THE UNIVERSITY OF MICHIGAN INDUSTRY'PROGRAM OF THE COLLEGE OF ENGINEERING Absolute (d, a) Reaction Cross Sections and Excitation Functions Oswald U. Anders This dissertation was submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the University of Michigan. April 1957 IP-216

10-00 "..dllr, r u -1, Ask'40, 0%!, Q

ACKNOWLEDGMENTS It is a great pleasure to thank Professor W. W. Meinke for his continued support, guidance and cooperation during the course of this work, his willing aid in procuring equipment and preparation of the manuscript. The suggestions and advice of Professor L. O. Brockway, Professor K. Fajans, Professor J. Halford, Professor W. C. Parkinson, Professor C. Rulfs and Professor P. V. Hough are truly appreciated. The willing assistance of Mrs. Rosemary Maddock, Mr. Morris Wahlgren, Mr. Ronald Shideler, Mr. George Grindahl, Miss Alice Burton and Miss Susan Holbrook in various phases of the problem is gratefully acknowledged. Thanks also go to Mr. William Downer and other members of the University of Michigan cyclotron crew in arranging for the bombardments and to Mr. John Mannlein, Mr. Robert White and Mr. Hermann Roemer for the help in the construction of various equipment used in the research. The aid given by the Engineering Industry Program in printing the thesis is appreciated. The author is indebted to the U. S. Atomic Energy Commission for partial support of this research. Thanks be extended to Mr. B. Falk and the Arnold Engineering Company for providing thin metallic foils for the targets. ii

Last but not least the author wishes to express his sincere gratitude to his wife, Edith, who assisted in the reading of the counters during many small hours following the bombardments, in the calibration of the integrator, in many calculations and in the typing of the final manuscript. Isotopic tracers were obtained from the Isotopes Division of the Oak Ridge National Laboratory and by several bombardments obtained at the Argonne National Laboratory. iii

TABLE OF CONTENTS Page ACKNOWLEDGMENTS ii LIST OF TABLES x LIST OF FIGURES xii ABSTRACT xvi Part 1 INTRODUCTION 3 I. Nuclear Reaction Cross Section 5 A. Definition 5 B. Derivation 6 II. Experimental Approach 10 Part 2 INSTRUMENTATION AND TECHNIQUES 11 I. Determination of the Energy and Number of Deuterons 11 A. Cyclotron 11 1. Layout 11 2. Beam Energy 14 B. Techniques and Instruments Used to Induce and Control the Nuclear Reactions 16 1. Bombardment Chamber 16 2. Negative High Voltage Supply and Suppressor Ring 18 5. Current Integrator 20 a. The Principle of Operation 20 iv

Page b. Layout and Construction 23 c. Calibration of Current Integrator 27 II. Determination of the Number of Parent Nuclei in the Target 31 A. Targets 31 1. Target Preparation 33 a. Metal Foils 33 b. High Vacuum Evaporation 33 c. Substrates and Absorbers 36 2. Target Measurements 36 a. Weight 36 b. Area 37 c. Evenness 37 B. Calculations 42 III. Determination of the Number of Product Nuclei 43 A. Radiation from Product Nuclei 43 B. Chemical Separation 45 1. Methods for Carrier-free Chemical Separations 45 a. Recoil Methods 46 b. Solvent Extraction 46 c. Simultaneous Precipitation or Scavenging 47 d. Ion Exchange 48 2. Determination of the Chemical Yields 49 C. Identification and Measurement of Radioactive Reaction Products 53 v

Page 1. Beta-ray Counting 53 a. Proportional Counting 54 b. Los Alamos-type Thin Window Proportional Counter 61 c. 4w Proportional Counters 64.d. 4w Counting Methods 72 Thin Films Used as "Weightless" Source Mounts for 4w Counting 72 4w Sample Preparation > 76 Gold Evaporator 78 Need of Conductive Source Mounts 78 Amplifier 82 2. X-ray Proportional Counting 86 a. X-ray Proportional Counter 86 Interaction of X-rays with the Counter Gas 86 Design Considerations 88 Construction Details 89 b. Electronics 95 c. Performance 98, X-ray Spectra 98 X-ray Counting 98 3. Gamma-ray Scintillation Counting 103 a. Principle of Operation 103 b. Design Considerations 104 c. Construction 104 d. Gamma Counter Performance 108 vi

Page D. Absolute Counting Methods 109 1. 47 Counting 109 a. Self-absorption 109 b. Background and Dead Time Corrections 111 2. Absolute X-ray Counting 112 a. Calculation of the X-ray Counter Efficiency 112 b. Measurement of Absolute Disintegration Rates by Means of the Coincidence Method 113 Coincidence Method 114 Absolute Disintegration Rate Determination for an y88 Sample 115 Attempts to Calibrate the X- ay Counter with Snll13 and Mn54 Samples (Decay Schemes) 121 Part 3 PROCEDURES AND RESULTS 127 I. Bombardment Procedure 128 II. The Zr(d, a)Y Reactions 130 A. Target Preparation 133 B. Chemical Separation 134 C. Tracers 134 1. Yttrium-88 134 a. Identification of Y88 135 Yttrium-88 Decay Scheme 139 b. Counting of Y88 Tracers 141 2. Tracer Y90 141 vii

Page D. Record of Bombardments 143 1. The Zr92(d, a)90 and Zr94(d, a)Y92 Reactions 143 2. Determination of the Zr90(d, a)Y88 Cross Sections and Excitation Function 152 E. Resolution of the Decay Curves 155 III. The Mo(d, a)Nb Reactions 158 A. Targets 161 B. Chemical Separation 166 C. Tracer for the Mo(d, a)Nb Reactions 166 1. Procurement 166 2. Identification and Use as Tracer 168 D. Records of the Bombardments 170 E. Resolution of the Decay Curves 176 IV. The Ti(d, a)Sc Reactions 180 A. Targets 184 B. Chemical Separations 185 C. Tracer for the Ti(d, a)Sc Reactions 188 D. Record of the Bombardments 189 E. Resolution of the Decay Curves 193 V. The S4(d, a)P32 Reaction 195 A. Target Preparation 198 B. Chemical Separation 202 C. Tracer for the S34(d, a)p32 Reaction 203 D. Record of the Bombardments 205 E. Resolution of the Decay Curves 207 viii

Page 56 21 VI. The Fe5 (d, a)Mn5 Reaction 212 A. Target 212 B. Chemical Separation 214 Part 4 DISCUSSION 219 I. Nuclear Shell Model 219 A. The (n, y) Reaction Cross Sections 220 B. The (p, n) Reaction Cross Sections 223 C. The (d, a) Reaction Cross Sections 223 II. Considerations for Future Experiments 228 III. Summary 235 APPENDIX 238 BIBLIOGRAPHY 242 ix

LIST OF TABLES Table Page I. Range-energy relationship for deuterons in various materials 15 II. Elution properties of various elements from a small column charged with Dowex 2 (200-400 mesh) 51 III. Data for the investigation of the dead time vs. voltage relationships for proportional counters 60 IV. Thin film materials used for sample supports of A- and x-ray emitters 75 V. Difficulties encountered with the operation of a 4w proportional counter 84 VI. Data for the determination of the absolute decay rate of an y88 sample 120 VII. Data for the determination of the absolute decay rates of a Sn113 and a Mn54 sample 122 VIII. Properties of yttrium isotopes encountered in this research 132 IX. Chemical separation of yttrium from zirconium targets (carrier-free) 136 X. Results of the Zr(d, a)Y cross section determinations 146 XI. Arrangement of zirconium targets for bombardment 13 150 XII. Chemical separation for yttrium in bombardment 15 151 XIII. Arrangement of zirconium targets for bombardment 6 153 XIV. Properties of niobium isotopes encountered in this research 162 XV. Chemical separation of niobium from molybdenum targets (carrier-free) 164 x

Table Page XVI. Chemical separation of niobium from zirconium targets (carrier-free) 167 XVII. Results of the Mo(d, a)Nb cross section determinations 173 XVIII. Arrangement of molybdenum targets for bombardment 15 175 XIX. Properties of scandium isotopes encountered in this research 182 XX. Chemical separation of scandium from titanium targets (carrier-free) 186 XXI. Results of the Ti(d, a)Sc cross section determinations 194 XXII. Summary of the half-lives reported for phosphorus-32 197 XXIII. Chemical separation of phosphorus from zinc sulfide targets (carrier-free) 200 XXIV. Results of the S34(d, a)p32 cross section determinations 206 XXV. Arrangement of zinc sulfide targets for bombardment 3 208 XXVI. Extraction of various elements from 7.75 N HC1 solution by an equal volume of isopropyl ether 215 XXVII. Chemical separation of manganese from an iron target (carrier-free) 216 XXVIII. Elements excluded from Investigation because of half-lives of (d, a) reaction products or low abundance of parent isotopes 250 XXIX. Elements excluded from investigation because of decay curve analysis problems 231 XXX. Data for the beryllium absorption curve of x-rays emitted by yO8 and Mn54 240 XXXI. Data for aluminum absorption curve of Pm47 beta-rays 241 xi

LIST OF FIGURES Figure Page 1. Schematic representation of the layout of the cyclotron vacuum system 12 2. Bombardment chamber 13 3. Faraday cage and target probe assembly 17 4. Circuit diagram of negative high voltage supply 19 5. Current integrator rack 21 6. Current integrator top 21 7. Block diagram of the current integrator 22 8. Current integrator panel 23 9. Current integrator bottom 24 10. Current integrator, circuit diagram 25 11. Current integrator power supply, circuit diagram 26 12. Block diagram of apparatus used to calibrate the current integrator 27 13. Calibration curve of current integrator 29 14. Dimensions of aluminum target frames 32 15. High vacuum evaporator 34 16. Beta gauge 39 17. Aluminum absorption curve of Pm47 P-rays 40 18. Typical scanning data obtained with a sulfur target 41 32 19. Hood with 32 separation equipment 50 20. Small ion exchange column 50 xii

Figure Page 21. Counting room 55 22. Pulse height dependence of dead time for proportional counters 59 23. Schematic diagram of thin window proportional counter 63 24. Plateau curve obtained with thin window proportional counter 65 25. 4w proportional counter 66 26. Schematic diagram of 4w proportional counter 69 27. 4w counter plateau curves with gold plated sample 71 28. Gold evaporator 77 29. 4w counter plateau curves with non-gold plated sample 80 30. X-ray counter with preamplifier 88 31. Schematic diagram of x-ray proportional counter 91 32. Plateau curves obtained with argon-methane filled x-ray counter 92 33. Linear absorption coefficients for x-rays in argon and krypton 94 34. Diagram of mercury pump and manometer for filling the x-ray proportional counter 95 35. Circuit diagram of preamplifier for x-ray counter 96 36. X-ray spectrometer setup 97 37. SpeSgra of x-rays emitted by Cr51, Zn65 and yoo 99 99 38. Spectra of x-rays emitted by Mn54, Zn65, 88 Cd109 nd Sn113 (photographs) 100 xiii

Figure Page 39. Schematic diagram of the x-ray counting setup 101 40. Schematic'diagram of scintillation counter 105 41. Spectrometer Equipment 107 42. Gamma-ray spectra of Co 88 and Nb92 108 43. Beryllium absorption curves for x-rays emitted by Yo0 and Mn54 114 44. Decay scheme of y88 116 45. Block diagram of x-ray - y coincidence counter 117 46. Block diagram of y-y coincidence counter 119 47. Gamma-ray spectra of Sn113, Mn54 and Nb95 123 48. Chart of nuclides for zirconium region 131 88 49. Decay curves of y88 taken with x-ray counter and 47 counter 138 50. Gamma spectra of 88 60 4 51. Complex P-ray decay curve of yttrium from bombardment 7 144 52. Excitation function for the Zr(d, a)Y reactions 156 53. Chart of nuclides in the molybdenum region 159 54. X-ray spectra of 88 and niobium fraction from bombardment 14 160 92 55. Nb9 x-ray spectra 160 56. Gamma spectrum of Nb95 tracer 169 57. Complex P-ray decay curve of niobium from bombardment 18 172 90 96 58. Resolution of the Nb 9and Nb6 components of the niobium decay curve from bombardment 18 176 xiv

Figure Page 59. Gamma spedtra of niobium sample from bombardment 18 177 60. Excitation function for the Mo97(d, a)Nb9 reaction 178 61. Chart of nuclides in the titanium region 18.62. Complex B-ray decay curve of scandium from bombardment 17 191 63. Gamma spectra of scandium, ten and sixty hours and 21 days after the end of the bombardment 19 192 64. Chart of nuclides in the sulfur region 196 65. P-ray decay curve of 32 from bombardment 1 over 11 half-lives 199 66. Excitation function for the S34(d, a)p32 reaction 209 67. Complex beta decay curve of p32 P3 tracer 210 68. Chart of nuclides in the iron region 213 69. Schematic diagram of nuclear level systems with spin-orbit coupling 221 70. Capture cross section for 1-Mev neutrons 222 71. (d, a) reaction cross sections for 7.7 Mev deuterons 225 xv

ABSOLUTE (d, a) REACTION CROSS SECTIONS AND EXCITATION FUNCTIONS ABSTRACT The purpose of this research was the determination of the cross sections of various (d, a) reactions induced by the 7.78 + 0.05 Mev deuterons from the University of Michigan cyclotron. Special emphasis was given to reactions involving closed shell nuclei as target isotopes. Experimental techniques were refined to permit precision measurement of the cross section values. The investigation involved the bombardment of thin targets of metallic zirconium, molybdenum and titanium as well as of zinc sulfide, subsequent chemical separation of the product nuclei, identification of the products by x-ray and y-spectroscopy and determination of the absolute disintegration rates of i- and x-ray emitting isotopes. The instruments constructed during this research included: a current integrator permitting measurement of the intensity of the cyclotron beam with 1% error; two 4w P-ray proportional counters, a four-inch diameter x-ray proportional counter filled with krypton at two atmospheres pressure and a beta thickness gauge for thin films. xvi

Procedures were developed to calibrate the current integrator, to determine, with an accuracy of 1%, the evenness of target foils by a non-destructive method, to prepare cyclotron targets of materials sublimating below 15000 C by high vacuum evaporation, to purify various reaction products by using carrier-free chemical separations with decontamination factors of 105 and better. Methods were devised to determine the yield of the chemical separations with radioactive tracers, to identify various isotopes by their 7- and x-ray spectra, to count absolutely x-rayS and P-rays with proportional counters as well as with coincidence set-up, and to prepare thin films used for mounting the 4w counter samples. The absolute cross sections a, in millibarns, were measured for the following (d, a) reactions at the deuteron energy V + 0.05 Mev. Reaction a V Zr90(d, a)Y88 * 2.34 + 0.28 7.56 Zr92(d, a)Y90 3.79 + 0.26 7.56 Zr94(d, )y92 ) 4.01 + 0.28 7.56 Mo92(d, a)Nb90 2.95 + 0.14 7.71 Mo97(d, a)Nb95 * 2.35 + 0.(14 7.71 Mo97(d, a)Nb95m 0.98 + 0.09 7.71 Mo8(d, a)Nb96 2.53 + 0.12 7.71 Ti46(d, )Sc44 52.4 + 2.6 7.71 xvii

Reaction a V TiL48(d, )Sc46 28.5 + 1.7 7.71 S34(d,' a)P2 330.3 + 23.1 7.73 Excitation functions were determined for the reactions with asterisk. The decay scheme of y88 was investigated and a metastable intermediary state discovered. Similar metastable states were found for the decay of Mn54 and Sn1. Absorption curves in beryllium were experimentally determined for x-rays emitted by Y88 and Mn54. xviii

Part 1 INTRODUCTION The study of nuclear reactions is one of the approaches to the knowledge of the properties of atomic nuclei. The quantity most significant for such investigations is the nuclear reaction cross section. It is a measure of the yield of the reaction and gives an estimate of the ability of the target nuclei to undergo a particular kind of transition. In recent years many articles have appeared dealing with the study of nuclear reactions and the measurement of the corresponding cross sections. A distinction is made in this connection between the so called "absolute" cross sections and the "relative" cross sections. Relative cross sections are values giving the yield of a certain reaction with respect to another standard or reference reaction. The two reactions may involve different elements or isotopes or vary only in the interaction energies. Absolute cross sections, on the other hand, do not require such reference standards and are given in the absolute units of an area. Many workers in the past were interested only in cross section values sufficient for rough theoretical treatment and spent much of their effort on the measurement of relative cross sections, using the few previously determined absolute values as their reference standards. Although sufficiently good relative data can be accumulated in this 1

2 way, large errors of the normalized values cannot be avoided without sufficient attention to details. Many of the measured values have inherent errors much larger than estimated by the investigators. Neglect of the losses of reaction product due to recoiling and the uncertainties introduced by bulky counting samples emitting P-rays contributed to these optimistic error estimates. In some work carried out with "standardized" GeigerMdller counters no correction was applied for the sensitivity of the counters to y-rays emitted in coincidence with the P-particles or to the effect of backscattering from the supports of the counting sample. For some studies the side reactions that may have produced the same product as the reaction under investigation were neglected. Some workers based their measurements on relatively thick targets for which they had no criterion as to their uniformity. Others encountered difficulties with the energy definition of the thresholds of charged particle reactions. This was due largely to uncertainties of the energy of the accelerator beam used. In some work the measurement of the bombarding particles proved difficult and yielded erroneous results due to the neglect of charge losses from secondary electrons emerging from the target and detecting device during bombardment.

3 Some investigators were very successful in minimizing one type of error, but neglected others completely. It is in the light of these facts that previously reported reaction cross-section values have to be seen. Some authors aware of them may have quoted errors of the order of 100 or more, and their values are still to be considered good. There are very few authors, however, reporting standard errors of less than 20% for their determinations of absolute cross sections. The following few laborious attempts performed by whole groups of investigators to establish absolute reference standards are among the most successful: Excellent work on the C12(p, pn)Cll reaction,used as primary standard by many investigators, has been reported by Aamodt, Peterson and Phillips (1) as well as McMillan and Miller (86) of the University of California Radiation Laboratory, furthermore by Hintz and Ramsey (63) of Harvard University. They quote a probable error of 14% and attribute more than half of it to the uncertainties in establishing the absolute decay rate of the C1 reaction product, by means of a calibrated Geiger-Miller counter. Self-absorption is again the major source of error in the work by Crandall, Millburn, Pyle and Birnham (25), who investigated the C12(x, xn)Cll and A127(x, x 2p n)Na24 reactions (x = p, d or a). With foil thicknesses of polystyrene up to 0.015 inch and aluminum up to 0.010 inch

4 one cannot overcome the self-absorption difficulties even using a 4w counter. Errors up to 10% standard deviation were inherent in this work. Some very good determinations of relative cross sections and their dependence on energy were reported by Ghoshal (44, 45) for silver and copper for the medium energy range and by Meinke, Wick and Seaborg (90) for thorium and uranium for bombarding energies of several hundred Mev. Errors of approximately 20% for the relative values were quoted for this latter work. The present research attempted to avoid many of these difficulties as far as possible. New techniques were developed and various electronic equipment constructed and acquired to aid in this attempt. Some of the investigations reported in the literature try to correlate measurements of cross sections for a certain type of reaction covering a wide range of atomic numbers. Measurements of this kind were published by Cohen for the (n, 2n) and (n, p) reactions (22), by Hughes and Sherman (66) for the (n, 7) reaction, by Harvey (58) for the (d, p) reaction, and by Blaser, Boehm, Marmier and Scherrer (13) for the (p, n) reaction. Insight into certain structural properties of the nuclides Nan be gained from these articles. No such systematic treatment has been given thus far to the (d, a) reaction. The relatively low yields of this reaction prohibited extensive investigation. The few

5 absolute cross sections reported thus far were insufficient in number to permit a meaningful correlation. The purpose of this research was thus, to study the (d, a) reaction of various nuclides by determining the reaction cross sections for 7.7 Mev deuterons and measuring their energy dependence for lower deuteron energies where possible. The excellent resolution of the deuteron beam available at the University of Michigan Cyclotron, having a known energy of 7.77 + 0.02 Mev, was of great help in this respect. Particular attention was given to the so called "magic number nuclei" to investigate possible effects of nuclear structure on the (d, a) reaction yields. I. Nuclear Reaction Cross Section A. Definition If a beam of particles is directed at a layer of matter, the effect of this layer is additively composed of the effects of the individual units (nuclei) in this layer. These act as individual scattering centers. The Cross Section of a nuclear reaction is thus defined by: Number of events of a given type per unit time per nucleus Number of incident particles per unit area per unit time (1)

6 For charged particle reactions of the type: a + X -- C* -+ Y + b where an intermediate "complex nucleus" is formed, there exists a threshold dependent on the Coulomb repulsion between target nucleus and projectile. In this case reactions will require a minimum energy of the bombarding particle, since no compound nucleus can be formed unless the entering particle makes contact with the target nucleus. The cross section defined by equation (l)is dependent on the interaction energy. The resulting "excitation" function is a characteristic of the reaction. The interaction or excitation energy is the energy of the bombarding deuterons and is measured in million electron volts. B. Derivation The cross section of a deuteron induced nuclear reaction can be written in the form: Number of Nuclei formed by the Reaction (Number of Parent Nuclei/area)(Number of Deuterons) (2) The dimension of the cross section is then an area measured 2 -27 2 cm 10 cm in ncl or better in millibarns = nulu nucl 1eus nucl Zeuss

7 The number N of nuclei produced by a particular reaction per second is thus: N = I (3) p A where n is the number of parent nuclei, A is the target area and I is the beam strength in deuterons per second. If N is the number of product nuclei produced only by the reaction with cross section a and lost only by decay, and t is the time during which the amount of the product changes: dNi n dN= I a X(4) whereX is the decay constant related to the half-life by X= (loge 2)(tl)-1. If the bombardment is constant during t, and if the transformation of parent atoms is small so that I, n and a are essentially constant, the number of product atoms present at any time during the bombardment is given by the solution of above equation (cf. appendix) X A (5) N = (I A) (1 - e ) (5) At the end of a constant bombardment of time T I a (n/A) -T I (1 - e ) =N (6) and at any time after the end of the bombardment: I a (n/lA (1 X) et e N e'st (lU-e )& (7) The total number of product nuclei formed by the reaction considered is then:

8 NT = NpT = I a n/A x T (8) If NT is known the cross section can be calculated from NT () (n/A) I T If IT is obtained from the number of micro-ampere hours, Q, and the number of elementary charges on the bombarding nucleus one obtains 3600 x 10o-6 (0) I T = (Q/Z) x 1 (10) 1.60 x 10-b and 4.4 x 1017 NTZZ Q (n/A) (11) NT can be obtained by solving equation (6) for Io(n/A) and writing N XT NT = I o (n/A) T= - (12) 1-e XT l-e"( If the bombardment is not constant, it can at least be divided into intervals of equal length during which I can be considered constant. In this case m I. a- (13) _,1 A - A t -xt No = X (1 - e ) e i i=l

9 or n m ( A 1 x At'.Xt () N A (1 -e el) I i e(14) 0oi=l where m is the number of intervals and t is the time i from end of the i th intervals to the end of the bombardment. from equation (10) follows for this case: Qt 1 j6oo x 1O'6 = j 0 (.5600 (15) i z A t 1.60 x lo0- ) Qi being the number of micro-ampere hours in the i th interval. Equation (14) can then be written: n m 600 x 1o-6 aA (6) 0N l19 (l e Xe(16) 1.60 x 109 ) Zt (1 - e t) i=l Substituting Ci/K for Qi where K = counts per microampere-hour, the calibration constant of the Current Integrator, and Ci the Current Integrator counts collected in interval i, we obtain: =,l17 N At K (17) a = 4.44 x 10'17 No at XK (17) m (A) (l- e xt toCie ) 3=l1 i=l

10 II. Experimental Approach The cross section, defined by equation (2) can serve as a guide for the quantitative experimental approach to nuclear reactions. The efforts of this research may thus be classified under three main headings: 1. Measurement of the energy and number of deuterons 2. Determination of the number of parent nuclei 3. Identification and determination of the number of product nuclei. The determination of the bombardment energies involved mainly calculations using the range-energy data in the literature (6, 102) while the determination of the number of deuterons entering the reaction was reduced to the measurement of the charge collected in a Faraday cup positioned behind the target. The problem of determining the number of parent nuclei included the preparation of targets suitable for bombardment and the collection of dependable data on size, weight, and evenness of the targets. Values for the isotopic relative abundance were taken from the literature (65). Most of the necessary experimental techniques, however, had to be developed for the identification and measurement of decay rates of the radioactive reaction products. Several new instruments were designed and built and new methods developed for chemical purification and counting of the reaction products.

Part 2 INSTRUMENTATION AND TECHNIQUES In this chapter the instruments and techniques em - ployed for measuring the (d, a) cross sections are described in detail. Design and performance data are given for the instruments and the principles of the procedures are discussed. Some findings made when testing the equipment are reported. I. Determination of the Energy and Number of Deuterons A. Cyclotron The deuterons for the research were obtained from the 42-inch, fixed-frequency cyclotron of the Physics Department of the University of Michigan. 1. Layout: A diagram of the topography of the cyclotron and its associated equipment is given in Fig. 1. Deuterons are produced in the tank of the cyclotron by an electric arc and accelerated to a maximum energy of approximately 7.8 Mev. When the maximum energy is reached the beam is deflected out of the tank by the deflector plate at B and conducted inside the pipe E to the Focusing Magnet F. Particles of 7.778 Mev are selected and magnetically focused to the target plane in the scattering chamber M. The bombardments for this research were carried out at the position R in front of the Faraday cage T. This is inside the 11

12 CYCLOTRON LAYOUT - SCHEMATIC T A CYCLOTRON TANK R B WINDOW BOX _ P C PROBE PORT ---- 0 D SLITS E 6 INCH PIPE \-K F FOCUSING MAGNET G 6 INCH CHANNEL H VACUUM PUMP INTAKE I 4 INCH PIPE J SLITS - H K' 2INCH PIPE G L SLITS F M SCATTERING CHAMBER N SCATTERING TARGET O MONITOR P 2 INCH PIPE Q BOMBARDMENT CHAMBER R TARGET S SUPPRESSOR RING T FARADAY CUP U VACUUM PUMP INTAKE V 6 INCH PIPE W SLITS D X ANALYZING MAGNET Y SLITS Z EXIT CHANNEL Fig. 1 Schematic representation of the layout of the cyclotron vacuum system

13 Fig. 2. Bombardment Chamber Bombardment Chamber Q (Fig. 2), and situated approximately 2 ft behind the focusing plane. At the position R in the bombardment chamber the diverging beam covers an area of 2 approximately 2 - 3 cm and has a maximum strength of 1 - 2 micro-amperes under standard operating conditions. Behind the target position is the suppressor ring S, the Faraday Cup T and other equipment not related to this research.

14 2. Beam Energy: The energy of the deuteron beam of the cyclotron at the focusing plane in the scattering chamber was determined in the Fall of 1956 by Olexa Bilaniuk and found to be 7.778 + 0.005 Mev (12). For the collimator apertures used in this research a straggling up to 50 Kev (width at half maximum) was assumed (7). This 7.8-Mev deuteron beam is capable of inducing (d, a) reactions in nuclei up to the rare-earth region, but its energy has been found to be low enough to lie below the threshold of the (d, an) reaction for elements as low in the periodic table as sulfur. When bombardment energies lower than 7.8 Mev were desired, stacked-foil techniques were employed (88). The resulting energies were calculated from the rangeenergy curves of deuterons in the anteposed absorbers. The range-energy relationship for Mylar, aluminvum, titanium, copper, zirconium and molybdenum used for this calculation are given in Table I. The values for aluminum and copper were taken from the literature (6, 102). The values for Mylar were obtained from Prof. W. C. Parkinson (94) and the values for titanium, zirconium and molybdenum were * obtained by interpolation. This interpolation was done by the following method: The proton range-energy relations for molybdenum have recently been published in a Russian journal (107) but were inaccessible to the author.

15 r-o OH c 0 H & L 0> oS [".0 0 co C 03 H C M r<'\ -L 0 OH p Cd H HCM C.- LC O Cl L0 Ln -.' S to r^ V o0 co c4 o ^ (c 0 V.0... 4) 0 O bO.i) E-'-I b (0 o3 o 1 0 V2o i LC M n O O O CH OO H t< G O K 4 H C 4 LC - 0.. U) H' q O * O l'H 30J... - -PbO [>- C Ln 4 L O C ci' H C [ Ln KG O) ) Ho GO [ 0 0 L H 0> EH CM ["N -r 0G c I 0 a, I H ci) -p h( oci)

16 The values of deutron range were calculated and plotted at. wt. vs. the atomic number for aluminum, copper and silver. Slightly curved lines were obtained for the points of corresponding energies. The ratio R range/at. wt.) element (range/at. wt.) copper was calculated and a plot of this ratio vs. the deuteron energy prepared. A smooth curve was fitted to the points of this plot and the value of this curve at a particular energy taken as the interpolated R. Multiplication by the range value for copper and the atomic weight of the element at.wt. yielded the range values of deuterons of the given energy in the element as reported in'Table I. (range in element)E = R (range in copper)E at. wt. The uncertainties incurred by application of the extrapolations being of the order of + 5% are smaller than the beam energy attenuation in the targets used. B. Techniques and Instruments Used to Induce and Control the Nuclear Reactions 1. Bombardment Chamber. The Bombardment Chamber (Fig. 2) contains the Faraday cage mounted on Teflon insulators (Fig. 3). A Suppressor Ring is mounted immediately before the mouth of the Faraday cage. The entire unit is mounted on a small wagon

17 Fig. 3. Faraday Cage and Target Probe Assembly by which it can be removed from the path of the cyclotron beam, when not in use. The cage as well as the suppressor ring are connected via shielded flexible cables to two Kovar-glass seals and Teflon insulated high voltage connectors on top of the Bombardment Chamber (Fig. 2). The target, mounted on an aluminum frame fits into the target probe assembly, which is placed in front of the suppressor ring. For introduction of the target probe into the evacuated bombardment chamber a vacuum lock is provided as can be seen in Fig. 2.

