THE UNIVERSITY OF MICHIGAN INDUSTRY PROGRAM OF THE COLLEGE OF ENGINEERING DETERMINATION OF (d, alpha) REACTION CROSS SECTIONS Kenneth Lynn Hall This dissertation was submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the University of Michigan. The research was partially supported by the Atomic Energy Commission under University Project No. 7, Contract No. AT (11-1)-70. November, 1955 IP-134

ACKNOWLEDGEMENT We wish to express our appreciation to the author for permission to distribute this dissertation under the Industry Program of the College of Engineeringo

.,lb. t r a c at 4t; c~,f' maes iX i bsolute (d,alpha) "-eaotion woss S',ections tei ment of U- i "ob em rit ion of Apparatus e.;i..t j.j, 30 X ~_LLo A 5 e ki i "i 6 e {i' b o r i c','' } ~J r J _'> t i o n':' O<erationz 4 -, " etal S rair tor 24 i.' e @ t r i'-a 1.. ~ o i:'_ 2e SI ctric a R r'l 26. acuum -,.'ste m 28 I;L aFra is ri T s 33' - ct i t n 34.j.i!,. i'ar" ) C e 44 i',ii

iage Suppression o'f 2eco -dar-y lactrons 44 1, i'reliminary Experiments 45; 2.:secondary l'ectrons Emitted 48 from t-'e Cup 3,.econldary!! ectron2 llitied a49 from t! t e rarget irimary Calibrat ion of th;e 2eam 52 Inte grator Io D3ire;ct; lrelent of3 Carrent 54 2* o tent etiomletetr ethod 5c J.condar Ca ihrations 62. Taret; I::reparation 63?Y!a o. ~_esium 6 2f1 f r 67 Cadmnium 0,..,lei~cal, e.aratios~ 72.at1l ts 75 F >los*;l,3oy r n iltl~r, -r r.Ur S s 7 C; ancaui i s.l f - u at n i w&,i Tar: Ft Ls'1 ilver from a'ad inum'. ar. r,4"es 3-3.:Vsotin-l tllr U Vteta C 7ountin L a + i}tv,7i 7 C o 1 1 ti, di i: ii c i n 7 c nt v iii

Pa e 41T Proportional Counting 03 1. i'lastic Film l;reparation 95 2. Film Thickness t 9easurer:ent 3. Film bs0orption Correction 102 4. Jie1f Ai sorption 1i0. ~f:ect of a Conducting Coating 110 6. Effect of'ain, "Trig-r and 114 Vo 1 tae ettings 7.'ead Time Correction 117 Kc. Counting of National Bureau 124 of Otandards Sanimpes Expori:iental esults 128 A. Derivation of D2Jquatioi2s 12, Cross - ction 12 Chenical Yield 133 3. oi- ardment I —roceeure e140 C,*; ~iagnesium BUomL1 ard (i-Jments 141 Target 1'reparation 143 aii cal I- ocedure 145r sisolute -i'eta Coun;ting 146 An rJ..sis of iJCa' Curves 151 Ha 1~ e?'s'i., 152!" 22 124 Cross Jection for'a22 and Ta24 155 )o.'ulfuxr 3Bombardments 158 Target i'reparat io n 160 Chemical r'roce:.re 160 9 K.:solute I3eta Couinting 163 iv

Page Analysis of i)ecay Curves 166 Cross Section for P32 167.'Litanium BoDbardments 169 Beam Detemrination 172 TarSet PIreparation 172 Chenlica1 - roced1re 173 Absolute -3eta Counting 173 Analysis o-f:ecay Curves 177 Cross,ection for c6 1,1 0 F. Cadmium Bombardments 181 raret i -reparation 1 83 c-em icl - rcedre 1 ii3.bsolute u Ieta Countin. 185 rInalysis of Jecay Curve 185 CJross.Sbection for "']..2 hr,0, and g12 15 G. Loss of tecoil Auclei Sci'H. Enerty of tue 9euteron i3eamr 19 uiscussion 191 191 Part II: Mass Assinrmeent of A 104 introduction 1 5 ExperimiRental r rocedures 1r'6 o Targ et 196. C-lemlicai yeparations 197 C. Cointin 200 v

Page Experimental Results 201 Discussion 203 Appendix I 205 Appendix II 208 Bibliography 212 vi

LIST 0? TAB-I'I' Table o_. Title Pa I Target Tlaterials Ii Spectographic Analysis of Foote 70. itani m III S.,pecto,ra1p.iic Analysis of Johnson, 72 MJatthey Cadmlium I'cods L IV S-rface i)ensity Data for Films made 101 from 33 lper cent Zapon Solution V Angliquot Reproducibility 108 VI Sunmmary of Bombardment iData for 144 I g(d, ) Na VII.nalysis of LecaJ Curves of Jodiu 153 VITI Suramarv,7 of Malf -ives of Ia4 154 IX Propasation of E rrors in Cross 157 Section for 1( d, a)Na Summarv of Bombard.ment iData for 161 S(d, c)P u1 mmnary of rombardm;ent Data for 171 Ti(d, a)e ZISI Surtmulary ofa Bombardment J-at1a for 184 Cd(d,c)A.g i{I! V-ass AnalIsis of Cadmium Cxide 197'Tar' et [.lat r ials KULV Relative Y ields of Stiilv —er Isotope 201 -iTV,Ratio of t.le Aelative Yields from 202 SE nr iched (Bombard;iment 5) and N'ormal Cadm:iuin Targets(Bombardments 1 arid 4) vii

LIST OF FIGURES Figure No. Title Page 1. Cyclotron Layout - Schematic 7 2. Cyclotron Bombardment Chamber - 13 Front View 3. Cyclotron Bombardment Chamber 14 Top View 4. Cyclotron Bombardment Chamber - 15 Side View 5. Faraday Cup and Target Probe 16 Assembly 6. External Focused Beam Bombardment 16 Chamber for Cyclotron 7. Cyclotron Probe Head 19 8. Current Integrator - Schematic 21 9. Power Supply for Current 22 Integrator - Schematic 10. Metal Evaporator 25 11. Types of Heaters for Mletal 25 Evaporator 12. Electrode, Heater and Target 25 Assembly for Iietal Evaporator 13. Evaporator Control Panel - 27 Schematic 14. Evaporator Vacuum System 30 Schematic 15. 4T Proportional Counters 37 16. Cylindrical 4Ir Proportional 37 Counter - inside housing 17*. Nodification of Model 162 for 40 Proportional Counting 18. Integrator-Rlonitor Comparison 47 viii

