72 IRRADIATED PURE WATER CHEMICAL PHENOMENA IN IRRADIATED PURE WATER BY P. BONET-MAURY Institut du Radium, Facult6 des Sciences de Paris, Laboratoire Curie, 11 Rue Pierre Curie, Paris 5e, France Received 12th March, 1952 Densely ionizing radiations (natural a-particles) decompose water, following a simple mechanism in which secondary reactions interfere only to a small extent. The action of sparsely ionizing radiation on water, for example X-rays, leads to a very small radiolysis. Dynamic equilibria between the formation and decomposition of products are observed and establish an equilibrium concentration, independent of dose, within certain limits. The heavy accelerated particles behave as if they contained an X-ray and an a-particle component; the radicals of the X-ray component can react with the H202 or H2 from the a-component and this picture is perhaps applicable to all ionizing particles.For a linear ion density of the order of 200 ion pairslp (corresponding to the equidistance of the radicals H and OH) the yield per particle passes through a maximum and this critical density appears to mark the transition of the r-particle chemistry to the X-ray chemistry. The radiochemical reactions depend without doubt on the initial distribution of ion- ization in the water, no matter what the ionizing particle is, and this fact plays a major part in the chemistry of radiations. The numerous biological effects depend in the same manner on the ion density and the different biological efficacy of certain radiations can be explained by the difference in the primary chemical effects.The irradiation of a medium containing a high percentage of water (aqueous solutions, living organisms) primarily brings about chemical processes in the aqueous phase. The great importance of a satisfactorory knowledge of the phenomena for radiation chemistry and for radiobiology is thus apparent. Ir- radiated pure water * constitutes a particularly simple radiochemical system both * We mean by the term pure water, water distilled at least twice (preferably in a quartz apparatus) and deoxygenated by degassing under vacuum. Aerated water is a solution of oxygen and constitutes one of the simplest radiochemical system in aqueous solution.P . BONET-MAURY 73 cxperimentally and in the interpretation of the phenomena.The radiolysis produces only three molecular species, Hz02, H2 and 0 2 , and only the first two arise in the primary effect, oxygen appearing only slowly due no doubt to the secondary decomposition of H202.1 Experiment shows that the formation of oxygenated water and of hydrogen follows very different laws when the water is irradiated with densely ionizing rays as, for example, K-rays from radium, or with sparsely ionizing rays, for example, when X- or y-rays are used.14 5 Before examining the experimental results it is convenient to define the term radio- chemical yield. THE RADIOCHEMICAL YIELD.-The formation of H202 or of hydrogen is character- ized by its radiochemical yield or the number of molecules produced by the absorption in the water of a certain quantity of energy, q. Taking q as 100 eV the yield is defined as G ; if q corresponds to the energy of formation of an ion pair the ionic yield M/lV is obtained.Now the energy of ionization in air varies 2 from 31 eV to 36 eV depend- ing on the nature of the radiation and, for water, the yield is not accurately known, but the differences in values obtained nevertheless do not exceed 10 %. We there- fore propose a mean value of q = 33.3 eV, which is convenient because M/N =- G/3. Also, the yield per particle has been used recently for a-rays, deuterons and protons.3 In the publications on radiolysis of water the term “yield” has been given a variety of meanings and this has resulted in a certain confusion in the comparision of numerical results. The ionic yield M/N can nave, following various authors, five distinct meanings.This is the number of water molecules decomposed per ion pair; the calculation supposes a knowledge of the primary reaction in the radiolysis and the value of the yield depends on the choice of this reaction. (ii) YIELD IN TERMS OF FORMATION.-((I) Of oxygenated water or (M/N)H,O,.