18 Before the actual cross section determinations were carried out for this research the bombardment chamber was thoroughly overhauled. The chamber and all its contents were thoroughly cleaned and new, more-flexible cables installed. The surface of the Teflon insulators was shaved and the solder joints rechecked. Intermittent check of the equipment in the time during which experiments were carried out insured against failures. 2. Negative High Voltage Supply and Suppressor Ring. During the course of the bombardment secondary electrons are freed from the target and the walls of the Faraday cage by the impact of the deuterons. When electrons from the target are permitted to reach the Faraday cage or those produced in the Faraday cage permitted to escape it, wrong beam-current readings are obtained. It has been reported by Hall (49) that current readings as much as 27% too low are obtained in this case. A suppressor ring placed between the target and the Faraday cage and charged with a negative potential exceeding 200 volt prevents secondary electrons from entering or leaving the Faraday cage. For the experiments carried out during this research a negative potential of 1000 volt was applied to the suppressor ring during the bombardments. A special negative power supply was designed and built to supply this negative potential. Fig. 4 gives the circuit diagram of thia unit. In Fig. 5 it can be

19 FI - 109 2x 2-A 784 2.5Fv 60,, TR-002 x... 6.3v g _ 4.7M - 270K |pf ~ 10 0V *I/1', *OImf 270K 1600v IOM 250' --- 90M 2C53 500K 0K IM 6.3 v ~72939^ 7134N N lOOK 560K. K LiXWN' 0250K lOOK 560K Fig Circuit Diagra Negative High Voltage SuppyIM Fig. 4. Circuit Diagram' of Negative High Voltage Supply

20 seen as the second panel from the top. The unit proved very dependable and no difficulties were experienced with it. 3. Current Integrator. The charge collected by the Faraday cage during the course of the bombardment is measured by the current integrator. This charge divided by the charge of one deuteron (+4.862 ~ 10-10 esu.) represents the total number of deuterons which penetrate the target, with an error of less than 0.001. This is the case since only very few interactions take place between the deuterons and the target nuclei, and almost the entire beam will emerge from the target attenuated somewhat in its energy but not in intensity. The current integrator is connected to the Faraday cage by a low-capacity coaxial cable (UG 62/U) and mounted in a cabinet together with a Sorensen type 1000 S AC line voltage regulator and other auxiliary equipment. (Fig. 5) a. The principle of operation of the current integrator is best described using the block diagram of Fig. 7. If S is closed one has a conventional feedback amplifier on the left with an output load consisting of Rm plus the resistance of the meter. The voltage developed across this load is: e3 = A e1 i1 - el, but A el + el = R1 i or e(A + 1) = R1 thus: e3 = Ae1 = x R1 i1, where A is the gain of the

21 *El 1, R 0 4CO,) Cbo bO

22 ei AMPLIFER R OUNTER Fig.' 7.Block.;Diagram of theC -urregnt'T tgttor amplifier -without feedback.'If A is sufficientlyl -arge, A-~~~~~~~ 1 " I RCU I) RIi i"! then el is'small..:' Inithis case,:the input terminal is held approximately at ground and the input current is A.+';..** e''.,*-;..*... -,,*,,. —.Y —.....' i1 ='A; x w; or approximately ii - eg/RL. the meter current through the apq r If i is very much less then si-~ than th&voltagehacross the capacitor teld 7,pBo atiaground: andf.t'h'e crr et:i en}"pati t.is,.\...,. i,Rl the' metAer current throughia thec'apa t-l o r I f i is very1 will be Vc = (l/C)J idt'. Since e3 = i R and i e i Rm/Rl) then V = (1/) (R/Rm) i dt. "f..3. h sich S:',~.~:-?o/nd, he,,:apiirwl r

23 Thus in effect the input current is amplified by a known factor R1/Rm and integrated by the capacitor C. b. Layout and Construction. The unit was layed out to accomodate the original version of the circuitry as obtained from S.. Rankowitz (l1.) at Brookhaven National Laboratory (Fig. 6), The: instrument has,10 ranges and is able to integrate currents as low as Q.01 micro-ampere and as high as -100 micro-ampers at full deflection of the "Current Indicate'"meter (Fig. 8). Fig. 8. Current: Integrator -Panel When the instrument was first tested it was impossible to adjust it to indicate "no current". The wiring was thoroughly checked but with no avail. As precision parts were used throughout, component failure was quite unlikely. Introduction of a feedback loop, changing the values of several resistors and addition of a 75 ohm potentiometer for fine control adjustment of the "zero integrate" potenti

24~~~~~~~~~ I:~~~~~~-p: ii —i~ii- "::'' —-:: -:iil:: i::iii-ii- i:' —-i: —-~i- ii-i-i-i~: i —i~ii0i rq:::: ~~0 4i~i-:):Lxd ~~~~iiiiiiiii ~ ~ ~ ~ ~ b::i -:::~~~~~~~~~~~~~~~~~~~~~~~~~~i *~;~ ($>ii- - iii~ i::i::::~~~~~~~~~~~~0 G::j~~D"S~"S~:::_ -:::i ii rriii:ii:i ii~-iiii::i-i ~ H ~

25 IBs ~~~ WOaLI oR2o X~so.~j o= Juro iivano M'XQ 1_^o t~ ~- 11 Ro i-, i- ~ *:r 3 -- w o 7 —vvA =: oE, n 30 - A 3.LO~O I 8 t3iV88lt1V MM')E09 iNl 00 9. 1 —---- C cr'go^^ — — I,,,.g4 < 5~~ to~~~~ — Aw- * L o\\ t 0Z0 0~ IwS < z ro- oH, ____1 _l.. _ m mm 0r,/) 0 yO< Oo3 J-lNI MM OI - LL~ 0 > JIll - - ln <oi <n ^2\ 0 >?l — OD Yq ~ ~(/InU r^' 3is!F D O (, mm | %I A 0 MM%1 A009 3LVOIONI MMO J| 1 A 9889 — )10.~n N03HiAtV- lI MM4-41 -Hn X

26 H,, zo I (_D —-^-[ L _-o L.J - L,-I S oi-E ^ <1:o u o;'".g 04 4 Z4 <[-0 0 >~/ ^2 ir > Kffl f I —- _. r' 0 ^'+ o0o 0H co o w o &'S~ ^ —----------------------------- I. —-., o —! — 1-,0'.0 )r 0 +- -p 0p 0 -ii_ c|i:, + acc t(9~rJL 0Z 000 = _ _ __ * ^ _,^ ^ ^. - o0I LOdN, z > - II _____ ___________________________________ + W 0'S~ S~ 5 0 0 zz2 8z > r ~~cu Q~~~~~40

27 ometer, resolved the initial difficulties. A view of the chassis wiring is given in Fig. 9 and a close-up of the panel in Fig. 8. Fig. 10 and 11 give the diagram of the corrected circuits as used in the present unit. c. Calibration of Current Integrator: The Current Integrator built in the Fall of 1955 was calibrated against a known current. The method employed is best seen from the block diagram of the calibration circuitry shown in Fig. 12. The Weston standard cell was compared with a wall (GAL STANDARD POTENTIOMETER STORAGE CELL. BATTERY' HI -MEG. I STANDARD L FARADAY SUPPLY INTEGRATOR Fig. 12. Block Diagram of Apparatus Used to Calibrate the Current Integrator. mounted standard to within + 0.0012. The latter had been calibrated by the National Bureau of Standards. The Potentiometer was a Rubicon type B, operated with a Leeds and Northrup type R - 2500 galvanometer. This combination

28 was found to be able to measure a 1 millivolt drop of potential across a 10,000-ohm standard resistor with an accuracy of 0.02% standard error. A Leeds and Northrup type 4040 resistor was used for the standard. Its resistance had been measured in 1954 by the National Bureau of Standards and found to be 10,005.2 ohm + 0.1 ohm. Six drycells connected in series and parallel were used to drive the potentiometer. The High-Meg resistance box contained 13 glass-sealed high meg ohm resistors ranging from 10 x 10 to 104 x 109 ohms (Victoreen Instrument Co., Cleveland, Ohio) mounted on Steatite selector switches and enclosed in an airtight desiccated copper box with two (UG 496/U) cable connectors as output (Fig. 5, top panel). A very stable high-voltage power supply was used for supplying the calibration current.:This unit had been built in this laboratory by W. Cassatt (17) in 1953. During the calibration procedure the Faraday cage was connected to the same input connector on the current integrator as the standard resistor. Excellent linear relation between current integrator counting rate and calibration current was found for all ranges of the unit. Fig. 13 is the calibration curve for the range used exclusively during this research, representing a slope of 64.79 counts per minute per micro-ampere. The factor K of equation (17) is thus: K = 64.79 + 0.6% counts per micro-ampere hour.

29 I I 1 70 CALIBRATION OF CURRENT INTEGRATOR 6050 LJ >40 LI n - 30 _0 / / 20 I0-.1.2 3 4.5.6 7 8 9 1.0 CURRENT (MICROAMPERES) Fig. 13. Calibration Curve of Current Integrator A least-squares analysis of the ten calibration points gave a standard deviation of 0.6% from perfect linearity. Since the deviation was found to be random, currents up to 1 micro-ampere could be integrated during the course of this research with standard errors of less than 1%. With

30 a half hour warm-up period before each bombardment run the drift characteristics were found to remain well within this 1% error for bombardments of two hours and more. The performance of the instrument was very satisfactory after an initial failure of the millisecond relay due to exceedingly strong surge currents had been corrected.

31 II' Determination ofthe Number of Parent Nuclei in the Target The second factor entering into the cross section equation (2) is the concentration of the number of parent nuclei in the target. This value depends largely on the dimensions of the targets whose measurements are facilitated by the use of carefully prepared targets. A. Targets Two factors enter into the considerations for preparing targets for the cross section measurements: When penetrating the target the deuterons are slowed down. For good energy definition of the cross sections it is, however, important that nuclei lying on the bottom of the target should undergo nuclear reactions with deuterons of more or less the same energy as those-lying on the top. Thus, the target used for experiments intended to measure cross sections should be thin. Too few nuclei exposed to the beam in extremely thin targets, however, will result in too few interactions and too little reaction product for measurement. The optimum target thickness must thus be a compromise and was found for this research to be approximately 0.0001 inch or 2 - 5 mg/cm2 During the nuclear reaction the momentum of the bombarding deuteron is transferred to the compound nucleus

32 and part of it'to the reaction product nucleus. This momentum might be strong enough to make the products leave the target itself. It was found (56) that such recoils emerge primarily from the back of the target and are stopped completely by a 2-mil Mylar film placed immediately behind the target itself. For the present research 1-mil Mylar substrates were used for this purpose. 2-3/4" o -1-3/8" —2-3/4" 37 R=U. 1321 1-1/2"' Material: Aluminum 1/16" Thick Fig. 14.'Dimensions of Aluminum Target Frames

33 1. Target Preparation. There are essentially two ways of preparing thin and even targets: the use of thin metal foils and the deposition of the target substance onto a substrate by high vacuum evaporation. a. Metal Foils. Refractory metals like zirconium, titanium and'molybdenum could not be evaporated in the metal evaporator available in this laboratory. With the help of various industrial firms it was, however, possible to obtain thin foils of these materials which could be used for targets. Disks were cut from the foils with a 1.474 + 0.002-inch diameter punch die. These disks were glued onto a 1-mil Mylar film serving as the substrate and mounted on aluminum target frames. Mounting the targets on 1/16-inch-thick aluminum frames prevented deuterons which did not penetrate the target from reaching the Faraday cage. To minimize interaction of target nuclei with any neutrons produced by the Al(d, n)Si reaction in the frame, targets were cut to just fill but not overlap the aperture of the target frame and mounted on the beam-exposed side of the frame. The dimensions of the target frames can be seen in Fig. 14. b. High Vacuum Evaporation. Targets for the S(d, a)P reaction were made by high vacuum evaporation. Their preparation posed considerable difficulties. Materials to consider for these evaporations were: elemental sulfur,

34 Fig. 15. HTigh Vacuum Evaporator lithium sulfate, cadmium sulfide and zinc sulfide. Only very uneven targets could be prepared with elemental sulfur, they were also found to evaporate during the bombardment with deuterons (54). Efforts to evaporate lithium sulfate, a substance producing no long-lived radioactive side products on deuteron'bombardment, failed after repeated attempts. Reasonably good target deposits were obtained, however, with cadmium sulfide. A technique was evolved at last for preparing targets for the study of the S(d, a)P reaction by evaporating zinc sulfide onto the roughened surface of a Mylar film.

35 The zinc sulfide did not stick to the smooth Mylar film. Proper roughening of the Mylar surface was accomplished in the following way. A 0.001-inch-thick sheet of Mylar was stretched over a glass plate with the aid of masking tape. The Mylar was then rubbed with a cotton wad dipped into a paste of fine pumice to produce an evenly roughened surface. Squares fitting the target frames were cut out of the roughened Mylar and mounted onto the frames with the roughened side facing away from the frame. For the evaporation a collimator mask consisting of an identical target frame with leveled edges was placed over the Mylar and this assembly mounted, mask down, onto the frame of the highvacuum evaporator (Fig. 15). An aluminum block having a raised circular surface was placed onto the target frame so that the polished surface touched the smooth surface of the Mylar to serve as a heat dissipator during the evaporations. Chemically pure zinc sulfide was placed in the tungsten boat serving as filament and evaporated unto the Mylar in high vacuum. Eight consecutive evaporations with intermediate rotation of the target were necessary to deposit approximately 2 1.5 mg/cm of zinc sulfide on the Mylar, suitable for bombardment targets. The weight of the deposit was determined by weighing the target frame before and after the deposition of the zinc sulfide. It was thus necessary to prove that during

36 the evaporations the target frame with the Mylar substrate did not change its weight. This was done by exposing an empty target frame with Mylar substrate to the same process, but with no zinc sulfide in the tungsten boat. The weight before and after this procedure remained the same within 0.006%, thus proving the suitability of the method. c. Substrates and Absorbers. Mylar film O.001-inch thick was chosen as substrate for most of the targets. The primary reasons for its choice were its relatively great durability toward radiation and its easy destructability during the chemical separations. The radioactive products due to the deuteron bombardment of Mylar (F and N 13) could be easily separated during the purification of the reaction. products. Aluminum foils of various thicknesses were used as substrate for the excitation function runs. This was advantageous, since aluminum could serve simultaneously as beam energy attenuator and catcher of the recoils. It was found able to withstand the greater beam intensities employed for these experiments. Some interference from aluminum in the chemical separations could be tolerated for these runs. 2. Target Measurements a. Weight. The average thickness of the targets was determined from their weight and area. The weights of the foil disks were found by direct weighing with a semi-micro

37 balance, which permitted weighing within 0.02 mg, representing an error of less than 0.1% for the targets. The weight of the zinc sulfide& deposits was determined with a regular chainomatic analytical balance. The empty target frame with Mylar substrate was weighed before the evaporation and then reweighed after the ZnS was deposited. The difference in weight represented the weight of the ZnS deposit with a standard error of approximately 0.5%. b. Area. The area of the metal foil disks was determined by the dimensions of the stamp die used for cutting the disks. The area of the target deposits prepared by high vacuum evaporation was determined by the dimensions of the collimating mask used. The area of the targets was thus known with an error of less than 0.1% standard deviation. c. Evenness. Erroneous results may be obtained for the reaction cross section, if uneven targets are used with a non-homogeneous beam. The cross-section value would be too large, if the center of the beam struck a part of the target which is thicker than the calculated average. It was thus necessary for this research to prepare even targets and obtain an estimate of the thickness variation for each individual target. Several methods are available to determine the evenness of target films of the type described above. The more accessible ones are less reliable and usually destructive. A crude way would be direct measurement of the

38 thickness at different parts of the target with a good micrometer. A destructive method consists of cutting a prototype target into several small sections and determining the weight and area of each (50). The variation in the thicknesses of the sections will serve for an estimate of the thickness variation in the destroyed sample. The same variation is assumed to occur in the other targets made from the same batch. For the present research a reliable sensitive, and nondestructive method was sought. The search for such a method resulted in the design of the beta gauge,shown in Fig. 16, for localized measurements of thin films (4). 147 The instrument employs the weak 0.23-Mev P-rays of Pm 7 and allows an estimate of the evenness of films a few mg/cm2 thick by reproducibly scanning the sample with a collimated parallel 3-ray beam 3/52-inch in diameter. 147 The B-ray point source of Pm47 was prepared by a method reported previously (3). Approximately 5 millicuries of carrier-free promethium chloride were deposited in a circle about 0.05 inch in diameter on a 0.04-inchh thick Lucite disk.. The source was covered with Cellophane tape and mounted in a 7/8-inch-thick Lucite collimator with a smooth-walld hole of 3/32-inchi diameter. The details of the source and collimator assembly are given in Fig. 16.. A microscope-type movable stage mounted on the source assembly allows reproducible scanning of the

0 0..'0 "-I F-h: -4-) 0 o3 rH- *0 o -, C) H- Q) ~.- P I rd ci ~ TJ -i., bs- r-H-i Cio \ r-10 4 O r -P 0 0 vJ ObE~..,, bO Q) o - Co - Cd OP., *o Q. Cd o -Pl V~ mH p _ q~

40 c/m 13,000 12,000 ALUMINUM ABSORPTION CURVE OF 111,o000 1~ Pm147 BETA RAYS 10,000 9,000 8,000 7,000 6,000 5,000 4,000 3,000 2,000 1, 000 I I I I I I I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 mg/cm2 Al Fig. 17. Aluminum Absorption Curve of Pm147 P-rays thin-film samples. A thin-window Geiger-Miller tube is suspended above the collimator hole and is protected by a 0.04-inch-thick plastic disk with a 1/4-inch central collimating hole. A typical aluminum absorption curve for the 0.23-Mev 147 - particles of Pm4 taken with the gauge and standard aluminum absorbers is given in Fig. 17. From this curve it is seen that the instrument is most sensitive for films 2 up to 6 mg/cm. In this range it is possible to discover thickness variations from point to point of less than 0.1 mg/cm. Fig. 18 shows the performance of the gauge

41 E cTi E LU 0 u 5.83k ~69% rn Z < 5.570O >68% 68%0 4 w2 v Y 5.56> > * D >67%- i -5.42LU o67% - I% of total 5.27 D Hz66%/ Ithickness z5. Z66% Z O 1.0 1.5 2.0 2.5 3.0 D LU RELATIVE POINT POSITION (INCHES) < Fig. 18. Typical Scanning Data Obtained with a Relatively Uneven Sulfur Target. Nine points 1/4 inch apart were scanned three times in random order. (X) Scan No. 1; (v ) Scan No. 2; (a ) Scan No. 3. Counting time was 10 min per point giving a reliable error of 12 per count. with a relatively uneven target of elemental sulfur deposited on Mylar. The thickness variation of all cyclotron targets used in this research was determined with the Pm147 gauge. The metal foils could be measured directly, the variations in the ZnS deposits were determined by difference. For the latter, fifteen measurements were taken on the Mylar substrate at points distributed over a 25 cm2 area. Zinc sulfide was then evaporated onto the film and counts again

42 taken at the identical 15 positions. The difference of the counting rates was calculated for each position and the data evaluated by the least-squares method. A standard deviation from perfect evenness of less than 1% was obtained for several ZnS deposits by this procedure. B. Calculations The total number of nuclei in the target given by the chemical atomic weight and Avogadro's number, the weight and area of the target, as well as the isotopic abundance of the nuclei of the parent species in the naturally occurring element enter into the calculation of the parent isotope concentration for a nuclear reaction. Data for weight and area were obtained experimentally as described above. The chemical atomic weights were taken from the International Table of Atomic Weights (125), the isotopic abundances from the Brookhaven National Laboratory Report on Neutron Cross Sections of July 1955 (65). The concentration of the parent nuclei in the target was then calculated by: n/A _ (weight) (isotopic abundance) (6.02 ~ 102o) (atomic weight) (area) (chemical atomic weight)

43 III. Determination of the Number of ProductNuclei The third factor entering the cross-section equation is the number of nuclei produced by the reaction. From the development of the cross-section formula as given in section I-B of the Introduction, it is seen that this amounted to the determination of the absolute disintegration rate of the product nuclei at the end of the bombardment. The methods and instruments used for purifying, identifying and counting the nuclei produced by the (d, a) reactions are described below. A. Radiations from Product Nuclei An excess of either protons or neutrons in a nucleus can, according to FaJans' and Soddy's "Displacement Law" of 1913 (35, 118) be corrected by the emission of an appropriate B-particle. Nuclei with neutron excess emit negatrons (-), nuclei with proton excess emit positrons (P+). The energies of the p-rays emitted by a particular nucleus vary from very small values to a certain maximum energy characteristic of the emitter. The distribution of these energies (P-spectrum) can explain certain nuclear characteristics, and is at the present the subject of extensive studies carried out by many authors (e.g. see 68 and 71). For the positron emitters there exists an alternate path of decay. The nucleus can capture an electron

44 from the electronic shell of the atom and by this process achieve the same effect as by positron emission. As the hole in the electronic shell is filled, a characteristic x-ray is emitted indicating the capture process. In case the product nucleus from a 3 -decay or electron capture process is left in an excited state, it will lose energy by y-ray emission to go to a less excited, more stable form of the same isotope. The three types of radioactivity encountered in the middle of the periodic table are thus: P-rays, x-rays and 7-rays. To count the absolute disintegration rates of n-ray emitting reaction products the self-absorption of n-particles in the source material has to be considered. Electrons and positrons emitted as P-rays strongly interact with matter and have relatively short ranges in solids and liquids. When counting P-rays emitted by a source containing much solid matter, many of the n-particles will never even leave the source and enter the sensitive area of the counter. It is thus of great importance to use only weightless or "carrier-free" F-ray sources, when their absolute disintegration rates are to be determined. X-rays and y-rays are more penetrating kinds of radiation. Self-absorption in the source is of minor importance when counting them. The very fact of their greater penetration through matter, however, requires

45 special techniques for their measurement. Much more detector material has to be used to bring them to interact, posing problems of large background rates of the counters, partial interaction of the radiation with the counter material) etc. B. Chemical Separation The product nuclei of the (d, a) reaction are chemically different from the target element and any other nuclear species produced by the interaction of 7.8-Mev deuterons with the target nuclei. It is thus possible to separate the (d, a) reaction product from the target material and radioactive impurities by chemical means. The determination of absolute disintegration rates in a sample of 3-ray emitting reaction product can only be carried out on bulk-free counting samples. The chemical separation steps must thus form a "carrier-free" procedure and yield the (d, a) reaction product in pure and bulk-less form. In addition it is necessary to know the yield of the chemical separations employed. 1. Methods for Carrier-free Chemical Separations Primarily four methods are available for the chemist to separate a small amount of activity from macro amounts of impurities by carrier-free procedures: recoil methods, solvent extractions, simultaneous precipitation and ion exchange.

46 a. Recoil Methods. Recoil methods were employed already in the early stages of nuclear chemistry and seem to gain increasing interest in the recent literature (91, 120). They are based on the principle that upon the emission of radiation (in particular a radiation) or after the impact of an accelerated particle, the recoiling daughter nucleus has enough momentum to escape from the sample. This carrierfree product may be collected in some catcher material. In another case, recoiling nuclei might break their bonds in insoluble substances and form soluble ions which can be eluted from the original molecular substance (24, 121). These methods are extremely suitable for the characterization of short lived isotopes. They cannot be used, however, for absolute yield determination of nuclear reactions, since not all reacting nuclei will recoil in the right direction to escape the original sample and be collected by the catcher. b. Solvent Extraction. Solvent extraction methods are relatively common for carrier-free separation procedures. They make use of the different solubility of certain species in various solvents. As this behavior is the same for micro and macro amounts, rather good chemical separations can be effected (73). For the present research solvent extraction procedures were avoided. They seemed to offer too few real advantages to offset the risks of spills when performing them in a hurry and under pressure.

47 c. Simultaneous Precipitation or Scavenging. Microamounts of radioactive species will precipitate from a medium in which they are insoluble together with an added carrier of different chemical nature which is equally insoluble in the medium (33,34)The latter forms the bulk ofthe precipitate which can be centrifuged and filtered. When employing such a method it is well to keep in mind that every precipitation from a radioactive solution will carry some adsorbed traces even of soluble radioactive materials (coprecipitation). The amount of these impurities may vary greatly, however, depending on the physical form of the precipitate. The latter Often depends on the way the precipitation was carried out. Gelatinous precipitates are usually good carriers, micro-crystalline precipitates not (103). In working with small amounts of a desired species when losses cannot be affordedprecipitates intended to carry the desired activity should be the only ones formed in a solution. In the case that this is not possible,it was found that homogeneous precipitations as pioneered by Willard (126, 127) carry least of a desired activity kept in solution. Homogeneous precipitations have the additional advantage of being able to precipitate only one of two species, where both would be precipitated, if any excess of reagent were added.

48 d. Ion Exchange. Ion —exchange techniques have found wide application during the past decade as they opened up an entirely new field of separation chemistry. They permit dependable and easy separation of species virtually inseparable by classical methods (61, 74, 75, 93), For the present research ion exchange procedures had the additional advantage of equal applicability to macro and trace amounts. When overloading of the resin bed was avoided, very clean separation could be achieved. Ion exchange steps were mainly used for the final purification of the reaction product rendering it in carrier-free form suitable for absolute P-ray counting. It was necessary to use conductivity water when preparing the eluents, Many of the ion exchange steps carried out employed Dowex-2 resin in the hydrochloride form. This anion exchange resin, obtained from the Dow Chemical Company, Midland, Michigan, was purified by washing with various solutions in the following sequence: Water, conc.HCl, water, 2 M HC1, water, 2 M HC104, water, 2 M NH40H, water, 2 M NaOH, water, conc.HC1. The resin thus purified was stored in 6 M HC1 and used in small portions over a period of more than two years, giving very satisfactory performance. For each The resin purified by a similar procedure is now available commercially from the Bio Rad Laboratories of Berkeley, California.

49 separation step the resin was packed into small columns and washed again with distilled water and conc. HC1. Fig. 20 gives the dimensions of the columns used with Dowex 2. Table II indicates the ion exchange behavior of many metals in such Dowex-2 columns (61). Fig. 19 shows the experimental apparatus to study the elution rate of p32 from a Dowex-50-Fe(OH)3 column as described in the procedure for the separation of phosphorus from zinc sulfide. 2. Determination of the Chemical Yields. When the product of the (d, a) reaction is obtained in carrier-free form, the question remains: How much of the original amount of reaction product was lost during the chemical purification and how much was actually incorporated into the sample ready for counting? An: excellent method to determine the yield of a carrierfree separation is the isotope dilution technique. A known amount of a carrier-free isotope of the reaction product is added to the undissolved target. Complete isotopic exchange between this tracer and the bombardment product is effected in a homogeneous solution under rigorous chemical conditions. The amount of tracer isotope mounted with the purified reaction product on the counting sample is determined.:The amount of tracer lost in the separation process will then be proportional to the amount of reaction product -lbst in the separation procedure.

50 0 z Z CD z 0) O $w - OC - U)~~~~~~~~00 H E E FP4 cr F-" rd bO CQ

51 Table II. Elution Properties of Various Elements from a Small Column Charged with Dowex 2 (200 - 400 mesh) (61). Eluting Agent Elements Eluted by \5 ml. Eluting Agent 12 M HC1 Alkali Metals, Alkaline Earths, Rare Earths, Sc,.Y, TiIII V11 III IV Ni AsIII AsV SeIII SeIV TI Pb II As, As, Se Se17 T1 Pb III II III II BiI Y Cu slowly, Al, Cr, Fe Mn 6 - 9 M HC1 TiIV VV, Ag*, Ta*, PtII, Zr, Hf 3 - 6 M HC1 Fe Co, Ge, Ge, b 1 - 3 M HC1 Zn, Ga, MoVI In, Sn, Te, Te Au (as AuCl2),PbII Below 0.01 M HC1 SnII HII BII Sb slowl H,Bi, Sb slowly 3 M HClO4 Po, Cd, SbI Sb slowly 1 M NH4OH Pd, Ag, Sb"I SbV 1 M NaOH W Not eluted Tc*, Ru, Rh, Re, Os, tr, Pt V Au1 Tli * In trace'concentrations only

52 It may be pointed out that complete isotopic exchange is a stringent requirement for the successful application of this method. It can be achieved only in homogeneous liquids under severe chemical conditions such as strong acids and oxidizing media. Radioactive isotopes to be used for yield determination of the separations of this research had to meet the following specifications. The half-life of the tracer must be either considerably longer or shorter than that of the reaction product to permit easy resolution of the decay curves. If there is only one reaction product, the tracer must not be identical with it. The tracer must be chemically pure, free of contaminating activity and available in carrier-free or partially carrier-free form. If the tracer used is identical to one of several reaction products at least two bombardments will be necessary for the cross-section determination. The first bombardment will establish the ratio of the components in the reaction product at the end of the bombardment without addition of tracer. The second experiment, with tracer, will determine the ratio of the components as if the tracer had also been produced by the bombardment, The amount of isotope identical with the tracer, but produced during the second bombardment can then be calculated from the count rate of the other component and the ratio determined by the first bombardment. The chemical yield is established by comparing the. amount of tracer

53 carried through the separations with the amount of tracer isotope initially added. It is unneccessary to determine the count rate of the tracer absolutely or with the same counter with which the reaction product is counted; as long as it is possible to establish the ratio of tracer material carried through the chemical procedure to the amount originally addedthe chemical yield can be determined. The various tracer methods employed during this research as well as the identification of the tracers used will be described with the individual bombardments in the experimental part. C. Identification and Measurement of Radioactive Reaction Products Various counters were employed in this research to identify the reaction products and determine their decay rates. The operation and design of these counters as well as their use is described below. 1. Beta-ray Counting. Most of the reaction products encountered were B-ray emitters. Particular emphasis was thus given to counting this type of radiation. Relative P-ray counting was carried out with a Los Alamos type thin window proportional counter; absolute beta counting with two Borkowski-type 4w pro

54 portional counters. For aliquoting and estimations of fray activities Geiger-MTiller counters were used. a. Proportional Counting. The proportional counter consists of a cylindrical or rectangular chamber whose walls are made of conducting material. A thin wire serving as anode forms the axis of the chamber. This center wire is insulated from the conducting wall and can be charged to a potential of several thousand volts. As counter gas very-pure methane or mixtures of argon and methane are used. The design of todays proportional counters is essentially the same as was developed by Geiger and Miller in 1928 (42). The ionization-amplification in the counter precedes by the following steps: A particle enters the chamber and produces primary ionization of the counter gas in the vicinity of its trajectory. The "secondary" electrons thus produced are attracted to the anode wire and accelerated by the electric field. When their kinetic energy reaches a value great enough to ionize another gas molecule on collision, a'second generation of electrons is produced. Both electrons travel now towards the anode, colliding with new gas molecules as they gain momentum. This avalanching continues until all electrons are collected by the anode, resulting in a sudden drop of the potential of the wire and producing an electric pulse to the number electrons collected.

