"Science, Faculty of"@en . "Physics and Astronomy, Department of"@en . "DSpace"@en . "UBCV"@en . "Heiberg, Severin Andreas"@en . "2012-02-07T22:48:01Z"@en . "1954"@en . "Doctor of Philosophy - PhD"@en . "University of British Columbia"@en . "Two boron trifluoride proportional counters, one containing normal isotopic boron and the other boron enriched to 96% B\u00C2\u00B9\u00E2\u0081\u00B0, have been irradiated with 4.87-Mev neutrons from the D +.D reaction. In addition to the reactions B\u00C2\u00B9\u00E2\u0081\u00B0(n,\u00CE\u00B1)Li\u00E2\u0081\u00B7and B\u00C2\u00B9\u00E2\u0081\u00B0(n,\u00CE\u00B1)Li\u00E2\u0081\u00B7* with Q-values of 2.79 Mev and 2.31 Mev respectively, two other reactions have been observed, (i) F\u00C2\u00B9\u00E2\u0081\u00B9(n,\u00CE\u00B1)N\u00C2\u00B9\u00E2\u0081\u00B6and F\u00C2\u00B9\u00E2\u0081\u00B9 (n, \u00CE\u00B1)N\u00C2\u00B9\u00E2\u0081\u00B6* with Q-values of -1.43\u00C2\u00B10.15 Mev and -1.77\u00C2\u00B10.15 Mev and (ii) either B\u00C2\u00B9\u00E2\u0081\u00B0 (n,p)Be\u00C2\u00B9\u00E2\u0081\u00B0 or B\u00C2\u00B9\u00E2\u0081\u00B0 (n,t)Be\u00E2\u0081\u00B8 with a Q-value of +0.35\u00C2\u00B10.20 Mev. Due to the presence of these two reactions, the analysis of complex fast neutron spectra by the use of such counters is not feasible. The Q-values for reaction (i) yield a value of the N\u00C2\u00B9\u00E2\u0081\u00B6 mass of 16.01110\u00C2\u00B1.00020 MU.\r\nNeutrons from a pulsed deuterium beam impinging on a tritium target were used to bombard a boron trifluoride proportional counter containing the normal ratio of B\u00C2\u00B9\u00C2\u00B9 to B\u00C2\u00B9\u00E2\u0081\u00B0. The half-life of the activity and the energy of the particles emitted indicated that they were due to the immediate breakup of Be\u00E2\u0081\u00B8 into two alphas after the 0.89 sec. beta decay of the Li\u00E2\u0081\u00B8 formed by the B\u00C2\u00B9\u00C2\u00B9 (n,\u00CE\u00B1)Li\u00E2\u0081\u00B8 reaction. The process was found to have a cross section of the order of 10 millibarns for 14-Mev neutrons.\r\nThe angular distribution of the non-resonant gamma radiation from the proton bombardment of C\u00C2\u00B9\u00C2\u00B2 has been measured and found to obey the relation:-\r\nI(\u00C9\u00B5)\u00CE\u00B10.02\u00C2\u00B1.02 + sin\u00C2\u00B2\u00C9\u00B5, \r\nfor a proton energy of 1,580 kev."@en . "https://circle.library.ubc.ca/rest/handle/2429/40532?expand=metadata"@en . "REACTIONS INDUCED BY FAST NEUTRONS IN BORON TRIFLUORIDE AND THE ANGULAR DISTRIBUTION OF THE NON-RESONANT GAMMA RADIATION FROM THE BOMBARDMENT OF CARBON WITH PROTONS by SEVERIN ANDREAS HEIBERG A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in the Department of PHYSICS We accept this thesis as conforming to the standard required from candidates for the degree of DOCTOR OF PHILOSOPHY. Members of the Department of Physics THE UNIVERSITY OF BRITISH COLUMBIA November, 1954 ABSTRACT Two boron trifluoride proportional counters, one containing normal isotopic boron and the other boron en-riched to 96% B1^, have been irradiated with 4.87-Mev neutrons from the D +.D reaction. In addition to the 10 7 10 7 ^ reactions B (n .^OLi and B (n , \u00C2\u00AB 0 L i with Q-values of 2.79 Mev and 2.31 Mev respectively, two other reactions have been observed, (i) F 1 9(n,o<)N 1 6 and F 1 9 ( n , * ) N 1 6 * with Q-values of -1.43+0.15 Mev and -1.77+0.15 Mev and ( i i ) either B 1 0(n,p)Be 1 0 or B 1 0(n,t)Be 8 with a Q-value of +0.35J10.20 Mev. Due to the presence of these two reactions, the analysis of complex fast neutron spectra by the use of such counters is not feasible. The Q-values 16 for reaction (i) yield a value of the N mass of 16.Olliot.00020 M0. Neutrons from a pmlsed deuterium beam impinging on a tritium target were used to bombard a boron trifluoride proportional counter containing the normal ratio of to B^. The h a l f - l i f e of the activity and the energy of the particles emitted indicated that they were due to the im-mediate breakup of Be 8 into two alphas after the 0.89 sec. beta decay of the Li** formed by the (n,<>0Li8 reaction. The process was found to have a cross section of the order of 10 millibarns for 14-Mev neutrons. The angular d i s t r i b u t i o n o f the non-resonant gamma r a d i a t i o n from the proton bombardment of C has been measured and found t o obey the r e l a t i o n : -I ( e)\u00C2\u00AB*0.02\u00C2\u00B1.02 + s i n 2 6 9 f o r a proton energy of 1,580 kev. THE UNIVERSITY OF BRITISH COLUMBIA Faculty of Graduate Studies PROGRAMME OF THE FINAL ORAL EXAMINATION FOR THE DEGREE . OF DOCTOR OF PHILOSOPHY of SEVERIN' ANDREAS HEIBERG B.Sc. (Alberta) 1948 M.Sc. (Alberta) 1950 THURSDAY, APRIL 7TH, 1955 AT 3:00 P. M, IN ROOM 301, PHYSICS BUILDING COMMITTEE IN CHARGE Dean H.F. Angus, Chairman G.M. Shrum J.G, Hooley K.C. Mann R.F. Osborne J.B. Warren H.B. Hawthorn W. Opechowski T.E. H u l l G.M. Volkoff External Examiner - D. B. Scott University of Alberta THESIS REACTIONS INDUCED BY FAST NEUTRONS IN BORON TRIFLUORIDE and THE ANGULAR DISTRIBUTION OF THE NON-RESONANT GAMMA RADIATION FROM THE BOMBARDMENT OF CARBON WITH PROTONS ABSTRACT \" Two boron trif l u o r i d e proportional counters, one containing normal isotonic .boron ;and^ the, other boron enriched to 96$ BlO ? have been irradiated with 4.#7-Mev neutrons from the D(d,n)He3 reaction. In addition .to; the-reactions B l O ( n , ^ ) L i 7 and B-W (n,\u00C2\u00AB)Li'*with Q>values of 2.79 Mev and 2.31 Mey respectively, two other reactions have been ob-served: (i) F^n/aON 1' and F i 9 ( n ^ N l 6 * Q_ values of -1.43- 0.15 Mev and -1.77- 0.15 Mev and (i i ) either BlO(n,p)Be 1 0 or BlO( n,t)Be 8 with a Q-value of +0.35^ 0.20 Mev. Due to the presence of these two reactions, the analysis of complex fast neutron spectra by the use of such counters i s not feasible. The Q-values forjreaction (!) yield a value af the Nl6 m a s s of 16.QJlllot 00020 -MU* Neutrons from a pulsed deuterium beam impinging on a tritium target were used to bombard a boron t r i -fluoride proportional counter containing the nor-mal ratio of BH to B 1 0. The h a l f - l i f e of the activity and the energy of\"the particles emitted indicated that they were due :to the immediate breakup of Be 8 into two alphas after' the 0.89-sec. beta decay of the L i 8 formed by the B 1 1(n ip<)Li t J reaction. The process was found to have a cross section of the order of 10 millibams for 14-Mev neutrons. The angular distribution of the non-resonant gardma radiation from the proton bombardment of C 1 2 ha\u00C2\u00A3 been measured and^found to obey the relation 1.(9) o<0.02 *.02 + sin^ 9 for a proton energy of 1,580 kev. GRADUATE STUDIES F i e l d of Study - Physics Electromagnetic Theory Electronics Quantum Mechanics Nuclear Physics W. Opecnowski F. K. Bowers G. M. Volkoff K.G. Mann Other Studies: D i f f e r e n t i a l Equations - T.E. H u l l Theory of the Chemical Bond - C, Reid Radiochemistry M. Kirsch K. Starke Topics i n Physical Chemistry J.G. Hooley and others ACKNOWLEDGEMENTS The author i s indebted to Dr. J. B. Warren for his kind supervision of the work which is described in this thesis. It is a pleasure to acknowledge the help and suggestions of the many staff members and research students whose work on the building and development of the University of British Columbia Van de Graaff generator and the 50 kilovolt acceler-ator during the past years has made much of the present work possible. I wish to thank Dr. C.A. Barnes for his guidance during the early stages of this work which were performed under his supervision, and Dr. D.B. James whose advice and assistance in running the Van de Graaff generator and in making the measurementswas invaluable. Finally I wish to thank the National Research Council for financial assistance in the form of a studentship during a part of this work. TABLE OF CONTENTS Chapter T i t l e Page I. INTRODUCTION 1 II. REACTIONS FROM THE FAST-NEUTRON IRRADIATION OF BORON 3 (a) THE DETECTION OF NEUTRONS 3 (b) THE MEASUREMENT OF THE ENERGIES OF FAST NEUTRONS 5 (c) NEUTRON CAPTURE REACTIONS WITH BORON TRIFLUORIDE 7 (i) Possible Reactions 7 ( i i ) Experimental 9 1. The Boron Trifluoride Counters 9 2. The Neutron Source 10 ( i i i ) Results 12 1. Identification of the Reactions 12 2. The F* 9 Reaction Group (5) 13 3. The B Reaction Group (6) 15 (iv) Conclusions 15 (d) THE DETECTION AND CROSS SECTION OF THE REACTION B (n ,oc)Li 16 (i) Previous Evidence for the Reaction 16 ( i i ) Theoretical estimate of the Order of Magnitude of the Cross Section 20 ( i i i ) The Beta Decay of L i 8 28 (iv) Experimental 29 1. The Neutron Source 29 2. The Boron Trifluoride Counter 32 3. The Monitor 34 4. Beam Modulation 35 5. Amplification and Counting of the Pulses 36 6. The Time Sorter 36 7. Procedure 39 (v) Results 41 (vi) Conclusions 43 Chapter T i t l e Page III. THE ANGULAR DISTRIBUTION OF NON-RESONANT GAMMA RADIATION FROM THE PROTON BOMBARDMENT OF CARBON 44 (i) The Reaction C 1 2(p,*p')C 1 2 44 ( i i ) Experimental 45 1. Target Preparation 45 2. Beam Current Measurement 46 3. The Scintillation Counter 47 4. Procedure 48 ( i i i ) Results 50 1. Solid Angle Corrections 50 2. The Angular Distribution 52 (iv) Conclusions 54 Appendices I. THE USE OF AN ELECTRODELESS RADIO-FREQUENCY DISCHARGE AS A SOURCE OF DOUBLY CHARGED HELIUM IONS 55 1. General 55 2. Excitation Processes in the Discharge 56 3. Experimental 60 4. Performance of the Source 61 II. A HELIUM GAS LEAK 65 III. A HYDROGEN LEAK USING DIFFUSION THROUGH NICKEL 67 IV. A SAFETY VALVE FOR THE VAN DE GRAAFF GENERATOR ION SOURCE . . . . 