximate mean energy for 3H /3-particles) results in only one or two electrons being produced in the photocathode of the photomultiplier tube. Because of the statistical nature of the proce~s,~ some events will not be observed at all, and therefore the maximum theoretical counting efficiency is less than 100 per cent.5 (typically 60 per cent.for SH with two tubes in coincidence). The electrons are accelerated through a series of dynode stages that produce between two and six electrons for each electron input. This multiplication may amount to a 105-fold increase over eleven stages. Spontaneous emission from the photocathode results in a background noise output from the photomultiplier tube. By using two photomultipliers in coincidence, the single-noise event in one tube only can be rejected, but excessive noise will result in accidental coincidences that can be reduced by cooling to below ambient temperature. The external background of gamma and cosmic radiation will produce events in the scintillator that can be reducedOctober, 19711 AS AN ANALYTICAL TOOL 683 by using:an anti-coincidence shield.Added materials may be chemiluminescent and produce single photons, but this effect is largely eliminated by using the coincidence system. Chemical impurities may interfere with the transfer of energy from solvent to solute to produce “chemical quenching,” or they may absorb the light emitted from the solute molecules to produce “colour quenching.” Both effects will reduce the light output of the scintillator6 and thus reduce the counting efficiency for the sample. The two types of quenching have different effects on the output spectrum of the counter, and chemiluminescence will give yet a third change in the spectrum. It is these spectrum changes and shifts that form the basis of most techniques for the determination of counting efficiency and the detection of chemi- luminescence.Chemical quenching can be avoided with high-energy /I-emitters (/Imax. > 0.15 MeV) by the use of Cerenkov counting. The sample is counted directly without a scintillator and the light output is produced when the speed of the /I-particles exceeds the speed of light in the medium.’ This technique, although simpler, is still subject to colour quenching and chemi- luminescence. These techniques are discussed later. INSTRUMENTATION A large number of complete instrument systems are commercially available for use with liquid scintillators. They vary in complexity from single-sample instruments to systems handling hundreds of samples with an output of efficiencies, corrections and the disintegration rates for up to three isotopes.The choice depends upon the application and the variety of isotopes encountered in a particular laboratory. The basic components of a typical coincidence system are shown in Fig. 1. The photo- multiplier tubes are normally contained in a temperature-controlled box, which also contains the sample changer. The choice of the temperature is a compromise between the lower backgrounds obtainable at 0 “C and the miscibility of some samples at reduced temperatures. At present, the majority of new systems operate at temperatures sljghtlybelow ambient to ensure that the sample is homogeneous. The system may contain a number of independent channels (usually three), each with a separate amplifier and channel-width controls.This enables two or perhaps three isotopes to be counted simultaneously. Some systems incorporate an auto- matic external source of y-radioactivity. Following an initial sample count, the source is automatically transferred from a shielded container to a position close to the sample and a second count in the three channels is obtained. This second count with the external source enables an estimate to be made of any quenching present; details of the method are discussed later. The output data are normally printed on to a paper tape and, if necessary, can be punched on to tape for computer analysis. In some machines the computer is built into the system and the processed results are produced directly. SOLVENTS, SCINTILLATORS AND ADDITIVES The purpose of the solvent is (a) to provide a medium for containing the sample, (b) to transport the energy from the source of radiation and (c) to contain the solutes and allow for the emission of light.The scintillation solute transforms the energy into light, but if the wavelength is unsuitable for the photomultiplier tube then a secondary solute can be used as a wavelength shifter. The two basic solvents used are alkylbenzenes, such as toluene and xylene, and aliphqtic ethers, the most common being 1 + 4 dioxan - water. Although