60 DOSIMETRY I N THE P I L E THE PROBLEM OF DOSIMETRY IN THE PILE BY 3. WRIGHT Atomic Energy Research Establishment, Harwell, Didcot, Berks. Received 12th February, 1952 Since energy absorption from pile radiation is high enough for calorimetric measure- ment at reasonable pile powers, such measurements when complete should provide a basis of dosimetry for pile radiation chemistry. Monitors are required for each individual irradiation so that the dose can be related to that received in unit time in the calorimeter. Nuclear physical monitors are not entirely satisfactory, but it may be possible to use as monitors chemical systems whose behaviour has been studied under carefully controlled conditions. Experiments on the irradiation of various systems are described, and it is shown that the oxidation of ferrous sulphate in 0.8 N H2S04 provides a suitable monitor for short irradiations.Experiments in which this system is used to study transient pile phenomena provide information about one component of pile radiation-the gamma radiation resulting from decay of fission products. The results show the importance of attaining steady pile conditions and emphasize the fallibility of megawatt-hours as a measure of dose. Certain results suggest that the chemical effects of neutrons and of gamma rays are not always additive for a mixed source, but further confirmation by more direct experiments is being sought. The experiments described in the present paper have been designed and monitored in such a way that they can be related to the calori- metric measurements when these are complete.It will then be possible to derive G values for pile radiation for the various systems studied.J. WRIGHT 61 As a source of neutrons for radiation chemistry, a pile has many advantages over sources hitherto available. A high ratio of neutron to gamma energy is provided, and this ratio can be varied over a wide range; energy absorption from pile radiation is high enough to permit calorimetric measurements at readily obtainable pile power levels; and the large volume within which a high dose rate is available may be useful in recovering measurable amounts of a substance produced only in low yield. Unfortunately, the nature of the source raises several difficult problems associated principally with dosimetry and the adoption of suitable monitors.Some of the problems encountered in quantitative work in pile radiation chemistry are set out below, and some of the experimental work undertaken in an effort to find suitable monitors is described. Attention is confined to thermal neutron reactors and, though most of the remarks apply equally to other types of pile, they are directed in the first place to natural uranium, graphite moderated piles of the type available at Harwell.1 THE BASIS OF DosIMETRY.-Tn experiments with gases the ion pair yield M/N of a product is a convenient way of expressing the result of radiation chemical action. The number of ion pairs formed in unit volume of the gases by a certain quantity of radiation may often he determined by experiment and, in certain cases, the concept of ion pair yield has been most useful in the interpretation of radi- ation chemical effects in gases.2 The possibility of contributions to the yield from excited molecules makes the concept of ion pair yield in radiation chemistry of less value than the corresponding concept of quantum efficiency in photo- chemistry and, like the latter, it is capable of theoretical interpretation only if the secondary reactions are related to the primary act in a simple stoichiometric manner.Moreover, in condensed systems the ion pair formation cannot usually be measured directly. For such systems it is better to express the yields on the basis of energy absorbed by the system, and the G unit (molecules converted or produced per 100 eV absorbed) proposed by Burton 3 is convenient for this purpose.Few radiation sources are sufficiently strong to permit direct measurement of the energy absorbed, though Stahel4 and his collaborators have used micro- calorimetric techniques for this purpose. For most radiations it is possible to use the relationship between energy deposition in a condensed system and ion- ization produced in a small air-filled cavity within the system. I n practice this involves the construction of an ionization chamber such that most of the ioniia- tions are caused by particles having their origin in the chamber