55 Fig. 21. Counting Room. Left: thin window proportional counter housing and PA-2 amplifier etc. Center: Borkowski-type 4w proportional counter in lead housing with Nuclear Chicago Ultrascaler. Right: same as center. Counter shielding closed. The field gradient around the anode is very large. Most of the avalanche thus takes place in the direct vicinity of the wire. This results in an excess of positive ions in the vicinity of the wire which are repelled from it with great violence. As the positive ions recede from the wire and travel toward the wall, the negative pulse is further increased by induction. The second part of the process requires approximately hundred times as much time as the collection of the electrons, so that the total pulse has a steep initial rise turning into a slower further rise until the positive ions reach the walls and the potential at the anode is restored. The

56 size of the pulse is "proportional" to the number of secondary electrons which again is proportional to the energy of the entering particle. As the positive ions strike the wall they may liberate a new set of electrons as so called delta-rays. These move rapidly towards the center wire and thus renew the process previously terminated. To suppress this "double pulse" effect organic substances are admixed to the counting gas or used themselves for the counting gas. The organic molecules exchange electrons with the positive ions so that virtually only ions of the organic molecules reach the counter walls. The energy which would otherwise be used for liberating electrons from the wall is now used to decompose the molecule of the "quencher". The probability of producing delta rays despite the quench gas is increased with the size of the avalanche which in turn is dependent on the anode potential. Double pulsing will thus determine the upper limit of the potential with which the chamber can be operated. One characteristic of every counter is its voltage plateau. This plateau is the range of anode voltages for which the counter is able to produce distinguishable pulses for every particle entering the counter during its sensitive time. The limits of the plateau are voltages too weak to permit the formation of a distinguishable pulse for some of the particles, and potentials for which the quenching

57 gas becomes unable to successfully control the formation of delta-rays. As the polarization potential of a proportional counter is increased the field will first reach the strength to produce avalanches. This voltage is the "threshold of the counter". As the potential is increased further, the pulse height is approximately doubled for every 200 volts until Geiger region is reached and very large amplifications due to multi-avalanche production are obtained. At some potential along the way the counter will be producing pulses large enough to be recognized by the electronic circuitry connected to it. This potential will depend on the sensitivity of the amplifier and shall be called the "threshold of the counter-amplifier combination". As the potential is increased further the "plateau" is reached, i.e., the counter produces recognizable pulses for every particle entering its sensitive volume and the count rate does not vary with the increase of the potential. The upper limit of the usable plateau of a counter-amplifier combination is generally the overloading of the amplifier which is unable to handle an excessive range of pulse heights and will block or double trigger the scaler. (Note: this double triggering of the amplifier is due to the differentiation of overloading pulses and is not the same as the double pulsing due to failure of the quenching process in the counter.)

58 The'insensitive time of the counter-amplifier combination, i.e., the time during which a pulse occurs, is amplified and the counter potential restored, is of the order of a few microseconds. The dead time for a counter-amplifier combination is dependent on the pulse height. The results of an experiment carried out with one of our 4w-proportional counter to prove this effect can be seen in Fig. 22 and Table III. The cause for the effect is the size of the pulse which has a length somewhat related to its height. The advantages of proportional counting over Geigercounting are manyfold.- Longer plateaus can be obtained, and highe'rcounting rates accomodated; the energies of the ionizing particles can be measured by the height of the proportional pulses and less quenching material is required for very dependable performance. Certain requirements on the electronic circuitry, however, made the use of the proportional counter practically impossible until recent years. The sensitivity of the amplifier must be great enough to recognize the smallest pulses from the counter produced by a single ion in the counting gas. As the smallest practical proportional pulses are of the order of a few millivolts and the trigger pulse of the scaler of the order of one volt, a minimum amplification of about 1000 is required. Electronic noise in the amplifier must be kept well below this one millivolt level so as not to be confused with true pulses.

59 x/ 4 / 15-10 PLATEAU CURVES /o 24* 104 x/ 24 I 4 o XK.X - 14 w X/ — X - "w "' Dm 2 m/ -x - / X x x 1/ a: i 0 / o2 I <o 50 F DEAD TIME / 0 40 4 prop. counter2 C ) 01 3 / a* 1 o - oo 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 ANODE POTENTIAL (kilovolts) Fig. 22. Pulse Height Dependence of Dead Time for Proportional Counters ANOE OTETIL (ilvols~ Qi.2.PleHih eedne rDa iefrPootoa UJ~~~~~~~Cu~~r 50

60 \ O _ 0 c'\ co c0o 0 r-PH r-l 04 C\ 0I K CO 0 O 04 t0o.' —- CO rO Ln O 0 C\0 4- r'- rH r- O. -H - 4 > O 0 O U2.,1 4- o m cz cQ rf Ln \1S 0o;*H c 0 O 0> 04 L> o- 0 o ) C01 0> c n n n LC n n OH Q 0O C\j C\J COl C\j CJ C\j C\U O4 ~ ^ 0 ~ H -p0 C n 0> O 0 -t H CO 4 L kn q 4C ) 0 0 0 4 00 00 kco )' n n o0 (D CO KO - O rc\ 0 k' 0 r 0 {H O " \ " \, \, 0% \, %'% o. t?-) 041 0 C l r \ c\j n 4 n 0c\I 0 o 0 H CM *H _-i 0 9 0 4 rN 0 0 04 0m H *H 0 C0 4 4 n H 0 4 0> 04 kC ) 0 Q o - r — o 9 a o ci) c^ 0 4 ^ o Lo 4 L 4 -O O r04 O n 0 — 04 0O 4 04 C\ 4 H *H Hr- H 1H H H H H H H H 4 C O (D ~0 0 4 0 H H O r-i r ^ -t- t DO ^-! t C!J H t- r0 r r - r r H O 0 H O 0 r3 0>-S-I H 0... O - 4 C H k\ -P L>n n 0 4 rK C\ H 0 V'\ EH?O c O 4O jt- r, I C\J Hl H H I H H H H Hrd H H H L H o) C 4) H i2 4 4 O 4 C0 4 C0 00 4 Cr3 00 O O - O O L H -P O O O O -.O CO r H 0> 0 H c0 L> 0 0 - -I rn n C n -3C O r 4 4 c4 H O r-t —tc r- HH rH ffi ~ ~ ~ z~ I^ ooa ^VOo

61 The primary'ionization of the counter gas when counting P-particles may vary by a factor of 100 and more., The range of pulse- heights presented to the amplifier will be of the same magnitude. If plateaus of the order of 800 v are to be obtained and the pulse height is approximately doubled by a 200 volt increase of the counter potential, the amplifier will be required to recognize separate pulses varying in size by factors of 2000 or more. For the largest pulses, the amplifier must not overload and double trigger the scaler. The dead time of the amplifier must be small to make best use of the properties of the proportional counter. The scaler unit must be fast enough to handle count rates up to 500,000 counts per minute without appreciable loss. These requirements could not be met until the advance of the electronic age, but even today the cost of the highly refined equipment prohibits the use of the method in many cases. b. Los Alamos-type Thin Window Proportional Counter. Most of the relative B-ray counting carried out during these investigations made use of a Los Alamos-type thin window proportional chamber (76). A diagram of this counter is given in Fig.23. The instrument was set up in the Fall of 1955 and is being operated in conjunction with a modified Los Alamos type PA-2 amplifier and S C4 scaler (Fig. 21, left) custom built by Trott Electronics Company, Rochester,

62 Explanation of Fig. 23 1. High voltage connector UG - 560/u 2. Connector support 3. Left shelf rack 4. Counter body 5. Right shelf rack 6. Polystyrene insulator 7. Cover 8. Aluminized 1/4 mil Mylar window 9. Window frame 10. Shelf stop 11. Polystyrene insulator

63 OH edd 0 0 0 r 0 bO C) bC0 rrc l / / \ ^~~~~~~~~~~~~~~~~~h / / \ *H~~~~~~~~~~~~~~F ^ ^v

64 New York. Plateaus obtained with this combination at two different settings of the amplifier are seen in Fig. 24. During this research the counter was operated at a 4400 volt anode potential employing 99-mol-percent methane (Phillips Petroleum Company, Bartlesville, Oklahoma) as the counting gas. The counter was shielded by a two-inch lead housing to reduce the background to approximately 56 counts per minute. Very reliable performance was obtained with the counter-amplifier combination. It can be seen at the far left of Fig. 21. c. 47-Proportional Counters. Absolute counting of P-particles was carried out with two Borkowski-type 47wproportional counters (122). An aluminum disk 0.010inch thick with a 3/4-inch-diameter hole in the center covered with a thin film on which the source is mounted separates two identical halves of the counter. The resulting two chambers are connected in parallel to the same input of the amplifier so as to add their counting rates and register coincidences as single counts. They are thus able to detect practically every P-particle emitted by the sample in any direction. The loss of P-rays striking the 10-mil aluminum sample support is negligible if the radioactive sample is deposited in a small circle of approximately half the diameter of the sample plate aperture. Gamma rays emitted in coincidence with the

65,O~~0 0 S o 0 r2 03 0 0 P0 \ 00 0 N CD rf)4 o.o 0 —. 0 I\I 2@ m>; = 0 0 0 I 0 3JV DNILNfoD 3AI1-13 1 C I Z O. iO To o 0 a < o \, U (L. (L LL ~.,. _ _ oI 31V 8NIlNOD 3a: V3.

66 Fig. 25. 4X-Proportional Counter. Left: counter opened. Right: counter inside housing. P-rays during the decay process will only increase the size of the n-ray pulse when interacting with the counter gas. The singularity of the pulse, indicating one disintegration will, however, be preserved. True 100 efficiency of the counter for detecting disintegrations by which 3-rays are emitted is closely approximated in the case of relatively strong n-rays and if "weightless" samples are used. In the case of very weak P-rays self-absorption in the thin film sample mount has to be considered and the necessary corrections applied. Fig. 25 shows two photographs of one of the 4w counters. The original design of the counters by Borkowski (122)

67 specified operation with a mixture of 90% argon and 10% methane at an anode potential of 1700 volts or methane counting gas at 2600 volts. For this purpose a 1-mil stainless-steel wire was used as anode and the counter evacuated and filled with gas for each count rate determination. To make the counters more rugged and suitable for routine internal counting of many samples, the design was modified for this research. The counters operate now with 2-mil stainless steel wires at.4000 volts chamber potential. They are used as gas-flow counters using 99gmol-percent methane (Phillips Petroleum Company) as counting gas. The design of the 4w counters is seen in Fig. 26. When operated with suitable amplifiers, plateaus of the type seen in Fig. 27 can be obtained with these counters. The two units used were completely interchangeable so that exactly the same count rate was obtained when counting a sample of very long half-life in each of the two counters. This interchangeability together with the long plateaus obtained and the counting of long-lived secondary standards which had been compared with a P3 standard obtained by the National Bureau of Standards were the tests for proper operation of the counters and the basis for assuming absolute counting of B-particles in the counter.

68 Explanation of Fig. 26 1. Top chamber 2. Bottom chamber 3. Teflon insulators 4. Sample plate (10-mil aluminum disk with a central aperture of 3/4-inch diameter) 5. Seal insert 6. Brass nut 7. Hose connector 8. High voltage cable 9. Kovar seal 10. Cap screws, No. 10 - 24, Allen Head, 1/2 inch long 11. I"0" ring 3/4 x 9/16, 3/32 cross section 12. "0" ring 3 1/4 x 3, 1/8 cross section 13. 0.002-inch stainless steel wire

69 Fig. 26. Schemati Fig. 26. Schematic Diagram of 47r-Proportional Counter

70 Explanation of Fig. 27 Amplifier sensitivity Correction to ordinate X 1 millivolt 0 0 2 millivolt +500 * 4 millivolt +1000 v 6 millivolt +1500 a 8 millivolt +2000

71 0 o LO _ X o 1I n 1 - I) I I I LC -p -xO ^- O ( 0 C) ^? _C LO -0 o 0 0 0 - o- 5o o o 0- C I-.- oI W A SlN -H -)I NI Ha SINNO -B r ^^^_ ~^ ^~ - " \ ^ ~~~~ -~ - ~~~~~~~~~~~~~~~r, - ^ \ -

72 d. 4T-Counting Methods. Considerable time and effort was spent in the present research to develop optimum procedures for 4w counting. These ensured fool-proof and reliable routine operation of the counters over the long periods of time required for following the decay curves of the samples. Thin Films Used as "Weightless" Source Mounts for 4w Counting. Absolute P-ray counting is only possible if the self absorption is negligible. When bulk-free sources are used it is important to mount them on "weightless" films or foils to make best use of the 4w counter. Various materials are being used for this purpose and are listed in Table IV together with their supplier, minimum thickness and method of preparation. During the present research only Zapon films and VYNS films were used for 4w counting, the former for tracers, the latter for reaction products. The advantage of the Zapon is its relative resistance toward acids and bases but easy destructability by conc. H2SO4. The advantage of the VYNS is its complete inertness toward practically all inorganic chemicals, and its great strength and durability. The method used for preparing the Zapon films is desribed in great detail by Hall (51). The method for preparing the VYNS films is given below. The latter is an adaptation of the method reported by Pate and Yaffe (95).

73 A pail is filled with water to the brim and the water surface cleaned by moving a metal bar across it. Two metal bars are placed parallel to each other close to one side of the pail. Approximately five drops of a 1/3-saturated solution of VYNS in cyclohexanone are the; placed at, equal distances between the two bars. The bar next to the openwater surface is now moved lengthwise along the edge of the other bar to distribute the solution evenly in the channel between them. The bar is then momentarily lifted from the water and some of the solution permitted to spread across a part of the water surface. The bar is then lowered to touch the drying part of the moving film and pulled across the water surface drawing the film along at a desired rate. The rate of this motion determines the thickness of the film. An even film area is selected by the' color of reflected light and an aluminum frame slightly moistened with the film solution is placed on the film to cover this area (95). A woodsplint moistened with cyclohexanone serves as a knife to free the frame from the surrounding film. The excess film is removed and the frame with the film lifted from the water with a rolling motion. Several layers of film can be built up in this way. Care must be taken that no water droplets and air bubbles are trappled between the layers as the frame is lowered onto a new film on the water.

74 Notes to Table IV A. Casting drops on water as described by Brown, Felber, Richards and Saxon (15) and Hall (51). B. Teflon-30 solution is painted on aluminum foil and baked at 550~. The aluminum is then dissolved away by dilute HC1. C. Solution is pulled across clean water surface as reported in section the text. D. Obtained from manufacturer.

75 cTl -bObO bD ~H *H 0 H - C H l? o.i o0 to 5 t t p 4O O co 4~!. * P - P-, — o o 0\ 4c c)to t o C',, 0 4) Q rl (DQ H 0 0 *>C) r Co C2o CM*P ^ T-I m-100 01 0 o o 0 c rI' -i -* 0 C a) O0 I ci t cl t-o4 -P O+ >4 C) ^ cd MH 1.H*-r ^ 0 d rD 4) 4 *Hcli cl a Q L to cl C O 0 f O O ^ O 4- 4)->dC I0 00 ^ 0 0 C 4P ) c o*H* = o O o l ~I r- 4) rC - 0 r- 10 *4-HZ r L ci 2H H )DH PH i C c*H VO cd > <D?- O cO c^ - ~ i+^ 2 A P r1H HO c i OU o S *Hr O Fm 0 0 Ch h 03 H -H Oc) 4O)C 0 Co p C- r ) c1 ) h r OC\ H - ~ 0 0 CO ) H H H H H C H ^ ^ ^ a) ^ P^ anH, H H C Ln \) H H HH o rd 4) CM H CM <D O b Q < < ct < f Q Q v Q D SC *,-1 i tM. gH H to3 *H C C 0 *H H * H t':b *H H H) C) v \ N \ 10 H F\q 4 10 \. 120 H 04 12 0 4 4*HH - t o o 0 6 0 0 I 0 1 o. o o 0 4-) c O 0 -r a) I C); 4-)t 1204) )o 4O, 4 )t4)o -4)o C o4'-) H E-l E- o rl C CEM r ci)~~~~~~~~~~~~~~~ ~ mt'0 H C) * Q O H O H > -z O U w O t O O ( (I) 0C 0 O c Co 0, (D ZK ~ Q) z C Q Co ( ( r-f Cti ~;^ ^0 ^^ A ^^ ^?-,{^^ OOr-1^ z Oj ~ CO ^ O P O c ft P OP^O z O? (DA f Cd ^(D O t (U O < PO O (D CD C0 00 U OO z Ql 2; u -C S QSO o Ql rl LQ Z; < rpqo Q vD ^ cM * ^ O rH O1. O - O V oO ts 0( t O X 3t V O g - UP V ~, cr3 V >{ >) M Ec 0 0? V^ U A <O

76 4n-Sample Preparation. Aluminum source disks with two layers of film giving a brown-yellow coloration in reflected light were used for 4w counting. The sources were deposited on the film by means of a micropipet. Several drops placed individually in the center of the film area could be deposited simultaneously to permit faster evaporation. Sufficient distance between the drops and the aluminum frame was kept to minimize the number of B-rays striking the aluminum. The samples were dried with a warm air blower and heat lamp. The dry samples were covered with one layer of film prepared by the same method as described above but using a larger frame. Frosting the sample surface onto which the film was to be placed helped the film to stick better. Similar one-layer films were prepared and gold plated. One of them was deposited on top and one on the bottom of each sample so that the gold layers formed the conducting outside surface. A "Q-tip" saturated with cyclohexanone served to remove excess film and ensured electrical contact between gold layer and aluminum disk. Gold plating of the 4w-sample films was carried out either in the high vacuum evaporator mentioned above or a specially built gold evaporator described below. The mounted films were placed onto the frame of the evaporator and gold evaporated from the filament. The progress of

77 0 0 0 Cd F-i ~~O~~~~~ 0 4 —.,Oz- -p.-. -%N0 n. - o ro - Cd icw z Z o r= 00 6 -L I U ^ 0 j ^ / jlv j 1 =,,. < a-. m a. I N 0 sxo / ~, "' 0) co J L 5 w____ ___ _ j:0, I o 1 0 IC -- ^ -- -^ -- ^ - ---- o 0 ------ ^ --- D

78 the deposition could be followed during the evaporation and the process stopped when a desired shade of blue transmitted light was obtained. Gold deposits of approximately 20 micrograms per square centimeter were used for the 4w samples. Gold Evaporator. A new induction heated evaporator somewhat similar to that described by Pate* was built to reduce the time required to goldplate the 4w sample films. Fig. 28 shows a diagram and photograph of the apparatus. The system consists of an evaporation chamber with ground glass cover for introduction of the sample frames. The neck of the chamber receives a hemispherical platinum or tungsten crucible, which can be heated by the external induction coil made of 1/4-inch copper tubing. The chamber is connected to a liquid air trap. A Welch Duo-Seal High vacuum pump is able to pump the whole system down to 0.1 micron in two minutes. By using an all glass system no leak problems were encountered. Need of Conductive Source Mounts. At the beginning of the investigations there was considerable uncertainty as to the need of a conductive source-mount. Results from a number of laboratories seemed to indicate that much depends upon the size of the aperture across which the film bearing the sample is stretched (52, 59, 96, 122), Sample plates originally used in this laboratory had openings of 1/2-inch diameter, for which some experiments * Pate, B. D., Disintegration Rate Determination by 4w Counting. Me Gill University Ph. D. thesis, April 1955.

79 seemed to indicate that absolute counts could be obtained without conductive coating. At sufficiently high anode potentials these samples would discharge, however. Sporadic discharge and scattering of data were observed occasionally even at plateau voltages. The ffects seemed especially pronounced for weak beta- and x-ray emitters (48), To investigate these phenomena a sample of Y88 (a longlived emitter of x-rays and weak positrons) was prepared by depositing the isotope in carrier-free form on a nonconducting layer of Zapon film and covering it with a film of the same material. A sample plate with 1/2winch aperture was used for this experiment. The sample was introduced into the counting chamber and the dependence of count rate vs. anode potential investigated. A plateau with a definite maximum was obtained and an abrupt uncontrolled rise in counting rate occurred at a relatively low applied potential. To investigate this further, one-minute counts were taken in succession at 100-volt intervals indicating a maximum in the plateau region. When the same counts were taken without interruption and before the breakdown potential had been reached on the previous curve, a curve with a maximum was again obtained, but the count rates were consistently far below the corresponding ones of the first curve. A third curve taken without interruption again lay below the previous one and thus successive curves. A certain

80 l__ ~ IV IIII 19,000 18,000 LU. z 17,000 n 16,000 15,000- /- E —L - 14,000 3600 3800 4000 4200 4400 4600 ANODE POTENTENTIAL (VOLTS) Fig. 29. 4w-Counter Plateau Curves with Non-Gold Plated Sample

81 limiting curve seemed to be approached after several curves were taken. This can be seen from Fig. 29. Interruption of the successive counts, when the anode potential was at the upper end of the range resulted in a much lower following curve, while hesitation at a lower voltage resulted in a higher plateau curve. When the voltage was turned off for several hours the original curve could be reproduced. This time and voltage dependent hysteresis effect can be explained in the following way. Charge is building up on the sample by drifting positive ions striking the non-conductive film (122). The effective potential between source and anode is thus decreased below the edge of the plateau and the count rate decreased. As this effect is dependent on the rate of production of positive ions in the proportional counter which again depends on the size of the avalanche, a maximum in the plateau curve is to be expected. Another phenomenon was traced to the same cause. When the voltage on the counter reached the breakdown potential, a discharge started, then accelerated until the scaler jammed. Reduction of the high voltage several hundred volts below the "breakdown" potential would not terminate the discharge. When the high voltage was turned off completely after a discharge and then turned on again, discharge started at a much lower potential than the first time. Longer waiting after the voltage was turned off raised the breakdown potential.

82 The same sample was next gold plated on one side without improving the above characteristics. When the other side was alsogold plated, however, a reproducible plateau as seen in curve VIII, Fig. 29 was obtained. No maximum was observed in this plateau and no discharge occurred up to 5000 volts. (The relatively large slope of this curve is due to double triggering of an overloading amplifier and has since been rectified). The result of this investigation proved the necessity of conducting source mounts for 4w samples. All samples used for the cross section experiments reported in the experimental part were goldplated and no scattering of data or sporadic discharges was experienced with them. As a result of these findings 3/4-inch-aperture sample mbunts were used facilitating the sample preparation for the bombardment runs. Amplifier. The amplifier is the most sensitive part of a proportional counter set-up. It must distinguish pulses from the proportional chamber varying in size by a factor of 2000 or more and introduce no spurious counts. Its dead time must be small and the amplifier responsive even at very high counting rates. It took considerable effort to find the best amplifier scaler combination for the two 47 counters. The final choice fell on the Nuclear-Chicago Ultrascaler Model 192X which incorporates in one unit an excellent amplifier with

83 good overload characteristics, a very stable 5000-volt power supply, and a fast and dependable scaling unit. Several other features included in this unit made routine counting less tedious. Difficulties were, however, experienced even with these amplifiers. It took several months to bring the two units to operate properly. Part of the difficulties may be ascribed to the higher humidity and temperature in the counting room during the summer months, part to tube failures and maladjustments of certain components. When replacing all 6AK5 tubes in both units with the more dependable type 5654/considerable improvement in reliability could be achieved. It was found advisable, however, to check the tubes of the amplifier every two months and make the necessary replacements. It took the author considerable time and effort to set up the proper procedure for 4w counting and learn the remedies for the shortcomings enumerated in Table V. After this experience was gained, however, no more serious delays were encountered due to failure of the 4w-counting equipment which is now considered among the most dependable apparatus of the laboratory.

84 0 4 )o i 0 4 Q rO O o o o i, 4)'^ 0O CP 4-C ) d P o p,- o 0 H p P O H b H U d o 3 U[O 4),. H 0 0 " p o > o CH H rH dH.H 4) p O- d 0 1-r r( M g -H -P ~ O d 4 4 d0 O o;p o m ~ ~o 4) 0 4> 4) 0 - C rO 0 "C)0 0 4)C Uc OCO p0~~)C Xi 32 H O)0 0 ~ H *H^ p 0,p'a H D o0 l 0 ci 04 4 C *- o o a,-p )R r m op o v,. a s - ~ ~ ~ ~ ~ n,. d. \ 0 0 0 I 4 *r 4 ci) 4 H H i ) cQ 02 0 UH P HH Q'd 1 4 Sa) p P S Q ^g 0~ I:: ~ o *H o o o ~ ~ o.el d) o 0o C) 4 c C) C O- ced'- ~d er m 4,-t-," 4, 0 ) -~ (D CO0 O. r 1. rO:d:: ~) j o., -d 40! H t O ~ 0 *1 q ( ( -O -O P VQ ( C 1 O,n'H'H *H hO. cH r F-i bO 4P.0 4o) H co 0 H'H pH CH F 4) F-~rPi c C 4) 0'H CH ai) $ Cd'H._. o.. v p o 4, 4 C:Hb c. *Hl et l CH F- O'H'HFi A:r C'd 0 t'p P A rO - b 0 0'i3 ed o ~. ~ o o E*e o o o, V1 o C4 O X z C -c O Cd i 0 C)'. q Z 0 C 4o o o ~,. ~ ~ O t4.d O2 <E P ed ~ 0 i ):4.O..P -P) )' ( @>. 44 - Or 0 C c C O r0 A H 0o C~d C CCp o,-p 4 ~r,- P o o'd dH 3 i-I o R - o ~ hD be *H C^^H~~ r o ~ o ~. o 4 )d o 00. g 0 40 O C H. Fl 4)4 4F OP 0 0 0 P -P 0r^ P ft Cd' 4) OH W P 4'-4 COH OH O C t bO C)4) CHO C) 0 ) O0 C' H H o 4- 0 4, C),0> 0 0 C C q rO 4)bO 4) C) P o o o 4 EQTJ -P hO F 0: H CO O 00V H^'< S cd ~0H 0CO2P 0 0 0C) C

85 C),P C)a C)d o - l^ o.. o o O CI A O 43U i c0 O hP 0 p P I O 4 U - H > r 0 0 OP (D O V O ) a) Q. ) C <C) 0- 03 0. oz o ~ ~o V w 1 ( Q.H IQ 0 A 0 C C 0 U) ) r4 A 4 o C ct'H bO *H O U) 0D H *H0 o H n t ) t A H * H Ci C 0 ^ rd 0 -U Ord P rr 0O O H 2 *r -P 4 1 J ft Q u rV q r UD D i 0 A'(D bD hO 4 H ci)a 0 )ro o o3 (~ c 0 H ( a0 r T OM' H -RPH 0 U 0) i H H (i cii H C 03 0 CO O 0 > -Z

86 2. X-ray Proportional Counting. In the course of these investigations various isotopes were studied which decayed by the electron-capture process. It was attempted to identify the radiations by their characteristic energies, as well as to count absolutely the x-rays emitted by a given sample (10, 39, 104, 105). An x-ray proportional counter was built for this purpose and various electronic equipment was constructed or acquired to operate the counter in conjunction with a pulse spectrometer. a. X-ray Proportional Counter: The operation of an x-ray proportional counter is very similar to the P-ray proportional counter described in section III-C-1 (26, 40). Some significant differences, however, shall be pointed out at this time. Interaction of X-rays with the Counter Gas. X-rays do not cause primary ionization but produce in the gas of the counter ionizing particles by either of the following three processes: 1. Photoelectric Effect: The x-ray entering the counter collides with a gas atom and ejects an electron from one of the outer electron shells of the atom. This photo-electron having a kinetic energy practically equal to the energy of the x-ray now acts like the primary

87 particle starting the counting process described before. The counter pulse, proportional to the amount of primary ionization in the counter gas, will be proportional to the energy of the x-ray. 2. Compton Effect: The x-ray entering the counter collides with a gas atom and transfers its energy to the atom which uses the energy for the ejection of an orbital electron. This Compton electron will behave as in the case above. The rest of the energy of the original x-ray will be emitted as a so called scattered quantum. This quantum behaving similarly to the original x-ray may collide with another gas atom and repeat the Compton process or eject a photoelectron. In the two cases mentioned so far the entire energy of the original x-ray is eventually transferred into ionization of the counter gas and a pulse proportional to the original x-ray energy produced. Pulses of this energy will fall in the so called "Photopeak" of the x-ray spectrum. 3. Photo-Effect with Escape-X-ray from Counter Gas: The third process by which primary electrons can be produced in the counter gas is primarily important for x-ra~ys having a greater energy than the binding energy of the k-electrons of the counter gas. Such x-rays may transfer their energy to the tightly bound K-electrons of the counter gas atom on collision. The ejected photo

88 electron will have a kinetic energy corresponding to the energy of the initial x-ray minus the binding energy of the IC-electron. The ionization it will produce in the counter will be proportional to this energy difference and the counter pulse smaller than that produced by the other two processes mentioned above. As the position of the ejected K-electron is filled again a new x-ray characteristic to the counter gas will be emitted. This very penetrating x-ray may not react any further with other gas atoms and escape from the sensitive volume of the counter. -The pulses created by this process will fall in the so called "escape-peak" of the x-ray spectrum corresponding to an energy equal to the energy difference between the initial x-ray and the characteristic x-ray of the counter gas. Design Considerations: X-rays are more penetrating than s-particles. An x-ray proportional counter thus must Fig. 30. X-ray Counter with Preamplifier

89 have a greater sensitive volume and contain more counting gas than a 3-raycounter. The former can be realized by building a bigger counter, the latter by putting the counter gas under pressure. The use of a heavier counting gas will also increase the efficiency of the counter. Construction Details. The design of the x-ray proportional counter built for this research is given in Fig. 31. In the Fall of 1954 the original design obtained from the Brookhaven National Laboratories was modified so as to simplify the machining and soldering and the counter built by the Physics Shop of the University of Michigan. A photograph of the counter is seen in Fig. 30. The counter was initially operated with a 90% argon, 10% methane mixture filling and connected to a commercially available scaler. Plateau curves obtained for different x-ray emitters by means of this combination are shown in Fig. 32. The counter was tested for leaks with a filling of three atmospheres argon-methane mixture, by submerging the entire counter in water. While all solder joints were found tight the Oo010-inch beryllium window obtained from the Brush Beryllium Company of Cleveland, Ohio, was found to contain pinholes through which gas escaped. This had been suspected since the thresholds of the counter when operated with the scaler had dropped considerably during the previous weeks.

90i Explanation of Fig. 31 I. 4-inch brass tubing II. 1 1/4-inch Kovar tubing III. Beryllium window made leak-tight by spraying \27 mg Glyptal lacquer on 2-inch Beryllium disk 0.01 inch thick. Window cemented to surface with Apiezon W wax IV. Stainless-steel plate to hold window in place V. Rubber 0-ring used for gasket VI. UG 560/u H-V connector VII. 3/32-inch Kovar tubing VIII. This space machined to fit plastic counting shelves IX. Edges rounded off with ball of hard solder X. Glass to Kovar seals XI. 0.004-inch stainless steel wire XII. 0.04-inch Ni tubing spot connected to counting wire XIII. Clean soft solder joints. Caution: avoid flux getting into counting chamber XIV. Mill surface after all soldering is done XV. Base soft-soldered to tube.