70 V. CALCULATIONS FOR THE THEORETICAL PREDICTION OF THE B i : L(n , o O L i 8 CROSS SECTION 72 VI. CALCULATION OF THE EXPERIMENTAL CROSS SECTION OF THE Bi:L(n,\u00C2\u00B0<)Li8 REACTION 76 BIBLIOGRAPHY 79 LIST OF ILLUSTRATIONS Plates I. C 1 2 ( p , y p ' ) C 1 2 apparatus 45 Figures 1. Target and counter assembly 10 2. Pulse spectrum from the bombardment of enriched boron trifluoride with slow neutrons 11 3. Pulse spectrum from the bombardment of enriched boron trifluoride with 4.87-Mev neutrons 12 4. Pulse spectrum from the bombardment of unenriched boron trifluoride with 4.87-Mev neutrons 13 5. Pulse spectra with the slow neutron background subtracted 14 6. Energy level diagrams for B 1 0 + n and F* 9 + n 15 7. Hammer tracks produced by cosmic rays 17 8. Energy level diagram for B 1 + n 17 9. Schematic diagram of the 50 kilovolt accelerator 30 10. Construction of the boron trifluoride counter 33 11. Construction of the neutron monitor 34 12. Circuit diagram of the synchronizer 35 13. Schematic diagram of time relations 36 14. Block diagram of the experimental arrangement 37 15. Circuit diagram of the time sorter 38 16. The decay curve of the delayed activity 41 17. The amplitude distribution of the delayed pulses 42 18. Energy level diagram for C + p 44 19. Excitation function for the reaction C 1 2 (p t Yp\u00C2\u00BB)cl2 44 20. Experimental arrangement for C 1 2(p , Y p 1)Cl2 46 21. The photomultiplier and head amplifier c i r c u i t 47 22. The sodium iodide crystal mount 47 23. Curves showing the dependence of the gamma ray energy on the proton energy 48 24. The gamma ray spectrum 49 25. Me&hod of calculating the solid angle corrections 51 26. Angular distribution of the non-resonant gamma rays 53 27. The radio-frequency ion source 60 28. The oscillator c i r c u i t 61 29. The spectrum of ions obtained using helium 62 30. Construction of the helium leak 66 31. Power consumption of the helium leak 66 32. Construction of the nickel tube hydrogen leak 68 33. Power consumption of the hydrogen leak 68 34. Construction of the \"ball protector\" 70 I INTRODUCTION This thesis describes the results of some work carried out on two unrelated problems in nuclear physics. The use of boron trifluoride counters for the detection of slow neutrons and i t s use in the form of a paraffin-shielded long counter for monitoring fast neutron fluxes i s well known. The reactions proceeding when fast neutrons irradiate such a counter have not been carefully examined hitherto. It was 10 7 thought that i f the reaction B (n,\u00C2\u00AB<)Li were the dominant process| then the pulse height distribution might serve to indicate the energy of the incident fast neutrons. In chapter II a study of the pulse height spectrum for monokinetic fast neutrons i s described, which was found to be so complex as tomake this method of l i t t l e value for the measurement of neutron i energies. However i t did lead to an assignment of the various groups in the distribution to reactions involving B^, B ^ and 19 F . Further, a study was made of delayed pulses occurring after the beam of neutrons i s switched off, which are initiated only by very fast neutrons (E n>7.3 Mev). These pulses are considered 11 8 to be evidence for the reaction B (n,*<)Li . A considerable amount of work in nuclear physics has been undertaken to measure the spectra of gamma rays occurring when light nuclei are bombarded with protons and deuterons. Nearly -1--2-a l l of this has been concerned with the resonant yield of gamma rays, since, after a l l , this gives the most direct picture of the energy levels of the light nuclei involved. Very recently however, in one or two laboratories, atten-tion has been turned to the non-resonant yield; for example, at the University of British Columbia the non-resonant gamma radiation from the reaction 0 1\u00C2\u00AE(p,y)F 1 7 has been studied, (WARREN 1954). Chapter III concerns i t s e l f with 12 an extension of this work to the case of C , describing a measurement of the angular distribution of the non-resonant 12 12 gamma radiation from the reaction C (p,yp')C II REACTIONS FROM THE FAST-NEUTRON IRRADIATION OF BORON (a) THE DETECTION OF NEUTRONS Among the methods used for detecting neutrons perhaps the most valuable depends on the interaction of neutrons with boron, specifically the interaction with the B^^ isotope, 10 7 B (n,<*)Li . This reaction has been intensively studied, (AJZENBERG 1952). The alpha particles produced are not mono-energetic since the reaction may go either to the ground state 7 of L i (Q=2.79 Mev), or to the f i r s t excited state (Q=2.31 Mev). For thermal neutron bombardment the partition ratio: i s 93:7 in favor of the excited state, (HANNA 1950; BICHSEL 1952). The cross section is 3,390 barns for thermal neutrons of 0.025 ev energy and varies as the reciprocal of the velocity for energies up to 10 kev, (HUGHES 1952). Neutron counting using this reaction involves the detection of the alpha particles emitted. This is generally accomplished by using boron trifluoride as the f i l l i n g gas of a proportional counter, since i t is the only gaseous compound of boron that is stable at ordinary temperatures and has the properties required of a counter gas. Proportional operation w i l l en-sure that processes leading to single electron ejection, such as gamma interactions, w i l l not produce pulses of appreciable size. This is an important requirement for a neutron detector since a neutron flux is usually accompanied by a large gamma ray flux. -3--4-The sensitivity of these counters depends strongly pn the energy of the primary neutrons due to the \" l / v \" variation of the neutron cross section of the boron 10 isotope. Various arrangements of paraffin moderator have been tried in order to get a counting rate that i s proportional to the number of incident neutrons and independent of their energy, (HANSON 1947). For this purpose a long cylindrical proportional counter i s used, the boron being introduced as boron trifluoride gas. Since natural boron contains 19% boron 10 isotope and 81% boron 11 isotope, often boron enriched in the boron 10 isotope is used for higher sensitivity to thermal neutrons. Such a fast neutron monitor is shown in f i g . 11. Fast neutrons incident on the face of the monitor are slowed to thermal energies by c o l l i -sions with protons in the paraffin. Some of them w i l l then d r i f t into the boron trifluoride counter and produce pulses. As long as the counter i s appreciably.longer than the mean free path of the neutrons in paraffin, the fraction of the neutrons that produces pulses does not vary greatly with the neutron energy. -5-(b) THE MEASUREMENT OF THE ENERGIES OF FAST NEUTRONS One of the most awkward practical problems in nuclear physics at theypresent time is theymeasurement of the energies of the neutrons emitted in nuclear reactions, and no really satisfactory technique has yet been devised for this purpose. The observation of recoil protons in cloud chambers (BONNER 1934) or in photographic emulsions (GIBSON 1948) has been used for most of the measurements that have been made on fast-neutron spectra. The laborious reading and evaluating of recoil proton tracks i s a serious drawback to this method. Since the distribution of recoil protons from the scattering of neutrons i s isotropic, i f the neutrons are monoenergetic there w i l l be equal numbers of protons in any unit energy intervals up to the energy of the neutrons. Differentia-tion of the recoil proton energy spectrum w i l l then indicate the neutron energy. Thus proportional counters f i l l e d with hydrogen or hydrogen compounds (TUNNICLIFFE 1952) may be used for measuring neutron energies by the measurement of the Y recoil proton energyjspectrum. Likewise a hydrogen-containing organic s c i n t i l l a t o r such: as stilbene may be used for this purpose (SEGEL 1954). However, while this i s a simpler and quicker technique, the accuracy of the method i s limited,due to the differentiation process required, and consequently i t is not too useful for the examination of a complex neutron spectrum. This drawback may be eliminated by using a coin-cidence counting scheme to select the direction of the recoils -6-that are accepted (BEGHIAN 1952; DRAPER 1954). \"Time of f l i g h t \" schemes have been used (JAMES 1951) for cases in which the advent of the neutron was coincident with a gamma ray. Such arrangements have a low sensitivity for reasonable resolution and involve complicated electronics. Threshold detectors (KLEMA 1948) are useful for discriminating between neutrons above and below a given energy but the discrimination levels available are limited and the applications of such counters are obviously relatively few. A more attractive possibility would be to measure the total disintegration energy resulting from a neutron-induced reaction giving rise to charged particles. A reaction giving only a single peak in the charged-particle energy spectrum would be preferable; but even i f the reaction were to proceed also via an excited state of the residual nucleus, so that two peaks could be obtained for monoenergetic neutrons, i t might well be possible to make an unambiguous energy assign-ment. Such a method avoids the tedious track measurement of the photographic-plate technique, does not involve the differentiation process required by proton-recoil counter measurements, and also gives the possibility of making coin-cidence measurements. Of course, i f several neutron groups are present, overlapping becomes serious and an unambiguous assignment of each peak to a particular neutron group cannot be -7-made. The method has been developed successfully for q 3 neutrons up to 1 Mev energy using the reaction He (n,p)T (BATCHELOR 1952). Another reaction which has been investi-20 17 gated i s Ne (n,<<)0 (FLACK 1953), but owing to the reaction 17 proceeding; to the f i r s t excited state of 0 with neutrons of energy above 3.5 Mev, the spectrum was complicated. 