L,-ure i0o. P itle a te 19. Suppressor Ring Curves 5v 20. Calibration Schematics 56 21. Calibration of Beam Integrator 60 22. Chemical Separation of Sodium from 77 }<ea <L;es ium 23, Cihemical Jeparation of -hosporus S1 from SIulfur 24. Cl-emical separation of Scandium from 03 Titanium 25. Apparatus for laz.ing Thin Plastic 97 26. sou.Intiins; th-e Film on a LT Counting 97 Y'late 27. Ir nim!! t e _ounted Film 97 20. T ransi s n of i3 -iadiat ion 103 22f?. Filonim o sor tion Clrve ofi iDS$ e 107 3o. Self;`o6sorption Curve 0o1 Co 110 31. I lateaiu Curves 113 32. Sample for 4Tl CJounter 115 33., is-cri u..inination Curves of 32S I32 116 116 34. Decay of In 121 35. Coincidence Corr ction 123 36.,ecay of?-3 i32 127 37.?uclide ohart of the >Iagnes ium Region 142 31. discrimination Curves of Sodiunm —- 147 22 39 3Discrimination Curves of dodium- - 148 22 40. Film;' sorption Curve of Sodium 149 ix

.o ur:.'itle r afie 41. Decay of.odium 150 42. 1iuclide Chart of thle Sulfur {egion 159 3 2iiear0f 32162 l43. Deca y of PLhiosphorus-pure 162!41. iiscrimination Ciurves of Phosp-horus 164 t435. JOeca OI' o:iosp'horus-P-32 plus 1'33 165 46. Deca y of 33 Tracer 168 47. t:. uclide Chlart of t!-e 1iitanium Region 170 48*. Jiscriiminaton C urves of,candium — 174 70o c4 4~'9. Discrirninatign Curves of Scandium — 175 100,; Sc4O 50. Film Absorption Curve of Scandium 176 1#:Decay of Scandium 17 c', 52. Calcilated )}ecay of:3c44 47' 4+ 179 53. 3Juclide Chart of t'he Cadmium A:-ejion 1:2 54. Jiscrimination C-Orves of Silver 186 j55. Decay of Silver 187 56. Chem, ical Separation of Silver from 199 Cadmium X

DETEi NT! NjlTI OF (d,alpha) REIACTItON CROSS oECTIONS by Kenneth Lynn Hall ABSTRACT The objectives of this research were to assemble the necessary apparatus and establish procedures for the accurate measurement of (d, () reaction cross sections, and to apply these techniques to determine several cross sections using the 7.8 MIev deuterons from the University of TIichigan cyclotron. The measurement involved the bombardment of thl-in targets, subsequent chemical separation of the product nuclei, and determination of the absolute disintegration rate of the P-ray emitting products. To obtain the irradiations a bombardment chamber was designed and installed as an integral part of the cyclotron vacuum system. The deuteron beam was stopped in a Faraday cup, after traversing the target. The beam current was rneasured by means of a current integrator which was built for this work. A metal evaporator was assembled for use in fabricating thin targets. After bombardment, the targets were dissolved, and the low yield (d, a) reaction products were Themically separated from relatively large amounts of (d, n) and (d, p) radioactive products. ihis was achieved without the addition of xi

inactive carriers in the case of three of the four elements studied. Absolute counting of the separated products was accomplished by application of the techniques of 4rT proportional counting. The experimental (d, a) reaction cross sections for the formation of the isotopes indicated are: 22 + Na22 a 0.094 - 0.004 barn at 7.8 0.1 Mev Na24: 0.151 0.006 barn at 7.8 0.1 Mev p32 r i 0.3 - 0.2 barn at 7.7 0.1 Mev Sc46: ~ = 0.00044 + 0.00033 barn at 7.0 - 0.8 Mev 1.2-hr Agl04: ( = 0.0017 + 0.0002 barn at 7.8 0.1 Mev A =2: 0.00049 0.00002 barn at 7.8 0.1 Mev Ag1 a = 0.0004 0-.00004 barn at 7.8 0.1 Mev The errors quoted are estimated standard deviations. The 46 result for Jc46 quoted above is an average of two deterrmiinati w:iich differed from one another by a factor of four. The hal life of Na24 was evaluated by the method of least squares and found to be 14.93 - 0.04 hours. Separated isotopes were bombarded to assign the 27-minute and the "1.2-hour" periods in silver to Ag104 xii

PART I:` 3SLUT;, (d, alpha)'iiOTICOh i~COSS SSOCTIONS i l TRODL L C T I O( The study of the properties of the nucleus and of the reactions which it may undergo has received an enormous amount of attention during thie past tw-o decades, and it appears that the fi:>'s'.... continue to increase in popularity at an even greater rate than it has in the past. Thus the expenditures of t'he United States dAtomlic Energy Cormmission have quadrupled since 1948, (1) and in seven years the dollar volume to be spent on nuclear reactors is predicted to undergo a tenfold increase. (2) Particle accelerators are fairly common on colleve campuses nowradays (3) so t.at high energy sub-atomic projectiles are available to many vworkers for trie purpose of producing radioactive species for study or application, or for the purpose of investigating tlie details of the reaction betw een the pro jectile and the nucleus. In speaking of nuclear reactions one of the important qutestions to ask is "<Lhat will thne yield of a certain nuclear species be if bombarded under a given set of' conditions?" In the field of reactions between molecules the organic chemist is similarly concerned with the chemical yield, i.e., t[ae percentage of the starting materials which react to give a specific product among thle various by-,rod lcts which are produced in the same reaction mixture. Fursuin-. this analosy a little furter, t1he rate of production of a given chem-ical species is nany times expressed -1

-2in terms of the concentration of the reactants, and a rate constant characteristic of the reaction. A similar situation exists in the reactions between nuclei and high energy bombarding particles, with the important difference te at in this case equilibrium conditions do not exist and the reaction must be discussed in terms of individual collisions. The rate of production of a given nuclear species is expressed in terms of the "concentration" of nuclei and incident particles, and a constant of proportionality characteristic of the reaction (cf. the rate constant for chemical reactions). This constant is known as the cross section and it may be defined as the probability ttiat a given nuclear reaction will occur. There may be several possible courses for the reaction to follow (as in the chemical case), but the cross section osually refers to only one. Jist as in writing cr-iemical equations a nuclear reaction may be written with the reactants on the left and the products on tie righit: Te122 + d =b120 + In words, this means thiat one nucleus of tellurium-122 reacts with a deuteron and a nucleus of a new element, antimony-120, is produced with the release of an alpha particle. A shorter notation is usually preferred, so that the above reaction is written as Te122 (d,c)bl20

-3These symbolizations refer to a single event (cf. chemical equations where each symbol refers to one rmol). The heats of nuclear reactions are trem:endous when compared with ordinary chemical reactions. Thus the "4-t"! for the above reaction is exothermic by 9.0 Mev per nucleus, or 2.1xlO1 cal/mol.! ate constants for chemical reactions are dependent upon the temperature of tlhe reactants. 1 Rememberinn t'hat temperature may be related to the mean kinetic energy of the molecules, a change in the temperature produces a ci-ange in the averace energy -with which the reacting amolecules collide. Thus the rate constants miay be said to depend upon t he,mean collision energy. The nuclear case is analogous in ti-'at the cross section is a function of the energy of the incident particle. A graph exhibiting t:i-s dependence is called an excitation function. There have been many studies of thlese functions carried out during the past two decades. Inr 1944 Clarke and Irvine (4) presented a bibliography of experimental excitation functions produced -uy cyclotron accelerated deluterons. The eleutron Cross Section Advisory Group of the Atomic Erergy Commnission (5) has more recently compiled neutron excitation functions and thermal neutron cross section data. A program is currently undern-ay at Los Alamos under tile direction of' R. F. Taschek (6) to assemble the existing experimental ch-iarged particle cross sections. Many of the excitation functions appearing in the literature have been

-4in terms of relative values of the cross section, as the shape of the curve was the main object of interest. Sometimes an absolute scale has been established by normalizing thle relative curve to one point which is an absolute cross section at a particular energy. Other times each point was determined assolutely. Less attention has been devoted to the careful measure-mrent of absolute cross sections at an accurately known energy, especially for low-yield reactions such as the (d,a), (p,a), (a, Y) etc. Trhis may be due in part to the experimental difficulties attendant in these measurements. The present work is concerned with such absolute determinations for thae (d,o) reaction using 7. 8 yev deuterons of the University of 1lfichigan cyclotron.