91 0;ft,c4~~ Ig 0 C) bO IP'I CC oo.j i'(x

92 30 Zns65 28,-26 S/xSe x 7x 26 >XI Pdl03 —/ 20 x::r~I I I I 2.9 30 31 32 33 34 ANODE POTENTIAL (kilovolts) Fig. 52. Plateau Curves Obtained with Argon - Methane Filled X-ray Counter

93 To seal the pinholes of the beryllium window a solution of red Glyptal lacquer in acetone was sprayed onto the disk in several layers. Intermediate drying was effected with the use of a 1500-watt heat lamp. The gain of weight of the window disk due to the lacquer was found 2 to be 1.35 mg/cm2 After replacement of the window, the counter was again tested for leaks. After several hours immersion no bubbles were observed either on the window or at any other point of the counter. Another test period of one month showed no change in threshold potential. The counter was then considered leak tight and ready for its permanent filling. The linear absorption coefficients for various x-ray energies were calculated from the available literature data(Fig. 33). It was estimated that for a 4-inch diameter counter practically 100% counting efficiency could be obtained for x-rays below 25 kv, if a 2-atmosphere krypton filling was used. A mercury pump and manometer assembly was then constructed according to the diagram of Fig. 34 and the counter filled with 151.15 mm Hg krypton and 15.2 mm Hg of 99%pure methane (Phillips Petroleum Company). Extreme care was taken not to contaminate the counter during the filling operation. Tests of the 2-atmosphere krypton-filled x-ray counter proved inconclusive since the plateaus as obtained

94 og - I I, E I10 _ \ z o and' KI Krypton L. C.)\ I 0.. 0 c,, Arg o n zI0 1O 20 PHOTON ENERGY (Kev.) Fig. 35. Lnear Absorption Coefficients for X-rays in Argon and Krypton under Atmospheric Pressure

95 Split pipe as support IS I C I_~150cc Pump Fig. 34. Diagram of mercury pump and manometer for filling the x-ray proportional counter with 2 atmospheres of krypton. with the Commercial scaler seemed to lie above 4900 volts. b. Electronics. Special electronic equipment to operate the x-ray counter and make best use of its potentialities was thus required. Particular emphasis had to be placed on the linearity and the overloading characteristics of the amplifiers. The preamplifier was designed with the assistance of Prof. Hough of the Physics Department and built according

96 to the circuit'diagram of Fig. 55. Fig. 30 gives a picture of the preamplifier connected to the x-ray counter. The preamplifier operates as a simple cathode follower of gain 1 with a large overload safety factor. The unit has shown very dependable performance for more than a year. The purchase of a suitable non-overloading amplifier for the x-ray counter proved difficult. It was first intended to build a double delay-line amplifier according to the design of Francis and Bell (38). The construction of _ 6J4_ _.,- +290 Volt )Z ^ -..-+ 15 K- 2W Input 15M f 15 70g.Ol 15M f Out 2.2M IOKV IM 8201 IM |.01 - 10 KV HV Fig. 355 Circuit Diagram of Preamplifier for X-ray Counter this circuit was contracted to the Beva Laboratory of Trenton, New Jersey, in early Spring of 1955. When insurmountable difficulties appeared in obtaining the unit in August 1955 a commercially available non-overloading amplifier was purchased. The design of this unit followed

97 the layout of Chase and Higinbotham (20). The amplifier together with a highly regulated 5000-volt power supply was obtained from the Radiation Instrument Company of Silver Spring, Maryland, and put into operation at the end of 1955. The amplifier as now available has a maximum gain of 40,000. The stability is excellent and the overload characteristics sufficient for obtaining good x-ray spectra with 14% half-width resolution for Sr K-x-rays. The linearity of the amplifier seems very good and distinction between the x-rays from Y88 and Nb92 is easily obtained (Fig. 55). Fig. 36. X-ray Spectrometer Setup One important feature of the RIC 107P non-overloading linear amplifier is a pulse-height selector permitting the introduction of a certain trigger bias. Pulses higher than this bias level are recognized and a square pulse

98 produced for direct counting in a scaler. The circuit of this pulse height selector was changed slightly to operate with the auxiliary equipment available. After these adjustments were made trouble-free operation was realized. c. Performance. X-ray Spectra. The x-ray spectra shown in Fig. 37 were obtained with the above equipment. The output pulse of the amplifier was analyzed by a single-channel pulse-height analyzer (Atomic Instrument Company No. 510, seen in Fig. 36) and recorded with a Leeds and Northrup Speedomax recorder. Spectra of the type shown in Fig. 38 were obtained by displaying the output of the linear amplifier on the screen of a Tektronix-type 514D cathode-ray oscilloscope and photographing the pulses with a Polaroid Land camera. Identification of the x-ray emitters was possible by either of the two procedures. X-ray Counting. When the decay rate of an x-ray emitter was to be determined a more quantitative method had to be found for counting the x-ray pulses. It was originally planned to scan the spectra by recording counts manually or semi-automatically with a step-sweep mechanism and to integrate the counts under the peaks at the desired energy. This was found to be an extremely laborious undertaking, especially for samples of low activity. Bad

99 y88 8- X-RAY SPECTRA 7 F- 6 Cr51 Zn65 oi 5 t,', o i 4' s'''', -n4 I;; ~z I I \_1 I II I I 3 < 2 0 2 4 6 8 10 12 PULSE HEIGHT (ARBITRARY UNITS) Fig 0 2 4 6 8 65 1288 Fig. 7. Spectra of x-rays emitted by Cr51, Zn65 and 88 as obtained with the krypton filled x-ray counter.

100 o ap O T 0 ) 0 4 (P 4 —! H Cd p gi ) I 4-3 CO or'-I

101 U,$ 0.& ~ S 8 V)c ~n X o 5 C')C --- I 0.0 H o~oh c a~ C < ~o OI' It 0 4_ 0~~~~~~~~~~~~~~0: cr~~~~~~~~~~~~~~~~~~~~~~I-I'Q)2C OL ~ b I —- - -- I I 0 4-r -w a-~~~~~> x > QL~~~~~c x l I I I, a.<~ ____ OS

102 statistics caused by the small number of counts and small drifts in the equipment over the long time intervals necessary to count low levels of activity, limited this method. To overcome these difficulties, a procedure was evolved whereby that part of the x-ray spectrum falling on the desired peak could be obtained in a single measurement. The peak was first located by a rapid sweep of the spectrum. By judicious setting of the pulse-height-analyzer base line on the high energy edge of the peak and the pulseheight selector on the low energy edge, it was possible to obtain two simultaneous counts, the difference of which gave the counting rate of the peak. A background count, with the same setting but without the sample/is also required since the background for a typical x-ray peak counted by the above method is of the order of 300 counts per minute. This represented approximately 25% of the total background of the large counter. A schematic diagram of the x-ray counting setup is seen in Fig. 39. When care is taken to keep the counting geometry constant, the counting rates obtained with this procedure are reproducible to within the statistical limits (cf. Fig. 48), Slight drifts in the equipment will introduce errors only in the valley regions of the spectrum, but these errors are generally much below the statistical error of the counts.

105 3. Gamma-ray Scintillation Counting Gamma-ray counting served in this research for the identification of tracers and bombardment products. Relative count rates of tracers were determined and calibration of the x-ray counter carried out with y-ray scintillation counters. Extensive work was performed with the scintillation spectrometer to verify certain decay schemes and follow the decay of y-ray emitters. a. Principle of Operation: Gamma-rays are a very penetrating type of radiation. Dense materials have to be employed for their detection and characterization. In addition to gas-counters, operating as proportional or Geiger counters(109) filled with krypton or xenon, most y-ray counting today uses the scintillation process. Solid scintillators containing preferably a high atomic weight constituent are employed. The principle of interaction between the scintillator material and the y-rays follows in its initial steps closely the types outlined in section III-C-2 for the x-ray counter. The essential difference is, however, that the energy of the ejected electrons or delta rays is not used to produce an avalanche, but is converted into fluorescent light which can be "seen" by a photomultiplier tube. The electric pulse from the photomultiplier is proportional to the intensity of the light radiation which in turn is proportional to the amount of energy spent by the y'ray inside the

1o4 scintillator. The probability of part of the y-ray energy escaping after previous Compton scattering is great for small scintillators and hard y rays. The spectra will thus contain a photopeak and a broad, more or less continuous plateau due to the Compton effect, cf. Fig. 54. This Compton plateau is separated from the photopeak by a valley representing the difference in energy between the total energy of the y-ray and the minimum energy of the photon scattered backward by the Compton effect (110), b. Design Considerations. Scintillation probes, including scintillation crystal, photomultiplier tube and preamplifier are now commercially available. Fig. 40 is a diagram of the generally used construction. The thalliumactivated sodium-iodide crystal used as scintillator has to have one highly polished side which can be optically coupled with the face of the photomultiplier. The other sides reflect the light and concentrate it onto the polished side as diffused illumination. For obtaining the experience of packaging a large rather hygroscopic sodium-iodide crystal a scintillation probe was constructed during this research. c. Construction. The cylindrical sodium-iodide crystal had a dimension of 1 1/2-inch diameter and 1-inch height. It was polished and packed by the following procedure:

105 PRE AMP LIGHT TIGHT TUBE SOCKET SING LIGHT SEAL (BLACK FELT) PHOTO MULTIPLIER TUBE DU MONT K 1177 LUCITE LIGHT PIPER _^ /, _^ / GLASS PLATE MgO 24 ST Al Na CRYSTAL Fig. 40. Schematic Diagram of Scintillation Counter

106 A polished aluminum can, having a 0.005-inch-thick bottom was introduced in a drybox together with the crystal and other equipment. The crystal was freed of its saw-marks and roughened on all but one side with emery paper. This was followed by a rinse in xylene to remove grit and excess sodium iodide. The optical side was then polished with cloth saturated with n-butanol until an even highly polished surface was obtained. The crystal was again rinsed in xylene and mineral oil and finally wiped dry with lens paper. The polished crystal was then mounted in the aluminum cup containing a thin layer of dry magnesium oxide as reflectant. More magnesium oxide was filled along the sides of the crystal and firmly packed into place. The surface of the crystal was cleared of magnesium oxide with a hairbrush and a drop of Dow-Corning Fluid 200 placed on its optical side. A 1/16.inch-thick soft glass window was then placed on top of the packed crystal making optical contact with its polished side and cemented in place with bonding agent type R-313. * The packaged crystal could now be optically coupled with a Dumont 6292 photomultiplier tube by means of a Lucite "lightpiper" and Dow Corning 200 fluid. The lightpiper was constructed from a cylindrical piece of Lucite whose one side was contoured to fit exactly the face of the photomultiplier. * Carl A. Briggs Co., 11616 West Peko Boulevard, Los Angeles 64, California.

107 Fig. 41. Spectrometer Equipment

108 The photomultiplier with mounted crystal was introduced in a light-tight housing containing a Co60 Nb92 y88 Fig. 42. Gamma-ray spectra of Co60, y88 and Nb92 taken at the same time with the same amplifier and oscilloscope settings (cf. Fig. 51) magnetic shield for the photomultiplier as well as a preamplifier. This "probe" was placed in a 2-inch-thick lead shield and used as a gamma counter (Fig. 41 foreground). d. Gamma Counter Performance: Several additional y-ray counters of the same design as described above were employed during this research giving generally very good performance and requiring a minimum of maintenance. Various spectra were taken with these counters in conjunction with the pulse spectrometer (Fig. 41) as can be seen from Figs. 42, 47 and 48. Whenever unknown y spectra were investigated

109 a known standard was counted immediatly before or after the unknown to serve for the energy calibration. D. Absolute Counting Methods To determine absolute cross sections the total number of nuclei produced during the bombardment has to be known. This number can be calculated, if the bombardment product, or a known fraction of it,is counted "absolutely", i.e., if for every disintegration occurring in the counting sample a count is registered or otherwise indicated. 1. 4w Counting From the earlier discussion of the 4w counters in section III-C-1 it was seen that absolute P-ray counting is achieved, if the self-absorption of the source is negligible. To realize this, only bulk-free sources were counted in the 4w chambers. It was thus possible to register all disintegrations by which a P particle of relatively high energy is emitted. a. Self-absorption. No corrections for selfabsorption were deemed necessary during these investigations Q3 2, yOO 90 44 47 48 for the counting of Y92, 90 Nb90 Sc S and P2 46 Self-absorption corrections were applied, however, for Sc 33 and Nb95. The corrections necessary for Sc were determined by Hall (55) for films of'the type used in this research.

110 The self-absorption correction for the P33 radiation was approximated with the 0.23-Mev P-rays of Pm147, The procedure was the following. The count rate of the Pm147 source of the thin-film gauge was measured for a certain setting of the Geiger-MUller tube (the window collimator was removed). A copper absorber was placed over the source and the P33 sample placed on top of the copper plate. The count rate of the P33 sample was then established with no interference of B rays from Pml47 The copper absorber was now removed and the Pm147 plus P3 count rate determined. The difference represented the count rate due 147 to Pm14 alone. From this and the count rate of the Pml47 source without absorber, the fraction absorbed by the P3 sample was established. It was assumed that half of this fraction would have been absorbed by half the thickness of the sample, which was the average absorber facing a P33 P ray emerging from the sample. This fraction was then applied as the selfabsorption correction for the P3 radiation, since the 147 353 energies of the Pm14 and P33 betas are almost identical. The self-absorption correction for Nb95 was established by approximation. The fraction of Pm 17 rays absorbed by the sample was determined as described above and the aluminum absorber thickness equivalent to the Nb9 sample determined. The fraction of the range of the 0.l6-Mev e raysof Nb95 represented by this thickness was calculated.

111 Applying a Feather analyzer (36) the fraction of b95 P rays absorbed by this thickness of aluminum was computed. Half-of this fraction was then applied as the self-absorption correction for the Nb5 sample. The error inherent in the corrected absolute counting rate of the Nb95 samples is estimated to be \2%, while the error introduced by the self-absorption correction in the case of P33 is less. b. Background and Dead Time Corrections. The customary corrections for dead time and background were applied for all counts. The first affected very high counting rates, the latter mainly the low counting rates. It was attempted to prepare samples of sufficiently high counting rates during most of their decay, so that the daily variation of the background would have little or no influence on the precision, even if it had been neglected. With backgrounds of the order of 90 c/m and a variation of 10 c/m, it was attempted to count samples with activities greater than 1000 c/m. This was found to be of particular importance, since it was observed that dust would collect in the inside of the counting chambers while remodeling of the Chemistry Building was in progress and effected rises in the background rate by approximately 10% during the period of several days. The advantage of higher counting rates as far as statistics and counting time are concerned is apparent.

112 Practically all counts taken during these investigations with the P-raycounters had a standard error of less than 1% uncertainty. When the decay had progressed to count rates much below 1000 c/m careful background determinations were made before and after a sample was counted in order to preserve the precision. As a rule of thumb it was found that a background count of the same length as the sample count would result in an absolute uncertainty smaller than that in the sample count, so that the statistical error in the background could be disregarded.'It might be of interest to note in this connection that for all counters employed in this research the background increased by a factor of about 40 on August 31, 1956, a few days after the explosion of a Russian hydrogen bomb over Siberia. 2. Absolute X-ray Counting. In the course of the investigation of the Zr90(d, a)Y88 reaction it became necessary to measure the disintegration rate of the 88 product, an isotope decaying primarily by electron capture. This required the calibration of the x-ray counter or, in other words, a determination of the efficiency of the x-ray counter for Sr x-rays. a. Calculation of the X-ray Counter Efficiency. To get an approximate idea of the geometrical efficiency, the

113 counter design was sketched on paper and the angle subtended by the counter window with its origin in the plane of the source drawn in. An average angle of 85~ appeared a good first approximation. The surface of a sphere of radius 1 is 4w. The surface of the segment of that sphere subtended by a rotated angle of 85~ is S = 2w * 1 * (1 - cos 85) 0.526. The fraction of the total surface subtended by the angle is tihus 0.526/4 or 0.131 i.e. the superficial geometrical efficiency is 13.1%. It was estimated from the drawing that the average length of the trajectory of the x-rays in the gas is approximately 2.6 inch corresponding to an 80% gas efficiency. From the beryllium absorption curve shown in Fig. 43 it was seen that Sr K-x-rays from Y88 are 97% transmitted by the 10-mil window making the efficiency for 88 K-x-rays approximately 0.131. 0.80 ~ 0.97 = 0.101 or 10.1%. In some experiments only the K-x-ray photopeak of the Y88 was counted. The percentage of all counts that are found in the photopeak was determined to be approximately 94%. This resulted in the calculated efficiency of the counter for the photopeak of Sr K-x-rays of approximately 9.4%. b. Measurement of Absolute Disintegration Rates by Means of the Coincidence Method. Attempts were made to determine the efficiency of the x-ray counter

114 W X 0o., 90 - 0 F 70 eo - 0 I40 \ \ I,,z 30 \ I 20 x-rays emitted by Y and Mn5. experimentally. For this purpose the absolute disintegration rate of a Y8 sample was determined by coincidence measurements. Coincidence Method. The absolute disintegration rate of an isotope emitting two or more radiations in coincidence can be established, by counting the radiations simultaneously in two counters and measuring the coincidence rate of the pulses from these counters (112, 115). It is not necessary in this case to know the geometry and efficiencies of the two counters as the following considerations show: If the efficiency of the first counter, counting only one type of radiation, is m for a sample in a given

115 position, then the absolute disintegration rate of the source is: D = m A where A is the count rate recorded in the counter. If n is the efficiency of the other counter, counting the other type radiation, then the absolute disintegration rate of the source is D= n B B being the count rate recorded in the second counter. The probability that a disintegration recorded by counter (a) is simultaneously detected in counter (b) is also n. The coincidence rate is thus C = A/n or from this D D mA ~ nB A. B D mnC C It is thus only necessary to establish the count rates in each of the two counters and the coincidence rate between them, to arrive at the desired absolute disintegration rate of the sample. Absolute Disintegration Rate Determination for an 88 88 Y Sample. In the case of y8 there are three different y-rays emitted in coincidence with the x-ray during an electron capture process (Fig. 44). The branching ratio between the two possible de-excitation paths was measured and found in agreement with the literature (97) to be of the order of 1%. For the purpose of the following

116 y88 A(4r 105d (3) 1.5x10 o6sec /0.2%+ 0.908 / 0.83 Mev Mev (2+) 2.76 1.85 Mev Mev 1% 99% o+ Sr88 (stable) Fig. 44. Decay Scheme of Y88 considerations it was assumed that this ratio was negligible and practically all transitions followed the path of the y-ray cascade. To determine the absolute disintegration rate of 88 the Y88 sample, coincidence measurements were carried out between the x-rays and the O.9-Mev 7 rays, the x-rays and the 1.83-Mev 7 rays, and between the 0.9-Mev and 1.83-Mev

117 Linear Single channel lPr eamp a mplifier P-H analyzer Preamp -ry Scaler I Scintillation I y counter i ceCoincidence used for te x yanalyzer Anticoine Chnnl rcodete aypusepodcd by t hscoler X-ray proportional counter _ Scaler Preomp I thel_ x a pSingle channel P-H analyzer Sc aer -- m __ Non -overlooding linear amplifier Pulse height selector BLOCK DIAGRAM OF ANTI - COINCIDENCE CIRCUITRY FOR ABSOLUTE X-RAY COUNTING Fig. 45. Block diagram of x-ray — Y coincidence counter y rays. The block diagram of Fig. 45 indicates the circuit used for the x-ray-7 coincidence measurements. Channel 1 recorded the y-ray pulses produced by the Y rays of the desired energy in the scintillation counter. The channel width and base line of the pulse height analyzer was set for this purpose to include the entire photopeak, but exclude y rays of different energy. Channel 2 recorded all pulses produced in the x-ray counter whose heights were greater than the lower edge of the x-ray photopeak. Channel 3 counted all pulses from the x-ray counter whose heights exceeded the photopeak x-ray pulses. The true count of the x-rays in the photopeak was

118 obtained as the difference in the count rates recorded by channel 2 and 3. Coincidences between channel 1 and 2 might be due to a coincidence between an x-ray and a y ray of desired energy. In this case the coincidence should be considered. It might also be due to the coincidence of an x-ray with a high energy 7 ray which interacted with the scintillation crystal by a Compton type interaction producing a pulse of the same energy as the photopeak of the desired 7 rays. Since the y rays in the case under discussion are in cascade,this effect does not disturb the result of the measurement. The coincidence count might, however, be due to the coincidence between a y ray of desired energy and a y ray of the other energy detected by the x-ray counter. Such pulses interfere with the measurement and should not be considered. To avoid counting these coincidences channel 3 was coupled in anti-coincidence with channels 1 and 2 so that y-ray pulses produced in the x-ray counter and triggering channels 2 and 3 simultaneously would not be considered in the coincidence rate. Corrections were applied to all observed rates for background and random coincidences. The latter correction was determined experimentally by using the same circuitry, but counting two different sources simultaneously in the two counters. The counters were separated by about 3 ft and several lead bricks were placed between them for these determinations.

119 Phototube Phototube' ~Linear ----— ^ ando and~ - -i' Linear A f -r Prenp r L p ap i I *Amplifier, Requlated Power Supply Regulated Power Supply Singe O I Coincidence Sin le Channe PulseHM -----------------— r Pulse Height Analyzer Analyzer r! Scal- r Scaler Scaler Fig. 46. Block diagram of y-y coincidence counter For the y-y coincidence measurements the circuit given in Fig. 46 was used. One of the scintillation counters counted only the photopeak of the 1.83-Mev y rays, the other the photopeak from the 0.9-Mev y rays. The counter sensitive to the 0.9 Mev y rays, however, recorded some of the Compton pulses from the 1.83^Mev 7 rays as well as the 0.9-Mev photopeak. Corrections for the Compton fraction had to be applied. The value of this correction was determined experimentally from the y-ray spectrum of the source taken with the same counter in the same geometry just before the coincidence measurements

120 0 a Q O H) H c) 0 CO CO cd GO cH H a O5,d rO 0oo > 4 - PP + S (+ o rcV o0 Q0 cO CO O kO i k O CO -- Co0 o L-' 4-) H b - O en (D o.p U a o b- o h O kO -P a O GH ci.. - P'. 0 H O 0 0 H O -PbOCd 0a ~ 0 0'- - O-l Q C d -jC * 0 o 0 (D -H -P 0 E..co Cd -P 0 H 0 <D +3~~~~~r1 H1, 0 rs\ 4 C — d cCd \: rc-\ rC\ r -P H H H H H I I H t 0) -P (D 0 0 0 0 > H @ Z z S: o 00 N Cd c 0> o oo c tf EH ) *0 * o* * C 2o 0 0 0 H 1 * Q ^.,,, 0 H CO dCd rO a0 rd aI 0 2 ( 2 2 C 2 D C d D C os 03CQ 0353 35 0 3 0303 03 0 0 0C0 C cd ct ct c( c^ cd c^ ct Ca a U c 0 I I I I I I I I > * I > O X X - X -- X 0r c - 0 c c O

121 were performed. The Compton plateau having a higher energy than the 0.9 Mev photopeak was extrapolated to the energy of the photopeak and its values for the range of the photopeak integrated. It was thus possible to determine the number of Compton pulses stemming from the 1.83-Mev y-rays and having the same height as the photopeak of the O.9-Mev y-rays. The results of the three attempts to determine the absolute disintegration rate of the Y88 sample are given in Table VI. It is seen from Table VIthat the absolute counting rate of the 88 sample as determined by the x-ray — coincidence methods is about 1.5 times the decay rate found with the y-y coincidence method. The values of the efficiency of the x-ray counter vary accordingly. Attempts to Calibrate the X-ray Counter with Snll3 and Mn54 Samples. To verify the efficiency values obtained for the x-ray counter from the above data, x-ray —y coincidence measurements were carried out for samples of Sn13 54 and Mn. The x-ray spectra of both isotopes indicate only one y-ray emitted during each electron capture process. (Fig. 47) It was thus expected that these y-rays would be emitted simultaneously with the x-rays and coincidences between them obtainable. The results of the coincidence measurements carried out with the two samples are given in Table VII.

122 0 0 0 0 00 L DO - in cJ H ^ * Orl C) CJ C 1 00 0 0' — O r - | c~ 0 0' r in <M en S a) o OJ Cn o 0 a CHX C n O V C a p -P 0 4 K 0 0l * O 0 -, 0 od rc*H. - 0 0 0> 0= dc4 O O4r C q C0 K>\ C ^ o ^ CM -.H C O O c n ~ c H 0 D oD 0o 0 0 0 -) fi rQ K0 t -c h- & KP Fr'C r-4 4 0rPO~ cD t- m K KG 0 z K K> H H rc 6 c6 cr,p orp -p H0 j -- "Cd -- d - IO rcl 0 OD C^ 0^ H r.. 1 1 4 a) o rrf K > n ^ OO r r tH t-n

123 Fig. 47. Gamma-ray spectra of Sn113 Mn54 95 and Nb95 taken with identical settings of amplifier and oscilloscope A comparison of the values obtained for the x-ray counter efficiency for the x-rays emitted by the Sn11 and Mn4 and Y samples shows their great discrepancies. (The great fluctuation for the two values of the Snll3 experiments is due to the small number of true coincidences that could be collected). To facilitate an explanation for these discrepancies the efficiency values of the y counter might be compared for the 0.9-Mev y rays of y88, the O.26-Mev y rays from

124 Snll and the 0.84-Mev 7 rays from Mn54 For the sodium iodide crystal employed,a greater efficiency would be expected for the 0.26-Mev y rays from Snll3 than for the other two, of which again the 0.9-Mev 7 ray from Y88 should yield the smaller efficiency. Comparing the experimental values it seems strange to encounter a variance of 10 and 100 in the opposite direction. The following explanation is proposed to clarify the situation: after the electron capture process the nucleus is in a highly excited state. Two processes have to take place before a stable state is reached: the level of the electronic shell has to be refilled with emission of a characteristic x-ray; nuclear excitation energy has to be released by the emission of y rays. To explain the experimental data it is assumed, that these two processes can be delayed with respect to each other. The x-ray is emitted after a very short time interval compared to the resolution time of the coincidence analyzer (1.94 microseconds). The nuclear de-excitation may, however, be delayed. In this case there exists an intermediary state, decaying randomly with a certain half-life. As point in proof it may be mentioned that several cases of metastable states as intermediaries in electroncapture decay processes have been reported in the literature (46, 67), Metastable states seem to be encountered in all three of above cases. In this event the

125 number of coincidences observed is only due to those y rays which are emitted while the coincidence analyzer is still sensitive after being triggered by the electron capture x-ray. The shorter the half-life of the metastable state, the more y rays will still be considered and the truer the obtained absolute rate. The latter, however, will always be higher than the actual disintegration rate. For 88 a fortunate case - two y rays are emitted in cascade (Fig. 44), giving the true disintegration rate within 1%, if no intermediate of appreciable half-life is assumed for the cascade emission of the y rays. This assumption is based on the agreement of the values obtained by the x-ray —y coincidence measurements seen in Table VI. The difficulty in determining the absolute rate even from the 7-7 coincidences is the precision with which the Compton background can be determined. This background has to be subtracted from the 7-ray counting rate. As the part of the counting rate which is due to Compton pulses is rather large, an error of about 8% is inherent in the absolute rate determined for above Y88 sample. From the experiments described above it appears that approximately 52% of the v rays are emitted during the resolving time of the coincidence analyzer for the Y88 This would indicate a half-life of 1.5 x 10 seconds for the metastable state. The half-lives for the intermediary stater in the decay of Sn3 and Mn would be of

126 the order of 85 and 41 microseconds respectively. These estimates were made with assumption of a 6.9%x-ray counter 88 efficiency as found from Y8 For the absolute counting of Y88 in the present research it was decided to use the value obtained from the r-y coincidence measurements as the true disintegration rate of the Y sample and assume the x-ray counter efficiency to be 6.9% for the photopeak, The value thus obtained experimentally agrees reasonably well with the calculated efficiency of the x-ray counter.

Part 3 PROCEDURES AND RESULTS The instruments used for determining the (d, a) reaction cross sections have been described in the previous sections and a general discussion of the procedures for the measurements given. A detailed report of their application to the experiments will be presented in this part. The experimental methods and results are treated for each reaction investigated. The choice of the target nuclides was governed by the following factors: 1. The value of the experimental results for theoretical considerations. 2. The ability of the 7.8-Mev deuterons to induce the reaction. 3. The availability of the parent isotopes. 4. The detectability of the product nuclide. 5. The knowledge of the decay scheme for the product nuclide. 6. The possibility of resolving the expected decay curves. The final choice made an optimum compromise of all the factors involved. 127

128 The reactions investigated in this research were the following: Zr94(d, )y92 * Zr92(d, a)Y90 * Zr90(d, )y88. Mo98(d, a)Nb96 * Mo97(d, a)Nb95m Mo97(d, a)Nb95 Mo92 (d, )Nb90 Ti48(d, a)Sc46 Ti46(d a)Sc44 S34(d, )p32 * The absolute cross sections for 7.7-Mev deuterons have been determined for all these reactions. Excitation functions were measured for the reactions with asterisks. Two of the above reactions had been investigated previously by another worker (47), but warranted a reinvestigation as the errors reported for them were rather large. I. Bombardment Procedure Targets were prepared by the methods reported in section II-A of part 2, and placed in the slots of the target probe assembly (Fig. 3). Several collimators consisting of empty aluminum target frames were placed before the target to reduce the

129 activity induced in the target frame, and behind the target to reduce the effect of secondary electrons emitted from the target. The current integrator was warmed up for approximately one hour and tested if it was functioning normally. Current integrator readings were taken manually at various times during the course of the bombardment to have an estimate of the variation of the beam strength during the bombardment. After the bombardment the drift rate of the instrument was determined to apply a correction to the final currentintegrator reading. This correction was usually less than 0.02%. Occasionally a 1/2-mil Mylar scattering target was kept in the path of the beam (Fig. 1 N) during the bombardment to facilitate the focusing of the beam on the target. The reduction of the deuteron energy by this scattering target was calculated with the aid of Table I and considered for the final cross section values. After bombardment the target and the substrate were dissolved and the (d, a) reaction products separated. The final counting sample was prepared on VYNS-films (section II-C-1 of part 2) and its decay followed with the 4w counters and the x-ray and y-ray counters.

