44 GENERAL DISCUSSION GENERAL DISCUSSION Prof. F. S. Dainton (Leeds University) said: In the features of the primary act which are discussed in the first three papers very little has been mentioned con- cerning the possibility of light emission by liquid and gaseous systems subjected to ionizing radiation apart from the fact that radiative recombination and radia- tive electron capture processes are intrinsically unlikely. An exhaustive experi- mental study of this subject would be worthwhile for two reasons. Firstly because a close study of the spectrum of the light might enable the emitter to be identified and thus provide unambiguous evidence for the formation of excited species and for the recognition of their chemical nature. Clearly with the limited source strengths usually available such experiments are much more likely to be successful in gases than in liquids, and some years ago Dr.Burcham and I made observations on the spectra of the luminescent 600 keV proton beams in air and dry nitrogen, which served to identify excited nitrogen molecules in both gases and to show that nitric oxide was rapidly formed in the former. We also attempted to carry out a similar experiment with water vapour, which was for various technical reasonsGENERAL DISCUSSION 45 unsuccessful, but which we hope to repeat with a better reaction cell. In 1949, Dr. Collinson and I attempted to take the spectrum of the feeble luminescence generated when a 400 mc radium source was immersed in water contained in a spherical quartz flask. This luminescence was just visible to the dark adapted eye, but on further examination proved to be merely the Cerenkov radiation.Repeti- tion of the experiment with an aqueous ferrous sulphate solution in place of the water demonstrated the capacity of the system for self absorption of that part of the Cerenkov spectrum lying at wavelengths less than 2900A. Since the overall chemical change induced in ferrous sulphate solutions by absorption of light of these wavelengths may be represented by Fe2f f H20 --f Fe3+ i- OH- + 4H2 which is superficially the same as the radiation induced reaction, this experiment raised in our minds the question as to whether any significant fraction of the yield of radiation reactions is due to secondary photochemical processes. This is the second reason for studying the light emission.For the vast majority, the photo- chemical fraction of the yield stimulated by the Cerenkov radiation is zero or very small. Recently, however, it has been suggested by Richards that in a system in which Cerenkov radiation is non-existent, namely Po c( particles an ultra-violet emission is produced which is strongly self absorbed and in so doing induces a photochemical change of similar nature to the radiochemical change. Since the existing publications by Dr. Richards are brief can he tell us whether he has any evidence which indicates (i) what proportion of the total chemical change is due to this effect, (3) what are the magnitudes of the quanta emitted? Dr. E. W. T. Richards (Glasgow University and A.E.R.E. Harwell) said: In any satisfactory theory of the effects of alpha particles irradiation of solutions it is necessary to postulate some mechanism by which the energy dissipated by the particle is spread throughout the bulk of the liquid.This is especially the case where experiments have been carried out using dilute solutions. From the evidence of physical experiments on the ionization produced in liquids by alpha-particles, it would appear that 99 % of the ions recombine in a time less than 10-7 see. It is therefore necessary to consider some process other than simple ionization to explain the chemical changes which occur. Various theories which make use of free radicals have been put forward, but the method of production and their spacial distribution do not appear to have been fully elucidated. If the radicals are produced directly from the original ions, then it is difficult to see why they should have a longer life time than that given for the ions in Jaffe’s original column theory, in which the electrostatic forces between the ions were not considered.If this is the case, then chemical effects which have a time constant long compared with this cannot be explained by these theories. It has been shown that when a solid is bombarded with alpha-particles quanta are emitted.1 These quanta are strongly absorbed by the parent material and therefore the percentage number which escape is small. If such an effect was present in liquids it would be a mechanism whereby radicals could be produced having a lesser concentration than that of the original ionization.It should be noted that whilst the volume over which the radicals would be produced is large compared with that of the ion columns it would still be very small compared with the bulk of the liquid. Experiments have been performed to test the validity of this theory and positive results obtained.2 Some evidence of the presence of the quanta have also been obtained using purely physical methods and it would appear that the wavelength involved is between 1500-17OOA. The quanta are strongly absorbed by water vapour but transmitted to some extent by crystalline quartz. However, the possibility that the radiations concerned are K or L X-rays cannot be entirely ruled out. 1 Richards and Cole, Nature, 1951, 167, 286. 