STATE7 —3 I~T CF THE-I Fi PROBLEM The objectives of this research were (a) to set up the apparatus and establish procedures by which absolute cross sections may be routinely determined with an overall accuracy of i+ per cent, and (b) to apply these techniques to the reasuremient of the (d,e) cross sections of several nuclei.

DESCriPTiOnL OF APPARATUS Ao Bombardment Chamber The 42 inch University of Michigan cyclotron produces 7.8 M;ev deuterons in high intensity beams. The external beam is as great as 100 microamperesjust beyond the deflector. An additional system was installed within the last two years to bring the deflected beam out to a "'scatterin5 chamber" via a focusing magnet, and finally through an analyzing magnet to a detector. The layout of the cyclotron vacuum system as it now stands is shown schematically in figure 1. It was known th-;at the nuclear chemistry group was to have a bombardment chamber incorporated into the sostem and it fell to the writer to design, test, and install the chamber. It was to be placed somewhere between the two magnets F and X. Simple geometrical considerations demonstrated thtat to take advantage of the focusing magnet the chamlber must be behind the focal plane and as close to it as possible. Slits mnay t`hen be placed at L (the calculated focal plane passes through the center at N) to select various combinations of energy definition and beam intensity. Referring; to rigure 1 the deflected beam emerges from thie cvyclotron vacuum tank A throigh the window box B, passes t'e vertical and horizontal slits D into a 6 inch diameter iron pipe E. A focusing i. agnet F directs the beam past the vertical and horizont al slits J and L towvrard the scat tering chamber F~ on the other side of thle shielding wall. -6

17 CYCLOTRON LAYOUT - SCHEMATIC A CYCLOTRON TANK -' -R B WINDOW BOX P C PROBE PORT 0 D SLITS L E 6 INCH PIPE \ a F FOCUSING MAGNET G 6 INCH CHANNEL H VACUUM PUMP INTAKE 1 4 INCH PIPE J SLITS h- H K 2INCH PIPE G( L SLITS \ 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 X ANALYZING MAGNET Y SLITS Z EXIT CHANNEL Fijs I. Schematic representation of tle layout of the cyclotron vacuum system.

Thle slits L consist of a set of collimators containing rectan7oular holes all 1/4 inch high,but of several horizontal dimensions from 1/8 to 3/4 inch. They were machined from 1/8 inch lead sheet and mounted on a brass collar which slides over the 2 inch pipe K. Nformally the 1/2 inch horizontal aperture was used in this rwork. (This choice was more or less arbitrary, but it was based upon the observation that whenever the 1/8 and 1/4 inch slits were used, thle maximum beam was roughly 0.05ja, whereas as much as 0.5,,a has been observed with the 1/2 inch slits.) The remaining apparatus pivots about the scattering chamber, riding on a circular railroad track. This system may be oriented to zero degrees when the chemistry bombardment chamber Q is used. The Faraday cup T and target R retract when not in' use so that the beam can pass into the analyzing magnet X for thie physics experiments. Vacuum pumps located at H and U are equipped with valves so th at various parts of the pipe system may be let down to air independently. One other control over the beam (in addition to the deflector and magnets) is a variable 10,000 volt D. C. power supply which is applied to vertical deflector plates so the beam may be aimed up or down as it enters the focusing magnet. Design The design of the bombardment chamber with the associated Faraday cup and target arrangement was the subject of much consideration. Its primary purpose was

- 9 to fulfill the requirements dictated by this research, but in addition it was to be made versatile enough to permit other problems to be investigated with a minimum of modification. Thus tlhe angular distribution of recoil nuclei, excitation function studies, and the bombardment of radioactive targets were among the problems for which the chamber was designed. Furthlermore, it had to be integrated with the cyclotron system such that only a few minutes would be required to change from a chemistry bombardment to a physics experiment. With this in mind a design was worked out which meets these demands and works quite satisfactorily in practice. A search of' th[e literature revealed that many diverse arrange-ments of thie target system have been used in previous work. Two ways are apparent of arranging the target with respect to the Faradar cup to measure the number of incident particles. The target may be placed inside of the cup or in direct contact with it. (7 - 13) In the other arrangement the target foil is placed in front of tlhe Faraday cup so teat thle beam passes t.hrough the target and is caught in thae cup. (4, 10, 13 - 20) Panofsky (10) found no difference in results when the two arrangements were tried. In either case some provision must be made to stop secondary electrons from either escaping (if produced inside) or entering the cup (if produced outside). While some workers have ignored this detail (12, 15, 20) most of them have made

use of an electric field to suppress tlhe secondaries. Others hnave used magnetic fields, e.g. the fringe field of the cyclotron, to insure that the electron orbits are spirals which are small compared with the dimensions of the collector. Since the target must be removed from the chamiber rapidly after bombardment and because of othler requirements, it was decided to use the scheme in which the target is placed before the Faraday cup with a suppressor ring between. The Ailachijne drawings of the bombardmrnent chamber are shown in figures 2, 3 and 4, and the p!hotographs of the completed chamber are showrn in figures 5 and 6. TIhe construction was tmade of brass, with all joints silversoldered except as noted below or where O-rings were used. Before installation the entire assembly was vacuum tested, first:,by filling with water under pressure and secondly with a helium leak detector. Ten inch diameter brass pipe used in the construction made the chamber large enough to accomodate apparatus to be used in future work as well as to permit the retraction of the Faraday cup assembly to the side of the chamber when not in use. Both the cup and target are on the end of probes manipulated from the outside. Geometrical considerations show that a cone of scattered particles defined by the center of the scattering chamber as its ape, and the end of the 6 inch brass pipe V, figure 1, as its base, is undisturbed by the protrusion of the Faraday cup when in its retracted position.