walls which must be of material equivalent to that of the condensed system. With neutron sources, though this is still possible in theory, it is no longer so accurate nor so convenient in practice.Because of the large and irregular differences in neutron scattering and absorption cross-sections between the elements, the relation between ioniiation in an air-filled chamber and energy deposition in a condensed system varies considerably with the nature of the latter. It varies also with neutron energy and hence with neutron spectrum of any non-homogeneous source. Further com- plications arise from variations in the quantity of gamma radiation present in any composite source. The theoretical basis for relating ioni7ation in a gas-filled chamber to energy deposition in biological tissue has been investigated, especially by Gray5 and by Aebersold6 and a more general relation applicable to solid systems rich in hydrogen has been derived by Gray.7 From this work it is clear that the materials used for the walls and the gas in the ionization chamber must be changed when the nature of the condensed phase is altered.To avoid these complications and uncertainties it is clearly desirable to measure directly the energy absorbed in a system whenever possible. Fortunately, jt is possible to make direct calorimetric measurement of energy absorbed in various materials in the pile at reasonable power levels. Measurements have been made in the Harwell pile, but the results so far obtained for the different materials are inconsistent with their known physical properties such as mass absorp- tion coefficients for gamma rays and neutron scattering cross-sections. The62 DOSIMETRY IN THE PILE inconsistencies are such as to cast doubt on some details of the experimental procedure and they emphasiire the difficulty of designing calorimetric apparatus suitable for radiation dosimetry.To avoid giving values which may be mislead- ing, the energy absorption values will not be quoted until these anomalies are removed, and in the present paper chemical yields are expressed in terms of thermal neutron dose rather than as G values. The experiments have been monitored in such a way that they can be related to the calorimetric measurements at a later stage. These calorimetric measurements are being made by members of the Pile Group at A.E.R.E., Harwell, and it is hoped that the results will be published later by Mr. F. W. Fenning. The methods used for monitoring irradiations are therefore considered.MONITORING.-The nature of the calorimeter and the necessity of taking measurements over a relatively long period of time make it unsuitable for monitor- ing individual short irradiations. Although calorimetric measurements provide the ultimate basis for calculating radiation chemical yield, we need a monitor whose functions are to ensure that the nature of the incident radiation is constant and to provide a means of relating the dose received in the individual irradiations with that received in unit time in the calorimeter. For reasons which will appear in discussing the experimental results, integrated megawatt-hours derived from the pile control instruments are a most unsuitable measure of dose ; the monitor must accompany the irradiation vessel itself.No completely satisfactory scheme of nuclear physical monitors has so far been devised for radiation chemical purposes. Thermal neutron monitors, depending usually upon activation by some (n, y) process, measure a component of the radi- ation which contributes little to energy absorption processes in many chemical systems (including especially water and organic compounds), “ Threshold ” monitors, depending on activation processes which occur only with neutrons of energy above a certain minimum, measure only the high energy neutrons of which there are relatively few in the normal pile spectrum. There are no monitors for neutrons of a few kilovolts energy which contribute most to the energy deposi- tion in hydrogenous materials, but by using “ resonance ” detectors it is possible to measure the density of neutrons having energies in a narrow energy band, usually in the region of a few electron volts.A combination of all three types of monitor provides the best assurance against changes in neutron spectrum and, in the work described below, gold was used as thermal neutron monitor, the 32s (12, p ) 32P reaction as threshold monitor for neutrons above about 2 