2 Dee and Richards, Nature, 1951, 168, 736.46 GENERAL DISCUSSION Dr.N. Miller (Edinburgh University) said: The experiments reported by Dee and Richards 3 were of considerable interest to us in Edinburgh in view of the fact that we had spent much time in the study of the chemical action of alpha- particles on ferrous sulphate solutions. We set out at once to confirm their obser- vations, but unfortunately found not only that we could not repeat their chemical experiments, but that sensitive physical experiments, designed to detect directly any ernission of ultra-violet light from water under alpha-bombardment, also gave negative results. In our attempts to repeat their chemical experiments, we used an apparatus resembling that reported in their publication as closely as possible, with the excep- tion that our 20 mc polonium source was covered with a mica sheet of about 1.5 mg/cm2 mass thickness to avoid the health hazards associated with the use of a bare polonium source of this strength. The method of covering the source will be explained in detail in another publication from this laboratory.The quartz plates used were (i) a fused quartz microscope cover slide, 1/2 mm in thickness, which showed over 90 % transmission down to 2000& (ii) slides made by cutting and polishing crystalline quartz, which showed similar excellent properties of near ultra-violet transmission. Maintaining a thin film of water on these slides was found to be exceedingly difficult: in fact we are somewhat sceptical as to whether Dee and Richards themselves satisfactorily solved this problem for the long period (11 days) of their irradiations.It proved a relatively simple matter, however, to maintain, at least for 24-hr. periods, uniform films of dilute sulphuric acid which could be shown by weighing to have a mass thickness less than the range of the alpha-particles. Carrying out many such irradiations for 24-hr periods, and using analytical techniques (spectrophotometry, cf. our paper above) much more sensitive than those employed by Dee and Richards, we were unable to detect any significant oxidation of ferrous ions. From the results of our work on the direct oxidation of such solutions by alpha-particles, we were able to say that any oxidation of this type must have been less than 1/300th of that observed under direct alpha-bombardment.(Reference is made here to conditions under which as many alpha-particles entered an “ infinitely deep ” solution as entered the thin layer of acidified water, the geometrical factor in the latter case being un- corrected for.) In a direct search for any such ultra-violet light emission, Mr. L. 0. Brown of our laboratory next carried out the following physical experiments : (i) The apparatus used in the first type of physical experiment resembled closely that employed in the chemical experiments just described, save that underneath the quartz, in place of the solution, was a thin coating of Apiezon grease, followed in succession by a Perspex light guide and the photocathode of an E.M.I. Type VX 5013 photomultiplier. Much weaker sources (ca.100 pc) were adequate for these physical experiments. These sources were not mica-covered, and the alpha- particles from them were collimated by placing the Pt discs on which the polonium was plated at the ends of brass tubes, 2 mm diam. and 10 mm long. All of the operations were carried out within a light-tight box, which was kept inside a dark room and only opened under red light, to avoid “ memory ” effects in the tube. The purpose of the Apiezon grease was to convert by fluorescence any ultra- violet light emitted to light of a wavelength which could pass through the envelope of the photomtuliplier. Bolton and Williams 4 have recently shown, using a similar type of photomultiplier, that light down to 972A can readily be detected in this way. The experiments were started with a film of water, thick enough to stop the alpha-particles completely, on the top of the quartz, and counting was carried out continuously as the water evaporated.After careful cleaning of the quartz, it could be shown’that such a water-film evaporated evenly and went through 3 Dee and Richards, Nature, 1951, 168, 736. 4 Bolton and Williams, Natirrc, 1952, 169, 316.GENERAL DISCUSSION 47 a phase in which it showed interference colours. There could, therefore, be no doubt that for some time during the evaporation the film was thin enough to allow the alpha-particles to penetrate. At the start of the experiments, with the thick water-film, a little light was detected, but it was established in a separate experi- ment that this was due to the passage of the alpha-particles through the air above the slide, a well-known phenomenon.5 At the end of the experiments, when the alpha-particles were being stopped in the quartz, about 5 times as much light was emitted, the increase being due to scintillations produced in the quartz itself.The interesting part of the experiment, of course, was the investigation of the intermediate region, when the water-film was thin enough to allow the alpha- particles to penetrate. If Dee and Richards had been right, the counting-rate should have gone through a very large maximum in this region. We observed no such behaviour. As the water evaporated, the counting rate merely rose smoothly to the value for quartz alone. (ii) In the second type of experiment the alpha-particles