130 II. The Zr(d, a)Y Reactions The products of the (d, a) reactions on zirconium include four yttrium activities: the 17-minute Y94, the 3.5-hour y92, the 64-hour Y90 and the 105-day Y88 This can be seen from the chart of the nuclides for the zirconium region as given in Fig. 48. The Y94 decays rather rapidly with emission of a very strong 5.4-Mev v ray and was not observed during these bombardments. It had usually decayed out before the first counts were taken on the 4w 92 counters. The 3.5-hour Y decaying with emission of 3 rays having maximum energies of 3.6, 2.7 and 1.3 Mev and several 7 rays in coincidence with them could be well observed and the cross section of its formation determined. Yttrium-90 decays with emission of one P ray only, having a maximum energy of 2.27 Mev. The 105-day Y88 decaying by an electron capture process and the emission of a weak 0.835Mev positron merited some special attention in this research as it posed various interesting problems. A summary of the characteristics of these isotopes with the literature references is given in Table VIII. Cassatt (18) showed that it was possible to obtain reasonable yields for the Zr(d, a)Y reactions with the 7.8-Mev deuteron beam of the Michigan cyclotron. He was able to identify all half-lives expected and resolved them from the gross yttrium decay curves. An attempt to

131 %D a 0 C a) E0r ~E b fe^S| Is t I ^ ~ ^< u d) %0 S =^^s~sl * A s I A -p H 4-o 4o-.dO n', M - ^. *~ ~ ~ ~ ~~~~O I> ~ O(.cI*(:C., (D o ^^^^^*^-WL-lOT-l&^^-,J OH 4 04-3HH -- - -- -- 1-Ji —iiii f -^-^H <O.D r -t~ 0 Q-H 02 N W idD~(~~1I ~ 0) 16" a) (1 ) 0t3 Q ) Z:~r s^*a^ -^ ^*H~ra-HM~ocD^ oc~ -D T1, ^ *1.0c I0P PI> P.o -P I - a) | 01) <* - C < Q)^'!-or-:!,r.,~. > —P-' ^ s ~ ~, z w ~~c, O ~re ~rl ~~OO t>' ~ a)o a) o ~o -,. z oa Cd~~~~~~~~C ift(P NT Vli8fi1 *(M I 5Co~~ <uccO Qu_,0I "'^*^0)~o:^^ ^^^^. ~. ~ ~ o. r0OH o 0.!' ^'S.^5* f^^ "|<^ <-?'^ * ~ 6 A | (1) 4) ct 35 r ) a 0>8'"0'(3 0>^>6:-l ^1<?>.S'"<;; * rd I> 4 >$-H O 4-3 C) 0 4,) In H r n,-t a, Ga, d' ~ ~ ~~~~ -I r r =C dE'(-' 0^~~~~o -HI Q 0 0 a ~~~~~~ W a ~~~~~~~~~% C Id ~~_ in -P a) q_, C) 03-r c6Pls a) a rdW oq[ z 4 _H Cd I 1'~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ HiTT a). 4 —l~' H~B~>3o + oi-a) C: -o Qa m I 44-> o' pi^^ ^K csso sg ^ ^^ ^c u ^-^a,.H' b) in v Nr N 40 3" in bo (1) cd a) rd qb)E-erd -, CT J- ~ o ~ e i'SS.^g c 03 53 4 | —: ot.L OD a).....iF-.,-i H~) 0 -a CY a)'04 9: d a) a)^S< ^>~o ^ rd -0 — "^-^~ %W 40^^ in V. 4R * 4 <D (D ^ ^ 02) Z<C >- lpP'1-'~^TJ~riO) a) 0 O10Cy 0 a)^~r ~t 4.- ) 0 a ) ^M z^ ^ V)^_ g^^^ ^ ^-H o* -P~a-HO^(DO, ~';' - —!" —i'?::;:-,: T;:::i~~i,!'~;; ~~~~~~~~~ n oC ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~W- 3C rl a) 4 a~3 ~~I ~~ 21 1% V) o, 9~~~~~~~c; crt ~~~~c, ~-H a)' 4c~ i 4 40 "d PO 4-) a) co 0 rd- - L~~C.~~S P;~r( u3C.)Cd 0 E P a) 4- ~' ra a SC~1 )C ^N^^<i^^^ ^^;^:^^ 0^ P CH rd ^ P -::.:;.. x^ )4^l W aF -/;.-P-:.r.-..a O~~_ N a) Cd OH -P. 4-oC1 3 4-3 I~: o ~ - C.) C) 40 CH H4c -, 1 T ~!~~~~~D 4 m Cd a) IQ0 a) iii~i~i ~ rC)'ra z ao 3) a) d = i;~;: o, I!; " Cd -- 0d 4 —') a). 5- — i CaS 1 4-~)8? "- C-) I'l a) 0,l ia~s a c0 "d o r r- ~ P..,-H q P SW.... Z rd r ~ d 4-)M d,'P O rd r\~~~~~~~~~~~~~~~~~~~~d a:, o D 4:j-: —...:_i -—' -.'. ~-."' IHI^ ^ ^ —'..:c.:...'H-.P ^^il f,'- ~ ~,a > n Q'.... ~ i;...;....:.:.-:..,:::: 0 r - ^^' ~,,,,'s-.C "~"~::(.i:;/~:;:'~-iL!i:. a. ed Q ~4 o ~) n a~ o?..s N" s 7".r-."~'.'.H 04 a) 0 C'v - CJ a) 4-3 CJ2 a ~CC oil ty Cdiii ~ a)~~~~ a,1 s~ ~~' o ~,HT. OD W a 0 03 P 4 cm g IL00 z 1) v:~I:~~~~~ -~~~~~~~~rd a, a, ~~~~d 4 a) 0 Cd d ct e 4 rs r Cd YI.'u:":'D III —O b~'z ~a)' " Fe.' -~ ti kg ( C) aD "i 4 P In co q-i co c -95 M, Q ~

132 CQ OH c 0 4 -D O~ ~ CO CO CtJ 03 - OJ/2 Ol Ci CMC 0 O H H H.. H 0o b ( o 0 0 r 0 - 00 0 -.OH o-l.tO 4D 0 o, z: O, ar a a 3, c =, I,r, H o 0 *H* + c O + r-i r-.H C 4o 4o 4 olO Hn 60 O, ~0 0 0 OH O H0.4 Ci co o~ co ~ u ~u I~ ~O Oo H) H* 0t 0 Cl)'H r'lcri> C H o' Q C + 0 OH*4 4 ) H Cl) cl) CO 4 3 CM 1-4 4 4 0 44 444NC) 44)0> C ) H 5'0 H' 0 9* H 0 CC * 4'H 0>) 4+O1 CO HI-d 0.4 H KOl\44 -1 + 0 0 0 H *

133 evaluate the cross sections, however, had not been made until this research. A. Target Preparation Some attempts were made to obtain thin uniform zirconium targets by evaporating zirconium metal according to a method described by Lillie and Conner (77). The results were rather disappointing. Zirconium foils, 0.00013-inch thick, were rolled by the Arnold Engineering Company from a 0.002-inch stock obtained from the Foote Mineral Company of Philadelphia, Pennsylvania. A spectrographic analysis of these foils was performed by LEDOUX and Company Inc. of Teaneck, New Jersey on April 4, 1957, with the following result: Zirconium..............99.85% Hafnium................none detected, less than 0.1f. Circular target disks of defined area were cut from the foils with a stamp-die. Their average thickness was determined within 0.3% by weighing on a semi-micro balance. To check the uniformity of the foil-thickness the absorption of the Pml47 3 rays of the thin-film gauge was determined at five points. The variation of the amount absorbed remained within 2% for the majority of the foils used.

134 B. Chemical Separation The (d, a) reaction products were separated from the bombarded targets by the carrier-free procedure given in Table IX, The procedure adheres to the following outline. The tracer, previously deposited on Zapon film and counted, is dissolved together with the bombarded target and isotopic interchange obtained in a homogeneous solution under severe conditions. Primary separation from the bulk of the target and radioactive side products is achieved by two types of simultaneous precipitations. The carrierfree yttrium fraction is finally separated from its zirconium carrier and any interfering activities by an anion -exchange step. The carrier-free yttrium activity contained in the eluate is regained by evaporation of the solvent. C. Tracers To determine the chemical yield for each particular separation following a zirconium bombardment tracers had to be used. 1. Yttrium-88 The 105-day y8 decays predominantly by the electron capture process and was well suited for this purpose. Preliminary bombardments indicated that the y88 count rate obtainable in the 4w counters from the yttrium product of

135 a two-hour bombardment represents only 1/10,000 of the counting rate of the product 4 hrs. after the bombardment. 88 This fact made the use of Y as tracer very convenient. To obtain enough pure Y88 activity for tracing the chemical yields a bombardment of several grams strontium oxide was performed with the 21-Mev deuteron beam of the cyclotron at the Argonne National Laboratory. Yttrium88 was produced in relatively large amount by the efficient Sr (d, 2n)Y88 and Sr87(d, n)Y88 reactions. The shortlived activities resulting from this bombardment were permitted to decay out for 11 months. The yttrium-88 was then separated from the target material in carrierfree form. The procedure for this separation involved 00 several simultaneous precipitations of the Y8 with zirconium carrier followed by an ion exchange separation of yttrium from zirconium similar to the procedure given in Table IX. 88 88 a. Identification of Y. The resulting Y88 was identified by its x-ray spectrum seen in Fig. 37 and Fig. 38. A half-life determination over a period of several months (Fig. 49) with several samples resulted in values of 105 + 1 days agreeing with the literature (97). 88 Investigation of the y-ray spectrum of Y resulted in the picture seen in Fig. 42 showing the 0.91-Mev, and the 1.87-Mev lines. 6The Co spectrum taken with the same amplifier and The Co spectrum taken with the same amplifier and

Table IX Chemical Separation of Yttrium from Zirconium Targets (carrier-free) Element Separated from Decontamination Factor Y (carrier-free) d. bombarded Zr -105 Time of Separation Chemical Yield ~3.1 hours 85% Chemicals and Equipment 4 Pyrex beakers (50 ml.); centrifuge; centrifuge tubes 40 ml., 15 ml., (Prex); Lustroid tube (10 ml.); micro pipette 50; platinum wire; small anion exchange column (Fig. 20); medicine droppers; calcium carrier (10 mg/ml); zirconium carrier (10 mg/ml). Dowex 2 (200-400 mesh); conductivity water; cone. H2S04; cone. NH4OH; cone. HC1; 30% H202; HCl-gas tank. Procedures (1) Cut the tracer yttrium, depositedon Zapon film, from aluminum sample plate and place in small beaker, add bombarded zirconium foil plus Mylar substrate. Add 1.5 ml. cone. H SO and a few drops 30% H202. Heat to fumes until dissolved (cool and add more H202 at intervals). (Note 1) (2) Transfer clear solution to a 40 ml.centrifuge tube, add 30 ml.H20 and precipitate zirconium hydroxide with cone. NHhOH. Stir with platinum wire; centrifuge and wash twice. (3) Dissolve precipitate with a minimum amount of conc. HC1, add 3 mg calcium carrier and one drop NbCl carrier (10 mg/cm). Transfer to 10 ml. Lustroid tube. (4) Precipitate CaF (and YF ) with 3 ml. cone. HF. Centrifuge (water in centrifug~ cups ) and wash twice. (5) Transfer precipitate to a 15 ml. centrifuge cone, centrifuge, decant,add 0.5 ml. cone. H2SO4. (6) Heat to fumes to drive off HF. Cool. Add 3 mg. Zrcarrier and dissolve residue in 10 mlo warm water. (7) Precipitate zirconium hydroxide with cone. NHOH. Wash three times with conductivity water. (Note 2) (8) Dissolve precipitate with a few drops cone. HC1 and saturate with HC1 gas.

137 Table IX (continued) (9) Transfer solution to a small anion exchange column charged with 1 ml. Dowex 2 resin saturated with cone. HC1. (10) Adsorb zirconium onto the resin slowly. After liquid level reaches resin bed add 5 drops of conc. HC1 and permit to soak in. (Note 3) (11) Elute carrier-free yttrium reaction product with 5 ml. cone. HC1 at a rate of 1 drop in 7 seconds. (Note 4) (12) Collect eluate when activity starts coming through. Evaporate to near dryness and plate for counting. Notes (1) A homogeneous solution is obtained under rather severe conditions. Complete isotopic interchange between the Yoo tracer and the reaction product is thus accomplished. (2) This washing removes the calcium carrier and must be done thoroughly,if carrier-free yttrium is to be obtained. (3) Try to wash the walls of the column free of activity with these 5 drops. (4) The yttrium is not absorbed by the resin in hydrochloric acid but is easily eluted, while zirconium and niobium are strongly absorbed from a conc. HC1 medium.

138 8000 ~8000~ DECAY OF YTTRIUM- 88 7000 (X-RAY COUNTER) 6000 5000 4000 3000 105d Y88 2000 00. J ~ 6000 o 5000 (4r COUNTER) 4000 Fa 3000 O ^^^o^^ 105d Y88 2000 H 000 0 9 18 27 36 45 54 63 72 81 90 99 108 117 126 TIME (DAYS) 088 Fig. 49. Decay curves of y taken with x-ray counter and 4r4 counter (two different samples)

159 oscilloscope settings served for the energy calibration. Yttrium-88 Decay Scheme. In 1955 Alburger and Sunyar (2' reported that they could not observe any of the 2.76-Mev y rays reported by W. C. Peacock and J. W. Jones (97) (cf. Fig. 44) The picture of the spectrum obtained with 88 our y88 (Fig. 50), however, clearly indicates the presence of this y ray. This picture as well as that of the Nb92 spectrum seen in Fig. 42, was taken by exposing the film for two minutes to the image of the two lower y rays, then adjusting the trigger level of the oscilloscope to block out the traces of these energies and exposing the film for an additioralhour to pulses resulting from y rays with energies above about 2 Mev. The Co standards taken immediately after with the same amplifier and oscilloscope settings served for the calibration of the energies. Manual counting of the peaks indicated a ratio of 0.0027 between the 2.76 andO.9-Mev peaks. The latter was corrected for the Compton pulses stemming from more energetic 7 rays. The experiment indicated a branching ratio of approximately 1 - 2%, if the efficiency of the 1 x 1.5-inch NaI crystal is taken into consideration for the two energy ranges. These findings together with the experimental results reported in section III-D-4 of part 2 verify the decay scheme for y88 as given in Fig. 44.

888 co60 Fig. 50. Gamma Spectra of y88 and Co

14i b. Counting of Y88 Tracers. The identified Y88 was used for tracing the reaction yields of the Zr94(d, a)y92 (d, x)88 and Zr92(d, a)Y90 reactions. The desired amount of Y was deposited on a thin Zapon film and covered so as to form a weightless source. The decay rate of this source was determined by counting on the first shelf of the x-ray counter. For the first few bombardments manual counts were taken in five-minute intervals covering the entire spectrum with a one-volt window. The area under the main peaks was integrated and this value used as the count rate of the tracer sample at that day. Errors of at least 3% were unavoidable by this method. For later bombardments the tracer Y88 was counted by the more exact and convenient method of counting the photopeak by difference as was described in section III-C-2 of part 2. 2. Tracer Y90. It was of special interest to determine the cross section of the (d, a) reaction for the magic number nuclide Zr90. The reaction product for the Zr90(d, )Y88 reaction is here the same 105-day Y88 described above. To trace the chemical yields for this reaction, Y9 was used as tracer. The latter was isolated from several millicuries of SrY90 by 10 coprecipitation steps with zirconium hydroxide and the final separation from zirconium by the ion exchange steps mentioned above (Table IX).

142 The carrier-free Y thus obtained was plated on Zapon film and identified by its half-life of 64.18 + 0.1 hrs. over a period of 25 days. To establish chemical yields the Y90 tracer as well as the separated samples from the bombardment were counted with the thin-window proportional beta counter in the same geometry.

143 D. Record of the Bombardments Six bombardments were obtained to determine the cross sections and excitation functions of the Zr(d, a)Y reactions. 1. The Zr92(d, a)Y90 and Zr(d, a)Y92 Reactions. The products from these two reactions are P-ray emitters and could be counted absolutely with the 4w counters (Table VIII). The dependability of the chemical separation procedure given in Table IX was proven by bombardments 4 and 7 and the relative yields for the formation of Y92, Y94 and Y88 established for a two-hour bombardment with 7.8-Mev deuterons. The 4w counter decay curve for bombardment 7 is presented in Fig. 51. The three components into which the curve is analyzed can be readily identified from the gross decay curve. The result of this bombardment helped to estimate the amount of Y88 tracer that should be used for the absolute cross section experiments to permit resolution of the decay curves and a minimum error in the yield determination. 88 It was found that an 88 sample counting between 10,000 and 15,000 counts per minute on the x-ray counter would well serve this purpose.

0 I-._CO z ed (00 0 0 (0 h-i // - i^ - o// o -o u o 0. O -D'1I8 1 O ~~. -c w 0f 10 C.j ___ _ —-- ~~ oj Q F0i 0 0 0 N 03inNIIM 03d SINH0

145 The absolute cross section for the Zr92(d, a)Y90 reaction for 7.56 + 0.05 Mev deuterons was measured by two experiments. In bombardment 10 a zirconium foil was exposed to the full energy beam behind a 1/2-mil Mylar scattering target. The primary experimental data for the experiment are reported in Table X. The major source of error for this experiment was the uncertainty in the chemical yield. This was primarily caused by the diffi88 culties of counting the x-rays from Y which were counted by integration under the photopeak of the K x-rays. This error could be reduced in bombardment 12b where the counting of the tracer was performed by the method described in section III-C-2 of part 2. The agreement of the values obtained for the cross sections from the two bombardments: 3.76 + 7% and 3.82 + 6% millibarns respectively, is considerably better than indicated by the estimated experimental errors. The two experiments constitute the only case where a double determination of a cross section was performed, but they may serve for an estimate of the precision of these investigations. The absolute cross sections for lower deuteron energies (6.97 and 4.78 Mev) were established by bombardments 11 and 12a. Difficulties were experienced with the resolution of the decay curves for bombardment 12a due to the low counting rate of the Y90 component. A better yield determination for this bombardment partially off-set the effect of this error on the cross section value.

146 02 r9- O O1 0 le ~'X Ck ~ O -t O H - t H H + + + CH O 0 0o r' a> 4 0 b ^ ^ f o +o + G m m -c- in o H D t- 1 O n W H 0 0 ^ C\ H 0 0 0 0 I m T I c Iu 4,!2 - c rdC o' N o QXMo C-. 0 00 ~. Q N. E r I-P H -Cp (D C) L0 o 0C +I Ln +1 +1 k9 +j +1 <i SbOo\0, i H H K) H 0> GO'Co M (\ H C....... o o E > r> - 4 -o - UTr n ^ a) a 4 C "~0 4O. +. H 0 W 4 p) 0-L 0 0 n 0 0 0 n 0 CM: o' H 0 + + +1 1 p 0 0 O0'C~ O C^ O +Ln> 4 r * O P= 0c -4 H H 4 0 0 H Cr M o 0 Td ~ 4 ^ d a) ~~~ ~ H~~~ a4 __ ^ tS. N -, _ ga. I+ + 1i +1 +1 1 a)+ o N. cl +L M + r p o c a) a)o n \t- 0 o o ) Oa rl O +1 a) HH H H m.O O I OI+11 +1ro ) o0 0 i I C +I I 00 0O O a oo o o fEI~ 00 0 0 0 C~ X =rS S= ^r -T t *. H 4 a. 0<) oao a) 0 E-o cu a) mr-t -Pa)m - - rt rta a) O 00 0 0 I H H i H boa) CD0 o t \<0 t —- M Go 0> 0 Gd O c Ln Ln Ln 0- c o O -n O C OD> t~FlD L -t Cj O O a o 4ra) C H H r-H r- H H (D) p., oQ,D a) ( Q a a) a) 1 a) a) 0 CCM \ H' \ CM \ ao o oj r rf r f rl rr OF9rl -l - rl -t - rl -

147 d o o h- c cu l0 o o Cd ^ ^ b 0 0 N CM HCM KN\ 0 0 - 0 >- rH H + +|+ +1 iH (H CM (H b Hr + H +| +$ +1 o$ n> o 0+ 1 +1 +1 +1 H 0o + 0 C\ 0> CM C\ o CO \ 0 0 H c0> N C 00 M O O 0>D n O s o c Mo 0 0 o C o U2CM..... -P(I) 0a0) OQ. 0000 o kID r- -I Oh CO COO O r" - OO 00 0 v +1 CM +1 CM CM |1 CM CM +1 +1 +1 +1 O $ \ LB +1 LB +|I +| 0> +| +| CO 0 LB u -?-cOV H 0 H 0>'O > CM CM O CO H 0> 0 S LB LB LB 4 4X 4'4 L 4F4 LB B cd -HS F Ln n n t n n n m rn oP r r r r r^r r i rC0 +- +1 $1 + + + + + +1 +1 + +1 0 *, r O c 0 0 0 LB kN D 0 0H CNM t-4 O Q Td o \ 4cO COc CM t>- o Ln Ln -- H 0) cd 2 v j ^o ^ ^ ^ #. C' Cd +1 + 0) c o CO H 0 0 H L + -P 0 P O Ln -4 C C f H *H 0 4 CM -P CM 00 0 0 0> CO / \o - rd 0 "^ ^ a 0 4 0 0.> 0 04rX - r 0 r- r- c I O Cd N r. 4 +,1 +, + + 4s 0 5 4 O H O I O 0 0O O H. X o o o o 0 co co 0 c m U E o o o M C -r-H *H O1 0 L Ln co C Cn CL 4-) 0 00 H H H H H H 000 0 C 0> 0 0> 0> C~j 0 O ) o ~ ^ H o 0 I 0 c, 0 I 1 -P-P Ln 0 co -C- CM C CCd >b0 O Q O O O Q O OO 0 0 0 (u) 0 0 (D -P P0 4t3........ Cd... Cd. C Cd Cd Cd! (C2 Cd 0 a, i-l i- a-l i-l t-l r- rDr

148 The same three bombardments also established the cross section values for the Zr9(d, a)Y9 reaction at the three energies. In bombardment 12 two foils (12a and 12b) had been exposed simultaneously to the deuteron beam, foil 12b, however, could only be worked up after its 92 component had decayed out already. A recheck on the value obtained by bombardment 10 for the Zr94(d, a)Y92 reaction with 7.56 Mev deuterons was thus not possible. The experimental errors as reported in Table X are comparable to those reported for the Zr92(d, aYy90 reaction measured by the same bombardments. The only significant difference is the smaller error reported for the count rate from bombardment 12a. The improvement over the Y90 component was mainly due to the statistics. The decay of the reaction products from the bombardments was followed with the 4w counters and in the order of 200 counts taken for each sample. This facilitated the resolution of the decay curves into the three components. (no additional half-lives were observed in the purified samples.) Scattering of the points remained largely within the statistical variation of the individual counts, which could be kept well below 1%. Bombardment 13 was carried out in addition to the absolute runs described above. It represented a typical stacked foil experiment for determining the excitation functions of the two reactions. Table X gives the results normalized to the values from the absolute experiments.

149 Seven zirconium foils were interleaved with aluminum foils serving as absorbers (attenuators) of the cyclotron beam and as catchers of the recoils. Table X gives the arrangement of the target foils and absorbers as employed in bombardment 13. The chemical separation of the yttrium from the bombarded targets could be simplified for this relative experiment. It was only important that the chemical yields should be the same for the foils exposed to different bombarding energies. This was achieved by the procedure given in Table XII. The carrier-type separation could be performed rapidly to permit counting of the Y92 components of all samples and provided maximum uniformity of the chemical yields. It is estimated that the variation of the chemical yield from foil to foil did not exceed +5% standard error. The samples obtained from this bombardment were counted in sequence with the -rgyproportional counter for 22 hours continuously. An absorber of 640 mg/cm2 aluminum was employed which favored strongly the 3.6-Mev P rays from Y92, rejecting most of the 2.27-Mev Y90 activity as well as weaker B rays from other impurities. Relatively good decay curves with 3 - 4 hour half-lives were obtained and easily resolvable into their Y92 component. After 22 hours the 3.6-hour Y92 components had largely decayed out and the samples were then counted through a 172 mg/cm2 aluminum absorber. This aluminum absorber permitted the strong 2.7-Mev m rays of Y90 to be

150 Table XI. Arrangement of Zirconium Targets for Bombardment 13 Target or Absorber Thickness (mg/cm ) Average Deuteron Energy.....- -ev-.... —Mev Mylar 1.76 Zr I 3.56 7.56 Aluminum 14.75 Zr II 3.49 6.37 Aluminum 11.32 Zr III 3.43 5.02 Aluminum 8.93 Zr IV 3.42 3.89 Aluminum 8.93 Z:r V 3.52 2.48 Aluminum 8.93 Zr VI 3.57 0 * Aluminumn 21.77 Zr VII 3.53 0 * * These two targets served to check the presence of reaction product from neutron induced reactions.

151 Table XII. Chemical Separation for Yttrium in Bombardment 13 Element Separated from Decontamination Factor Y d. bombarded Zr 103 Time of Separation Chemical Yield 15 min. 90% Chemicals and Equipment Lusteroid tubes,40 ml; centrifuge;, hot water bath; razor blade. HF; yttrium carrier (10 mg/ml); calcium carrier (10 mg/ml). Procedures (1) Dissolve the Zr foil plus the aluminum substrate in dilute HF solution in a Lusteroid tube. Add 5 drops yttrium carrier plus 1 drop calcium carrier. Centrifuge for 4 minutes, decant. (2) Flatten the bottom of the Lusteroid tube by dipping it in boiling water and pressing it against a lead brick. (3) Cut off the bottom of Lusteroid tube, containing the precipitate, with a razor blade. (4) Dry precipitate under a heat lamp and mount on sample card covered with 1/4-mil Mylar.

152 counted, but reduced the background of impurities due to incomplete chemical separation to a negligible minimum. Each sample was counted three times daily in intervals of approximately 8 hours over a period of 15 days. Very straight decay curves of the 64 hour Y90 were obtained. 2. Determination of the Zr90(d, a)Y88 Cross Sections and Excitation Function. The determination of the Zr90(d, a)Y88 cross sections and excitation function proved considerably more difficult than the cross section determination discussed above. The product has a relatively long half-life and very little of it, in terms of activity, could be produced during the regular two-hour bombardments (Fig. 51). It was thus necessary to obtain a very much longer and more intense bombardment for this determination. With the special cooperation of the cyclotron crew it was possible to arrange for a 10-hour bombardment which was used to bombard simultaneously five zirconium targets in the stacked-foil fashion. Their arrangement is seen from Table XIII.Unfortunately the Mylar substrates of all but the first target were damaged so severely by the beam that they broke, wrinkled and bent backward, resulting in higher actual bombarding energies for the last three foils, as well as loss of recoils. Fortunately the substrate of the first target remained intact and that of the second sufficiently so to be dependable for the cross section determinations at those energies. In

153 Table XIII. Arrangement of Zirconium Targets for Bombardment 6 Target or Absorber Thickness (mg/cm ) Average Deuteron Energy Mev Mylar 1.76 Zr-6 3.488 7.56 Mylar 3.52 Zr-5 3.480 7.09 Mylar 3.52 Aluminum 10.78 Zr-4 3.515 4.7 Mylar 3.52 Aluminum 10.78 Zr-3 3.597 2.7 Mylar 3.52 Zr-l 3.479 1.7 Mylar 3.52

154 addition the l/2-mil Mylar scattering target used burned through during the latter part of the bombardment making the initial energy somewhat uncertain. The targets were permitted to "cool" for several,g 92 months to permit the Y0 and Y to decay completely. After this they were worked up and the chemical yields of the yttrium product traced with Y90. The decay of the 90 Y9 tracer was followed for one month to establish the chemical yields. Finally the samples were counted with the x-ray counter over a period of several additional months. It was possible by this method to determine the relative counting rates of four samples so as to establish a relative excitation function of the reaction. Large errors, however, had to be taken into consideration due to the partial destruction of the Mylar substrates resulting in a "curve"'of which actually only the first two points are known to within 10%. For the determination of the absolute cross section of the reaction from the two relatively undestroyed targets it was necessary to measure the absolutte disintegration rate of y8 In effect this amounted to a determination of the efficiency of the x-ray counter for Y x-rays as was described in section III-D-4 of part 2. With the calibrated x-ray counter efficiency for the 88 Y photopeak the absolute counting rate of the samples from bombardment 6 could be determined with an estimated error of approximately 10%. The cross-section values

155 calculated from the experimental data are presented in Table X. Two points of the excitation curve are entered into Fig. 52. For the other values the uncertainties were too great to warrant entries. E. Resolution of the Decay Curves The resolution of the decay curves obtained for the yttrium products was carried out by the following method: A large graph of the gross decay curve, corrected for background and coincidences, was plotted on semi-log paper. The longest half-life (88) was extrapolated to the end of the bombardment. The value of the extrapolated line was rea'd from the graph for each time a count had been taken during the decay and subtracted from the gross decay curve. A plot of the difference resulted in a twocomponent curve. Small corrections could be applied to the extrapolated 88 line to straighten out a slight curvature in the curve of the next shorter half-life (Y90). The same process was repeated for the two-component curve until a straight line was obtained for the third component 92 (Y9 ). The counting rate of a particular component at the end of the bombardment was then calculated from the resolved decay curve in the following way: arbitrary points, lying on the decay curve drawn with the "known" half-life of the component through the resolved points, were chosen.

156 Excitation Functions /-O Zr90(d, cc)y88ee / Zr92(d,a)Y90 x /0 I0 Zr94(dc )Y92 o / Jol~ // I~~/ -2 /4 5 DU R E RG/,,I // 2 0415 6 7 8 DEUTERON ENERGY (Mev) Fig. 52. Excitation Functions for the Zr(d, oa)Y Reactions. im~~~~~~~~~ Fig. 52. Excitation Functions for the Zr(d, o~)Y Reactions.

157 With the aid of the decay constant the count rate at the end of the bombardment could be computed from these points and the obtained values used to estimate the error. It is believed that the much more laborious analysis of the decay data by means of a least-squares method would not result in error limits exceeding those quoted, since very good data was available for the analysis. The resolved points scattered little around a line fitted to them representing a half-life in good agreement with values reported in the literature.

158 III. The Mo(d, a)Nb Reactions The molybdenum reactions were considered of interest as a comparison to the reactions involving zirconium. The atomic number of the target element is two units larger than that of zirconium. The chart for the nuclides of the molybdenum region is shown in Fig. 53. The (d, a) reaction products from molybdenum include seven daughter isotopes with 11 different activities. Of these only four isotopes were found in this research and only four cross sections determined. Niobium98 with a half-life of 30 minutes, as well as the isomers Nb94m and Nb90m with 6.6-minute and 24-sec.half-lives respectively had decayed out before the samples could be counted. The 20,000-year Nb9 was not obtained in sufficient quantity to be detected. Niobium-92, although not resolved from the 4w B-decay curves, could be identified by its characteristic x-ray with the x-ray spectrometer. Its spectrum can be seen in Fig. 54. To verify this spectrum, a sample of the 10-day Nb92 was prepared by deuteron bombardment of zirconium and subsequent ion exchange separation. (Niobium is eluted from a small Dowex 2 column with 4 M HC1, while zirconium and yttrium, the other two products of a zirconium bombardment, are eluted by 7 M and 12 M HC1 respectively (cf. Table II).

159..,............ Cd ~F O0j y. 0) 0 o 0 E 0$ 0 S0 H"::Ho.';! i i?:. Co~~~~~a,,s lB~~i.0, $ 0 C2E- P I T,.. ~:i::BC: r~ n, NI.,4't~ ZZ:? 6! I ~o ~'.I'it W!_.,ii.4 iji~. ~'l'~, ~~::. -. i )o ~-.. ~r T o -(O aa riSI.~ iii~~~cij~ I~ P: b re

160 y88 MoNb-14 Fig. 54. X-ray spectra of 88 and niobium fraction from bombardment 14 taken 12 days after the end of the bombardment. 9 ~92 y88 Nb92 Fig. 55. Nb92 X-ray Spectrum Fig. 55 shows the x-ray spectrum of this sample. It is essentially the same as the niobium spectrum in Fig. 54. 92 The 13-hour activity also reported for Nb92 could not be identified. By resolving the niobium P-decay curves obtained with the 4w counters into only four components (Fig. 57 and 58), any contribution by the 13-hour Nb92 was added to the Nb90 component, decaying with a 14.6-hour

161 half-life. The contribution of the Nb92 was thus included in the calculation of the Mo92(d, )Nb90 cross section making this value an upper limit for the (d, a) cross section of the magic Mo92 parent. The following niobium isotopes could be identified in the P-decay curves of the molybdenum bombardment products: The 23-hour Nb96, the 35-day Nb95, the 84-hour Nb95m and the 14.6-hour Nb90 emitting a 1.5-Mev positron together with three y rays. A list of the properties of these isotopes is seen in Table XIV. A. Targets Like zirconium, molybdenum is a rather refractory metal and could not be satisfactorily evaporated with the equipment available for this research. Metal foils were thus used for the targets. Through the courtesy of the Arnold Engineering Company foils of pure molybdenum metal were obtained with a thickness of approximately 5 mg/cm2 (0.13 mil). The targets were prepared by cutting the foil into disks by means of a stamp die and mounting them onto a Mylar substrate which was stretched over the aperture of an aluminum target-frame. The average thicknessof the disks was determined by weighing with a micro-balance to within 0.1% and the uniform thickness determined by means of the Pm147 gauge as in the case of zirconium.