the dioxan base produces a scintillator with a lower scintillation efficiency than that of the alkylbenzene bases, it has the advantage of being miscible with water. All solvents must be of the highest purity to avoid quenching effects which can be caused by trace amounts of impurities.Typical primary solutes are 2,5-diphenyloxazole (PPO) and 2-(4‘-t-butylphenyl)-5-(4”-bi- phenyl)-lJ3,4-oxadiazole (Butyl-PBD) . The two common secondary solutes (wavelength shifters) are 1,4-bis- (5-phenyloxazol-2-yl) benzene (POPOP) and 1,4-bis- (4-methyl-5-phenyl- oxazol-2-y1)benzene (DM-POPOP). The solutes must be soluble at the operating temperature of the system and also must not be precipitated by the addition of the sample, e.g., water. Radioisotopes are most commonly prepared for counting in aqueous solutions, so that water miscibility is an essential requirement in many scintillators. Dioxan-based systems are therefore most suitable for this purpose, but naphthalene is usually added to increase684 GIBSON AND LALLY: LIQUID SCINTILLATION COUNTING [A%a&Si!, Vol.96 the energy transfer from the solvent to solute and hence to increase the counting efficiency. Many popular dioxan-based scintillators originate from the Brays solution , comprising naphthalene, PPO, POPOP, methanol and ethylene glycol in dioxan. Ethanol and mono- methyl and monoethyl ethers of ethylene glycol can be used to reduce the freezing-point of dioxan scintillat 0 ~ s . ~ $lo Toluene-based scintillators can be diluted with various polar solvents, such as ethanol and methanol, to increase water miscibility. This system was designed mainly for organic solvents, but now that solubilisers are freely available, a wide variety of sample materials, e.g., blood, urine and biological tissue, can be counted in toluene-based systems.Typical solubilisers are hyamine hydroxide, NCS (Nuclear Chicago) and the “Bio-Solv” range (Beck- man Instruments). Other additives such as Triton X-100 enable emulsion counting to be performed in toluene systems.ll Gel counting is a technique to be used when the sample is insoluble in the scintillator or in any convenient solvent. Normal dioxan-based scintillators can be converted into gels by stirring in finely divided silica to provide a thixotropic phase in which the insoluble material is suspended. The choice between buying ready-made scintillators and preparing one’s own depends upon the number and variety of samples to be a n a 1 y ~ e d . l ~ ~ ~ ~ Shelf-life is an important consideration. Dioxan-based systems should be kept under nitrogen to prevent oxygen absorption and in tightly stoppered vessels to prevent loss of solvent and the subsequent crystallision of naphthalene, both of which reduce the counting efficiency.Toluene-based scintillators are not affected by either of these phenomena. TECHNIQUES This section covers the factors involved in selecting the technique, sample preparation (homogeneous, heterogeneous or for Cerenkov counting) , standardisation and data handling. Initially, when a new isotope or material is to be counted, various factors must be considered: (a) the energies and types of radiation emitted by the sample, e.g., a-, B- or y-radiation; (b) the chemical form of the sample and whether chemical preparation could improve the effectiveness of the method; (c) the choice of the scintillator system and the possible inclusion of additives to obtain miscibility, reduce quenching and improve the counting efficiency; (d) instrumental conditions to ensure adequate sensitivity and an accurate assessment of the efficiency and background; and (e) the method of data analysis should be included in the selection of the technique.After choosing the technique, it is finally necessary to ensure that conditions remain constant by regular checks with calibrated standards. SAMPLE PREPARATION FOR HOMOGENEOUS SYSTEMS- Sample preparation should normally be kept to a minimum so as to avoid losses from incomplete chemical recovery and isotope effects caused by the different properties of the natural and radioactive isotopes. Chemical impurities introduced by processing may also introduce unknown quenching effects.Aqueous systems can be added directly to dioxan-based scintillators and to toluene- based scintillators when a solubiliser is added.14J5J6 Biological materials such as blood, urine, salts, sugars and the alkali digests of plasma and tissue samples can be incorporated into toluene-based scintillators with Bio-Solv (BBSB and BBS3, Beckman Instruments Ltd.). BBS2 is an acid solubiliser for alkaline tissue digests and aqueous solutions, and it must be neutralised before