---The formation of hydrogen peroxide appears to depend in certain cases on the concentration of hydrogen in the irradiated water ; 3 this concentration, determined by Henry’s law, varies with the hydrogen pressure in the gas phase above the water. The yields of H202 obtained by diRerent authors are only comparable for identical hydrogen concentrations and these depend on the conditions of the experiments, e.g. thickness of the water layer, agitation (mechanical, or by gas), irradiation in a closed vessel or one open to the air, under vacuum or under pressure of hydrogen, relative volume of the liquid and gas phases, etc.(b) Of total gas or (hf/N) H,+O,.-Certain authors have measured the total volume of the gas formed without analyzing it. This yield, like those following, depends on the pressure of the gas phase above the water and a description of the experimental arrange- ment is essential. (c) Of hydrogen or (h.f/Jv)H2.--It appears that at the beginning of the radjolysis, the yield is equal to that in the former case (b), the oxygen appearing only as a secondary product. (i) YIELD IN TERMS OF DECOMPOSITION OF WATER OR (M/N)H,o. (4 Of oxygen (WN)o2. (A) PRIMARY REACTIONS IN THE RADIOLYSIS.--AS has already been shown, these reactions depend principally on the density of ionislation caused by radiation * and the phenomena may be contrasted if the radiolysis by natural a-rays and by X-rays is considered first.(i) FORMATION OF HzOz IN PURE WATER.-(^) Raysfrom Rn (6.5 MeV).-The quantities of hydrogen peroxide formed are proportional to the dose, with a constant yield of fornia- tion, i.e. ( M / N ) H ~ o , = 0.3 or G = 0.90. Within certain experimental limits this yield does not vary in practice with the radiation intensity, the temperature, the pH or the introduction of oxygen.5 (h) X-rays.-For doses up to 2 x 1011 ion pairs/ml the observed yields are very small 3 or 78i-0.5* 6 With a sensitive method we have not been successful in detecting HzOz * The mean linear ion density is onlv a crude estimate of the true distribution of ion- izations.These ionizations are grouped as 2, 3 or more and the mean spacing diminishes rapidly towards the end of the ionizing trajectory; for evaluating this spacing a mean grouping of 3 ion pairs is taken. The distribution is further complicated by the presence of &rays often representing more than half of the ionizations.29 4 -f The H202 is complexed with titanic sulphate and the peroxide is determined by means of ferrous sulphate and thiocyanate. The sensitivity of the method is 0.1 y/ml or 1.8 < 1015 molecules/ml. C’74 I R R 12 D1 A T E D P U R E W A 7 F K in water irradiated under vacuum ; 6 with 1 x 108 r (0.87 x 1018 ion pairs/nil). Cal- culation shows that the formation yield is, in this case, less than 0.02.The introduction of oxygen largely increases the yield which varies with a number of factors, e.g. dose, intensity, temperature and pH.79 3 (c) Accelerated heavy particles.-With protons (1 a 8 to 7.5 MeV) the yield decreases with dose as for X-rays and oxygenated solutions. The initial yield for small energies (1.8 MeV) is similar to that for natural cx-particles, i.e. Mihr-9-3. The initial yield FIG. 1. falls when the energy of the protons is increased (0.12 for 7-4 MeV) and the yieldper particle passes through a maximum for an energy of the order of 6 to 7 MeV (fig. 1). This important phenomenon, discovered by Toulis, is also found for dei:terons (maximum around 13 MeV) but has not been observed for a-particles, at least up to energies of 34 r"" MeV.Table 1 and fig. 2 clearly show the correlation between the initial yield * and ion density, although the experimental results are very limited in number and new experiments are very desirable. particles from Rn.-It has been known for a long time that the formation of hydrogen is proportional to the dose and the yield is constant but this agreement is not borne out by the values which vary, according to the authors (table l), from 0.8 to 0.6. (b) X-rays.-As for oxygenated water the observed yields are very small (3.6) and this is ofien attributed to presence of impurities in the water. Even when carefully purified, it can still contain con- centrations of the same order as that of the products of rddiolysis. We have undertaken, with Patti,s a series of experiments for measuring the liberation of hydrogen, with the apparatus (ii) FORMATION OF HYDROGEN.