encountered only liquid water and water vapour in the whole of their paths.Here the polonium source, with its brass collimating tube, was inset into the bottom of a Perspex light guide. The light guide was suspended above the water, and the base of it, which was covered with a thin layer of Apiezon grease, was separated from the water surface by a gap varying between 1 and 2.5 cm, depending on the water level. The photomultiplier, suspended upside down above the system, looked into it from the upper end of the light guide, i.e. from behind the source. The whole of the apparatus could be evacuated before the experiment began, and de-aerated water introduced from an adjacent vessel. Some experimental difficulties had to be overcome here, which will be described in detail in a later publication, but no light emission was detected which could be ascribed to the water or water vapour whatever the distance between the bottom of the light guide and the water surface.The very small light emission observed was noted also when the end of the alpha- particle collimating tube was stopped up so that no alpha-particles escaped, and was probably to be attributed to scintillations caused within the Perspex light guide by y-rays from the polonium. Rough calculations based on a knowledge of the strength of the source and of the intensity of the y-emission 6 from Po210 showed that the counting-rate was of the right order of magnitude to be accounted for in this way. These experiments have now convinced us (i) that the chemical effects observed by Dee and Richards, whatever their cause, could not have been due to U.V.light emission from the water-film, and (ii) that no appreciable emission of light of wave- length greater than 1800 A, the upper absorption limit of water, takes place when water is bombarded with alpha-particles. With regard to possible light emission below 1800 A, one cannot, of course, make such categorical statements, as present knowledge of the transmission of liquid water and water vapour in the vacuum ultra-violet is very limited. We estimate, however, that any appreciable emission of such light from the alpha-particle tracks would have been detected in the second type of experiment described above unless the linear absorption coefficient of liquid water for the light were greater than 105 cm-1, a very high figure.If the absorption coefficient were indeed appreciably greater than this, the quanta would be absorbed so close to the ion column in liquid water as barely to affect the radical density. Quite apart from these considerations, Mr. Wilkinson and I find that we can explain the results of our studies on the oxidation of ferrous sulphate solutions by alpha-particles adequately without postulating any such effects. This evidence mill be dealt with in the publication which Mr. Wilkinson and I are now preparing on our chemical work with alpha-particles. 5 Greinacher, 2. Physik, 1928, 47, 344 ; Kara-Michailova, Sitzber. Akad. Win. Wien (HA), 1934,143,15 ; Audubert and Lormeau, Compt. rend., 1949,228,318 ; Lormeau, Compt. rend., 1950, 230, 956.6 Grace, Allen, West arid Halban, Pmr. P/zysic. Sor. A , 1951, 64, 493.48 GENERAL DISCUSSION In conclusion I would like to point out that in spite of the criticism to which their views are susceptible, Professor Dee and Dr. Richards have rendered a valu- able service to radiation chemistry by attracting attention to the possibility of effects such as they postulate. EfTects of this kind may quite possibly be important in other non-aqueous systems, Dr. M. Magat (Paris) said: Prof. Audubert asked me to report a few experi- ments performed by two of his co-workers, Miss Lautout and Dr. BUSSO, which concern the findings of Prof. Dee, Dr. Richards and Dr. Cole.73 I n the experiments of Lautout and Busso photocounters, sensitive in the range of 2000- 3200 8, were used ; these recorded about 1 kick per 105 photons, which represents a sensitivity about 107 times larger than that of Prof.Dee and Dr. Richards. Using a polonium source of 23 mc, they observed a weak light emission of a quartz plate bombarded by cc particles which was about of the same order of magnitude as the light emission of air, previously detected and investigated by Audubert and Mrs. Lousteau.9 Tn this last case the quantum yield was of the order of 1 light quantum/ 107 ion pairs, that is about lo7 times smaller than the effect reported by Dee and his group, and much smaller also than the emission produced by irradiating quartz with X- and y-rays. The emission from the quartz plate could be separated from that of air, by working either in vacuo or by covering the quartz plate with thin layers of A1 or Au.Tn this last case, according to Richards and Cole, the light intensity ought to increase due to the emission of the metal film or at least remain constant. The experimental results obtained were absolutely to the contrary- the light emission decreased in the classical way when the thickness of the absorbing layer increased, i.e. in the ratio of energy dissipated in quartz, when compared to that obtained in vacuum. No method could be found for spreading stable, 40p thick layers of water on a clean quartz surface, such as were used by Dee and his group. Instead the light emission of a thick layer of water when bombarded by CI particles was investi- gated. Should an emission, representing about 10-6 of