counting. BBS3 is a general-purpose solubiliser for all types of aqueous samples and for blood and plasma.16 NCS (Nuclear Chicago Corporation) and Hyamine 1OX can also be used for tissues and purified biological material.17 However, the quaternary ammonium base of hyamine hydroxide is strongly chemi1uminescentl8 and should be used with caution.It is necessary always to use some met hod to check for chemical and colour quenching and chemiluminescence. Materials that are strongly luminescent or produce severe quenching require further treatment. Colour quenching can often be eliminated, or at least reduced, by digestion withOctober, 19711 AS AN ANALYTICAL TOOL 685 hydrogen peroxide and perchloric acid. This method has been used successfully for soft tissues, solid biological materialsfQ (e.g., teeth and bones) and for the determination of radio- activity on filter-papers.20 Isotopes investigated include S2P, S5SJ 45Ca, 55Fe and 57Co. Alternatively for 3H and 14C, complete combustion of the sample to water or to a soluble carbonate produces a simple counting method.The Schoniger oxygen-flask method21 was the forerunner of this technique and many modifications have been made.22 An automatic version of the instrument is now available commercially as a “Tritium Oxidiser” (Packard Instruments Ltd.). Recovery experiments to investigate the chemical yield and any isotope effect are an essential part of any combustion experiment. SAMPLE PREPARATION FOR HETEROGENEOUS SYSTEMS- Insoluble materials and other samples that cannot be processed chemically can be measured as suspensions in gels or as emulsions. If necessary, a solid support such as a filter-paper can be used. A reduction in counting efficiency may occur through self-absorption in particles, supports, etc., and standardisation may be difficult.Suspension counting in gels is useful for incorporating reasonable amounts of precipitates that are otherwise insoluble in liquid scintillators. A transparent gelling agent such as Cab-0-Sil, a finely divided silica powder, is mixed to give a thixotropic gel that is fluid when shaken but firm when at rest. This technique has been used for barium and strontium carbonates, perchlorates, etc., and for an iron ferriphosphate complex for the determination of 55Fe and 59Fe (Eakins and B~owII~~) and 239Pu and 241Pu (Eakins and L a l l ~ ~ ~ ) . Emulsion counting can be used for incorporating large volumes of aqueous samples into scintillators. An example of this technique is the use of Triton X-100 detergent with a toluene-based scintillator.12 This technique is very sensitive to pH, temperature and salt concentrations.SAMPLE PREPARATION FOR CERENKOV COUNTING- In the simplest form no sample preparation is required and the solution is placed directly into a counting phial. The efficiency is improved by effecting an increase in the refractive index (to reduce the energy threshold) and this may be necessary for p-emitters of lower energy. The introduction of a wavelength shifter will further increase the efficiency.’ Chemical quenching is eliminated in this method, but colour quenching and chemiluminescence are still important. Quenching can be reduced by using the decolorising techniques discussed above. Standardisation is carried out either by adding an internal standard or by using a high-energy y-emitter to produce photoelectrons in the solution.The major advantage of this technique is the high sensitivity with large volumes, or for flow monitoring without changing the liquid passing through the detector. STANDARDISATION- The counting efficiency for a high-energy /3-emitter can be nearly 100 per cent., and the effects of quenching are then small. If all samples in a particular experiment have the same composition, then a simple standardisation technique is all that is necessary. However, for low-energy /3-emitters in a wide range of materials, either chemical or colour quenching, or both, will normally be present. The three methods most commonly used are the use of an added internal standard, an automatic external standard25 (usually a long-lived y-ray emitter), and a channels ratio method.26 The advantages and disadvantages are summarised in Table I.The normal procedure is to choose the method, e.g., channels ratio, in which the efficiency is plotted against the ratio for a series of standards with different amounts of a quenching material. The ratio obtained during an experiment can then be converted into an efficiency by the use of this graph. Such a calibration curve is necessary for each isotope, scintillator and instrument setting. A similar technique is used with the external