10 7 This investigation was undertaken to see i f theB (n,\u00C2\u00A90Li reaction might be suitable for the/ieasurement of neutron energies in this manner using the boron trifluoride proportional counters available in the laboratory. (c) NEUTRON CAPTURE REACTIONS WITH BORON TBIFLUORIDE (i) Possible Reactions used for the measurement of neutron energies the peaks from this process must be sharply defined and must stand out clearly above the background pulses from other reactions. Table 1 l i s t s the charged particle reactions that can occur when B^, and 19 F are bombarded with fast neutrons. Some of these may have a threshold higher than the energy of the neutrons being used^in which case they w i l l cause no d i f f i c u l t y . Only reactions lead-ing to the emission of charged particles other than electrons are considered because neutrons, gamma rays or beta rays w i l l not dissipate enough energy in the counter in the form of ioniza-tion to form pulses of appreciable size, so,for example, the If the fast neutron interaction with B 10 is to be is not l i s t e d . -8-Reaction Qm Cross Section Reference (Mev) for neutron source mentioned B i o Ln,p)Be .226 3 mb. fast pile neutrons EGGLER 1948 B i o (n,d)Be -4.35 39 mb. En=14 Mev. RIBE 1954 B 1 0 x 8 (n,t)Be .232 10 mb. fast pile neutrons CORNOG 1941 B10< Cn,He 3)Li 8 -15.8 B 1 0 Cn,oOLi7 2.79 200 b. Thermal neutrons HANNA 1950 B 1 0 7* Cn,oOLi 2.31 3100 b. Thermal neutrons HUGHES 1952 B 1 1 (n,d)Be -9.0 B 1 1 (n,t)Be -9.56 B 1 1 Cn,\u00C2\u00AB<)Li -6.63 10 mb. E n = 14 Mev. loc. c i t . F 1 9, (n,p)0 1 9 -3.72 14.5 mb. Be 9(d,n)B 1 0 JELLEY 1950 F19< (n,d)0 1 8 -5.75 F 1 9\u00C2\u00AB Cn,t)0 1 7 -7.53 \u00E2\u0080\u00A2 F 1 9 (n,He 3)N 1 7 -16.3 F 1 9, (n,oON16 -1.4 9 10 82.6 mb. Be (d,n)B JELLEY 1950 Table I. Neutron-induced reactions in B 1 U, B and F-La y i e l d - ing charged particles. Q-values are computed from mass values as given in SEGRE \"Experimental Nuclear Physics\" Vol. 1, /35\"3, -9-Some charged particles may be produced by the interaction of neutrons with nuclei in the walls of the counter. However these w i l l have traversed varying amounts of metal and hence wi l l exhibit no group structure. The branching ratio of the formation of lithium in the ground and f i r s t excited states has been measured as a function of the neutron energy from thermal energies up to 4 Mev. This ratio rises smoothly from 0.06 at thermal energies (BICHSEL 1951; CUER 1951} HANNA 1950; RHODES 1952) to 2.3 at En=1.9 Mev and f a l l s to 0.8 at En=2.6 Mev (PETREE 1951). The Q-value for the reaction going to the ground state of L i 7 is 2.79 Mev while that of the reaction going to the 478-kev ex-7 cited state of L i i s 2.31 Mev. Thus, although alpha particles of two energies are produced, the complexity of the spectrum is not too serious since the branching ratio is well known. The process has such a high cross section (3390 barns) for thermal neutrons that pulses from slow scattered neutrons above the 0.2-ev absorption peak of cadmium would be expected to mask a l l pulses under 2.5 Mev that may occur due to other reactions, in spite of any cadmium shielding. ( i i ) Experimental 1. The Boron Trifluoride Counters Two counters, one containing normal isotopic boron and the other containing boron enriched in B1^, were irradiated F i g . 1 . TARGET ft COUNTER ASSEMBLY Liquid Oxygen Beam H BF3 COUNTER Cadmium i 1 ^ Cap Paraffin L a n . 0.040\" Cadmium a HT JT To Amplifier D,0 -10-with 4.87-Mev neutrons from the H2(d,n)He3 reaction. The counters were supplied by Atomic Energy of Canada Ltd. The counter cathodes were of copper,' of 1\" O.D, and 7/8\" I.D. chemically cleaned before baking and f i l l i n g . The central tungsten wire anode was kept normally at ground potential. It was shielded at each end by grounded guard rings that limited the effective length of the counter to 33 cm. The enriched counter was f i l l e d to a pressure of 59.8 cm. of mercury at 24.9\u00C2\u00B0C with boron trifluoride obtained from a complex. It was enriched to 96% B 1 0. The unenriched counter was f i l l e d to a pressure of 60.0 cm. of mercury at 30.5\u00C2\u00B0C with ordinary boron tr i f l u o r i d e , and so contained B 1 0 and B 1 1 in the normal isotopic ratio of 19:81. The counters were operated at a potential of 1850 volts negative with a gas amplification of approximately 30. They were housed in l/16\"-thick brass tubes and wrapped in 0.040\" cadmium sheet to reduce the number of events due to slow scattered neutrons. 2. The Nautron Source The neutrons from the H2(d,n)He^ reaction, Q \u00C2\u00AB= 3.26 Mev, (TOLLESTRUP 1949) were generated by allowing 1.67-Mev deuter-ons accelerated by the Van de Graaff generator to impinge on a heavy-ice target about 30 kev thick. The target was pre-pared by opening the tap nearest to the heavy-water container (see f i g . 1.) and allowing the vapour from the water to reach Q = 2.3I 1000-_i LU 2 2 < X o (/) H Z Z> o o 500-Z 3 4 ENERGY (Mev) Fig\u00C2\u00BB S. D i f f e r e n t i a l puise height spectrum from the bombardment of enriched boron t r i f l u o r i d e with slow neutrons. Spectrum from unenriched counter i s s i m i l a r with smaller amplitude. -11-equilibrium pressure, (17 mm. of mercury at 20\u00C2\u00B0C) The tap was then closed and the second tap opened, allowing 3 5 cm. of heavy water vapour to be ejected through a very small hole onto a copper target-blank cooled by liquid oxygen. The process was repeated until the desired target thickness was obtained. The target thickness produced in this manner was measured experimentally. A similar heavy-ice target was la i d down on a thin target of calcium fluoride. When this was boroarded with protons, gamma rays were produced through the reaction F (p^JOO . The excitation function for this reaction was determined in the neigh-borhood of the accurately known 873.5-kev resonance. The increase in the incident proton energy required to produce resonance was considered to be equal to the energy lost by the protons in traversing the heavy-ice layer. The target thickness for 1670-kev deuterons could then be c a l -3 culated and was found to be 4 kev + 20% for every 5 cm. of deuterium vapour. The pulses from the counter were amplified and analyzed with a 30-channel Marconi \"kicksorter\". The neutron flux was monitored by another enriched counter embedded in paraffin in the manner described by Hanson et a l . (HANSON 1947). This monitor i s shown in f i g . 11. The target and counter arrange-ment are shown in f i g . 1. The target was kept at a position 3 0 P O O 2 0 , 0 0 0 _J UJ 2 Z < I o \ to | 1 0 , 0 0 0 o o .>*-B'\u00C2\u00B0(n,^)Li,# Q = 2.3I (En~THERMAL) \u00E2\u0080\u00A2 x 2 0 ' rB,0(n,p)'B.e'0 o , o 0 r e B (n,t)Be8 Q=0.35 rB,0(n,o<)Li7 Q = 2.79 (E^T HERNIAL) ^F,9(n,o<:)N . Q=-l.77 En=4.87 B ,0(n,^)Li7*Q = 2.3l E\u00E2\u0080\u009E=4.87 r B,0(n,oc)LiT Q = 2.79 En=4.87 2 3 4 5 6 7 8 E N E R G Y ( M e v ) gig. 5. Differential pulse height spectrum from the bombardment of enriched boron trifluoride with 4.87-Mev neutrons. -12-of 0\u00C2\u00B0 relative to the deuteron beam direction. ( i i i ) Results 1. Identification of the Reactions A 6 cm.-thick paraffin moderator was placed in front of the target and the cadmium cap was removed from the end of the counter to admit slow neutrons. The resultant slow-neutron pulse height spectrum obtained with the counter enriched in B^^ is shown in f i g . 2. Two groups only can be seen. The f i r s t , at an energy of 2.31 Mev corresponding to the reaction B^(n,o<)Li 7* for thermal neutrons (Q=2.31 Mev), was used to establish the energy calibration of the counter. The second and smaller group, at an energy of 2.79 Mev, corresponds to the ground state transition. The cadmium end plate was replaced and the paraffin removed. The two counters were then in turn irradiated with 4.87-Mev monoenergetic neutrons. The resultant pulse height distributions are shown in f i g . 3 and f i g . 4. Six groups were observed:-(1) at 2.31 Mev corresponding to the reaction B]\"^(n,oc)Li7;*e with slow neutrons, ( f i g . 3 and f i g . 4) (2) at 2.79 Mev corresponding to the reaction 10 7 B (n,oc)Li with slow neutrons, ( f i g . 3). (3) at 7.2 Mev corresponding to the reaction 10 7* B (n,oc)Li with 4.87-Mev neutrons, (f i g . 3). (4) at 7.7 Mev corresponding to the reaction 8 , 0 0 0 B ' \u00C2\u00B0 ( n \u00C2\u00BB L i 7 Q= 2.31 (En~THERMAL) F\"(n,<)N E n \u00C2\u00AB4.87 B'\u00C2\u00B0(n,p)Be'\u00C2\u00B0 B,6(n,t)Bett Q=0.35 E n=4.87 3 4 5 E N E R G Y ( M e v ) 8 F i g . 4. D i f f e r e n t i a l pulse height spectrum from the bombardment of unenriched boron t r i f l u o r i d e with 4.87-Mev neutrons. -13-10 7 B (n,\u00C2\u00AB0Li with 4.87-Mev neutrons. (Fig. 3.) (5) at 3.1 Mev corresponding to a reaction with a Q-value of -1.77 Mev. (Fig. 3 and Fig. 4) (6) at 5.22 Mev corresponding to a reaction with a Q-value of 0.35 Mev. (Fig. 3 and Fig. 4) The reacting nuclei from which groups (5) and (6) arose were identified from the relative yields of these reactions in the enriched and unenriched counters. Runs with the two counters were compared by normalization to the same number 10 11 19 of monitor counts. The two contained B , B and F in the ratios 96:19, 4:81, and 1:1 respectively. By a comparison of the observed yields to these ratios, group (5) was assigned to F 1 9 and group (6) to B 1 0. 19 2. The F Reaction Group (5) To determine the Q value more accurately, the region of the distributions from 2.5 Mev to 4.0 Mev was repeated with the \"kicksorter\" set at 60 kev per channel using both counters in turn. The contribution of the two slow neutron groups (1) and (2) was subtracted using data obtained using slow neutrons as for f i g . 2 and normalizing the amplitude of the B (n,o0Li slow neutron peaks before subtraction. The groups so obtained are shown i n f i g . 5. A Q-value of -1.77+ 0.15 Mev was obtained from this data. Group (5) 6 0 0 \u00E2\u0080\u0094' 1\u00E2\u0080\u0094 *\u00E2\u0080\u0094//\u00E2\u0080\u00A2 1 ' \u00C2\u00AB\u00E2\u0080\u0094 3 O 3.5 4 . C T ' 5 .0 5.5 6 . 0 E N E R G Y ( M e v ) g i g ' 5\u00C2\u00AB D i f f e r e n t i a l pulse height spectra with the slow neutron background\" subtracted. ^=4.87 1,11 -14-i s probably due to the reaction F 1 9(n,o<)N 1 6, Q m = -1.4 Mev (AJZENBERG 1952), since the Q-values of other neutron 19 induced reactions in F are too low to f i t the data. Using the value adopted by Ajzenberg and Lauritzen for the mass of N 1 6, (N 1 6-0 1 6 - 10.3 + 0.5 Mev) the measured Q-value corresponds to alpha emission to an ex-1 c cited state in N at 0.37 Mev above the ground level. An excited state at 0.3 Mev has been reported (WYLY 1949) X 5 X 6 based on a study of the N (d,p)N reaction. There is evidence of a group at 0.34 + 0.12 Mev above the main group, which could correspond to the ground state transi-tion. The yield for this group was about 4% of that for 19 the main F peak. This measurement bears on the mass 16 of N for which the experimentally determined values have shown large discrepancies. The results of other measure-ments of the F^(n,\u00C2\u00AB0N'^ reaction energy are: Q = 0.73 + 0.25 N 1 6 - 0 1 6 = 9.64+0.28 Mev (BLEULER 1947) Q - -1.2 + 0.9 N 1 6 - 0 1 6 =10.1+0.9 Mev (JELLEY 1950) The present measurement gives 16 1 R Q - 1.43 + 0.15 N - 0 1 6 = 10.3 + 0.2 Mev. 1 fi 16 in good agreement with the value N -0 =10.3+ 0.5 Mev adopted by Ajzenberg and Lauritzen on the basis of a l l previous 16 data bearing on the mass of N . -15-3\u00C2\u00BB The B 1 0 Reaction Group (6) A similar expanded pulse height distribution was taken in the region of 5 Mev to 6 Mev and a Q-value for the 10 B reaction of 0.35 + 0.20 fias obtained, ( f i g . 