-11 The Faraday cup assembly is -mounted on teflon slides whlich en;a-le in grooves in the brass floor of the' chamber for easy positioning oI thle cup. The probe consists of a 3/4 incli brass pipe screwed into a part of the teflon structure. T7his pipe slides in and out of the chamber through an O-ring port. The suppressor ring was cut from 1/4 inch brass plate, and a hole was machined in th!e center to match the inlside diamleter of the Faraday cup. It is mnounted permanently on the front of thie cup and insulated with teflon. Ellectrical leads are bros.ht into tlie chilartber via lkovar seals (Stupakolf Ceramic anr Vrinufacturing Company, Latrobe, Pennsylvania, iNo. 95.1029) whlich were scft-soldered over thie 1/2 inch lholes atop thie chamber. The cup and ring are provided with binding posts, and -10 inch lengths of ZG-62/U cable are attached at one end to tile posts, tile othler end being soldered to thie kovar seals. To protect the outside of thle seals from breakag:e and dirt, removable'brass nhousings are provided. The outer terminals of the kovar seal were soldered to thie pin of ca le connectors mnounted on the brass hlousings. Type UC-560/U (teflon insulated) was used.Vwith the Faraday cup while UG-496/U was used with thie suppressor ring. The entire cup assemble ar-d connecting wires were scrupulously cleaned of dirt, srease, and lint before final assembly, witlh special attention paid to the teflon parts and the electrical contacts.

-12Tne target pro be may be itrtroduced through a side c a:i-ber so as not to disturb thne mLain vacuum. The vacuum lock is constructed of 6 inch briss pipe, and a vacuum valve with a 4 inch opening isolates it from the main vacuum system. A 3/4 irch modified steam valve was softsoldered into a pipe fitting in the side of t1e 6 inch Ti,pe to serve as t'h-e conn-ection for tihe vacuwn pulmp. (For details see th-e section describing tKle metal evaporator. ) A Cenco-Iypervac 4 iechanical pump (Central Scientific Companly, Criicago, Illinois) iras used to evacuatle the sidie cuamber in about 10 -minutes. i 3/8 incll IHoke valve (oriented with thle arrow pointing out) was soft-soldered into a pipe fitting p)rotruding fromr the 6 i 1ch pipe. This provides a meoans of letting air into tie c-la-mnber slowly without ru).pt, ring thle delicate targets. An O-ring port in the end plate of the side ch-ia.mmber m~aintains the vacuuln seal arnd supports th e target probe. The end plate must be removed before thle robe can be inserted or writhidrawn, but. this can be done in a matter of seconds. This design is seen to fulfill the original objectives, since it las given very satisfactory service over t-ie past yrear. The brass probe ead wit"h provision for eighit separate tar'get mounts engages in the Flaraday cup assembly. The eiLglt tar-get positio:is vary from 1 1/2 to - 2 2/3 inches in front of thle cup to accomodate collibrmators, absorbers, targets, etc. The positions were assigned numbers 1, 2,...8

-13FARADAY CUP rRC)NT \/IF\,/ TARET PRDE - FRONT Zr j SL\Jeq VOLOa&'ku SOLDeft_ -7 A 3" com Z7 Lucii-E WjNC)C)\,V DIA. ------ k — VACUUM CHAMBEk Fizowr V)r —w EF-cTiDN A-A als, HCILSS f"DIA, 914OLES`V4600M qATF- (CL,4PPBP\) 7 7 7 7 777,Le,,777777, 3 E71A (Imtusglot4 PoNq t5 \b q WOVE Z4 /./Z Z.7 777= 7-F r/

-15ComRESSIOI\ RiNq TAKGET PP~oE - ToP FiAA u p ~,~o l ~2 HOLES, 01A.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~S~E 6EIC J,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ k, 3'~~~~~i tH~ -VAC- UVM — aATC~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.~ — COM PESS ION R I N +50'0"Rlmq qRP "3 3' - 1 61 I:Q57 O RIqAjPoV 3"- - ~ 2$2~~~~~~~214 /,4,~ S~~~~l'I~ ~~e~e l~g c -~ — -4 x s I *d" 4! I,

— RRqE-r FARDmy Cup - 3iDE FHllot A61- oven ECTIOt4'DE)'F(WS-4 ALI. OV6R is 01 x A A 7.'7, 7 7- 77 Ff /w/.3" 312. 7 N5uLATM Sj" Luc 1,17E WINDOW HOLES DIA, 2"" J-) r (40LC-s Fo 010VAR TEALS A y' 3" FLDoR tAVST Be HORIZONTAL Smr-AcE FITS AWNST r,"fklNq

-~16-~~~~~~~~~~~~~~~~~~~~~~~~~il RIM Fig,, Faaday Gapand Targt Probe ssembly ME-~... INI-1-1,1-11-~~~~;-~ Fig 6 xtenalFocsedBea Bobarmen Chmbe fo Cyloton

for easy referene, position 1 being closest to the Faraday cup. Bo 1iindow Box Probe In the early phase of this work, a numnber of bombardmierits had to be obtained in order to test chemical proceduires, to explore the order of magnitude of (d,Q) yields that could be expected, etc. Beam currents of 50 - 100l a are available at the window box probe,(C of fi:ure I), making it easy to obtain large amounts of activity. The probe for use in this position was made from 3/4 inch brass pipe in which the target was clamiped to a copper ahead. -.uhen the probe was with.drawn from thie cyclotron after being bombarded, it was extremely radioactive - many tar-ets were obtained which m-easured 2 roentgens per hour at a distance of one foot! T2he need thus arose for a new probe design from which the target could be quickly removed. Fast removal would result in lower radiation exposures for personnel since most of tihe ind-uced activities are in the copper probe head.:'loreover, t'he investigation of short-lived activities would be aided bv thle use of a quick opening device. The result of these considerations was a new probe desiigen, the details of w iich are a w.wr in figure 7. The essential featlure is th'.at a single Phillips head screw secures the hinged clamp against the probe head thus hlolding the tartet foil tightly. A spring normally forces the clamp away from the head so that after bombardment the screw is loosened and t'he clamp srirings away releasing the

target, To prevent tile screw from flying out of the clamp, thle threads were macrhinred off for 1/16 inch close to tihe head of the screw. Thie entire operation of releasing the target from the Probe and dropping it into a beaker can be done at a distance of several feet by employing a Phillips head screw driver mounted on the end of a long aluminum rod. Tihis new design hias been in use for about three years, and has g;iven very satisfactory service. Spare clamps, springs, and screws allow one probe to be used many times before it must itself be repaired. C. Beam integrator There are two principle methods for recording the current produced by the incident deuterons. They may be cal ght in a Faraday cup and the current amplified and fed inito a pern recorder. OCt erwise some form of current integrator may be used to record t Le total chiarge collected.:rom t.he point of view of accuracy, Kelley and Segre (1i) claim only _ 0.5 per cent by integrating under the curve of current vs time, whereas + 0.1 per cent has been reported for many of the integrators reported in the literature. ( 3, 10, 13, 21 - 34) Since an integrator patterned after thre one described by'watts (30) had previously been used with the University of MIichigan cyclotron b'y King (35) and was available to us, it was decided tilat this would be the logical starting point. The only mo ifications foreseen erere tile improveiaenrt of its stability and the extension of its range up to ~5fa.