standard method, but the internal standard will give the efficiency directly for each sample. The background of the counter is also affected by quenching, and for low-level counting it is necessary to know the background for various values of the channels ratio.The presence of chemiluminescence introduces an increased variable background and can lead to erroneous results if it is significant compared with the radioactivity of the sample.686 GIBSON AND LALLY: LIQUID SCINTILLATION COUNTING [Analyst, Vol. 96 DATA HANDLING- The output from manual instruments can normally be analysed with a desk calculator. Automatic systems that process hundreds of samples per day can produce five or more items of information per sample, and manual processing becomes tedious. The simplest technique normally involves the use of a small desk-top computer,27 which can be programmed to take information about the sample (perhaps counts in three channels), the external standard (three more counts) and the time of the measurement, and produce the mean disintegrationrate for that sample.The operator of the computer will have prepared a suitable polynomial fit for the efficiency versus ratio curve, and the data can be transferred by hand or by paper tape. The use of more sophisticated computers may be necessary for larger outputs or for a wide variety of samples.28 If each sample is different in radioisotope, scintillator or type of quenching, then manual methods are usually the most efficient. The use of both the external standard and the channels ratio methods for each sample will normally reveal the type of quenching or the presence of chemiluminescence, and it is good practice to compare the efficiencies determined by these methods by using statistical tests.TABLE I METHODS OF DETERMINING COUNTING EFFICIENCY Method Internal standard Automatic external standard Channels ratio Advantages Gives individual results Best method for highly quenched samples Only reliable method when solid support Corrects for both colour and chemical material is present quenching Automatic with no handling problems Only short repeat count needed Composition of sample unchanged Only one count needed No handling of the sample required Composition of the sample unchanged Independent of sample volume Independent of inhomogeneity in the sample { CONCLUSIONS Disadvantages Possible errors when measuring small amounts of standard solution with a pipette Second count needed, i.e., time consuming with large numbers of samples Sample cannot be re-counted Not suitable for dual-label samples Dependent upon sample volume and the Sample must be homogeneous Poor accuracy with highly quenched Instrumental costs and maintenance accurate positioning of the source samples Long counting time needed for accuracy with low-activity samples Poor accuracy with highly quenched samples Needs a t least two channels Liquid scintillation counting is the accepted technique for many radioisotopes in a wide range of chemical forms.It lacks sensitivity for very low levels of a-activity and cannot compete with internal gas counters used for natural tritium levels. This leaves a wide field of analytical application in chemistry, biochemistry and medicine. The presence of quenching agents in most samples can normally be detected, and their effect either reduced or corrected for by the choice of suitable standardisation methods.Similarly, chemiluminescence can be detected and reduced by using a different technique. Improvements in the future may come from increased photocathode efficiency and from improved chemical techniques, with the best use of scintillators and solubilisers to reduce unwanted effects. Use of the technique is essentially a practical problem, which presents new facets with each new type of sample. Appendix ISOTOPES MEASURED IN A LIQUID SCINTILLATION COUNTER This appendix is intended as a preliminary guide to the versatility of the technique and gives some idea of the sensitivity of the method. The information in Table I1 includes a range of isotopes and the matrix from which they were extracted.The preparation tech- nique can be obtained from the references in the final column, but brief details of the additives are given, together with the scintillator used. The efficiency of counting is given only as aOctober, 19711 AS AN ANALYTICAL TOOL 687 guide, but it can be used to give an approximate indication of sensitivity if the background is taken as typically 10 to 25 counts per minute for most systems. The reference list is fairly limited considering the vast literature on this subject, and represents methods, techniques and theoretical information that we have found to be useful in the theory and application of liquid scintillation counting. TABLE I1 ISOTOPES MEASURED IN A LIQUID SCINTILLATION COUNTER Isotope Matrix 3H 3H 3H 3H 3H 14c 14c 3 S P 3% 35s 36~1 4sCa ssFe S°Fe 63Ni OOSr O O Y 1311 147Pm aloPb pu (4 34lPu 33spu 1.2. 