-(a) CC- * When the formation yield remains constant with increasing dose it can be considered to deal with the primary reaction.When the yield diminishes with the dose, its deter- mination at the beginning of the irradiation for doses as small as possible is of particular interest, because recombination reactions are also small and the primary formation re- action can be observed almost without complication ; this is termed the initial yield in table 1.P. BONET-MAURY TABLE 1 ion energy initial, ionic yield MJN in MeV density radiation (p.i./p.) 4600 fission of' boron c( = 1.522 a-particles and Li =0.88 H 2 0 ~ H2 yield con- - stant with Li nucleus) Rn 4000 a-particles froin 750 protons part ides deuterons protons 200 -deuterons ,&rays from .electrons I (tritium) 160 protons 109 X-rays I SO X-rays 10 electrons 6.5 1.8 31 8 12-14 6-7 0.005 7.5 0.2 0.017 1.0 dose 0.3 0.3 0.18 - 0.12 0.1 1 ._ 0.12 3.000 -0.0 < 0.02 - 0.8-0.6 - 0.25 0.18 - __ 0.1 3-0.03 0.15 0-4 0.03 0.13-0.03 0.17-0.07 shown in fig. 3. This apparatus has a volunie of 50 rnl 75 authors Bonet-Maury Duane and and Lefort 5 Scheuer 1 Nurnberger 1 Lanning and tind 1 Lefort 1 Toiilis 3 - 7 ) Toulis 3 9 9 Allen 1 2 7 Allen 1 7 9 Toulis 3 Fricke 5 Fricke 5 Toulis 3 and Patti 3 Bonet-Maury Bonet-Maury and Patti 8 I Allen 1 for the liberated gas and enables the irradiat~on of 3 1n1 of pure water contained in a quartz vessel to be carried out through FlG.3. an aluminium window F. the degassing and irradiation. The water is vigorously stirred by a magnetic stirrer M during The water vapour is condensed by a liquid nitrogen , _. - F~G. 4.76 IRRADIATED PURE WATER trap P and the final pressure is measured by means of a McLeod gauge reading to 10-4 niin Hg. The detachable vessel A is provided for analysis of the gas by chemical ab- sorption and the electrodes E for obtaining an electric spark to allow of the recombination of hydrogen and oxygen. We have observed, as have preceding authors, a small but quite measurable liberation of gas and a reasonable proportionality with dose up to 1 x 1018 ion pairs/ml(1.6 x 106 r). The yield (M/N)H, = 0.03 (fig. 4). Since the water, after high doses, becomes “ purified ” by the irradiation, this observed yield may be attributed to the radiolysis of water.to ‘8 Hz ;yield I I (c) Accelercrted particles.- The observed yield for protons of energy 7.4 MeV is pro- portional to the dose up to 13 x 1019 ion pairs/ ml, and is of the order 0.15. For x-particles of 31 MeV, MIN -= 0.25 compatible with values found using radon. For the formation of hydrogen the same correlation is found between the yield and the ionization density of the irradi- ation (fig. 5 and table 1). DISCUSSION Over all, these results show that the “X-ray chemistry” of water is very different from the “ ct-particle chemistry ”, the latter being apparently much simpler. In effect, not only does the a-yield remain constant but also the concentrations of radiolysis products obtained, even with small doses, are easy to measure without par- ticular precautions.On the contrary, even with high doses of X-rays (several million r) concentrations of the order of those of the impurities in the carefully purified water, are obtained, that is < 10-6. As the most sensitive analytical methods detect with difficulty less than 0-1 ylml the use of much more powerful sources of X-rays must be envisaged for further progress. The contrast between the u.- and the X-ray chemistry has already been inter- preted in a reasonable manner in terms of the chemical theory of Weiss 11 and the physical theory of Lea and Gray.4.12 If, following these theories, the H and OH radicals are distributed originally in the same way as the ions from which they were derived the initial concentrations of these radicals will be very different when the cc-particles and the X-rays are used.The ct-particle chemistry is characterized principally by the initial separation of the H and OH radicals, leading to the combination processes : OHi-OH - Hz02 4 51 kcal (initial OH concentration, 1 M) ; H+H = H2 + 103 kcal (initial H concentration, 8-7 Y 10-3 M) ; whereas the X-ray chemistry corresponds to a uniform distribution of both radicals favouring the recombination : H -t OH H20 + 11 8 kcal (initial concentration 7 ‘<, 10 5 M for an electron of 60 kV). Calculation shows in addition 4,12 that the reactions in cc-particle chemistry are effected principally inside each column of ions which constitutes an isolated system and the negligible influence of intensity of the a-radiation confirms the small importance of reactions between columns of ions.This chemical indi- viduality of the ionized trajectory must disappear for X-rays and the reactions occur throughout the bulk of the water, facilitating decomposition reactions because of the excess of radicals. The influence of radiation intensity in X-rayP. BONET-MAURY 77 chemistry confirms this because the mean primary concentrations of radicals depends on the mean ionic concentration, that is to say, the intensity. In spite of this simplicity, without doubt exaggerated, this interpretation allows equally well an explanation of the phenomena observed with accelerated heavy particles ; these latter form a progressive transition between a- and X-ray chemistry.In fact, a-particles of several MeV give yields of the same order as those from natural cr.