5) How-ever in the case of group (6) %he reaction i s not uniquely identified since their are two reactions which could f i t the experimental Q-value. B 1 0(n,p)Be 1 0 Qm - 0.226 Mev (AJZENBERG 1952,EGGLER 1948) B 1 0(n,t)Be 8 Qm = 0.232 Mev (AJZENBERG 1952,PERKIN 1951 RIBE 1954). Be 8\u00E2\u0080\u0094>2\u00C2\u00BB< Q - 0.095 Mev Thetbtal energy released in (n,t) reaction i s 0.33 10 Mev. Other neutron-induced reactions in B yielding charged particles have Q-values that do not agree with the position of this group. (iv) Conclusions Several groups are present in the pulse height distribution of boron trifluoride bombarded with neutrons in the energy region of 5 Mev. These have yields of the same order of magnitude, so that the spectrum,, is too com-plicated to be used for the measurement of neutron energies i f several neutron energy groups are present. The high 10 7 energy peaks from the B (n,<)Li reaction could indicate the energy of monoenergetic neutrons but the low yield and -16-the spreading of the peaks due to wall effect reduce the accuracy so that this method i s not suitable for the measure-ment of fast-neutron energies from the analysis of the pulse height spectra observed. The pulse group at 5.22 Mev i s due to either or both of the reactions B 1 0(n,p)Be 1 0 and B 1 0(n,t)Be 8. The groups at 3.10 Mev and 3.44 Mev are due to the reaction 19 i fi F (n,o<)N . The Q-value for the transition to the ground state i s -1.43 + 0.15 Mev, leading to a value for the mass of the N 1 6 nucleus of 16.01110 + 00020 MU. (d) THE DETECTION AND CROSS SECTION OF THE REACTION B i : L(n,o0Li (i) Previous Evidence for the Reaction Among the possible reactions l i s t e d in Table 1 11 8 the reaction B (n,<*)Li i s of especial interest in view of 8 i t s possible use to produce L i for neutrino experiments. This reaction, which could only be produced by bombarding B 1* with very fast neutrons, was f i r s t proposed by Burcham and Dee (LAWRANCE 1939) as a possible explanation of a number of delayed pulses that they had observed in a boron trifluoride proportional counter after i t had been exposed to a flux of 7 8 neutrons from the L i (d,n)Be reaction. They suggested the sequence: Fig. 7. . Two examples of hammer tracks produced a by cosmic rays in nuclear emulsions-i \u00E2\u0080\u00A2 From Powell and Occhialini, \"Nuclear Physics in Photographs\". -17-B 1 1 + n * L i 8 + o< + Qx Li\u00C2\u00B0 * Be + P\" + V> + Q 2 T| - 0.89 sec;. 2\u00C2\u00B0< + Q3 o L i i s an emitter of \"delayed alpha particles\" or what were known in the terminology of classical radioactivity as \"long range alpha particles\". These arise from excited o states of the daughter nucleus, Be\u00C2\u00B0, following the beta decay 8 ft of L i . The Gamow exponent for the decay of Be\u00C2\u00B0 into two alpha particles i s low due to the low nuclear charge, hence the breakup occurs very rapidly. The theoretical estimate of the lifetime of the ground state i s 10~ 1 6 seconds (WHEELER 1941). However, i t is found that 98% of the beta transitions go to excited states of Be R with excitation energies of 3 Mev or more (AJZENBERG 1952). At these energies the barrier factor i s lower and the excited states w i l l have even shorter lifetimes \u00E2\u0080\u009421 of the order of 10 seconds. The observed h a l f - l i f e of the alpha activity w i l l , therefore, be that of the relatively slow Q beta decay of the parent L i with which i t i s in equilibrium. This h a l f - l i f e i s 0.89 seconds (HUGHES 1947) which is sufficiently long to have made i t possible for Burcham and Dee to have observed delayed alpha pulses in their counter i f Q Li\u00C2\u00B0 had been formed in i t . Miss Lawrance irradiated a boron t r i f l u o r i d e - f i l l e d counter intermittently with fast neutrons and recorded a l l delayed pulses photographically, (LAWBANCE -18-1939). The h a l f - l i f e of the delayed activity was found to agree with that of L i 8 . The activity occurred i f neutrons from L i 7 (d,n)Be8 were used but not with those from Be 9(d,n)B 1 0, indicating that the threshold for the neutron induced reaction involved lay somewhere between the 4.5 Mev of the neutrons from the latter process and the 13.6 Mev maximum energy of the neutrons from the former. The threshold of the B^(n, L i 8 + He 4 + H1 . Hammer tracks of the same nature have been produced (PICKUP 1948) by bombarding boron-loaded emulsions with neutrons from 7 8 8 L i (d,n)Be . He attributes these to the formation of L i by 11 8 the B (n,\u00C2\u00AB<)Li reaction but makes no attempt to evaluate the cross section for the process, g L i i s useful in the study of the neutrino hypothesis. Attempts to determine whether or not the existence of a neutrino i s necessary for compliance with the law of conservation of linear momentum in beta-decay processes involve the measurement of the recoil momenta of residual nuclei, a d i f f i c u l t thing to do accurately because of their short range. The alpha-particle 8 8 breakup of Be immediately following the beta decay of L i presents a way of alleviating this d i f f i c u l t y , since the alphas have reasonably large range yet their paths are influenced by g the momentum of the Be from which they originate. The breakup 8 occurs rapidly enough so that the Be w i l l not have lost i t s recoil momentum. This unique situation was exploited by Christy and collaborators. (CHRISTY 1948) If the nucleus were at rest when the breakup occurred the two alpha tracks would l i e on the same line in a cloud chamber. However, the disintegra-tion occurs while the nucleus i s recoiling. If fission occurs -20-along the path of r e c o i l , the forward alpha track w i l l be observed to have up to 20% greater range than the other. If i t occurs at right angles to the recoil the ranges w i l l be equal but the directions of the two alpha particles w i l l deviate from 180\u00C2\u00B0 relative to each other by as much as 6\u00C2\u00B0. Both effects were observed but the main problem was the g production and introduction of the L i into the cloud cham-g ber. Christy f i r s t attempted to produce L i for his experi-11 8 ment by means of the B (n,o<)Li reaction. The cloud chamber was f i l l e d with methyl borate and irradiated with fast neutrons. No evidence was found for the formation of a 11 L i from the neutron bombardment of B and so Christy et 7 8 al were forced to use the reaction L i (d,p) L i which neces-sitated the use of a complicated mechanical arrangement for introducing the solid lithium hydroxide target into the cloud chamber after each bombardment. They concluded that the cross section for the B ^ reaction must be very small, i f , indeed,the process occurred at a l l . Since this reaction would permit the production of L i 8 directly in the cloud chamber gas, i t seemed of interest to corroborate Miss Law-rance's work and i f possible to measure the cross section. ( i i ) Theoretical Estimate of the Order of Magnitude of the Cross Section In the Bohr compound-nucleus treatment of neutron-induced reactions involving processes in addition to elastic -21-scattering, the interaction is considered in two steps: 1. The c o l l i s i o n i t s e l f in which the neutron enters the nucleus. 2. The competitive breakup of the compound nucleus into the f i n a l reaction products which may be alphas, protons, neutrons, etc., together with electromagne-t i c radiation which may s t i l l de-excite a nucleus when the excitation energy i s inadequate for par-ti c l e s to be emitted. The cross section for any given reaction (n,x) may be written as the product of two terms: k \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 % ^ E*the excitation energy of the residual nucleus. -27-petition would lower the value of (r*(n,<<) slightly by i n -creasing the value of -21 /\"x but the effect i s small due to the comparatively large value of J \" ^ . The value of the ratio: was found to be 0.135 resulting in a cross section for the 11 8 B (n, <<)Li reaction of about 80 millibarns. It must be emphasized that the above calculation i s only a very rough approximation. The values for \"C\" and \"a\" in the formula for the energy level density were inferred from values that had been calculated for odd values of the mass number. Not enough material is available to make a calculation of these constants for even mass numbers and i t i s expected that there would be some difference. The formula used for the energy level density is an extremely rough approximation to the general features of the densities of nuclear levels based on a s t a t i s t i c a l picture of the nucleus and w i l l be less accurate in regions which contain only a few levels as in the case of light nuclei. It i s g lik e l y that the properties of individual levels of L i w i l l play a v i t a l role in the value of the cross section for alpha particle emission since only a relatively small number of the lower levels are effective. Since these properties are not known their effects have not been considered and the continuum theory has been applied instead. The assump--28-tions on which the continuum theory i s based do not agree closely with experimental facts for such light nuclei as are involved in this problem, particularly when levels of low excitation are responsible for a major part of the cross section. Hence any calculation on this basis must not be regarded as anything more than a reasonable guess as to what order of magnitude may be expected for the cross section, which may be of value in designing the experimental scheme for the measurement thereof. ( i i i ) The Beta Decay of L i 8 L i 8 Be 8* + (3\" +13 Mev 2 \u00C2\u00A9< + 3 Mev The h a l f - l i f e of this activity has been measured to be 0.89 s e c , (HUGHES 1947). The beta spectrum indicates that about 90% of the transitions go to a wide state about 3 Mev above the ground level, 10% to states of the order of 10 Mev and 13 Mev, less than 5% to a 4.5 Mev state and less than 2% to the ground state. (AJZENBERG 1952, HORNYAK 1950). (See f i g . 