,sCLOTN?RoBE Hb4D u IC:-O?;EN,I. l FL CLAPP —,fA7,4.' AWLMIN,,M &J,, Qe-s SIDE rs tQICK-0 AA FRONT \/I SW TDP vIE\^ *. —42.... Z- - -....J 2 lb,,;,Rl —!. — 1 "'"/. k, U ~<"g1,....". 1 E - SC S.1/eli. P H INGE Tli PJMAIS THteAS4 gio!vGi, w4S SNaO AJ TRoB6 HBA -D pi htL,*.dP. DRILL 1" A. E eQ9Wr K3LH,'~1~~~~~~~S:3 -4 U LL ig. 7. Machine drawing of the cyclotron window box probe head. ~.?~-_cZ~'e~O....' —. —'- 7t 1 7,.-+~..L._~_1_~... Fam.L.z~j~PIQI~E ~ ~ ~ —-----— L-, —-— / IV U Fig/ 7- -/ drw/ of the cor',nd, box probe hea d..

-20According-ly, Trott Electronics Company of Rfoc'h-ester, New York, was contracted to make these clanges. However, the unit as received was far from satisfactory. Jilth no input current it counted at a rate corresponding to several microamperes. It was found tiiat the integrating condensers did not fully discharge after eaca count. Several errors in wiring were discovered. These and other difficulties necessitated extensive modifications before the integrator could be used. Description The circuit as finally adopted is shown schematically in figures $ and 9. Briefly the operating cycle is as follows. The deuterons strike th-e Faraday cup, extract electrons and cause one of the integrating condensers A...D to ch-arge. The grid of V3 thlen becomes positive, and V3 starts to conduct -when the grid potential reaches about 5 volts. This cuts off one half of V4 and results in a rise in thle grid potential of one half of V5, a univilrator circuit.'I is half of V5 starts to cond.uct, cu.ttin —, off the oth-er half. Tliis n.eans thiat tle grids of V1 and V2 rise and thljese tubes corndIct, disclargino the integratinf condenser. A positive pulse is also received on th'e grid of V6 w!-ich drives the register azrd records a count. V6 also actuates a relay (Potter and Blrurnfield,. ri~ncetonl,I aiana, >' lp'e LT!5L3, 10,000 o'un. ) wilQse closing time is -3 milliseconds and opening time is less than

CURRENT INTEGRATOR - SCHEMATIC.RED - +300 SW3 R29 R C R3 R25 270K 51K( 51 K ~L i 750 R27 /LO ~~~~~~~~~~.05A47 SW2a b I!26 M K 75K VUJ I P8B SM5LS V4 4 0 C5 U ow rr 5 - > IOK RELAY R4.j (1 R900,0K P24 0 7 500K 50 oo o_2AX7 6 - 60 9700K -R20 50K TRAFFIC R 9,7002K' V6 COUNTER CoD 7 2D21 OUTPUT E A 5803 "IR21 cu ~~ R~7 lW OM r Dsw IKW 3 -8 0 ~_R 2 R8 4M A 2 C R292 20K RI R23 R~29 10 RIO-RI9 IM220K 500 DUAL 500 50K WHI28 1.25.02 c\-.... GLASS MIKE 3.33 12WH M ~~~8200 -FCASS R32 VSC6 V 170 K ~ I 5 Opi. 596 6~~~~~~~~~~~~~~~~~~~ -- -'7 I'~~i~2' ~ 23 8 220K Fig,, 8I0 C r R38 inte.rator R36IM R37,22K ORANGE -150 I~~~~~

POWER SUPPLY FOR CURRENT INTEGRATOR- SCHEMATIC TI Fl IA SW4V9 (SB) I ~l/()o II 6X4 6AQ5 ~(SE) +300 v 7 1 I I5, IR IRVIII 3 m Q::)~~~~~~~~~~~~~~~~~~3m v 60 R41 --- C I0 R 45 115 v )60-Nrr —.,400v 150 K c), IR48 C 1022K-2W 6BH6 5 R 42 40.Lf 6 R 46 (300v) V8-' 20 K R 50 V8 [7 II I I'M 5vrr1, 5.O 5651 R44 R47 - Ve 1 I00 K < 50K K' "'~ ~~~~~~(CO L) F'o~ - 4 - _ I PI_1_ Vg__ ____ ____ __-'v,~~~~~~~~~~~~~~~~~~~~~~~~,. ~~~~~3 ~ V12 4. CHASSIS OA2 C8 60/uf C9 60Ff R 39 R 40 Vs15 K I470. - 150 v 6X4 I - Y ------- 2ma Fig, 9. Schematic diagram of the power supply for the current integrator.

-23 - 1 millisecond. The condenser is thus completely disc -ar ed. The counts;xay7 be recorded either on the I.ercury reg'-ister, or on a traffic counter (Streeter-amet Company, Chicago, Illinois, Model SC1-5) which stamps out tie number of counts at equal time intervals. Tihe 5: 03 electrometer t ubes were sealed in a light tight box along with a small bag of silica gel. The integrating condensers are similarly sealed and desiccated. All thle exposed input wires were cleaned and sprayed with KIrylon and the cable connectors kept scrupulously clean and dry. These precautions were found to be absolutely essential in order to reduce the drift rate (counting rate with zero input current) to a tolerable level. Now the drift rate corresponds to -+0.001/a on range B and — 0.001-/a on range D. The Faraday cup is connected to the integrator by about 67 fee t of RG —8/U cable usinig teflon insulated conrectors, U-59B,/U. The second input connector is for t ane calibration equipment. In order to maintain thle 3+ constant and at 300 volts it was necessary to -use a Sola constant voltage transformer and a Variac between it and the integrator. Even with these precautions 3+ drops several volts whenever the register is activated. It was noted later in the work that if -tie traffic counter and the initegrator are both plugged into tihe same line circuit, every time the traffic counter prints a number, a certain amount of extraneous charge appears on tihe integrating condenser.

-24OQeration It was found that it is necessary for the integrator to be warmed up at least two hours before reproducible results can be obtained. The B+ is then adjusted to 300 volts by means of R46 of figure 9. Referring to figure i the range switch SbTJ1 is set to E (shorting out the grid of V3), and the meter switch STW2 set to V3. The bias on V3 is adjusted by means of R2 so that the meter reads 8r/a. Next SJ1 is switched to D, and a small current is fed into the instrument. If the meter does not indicate 22 a when the register trips, R2i, R20, and/or R10-19 are adjusted until it does. Finally 3;W2 is changed to V4, and the meter is made to read fromn zero to full scale to indicate fractions of a count. This is done by varying R6 and R9a, The performance of this instrument is best discussed in terms of its accuracy and stability, and this will be reserved for the Experimental Procedures section on the calibration of the beam integrator. The calibration curves presented there are not linear except at very low counting rates, and this together with the difficulties discussed above led to th:e conclusion triat an entirely new integrator should be constructed. The design of Higinbotham and Rankowitz (26) was chosen, and tihe parts ordered, but there wasn't time'to complete the new unit for ase in this work. D., rdetai Evaporator A hig: vacuum metal evaporator was constructed for use in this work. It operates at pressures from 10-3 to 106mm

> W~~~~~~~iVU14=5LL@Lre~~~~~~~~~~~~~~~~~~~~~~v::~~:: utH.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ P~ Figi! Types of Heaters aor Aetal EvaapoYator A~ Vycor crcible B ~lund rn cruecibe C: Tungsten wire basket D: Thugsten vzire hetie/ a Fig 1O Metal Evaporator, Fi2 Electrode, Heater~ and Targe+ Assemaby for Aectal Evaporator