MeV, and the cadmium ratio for gold provided a reasonable monitor for neutrons of 4.8 eV at which energy gold has a strong resonance absorption. None of these monitors gives any information about the gamma component of the radiation and, in spite of the complications mentioned earlier, it is desirable to supplement the nuclear physical measurements by ionization chamber measurements.Unfortunately, no suitable chambers are available for use in the pile at normal power levels, whilst at very low power levels the gamma to neutron ratio is different. Undoubtedly the best monitor for irradiations of a chemical system would be another chemical system similar in nature and of known behaviour under controlled conditions. It is in the hope of building up informa- tion about a few such systems that the present work was started. EXPERIMENTAL MATERIALS.-GO~~ Foils, 1 cm square were cut from Ash’s Cohesive Gold Foil No. 4, prepared by The Amalgamated Dental Co. Ltd., and had an average thickness of about 2-5 mg cm-2. Sulphur pellets, $ in. diam. and $ in, long were prepared by melting Philip Harris’ pure crystallized sulphur and pouring the liquid into moulds.The Isotope Division, A.E.R.E., have used this material for several years and found that the only important activity developed by pile irradiation, other than from the sulphur, is a small amount of 24Na. Similar foils, 3 CM square, were used for the larger doses.J . WRIGHT 63 Ferrous sulphate.-A.R. ferrous sulphate was recrystallized from 0.8 N H2SO4 twice, the second time being immediately before preparing solutions for irradiation. Stock solutions were not kept more than one week. o-~~zenunthroline.-Hopkin and Williams' redox indicator o-phenanthroline was recrystallized twice from water and the monohydrate so formed gave colourless solutions stable for several weeks in diffuse light. Benzene.-A.R. benzene was redistilled and the middle fraction used.Mundelic acid.-May and Baker's L-mandelic acid, specific rotation - 158" to - 160", was used without further purification. Wuter.-All solutions for irradiation were prepared from specially purified water. Laboratory distilled water was redistilled from alkaline potassium permanganate and then either from alkaline manganous hydroxide (in early work), or from potassium bi-sulphate. The product was again distilled in silica apparatus and the distillate stored in closed silica vessels. All other chemicals were of A.R. grade and were used without further purification. IRRADIATIONS.-AI~ irradiations, except those of solid mandelic acid, were carried out in silica vessels inside a polythene carrier. The arrangement is shown in fig. 1. The polythene carrier was driven c by compressed air from the laboratory to a fixed position in the reacting core of the Harwell pile, taking 7 sec for the outward journey and 8 sec to return.A total of about 9 sec was spent in travelling inside the pile. While this led to small E uncertainties in the times of the shortest irradiations, no appreciable correction was necessary for calcul- ations based on the activity of the gold foils. Light signals, operated by a switch at the irradiation position, were used for accurate timing of the irradiation periods. The silica vessel developed beta activity due to the formation of 31Si, but there was no accompanying gamma radiation and the vessel could be handled with gloves for long enough to empty, wash, and refill. The weighed gold foils, - which accompanied every irradiation, were held in FIG.1 .--Arrangement of irradiation folded filter paper wrapped around the outside of the silica vessel. Irradiations of solid mandelic acid were con- ducted in aluminium cans placed in an experimental hole in the Hanvell pile and were accompanied by cobalt metal foils weighing about 2 mg for thermal neutron monitoring. ANALYsrs.-~ervous ion.-AII solutions were diluted with 0.8 N HzS04 after irradi- ation to give a total iron concentration of 10-4 M. 5 ml of diluted solution was added to 3 ml of 2 M sodium acetate solution containing 0.56 g/l. of NH4F. To this mixture was added 2 ml of 10-2 M o-phenanthroline solution and the optical density was measured 15 rnin later in a 1 cm cell of a Spekker photoelectric absorptiometer using llford filter 603.The concentration of residual ferrous ion in the irradiated solution was derived from a calibration curve constructed for synthetic mixtures of ferrous and ferric sulphates of total concentration 10-4M. Addition of F- was found to increase the stability of the solutions after colour development by complexing the ferric ion and retarding its conversion to ferrous o-phenanthroline. Too high a concentration of fluoride ion re- sulted in curvature of the calibration graph at the pure ferrous end. PHENOL.