7). The total alpha energy spectrum exhibits a peak that has been found by various observers to l i e anywhere between 2.4 Mev and 3.3. Mev with a width from 0.8 Mev to 1.26 Mev, (FOWLER 1937, RUMBAUGH 1938). The alpha energy width at this peak corresponds to a lifetime of the order of 10~ 2 1 sec. for the excited state at 3 Mev. Gamma decay to the ground state i s relatively slow so that breakup into two alpha particles occurs in effectively every \u00E2\u0080\u0094 2 9 \u00E2\u0080\u0094 case. The high-energy alpha particles observed seem insuf-ficient to f i t the beta spectrum, indicating that some gamma radiation occurs, (LI 1951). There is some evidence g of gamma radiation from the 4.9-Mev state of Be . Evidently there is some selection rule that forbids Q 8 the beta transition from L i to Be in the ground state, indicating that the angular momenta of the ground states of the two nuclei differ by at least two units. It has been suggested that this condition i s f u l f i l l e d i f there exists a level, 3 Mev above the ^SQ ground state of Q - g O Be and i f the normal level of L i is a P2 state, (WIGNER 1936). As previously mentioned, this beta-decay reaction lends i t s e l f peculiarly well to the study of the neutrino hypothesis. The angular relations and the relative energies g of the two alpha particles depend on the Be recoil momentum following the beta decay. Christy measured these quantities to determine which of the different interactions of the Fermi theory were involved, (CHRISTY 1947). (iv) Experimental 1. The Neutron Source The neutrons were produced by a beam of 50-kev deuterons impinging on a target of tritium adsorbed TRAP pump CADmiunn COUNTER Fig. 9. Schematic diagram of the 50 kilo-volt accelerator. - 30 -in zirconium, \u00C2\u00AB 3 + H 2 * He4 + n 1 + 17.6 Mev. With a deuteron energy of 50 Kev the energy of the neutrons emitted in the forward direction is 14.1 Mev. At this low bombarding energy the angular distribution of the neutrons is isotropic. (ALLAN 1950). The accelerator is shown in f i g . 9. A pressure of 2 x 10 mm. of Mercury was maintained by a D i s t i l l a t i o n Products, 275 l i t r e s per s e c , o i l diffusion pump backed by a Cenco Megavac forepump and a mercury diffusion pump backed by a Welch Duoseal forepump. Liquid air traps were introduced immediately above the diffusion pumps and on the magnet box to condense o i l and mercury vapours. The 50 kv. accelerating voltage was supplied by a f i l t e r e d half-wave r e c t i f i e r . A l l accessory power supplies and ele c t r i c a l appliances were mounted at the high potential. 110 volt AC power was supplied to them through an isolation trans-former . The deuterium ions were produced in a radio fre -quency ion source (THONEMANN 1948). The flow of gas to the ion source was controlled by means of a palladium leak. The discharge was excited by an amplifier oscillator with i t s tank c o i l wound directly around the discharge tube. - 31 -The ions were extracted by a DC voltage of up to 3 kv. applied between a tungsten probe at the top of the discharge tube and an aluminum extractor cone at the bottom. The extractor was covered by a glass sheath and a quartz tube to reduce the area of metal exposed to the discharge and so minimize the recombination of ions. In addition, the static charge accumulating on the quartz tube served to shape the electric f i e l d above the extraction canal in such a manner as to attain the maximum extraction efficiency. The ions passed through a canal 2 mm. in diameter and 1.9 cm. long d r i l l e d through the extractor. The potential in the discharge i s similar to that in a DC discharge with most of the potential drop occurring near the cathode. Since the majority of the ions are formed above this region they w i l l a l l have approximately the same energy. The ion beam emerging from the extractor was focussed by means of electrostatic lenses constructed from two highly polished steel tubes of 1\" inside diameter, each with corona rings. The lens gaps were approximately 1/8\u00E2\u0080\u00A2'. The bottom lens section was grounded. The upper one could be adjusted from 0 to 18 kv. negative relative to the 50 kv. terminal. The beam was deflected through 90\u00C2\u00B0 and separated -32-into i t s various components by a magnet constructed by Mr. J.S. Kirkaldy, (KIRKALDY 1951). Because of the high cost of tritium targets, great care was taken to prevent contam-ination of the target surface. The low energy of the deu-terons made i t doubly important to prevent the condensation of any o i l vapours on the target since a thin film of o i l could stop the 50 Kev deuterons from reaching the tritium. To ensure a high pumping speed at the target end of the vacuum system a mercury diffusion pump was installed between the magnet and the target. The beam was made to pass through a channel \ \" in diameter and 11\" long through a liquid air trap before reaching the target. A suppressor ring kept at -270 volts relative to ground prevented electrons knocked off the walls of this channel by the beam from reaching the target, and also kept secondary electrons from the target from escaping and being recorded as additional beam current. The target, supplied by the Atomic Energy of Canada Ltd., consisted of tritium nuclei adsorbed in zirconium metal on a tungsten backing. This was clamped to a thick water-cooled copper plate. The yield was of the order of 10 neutrons per sec. with a deuteron beam current of 100 microamperes. 2,000 microampere-hours resulted in a reduction of the neutron flux by a factor of about 3, and only a very slight darkening of the surface of the target. 2. The Boron Trifluoride Counter A long boron trifluoride counter was placed jjjg. 10. Construction of the boron trifluoride counter. -33-adjacent to the chamber at 0\u00C2\u00B0 relative to the beam direction and with the axis of the counter at right angles to the beam direction. It was supported on sponge rubber grommets inside a brass cylinder of 2 J \" diameter and 1/16\" Wall thickness. The whole tube was covered with a double layer of 0.020\" cadmium sheet to keep down the thermal neutron flux. (See f i g . 10). The counter was constructed and f i l l e d at the Chalk River laboratories of Atomic Energy of Canada under the auspices of Mr. Fowler. The cathode was of copper, of 1\" outside diameter and 7/8\" inside diameter, chemically cleaned before baking and f i l l i n g . The central wire anode was of tungsten and was kept normally at ground potential. Grounded guard rings at the ends of the counter shielded the anode and limited the effective volume of the counter to a length of 33 cm. An aquadag guard ring was painted on each of the glass ends of the counter around the anode lead and connected to the ground lead from the inner guard ring to prevent noise pulses due to electrical leakage across the surface of the glass from the cathode. The cathode was held at 1900 volts negative. The counter was f i l l e d to a pressure of 60.0 cms. of mercury at 30.5\u00C2\u00B0 C with boron trifluoride from a cylinder. This contained B 1^ and in the normal isotopic ratio of 18.6:81.4. H.T.- 1900 V.-0U1 PUT PARAFFIN PARAFFIN PARAFFIN PARAFFIN 22\"-.050\" SHEET HON. ROfcA^ \u00C2\u00A3 ALUIY1INUIY1. CADmium CAP. BORON TR.IFLOUR.IDE COUNTER. F i g . 11. Construction of the neutron monitor. -34-3. The Monitor A shielded \"long counter\" was used to monitor the neutron flux. (HANSON 1947) See f i g . 11. Neutrons entering the outer layer of paraffin are degraded and are adsorbed by the \ n layer of borax so that they are prevented from reaching the counter. In this way the effect of scattered neutron radiation from the walls and the floor is reduced. A cadmium cap i s placed over the end of the counter to protect i t from direct thermal neutron radiation. Fast neutrons entering the inner cylinder of paraffin are thermalized by collisions with protons and d r i f t into the 10 7 counter where they produce pulses through the B (n, \u00C2\u00A9<)Li reaction. In order that the counter should have the same efficiency for neutrons of a l l energies, i t should be longer than the mean free path in paraffin of the highest energy neutrons to be counted. This counter i s f i l l e d with boron trifluoride enriched to 96% of the B 1 0 isotope. The monitor sensitivity was calibrated using a 51 millicurie radium-5 beryllium source yielding 7.8 x 10 neutrons per sec. The monitor gave 4.6 counts for every 10\u00C2\u00AE neutrons emitted by the source at a position on the axis of the monitor 150 cm. from i t s face. This i s within 5% of the published efficiency for this type of monitor. From the variation of sensitivity of the counter with neutron energy as given by Hanson and McKibben, i t is estimated that the sensitivity for 14 Mev neutrons i s about 80% of the mean sensitivity for radium-Fig. 12. Circuit diagram of the synchronizer. beryllium neutrons. Therefore 3.7 counts are expected for 6 every 10 neutrons emitted isotropically from the Hf*(d,n)He^ reaction at a point 150 cm. from the face of the monitor. 4. Beam Modulation 11 8 In order that the B (n,<*)Li events might be observed against the background of pulses from the _ _ 7 B (n,\u00C2\u00AB<)Li reaction, i t was necessary to turn off the deuteron beam that was producing the neutrons and look for delayed pulses that could be attributed to the break-8 up of Be . The counter output had to be turned on short-ly after the beam was shut off and turned off again before the start of the next period of irradiation. This cyclic sequence of events was controlled by an electronic synchronizer. (See f i g . 12) The period of the cycle (7.65 sec.) was governed by a slow multivibrator. A relay in one of the plate circuits operated a switch that removed the extraction voltage from the ion source, thus stopping the deuteron beam during one half of the cycle. This method was used in preference to deflecting the beam since a deuteron beam w i l l produce a few neut-rons from any stop that i t may strike. A univibrator stage triggered by the closing of this relay introduced a short delay (0.09 sec.) before triggering another univibrator that opened a shorting switch across the POTENTIAL ON PLATE OF TUBE \" T l . \" -POTENTIAL OF PLATE OF TUBE \"T3>% POTENTlAL ON PLATE OF TUBE \" T5.\"-POTENT1AL OF COrfimON CATHODE OF TUBES \" V4.\"\"V5 .\"\"VG.\" CmiLLER. SWEEP ) \u00E2\u0080\u0094BEAM OFF (4.15 SEC.)