-26of mrercury and is powered by a transformer system capable of delivering about 250 watts to the heating element. The completed unit is pictured in figures 10, 11 and 12. Thin even films of various target materials may be prepared with this apparatus starting with metal chips, shavings, powder, or metallic compounds such as oxides, etc.'The films so produced are very uniform in their thickness, as indicated by the data presented' in the Experimental Procedures section on target preparation. Another use made of the evaporator is to evaporate gold onto thin Zapon films for absolute beta counting with the 4Ir counter. Electrical Controls Thae vacuum evaporator used in this work was patterned afLter one in use at the University of' California Radiation Laboratory. Dr. E. H. Fleming of that laboratory kindly furnished ts witri the drawings and specifications. The control panel was built in 1951 by J,. A. Cassatt and was only slightly modified for its present use. The schematic diagram of the panel as it now stands is shown in figure 13. The heater output is capable of delivering up to -13 volts w ien the Powerstat is wired for 135 volts output..! ZUUl't U" A thermocouple vacuum gauge, RCA type 1946, was sealed directly to the base plate of the evaporator by means of Apiezon J wax. (All attempts to make a vacuum seal through an O-ring connection failed. Presumably this was because

EVAPORATOR CONTROL PANEL- SCHEMATIC A.C OUTPUTOU 4 T2 VO$$URRENT f~ () ~2 H EATE R ~ —-----— 9OUTPUT 8A TI MI 0 A DIFFUSION PUMP HEATER 135 W CENTURY THERMOCOUPLE FORE PUMP OUTPUT VOLTAGE MOTOR 1/4 HP WHT BLK 4.4 AMP THERMOCOUPLE M4 HEATER CUFRENT BRN 3A ~il M3 T3 30ft THERMOCOUPLE TI POWERSTAT TYPE 116U MI SIMPSON MODEL 57 0-25 VOLTS AC VACUUM GAGE O-135V 75A M2 TRIPLET MODEL 331-S 0-25 AMPS AC RCA 1946 T2 GE TRANSFORMER C MODEL 9TM805AI Ii.5V 1.00 KVA M3 SIMPSON MODEL 57 0-100 MILLIAMPS DC T3 STANCOR P6134 6.3V 1.2A M4 MARION SERIES 55 0-15 DC MILLIVOLTS Fig. 13. Schematic wiring diagram of the metal evaporator control panel.

the base plate was mrachined from a cast in-,g so thiat the recessed surface originallb y provided in the base plate focr th]e gauge could not be polished sufficiently.) l ~icLeod gauge was used for thle calibration. 7Toth galiges were conu-ected to t.' vacuum system through holes provided in the base plate. T1e press-re was varied by isolatiu-. t'he vacuum systema at various stafes of evacuation. The pressure was tlhe same in the tw-o gauges and remained constant over a period of time muiich longer than necessary to take readinrs, as proved by repeated readin;s,. Calibration cutrves were tar:en for vario-us values of' the heater current in the 1946 tube. Sabsequent calibrations showed that the carve shifts as the tube ages, but not enough to affect the dayr-to-day operation of the evaporator. The thermocouple gauge becomes quite insensitive at pressures below 10-4 but since most evaporations do not require a vacuum better tuan this value, tule gauge serves the purtpose. Vacuulm si:stefn A sceiemlatic representation of t'he sysvtem is sown in figure 14. T'he vacluum s-Tstemn of t'le ori-inal unit was rebu.ilt to include a liq-uid nitrogen trap and a s-ystem of valves to'ypass th.Le iiffusion p imp. T he original CencoHyrvac forepump was replaced with the faster Cenco-Hypervac 4, wh.icih reduced the pump-down time for rough vacuum from 30 to 5.minutes. Thre inclusion of tihe cold trap was found to be essential for the rapid attainment of pressures in the range of 10-6nm, especially on humid sum er days. The

-29valves msake it unnecessary to cool tie diff sion pump between evaporationYos, as the pu1mping sys'em can be closed off w;E-ile air is being admitted to tCie main vac um chamber aLnd the diffusion pumnp alone can be isolated while "roughing out" thle -main chamb er. These modifications permit greater efficiency and more rapid operation - it takes only 10 to 15 mklinutes to obtain the va-cu —,m needed to proceed with an evapor ation' The pipe syst em consisted of ordiinary 1 inch brass plu..bin, fittings`joined toget er with silver solder. A short 1 inchl brass nipple was soft-soldered to tihe intake of trie diffusion pump. The threads of this nipple were tinned with soft solder, then coated with red Glyptal enamel To. 1201, (General Electric Company) and finally screwed into valve 3. A second coat of Glyptal assures a vacuume tighit conn- ection. The entire pipe system (including th e dil'ffusion pu.mp and cold trap) is attac.hed to tile base plate of the main vacuaum chamber iy means of a brass flange and an O-ring sealThie diffusion pump used, a VIFvI-10 (Consolidated Vacuum Corporation, Rochester, New York), was thoroughly cleaned out with C. P. ether and filled with Octoil. (Each time the bottom plug is removed the copper gasket must be replaced to assure a vacuum tight seal.) The fore pump was filled with light grade Cenco-Hyvac oil. As thne pump is broken in through usage, their heavy grade

EVAPORATOR VACUUM SYSTEM-SCHEMATIC VALVE VALVE FORE - I DIFFUSION LIQUID VACUUM PUMP UMP CHAMBER VALVE 2 (BY PASS) o VALVE 4 (AIR VENT) VALVE 1,2 WEBSTER SYLPHON VALVES VALVE 3 VACUUM GATE VALVE VALVE 4 HOKE NEEDLE VALVE DIFFUSION PUMP- VMF 10 FORE PUMP - CENCO'HYPERVAC 4 Fig. 14. Schematic represertation of the mnetal evapiorator vacuum system.

-31oil is necessary to prevent the vacuum from becoming soft as tlhe oil heats up. The valves used in constriucting thle bypass system are of three kinds. Valve 3, directly above the diffusion pump, is a commercial high vacuum valve. The air vent, valve 4, is simply a 1/4 inch Hoke needle valve properly oriented. The other two valves and the one used on the bombardrnent chamber (see page 12)deserve special attention. Several ways of modifying ordinary commercial valves for use in vacuum systems have been suggested. (36 - 35) However, ordinary steam radiator valves containing sylphon bellows can be modified rm-uch easier for conversion to serviceable high vacuum valves. (39) First, the fiber gasket was replaced with one of soft riubber (e.g. neoprene), Niext, the leaka;ge path around the screw in the gasket holder was closed off by soft-soldering the two parts together. The valve seat and the top of the body (w'here the bonnet screws down) were polished with emery clothe An O-ring was inserted between the top of the body and the sylphon flange. A piece of soft solder in the form of a ring was useful to act as a spacer in keeping the O-ring centered. For assembly the O-ring (coated with Apiezon T if desired) was placed in position on the body and the sylphon inserted withX tile gasket in place. Next the valve was pumped down so that atmospheric pressure held the pieces in place. The shaft was removed from the bonnet and screwed into the