-TO 10 ml of irradiated solution was added 5 ml of Folin and Ciocalteu's reagent 8 followed by 15 ml 10 % Na2C03 solution. The mixture was made up to 50 ml and stood for 20 min at 25" C . The optical density was then measured in a 1 cm cell of a Spekker absorptiometer using Ilford filter 607.The concentration was derived from a calibration curve constructed for standard solutions of A.R. phenol. SucRosE.-The optical rotation of the irradiated solution was measured at 25" C in a 2 drn cell, using a Hilger pohrimeter Model M I 1 3. B D f /hch vessel in polythene carrier. A, cotton wool packing B, spherical ground surfaces c, gold foil ,,, polythene carrier E, silica irradiation vessel.64 DOSIMETRY I N THE P I L E MANDELIC ACID.-A weighed quantity of irradiated solid was dissolved in ethyl alcohol and the solution was made up to known volume. The optical rotation was measured at 20°C. MONITORING.-~hel'rna~ neutro~ dose.-Weighed gold foils were irradiated with all solutions and were usually counted next day in a reproducible position 11 cm below an end-window G-M tube.Groups of gold foils were counted alternately with 204Tl standards. Decay of activity of the gold foils, followed over 5 half-lives, gave a value of 64.3 h compared with the accepted value of 64.6 h for the half-life. A small amount of short-lived activity from impurity was of negligible importance t h after irradiation. Absolute calibration for thermal neutron dose was made by irradiating weighed foils in the G.L.E.E.P. for a known time in a known thermal neutron density and counting the foils in the standard position together with 204Tl standards. Correction to this calibration was made for the measured difference in cadmium ratio for gold in the G.L.E.E.P. and in the normal irradiation position in the Harwell pile.Resonance neutron dose.-The cadmium ratio for gold was determined by irradiating a bare gold foil and a foil covered by 1 mm thickness of cadmium successively in the same position. Simultaneous irradiation of bare foils in another part of the pile in each case enabled corrections to be made for any variations of pile behaviour between the two successive irradiations. A brief study has been made of some factors influencing cadmium ratio so that the effect of changes in experimental conditions could be assessed. The irradiation position was in a part of the pile where changes of equipment were seldom made, and experience showed that it was sufficient to check the value of the cadmium ratio at intervals of about two weeks.Fast neutron dose.-Sulphur pellets were irradiated inside the irradiation vessel which was filled with water for these measurements. After irradiation they were dried, weighed and dissolved in concentrated HNO3. The solution was made up to 50 ml with water and a few days later, when the 24Na activity was negligible, the solution was counted in a liquid counter previously calibrated with a solution of known 32P activity. Thermal neutron reactions led to active species (e.g. 34s) with radiations too soft to register in the liquid counter, and covering the sulphur pellets with cadmium during irradiation did not alter appreciably the activity measured. The fast neutron dose was checked at intervals throughout the work. RESULTS AND DISCUSSION NUCLEAR PHYSICAL MoNIToRS.-The absolute flux calibration of the gold foils showed that a foil giving 1000 counts min-1 mg-1 under standard counting con- ditions, corresponded to a thermal neutron dose of 5.42 x 1012 neutrons cm-2 in the G.L.E.E.P. and to 6-06 x 1012 cm-2 in the normal irradiation position in the Harwell pile.With the pile in a steady state, variations of thermal neutron flux were seldom more than 5 % from the mean value over a period of about 1 h. Results expressed in terms of gold activity always gave a smaller standard error than those on a time basis. On the other hand, variations of flux of 15 % or more were observed from day to day for the same nominal pile power. The con- stancy for short periods was due to the permanence of equipment in the vicinity of the irradiation position while the larger variations over