3. 4. 5. 6. 7. 8. 9. Water Water Water Blood, plasma, Plasma, urine urine 14C-toluene l’C-fructose Organic Various compounds Vegetation Various Various Various Counting form Water Water emulsion Water emulsion Blood, plasma, urine Plasma, urine, emulsion 14C-toluene 14C-fructose Aqueous solutions Aqueous solutions BaSO, precipi- tate in a gel H2S04 on glass- fibre disc Aqueous solution of NaCl CaCl, in dibutyl Blood phosphate Ferrjphosphate complex in gel Aqueous solution Tetrapyridine- nickel dithio- cyanate Various 2-Eth ylhexanoic acid solution of SrCO, Plasma Plasma in gel Urine Di-2-ethylhexyl phosphate complex Aqueous solutions Aqueous concen- trate Urine, faeces, Ferriphosphate blood complex in gel Bone, liver, Acidic solution spleen, urine after in-vial oxidation N.S.Not stated. Scintillator PPO - p-bis(o-methyl- styryl) benzene in dioxan Triton N-101 in p-xylene Triton X-100 in toluene BBS3 solubiliser in toluene Triton X-100 Hyamine 1OX in toluene BBSl solubiliser in Triton X-100 in toluene toluene None +. wavelength shifter Dioxan 3 g I-1 of p-terphenyl in Dioxan toluene Toluene Dioxan Dioxan Toluene Toluene Toluene Dioxan Dioxan 0.4 g 1-1 of PPO in Toluene ethanol Counter tempera- ture/OC 0 to 25 17 to 25 4 2 0 N.S. 0 N.S. N.S. N.S. 12 N.S. -5 4 0 - 4 -2 N.S. N.S. 4 N.S. Efficiency, per cent. 23 24 27 37 30 20 88 75 25 50 65 78 85 85 19.4 33.4 65 83 95 85 95 97 86 21 85 Reference 14 15 11 16 29 16 11 7 7 30 31 32 33 23 34 35 36 37 38 24 39 REFERENCES Bell, C.G., and Hayes, F. N., Editors, “Liquid Scintillation Counting,” Pergamon Press, Oxford Branksome, E. D., Editor, “The Current Status of Liquid Scintillation Counting, ” Grune and Birks, J. B., “The Theory and Practice of Scintillation Counting,” Pergamon Press, Oxford and Gale, H. J., and Gibson, J. A. B., J . Scient. Instrum., 1966, 43, 224. Gibson, J. A. B., and Gale, H. J., J . Phys. E., Series 2, 1968, 1, 99. -- , IM. J . A$$. Radiat. Isotopes, 1967, 18, 681. Elriik, R. H., and Parker, R. P., Ibid.. 1968, 19, 263. Bray, G. A., Analyt. Biochem., 1960, 1, 279. Lerch, P., and Cosandey, M., Adv. Tracer Meth., 1966, 3, 107. and New York, 1958. Stratton, New York and London, 1970.New York, 1964.688 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. 35. 36. 37. 38. 39. GIBSON AND LALLY Polesky, H. F., and Seligson, D., Analyt. Biochem., 1965, 10, 347. Turner, J. C., Int. J . Appl. Rudiat. Isotopes, 1969, 20, 499. White, D. R., Ibid., 1968, 19, 49. Burke, C. W., Humphrey, K., and Beardwell, C. G., Ibid., 1968, 19, 666. Moghissi, A. A., Kelley, H. L., Regnier, J. E., and Carter, M. W., Ibid., 1969, 20, 145. Lieberman, R., and Moghissi, A. A., Ibid., 1970, 21, 319. “Beckman International Newsletter,” 1968, SH-68-1. Hansen, D. L., and Bush, E. T., Analyt. Biochem., 1967, 18, 320. Horrocks, D. L., Int. J . Appl. Radiat. Isotopes, 1968, 19, 859. Herberg, R. J., Analyt. Chem., 1960, 32, 42. Mahin, D. T., and Lofberg, R. T., in Branksome, E. D., Editor, op. cit., p. 212. Schoniger, W., Mikrochim. Acta, 1955, 123. Oliverio, V. T., Denham, C., and Davidson, J. D., Analyt. Biochem., 1962, 4, 188. Eakins, J. D., and Brown, D. A., Int. J . A$@. Radiat. Isotopes, 1966, 17, 391. Eakins, J. D., and Lally, A. E., Rep. U.K. Atom. Energy Autlz., AERE-R 6640, H.M. Stationery Takahashi, I. T., and Blanchard, F. A., Analyt. Biochem., 1970, 35, 411. Glass, D. S., Int. J . Aflfll. Radiat. Isotopes, 1970, 21, 631. Williams, M. A., Cope, G. H., Jackson, J. L., and Hill, P., Biochem. J., 1970, 118, 379. Figdor, S. K., Comp. Biomed. Res., 1970, 3, 201. Whyman, A. E., Int. J . Appl. Radiat. Isotopes, 1970, 21, 81. Willis, C. P., Olson, D. G., and Sill, C. W., Analyt. Chem., 1970, 42, 124. Lloyd, R. A., and Rees-Evans, D. B., Int. J . Appl. Radiat. Isotopes, 1965, 16, 393. Moghissi, A. A., in Branksome, E. D., Editor, op. cit., p. 86. Hardcastle, J. E., Hannapel, R. J., and Fuller, W. H., Int. J . A@pl. Radiat. Isotopes, 1967, 18, 193. Harvey, B. R., and Sutton, G. A., Ibid., 1970, 21, 519. Uyesugi, G. S., and Greenberg, A. E., Ibid., 1965, 16, 581. Bell, T. K., J . Clin. Path., 1967, 20, 629. Ludwick, J. D., Analyt. Chem., 1964, 36, 1104. Fairman, W. D., and Sedlet, J., Ibid., 1968, 40, 2004. Lindenbaum, A., and Lund, C. J., Radiat. Res., 1969, 37, 131. Office, London, 1970. Received April Sth, 1971 Accepted June 25th, 1971