162 o' *- >' 0' *' I* X +l 04 2 * *'U *) b Lnco + n+ > — +ric C o 0 o o 0 n 0 1 o Q r\H O r H O t^ m m 0 c H - Oc\ LfV:0 II o~'\ 0ooo tlc- r'\ 00 0..- O L O 0 0 0 HH 0 0 0 rd 0o ~o- Ln Ln 0000oo 0 0 rC ot- + 00 f~~Q)~~~~~~~ () v e (U 0 + 0 ~C + Ai 04 L( I 0 HCM ( i C I OI ~~Z ~ - op ~ Z 02 H0 ci ~ ~ - z — I 0- 0 c Q. - H a) 0 0 02 <D < || 02, 0 * 0 02,.d c-4 0 c2 H 2 O CQ 0 2 Pi H H ci 02 0 2 4) 0 CM Lf L 0 cm 0m Go

163 0 C) (^ I, \ o\ o>>cr\ o\ m ~ccr\ c d) iu 0o o H o o 0 oo O0 0 0 0 0 0 0 0 0 0 O H U) 4 H O G0C000\ O cu o 0-\ co 0 -H -C) C IH I 0O! H Q ~ U0 0) 0) rQ' 0) Cd o C Ia2* ) H Co r3 -CQ E^ H ~c

164 Table XV Chemical Separation of Niobium from Molybdenum Targets (carrier-free) Element Separated from Decontamination Factor Nb (carrier-free) d. bombarded Mo,105 Time of Separation Chemical Yield ~4 hours ~50% Chemicals and Equipment Pyrex beakers (50 ml.); centrifuge; centrifuge tubes 40 ml., 15 ml.; medicine droppers; micropipette 50A; platinum wire 6 inch; small anion exchange column (Fig. 20); molybdenum carrier (10 mg/ml); aluminum carrier (10 mg/ml). Dowex-2 (200 - 400 mesh); conductivity water; conc. H2S04, NH40H, HC1, HN03; 30% H202; HCl-gas tank. Procedures (1) Place tracer niobium in small beaker, add bombarded molybdenum foil plus Mylar substrate. Add 30 drops of conc. H S04 and a few drops H202. Heat to fumes until dissolvSd. (Cool and add H202 together with a few drops of conc. HNO3 at intervals.) (Note 1) (2) Transfer clear yellow-brown solution to 40 ml. centrifuge cone, add 30 ml. H20 and 2 mg aluminum carrier. Stir well. Make alkaline with cone. NH40H. Centrifuge and wash twice. (Note 2) (3) Dissolve precipitate with a minimum of conc. HC1, dilute and reprecipitate with cone. NH40H. Centrifuge and wash twice. (4) Dissolve precipitate in conc. HC1, add few drops of molybdenum carrier, dilute and reprecipitate with cone. NH40H. Centrifuge and wash twice. (5) Dissolve precipitate in minimum amount of cone. HC1 and saturate with HC1 gas. (Note 3) (6) Transfer this solution to a small anion-exchange column charged with 1 ml. Dowex-2 resin saturated with conc. HC01.

165 Table XV(continued) (7) Permit to absorb slowly. When liquid level reaches the resin bed add five drops conc. HC1 and try to wash glass walls with this. Permit to soak in and add about 1 ml. conc. HC1. (8) Elute the aluminum plus some salt (NH4C1) with 12 ml. of 7 M HC1. (Note 4) (9) Elute carrier-free niobium with 25 ml. of 4 M at a rate of 1 drop in 5 - 7 seconds. (10) Evaporate the niobium fraction to near dryness and plate for counting. Notes (1) Complete mixing of tracer and reaction product is achieved in the homogeneous solution. All elements are oxidized to their highest valence state. (2) Ammonium molybdate and ammonium pertechnate are soluble and thus not precipitated together with the aluminum. (3) During most bombardments the precipitation of salt, be it AlCl or NHhC1, was observed when the solution was saturated with HC1. This, however, did not interfere with the ion-exchange step. (4) The specifications given in the procedure have to be rigidly observed as they are rather critical.

166 B. Chemical Separation The (d, a) reaction products were separated from the bombarded targets by the carrier-free procedure given in Table XV. This chemical procedure followed in its outline very closely the ideas described for the zirconium-yttrium separation, and was found to be very dependable and effective. The elution of niobium from the Dowex-2 column as carried out in the last step of the procedure is somewhat difficult and only approximately 60% of the niobium activity is removed by the 25 ml.of 4 M HC1. Formation of a radiocolloid at this HC1 concentration inside the resin particles may be the reason for this behavior. At 50 ml.molybdenum starts' eluting. A safety factor of 2 is thus observed, if only the 25 ml. fraction is retained. Technetium was not found in the columns after elution with 50 ml. 2 M HC1, proving the precipitation steps a rather effective separation of niobium from that element. C. Tracer for the Mo(d, a)Nb Reactions The most readily available long-lived niobium isotope is the 35-day Nb95. 1. Procurement. To procure Nb95 in sufficient quantity it was separated from its parent Zr95 by the procedure given below. (Table XVI)

167 Table XVI Chemical Separation of Niobium from Zirconium Targets (carrier-free) Element Separated from Decontamination Factor Nb (carrier-free) Zr95 from fission 105 products Time of Separation Chemical Yield 4 hours 50% Chemicals and Equipment 2 Pyrex beakers (50 ml.); centrifuge; 2 centrifuge cones 15 ml.; medicine droppers; platinum wire 6 inch; small anion exchange column (Fig. 20); zirconium carrier (10 mg/ml). Dowex 2 resin (200 - 400 mesh); conductivity water; conc. H2S04 HC1, NH OH; 30% H202; HCl-gas tank. Procedures (1) Destroy oxalate complex with 1 ml. conc. H2SO plus 3 drops 30 H202. (2) Dilute with 10 ml. HO2 and precipitate zirconium-hydrcxide with conc. NH40H. Centrifuge, washonce. -Repeat step 2 once more. (3) Dissolve precipitate with minimum amount conc. HC1, saturate with HC1 gas and apply to small anion exchange column charged with 1 ml. Dowex 2 which had been saturated with con'. HC1. (4) Absorb slowly. When liquid level reaches the resin bed add 10 drops conc.-HC1 and try to wash sides of column wall. Permit to soak in. (5) Elute elements like alkalis, earths and zirconium with two 5 ml. portions of 7 M HC1. (6) Elute Nb95 tracer with 7 ml. of 2 M HC1 made up with conductivity water. Discard the first ml. and use rest as tracer.

168 A first attempt to identify Nb95 in the separated fraction 95 indicated that the old Zr95 sample on hand contained several extraneous activities. A new sample was thus obtained from the Oak Ridge National Laboratory. The latter was purified by the procedure given in Table XVI. 2. Identification and Use as Tracer. The tracer thus purified was identified by its y-ray spectrum seen in Fig. 56. The Nb95 tracer was not used for any experiments until all of the 84-hour Nb95m had decayed out, so as not to complicate matters unduely. The method used for tracing the niobium fraction of the bombardment product from molybdenum varied from that described for the zirconium bombardments. Three tracer samples were prepared and mounted on gold-plated Zapon film. Two had equal amounts of activity and one ten-times their strength. Relative counting rates for all three samples were obtained for the identical setting and geometry with a 7 counter counting the photopeak of the 0.74-Mev 7 ray 95 of Nb9. The two weaker samples were counted in addition in the 4w counter and their counting rate ratio established. One of these samples was used as tracer for the bombardment. and the other two preserved. 95 After all but the Nb9 component had decayed out of the sample from the molybdenum bombardment, counts were again taken with a y counter and 4w counter and the ratios of the count rates established. After subtraction of the Nb95

169 7NIOBI U M -95 y- spectrum si' a) CC 2 Q:-~~~~~~~PICL_ r a- ^~: _.1.2.3.1.S.6.7 Energy (M ev.) Fig. 56. Gamma Spectrum of Nb95 Tracer

170 fraction stemming from the Mo97(d, a)Nb95 reaction the chemical yield could be established immediately without further extrapolation to the time before the bombardment. This procedure has the advantage of being independent of the counters which may vary in their efficiency from day to day. It is also independent of the accurate knowledge of the decay constant of the tracer. These two facts make the method preferable, if enough pure tracer activity is available. The only requirement for the method is, that the tracer has been purified previously by the same method by which the sample from the bombardment is purified afterwards. D. Records of the Bombardments Four bombardments were obtained to determine the cross sections and excitation functions of the Mo(d, a)Nb reactions. Bombardment 5 served to check the dependability of the chemical separation procedure and to identify the reaction products. The relative yields of the product activities was established by bombardment 14 for a two-hour bombardment with 7.8-Mev deuterons. The ratio determined between the count rates of the Nb95 component and the other activities was to be used-in the yield calculations for the absolute cross-section experiment of bombardment 18. The molybdenum target was exposed to the full energy

171 beam in bombardment 18. The niobium product was chemically separated from the bombarded target together with 31,000 c/m 95 Nb added tracer. The separated niobium fraction was followed in its decay with a 4w counter over a period of several months. To establish the absolute count rate of the Nb95 component a self-absorption correction had to be applied. This correction was determined by the method described in section III-D-1 of part 2 and found to be 6 + 1% for the sample counted. The primary data and the cross section values for the Mo97(d, a)Nb95, Mo97(d, )Nb95m, M98(d, a)Nb96 and Mo92(d, a)Nb90 reactions are reported in Table XVII. An internal-conversion coefficient of 90% was assumed for the 0.23-Mev y ray of Nb95m No corrections were applied for any electron-capture processes accompanying the decay 90 of Nb90. The errors quoted in Table XVII were estimated in the same way as was discussed for the zirconium bombardments. The values are based entirely on the experimental data as obtained with the 4w counters. The decay curve obtained from bombardment 18 is seen in Fig. 57. The size of the points of these figures is in no relation to the experimental errors which were much less than the circles indicate. Only with gross data of this type was it possible to analyze the four component decay curve of the niobium reaction product with the small errors quoted.

172 o a) o 0 0 I- 0 z io o0': -E -c ~ o/ c o 0 0 0 0 O> - 0 LL 80 0 00L ct 00 0 0 00 0 00000000000 N 0 0 0 0 0 0 11, k o o -' c/ - < -o >-Ct 0 L OL.O O~~~~~L. 0~ 0 0Oq~0 0 0 0 o oCL 00 bl)0 003 8V g<M 0 0:L SI <^7- 0 <P - o dio^vr00.0000,00,,,,/ — o —I oo o 0

173 cr o o 0R o0 0 I 0P ~ ~ 0 H C 0 H - L fc b HH1 H +1 + 1I +1 +1 + rH In I +' In in + co r Ln H r r 10 r KO 0>\ LCn 0> C Ho 0 0 0 OC CM F 3 o0 H H H H -H i c <4 C o o1 |?I I C,C + 1 +1 +1| cr bOW N\ CM 0 CM +1 C +1 CM CM C Co czi-sc~~oo o crr n 1 o 1 on?-E I H;< I 4 4 4 4I ) - I 1 4-P a) 0 0 p, ~ a P O PH -H H R -Q o -P p4- LC LCn oc) P Ln\ 4-) LC Co cr3 c) * * 0'' 0 0 0 cc) c O CY) cr C(M H K\ LC-, M Co > c r3 H-, c r3 H o Co +| -1 + - +1 +1 +1 P +1 Co +1 C +1 C o oR;l- 0 o 0 Ln 4 -N\ CO 0n M o n O n c 00> Ln Lon O D L: C O:5 n I n I C>n OC 0 0 0 H p F-i 0 0 0> 0>.' 0> 0 I 1 d 10~d 1rOd 1 ~ 1 t Fl 03 C^ ^ *So rd f- tC- Co ~ CM C H- 0> 0> 0 0> 4-p o a) 0 ^. 0. 0 1. 04 o H H H C +'~, rl a,+,+,ia a) r4 ^ cF l 1 *L L1 IF L+ C I1> IC C 1 Sl +I F aj +1 C +4) E-i t- +1 +1 +1 +1 +1 E- c- E-H - E-i - tCH Co Co Co CO FC o r0 H H H 4-3Co H 0- P Eq h'O Oh (O Oh' C.O t S $1 +~l $1 $1 O, —I 1 $- $1 o H0 4H> Co H IH H H, - Co + H +1 +1. +1.o p 0 1 1 1 r —I 10-" a) " - Q o;5:E: Ln Ln Ln Mn a C L n>~1 Co (DO 0 0 (1) U?-(D ~ (D4 Co 4F-> C o H o 0 0 0 H CMOJ 0 0 0 4 ) b0~o H 4 4 H'l-=f -^- M OJ~nrf MH H crFl 1> - tLf LC Co CM b- b- crd3( 4-P Co 4 CO L IC L LC> IC CO CO CO 0 H- - H r H H-i r r H r rH H

174 Bombardment 15 was performed to establish the relative excitation functions for the Mo(d, a)Nb reactions. A stacked foil technique with aluminum absorber substrates was used as in bombardment 13 mentioned above for zirconium. The arrangement of the targets and absorbers for this experiment is given in Table XVIII.No tracers were employed for the yield evaluation, but strict conformity of the simultaneously performed separations permitted an error limit of not more than about +5% for the chemical yields. The six samples obtained from this bombardment were counted in sequence with the proportional counter. Counts were taken continuously for five days following the bombardment to permit resolution of the 14.6-hour Nb90 and 96 the 23-hour Nb. The remaining activities stemming /95 95m from Nb95 and Nb95m were followed in their decay for an additional two months to permit good resolution of the curves. The relative cross sections for Nb are normalized to the absolute value obtained in bombardment 18 and entered in Table XVII. A plot of the excitation curve for the Mo97(d, a)Nb95 reaction is seen in Fig. 60. Gamma spectra were taken twice during the decay of the niobium sample from bombardment 18. They can be seen in Fig. 59 indicating a strong enrichment of the 0.76-Mev Nb95 line with time.

175 Table XVIII. Arrangement of Molybdenum Targets for Bombardment 15 2 Target or Absorber Thickness (mg/cm2) Average Deuteron Energy Mev Mylar 1.76 Mo I 4.79 7.54 Aluminum 11.32 Mo II 4.85 5.61 Aluminum 14.75 Mo II 4.91 3.82 Aluminum 8.93 Mo IV 4.84 2.25 Aluminum 4.46 Mo V 5.06 1.0 Aluminum 4.46 Mo VI 4.80 0 * Aluminum 4.46 Mo VII 5.00 0 * Aluminum 14.75 Mo VIII 5.00 0 * * These targets served to check the presence of products from neutron-induced reactions.

176 E. Resolution of the Decay Curves The decay curve of the niobium product from bombardment 18, seen in Figs. 57 and 58, may serve as an illustration of the resolution problems encountered. The 35-day Nb95 and 84-hour Nb95m activities could be subtracted from the gross P-decay curve by the same procedure employed for resolving the yttrium curves described earlier. The resulting two-component curve could not be analyzed, however, by continued subtraction. The following consideration aided in its resolution: 9,000 - RESOLUTION OF Nb90AND Nb96 O-TIME-11-30-1956 20PM 80,000 - 70,000 60,000 - 50,000- oy j 40,000 - O l 1.2.3 A.6.7.8.9 1.0 -(X2-Xl)t Fig. 58. Resolution of the Nb90 and Nb96 components of the niobium decay curve from bombardment 18 (cf Fig. 57)

177 oo 00 H o p he rd 0.ra 0~-I O-H coi a) -p 0 0 m C rq o0 CO Cr. O) 0 Ct rd A t0 LC\ 0) *H (4 F.__ Cd

178 I,0Excitation Function for Mo'7(dcr)7Nb en / z 0.1 z Ol ~-I~~~~' to 2' / mX.01 - 0 I I A. I I I I 2 3 4 5 6 7 8 DEUTERON ENERGY (Mev.) Fig. 60. Excitation Function for the Mo97(d, a)Nb95 Reaction

179 The decay rate of a sample containing two radioactive species is given by: A = A1e + A2e 2t where XA is the decay constant of the longer lived, and X2 the decay constant of the shorter-lived activity. Transposition of above equation yields: Aet = A1 + Ae-(2 - t or AeXlt A + Ae t It is seen from the last equation that a plot of Ae t - Axt vs. e will yield a straight line with an intercept -6At at e = 0 or t= oo representing A1 and a value for t = 0 or e-~\t = 1 representing Al + A2, the sum of the decay rates of the two components at time 0. A plot of this type is given in Fig. 58 for the Nb90 and Nb96 components of the niobium decay curve from bombardment 18. For this type analysis the exact knowledge of the decay constants is very essential. Resolution of the two components was possible within approximately 1.5% for the two components. The labor involved in this analysis, however, was extensive.

180 IV. The Ti(d, a)Sc Reactions The titanium reactions were considered of interest partly for the challenging chemical problems and the fact that Ti5 containing 28 neutrons is a magic nucleus. Hall attempted the cross section determination of the Ti 48(d, )Sc6 reaction but obtained only rather uncertain values. His main difficulties rested with the chemical separation employed and the fact that no O.0001linch titanium foils were available to him. The chart of the nuclei in the titanium region is seen in Fig. 61. Six half-lives should be expected in the (d, a) reaction products on a titanium target. Five of these were observed in the present studies. Only the 20-sec isomer 46 of Sc46 had decayed out by the time the separated reaction products could be counted. The isotopes observed in the reaction products were the following: 48 a) The 44-hour Sc, emitter of a 0.64WMev P ray and several 7-rays. b) The 81-hour Sc 47 emitting a 0.44-Mev or 0.60 Mev negatron as well as a 0.16 Mev x-ray. c) The 85-day Sc46, emitting a weak 0.56-Mev P particle with two y rays. d) The 4.1-hour Sc, emitting a 1.47-Mev positron together with several y rays. e) The isomeric Sc decaying by internal transition with emission of a 0.14-Mev y ray with a half-life of 57 hours.

181 C**J! -, o. CI I - - - _ O il E """ a ~ |iaoo < 0 go OD 11) CY cmO F~~~~~~]i 0 W1,0 -I 0 in 00 10i i j b0 ^s?8 ^^Sr |~:.::~Qk W > ^ ^/.^ -^ ^:~~ C1:~ 9s ^ ^~~~~~~~~~~a |p (0j 0) j~ (P

182 4 E 4 i > C4 4 44 |4 C) aO CU c CU C0 4 D CUO C4 O( H r-I rr rH H H H H H H QC H a3 )) (Da) P. a 0,C.C0 H LF t o0 C d o H ( \4 O H +10 _ rH 0 0 0 HM H C0 U2i TC) Q) 0 0 O O - p o C o O 0 O4 0C) 0 + H H rH r I= P0: 00 cr3 o) a) 0* I2 C Mf O C + I0 rd Cd *0 44 c 44 4 O c- C\JCo404 C('04 ('4 0('4 a C C) ~- -- - + - O " I r-IC-i r^ H *H M OCQc^ ^- ^- GQL CCL ^-^C) rga ~ 0 0 1 -4 c 3 *H -P t rCl)d~~e C HQ 0 H 0 3tRt

183 CH 0 d) 0) C CM CM M rd r-I rH d) o 0 0 Lo hO Ln r- LO CO -'O -C -' ff~ o -U ^H 0 00 4 D 4 cRo 4 O O O O O O 0 rh 0 0 0 0 0 0 0 r - ~ r + ^ ^ c0 o 0 0 0 0 0 rl E b>- C 0 0 0 0 0 a c3 rd rH r- rH H H 0 CT) -H I I H i I K\ 0) 4 4 LC 0> 4 rd C2- 0o_ - co. (u rt l r c CM CM CM 0 H H H (D I-4I -P 0 0 0 44 r 4 H fH o C) ^ - (D X 0o - 4.4 H 0 00 0V 0 H Ecl

184 A summary of the decay characteristics of these isotopes is given in Table XIX. Although very good data were obtained for the decay curves of the Ti(d, a)Sc reaction products as can be seen in Fig. 62, resolution of the curve was only possible for 46 44 the 85-day Sc activity and the 4.1-hour Sc4. The com-bined components due to the 44-hour Sc48 the 57-hour Sc4 and the 81-hour Sc47 forming the 62-hour component seen in Fig. 62 could not be resolved into the three activities by the means available. This made the determination of the cross sections for the Ti50(d, a)Sc48, Ti49(d, a)Sc47 44~m and Ti(d, )Sc44m reactions impossible. The value re46 ported for the (d, a) reaction on Ti is to be considered only a partial cross.section value, since the amount of 44m ~Sc 44mcould not be measured and simultaneous decay of the 44 Sc by the electron-capture process was not entered into the calculations. Application of a correction for the latter effect would increase the value obtained for the (d, a) cross section by about 6 + 1%; the relative yield 46 44m of the Ti (d, a)Sc reaction is estimated to be of the order of 5% compared to the Ti46(d, a)Sc4 reaction. A. Targets Again, as in the cases of zirconium and molybdenum, thin titanium foils were used for the targets. The foils were rolled to a thickness of 4.4 mg/cm2 (0.15 mil) by the Arnold Engineering Company from 2-mil titanium stock, kindly supplied by the

185 Titanium Corporation of America. These foils were cut with the die, mentioned previously, and mounted on Mylar. The average thickness of the foil disks was determined by weighing with a micro-balance within + 0.1% and the uniformity of the thickness checked with the Pm 47 gauge (+ 3%). B. Chemical Separations The chemical procedure for separating the scandium reaction product from the bombarded targets combined a radio-colloid purification step with an anion exchange separation. The yields were relatively small for this separation, but the effected purification excellent (Table XX). After preliminary separation from vanadium, the bulk of titanium was reduced to approximately 1/20 by a homogeneous precipitation of Ti(OH)4 while the Sc+++ remained in solution. The anion-exchange step was carried out in a strongly oxidizing medium which permitted the titanium and vanadium to remain in their highest oxidation states. +4 +5 Ti and V were not eluted from the Dowex-2 column by more than 50 ml cone HC1 - KClO1 under these conditions (Table II). The final "radio-colloid" step (43) freed the carrierfree scandium from any macro amounts of NaCl and KC103 which accompanied it in the eluate from the anion-exchange separation. The carrier-free scandium activity was then obtained by elution from the filter paper with 4 M HC1.

186 Table XX Chemical Separation of Scandium from Titanium Targets (carrier-free) Element Separated from Decontamination Factor Sc (carrier-free) d. bombarded Ti 105 Time of Separation Chemical Yield 4 hours 50% Chemicals and Equipment 4 Pyrex beakers (50 ml); centrifuge; centrifuge tubes 40 ml., 15 ml.; platinum wire 6 inch; Erlenmeyer flask 250 ml.; medicine droppers; glass fritt filter funnel, coarse; Whatman 42 filter paper; micropipet 50A; small ion-exchange column Fig. 20. Dowex-2 resin (200 - 400 mesh): conductivity water; conc. H2SOh, HNO, HC1, NH4OH; 30% H 02; NaOH (solid pellets); NaHCO (saturated solution);KCI03 (crystals); HCl-gas tank. Procedures (1) Place tracer scandium in small beaker, add Mylar substrate. Add 2 ml cone H2SO plus a few drops of 30% H02. Add bombarded titanium foil, heat until dissolved. Oxidize purple solution with few drops of HNO (some TiO2 precipitates at this point if solution Us too hot, but this does not interfere). (Note 1) (2) Transfer to 50 ml centrifuge cone and precipitate TiO2 aq. with several pellets of NaOH. Wash twice. (3) Dissolve with cone HC1. (If solution is not complete, try adding water and heat gently). Centrifuge and transfer supernate to another centrifuge cone. (4) Add slowly a saturated solution of NaHCO until initial precipitate still dissolves on stirring.3 (Note 2) (5) Heat gently to effect homogeneous precipitation of Ti02.aq. Do not permit the pH to rise above about 5.5. (6) Repeat steps 3, 4 and 5 and combine the supernate with that of step 5. (7) Saturate combined supernates with HC1 gas (Cool!). Centrifuge off the NaCl. (8) Evaporate to 5 ml and repeat step 7.

187 Table XX (continued) (9) Add a crystal of KC10, shake and transfer to a small anion exchange column charged with 1 ml. Dowex-2 resin which has been saturated with conc. HC1 containing a few mg KC103 per 100 ml. (10) Permit to absorb. When liquid level reaches the resin bed add five drops of conc. HC1 with KC10 and permit again to reach the bed-level. Elute the scandium activity with approximately 12 ml. conc. HCl-KClO.1 (Check with monitor when activity is eluted). (11) Pour the eluate rapidly into a 250 ml. Erlenmeyer flask containing 30 ml. of a solution made of 11 parts 8 N NH OH and 1 part 30% H202. Shake vigorously. (Note 3) (12) After fumes subside, cool the clear solution and pass it twice through a double-layer of Whatman 42 filter paper positioned over a glass-frit funnel. (The scandium activity remains on the filter in radiocolloidal form). (13) Wash filter paper with 10 ml. of the mixture of step 11 to which about 3 ml. conc. HC1 has been added. (14) Wash with 5 ml. alkaline distilled water followed by a wash with 5 ml. conductivity water. (15) Elute the scandium activity from the filter paper by passing 5 ml. of 4 M HC1 twice through the filter paper some of the activity will still remain on the paper, but most of it will be eluted). (16) Evaporate to near dryness and plate for counting. Notes (1) The Mylar is dissolved first, since titanium forms a precipitate in hot conc. H2S04 solution containing peroxide. (2) Ti(OH) is precipitated homogeneously from a solution of pH I, while the Sc(OH)3 precipitates only from a pH of 7. (3) A sudden increase of the pH in a strongly oxidizing medium transforms any titanium present into the titanate ion whose ammonium salt is soluble, while the trace amount of scandium, being insoluble in the medium "precipitates" in the form of a radio-colloid.

188 At times it was possible to elute only about 3/4 of the scandium from the filter paper to which it is very strongly adsorbed. Relatively low yieldswere the results of this effect. C. Tracer for the Ti(d, a)Sc Reactions The most readily available long-lived scandium isotope is Sc with a half-life of 85 days. This isotope was obtained in carrier-free form from the Oak Ridge National Laboratory and stored for approximately one year. To ensure a pure tracer, the Sc46 activity was purified previously to bombardment 19 by the same chemical procedure as given in Table XX. Some inactive Ti foil was added to serve as carrier of the scandium during the first steps. The tracer Sc 46 was identified by its y spectrum and half-life. The spectrum is seen in Fig. 63. The method used for tracing the scandium fraction of the titanium bombardments followed in essence the method described above for the Mo(d, a)Nb bombardments. Two 46 Sc samples were prepared and the 1.12-Mev photopeak of the y spectrum counted. The counting rate ratio for 7 rays in the two samples was thus established. One of the samples (A) was used in tracing the chemical separation, the other (B) preserved for later comparison. After the short-lived activities of the bombardment product had decayed out, the ratio of the sample (C) from the bombard

189 ment product and tracer sample (B) was determined. At the same time the ratio between the sample (B) and the bombardment product from a previous bombardment, for which 46 no Sc tracer had been used, was determined. This ratio was normalized to the time of the end of the absolute bombardment and to the same activity of the shorter-lived reaction products. (Bombardments of equal lengths were used for this comparison). From these ratios the amount of 46 48 46 Sc46 present in sample (C) due to the Ti8(d, a)Sc4 reaction could be calculated and subtracted from the total 46 Sc activity. The remainder, stemming from the added tracer, was then used to determine the chemical yield. It is estimated that an error of less than 2% was introduced in this experiment by the tracing. D. Record of the Bombardments The absolute cross sections of the Ti(d, a)Sc reactions for 7.7-Mev deuterons were determined with the aid of two bombardments. Bombardment 17 served to check the chemical procedure and establish the identities as well as the relative yields of the reaction products. The latter was important for the yield determination of the tracer experiment of bombardment 19. Counts were taken with the 4w counters as well as with the y-ray spectrometer.

190 Two 4w-counter samples were made from the bombardment 44 product, one weaker to follow the short lived Sc and one approximately 29 times as strong to follow the longerlived components. The ratio of the counting rates of the 44 two samples was carefully determined after the Sc4 activity had decayed out, and the curves, corrected for background and coincidences, combined into one. This procedure permitted to count samples not exceeding 100,000 c/m and yet left a strong enough sample to follow the 85-day Sc46 component over many half-lives. The complex 3-decay curve of the scandium fraction from bombardment 17 is seen in Fig. 62. Many more points were taken than are indicated in the figure. Those presented are, however, typical for the data obtained. The titanium target was exposed to the full energy of the cyclotron beam in bombardment 19. After the chemical separation two samples were prepared. One was followed in its decay with the 4w counter, the other for a period of two days with the y-ray spectrometer. Spectra of this sample taken at various times after the bombardment can be seen in Fig. 63 giving an indication of their complexity. From a sequence of spectra taken by automatically sweeping the y-ray spectrum for a period of two days follpwing the bombardment, the decay of the 0.5-Mev positron annihilation line could be followed. It decayed with a half-life of approximately 4 hours, eventually going into a much longer half-life.

.1,9 t o om0 U -i~ E4 H 0 o Z -- - 0 — C- LO I- U Q c U"> c 0J CI- Q) U / - (n z 0r rc o o cr - o 0 c- i3 Cm C0D w Q 0 0 o 00 (0 - 0= 0~ o 000 0 w 0 0 8 0;I^~- _ oU') 0 -r C) R~ 0 z a) U) 0) >< x x x (0 U) C) r-f 0 _0 0 0 0 (0NI dUdSI n _oh l0 00'3 0 ~~~~0 0 Jiflnki i -- 3d I f O ~~ 3r'NI 83 S. r10

192 Ln c3 I ) zO rO r o v 4Of vH.rd 1d Cd IOO o o rd Cd C) o 4d Or Hbcq~ -cm

193 The results from bombardment 19 are given in Table XXI. 46 Self-absorption of 1% was applied for the Sc count rates as determined by Hall (55).The error estimates are again based on the principles discussed for the case of the zirconium bombardments. The somewhat larger error in the 46 count rate determination of the Sc46 is due to the uncertainty caused by using this isotope as the tracer for the chemical yield. As was mentioned above, the cross-section value reported for the Ti 6(d, a)Sc44 does not include the Ti46(d, a)Sc44m reaction and has thus to be considered as a partial cross section. This is in contrast to the case of the Mo97(d, a)Nb95 reaction mentioned earlier where the Mo9(d, a)Nb95m was included into the cross section value of the over-all reaction. E. Resolution of the Decay Curves Figure 62 represents the complex decay curve of the scandium product from bombardment 17 as obtained with the 46 44 4w counter. A partial analysis into its Sc and Sc components was obtained by the method of consecutive subtraction discussed for the zirconium bombardments. 2ln 47 48 Resolution of the combined Sc44m Sc47 and Sc4 yielded an average half-life of approximately 62 hours which could not be resolved any further with the means available.