a long period resulted from differences in pile loading or local disturbances near the controlling ionization chambers.The cadmium ratio for gold under the conditions of irradiation in the Harwell pile was found to be 2.70 (activity of bare foil divided by activity of cadmium covered foil). In the absence of the water-filled irradiation vessel the value fell to 2.61. No significant changes in these values occurred during the present series of experiments. The cadmium ratio for gold in the G.L.E.E.P. was 2.26. The high energy neutron flux under the conditions of irradiation in the Harwell pile was such that a thermal neutron dose of 1016 neutrons cm-2 corresponded to a 32P activity of 8-5 pc/g sulphur. In the absence of the water-filled irradiation vessel the value was 9.4 pc/g sulphur. OXIDATION OF FERROUS IoN.-Irradiations of ferrous sulphate were made in air-equilibrated 0.8 N H2SO4 solution and a series of irradiations of increasingJ .WRIGHT 65 duration were conducted up to complete conversion to ferric at each concentration. The results are summarized in table 1. TABLE RA RATE OF OXIDATION OF FERROUS SULPHATE ( 1 2 = thermal neutrons) (Fez numbers of ferrous ions) rate of oxidation _.__ - _ - ~ dose rate -- ~ after oxygen number of ('1 ~~~~~~ I in presence of air depletion irradiations (Fe2+1n]-l,~-l,-~2 number Of iiridations (Fe2+rnl-ln-km* x 10-3) x10 3) 7.04 6.7 1 6.75 6.82 7.06 6-82 6.85 6.77 6.73 7.98 7-79 4.06 2.10 1.05 0.62 0.27 FIG. 2.-Oxidation of 2-5 x 10-3 M ferrous sulphate solution ; dose rate, 6.70 x 1011 n cm-2 sec-1.3.34 L 0.10 3-35 1 0.04 3.31 1 0.06 3.24 0.07 3.21 0.21 3.10 - 0-10 3.47 - 0.15 3.03 - 0.21 - 3.14 J 0.10 3.14 0.07 3.44 I- 0.05 3.02 y 0.14 3-35 f 0.04 3.06 2 0-13 3.17 4 0-03 76 17 1s 25 10 6 15 9 7 7 7 16 7 8 I0 - b 80.5 9 ,d 100 4 f Y / For initial Fez+ concentrations from 1 0 - 4 to 10-3 M, a linear relation between dose and Fe2+/ml oxidized was observed up to complete conversion to ferric, but at higher initial Fe2f concentrations the relation was represented by two straight lines (fig. 2). The point of intersection of these two lines occurred at approxim- ately the same dose in all cases and was identified with the "oxygen break" where all the oxygen in the solution had been used and the aerated rate of oxidation gave place to the lower oxygen-free rate.When oxygen-equilibrated solutions66 DOSIMETRY I N THE PILE were used, the break did not occur till much higher doses were reached, whilst with nitrogen-equilibrated solutions, the break did not occur and the lower rate of oxidation was obtained from the start. Within the limits of experimental error, this rate was the same as that at higher doses with air-equilibrated solutions. The rate of oxidation of ferrous sulphate was independent of initial concen tra- tion of Fez+ within the limits of experimental error from 10-4 M to 3 x 10-3 M. The linearity of the graphs for an initial concentration of 10-4 M Fezi- to well beyond 90 %’ oxidation (fig. 3) enable extension of the lower limit of concentration independence to 18-5 M.The second part of tabIe 1 relates to experiments at different dose rates and in all cases the pile was allowed to reach steady state conditions before the series of irradiations was carried out. This is most important at low pile powers if changes in the incident neutron to gamma energy ratio are to be avoided. The P n x 16’~ 2 FIG. 3.-Oxidation of 10-4 M ferrous sulphate solution ; dose rate, 7-04 x 1011 n cm-2 sec-1. results show that the oxidation for unit dose was independent of the dose rate from about 8.3 x 1011 n cm-2 sec-1 to about 8 x 1011 IZ cm-2 sec-1. The mean value for the rate of oxidation in the presence of oxygen is 3.24 0.1 1 x 103 Fe2f ml-1 I T - 1 cm2, and in the absence of oxygen, I -88 + 049 i; 103 Fez+ 1111-1 12-1 cm2.When calorimetric measurements are complete it should be possible to derive G values for pile radiation which may be compared with the published values for gamma radiation 9 and for alpha radiation.10 5 x 1016 Fez+/mI have been oxidized. From the amount of oxygen present initially in aerated 0.8 N H2SO4,11 it may be calculated that this corresponds to approximately four Fez+ oxidized per oxygen molecule used. The number of ferrous ions oxidized per ml at the oxygen break agrees quite closely with the value obtained by Fricke and Morse 12 for X-rays and with Miller’s result 9 for gamma radiation (54 x 1016 Fe2+/ml) despite