- BEWYION (3.50'5EO-|~ BEhttl OFF o / $.551 < \ \ \ s j 1.345 C a 4 P I.G33 N'3 J='/2--2.211 F i g . 18. Energy l e v e l diagram f o r C^ 2+ p. T 1 1 1 r Q I 1 I 1 i i i i i i i _ 1.0 l.l 1.4 1.6 1.8 2.0 2.1 IA 2.5 2.8 EpCmEV.) F i g . 19. The t h e o r e t i c a l e x c i t a t i o n function f o r the re a c t i o n C 1 2 ( p , y p ' ) C 1 2 . I l l THE ANGULAR DISTRIBUTION OF NON-RESONANT GAMMA RADIATION FROM THE BOMBARDMENT OF CARBON BY PROTONS (1) The Reaction C 1 2 ( p , r p f ) C 1 2 When C is bombarded with protons of energy 1.3 to 2.7 Mev, gamma rays are observed whose energy varies with the proton energy, (WOODBURY 1954). Only radiative capture or elastic scattering are possible at such low bombarding energies. Measurement of the gamma ray energy shows that the radiation i s due to capture of the proton with the 13 emission of a gamma ray to the 2.369-Mev level of N , (AJZENBERG 1952), as shown in the energy level diagram of f i g . 18. Since this i s a virtual level, i t decays 12 predominantly by proton emission to the ground state of C , C 1 2 + H 1 * ( N 1 3 ) \u00E2\u0080\u0094 ^ N 1 3 * + y 12 12 which may be written more briefly as C (p,yp')C Woodbury et a l . measured the excitation function for this reaction at angles of 0\u00C2\u00B0 and 90\u00C2\u00B0. They found that the radiation was composed of a slowly-varying non-resonant component and a component that was resonant at 1.7 Mev. At 0\u00C2\u00B0 the non-resonant component was found to have approxi-mately zero yield. It is impossible to account for the -44--45-lack of yield at 0\u00C2\u00B0 i f a compound nucleus i s formed with definite angular momentum and parity; and further the shape of the excitation curve indicates that a resonant compound state i s not formed. It was pointed out by Christy that this corresponds to a single stage reaction of capture from the continuum. It i s suggested that the zero yield at 0\u00C2\u00B0 may be explained by a reaction in which p-wave protons are captured into an S state of a proton 12 around C in a radiative process that conserves the z component of the spin,i*no \"spin f l i p \" . This assumption could also account for the interference near the 1.7 Mev resonance. The expected angular distribution for such a 2 reaction is sin 0. The theoretical excitation function based on this picture i s plotted in f i g . 19 with parameters chosen to f i t the data obtained by Woodbury et a l . The measurements at 0\u00C2\u00B0 and 90\u00C2\u00B0 alone cannot be re-o garded as demonstrating that the distribution is sin 6. The object of this experiment was to measure the relative yield at several angles in order to determine the validity of this assumption. ( i i ) Experimental 1. Target Preparation Gold was chosen for the target backing due to it s low gamma ray background under proton bombardment and BE/\m GLASS TUBE- I \ LUCITE INSULATING RING -TARGET ROTATING HANDLE \ PROTON \BEAm \ m A Q N E T BOX - PLATINUM FOIL QUART! F0CU5SING PLATE TO Pump TARGET-NICKEL/ TUBE l.C.C. BULB BEN7ENE Fig. 2 0 . The experimental arrangement for observing the gamma rays from the' proton bombardment of -carbon. -46-i t s good thermal conductivity. A piece of gold 1.5 cm. square and 0.115 cm. thick, was clamped to the bottom of a liquid-air container. This target mount, shown in f i g . 20, could be turned relative to the target chamber by means of a handle at the top. The gold was electrolytically etched before use. Targets were prepared by freezing benzene vapour onto the gold, using the procedure described on page 10, for the preparation of heavy-ice targets. 2. Beam-Current Measurement The beam was focussed on a quartz focussing plate. The plate was then removed, allowing the beam to pass through small circular apertures in 1-mm. gold stops at either end of a brass pipe leading to the target chamber. This brass pipe was lined with thin platinum f o i l to reduce background from scattered protons striking the brass. The target and the liquid-air container on which i t was mounted, were insulated from the target chamber by a lucite ring so that the beam current incident on the target could be measured. A small battery held the target at a potential of 300 volts positive relative to the target chamber to prevent spurious beam-current measurement due to secondary-electron emission from the target. The position of the target could be adjusted slightly for f i n a l maximizing of the beam current. PHOTO CATHODE =t=8>u.F. -H.T. O-950 V. O+100V. OOUTPUT \f\u00E2\u0080\u00940 OUTPUT Fig. 21. The photomultiplier and head amplifier ci r c u i t . No. I. CRYSTAL Em. (olQ2 V - ALUMINUM CAN 4 ITIgO.. POWDER. -LUC1TE PLATE \u00E2\u0080\u00A2CLAmP SUNG. LIGHT TRAP V//////////\u00C2\u00A3 DYN0DE5. Fig. 22. The sodium iodide crystal mount. -47-Beam currents of 3 to 5 microamperes were used. The i n -tegrated current during each run was measured using an electronic integrator, (EDWARDS 1951). 3. The Scintillation Counter The gamma rays were detected by a thallium activated sodium iodide crystal, 2 inches long and 1.75 inches in diameter, used in conjunction with an EMI 6262 photomultiplier (GRIFFITHS 1953). This counter was mounted on a movable arm that could be rotated about a vertical shaft rigidly attached to the bottom of the target chamber in a position directly below the target. The out-put of the photomultiplier was fed via a cathode follower to a Northern Electric, type 1444, linear amplifier and thence to a biased amplifier and a pulse height analyzer. A crystal diode pulse limiter was added to limit the amp-litude of cosmic ray pulses corresponding to energy losses of over 7 Mev in the sodium iodide crystal. A soft-iron shield was placed around the photomultiplier to prevent gain changes due to the magnetic f i e l d of the beam-deflecting magnet. A second sodium iodide s c i n t i l l a t i o n counter was attached to the target chamber in a fixed position relative to the proton beam. This counter was used as a monitor -48-during the angular distribution measurements. Its output pulse was fed to an Atomic Instruments Company linear am-p l i f i e r and discriminator, type 204C. The discriminator was set to exclude a l l pulses below those from the 1.28 22 Mev peak from Na . Its output was fed to a scaler. Lead shielding shown in Plate 1 was placed around the counters to shield them from the X-ray background of the Van de Graaff generator. 4. Procedure The energy calibration and linearity check of the crystal were made by observing the positions of the 22 0.51 Mev and the 1.28 Mev radiation from a Na source and the 2.62 Mev radiation from a RaTh source on the pulse height distribution as shown by the kicksorter. The gamma ray of interest was identified by i t s characteristic photopeak. If the f i n a l state in the capture radiation 13 reaction i s indeed the 2.369 Mev level of N , the expected variation of the gamma ray energy with proton energy w i l l be given by: 12 Ey - 1 3 (Ep - 0.45) Fig. 23 shows the curves obtained for three values of the proton energy and clearly demonstrates this dependence of the gamma ray energy on the bombarding energy. G A M M A E N E R G Y ( M \u00C2\u00AB v ) Fig. 24. The.gamma ray spectrum from-the proton bombardment of carbon. . E.~ 1,325 kev. -49-It was'possible to observe the gamma ray when i t s energy was less than about 0.65 Mev due to the background 13 of annihilation radiation from the positron decay of N (T| = 10.1 minutes) formed by the C 1 2 (p, Y )N 1 3 reaction. The amount of residual annihilation radiation that remained after the proton beam had been turned off was not sufficient to agree with the activity during bombardment. This sug-gested that a shorter-lived positron emitter was being 17 formed as well. F emits positrons with a h a l f - l i f e of 17 66 seconds. F could have been formed by the reaction 0l6(p > Y i f 0*** were present in the target as an im-purity from water in the benzene or from the condensation of water vapour entering through leaks in the vacuum system. The presence of some activity when the back of the gold plate was bombarded suggests that the oxygen was due to the latter source. The presence of a 1.5 Mev gamma ray corresponding to decay to the 0.53 Mev excited state of F ^ supports the view that was the impurity involved. 17 The 0.53 Mev cascade gamma ray in F would augment the annihilation peak during the proton bombardment. The ^ target was allowed to warm up so that the benzene and water evaporated leaving only a deposit of carbon from benzene that had been cracked by the beam. This l e f t sufficient carbon for a target and reduced the annihila-tion radiation background relative to the non-resonant -50-gamma ray yield. The relative yields of the non-resonant gamma radiation were measured at angles of 0\u00C2\u00B0, 30\u00C2\u00B0, 45\u00C2\u00B0, 60\u00C2\u00B0 and 90\u00C2\u00B0 relative to the incident beam for proton energies of 1,460 and 1,680 kev. These runs were normalized to the same number of counts from the monitor, which operated on the higher energy gamma rays from the capture reaction C (p,J^)N . It was d i f f i c u l t to obtain a clear separation of the photopeak of the variable energy gamma ray from the Compton distribution and from back-ground radiation from the induced positron activity. The procedure followed was to regard the front part of the Compton distribution as background and to draw in the background as shown in f i g . 23 following the criterion that the photopeak corrected for background should be symmetrical. ( i i i ) Results 1. Solid Angle Corrections Due to the f i n i t e solid angle subtended at the target by the detector, the observed counting rate was in each case an average of the gamma ray intensity over a range of angles rather than being proportional to the inten-sity at the centre of the crystal. Corrections for this effect were computed on the basis of the assumption that the L. angular dependence of the intensity was of the form A + Bsin 6 where 6 is the angle between the direction of Fig. 25. Diagram, showing the method of calculating the solid angle correction. \" r \" i s smaller than the distance from the target to the centre of the crystal, to allow for the inverse-square-law variation of the gamma ray intensity. -51-the proton beam and the direction in which the gamma ray is emitted. On this basis the expected yield integrated over the whole crystal was determined and compared to that which would have been obtained i f the intensity had been the same everywhere as at the centre of the crystal. At the 0\u00C2\u00B0 position the crystal solid angle was approximated by the angle subtended at the target by a spherical cap covering the cross section of the crystal at an effective mean distance of \"r \" cm. from the target. The measured yield was then rep-resented by the integral of Bsin 0 du> over the spherical cap plus a small contribution due to the isotropic component. This integral i s given by: where j is the angle subtended at the target by \" R \", the radius of the crystal. For values of 9 other than 0\u00C2\u00B0 the value of the integral was determined by numerical integration. The surface of a vertical cylinder of radius \" r \" cm. cutting the crystal was used as the integration surface instead of the spherical cap. (See f i g . 