194.P rO 1'~C+1 + H LC\|^\ H C)H CO C\ C 1 H H H 0 $H boH i n _:I- 1 +1 0bO CM' D CM z H~~~~~r; (D L C -p -p -t - CQ 0 o V H 0 ^0E -1 1 HO P 2 00 I H d) d. *. co oy ~ ~ z Go I C1 G o; Eq n Eq Ln 0 410' 00 HE- Ln En C H>!C krCl' CM C, - I W) O= O 1 t 00 g E: F P ( r O O O O a +1 I +1 w >< S\ n r\\: ) K* N\ r 4\ 1 d) p i-l bo -P c i) -P. U0. 0 t- 0 OmZ H H H

V. The S34'(d, C)P2 Reaction The S34 reaction was considered of value for this research to establish the trend of the (d, a) reactions for non-magic number nuclei in the low-atomic weight range. Hall (54) attempted the determination of the cross section for this reaction, but experienced difficulties with the elemental sulfur targets employed, thus introducing rather large uncertainties in his values. The chart of the nuclides in the sulfur region is given in Fig. 64. There are three phosphorus isotopes obtainable by the (d, a) reaction on sulfur. Only one of them, however, has a long enough half-life to be observable after the chemical separations necessary to obtain it in carrier532 free form. Only p was thus observed in these investigations. This isotope decays by emission of a strong 1.71-Mev f particle without any 7 radiation. The half-life of P3 was of great importance for the analysis of the decay curves and had thus to be determined experimentally. The value of 14.221 + 0.005 days determined in this work may be compared to some of the values found in the literature as given in Table XXII. Most of the authors quoted in Table XXIIworked with p32 samples that were not pure and corrections had to be applied for the presence of S35 and P33. The majority

196 NUCLIDE CHART OF THE SULFUR REGION Cl33 Cl34 Cl33 Cl36 C137 Cl38 C139 2.8s 33.2m 75.4 4.4xly10 24.6 3729m 55.5m,___ + 3+)f f /r P, r S31 s32 s33 34 s35 s36 37 3.18s 5.018 0.750 4.215 871 d 0017 5.04m /9+ /3- 03,5 4 p28 29 30p3 p32 33 34 P P P P P P P.2,s 4.57s 2.5m 100 14.3 d 24.4 d 12.4s Fi g. /36 + /p- /3- _fFig. 64. Chart of Nuclides in the Sulfur Region

197 42 oH J d,0 - crrC\ cC rH toC o03 03 ) 0) 0 0 0:> >?H P-4-H.. H.-1 0 00 U) i 4-P 4) U) 0 a - U o) c0 4 ) 0 oa M P4->?f 0( 4-> 0 0 CO - C g, e, e c - O 0 8d rl -P2 r- -P t 0-2 )- 0 -.H 4O - - ~0 H O 2 c3 4-) C O 4-P - rdC(+- 0 0 0 4 0 0 00 0 cv 04u g HL U )L rC H r S H H U )c o0 CM 41.H ~ Hr C'~ — r —I or-l c c-'-I -0 C @PQ -P - 0 U 0QC u v2 0 U)CO a - 2 - o I U) 0. 0 H - 0 0 0 0 0 02? (D'''0 0 >0 Cd* v ^3 *H *H *H T-4 *0 E-4 O H H H H H H 0P O 0H H H H H H o 41-I i 45 4 Fi Q) 0 Cd rR a) U) a r,, U CM $ O 0 L H o0 00 C 0 00 02 > ^ H 0 02 1 I O'0 U) H!,^'- 03 0 r1H 41 0 0 0 0 cii 02 H U) 0 Ln 41l P 4 O^ 0- ^ H U 0 4) 32 * 4) U) ~ ^ c ^ 3 -P >5 < - aD o H U^ 0 U - 0 O ^ 0^ O H - O-. 03 ci 0 H N - CM AH 000 V H g ^.1ci r*H H *HH c O- 4 H U0 H X LN U)l 0 0 ><iTqr <:. - m LCOJ O H U) 0 0 0 0 0 0 H-l rH H H *o3 -t I -z z- -1 -- OJff ~H H H H H H H

198 of the determinations were made by following the decay of the P for less than six half-lives and carried out with instruments for which the reproducibility of the counting geometry, change of atmospheric conditions in the optical path and linearity of detection posed problems. The method used in this determination improved on the isotopic purity of the sample used, and avoided some of the errors by counting a carrier-free sample with a 4w counter. Following the decay for 11 half-lives increased the reliability of the experimental value (Fig. 65). A. Target Preparation Targets for the sulfur experiments were prepared by evaporating chemically pure ZnS onto roughened Mylar film in the high vacuum. (cf. section II-A-1 of part 2) 147 The evenness of the deposit was tested with the Pm' gauge and found to vary within less than 1% (the sensitivity limit of the gauge) over the entire target-area. The average thickness was determined to within 1% from the weight difference of the empty target frame plus Mylar substrate, and the weight after the ZnS had been deposited. The target area was defined by the area of the collimator-mask used in the deposition of the zinc sulfide.

199. DECAY OF P32 104 103 C)\ F- X _ X 0\,o, - I, I 800 1600 2400 3200 TIME (HOURS) Fig. 65. p-ray Decay Curve of p32 from Bombardment 1 over 11 half-lives

200 Table XXIII Chemical Separation of Phosphorus from Zinc Sulfide Targets (carrier-free) Element Separated from Decontamination Factor P (carrier-free) d. bombarded ZnS 10 Time of Separation Chemical Yield 6 hours 70% Chemicals and Equipment Pyrex beakers (800 ml); 1 Phillips beaker (250 ml); centrifuge; pH meter; 2 centrifuge tubes, 50 ml; Erlenmeyer flask, 1000 ml; 3 ion exchange columns, 25 cm long, 1 cm diameter. Dowex 50 resin (100 - 200 mesh); FeCl solution (1/2 satuProcedures (1) Place target into 250 ml. Phillips beaker containing the P3 tracer plated on Zapon film. Add a mixture of 20 ml. CC14 and 5 ml. Br and 5 ml. conc. HNO. Permit to react for 1/2 hour, shake at intervals. (N8te 1) (2) Heat gently and narrow down to near dryness, shake occasionally. (3) Add 3 - 4 ml. conc. H S04 to dissolve the Mylar (Polyester film) and the Zapon film of the tracer sample. Don't heat. (Note 2) (4) After clear solution is obtained, add water to hydrolize Mylar. Centrifuge off the precipitate. Decant and save supernate. (5) Add a NaOH pellet with a drop of water. The precipitate dissolves with little residue. Dilute, centrifuge and add supernate to that of step 4. (A milky solution results. Treat residue with a few drops H2S04 and add it to supernates. (Note 3) (6) Centrifuge for 4 minutes to get rid of the colloidal terephthalicacid. Add supernate into a 1 liter Erlenmeyer flask and dilute to approximately 500 - 600 ml. (depending on how much H2S04 had been used). (7) Pass the solution through a cation exchange column charged with 10 ml. 100 - 200 mesh Dowex-50 in hydrogen form. The metallic ions are retained by the resin.

201 Table XXIII(continued) (8) Evaporate eluate to less than 500 ml. Neutralize to pH 6.5 with less than 4 gr. solid Na2CO anhydr. Heat while adjusting pH. (9) Pass solution slowly through a Dowex-50 Fe(OH) column (Fig. 21 E) that has been prepared as described below. Wash with 50 ml. H20. Phosphorus is retained by the column. (Note 4)) (10) Elute phosphorus activity from column with 100 ml. of 0.125 N NaOH and pass it directly through a Dowex-50 column in its hydrogen state. Wash columns with 20 ml. H20. (11) Evaporate eluate to near dryness. Destroy dissolved column resin (r1 mg) with a few drops 30% H202 and plate for counting after H202 has been destroyed by gentle heating. Notes (1) The solution of Br in CCl oxidizes the sulfide to sulfate and avoids formation of colloidal sulfur which would interfere with attaining an initial homogeneous solution. Escape of phosphine is also avoided in the oxidizing medium. (2) Phosphoric acid is volatile when heated to the fuming temperature of sulfuric acid. (3) A homogeneous solution of the terephthalic acid monomers results in the alkaline solution. It is expected that only a negligible amount of p32 is carried by the reprecipitating acid before isotopic exchange with the P33 tracer has occurred. Steps 8, 9, 10 form a unique method of separating phosphorus activity from sulfur. (4) The Dowex-50 Fe(OH) column was prepared by the method given by McIsaac and Voigt. (85) Dowex-50 (100 - 200 mesh) obtained from Bio * Rad Company, Berkeley, California, was filled into the column and washed with distilled water. Conc. HC1 and again distilled water was passed through the column until eluate did not change red litmus. A solution 1 N in FeCl and 0.1 N in HC1 was then passed through the column until eluate turned yellow. It was then rinsed with water to free it from excess iron. 50 ml. of 2 N NH40H was passed through next,turning the resin purple in color. The column was finally rinsed with distilled water to remove excess ammonia.

202 B. Chemical Separation The chemical procedure employed for separating the phosphorus reaction product from the bombarded target is given in Table XXIII. The unique method reported by McIsaac and Voigt (85) for the separation of trace phosphorus from macro amounts of sulfur by the use of a kation-exchange resin in the ferric-hydroxide form was applied for the present purpose. It had been known previously that Fe(OH)5 precipitated from a FeC13 solution by ammonia is a good carrier for any phosphorus activity present as phosphate. McIsaac and Voigt produced this precipitate inside the particles of a cation exchange resin and found that trace phosphorus will be retained by the resin from a solution containing macro amounts of sulfate. As the Fe(OH)3 precipitated from a strongly alkaline solution is less able to absorb phosphate ions, the phosphate traces were found to be eluted from the resin bed by 0.125 N NaOH. Carrier-free phosphorus is obtained from this eluate by passing the solution through another column charged with a cation exchange resin in the hydrogen form. The steps preceding this separation as given in Table XXIII' aim to free the solution containing the dissolved ZnS target from its content of radioactive metallic ions and render a solution identical to the one employed by McIsaac and Voigt.

203 C. Tracer for the S34(d, a)p32 Reaction The only available long-lived phosphorus isotope other than P32 is P33 It has practically all the disadvantages a tracer can have. It is difficult to obtain at all' practically impossible to obtain in isotopically pure form and in sufficient quantity. The half-life of P3 *32 only 1.7 times that of the P has not been determined very accurately making the resolution of the decay curves difficult. The activity of P33 is only a relatively weak P with no 7 ray accompanying its decay. The determination of the cross section for the above reaction thus offered its greatest challenge in this tracing step. Phosphorus-33 is produced in a nuclear reactor by the (n, p) reaction on sulfur. It thus accompanies all P32 activity obtained by this method as an impurity of approximately 1% the initial activity. Due to the difference in half-life the P33 content of a decaying p32 sample is enriched with time, although its absolute amount decreases. To obtain sufficient P33 tracer for this research, m32 a 10-mc P2 shipment was obtained on April 5, 1956, from the Oak Ridge National Laboratory and permitted to decay until November 1956. The relative abundance of P33 in this sample was thus increased to approximately 66%. The phosphorus tracer was purified from any S35 content it might contain and other impurities by the chemical procedure given in Table XXIII.

204 The method for tracing the chemical yield of the bombardment was the following: Two samples of the P3- P tracer mixture were prepared and counted in the 47 counter. The self-absorption for the P33 activity was found as described in section III-D-1 of part 2 and the ratio of the counting rates determined. One of the tracer samples was used for tracing the bombardment, the other was preserved as reference and followed in its decay. The amount of P3 both in the tracer sample and the bombardment product was resolved from the decay curve, after corrections for selfabsorption had been applied. The chemical yield of the separation was thus established. The amount of P32 added with the tracer P3 to the bombardment product was calculated from the relative amount in the preserved sample and subtracted from the 32 fraction in the bombardment product to obtain the absolute disintegration rate of the fraction produced by the S34(d, a)p32 reaction only.

205 D. Record of the Bombardments Three bombardments were performed to investigate the S3 (d, a)p32 reaction. Bombardment 1 served to check the chemical procedure and determine the half-life of the bombardment product. The latter was followed for 11 halflives with the 4w counter. The decay curve of this sample is seen in Fig. 65. The half-life was determined by the least squares method applied to the logarithms of the count rates and found to be 14.221 + 0.005 days. From this result it could be assumed that no P33 was produced during the bombardment, which served as proof that the threshold of the (d, an) reaction was not exceeded by the 7.8-Mev deuterons even for elements as low in the periodic table as sulfur. The absolute cross section for 7. 8-Mev deuterons was determined by bombardment 16. The target was exposed to the full-energy beam and the phosphorus product separated several days after the bombardment. The tracer mixture of 33 32 P - P added to determine the chemical yield counted approximately 69,000 c/m in the 4w counter. The selfabsorption correction necessary for evaluation of the J P33 content of the purified sample from bombardment 16 was determined to be 4 + 0.5%. (Section II-D-5 of part 2.) Table XXIV gives the data and results from this bombardment. The errors reported were estimated from the principles discussed earlier (section II-E). They are in general *No- leas squares met Non-weighted least squares method was used.

206 Cd +1o o LCn 0 0 ~H rc\ H H CM r-\ + + 1 +| 4 Hrl 0 0 <0 K> 00c r \ r'\ cj k0o rc (j \S Wc H ) H -p 5= e u u u u u O 2 CO CO O O c CO O r- ~. I I.... c d n n o o oCH a)O r r H H H H H H p i I. 4o O h a, 0 o,*~~~P~ ~~c I) a) 0 CI) 4-P l -o. rQ o kp doo o HP H ^ ^ ^o 0g ~r^ -^- H m U)i 5:1 I+J +1 +1 + 1 +1 1 0 *H' I H 0O N LC n0 P 0 -p -P H 4 H Ln o \C 0V ) C r- ^ ^ ^ C) r c o0.0 co o - c\ Ci o P a) v c- r + ) O, d a) Ci 4 4C a) 0 U n LC L LO LC LCn Cd * ~1 +1 +J+I +1 +-1I 4 — 0) a) a U) 0 4-P ^a) r^Qo C) +-t- 1 U<a) H ^ 0 a) P! ~ ~~> v0~ ~ r I rI I I I 4 -4 b C) Ln C 00 Hr 0 N a) 0 0 r4 r4H wO O O U'A EPa, a)?-...... <4, a -P * * * 1H U) H H H H H O I Q a) O) a) a) a) O H pq

207 somewhat large due to the difficulties experienced with analyzing the decay curves. The error of the target thickness has to be ascribed to the uncertainties introduced by weighing the ZnS deposits by difference. Bombardment 3 served to establish the relative excitation function for the S3 (d, a)p32 reaction. The target arrangement for the bombardment is given in Table XXV. The targets were worked up simultaneously by the carrierfree procedure given in Table XXIII,and the purified p52 products counted over a period of several months with the proportional counter. A series of straight decay curves was obtained for the samples with little scatter of the individual points. A graph of the excitation functions is seen in Fig. 66. E. Resolution of the Decay Curves The difficulties in resolving the decay curves of the tracer and the sample from bombardment 16 may be illustrated by Fig. 67. The decay curve of the tracer is given to illustrate the amount of 33 contained in the six-month-old shipment from the Oak Ridge National Laboratory. Analysis of the curve was performed by the method discussed in section III-D for the case of the molybdenum bombardments. Although the half-life of p32 was experimentally checked in this research, uncertainties in the "known" half-life of P33 made the analysis difficult. To obtain the resolved curve seen in the insert of Fig. 67 a half-life of 24.2 days

208 Table XXV. Arrangement for Zinc Sulfide Targets for Bombardment 3 2 Target or Absorber Thickness (mg/cm2) Average Deuteron Energy Mev ZnS XI 1.39 7.74 Mylar 3.52 Aluminum 10.78 ZnS VII 1.594 6.55 Mylar 3.52 Aluminum 10.78 ZnS V 1.786 5.05 Mylar 3.52 Aluminum 10.78 ZnS XII 1.902 3.15 Mylar 3.52 ZnS IV 1.967 0.7 Mylar 3.52

209 0 Excitation Function for S2"(d,c) P" / 200 C) z 100 2 50 0 *,I I I i I I 1 2 3 4 5 6 7 8 DEUTERON ENERGY (Mev.) Fig. 66. Excitation Function for the S34(d, a)p32 Reaction

210 lOx 106 DECAY OF PHOSPHORUS (47r PROPORTIONAL COUNTER) 180,000 170,000- RESOLUTION OF TRACER PHOSPHORUS INTO 24.2d P33AND 1422d P32 _160,000- 0-TIME=11-8-1956;OAM 150,000 - 140,000; 130,000 120,000 1 20,000 100,000 — 0 80,000 05 I 2 3 4 5 6 7 8.9 1.0 e-aXt 0 10 x 105 HLI 0 Ll F -\ ~ () \0 0 0 0% 0 580h P33 o 341h p32 10 x 104 500 1000 1500 2000 2500 TIME (HOURS) Fig. 67. Complex Beta Decay Curve of P2 P33 Tracer

211 seemed to fit the data best. Taking this into consideration it may be that the error quoted for the absolute cross section in Table XXIV is somewhat optimistic.

212 VI. The Fe56(d, a)Mn54 Reaction For calibration of the x-ray counter it was desirable to have, a standard sample emitting characteristic x-rays of approximately 6 kev. Manganese-54, an isotope decaying by electron capture with the emission of a 0.84-Mev Y ray and a half-life of 290 days seemed well suited for this purpose. To produce a sufficient quantity of this isotope a bombardment of an iron foil was obtained at the Argonne National Laboratory with the 21-Mev deuteron beam. The (d, a) reaction on the natural mixture of iron isotopes yields a certain amount of short"lived products as can be seen from the chart of the nuclides in the iron region given in Fig. 68. Interference from these products, however, was avoided by permitting them to die out before the remaining Mn5 was used. A. Target The aim of the bombardment was to produce a relatively high yield of Mn4. To achieve this in a two-hour bombardment a relatively thick 10-mil pure iron foil was silver-soldered onto a thick copper backing plate to withstand the 100 micro-ampere current of the cyclotron beam. Special solder containing the eutecticmixture of silver and copper was flowed onto the copper backing plate with a flux consisting of 50~ H=B0 and 50~ Na2Br4 010 HO0

213.. * - -. ty 0 s t)w e ) o < onto I -p 0a~C = -'-" <0:C) ag 10 ^.o ES r2^i^ 00C o~~~ I 0 0 -" - li"J.o 0 U in a. n ~ a (0m~~ S104-)l0 co W W Io X O6 — u l W. 0a 0O cu L~~~~~~~W LW -

214 plus water and the iron foil laid on the liquid solder. Excess flux and solder was removed and the target exposed to the deuteron beam. B. Chemical Separation The chemical separation procedure used to separate the manganese reaction product from iron and cobalt as well as from copper, zinc, nickel, silver and cadmium combined an initial solvent extraction step with simultaneous precipitation and ion-exchange steps. The procedure is given in Table XXVII. The solvent extraction step employs a continuous extractor described in detail by Willard and Diehl (128) to remove all iron from the solution of the target with a minimum amount of handling. (The target had a radiation intensity of approximately 1 roentgen at a distance of 2 cm). This step was based on the data reported by Dodson, Forey and Swift on the extractability of certain metals from a hydrochloric acid medium by isopropyl ether as given in Table XXVI. The simultaneous precipitation steps as well as the anion exchange steps follow the principles discussed in section III-B-2 of part 2.

215 Table XXVI The Extraction of Various Elements from 7.75 N HC1 Solutions by an Equal Volume of Isopropyl Ether* Element % In Isopropyl Ether Layer III Fe11 99.00 CuII 0.00 CoII 0.00 MnII 0.00 Ni11 0.00 Al1" 0.00 Cr < 0.0 01 TiIV 0.00 VV 0.00 S0O in presence 0.30 of Fe P04 in presence trace of Fe Mo in presence trace of Fe V + IV trace C. Identification of Reaction Product and Side Products 54 The Mn5 reaction product was used as a standard for the x-ray counter. Its x-ray photopeak as seen in Fig. 37 and the r spectrum seen in Fig. 47 identified the isotope. * Dodson, R. D., Farney, G. J., Swift, E. H., J. Am. Chem. Soc. 58, 2573 (1936). The extraction of ferric chloride from hydrochloric acid solutions by isopropyl ether.

216 Table XXVII Chemical Separation of Manganese from an Iron Target (carrier-free) Element Separated from Decontamination Factor Mn (carrier-free) d. bombarded Fe, 10 Cu and Ag Time of Separation Chemical Yield 10 hours 50% Chemicals and Equipment Continuous extractor(128); blowtorch; Pyrex beaker, 200 ml., 50 ml.; filter crucible; Erlenmeyer flask, 250 ml.; porpous plate; aluminum foil; 2 centrifuge tubes, 15 ml. centrifuge; platinum wire, 6 inch; 2 small ion exchange columns (Fig. 20); medicine droppers. Dowex-2 resin (200 - 400 mesh); conc. HNO, HC1, NH4OH; conductivity water; iso-propyl ether; HCl-gas tank; iron carrier (10 mg/ml). Procedures (1) Heat target in hood and free iron foil from backing plate. (2) Place iron foil with adhering solder in beaker and add conc. HNO. Warm gently and permit copper and silver to dissolve. Decant green solution and store after ebullition ceases. (3) Dissolve remaining iron foil in dilute HC1. (4) After the iron is dissolved filter off any AgCl present, add 5 ml. conc. HNO3 and evaporate to dark yellow syrup of about 10 ml. (5) Transfer to bulb of continuous extraction apparatus and add 25 ml. conc. HC1 and 5 ml. H20 to make about 9 M in HC1. (6) Place approximately 175 ml. iso-propyl ether into 250 ml. Erlenmeyer flask; add a small piece of porous plate, and connect to extractor and place condenser on top of extractor. Wrap extractor bulb with aluminum foil and place in beaker with water. Heat the Erlenmeyer flask to boil moderately and permit to extract in dark over night.

217 Table XXVII (continued) (7) Evaporate extracted aqueous solution to near dryness, take up with few ml. conc. HN0O. Add 3 drops Fe-carrier. (8) Transfer to 15 ml. centrifuge cone and precipitate with NH4OH. Wash twice and dissolve with dil. HC1. (9) Precipitate with NH40H and wash twice. (10) Dissolve precipitate with conc. HC1 and saturate with HC1 gas. (11) Apply solution to an anion exchange column charged with approximately 1 ml. Dowex-2 resin saturated with conc. HC1. (12) Permit to adsorb. Add 5 drops conc. HC1 and soak in slowly. (13) Elute Mn54 with 12 ml. of conc. HC1. (14) Evaporate eluate to near dryness. Take up with 9 M HC1. (15) Apply to second column charged with approximately 1 ml. Dowex-2 saturated with 9 M HC1. (16) Permit to adsorb. Add 5 drops 9 M HC1 and soak in slowly. (17) Elute Mn54 with 12 ml. 9 M HC1.

218 Side products such as Cd9 and Co from the above bombardment were separated by carrier-methods from the solutions encountered in above procedure and were also used for standards. The x-ray spectrum for Cd109 is seen in Fig. 38.

Part 4. DISCUSSION I. Nuclear Shell Model In recent years vast progress has been made in establishing theories regarding the atomic nucleus. One of the most successful attempts in describing the behavior of nuclear particles inside the nucleus is presented by the "Nuclear Shell Model". The foundations for this model were laid by Hartree (57), Elsasser (28 - 32) and Bethe (10) in the 1930's in their attempts to find an "Independent Particle Model" for the structure of the nucleus. In 1948 Maria G. Mayer (81 - 2) pointed out several significant discontinuities in the trend of the properties of nuclei as a function of the atomic number, which seemed to indicate the existence of distinctenergy shells inside the nucleus. The new idea bringing forth the shell model was conceived simultaneously by Maria Mayer (83) and Haxel, Jenson and Suess (60) in 1949. Assuming strong spin-orbit coupling of the nuclear particles resulting in splitting of the nuclear energy levels, the shell model can account for the discontinuities of nuclear properties and permits the calculation of these properties by an adaptation to the independent particle model as known before. 219

220 The level scheme proposed by the Shell Model is given in Fig. 69 indicating the presence of relatively large energy-level spacings at the points for which discontinuities of nuclear properties had been proven. The discontinuities occur at the "magic numbers" 2, 8, 20, 28, (40), 50, 82, 126. A. The (n, y) Reaction Cross Sections One of the postulates of the Nuclear Shell Model is the relatively great stability of nuclei containing "magic" numbers of either neutrons or protons. The nuclear reaction cross section is a quantity related to the stability of the nucleus. Hughes (66) was thus able to prove the greater stability of closed shell nuclei with a plot of the capture cross sections for thermal neutrons as given in Fig. 70. In this plot certain isotopes such as Kr86, Rb87 Sr88; Xe136 Ba138 pb208 Bi209 etc., containing a magic number of neutrons, reveal significantly lower cross-section values. This reveals a certain inertness toward nuclear reactions for these nuclei and is an indication of their relatively greater stability.

221 "-1-% — l —-— (16)-[184 184184, Y23d3/~2 —--- (4)4_ ___ - ________________4 s_____2 (2)even _ —_',__:_ -3d -— < 2g —--— 2 ---- (8)- 6 h3, 0- 1 —- -i — 2-(12)even — 2 — ----- 3 — -3d 5 - (6)/___/ " --' 2g9 -- (10)-— \ k-1 i-2 -(14) - [126] -- 126 3p c=_- 3p — (2)— 3px- (4)_ ____^1 —(14 —---- [2f - (6) - 5 hfw 2f 2f — (8)-[100] odd 1h9/2 -- (10)3 s --— 3s3s/2 (2)2d ---- 2d - (4)— 2d 4 -_o < - -- -— 2d5/2 (6)-[64] even 1 ---- (8)------- 1 — lg9/2 ----- (10)-[-50] 50 - 2p (2)- [40] 3hw 2p -— 1 (6).- 38] odd -if2p- -2p3/2 (4)-' —---- \f /2 (8)-[28] 28 2hw (-2s 1 d -2 — (4)-[20] -20 even ild - 2s (2) [16] ---— ld5g (6)_ [14] h1 h _ — 1p (2)- [8] 8 odd -.-lp.2 (4)- [6] 0 ------ s (2)- [2] 2 Fig. 69. Schematic Diagram of Nuclear Level Systems with Spin-Orbit Coupling

222 - tc * 0 -< -- - 0 * (\J 0 U) 0) 0 0 0 n Oj CZC~ c - LI-, C * ^ * ^ ^ -- - ~^4) - o- - I I (Y * 5C) d 0\ ~ * *o _ _ _ _ _ _ _ _ _ _ __o __ o ^ 6 *.' - "4 * " h *?-1c __________________________*_______________________ PE

223 B. The (p, n) Reaction Cross Sections To illustrate the same principle by means of the (p, n) reaction Blaser, Boehm, Marmier and Scherrer (13) investigated the cross sections and excitation functions for the (p, n) reactions of many target isotopes. A plot of the (p, n) cross sections which they determined for 6.7-Mev protons vs. the neutron content of the target nuclei resulted in values for magic-number nuclei lying below the general trend. The values for 6.7-Mev protons, however, did not show significant deviation for all magic nuclei. Those deviating most significantly in their (p, n) cross section values belonged to so called "double magic" nuclei, featuring a magic number content of both neutrons and protons. C. The (d, a) Reaction Cross Sections Essentially the same type of investigation as mentioned above for the (p, n) reaction was attempted by the present research for the (d, a) reaction. It was expected that the deviation of the cross sections for magic-number nuclei would be much less significant than for the two reactions mentioned above and, that measurements giving values with an accuracy better than 10% would be required, to prove any significant difference in the crosssection values for closed-shell nuclei as compared to

224 their non-magic neighbors. This expectation was based on the fact that approximately 2.5 times as much energy is released when the 7.8-Mev deuteron interacts with the target nucleus than in the capture of a thermal neutron. (The relatively high-energy deuterons had to be used to give sufficient yields for the nuclear reactions). This energy effect is likely to overshadow any small rest energy differences in the magic-number target nucleus which are relatively important for the (n,y) case (32, 110). The most significant results were expected from the comparison of the cross-section values of a nucleus 90 having closed shells for both protons and neutrons, Zr90 with those of its non-magic isotopes. The results of this research appear to verify the expectations. Figure 71 is a plot of the (d, a) reaction cross sections for 7.7-Mev deuterons as determined in this research. The values for cadmium and magnesium were determined by Hall (43) and are included to round out the picture. The errors for the points in Fig. 71 are listed in the tables for the various reactions (Tables X, XVII, XXI, XXIV). Two closed-shell nuclides with a content of 50 neutrons 92 have been investigated. While the value for Mo9 does not seem to deviate from the trend established by the cadmium, molybdenum and zirconium isotopes, the value for the double-closed-shell Zr90 is seen to lie significantly below it.

-225 U I.I I I I I i - I 11 _ *$S (d,<) REACTIO N CROSSS ECTIONS FOR 7.7 Mev 102 * g DEUTERONS - Mg TI - r) - 0 - z- Ti 10 - jIO Mo Zr - Mo * d Cd Zr ~ Cde Cd 0.I I I I''I 1 iI I'.I I I 10 15 20 25 30 35 40 45 50 55 60 65 NUMBER OF NEUTRONS Fig. 71. (d, a) Reaction Cross Sections for 7.7 Mev Deuterons

226 The relatively high cross-section value for sulfur is not easily explained. Although the decay curve of the phosphorus reaction product as seen in Fig. 65 seems to exclude any admixture of P 3 stemming from the S36(d,an )P3.2 reaction, some of the P32 reaction product may'still originate by the neutron-induced reaction S32(n, p)P32 The latter is also suspected from the curvature of the lowenergy excitation function of the S34(d, a)p32 reaction, as is seen in Fig. 67. A comparison of this curve and those presented in Fig. 52 and Fig. 60, for the zirconium isotopes and Nb9 respectively, will readily elucidate this difference in character. The trend of the (d, a) reaction cross section values is indicated by the points in Fig. 71 and is seen to be distinctly different from that of the (n, y) reaction found in Fig. 70. While the non-magic number cross sections tend to greater values for increasing neutron content of the target nuclei, the (d, a) cross sections decrease with increasing atomic weight. Thermal neutrons do not have to overcome a potential barrier to strike the nucleus. The probability of interaction, i.e., the capture cross section, thus increases with the size of the target nucleus. -In the case of deuterons as bombarding particles a certain minimum energy is required before the projectile can penetrate to the nucleus of the target atom. This threshold energy is the

227 greater, the greater the positive charge of the target nucleus. For a given laboratory energy of the deuterons the threshold energy is exceeded more for low-atomic number nuclei than for those higher in the periodic table. The cross section, however, is energy dependent as could be seen in Fig. 52, and a steep rise of the excitation function is experienced in the vicinity of the threshold. This makes a slight excess of deuteron energy above the threshold account for a large increase of the cross sections (14). To obtain a plot more closely analogous to Fig. 70, the (d, a) reaction cross sections should be plotted for those laboratory energies for which the deuterons have the same kinetic energy when reaching the nuclear surface. Such a plot would partially compensate for the effect of the Coulomb barrier, which is not encountered in the case of neutron capture. To establish such a plot, very exact values for the cross sections must be known for the respective ranges of the excitation functions. The energy range that could be covered by the means available for this research, however, did not suffice for such a plot. It was thus only attempted to determine the absolute cross sections for one energy, and to find some relative values of the excitation functions to determine the character of these functions.