differences in experimental conditions, type of radiation, ratio of oxygenated to oxygen-free rates and dose rate. The oxygen break occurs when 60J .WRIGHT 67 These experiments showed that the system FeS04 in 0.8 N HzSO4 was suitable for studying pile characteristics. The amount of oxidation provided a simple integration of the effects of " normal " pile radiation * as measured by the neutron dose, and any deviations from this behaviour could be attributed to changes in the nature of the radiation (for instance, to changes in neutron to gamma energy ratio). The rate of oxidation was high enough to permit accurate measurements with irradiation of a few seconds at the higher dose rates. This was valuable in studying transient phenomena. For monitoring purposes it has been found convenient to use aerated solutions 10-4 M or 2 x 10-4 M in ferrous ion, and to irradiate to about 70 % of complete oxidation whenever possible. Reasonably reliable results can be obtained for 25 % conversion of an aerated 10-4 M ferrous solution and for 80 % conversion of an aerated 10-3 M ferrous solution, giving a factor of about 30 in dose.For FIG. 4.-Decrease of gamma radiation following shu t-down. A, shut-down from dose rate of 6.8 x 1011 n cm-2 sec-1 B, shut-down from dose rate C , shut-down froin dose ratc of of 0.62 x 1011 n cm-2 sec-1 0.27 x 1011 IZ cm-2 sec-1. higher doses, the upper limit can be extended by using nitrogen equilibrated or de-aerated solutions. PILE cmRAcTmIsrrcs.-The gamma component of pile radiation is derived from three sources-from fission, from (n, 7) processes, and from decay of radio- active species (particularly fission products).For a given irradiation position, energy from the first two sources will be proportional to thermal neutron flux in that region of the pile, but energy from the third source will vary in a complex manner with the past history of neutron flux in that region. The magnitude of this component is therefore of interest in considering the behaviour during pile power variations. It has been studied by irradiating ferrous sulphate solutions at intervals after shut-down of the pile. Some results are shown in fig. 4. In irradiations made a few seconds after the shut-off ro s had been lowered, an appreciable neutron dose due to delayed neutrons was observed. The * The term " normal '' pile radiation is used in the remainder of this paper to refer to the particular mixture of neutron and gamma radiation used in the ferrous sulphate studies described above.It constitutes a useful reference point in discussing variations in composition of the radiation but has no special significance or applicability to other piles.68 DOSIMETRY IN THE PILE observed chemical effect has been corrected for the contribution of normal pile radiation which, it may be assumed, accompanied these neutrons. The correction becomes negligible within a few minutes of shut-down, and only the first points on each curve of fig. 4 are affected. The three curves of fig. 4 show the fall in oxidation rate as the fission product gamma radiation decreased when the pile was shut down from three different power levels. Curve A was obtained after steady running at the power level at which the pile normally operates and therefore represents the decay of fission products whose initial intensity was characteristic of that power level.Curves B and C, on the other hand, were obtained after running at much lower power levels and are of different shape from A. Although the pile had been running steadily at these low powers for many hours the longer-lived fission products were still of intensity corresponding to much earlier high power running. By subtracting from curves B and C the contribution made by this earlier running, values can be derived which represent the contribution to the shut-down gamma radiation made by running at the lower power. The resultant curves are then of the same shape as A and may be superimposed by adjusting the ordinate scales in proportion to the dose rates before shut-down. The steep slope of the curves FIG.5.