-particles whereas for high energies the appearance of phenomena charac- teristic of X-ray chemistry is observed. Now this increase of energy involves a progressive spacing of the radicals, the initial concentration of which is lowered, whilst the separation of H and OH disappears. It is striking to find that the yield per particle passes through a maximum for an energy corresponding to a mean ion density of 200 ion pairs/p and a mean spacing of 15 mp for the OH radicals, that is to say, precisely the distance which separates them from the H radicals in a column of a-ionization. This is close to the value expected for the transition of columnar distribution to a sensibly homogeneous distribution of radicals produced by the absorption of X-rays in water.This interpretation involves, as a first approximation, a constant distribution of radicals along the length of the ionized trajectory and it certainly minimizes the part played by the 6-rays or the ends of the electronic trajectories. In order to take this into account more reasonably without entering into the details of the true distribution, as yet not well known, it is possible to represent schematically the trajectory of any ionized particle as composed of two parts : (i) " X-ray part ", where the loss of energy and the radicals are distributed in a sensibly uniform manner ; in the main they react to produce water by recombination.(ii) " a-part ", where the loss of energy is large; the two radicals are separated and give the compounds Hz and H202. The critical loss of energy corresponding to the transition from X-ray to a-particle chemistry, is of the order of 70 MeV/cm, i.e. about 210 ion pairs/p. If the %rays are considered, it is of the order of 140 MeV for a-particles (420 ion pairslp) and 64 MeV/cm (192 ion pairslp) for deuterons and protons. Now for natural a-particles the X-ray part is negligible but becomes important for a-particles of large energy. Conversely, the a-part plays only a small role for ordinary X-rays whereas it becomes the principal effect for electrons of small energy; the liberation of H2 which we have observed with X-rays is due to this a-component." For accelerated heavy particles the relative importance of these two effects varies with the energy; the yield per particle grows with the energy, then more slowly and passes through a maximum; it decreases subse- quently and then tends to the low X-ray yields (fig.6 ) . primary molecules (€3202 and H2) as their concentration increases with dose leads to reaction either between themselves or with the free radicals. Some secondary decomposition reactions will occur which in a general way will bring about a lowering of yield with dose. The parallel increase of amounts formed and de- stroyed leads finally to a dynamic equilibrium which manifests itself as a constant concentration independent of the dose. This equilibrium concentration will depend on a number of factors such as intensity and concentrations in the gas and liquid phases. The dynamics of these phenomena appear quite complex and the published experimental data are again few.It seems certain nevertheless that the equilibria of H202 and of H2 are established in a distinct manner, for different doses ; the equilibrium concentration of oxygenated water appears first. H202 EQUILIBRIA.-(~) NaturaZ a-particles.-No equilibrium has hitherto been reported, the concentration increasing linearly with the dose. * The corresponding yield of Hz02 will be of the order of 0.015, that is to say, at the limit of sensitivity of our method of chemical analysis. (B) SECONDARY REACTIONS-RADIOCHEMICAL EQUILIBRIUM.-The diffusion Of78 IRRADIATED PURE WATER (6) X-rays.