-32sylphon. Finally the bonnet was lowered into position with the solder spacer and screwed down tightly to effect a vacuum seal around the O-ring. Using this procedure there was no twisting of theO-ring which can cause leaks. The cold trap consisted of a 2 inch brass tube with a 3/4 inch brass pipe extending concentrically almost to the bottom. A flange with an O-ring groove was provided so that the trap can be cleaned. The cold trap was thermally insulated from the rest of the metal system with the rubber O-ring and teflon washers on the flange bolts. -n inordinately large amount of time was spent in leakhunting and in making vacuum-tight connections. Large leaks were found by filling the system with water under a pressure of several pounds, or by applying air pressure inside and soap solution outside. i spare vacuum gauge Ias found to be very handy when it came to hunting smaller leaks. The pressure was then checked in one section at a time, making use of' the valves eith'er assembled or with the bonnet and stem replaced'by rubber stoppers. A hydrogen leak detector was also used with some success, but the.helium leak detector is far superior in detecting minute leaks. Both wax seals and screw threads coated with Glyptal'have been used to effect vacuum-tight connections, but the only fool-proof method is to employ O-rings in properly constracted rosoves. Bolts were designed with O-ring grooves in the under side of their heads to seal off the

-33spare holes in thie base plate. Great care was exercised in making the grooves. Their dimensions were made to conform closely to those recommended by the manufacturer (Crane Packing Company, Bulletin 1',o. P-308) except that the groove width "Dtt was made equal to the O-ring crosssectional diameter. ill surfaces corntacting the C-rinr-s were honed or polished free from radial scratches and tool marks. It was found thlat the metal used must not be porous, i.e., castings carl not generally be used. Tee original bel jar provided was made of steel with a window in thle top. The welded joints made it impossible to work with, as leaks developed when a modification was at:;- empted. A Pyrex bell jar 10 1/~4 inches in diameter and 14 inches high, (Consolidated Vacuum Corporation, Aociaester, iMew York, ino. 572r) was found to be'very satisfactory. An L-shaped rubber gasket fits around the rim of the bell jar to make a vacuum tigraht seal with the base p late. IJo grease was necessary. An effective method of preventing the bell jar from becominv, coated with an opaque deposit of evaporated material is to apply a thin layer of Dow Corning High Vacuum Grease on the inner surface. Heaters and target assembly Several types of' heaters described in the literature (4.0 - 41) were tried. Figure 11 illistrates four different designs: a Vycor crucible with a tungsten heater coiled around it, a crCcible made from alundum (impure alundum was

-34used because the pure variety crumbled and would not stick to the conical wire frame), a tungsten wire basket, and a tungsten wire helix. These designs met with varying degrees of success, but t!ie most generally satisfactory heater for this work consists of a strip of 0.005 inch tungsten foil with thle sides bent up to form a boat. A large amount of material may be placed in the boat with little danger of it falling out, and a white heat is easy to obtain. However, each material to be evaporated presents its own problems and the solution is an individual matter. The heaters were clamped between steel plates mounted on steatite stand-off insulators, as illustrated in figure 12. The current was conducted into the vacuum chamber by commercial electrode assemblies (Consolidated Vacuum Corporation, NIo. 7299). Figure 12 also shows the substrate upon which the metal vapor condenses supported above the heater from a specially designed ring stand. In the case Of plastic substrates a 1/2 inch thick aluminum block is placed behind an&i in contact with it to help dissipate the heat. The choice of substrate is also an individual consideration, and will be treated along with t-he choice of heater in the Experimental Procedures section on target preparation. Operation Thle detailed instructions for use with this evaporator are given in Appendix I.

-35o 4nr P'roportional Counter?i en this research was undertaken there were no 4rr counters availaole comr-iercially. The choice of this type of counter for absolute disintegration rate determination will be explained later in thre Experirnental Procedures section on absolute beta counting. Th;ere are four different designs of 4Tr chambers described in th-e literature. European workers employed a cylinder with the sample placed on a support dividing th- e cy linder into two halves along its axis. (42 - 46) Two wires stretch~ed parallel to the cylinder axis above and below the sample support served as anodes. A?"pillbox" arrangemenrt has been used at the National Bureau of Standards and at Harwell. (A4o - 47) A rectangular chamber was divided into two ialves oy a sample plate, with the anode wires stretched above and below the plate and parallel to it. Borkowski (48) is responsible for the third design which consists of two cylindrical vhalves clamping the sample plate between them. (47 - 458) The anode wires are stretched along a diameter of each half. Finally two hemispheres placed tooetzier with the sample plate in between constitute the fourth type. (47, 49-52) T.h-e center wires are two small loops suspended above and below the sample. The two anodes were usually connected together in parallel and fed into a proportional or geiger counting unit. Some workers hlave made stl)dies of the coincidences between the two halves. (42 - l44) The counting gas flows continuously or a pump may be used to evacuate the chambver after which thie gas is intro

auced at the desired pressure. In the proportional counters both pure methane and a mixture of 90$ argon and 10/ methane have been used. (Trhe latter, sometimes called P10 gas, is reported to give a plateau at a lower voltage than pure methane.) (48, 53) The pressure is usually atmospheric, alth!ough l Borkowski (48) has conducted experiments in which the pressure was varied from 10 to 76 cm of mercury without noticeaJle (+ 0.271) chiange in counting rate on the plateau. Description A cylindrical (Borkowski type) and a spherical 4TT chamber were constructed for use in this work. Thley are pictured in figures 15 and 16. Thle cylindrical type was patterned after tIat used oy Borkowski (48) who very kindly sent us the machine Crawings. The two hemi-cylinders were 2 3/4 inches in diameter and 1 5/16 inches deep, machined from aluminum. These two halves are placed on top of one another, clamping a thin aluminum sample plate between them. A gas tight seal is provided Siy an 0-ring between the hemi-cylinders. The sample to be counted is mounted on a plastic film covering the hole in the center of the plate. The plate then rests just inside the O-ring. The electrodes are stainless steel wires 0.002 inches in diameter (obtainable from Sigmund Cohn, New York). They are stretched horizontally across each hal.f of the cylinder at a distance of 11/16 inches from the central plate. Since it is frequently necessary to change these wires the details

Fig 1" 4 pi Proportional Gounters left spherical; right cylind rica l. Fi -'~

of this operation are described in Appendix II. The other chamber, patterned after the ones in use at the 1iational Bureau of Standards (47, 49 - 50) consists of two stainless steel hemispheres 2 7/16 inches in diameter. The sample plate is clanmped between theme A fgas-tight seal is effected by two O-rings compressed agairlst opposite sides of the sample plate. (Here the O-rings actually form a part of thie inside spherical surface.) The anodes are made from 0.001 diameter wire, either stainless steel or platinum. They are formed into loops 3/16 inch in diameter and are suspe ided above and below the sample from teflon insulators 3/4 inch from the sample plate. 3oth loops and h;lemispheres are comimlercially available ( Nuclear Measurements Corporation, Indianapolis, Indiana). The loops can also be made in the laboratory, and the following directions for making thlem are given in Appendix II. Thie sample plates for these chambers are made from 0.010 inch aluminum sheet. Three 1/E inch holes spaced 120 degrees apart and 1/4 inch from the edge permit the flow of methlane from one half of the chamber to the other. }Plastic films can be mounted directly over a 1/2 inch central hole upon which is deposited the active source. Another arrangement is to lay across this hole a smaller disk (1 inch diameter) which itself has a central hole punched in it and across which the sample-bearing film can be secured. The latter is illustrated with the cylindrical counter in figure 15. Usually the active