-Variation in oxi- dation rate of ferrous sul- phate as the pile attains steady - state conditions ; dose rate decreased from 7.7 x 1011 n cm-2 sec-1 to 0.53 x 1011 n cm-2 sec-1 at zero time. at the start precludes extrapolation to zero time, but the first observed values are about 10 % of the oxidation rates obtained before shut-down. As an illustration of the practical effect of the gamma component due to residual fission product decay, we may consider the effect of changing pile power. Fig. 5 shows the rate of oxidation of ferrous sulphate at various times after the pile power had been reduced. The reduction of power took only 2 min and was complete at zero time on the scale: subsequently the pile power was maintained constant according to the control instruments.The decrease in oxidation rate with time after the power change is due to the decrease of gamma radiation as the fission product activities decay from the high level corresponding to high power running to the new, lower level. This residual gamma radiation is addi- tional to the normal pile radiation corresponding to the low power level and its value may be obtained from shut-down curves like A, fig. 4. Not only is the total energy deposition changing after alteration of pile power, but the proportion of that energy due to neutrons is also changing. The time required to reach stable conditions clearly depends on the magnitude of the power level change and may be 24 h or more in extreme cases.The broken curve in fig. 5 was calculated on the assumption that the chemical effects of normal pile radiation and of the residual fission product gamma radiation were additive, The observed values are lower than the calculated ones in allJ . WRIGHT 69 cases and this suggests that the chemical effects of normal pile radiation and of residual gamma radiation are not additive. Several other data obtained in work with pile radiation lead to the same conclusion but none is sufficiently well substantiated to be decisive. If the non-additivity of chemical effects is confirmed, it will imply that the yield of ferrous oxidation is liable to vary with changes in the proportion of neutron and gamma energy contributions.The rate of oxidation quoted in the preceding section would then be representative only of the particular mixture constituting normal pile radiation and would not apply to other piles or even to other positions in the same pile. This point is being studied at present by experiments in which the proportions of different types of radiation are being changed deliberately. The results of these experiments are clearly important in considering the use of chemical systems as monitors for pile radiation. More direct tests are planned in a search for evidence for or against additivity of chemical effects of neutron and gamma radiation. The result has practical as well as theoretical significance since correction is often made by subtracting a " blank " value derived from separate irradiations with one of the components of a mixed source (e.g., the gamma component of a Ra/Be neutron source).r" FIG. 6.-Diagram showing the re- lative positions of experimental hole and uranium channels in the Har- well pile. graphite ; uranium metal ; experiment a1 hole. The ratio of neutron to gamma energy deposition in a system varies according to its position in the lattice of the pile, being higher at positions marked A in fig. 6 than in position B. This is due to the varying thickness of graphite moderator between the irradiation position and the uranium metal. The magnitude of these changes and their chemical effects have been studied. The ratio of neutron to gamma energy deposition may be changed deliberately, being increased by surrounding the specimen with uranium and decreased by enclosing it in hydrogenous material.Studies may also be made at very low pile powers where the residual fission product gamma radiation from previous high power running is an appreciable fraction of the total incident energy. During experiments on transient pile phenomena, effects were noticed which illustrate the fallibility of megawatt-hours recorded by control instruments as a measure of dose. For several hours after a change of pile power, the thermal neutron flux at the irradiation position was found to be drifting although the power level registered by the control instruments was quite steady within 1-2 min of the power change. A drift in thermal neutron flux of 10 % in 3 h was often observed. This effect is partly due to changes in the " shadow " cast by the control rods on the control ionization chambers.When pile power is reduced, the gradual70 DOSIMETRY IN THE PILE reduction of temperature and of fission product “poisoning” result in an in- creasing reactivity of the pile which must be compensated by a gradual