-The experiments are difficult on account of the small primary yields.Toulis claims to have observed an equilibrium concentration of the order of 3 x 1015 molecules/ml (0.17 y/ml) which is greater than expected and which would correspond to a yield of the order of 43,000 molecules of H202 per ion pair. Allen, however, reports an equilibrium concentration of the same order for 1 MeV electrons. Our own cxperiments with Patti 8 have given a negative result, that is to say, less than lo15 molecules/ml for 1.8 x 1018 ion pairs/ml. (c) Heavy accelel.atectparticles.-For protons of 7.4 MeV the curve of forma- tion as a function of dose leads to an expected equilibrium concentration of 2 or 4 X 1019 molecules/ml or 600 ylrnl.3 (ii) H2 EQUILIBRTA.-(~) Natrrval a-particles.-The use of conccntrated solu- tions of radium salts for the preparation of Rn ampoules has led to thc old 0's- servation of the presence of this equilibrium for a pressure of the order of 706 inm m.The secondary appearance of oxygen in the gas mixture indicates a dccom- position of the H202 formed, which perhaps can be attributed to one of the three reactions : (i) H202 + H2 = 2H20, (ii) H202 4- OH = H20 -t "02, (iii) H202 + H = H2O 4- OH or HO2 1 H:. The first reaction is improbable; this is confirmed by the fact that elimination of hydrogen by evacuation does not modify the formation of H202.6 The second initiates the chain H202 + HO2 = 0 2 + H20 -:- OH, OH 4- H02 --- H20 + 0 2 (termination). (b) X-rays.-The equilibrium observed by Fricke is attributed by him to impurities in the water, but Allen has observed an equilibrium pressure for 1 MeV electrons (1-2 cm) and for 5 keV (10-20 cm) due probably to reactions of the or.-part of the electron trajectory (hot spots).Our experiments with Patti do not give any evidence of an equilibrium in the range of doses used (fig. 4). (c) Accelerated heavy particles.-An equilibrium is not expected for 7 MeV protons at pressures lower 3 than 1 atm. but 8 MeV deuterons lead to an equilib- rium pressure of 60 cm. The experimental study of these equilibria is again not very advanced and it seems most desirable to develop it further. INTERPRETATIoN.-The interpretation of the results is as yet not very advanced because these equilibria result from competition between several possible decom- position reactions, of which the respective probabilities depend on numerous factors.Physical factors determining the probability of erzcounter are potential barriers and, particularly, the number of collisions per second ; this number depends on the temperature, the concentration and the time of sojourn in the solvent cage, etc. The chemical factors, determining the probability of reaction at the moment of collision, are the state of excitation and the free energy change in the reaction.P . BONET-MAURY 79 The calculation is difficult and it can only be hoped that approximate solutions That given by Toulis leads to the division of the reacting species can be obtained. into three categories, in decreasing order of their probability of reaction : (1) probability about unity-interaction of radicals, H + H = H 2 H + OH 7 H20 OH + OH = H202 H02 + HO2 -- H202 f 0 2 H02 + OH = H2O + 0 2 . (2) probability about 10-3-interactions of the radical H with 0 2 and HO2, H + 0 2 = HOz H + H202 = H2O -1- OH; (3) probability about 10-5-interactions of the radical OH with H2 and H202, OH + H202 = HzO f HOz OH 1 H2 = H2O + H. 1 (a) Kernbaum, Radiiinr, 1909, 6, 225. (h) Duane and Scheuer, Radium, 1913, 10, ( d ) Allen, (e) Laming, and Lind, J. Physic. Chern., 1938, 42, 33. J . Physic. Chin., 1948,52, 479. 1229. (c) Nurnberger, J . Physic. Chem., 1937, 41, 431; 1934, 38, 47. 2 Fano, Nat. Bur. Stand., Report No. 1002, 1950. 3 Toulis,Tlie decomposition of water by radiation ; Berkeley and Toulis, The influence 4 Lea, Action of radiation on livirrg cells (Cambridge, 1946). 5 Bonet-Maury and Lefort, Compt. rend., 1948, 225, 1353, 1445 ; Nature, 1948, 162, 381. Bonet-Maury, Bull. Soc. Chim., 1951, 18, 333. 6 Fricke, Hart and Smith, J. Chenz. Physics, 1938, 10, 229. Fricke, Chemistry and Physics of Radiation Dosimetry, Sj~mposiim, 4, 1950. 7 Bonet-Maury and Deysine, Compt. rend., 1951, 232, 1101. * Bonet-Maury and Patti (in course of publication). 9 Gray, J . Chim. Pliys., 1951, 48, 472. 10 Risse, 2. physik. CIiem., 1929, 140, 133. 11 Weiss, Nature, 1944, 153, 743. 12 Dale, Gray and Meredith, Phil. Trans. Roy. Soc., 1949, 242, 33. of density of ionization on the decomposition of water and biological survival.