25) This was cut into vertical strips each of width R7/5. 9 was assumed to be constant over each str i p . The integral was then replaced by the summation: B ^ s i n 2 \u00C2\u00A9 - *Ok <** where t \ u j i s the solid angle subtended at the target by each -52-strip is given by: X (X 2 + R 2 ) f and: X = [ R 2 - r 2 (9 - f r49 is the width of each str i p . 2. The Angular Distribution Table 2 shows the solid angle corrections made for the angular distribution at a proton energy of 1,680 kev. An additional correction was made for the absorption of the e Area under photopeak Correction Relative Intensity 0\u00C2\u00B0 62 -45 17 30\u00C2\u00B0 206 -19 187 45\u00C2\u00B0 351 351 60\u00C2\u00B0 561 11 572 90\u00C2\u00B0 721 22 743 Table 2. Solid angle corrections for the angular distribution of the non-resonant gamma rays at an effective proton energy of 1,580 kev. -53-gamma rays in the gold target backing since different thick-nesses of gold were traversed at different angles of observa-tion. The gamma ray interaction cross sections for these corrections were obtained from the work of Davisson and Evans, (DAVISSON 1952). The corrected distributions for the two proton energies are shown in the graphs of f i g . 26. The effective proton energy, taking into consideration the target thickness, was computed from the position of the gamma ray photopeak and the theoretically predicted energy given by the relation: 12 E y (E n - 0.456) * 13 P This relation has been found to agree with experimental results, (WOODBURY 1954). The observed angular distribu-tions are given by: 1(6) = 0^;oo + sin 29 at E p = 1,365 kev 1(9) = 0.021^02 + sin 20 at E p = 1,580 kev These may be compared with the results predicted on the basis of the theory put forward by Christy and Woodbury to explain the properties of the non-resonant radiation. The theoretical expression for the intensity i s : , [(f)Mr4Egos4 + (f)Mr-s\u00C2\u00BB\u00C2\u00BB,SL7 Sm'g -54-Th is leads to the angular distributions: 1(e) = 0.003 + sin 2 9 at E p = 1,365 kev 1(9) = 0.016 + sin 2 9 at E p = 1,580 kev using the values of the parameters: P = 0.060 Mev = the proton resonance width at half maximum. E r = 1.70 Mev = the proton resonance energy. 6 = 2.4\u00C2\u00B0 = the relative phase of the resonant and the non-resonant gamma rays at E = E r. A = 1.1 = the relative amplitude of the resonant and the non-resonant gamma rays at E = E r. which gave the best f i t to Woodbury's measured excitation function at 0\u00C2\u00B0 and 90\u00C2\u00B0. The agreement with the observed angular distribution is within the experimental errors. (iv) Conclusions o The sin 9 angular dependence of the non-resonant gamma radiation assumed by Woodbury has been shown to be correct, in agreement with the explanation of the reaction as a single stage process of p-wave proton capture from the continuum into an S state of a proton about C ^ in a radia-tive transition involving no \"spin f l i p \" . -55-APPENDIX I THE USE OF AN ELECTRODELESS RADIO-FREQUENCY DISCHARGE AS A SOURCE OF DOUBLY CHARGED HELIUM IONS 1. General The classical nuclear disintegration experiments of Rutherford and others were a l l performed using alpha particles from natural radioactive elements and varying the energies by means of absorbing f o i l s . These sources give a high energy dispersion and a low yield. A one microampere beam of doubly charged helium ions would be the equivalent in yield of about two hundred curies of a natural alpha particle emitter. Singly charged helium ions are relatively easy to obtain in quantity. However i f doubly charged ions could be produced i t would double the maximum energy available from any given accelerator. Moreover, alpha bombardment of light nuclei should yield angular distribution data particularly easy to reduce and unambiguous in interpretation since the spin of the ingoing particle is zero. Appreciable beam currents of doubly charged helium ions have been achieved using arc sources. On the U.B.C. Van de Graaff generator a radio-frequency electrodeless discharge type of ion source is used. This type has been chosen for its low power consumption, low energy spread of the ions, simplicity, long l i f e , and (for hydrogen) i t s high percen-tage of atomic ions. If i t were found possible to get doubly charged alpha particles with a radio-frequency ion source -56-i t would greatly simplify switching the Van de Graaff genera-tor from operation with a proton beam to operation with an alpha particle beam. The output of such a source when used with helium gas was studied to see i f any doubly charged ions were produced. 2. Excitation Processes in the Discharge High frequency discharges are commonly classifi e d according to the mode of excitation. (BABAT 1947) (i) The E or electrostatic discharge which is ex-cited by the alternating potential difference between the ends of the c o i l . ( i i ) The H, or ring discharge, due to the induc-tion f i e l d . At 20 megacycles/sec. both types of excitation w i l l occur. The mechanism of the discharge consists of generation of electrons by single electron impact with gas molecules, and of loss of electrons by diffusion, d r i f t , attachment, or recombination. As shown in Table 3, the ionization potentials of helium are much higher than those of the gases that are apt to occur as impurities in the helium. This is due to the closed shell structure of the helium atom. With the potential difference between the ends of the c o i l of the order of 1000 volts a considerable fraction of the electrons w i l l reach -57-Ion Vi(ev) Ion V-^ev) He+ 24.5 N + 14.5 He + + 54.1 \u00C2\u00AB2 + 15.5 H + 13.5 0 + 13.5 H 2 + 15.6 \u00C2\u00B02 + 12.5 Table 3. The ionization potentials, , of some of the light-gas atoms. energies sufficient to remove both electrons from the helium atom. The problem i s one of getting a high enough electron density to produce a significant number of such col l i s i o n s . The probability of double ionization of an atom by electron impact is closely related to s a t e l l i t e line inten-s i t i e s in X-ray spectra which also involve the simultaneous removal of two or more electrons from the target atom. Druyvesteyn has calculated, on the basis of the classical ionization theory (THOMSON 1912), the probability of double ionization by single-electron impact relative to that of single ionization, (DRUYVESTEYN 1927), E x = the ionization potential for the second electron. T = the kinetic energy of the incident electron. and Yx are screening constants which are small for helium. q x = the number of electrons in the \"x\" sh e l l . -58-E x E x The f i r s t term, \u00E2\u0080\u0094 \u00E2\u0080\u00A2 (1 ) , gives the dependence T T on the incident electron energy and is slowly-varying. Its maximum value i s 1 at T = 2E^. The second term, ( z \"^x^ , (Z - y x) i s of the order of one and is always less than one. The last q x term, cr \u00C2\u00BB shows that W w v varies inversely as the square of the effective nuclear charge. For x ? K, i.e. for the case of one electron ejected from the K shell and one from another shell, the formula gives good agreement with results from X-ray data. In medium and heavy atoms W g g is too small to lead to any measureable X-ray intensities. For helium the formula predicts = 1/8 or 12.5% at T <= 2Ex. This may be compared with 16% (MILLIKAN 1921) and 10% (WILKINS 1922) measured for alpha particles using Millikan's oil-drop apparatus. The energy dependence of the double ionization as measured by these workers, agreed qualitatively with the prediction of Druyvesteyn's equation. Hence even for electrons with energy above 54 ev, the relative number of He + + ions produced w i l l be only of the order of 10%. In addition to ion loss by the ordinary recombination process, the loss of doubly charged helium ions may occur by collisions with neutral helium atoms, He + + + He >- 2He+ -59-Th is would amount to the transfer of an electron from the neutral atom to the doubly ionized atom with the disposal of only 5 e.v. of extra energy. At a pressure of 50 microns the mean free path for helium atoms is only 0.4 mm. so i t is conceivable that this process might seriously reduce the fraction of the original He + + ions that can be extracted from the source. When helium contains small traces of impurities the similarity theorem (MARGENAU 1948, LLEWELLYN JONES 1951) does not hold, indicating that a process of ion generation or loss is involved that is not dependent on X/p, the ratio of f i e l d strength to pressure, (LLEWELLYN JONES 1953). One such process that is possible is the ionization of molecules of an im-purity in collisions of the second kind with metastable atoms of helium. H eraet + H2 > H e + H 2 + + e The state involved is a ls2s, metastable state with excita-tion energy 19.77 e.v. This is also involved in another ion-ization process (ARNOT 1939) resulting in molecular helium ions. H emet + H e * H e 2 + + e Arnot and collaborators also showed that HeH* ions are formed + by the attachment of H 2 to He. H 2 + + He > HeH4\" + H This reaction is relatively unimportant as i t affeets only a small fraction of the ions that appear as an impurity in a helium discharge. PIRANI GAUGE OSCILLATOR CARBON PROBE a a (X _C2 D D D EXTRACTOR / f \ HELIUM LIQUID NITROGEN Fig. 27. The radio-frequency ion source. -60-3. Experimental The 50 kilovolt accelerator described in Part II (d) (p.30) was used for this investtgation. The ion