that observed by Dee and his co-workers exist in the range of 2000-3000 8, (as it should to explain the observa- tions on chloroacetic acid) it would have been easily detected, particularly since water does not absorb light in this region.The result was, however, entirely negative, the count being identical with the background. The light emission of liquid water irradiated by cc rays is at least 10 times weaker than that of air. Dr. J. Weiss (Dzuhanz University, Newcastle) said: T think that even jf one assumes a very rapid recombination of the primarily formed ions, as suggested by Dr. Richards, it is extremely unlikely that in solutions this will lead to an appreci- able light emission, the reason being that the life time of the excited state for emission in the ultra-violet is of the order of 10-8 to 10-9 sec, and it is therefore rather more likely that, in general, the excitation energy is dissipated by the per- turbing influence of the surrounding molecules, leading to loss of energy in some other form, e.g.by dissociation or predissociation processes. I should be very glad to have Professor Massey’s comments on this point. Dr. E. W. T. Richards (Glusgow University and A.E.R.E., Hurwell) said: It i s generally accepted that the ionization produced in a liquid by the passage of an alpha-particle may be expressed by the “ column theory ” of ionization proposed by Jaffe. Kramers has pointed out that in Jaffe’s original theory no account is taken of the electrostatic forces which must exist in such columns and that under normal conditions these forces have a greater effect than the normal diffusion processes.However, according to both theories there should be a difference in the field strength against ion current relationship depending on the direction of 7 Dee, Richards and Cole, Nature, 1951, 167, 286. 8 Dee, Richards and Cole, Nature, 1951, 168, 736. 9 Audubert and Lousteau, Compt. rend., 1950, 230, 956: Conipt. rend., 1950, 230, 1771 ; 1952,234, 330.GENERAL DISCUSSION 49 the applied field relative to tracks of the alpha-particles. Experiments were designed to test these theories. The ion current which could be collected from a liquid when a field was applied at right-angles to the paths of the alpha-particles was measured and compared with the case when the field was applied parallel to the particle tracks.Within experimental error there was no difference in the two cases, both with respect to the magnitude of the current and to the variation of the current with the field strength. Experiments were then undertaken to investigate the variation in the percentage ion current which could be collected as a function of the original ion density. This was carried out by inserting absorbers between the alpha-source and the liquid under irradiation. The relationship obtained was entirely different than that which is obtained from a gas. However, the results were in agreement with the current which would be theoretically predicted if all the current were due to the delta rays which had an energy of greater than 1 kV. It would appear that there is little or no contribution from the primary ion columns to the ion current which can be collected from liquids.Dr. E. Collinson (Leeds University) said : The Halpern-Hall theory predicts a decrease of stopping power in the condensed state, yet the stopping power of liquid water is about 1 3 % higher than that of water vapour. Are there any reasons in the light of our present knowledge why this should be so? Bearing in mind this anomalous behaviour is it not particularly dangerous to assume that W will have the same value of 28 eV for both the liquid and the vapour? Applying the Bethe formula to calculate numerical values of rates of energy loss, etc., Lea 10 rejects the value of 45 eV given by Mano for the average excitation poten- tial of water, and obtains a value of 69 eV which depends on the validity of the Bragg additive law. In fact it now seems certain that liquid water provides an important exception to this law 11 and the value of 45 eV, though based on an incorrect derivation, apparently gives calculated values nearer to those found experimentally. It would be of interest, therefore, to know how Cormack and Johns derived their figures of E = 80 eV. Prof. F. W. Spiers (Leeds University) said: The value of W = 28 eV for electrons in water has been derived, as Dr. Collinson remarks, from experiments in water vapour. This procedure, although unsatisfactory, would appear to be better than the simple assumption that the value of Wfor water is the same as that for air. Studies of effects depending on ion density are not yet sufficiently critical to distinguish between postulated values of W. Changes in W cannot be inferred from changes in stopping power on condensation, since the proportions of energy lost in ionization and excitation are not necessarily the same in the liquid and vapou r phases. The value of E = 80 eV for water used by Cormack and Johns is that given by Halpern and Hall and is derived by these authors from the generalized geometric mean of the atomic frequencies of their mutliple-frequency model. Its derivation is similar, therefore, to that used by Lea applying the Bragg additive law to a simple molecular model. 10 Lea, Actions of Radiations on Living Cells (Cambridge University Press, 1946). 11 Appleyard, Puoc. Camb. Phil. SOC., 1950, 47, 443.