movement of the control rods inwards. In the Harwell pile the control rods enter on the same face as the ionization chambers and, as the rods move into the reacting core, they partly screen the chambers from thermal neutrons in that direction. Since the current output from these chambers is being held steady for control purposes, the flux in other parts of the pile less affected by the control rods must increase so long as the rods are moving inwards. The results shown in fig.5 have been corrected for this effect by allowing for the thermal neutron dose recorded by the gold monitor. All other experiments were carried out under steady state conditions when this effect was negligible. The differences in control rod setting for the same nominal pile power probably account for most of the day-to-day variation in thermal neutron dose rate already noted. We must conclude that megawatt-hours recorded by the control instruments are unreliable as a measure of dose. I 5 1 FIG. 7.-Production of phenol in aerated solutions of benzene in water ; dose rate, 8.4 x 1011 n cm-2 sec-1 IRRADIATION OF BENZENE IN AQUEOUS SOLUTION.-h a search for alternative systems for monitoring purposes, a study is being made of the production of phenol by irradiation of saturated solutions of benzene in water first described by Weiss and his co-workers.13 A graph showing the behaviour of saturated solutions of benzene in air-equilibrated water is shown in fig.7. The same type of behaviour is observed as in the more concentrated solutions of ferrous sulphate. The oxygen break occurs at the same number of phenol molecules per ml as in Weiss’ experiments with X-rays and corresponds to about six phenol molecules per oxygen molecule originally in the water. Much more work is required on this system, particularly in de-aerated solution, before it can be recommended for use in monitoring pile irradiations. The other products, of the irradiation of bemene in water are also being studied, particularly the aldehyde which Weiss found after neutron irradiation.14 sucrose are irradiated in the pile a slow change occurs in the optical rotatjon of the solution.A graph showing the change in the specific rotation of a 10-2 M solution with time of irradiation in the pile is given in fig. 8. No detailed work on this system has been undertaken so far, and the chemical changes which give rise to this effect may be quite complex, but it may prove of value as a monitor for large doses and in cases where a result is required within a short time of the end of irradiation. IRRADIATION OF SUCROSE IN AQUEOUS SoLuTIoN.-When aerated Solutions OfJ . WRIGHT 71 IRRADIATION OF SOLID MANDELIC ACID.-AS an example of quantitative measure- ments made on a solid system irradiated in the piIe, fig. 9 shows the change which FIG. S.-Decrease of optical rotation of sucrose solution ; dose rate, 8.0 x 1011 n cm-2 sec-1. FIG. 9.-Change in optical rotation of solid mandelic acid. occurs in the specific rotation of I-niandelic acid when this material is irradiated in the solid state in the pile. The principal chemical product of this irradiation is benzaldehyde, but the steps leading to its production are not known.72 IRRADIATED PURE WATER The author wishes to express his thanks to Mr. F. W. Fenning for helpful discussions and advice on pile physics and to Mr. H. Morley and colleagues in the pile radiation chemistry group for assistance in the experimental work. This paper is being published by permission of the Director, A.E.R.E. 1 Harwell ; the British Atomic Energy Researcli Establishmetit (H. M.S.O., 1 952), appendix B. 2 Eyring, Hirschfelder and Taylor, J. Cliein. Physics, 1936, 4, 479, 570 ; Hirschfelder and Taylor, J . Chem. Physics, 1938, 6, 783 ; Lind, Tlie Chemical Efects of Alpha Particles and Electroiis (Chem. Catalog. Co., Inc., 1928). 3 Burton, J. Physic. Chem., 1947, 51, 613. 4 Stahel, Strahlentherapie, 1929, 31, 502 ; 33, 296. 5 Gray, Proc. Roy. SOC. A, 1936, 156, 578 ; Gray and Read, Nature, 1939, 144, 439. 6 Aebersold and Anslow, Physic. Rev., 1946, 69, 1. 7 Gray, Proc. Camb. Phil. Soc., 1944, 40, 72. 8 Folin and Ciocalteu, J . Bid. Clzem., 1927, 73, 627. 9 Miller, J . Chem. Physics, 1950, 18, 79. 10 Nurnberger, J. Physic. Chem., 1934, 38, 47. 11 Bohr, 2. physik. Chern., 1910, 71,47. 12 Fricke and Morse, Phil. Mag., 1929, (7), 7, 129. 13 Stein and Weiss, J. Chern. Soc., 1949, 3245. 14 Stein and Weiss, J . Chem. Soc., 1949, 3254.