228 A rather complete search for charged-particle-reaction excitation functions reported in the literature was undertaken during the initial stage of this research.* Very few excitation functions for (d, a) reactions were found in this search, thus pointing out the need for additional data of this kind. II. Considerations for Future Experiments Further experiments on the (d, a) reaction cross sections by the techniques presented in this report may be planned as an extension of the present research. The considerations entering into the choice of the reactions should be essentially the same as the ones governing the choices for this investigation. The energy of the available deuterons may exclude as target materials all elements for which the threshold of the (d, a) reaction is not exceeded. For the 7.8-Mev deuterons of the cyclotron of the University of Michigan this would exclude all elements with atomic number greater than 66 (dysprosium) as was found experimentally. The following elements had to be excluded, since the Anders, O. U., Meinke, W. W., Excitation Functions and Cross Sections, document No. 4999,American Documentation Institute, Washington 25, D. C., U. S. A. (1956).

229 (d, a) reaction products form stable nuclei: Be, C and the odd atomic-number elements: N, F, Na, Al, P, Sc, V, Mn, Co, As, Y, Nb, Tc, Rh, In, I, Cs, La, Pm, Tb. Investigation of the (d, a) reactions of these elements would have to follow a completely different approach. This would probably include either detection of the primary a particles emitted during the reaction, or the use of a mass-spectrometer. Of the remaining 44 elements some were excluded because the counting rates of the (d, a) reaction products would be too low to measure due to the low relative abundance of the parent isotopes and too long or too short halflives of the products. These elements are shown in Table XXVIII. Inability to bombard gaseous materials and count gaseous products with the equipment on hand excluded the following elements from consideration: He, Li, Ne, A, K, Kr, Rb, and Xe. Difficulties in analyzing decay curves of the (d, a) reaction products were anticipated for a number of other elements. The critical isotopes of these elements are shown in Table XXIX. Several elements posed difficulties in tracing their (d, a) reaction products through the chemical separations: No suitable tracer could be found for the C137(d, a)S35 reaction since all radioactive sulfur isotopes, but S3, have half-lives oelow 10 minutes.

230 Table XXVIII. Elements Excluded from Investigation because of Half-Lives of (d, a) Reaction Products or Low Abundance of Parent Isotopes Half-Life of (d, a) Element Parent Isotope Abundance Rectn Prodct --— ~- --- - -- - ~ ~Reaction Product B B10 18.8 4x 10-15 sec 18 0 ol8 0.204 7.4 sec 28 6 Si Si 92.17 10 years Si30 3.12 2.3 min Ca Ca4 96.9 7.7 min 42 Ca 0.64 3.9 x 109 years 44 Ca4 2.1 12.5 hours Ca46 0.0032 22 min Ca48 0.18 unknown Br Br81 49.4 7 x 10 years 106 Pd Pd 27.3 4.4 min, 42 sec Ru Ru99 12.7 105 years Ru100 12.7 104 years Ru101 17.0 2 x 105 years Ag Ag109 48.6 7 x 106 years Eu Eu153 52.2 80 years

231 03 4 —)~qo ~ ~ ~ ~ ~ ~ ~ ~~~a rd 0 d rd DH. +- 0o C H H Co - L O ~ \ O a) 02 F 6+0 F5 02 - 02 F 02 PN ~HO?0 ct U 0 Q T i 0 2 t 1 c CC) d a) C k C~ a) cd 0c 02 14I H — H ad c -3 o 0 a d r H C v E h as). 0 >o h T 0 H r r r- - - tE- c>co 4- oC ) c- - Ln 4 4 C rH X H Pc CJ rM HO HC H HO r c O a) u> 0 O 02 O C tzcO 0>4+... Co 0 () H H H H H Oi 4' rO T ci Cii o Z Z U2U2 CQ C l 00 0 0 PL1 mQ a) a) 0 U rC4-)' H ^ 0 2 -..=qO * 0 (D O0 C 02 cH o * ab C\0 0'H 0 c ) 4 ^ ^ i ^ 0 0 03 r.. O I _O-H -,, H f O H * * ~ * * ^ CM a) S 03 0 ) H 0 H > C1 s 2FO02 OH l O'LHO 02 0 r' 02 02 0 O' E I - O 3 0O C0 03 TO LHC H H 4 C H rH Lr rH CM 0 0 H a ) 02 0 a) 03 r0l2- Hl H l H H r CM C CM CM CM I H Hl rH H rlH H H Hr Ht rl rHl H ) ^ a) a ) a0 ) a) (a) a X Pi H X P H a ) eH X

252 Difficulties were anticipated in procuring tracers for the (d, a) reactions on Zn, Ge, Ce, and Pr. As tracer for the Zn(d, a)Cu reactions the 61-hour Cu7 should be used. This isotope can be obtained either by the Zn67(n,p)Cu6 7 reaction or the Zn70(d, an)Cu67 reaction, but both would necessitate the use of isotopically enriched targets. The Ge(d, a)Ga reactions may be traced by the 14.1-hour 72 Ga72 which would have to be produced by a preliminary bombardment of germanium. For the Ce(d, a)La reaction l4l no good tracer is available. La with a 3.8-hour halflife may serve to trace the Cel42(d, a)La 14 reaction but its procurement by the low-yield Ba1 8(a, p)Lal reaction is difficult and close timing of the bombardments has to be observed to make use of it. Without use of enriched Ce140 no pure tracer Ce141 can be obtained which would be necessary to trace the Pr 4(d, a)Ce139 reaction. The use of separated isotopes was, however, avoided in the present research due to their high cost and lack of availability. (d, a) reactions involving Mg, Cd (47) and Cr * had been investigated by previous workers, so that the final choice had to be made from the ten remaining elements. * Kafalas, P., Ph. D. Thesis, Massachusetts Institute of Technology (August 1954).

233 The cross sections of the (d, a) reactions on iron were not measured since the thin, high-purity iron foil necessary could not be obtained during the course of the investigation. 56 58 Nickel bombardments producing 77-day Co56, 71-day Co and 5.2-years Co were not performed, although thin targets could be produced by electrodeposition. The bombardment times necessary for sufficient amounts of reaction products would have been relatively long and the resolution of the decay curves of the products difficult. Gallium and selenium containing no magic number isotopes were of relatively small interest for an investigation of the effects of nuclear shell structure on the (d, a) reactions. This left only S, Ti, Zr, Mo, Sr, and Sb. The (d, a) reaction cross sections of the first four were investigated in the present research. Improvements of the data obtained in this research could be made by employing the various techniques developed for tracing the reaction products more methodically. In this connection it should be kept in mind that the method using a reference sample iS independent of the knowledge of the decay constants of the tracers as well as of the reproducibility of a given counter over a long period of time. The determination of the self-absorption of $ rays in the 4w sources may be studied in order to improve some

234 of the errors quoted. Finally, the analysis of the decay curves may be facilitated by the use of a computer, making the resolution of many-component decay curves quicker and less laborious. This would also permit the resolution of decay curves involving isotopes whose half-lives differ by less than 50 percent. Nevertheless, it is believed that the method of determining the (d, a) reaction cross sections has by now been sufficiently developed to obtain data precise enough for theoretical treatment. The work should now be expanded to include various other nuclides to permit a good quantitative correlation of the trends in the (d, a) reactions. More excitation functions should be measured which could be normalized to the absolute points. The techniques developed in the present research are applicable without alteration to the study of the (d, n) reaction and, with little improvement on the chemical separations, for all reactions for which the product is different than the target element. For such reactions more attention has to be given, however, to effects of side-reactions having the same product.

235 III. Summary The purpose of this research was the determination of the cross sections and excitation functions of various (d, a) reactions induced by the 7.8-Mev deuterons from the University of Michigan cyclotron. Measurements were performed for reactions of isotopes of zirconium, molybdenum, titanium and sulfur. Special attention was given to reactions involving closed-shell nuclei in order to detect any significant deviation of the cross-section values for these reactions when compared with values for non-magic number nuclei. These deviations were expected to be relatively small and precision measurements had to be performed to detect them. New techniques and instruments were developed for this purpose and others improved, which had been evolved by previous workers. It was thus possible to obtain absolute cross-section values with standard errors between 5 and 10 percent and reproducibility within 2 percent. The accuracy obtained for the cross sections measured in this research compared favorably with values reported by other authors in the literature. Blaser, Boehm, Marmier and Scherrer (13) quote errors of the order of 24 millibarns. Aamodt, Peterson and Phillips (1) report values for the C12 (p, pn)C11 reaction cross sections with errors of 7 millibarns. For this research the sensitivities were of the order of 0.1 - 2 millibarns.

236 The values were determined for the relatively lowyield (d, a) reactions. Determination of the cross sections was not possible by measuring the particles emitted during the reaction itself. The reaction products had to be purified by carrier-free procedures and counted absolutely with 4w proportional counters and coincidence methods. This added several steps introducing errors, which could be avoided by workers measuring cross sections by physical means only. For the present research no compromise could be made lessening the achievable accuracy of the measurements, because of the small magnitude of the effects sought. A compromise thus had to be made in the number of reactions investigated. The final selection presents a compromise between the available man-hours and the expected efforts required for precision measurements on a particular isotope. The instruments constructed for measuring the intensity of the cyclotron beam included a current integrator permitting readings with standard deviation of less than 1 percent and a negative high voltage power supply to provide a potential of -1000 volts to the suppressor ring. Chemical separation methods were developed to purify the (d, a) reaction products by carrier-free techniques and prepare carrier-free tracers used for determining the yields of the chemical separations. A beta gauge was designed and built for measuring the evenness of cyclotron targets of a few mg/cm thickness with errors of approximately 1 percent.

237 Various counters, including two 4r-proporticnal counters and a 4-inch diameter x-ray counter, were constructed to - gether with several electronic circuits. The performance of these counters was studied in detail and optimum conditions for proper operation found for them. Absolute counting techniques were developed for counting x-rays and A rays and experience gained in x-ray and y-ray spectroscopy using proportional and scintillation counters. Excitation functions were measured for several of the investigated reactions by the "stacked foils" technique and a rough correlation of the cross section values for 7.7-Mev deuterons attempted.

APPENDIX Integration of equation 4 dN I n dt A Transpose d + x N= I n dt A Let S -.+ S X N be the derivative of some dt combination. dN d aN aS We obtain S d + S XN - (SN) = + N d — d dt at at dS This holds if d dS then S-= X dt S n S = f dt t and r 0 xt S= e =e now + S N S (I T ) now adt 238

239 substituting xt + eXt tI a n for S (dt )+ X N By integration we obtain e N = e (I ) dN + C 0 N ex N = e (I ) XN + C 0 Xt 1 a n Xt t e N = (I -) e 3 X A o Xt 1 an kt 1 a n 1 t e N IAN A ( )( - ) N 1 a n( e at

240 Table XXX. Data for the Beryllium Absorption Curve of X-rays Emitted by Y88 and Mn54 Beryllium Absorber y88 % Transmitted Mn54 % Transmitted mg/cm2 * count rate** count rate** c/m c/m 44.8 17,091 + 8 100 3101 100 67.6 16,907 98.9 67.6 16,764 98.1 2873 92.6 89.6 16,776 98.1 89.6 16,583 97.0 2613 84.2 126.1 16,377 95.8 2239 72.2 178.8 16,129 94.4 1844 59.4 297.5 15,560 91.0 1177 57.9 423.6 15,069 88.2 825 26.6 591.8 14,136 82.7 361 11.6 724.8 305 9.8 1000 12,176 71.2 1000 12,140 71.0 116 3.7 1503 9,839 57.5 2188 8,415 49.2 2475 8,026 47.0 2475 8046 47.1 2846 6726 39.4 Beryllium absorber plus 44.8 mg/cm beryllium counter window. ** Count rate corrected for background statistical variation 1-2% per point,

241 Table XXXI. Data for Aluminum Absorption Curve of Pm147 Beta-rays Absorber Count rate mg/cm2- counts per minute 0.00 13,834, 1.46 9,544 2.72 7,235 3.25 6,594 3.59 6,191 4.66 5,343 6.18 4,143 13.9 1,557 21.8 836 25.1 729 39.7 566 46.6 532 54.3 505 1710 189.5

BIBLIOGRAPHY (1) Aamodt, R. L., Peterson V., Phillips, R., Phys. Rev. 88, 739-44 (1952). C2(p, pn)C11 cross section from threshold to 540 Mev. (2) Alburger, D. E., Sunyar, A. W., Phys. Rev. 99, 695702 (1955). Decay of Bi207. (3) Anders, 0. U., Nucleonics 13, No. 7, 46 (1955). Millicurie beta-ray point source. (4) Anders, 0. U Meinke, W. W., Rev. Sci. Instr. 27, 416-7 (1956Y. Beta gauge for localized measurements on thin films. (5) Anderson, G., Phil. Mag. 45, 621 (1954). Electromagnetic separation of spallation products. (6) Aron, W. A., Hoffman, B. G., Williams, F. C., U. S. Atomic Energy Commission, Rept. AECU-663 (May 1951). Range-energy curves. (7) Bach, D. R., Ph. D. Thesis, University of Michigan (1955s. An investigation of the low-lying levels in 0m. (8) Bayly, J. G., Can, J. Research 28A, 520-9 (1950). Calorimetric measurement of the disintegration rate of a phosphorus-32 source. (9) Beach, E. H., Ph. D. Thesis, University of Michigan (1952). Angular distribution studies of P 1(d, p)P2 reactions. (10) Bernstein, W., Brewer, H. G., Jr., Rubinson, W., Nucleonics 6, No. 2, 39-45 (1950). A proportional counter x-ray spectrometer. (11) Bethe, H. A., Bacher, R. F., Revs. Modern Phys. 8, 103 (1936). Nuclear physics. Stationary states of nuclei. (12) Bilaniuk, 0., Ph. D. Thesis, University of Michigan, Department of Physics, in preparation. 242

243 (13) Blaser, J. P., Boehm F. Marmier, P., Scherrer, P., Helv. Phys. Acta 24, 441-64 (1951). Anregungsfunktionen und Wirkungsquerschnitte der (p, n)Reaktionen (II). (14) Blatt, J. M., Weisskopf, V. F., "Theoretical Nuclear Physics," Wiley, New York, 1952. (15) Brown, J., Felber, F., Richards, J., Saxon, D., Rev. Sci. Inst. 19, 818 (1948). On making thin nylon films. (16) Cacciapuoti, N. B., Nuovo Cimento 14, 213-19 (1938). Determination of the decay constant of P52. (17) Cassatt, W. A., Jr., Decay scheme and characterization study of radioactive products of deuteron reactions. University of Michigan Ph. D. thesis, September 24, 1954. U. S. Atomic Energy Commission, Rept. AECU-2958, 49 (September, 1954). (18) Ibid., p. 77. (19) Cassatt, W. A., Jr., Meinke, W. W Phys. Rev. 99, 760-1 (1955). Decay scheme of Y.2. (20) Chase, R. L Higinbotham, W. A., Rev. Sci. Instr. 23, 34-6 (1952). A flexible pulse amplifier with good overload properties. (21) Chen, J. J. L., Rev. Sci. Inst. 21, 491-2 (1950). Test of rupture strength of thin plastic films. (22) Cohen, B. L., Phys. Rev. 81, 184-8 (1951). (n, 2n) and (n, p) cross sections. (23) Compton, A. H., Allison, S. K., "X-Rays in Theory and Experiment," VanNostrand, New York, 1935. (24) Coryell, C. D., Sugarman, N., "Radiochemical Studies: The Fission Products," McGraw-Hill, New York, 176-83 (1951). Szilard-Chalmers reactions. Principles of enrichment. Article by A. W. Adamson, R. R. Williams, Jr., PAPER 15. (25) Crandall, W. E., Millburn, G. P., Pyle, R. V., Birnbaum, W. Phys. Rev. 101, 329-37 (1956). C12(x, xn)Cli and A127(x, x2pn)Na24 cross sections at high energies.

244 (26) Curran, S. G., Angus, J., Cockroft, A. L., Phil. Mag. 40 (7), 36-52 (1949). Investigation of soft radiations by proportional counter. (27) Drabkin, G. M., Orlov, V. I., Rusinov, L. I., Izvest. Akad. Nauk SSSR Ser. Fiz. 19, 324 (1955), Columbia Tech. Transl. p. 294-37. Study of nuclear isomerism of Zn69, Se79, Se81, Nb95, Rh103 and Ba137. (28) Elsasser, W. M., J. phys. radium (7) 4, 549-56 (1933). Sur le principe de Pauli dans les noyaux. (29) Elsasser, W. M., J. phys. radium (7) 5, 389-97 (1934). Sur le principe de Pauli dans les noyaux. II. (30) Elsasser, W. M., J. phys. radium (7) 5, 635-39 (1934). Sur le principe de Pauli dans les noyaux. III. (31) Elsasser, W. M., J. phys. radium (7) 6, 473-76 (1935). Energies de liaison des noyaux lourds. (32) Elsasser, W. M., Phys. Rev. 51, 55 (1937). The selfconsistent field and Bohr's nuclear model. (33) Fajans, K., Beer, P., Ber. deut. chem. Ges. 46, 3486-97 (1913). Das Verhalten der Radioelemente bei Fgllungsreaktionen. (34) Fajans, K., Erdey-Gruz, T., Z. physik. Chem. 158 (A), 97-151 (1932). Ueber die Adsorption von Ionen, insbesondere der Radioelemente und Farbstoffe an salzartigen Verbindungen. (35) Fajans, K., Physik. Z. 14, 136-42 (1913), Radium 10, 61-5 (1913). The position of the radio-elements in the periodic table. (36) Feather N., Pro. Cambridge Phil. Soc. 34, 599-611 (1938). Further possibilities for the absorption method of investigating the primary beta-particles from radioactive substances. (37) Fogelstrtm, I., Westermark, T., Nucleonics 14, No. 2, 62-5 (1956). Production of P33 by neutron irradiation of sulfur. (38) Francis, J. E Bell, P. R., Kelly, G. G., Nucleonics 12, No. 3, 54-6 (1954). Scintillation counting today: spectrometry. Double-line linear amplifier. (39) Friedlander, G., Perlman M. L., Scharff-Goldhaber, G., Phys. Rev. 80, 1103-4 (1950). Conversion coefficient of the 35-kev gamma-ray of Te125.

245 (40) Frisch, 0. R., "Progress in Nuclear Physics, Vol. 3, Academic Press, New York, 18-62 (1953). Energy measurements with proportional counters. Article by D. West. (41) Geiger, H., Physik. Z. 14, 1129 (1913). Demonstration einer einfachen Methode zur Zahlung von alpha- und beta-Strahlen. (42) Geiger, H., Miller, W., Physik. Z. 29, 839-41 (1928). Das Elektronenzghlrohr. (43) Gile, J. D., Garrison, W. M., Hamilton, J. G., J. Chem. Phys. 18, 1685 (1950). Carrier-free radioisotopes from cyclotron targets. Preparation and isolation of Sc44, 46, 4 from titanium. (44) Ghoshal, S. N., Phys. Rev. 73, 417-18 1948). Excitation curves of (a, n); (a, 2n); a, 3n) reactions on silver. (45) Ghoshal, S. N., U. S. Atomic Energy Commission, Rept. UCRL-709 (July, 1950). An experimental verification of the theory of compound nucleus. (46) Goldhaber, M., Hill, R. D., Revs. Modern Phys. 24, 179-239 (1952). Nuclear isomerism and shell structure. (47) Hall, K. L., Determination of (d, alpha) reaction cross sections, University of Michigan Ph. D. thesis, September 24, 1954. U. S. Atomic Energy Commission, Rept. AECU-3126 (September 1955). (48) Ibid., p. 43. (49) Ibid., p. 48. (50) Ibid., p. 67. (51) Ibid., p. 95. (52) Ibid., p. 112. (53) Ibid., p. 124. (54) Ibid., p. 167. (55) Ibid., p. 176. (56) Ibid., p. 188.

246 (57) Hartree, D. R., Cambridge Phil. Soc. 24, 89-132 (1928). The wave mechanics of an atom with a non-coulomb central field. Theory and methods. (58) Harvey, J. A., Phys. Rev. 81, 353-64 (1951). Neutron binding energies from (d, p) reactions and nuclear shell structure. (59) Hawkings, R. C., Merritt, W. F., Craven, J. H., Proceedings of a Symposium "Recent Developments and Techniques in the Maintenance of Standards," held at the National Physical Laboratory (May 1951). (60) Haxel, 0., Jenson, J. H. D., Suess, H. E., Z. Physik 128, 295-311 (1950). Modellmaessige Deutung der ausgezeichneten Nukleonenzahlen im Kernbau. (61) Hicks, H. G., Gilbert, R. S., Stevenson, P. C., Hutchin, W. H., U. S. Atomic Energy Commission, Rept. LRL-65 (December 1953). The qualitative anionic behavior of a number of metals with an ion exchange resin, "Dowex 2". (62) Higinbotham, W. A. Rankowitz, S., Rev. Sci. Instr. 22, 688-90 (1951). A combined current indicator and integrator. (63) Hintz, N. M., Ramsey, N. F., Phys. Rev. 88, 19-27 (1952). Excitation functions to 100 Mev. (64) Hollander, J. M., Perlman, I., Seaborg, G. T., Rev. Mod. Phys. 25, 469-651 (1953). Table of Isotopes. (65) Hughes, D. J., Harvey, J. A. U. S. Atomic Energy Commission, Rept. BNL-325 (July 1955). Neutron cross sections. (66) Hughes, D. J., Sherman, D., Phys. Rev. 78, 632-3 (1950). Fast neutron cross sections and nuclear shells. (67) Hyde, E. K., Florence, M. G., Larsh, A. E. Phys. Rev. 97, 1255-58 (1955). Isomeric state of yo8 in the decay of Zr8. (68) Keister G. L., Schmidt, F. H., Phys. Rev. 93, 140-5 (1954. The second-forbidden beta spectra of Coo and Sc46. (69) Kemp, E. L., U. S. Atomic Energy Commission, Rept. LA-1207 (July 1950). Scaler model SC-3.

247 (70) King, J. S., Ph. D. Thesis, University of Michigan (1953). Angular distributions from the reactions C1)(d, p)C136, Sc45(d, p)Sc46, and V51(d, p)V52 as a test of the shell model. (71) Kistner, 0. C., Schwarzschild, A., Rustad, B. M., Phys. Rev. 105, 1339-41 (1957). Positron spectrum of 015. (72) Klema E. D., Hanson, A. 0., Phys. Rev. 73 106-10 (1948). A determination of the S32(n, p1p32 cross section for neutrons having energies of 1.6 to 5.8 Mev. (73) Korttim, G. F. A., Buchholz-Meisenheimer, H., "Die Theorie der Destillation und Extraktion von Fl1ssigkeiten," Springer, Berlin, 1952. (74) Kraus, K. A., Nelson, F., "Proceedings of the International Conference on the Peaceful Uses of Atomic Energy," Vol. 3, 113-25 (1956). Anion exchange studies of the fission products. (75) Kunin, R., Myers, R. J., "Ion Exchange Resins," Wiley, New York, 1950. (76) Larkins, J. H., U. S. Atomic Energy Commission, Rept. LA-_1238 (May 1951). Thin window proportional betachamber. (77) Lillie, A. B., Conner, J. P., Rev. Sci. Instr. 22, 210-11 (1951). Preparation of thin tritium targets. (78, Lockett, E. E., Thomas, R. H., Nucleonics 11, No. 3, 14-17 (1953). The half-lives of several radioisotopes. (79) Lyon, W. S., Kahn, B., Phys. Rev. 99, 728-30 (1955). Decay of Sc 7. (80) Mann, W. B., Seliger, H. H., J. Research Natl. Bur. Standards 50, 197-200 (1953). Refinements in radioactive standardization by 47 beta counting. (81) Mayer, M. G., Phys. Rev. 74, 235-9 (1948). On closed shells in nuclei. (82) Mayer, M. G., Phys. Rev. 75, 1969-70 (1949). On closed shells in nuclei. (83) Mayer, M. G., Phys. Rev. 78, 16-23 (1950). Nuclear configurations in the spin-orbit coupling model. Empirical evidence.

248 (84) Mayer, M. G., Jensen, J. H. D., "Elementary Theory of Nuclear Shell Structure," Wiley, New York, 1955. (85) McIsaac, L. D., Voigt, A., U. S. Atomic Energy Commission, Rept. ISC-271 (June 1952). A new technique for the separation of radiophosphorus from sulfur. (86) McMillan, E. M., Miller, R. D., Phys. Rev. 73, 80-1 (1948). Absolute cross section for the reaction Cl2(p, pn)C11 at high energy. (87) Meinke, W. W., Ph. D. Thesis, University of California, 46 (April 1954). High energy bombardment products of thorium. (88) Meinke, W. W., Ph. D. Thesis, University of California, 76 (April, 1954). High energy bombardment products of thorium. (89) Meinke, W. W., U. S. Atomic Energy Commission, Rept. UCRL-432 (August 1949). Chemical procedure used in bombardment work at Berkeley. (90) Meinke, W. W., Wick, G. C., Seaborg, G. T., J. Inorg. Nuclear Chem. 3, No. 2, 69-92 (1956). High-energy excitation functions in the heavy region. (91) Moseley H. G. J., Fajans, K., Phil. Mag. 22, 629-38 (1911Y. Radioactive products of short lives. (92) Mulder, D., Hoeksema, G. W., Sizoo, G. J., Physica 7, 849-59 (1940). Measurements on the period of radioactive phosphorus. (93) Nachod, F. C., "Ion Exchange: Theory and Application," Academic Press, New York, 1949. (94) Parkinson, W. C., private communication. (95) Pate, B. D., Yaffe, L., Can. J. Chem. 33, 15-23 (1955). A new material and techniques for the fabrication and measurement of very thin films for use in 4w counting. (96) Pate, B. D., Yaffe, L., Can. J. Chem. 33, 929-37 (1955). Disintegration-rate determination by 4w counting. Source-mount absorption correction. (97) Peacock, W. C., Jones, J. W., U. S. Atomic Energy Commission, Rept. AECD-1812 (March 1948). Decay scheme of Y88.

249 (98) Perkins, L. R., E. I. Du Pont de Nemours and Co. Polychemical Dept. Research Div., private communication to W. W. Meinke. (99) Preiswerk, P., Stahelin, P., Helv. Phys. Acta 24, 300 (1951). Term diagrams of several niobium isotopes. (100) Rankowitz, S., private communication to W. W. Meinke. (101) Rexroth, V. G., U. S. Atomic Energy Commission, Rept. LA-1239 (May 1951). Model PA-2 pulse amplifier. (102) Rich, M., Madey, R., U. S. Atomic Energy Commission, Rept. UCRL-2301 (March 1954). Range-energy tables. (103) Rieman, W. III, Neuss, J. D., Naiman, B., "Quantitative Analysis, a Theoretical Approach," McGrawHill, New York, 1951. (104) Rubinson, W., Bernstein, W. U. S. Atomic Energy Commission, Rept. BNL-132(S-11), 104-110 (July 1951). Emission of Pb x-rays in the decay of Po210. (105) Rubinson, W., Bernstein, W., U. S. Atomic Energy Commission, Rept. BNL-93(S-8), 98-107 (October 1950). Emission of lead x-rays in the decay of Po210. (106) Rutherford E., Geiger, H., Pro. Roy. Soc. (London) 81(A), 141-61 (1908). An electrical method of counting the number of alpha-particles from radio-active substances. (107) Rybakov, B. V., Zhur. Eksptl'.i Teoret. Fiz. 28, 651-4 (1955). Range of protons in medium and heavy elements. (108) Schmeling, P., Z. Naturforschun 10A, 79 (1955). Ueber die Halbwertszeit von Y 0. (109) Segr6, E., "Experimental Nuclear Physics " Vol. 1, Wiley, New York, 149 (1953). Geiger-MAller Counters. (110) Segrt, E., "Experimental Nuclear Physics," Vol. 1, Wiley, New York, 318 (1953). Energy momentum relations. (111) Siegbahn, K., ed., "Beta- and Gamma-Ray Spectroscopy," Interscience, New York, 24-51 (1955). Interaction of gamma-radiation with matter. Article by C.- M. Davisson.

250 (112) Siegbahn, K., ed., "Beta- and Gamma-Ray Spectroscopy," Interscience, New York, 201-223 (1955). The coincidence method. Article by A. C. G. Mitchell. (113) Siegbahn, K., ed., "Beta- and Gamma-Ray Spectroscopy," Interscience, New York, 259-60 (1955). Source and window technique. Article by H. Sltis. (114) Siegbahn, K., ed., "Beta- and Gamma-Ray Spectroscopy," Interscience, New York, 834 (1955). Measurement of disintegration rate. 47 scintillation counter. Article by J. L. Putnam. (115) Siegbahn, K., ed., "Beta- and Gamma-Ray Spectroscopy," Interscience, New York, 835 (1955). Measurement of disintegration rate. Beta-gamma coincidence method. Article by J. L. Putnam. (116) Sinclair, W. K., Holloway, A. F., Nature 167, 365 (1951). Half-lives of some radioactive isotopes. (117) Smith, W. G., Hamilton, J. H., Robinson, R. L. Langer, L. M., Phys. Rev. 104, 1020-25 (1956). Beta decay of Tb.-. (118) Soddy, F., Nature 91, 57-8 (1913), The radioelements and the periodic law. (119) Strohmaier, K., Z. Naturforschung 6A, 508-9 (1951). Ein einfaches Verfahren zur Herstellung grossflchiger durchsichtiger Haute aus Aluminiumoxyd mit einer Dicke zwischen 50 und 300 mu. (120) Sun, K. H., Pecjak, F. A., Nucleonics 14, No. 11, 122-6 (1956). Recoil separation of isotopes. (121) Szilard, L., Chalmers, T. A., Nature 134, 462 (1934). Chemical separation of the radioactive element from its bombarded isotope in the Fermi effect. (122) Thompson, W. E., U. S. Atomic Energy Commission, Rept. ORNL-1056 (November 1951). 2. Instrument research and development quarterly progress report for period ending April 20, 1951. Article by Borkowski. (123) Treybal, R. E., "Liquid Extraction," McGraw-Hill, New York, 1951. (124) Way, K., King, R. W., McGinnis, C. L., van Lieshout, R., U. S. Atomic Energy Commission, Rept. TID-5300 (June 1955). Nuclear level schemes A = 40 - A = 92.

251 (125) Weast, R. C., Hodgman, C. D., Wallace, C. W.,"Handbook of Chemistry and Physics," Chemical Rubber Publishing Co., Cleveland, Ohio, 1955. (126) Willard, H. H., Tang, N. K., Ind. Eng. Chem. Anal. Ed. 9, 357-63 (1937). Quantitative determination of aluminum by precipitation with urea. (127) Willard, H. H. Tang, N. K., J. Am. Chem. Soc. 59, 1190-6 (19375. A study of the precipitation of aluminum basic sulfate by urea. (128) Willard, H. H., Diehl, H., "Advanced Quantitative Analysis," Van Nostrand Co., New York, 56, 1950.

UNIVERSITY OF MICHIGAN 3 9015 02499 5592