source is shown in Fig. 27. The discharge tube is of pyrex, 15 cm. long and 1.8 cm. inside diameter. Pyrex has a recombination coefficient of 2x10\"5 compared to 1 f 0 1 amp. H* beam H 2 +,He + + beam He beam 0.04^ 0.12^ 4.5 /tA 0.07 0.20 ^ *A 25 //h TABLE 4. Composition of beam with helium purified by passage through carbon at liquid nitrogen temperature. -64-It is seen that the ratio of the mass 2 beam current to the H + current is now 3:1 whereas with impure helium the ratio was closer to 1:4. This reversal may be an indication that part of the remaining mass 2 beam is due to doubly charged helium ions but there is oao conclusive evidence to support this view. In order to identify He + + unambiguously i t w i l l be necessary to i n s t a l l the ion source in the Van de Braaff generator and see i f an alpha particle reaction can be induced with the mass 2 beam. It is quite possible that the ab-++ solute yield of He w i l l be much higher in an intense dis-charge in impure helium though the percentage yield w i l l be very small. -65-APPENDIX II A HELIUM GAS LEAK ^ The flow of hydrogen to an ion source i s generally controlled by means of diffusion through palladium. Helium however, does not diffuse rapidly enough through glass or metals to be supplied through a thermally controlled diffusion leak. Many mechanically operated valves have been devised to provide a controlled slow rate of flow of helium but in general these cannot easily be adapted for electrical control as is required for use on the Van de Graaff generator. A leak designed by Mr. R. Bowman in this laboratory f u l -f i l l e d most of the requirements as to compactness, power con-sumption, flow rate, economy of manufacture and electrical control but did not close tightly enough when shut off. This leak consisted 'of a tapered steel plug f i t t e d closely into a brass case. The difference between the radial thermal expan-sions of the plug and the cylinder caused thegas toleak between the two when they were heated. The leak herein described (See f i g . 30) uses the same basic principle but some changes in the design have achieved effective closing while complying with the other requirements. Von-HELlUm INLET s s ^ - T E F L O N WASHEP-i L SCREW DRIVER. SLOT VL NF -20 LOCK.NUT INVAR. NICHR.OME WIRE miCA SHIELD. 2 9 /sV DR.. HELlUm OUTLET. SCALE l \" = I\". F i g . 30; Construction of the helium leak. -66-By using only a small bearing surface at the end of a long invar needle a higher pressure per unit area was pro-duced when the leak cooled. In addition the use of the longitudinal expansion rather than the radial expansion pro-duced a greater separation for a given increase in tempera-ture. The helium reached the valve chamber through a small hole in the threaded end of the needle. A lock nut was Q required on the needle to avoid backlash in the threads. The heating element consisted of 0.010 inch nichrome wire wrapped around the body of the leak and separated from i t by a sheet of mica. The variation of leak rate with power input is shown in 3 f i g . 31. 45 watts produced a leak rate of 125 cm /hr. When 3 the valve was closed the residual leakage was less than 0.1 cm /hr. When the power input was changed the leak rate took about 15 minutes to settle down. This time could be reduced by over-adjusting the power for a few minutes and then returning to the desired setting. The leak that has been designed w i l l produce a controlled flow of helium of the magnitude required for the operation of an ion source, The dimensions, power requirements, and method of control of the leak are such as to make i t suitable for use on the Van de Graaff generator. -67-APPENDIX III A HYDROGEN LEAK USING DIFFUSION THROUGH NICKEL Considerable trouble has been experienced by the Van de Graaff generator crew due to the failure of palladium leaks for the ion source. Therefore a study was made to determine the f e a s i b i l i t y of constructing a hydrogen leak out of nickel tubing. The use of diffusion through nickel (SNOEK 1950) for the purification of hydrogen indicated that this would be possible i f the size and power consumption of the device could be reduced. Although nickel has the second highest diffusion coef-ficients of a l l the metals, they are much lower than those of palladium^so that higher temperatures are reqnired to get the same leak rates as those given by palladium leaks. The diffusion rate obeys the formula: R = AT^ ^ e- b/ T where 3 R = the diffusion rate in cm. (measured at normal tempera-ture and pressure) per sec. per cm. per mm. of thickness. MATERIAL - BR.AS5. F i g . 52. Construction of the nickel-tube hydrogen leak. S . 10 . 15 POWER. INPUT (WATTS ) F i g . 55. Povfer consumption of the hydrogen leak. -68-A = a constant, characteristic of the medium. T = the absolute temperature. p = the pressure in mm. of mercury. b = a constant. The values of the coefficients for a sample of commercial nickel are: b = 6,680\u00C2\u00B0K A = 0.93 x 10\"2 cm. (mm. of Hg)~^ deg.\"^ sec.\" 1 mm. (POST 1938) 3 2 Using these values, R = 0.11 cm. per cm. per mm. of thickness per hour at one atmosphere pressure and 750\u00C2\u00B0K. Grade A nickel tubing was obtained from the Superior Tube Co. of Norristown, Pennsylvania. The leak i s shown in f i g . 32. One yard of nickel tube was used, of 0.050 in . O.D. and 0.005 in. wall thickness. It was wound in a double helix, the open end soldered into a hole in the plate lead-ing to the hydrogen bottle, and the other end sealed and connected to the exterior through a Kovar seal. To increase the efficiency, the heating current was passed through the diffusion element i t s e l f so that no auxiliary heater was required. The leak was designed to operate with the high pressure on the inside of the tube and vacuum on the outside so that the only appreciable heat loss was through radiation. -69-The flow rate was determined by measuring the pressure drop across a constriction 1.17 cm. long and 6.1 cm. in radius with a Pirani gauge. Fig. 33 shows the variation of the leak rate with the power input with a pressure of 1 atmo-sphere inside the tube and vacuum on the outside. With a 3 power input of 15 watts the leak rate was about 50 cm. /hr. The mean temperature of the nickel tube as judged by the change in resistance, was 600\u00C2\u00B0C. No sagging of the c o i l was observed. This leak rate i s sufficient for the operation of a radio-frequency ion source. The dimensions and power consump-tion are comparable to those for palladium leaks. The d i f -fusion element may be replaced at a cost of one dollar, about 1/50 of the cost of replacing a palladium leak. 7.1 z m M E P v i A L - AiunniNum. SCALE \u00E2\u0080\u00A2. 3\" - l\". Fig. 54. Construction of the \" b a l l protector\". -70-APPENDIX IV A SAFETY VALVE FOR THE VAN DE GRAAFF GENERATOR ION SOURCE On one occasion the ion source on the 50 kilovolt accelerator became overheated and the glass discharge tube imploded. The consequences of such a failure of the Van de Graaff generator ion source would be more serious due to the presence of nitrogen and freon at high pressure in the tank. The possibility of a mishap of this nature prompted the design of a safety valve for the protection of the vacuum system. The original ion extractor had a plain axial canal dri l l e d through i t . This was modified as shown in f i g . 34. An enlarged cavity was made in the canal with a tapered seat in the lower end. The length of the canal was altered so as to maintain the same pumping speed. Under normal operating conditions the pressure difference between the discharge tube and the small cavity is insufficient to move the bearing ball in the side channel. If the discharge tube should break the increased pressure would blow the ball into the cavity where i t would settle into the tapered seat and close the extraction canal. The valve has been found to be too sensitive unless the slope of the side channel is at least 30\u00C2\u00B0. -71-Since this modified extractor was installed in the U.B.C. Van de Graaff generator, the discharge tube has broken twice. The valve operated satisfactorily on each occasion, the pressure in the vacuum system rising from -5 -4 about 10 mm. of mercury to 10 mm. -72-APPENDIX V CALCULATIONS FOR THE THEORETICAL PREDICTION OF THE B 1 1(n >Q0Li 8 CROSS SECTION The following formulae were used: For the energy level density of the residual nucleus with excitation energy \"E\" Mev: _ 2\laF w(E) = Ce For the potential barrier height \"B\": Zze 2 B = R For the function \"F x\": 2MX Fx = \u00E2\u0080\u0094 2 - ^ . \u00C2\u00A3 x ^ ( ^ x > - w < E > ^ E In calculating F ^ , the following values of the parameters were used: Ejj^x = 12.9 - 6.6 = 6.3 Mev B =2.88 Mev a =0.14 Mev\"1 C =0.54 Mev\"1 -73-The calculation of F^. E Mev 2^aF 2VaE\" e ' w(E) Mev\"1 Mev mb. n 2 2 1 .7A8 2.11 .1.14 1.82 5.3 460 11,100 2 1.058 2.88 1.55 1.49 4.3 420 11,200 3 1.30 3.67 1.98 1.15 3.3 295 7,600 4 1.50 4.46 2.41 .80 2.3 150 3 ,330 5 1.675 5.34 2.88 .45 1.3 33 ,230 Therefore: F^ = 33,230 * In calculating F n , the following values of the parameters were used: Emax - 1 2 - 9 M e v B = 0 a =0.20 Mev C =0.54 Mev -1 -1 -74-The calculation of F n. E Mev 2/aE 2\/aE e * w(E) Mev\"1 kR *n Mev n 2 \u00E2\u0080\u0094 n 2 1 .894 2.44 1.32 2.50 11.9 590 9,300 2 1.267 3.55 1.92 2.40 10.9 590 12,300 3 1.550 4.71 2.54 2.28 9.9 610 15,400 4 1.788 5.98 3.23 2.16 8.9 610 17,500 5 2.000 7.39 3.98 2.04 7.9 630 19,800 6 2.192 8.95 4.83 1.91 6.9 660 22,000 7 2.368 10.68 5.77 1.76 5.9 680 23,100 8 2.534 12.60 6.81 1.61 4.9 710 23,600 9 2.682 14.60 7.88 1.43 3.9 730 22,400 10 2.828 16.90 9.12 1.24 2.9 760 20,100 11 2.968 19.45 10.50 1.00 1.9 830 16,600 12 3.100 22.19 11.95 .69 0.9 910 9,800 13 3.225 25.15 13.60 212,900 Therefore: 2 F n = 212,900 \u00E2\u0080\u00A2 n H2 -75-C7*. = 600 mb. r If is calculated on the basis of the assumption /. "Thesis/Dissertation"@en . "10.14288/1.0085536"@en . "eng"@en . "Physics"@en . "Vancouver : University of British Columbia Library"@en . "University of British Columbia"@en . "For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use."@en . "Graduate"@en . "Reactions induced by fast neutrons in boron trifluoride and the angular distribution of the non-resonant gamma radiation from the bombardment of carbon with protons"@en . "Text"@en . "http://hdl.handle.net/2429/40532"@en .