-39deposit is covered with another layer of film to prevent contamination of the chambers. The counting ras used is 99 mool per cent methane (Phillips Pletroleum Cornany, Uiartlesville, Oklahoma) and the flod,~ rate is adjusted b}y moeans of a bubbler filled with Ior Corning- 702 fluid. The ch'amber is flushed for ab)out 10 Jiivutes at a rate of -10 bubbles per second after introduction of each samle. Th'.is length of time -:as determilned oy obsZerving tele countin rate of Sr-Y90 as a function of the le::-; th of flusIhing time..fter 3 1m-inutes Lhe satmn.-le counted wzithin the statistical error of the nornmal rate. JCare must be e:wercised w'rith the.?lapon films as they will ie rupt red e y d too irco-t =flushing raIte. Pho flow r-ite is reduced to -3 bubbles per second for the duration of the count. The cournters were osorat,:d in thle propiortional re'ion so teiait vexry high counting; rLtes ma. be a<;-asuread;si: eout the applica tion oi a L!.rge de d time correction. This is due to thle fact that t;e deAd tiimes of proportional counters are intrinsic:;llv low in co?.parison with thlose of Geiger counters: of the order, of 5 sec as compared with -200 Psec for Geiger counters. it [lucle'ar-Cliicago?:[odel 162 sc ling unit 1modified for Jrocortional counting] (54) was used in conjunction with the c —lid —rical chainl er. The amplifier staC-e was ~odii-ied by the aA-ition of two 1N3, diodes and a 1 megohm, 1/2 watt resistor in the -rid circuit of t-he

-40MODIFICATION OF MODEL 162 FOR PROPORTIONAL COUNTING VT-2 VT-3 6AH6 R 16 6AH6 C8 _ AC17 RIO = _ R 17 C7 C9 R 1 13'10 R14 1N38 MEG' + 1N38 Fig.,I. Ch ul. t a oie ran o crtion o b0 e l l1ar winchrhasent Co ipora1. i. ~ to e 2;capd in urnit circ-.it whi -h ia s Deep mo.id "i Hi P'or prop]"t ioriai countioi.<j

tuird 6AH6 tube (VT-3, per thre circuit diagrn of the Model l 12 scaler given in the accompanying instruction C:anual). The modified part of the circuit is shown in the figure 17 which is a portion of the iodel 162 circuit diagrarm. This alteration is to prevent the amplifier from overloading and double pulsing with very large pulses and hli,-gh counting rates. The high voltage was supplied by a Radiation Instruiment Development Laboratory Ilodel ~O dual high voltage supply. The amplifier and scaler for use with the spherical chatmoer was built by the Trott Electronics Company. The pulse alaplifier' is trhe same as designed by itexroth. (55) The scaler is a modification of the one dLescrilbed by Kenip.(56) (.Thle end window proportional counters used at Los'Alamos utilize the same amplifierscaler combination as reported by Rexroth and Kemp.) Th!e 4n chambers are placed in a copper box for electrical shielding, as shown in figure 16. At first heavy bare copper wires conrnected the cl~amber to the UG-/+96/U cable connector in the side of thne copper boxo Later f'iexiule rubber covered wire (,000 volt insulation) was found to stand up better. These wires were attached to the cable connector 3by clip leads so tlhat the chamber may be easily disconnected for changing samples, cleaning, or for using the top and bottom halves indepenuently. A short length of AG-11/U cable runs from the box to the proportional amplifier. This copper box arrangement is -used at the National Bureau of Standards for the purpose

-42of lowering the input capacitance and thlus reducing the resolving time anJ tne attenuation of th'e pulses entering the armpllifier. (57) Thle copper box is surrounded bly two inches of lead for sh-ieldini; purposes. Thie background is then ~100 counts/min on the cylidrical counter and 40 counts,/nin on the spherical counter. Performance A good deal of trouble was experienced in getting the counters to operate properly. This was especially true of thle spherical chan-ber. The difficulty was that after a pueriod of operation the voltage plateau would become almost nonexistent, except at very hij.h methane flow rates. The cylindrical chamrber seemed to consistently give better lookinog plateaus, and for this reason it was used almost exclusively in tite crons section work. But even it gave spurious counts which would sometimes jam the scaler. The several precautions that were found necessary for proper performance are treated below. Fjrst of all the methane must be flowving thiough the chambers at all times whien t.he high voltage is on. Otherwise a continuous discharge may occur and a brown deposit may accumulate inside t'e chamber. (58 - 59) If this happens, the anode wires and, jr chambers must be cleaned as described in Nfluclear Measurements Corporation Instruction Mlanual for their itodel PC-1 counter. Further details are given in Appendix II.

-43 - Special attention must be paid to any dust, lint, etc. which mrray lodge on the high voltage connections, for this will surely cause a discharge to occur. The gas connections are another po-sible source of grievance. It was found that the chamber must really be gas-tight in order for proper counting to result. Thus the Tygon (type 36(03) connecting tubing was wired tightly to the chamber connections. Apiezon'.J wax was used to insure that the clhamber connections were sealed. The chamber was frequently tested for bad leaks by watching the 1/2 inch head of oil in the bubbler wrlen the system is closed. Lastly a phenomena occasionally takes place that so far defies an explanation. When certain samples are counted for more than several minutes the counter goes into continr ous discharge, while othier samples, the background, etc., all count normally. Furthermore the questionable sample apyiears to count all right when the bottom half of the chiamber is used alone, but it jjams the counter if the top half is used alonel Perhaps it has something to do with tiie way in which the sample is prepared, because little trouble has been experienced ever since the procedure has been adopted of cleaning the sample plates with acetone and drying the final sample under a heat lamp and hot air current.

EXt-9~M2t ZIZ'iJf~TA'L }-;ROC~Thf, RS A. Beam Determination Suppression of Secondary Electrons As pointed out in the section describing the bombardment chamber, page 9, secondary electrons, or delta rays, produced when deuterons strike the target, collimator, etc., must not be allowed to enter the Faraday cup if produced outside of it. Otherwise the positive current collected would be smaller tl-han that produced by the deuteron beam. Conversely, electrons produced within the cup must not be allowed to escape or too high a positive current would be indicated. MiIost workers have taken secondary electron emission into consideration using 200 - 300 volts to suppress thAem in lieu of a magnetic field, although ranofsky (10) used C:,000 volts. Since the geometric arrangement was different in each case reported in the literature it was deemed necessary to experimentally determine the proper potential to use. To take into account the fluctuation in the deuteron beam, a scattering target of Mylar, a DuPont polyester film, usually 0.00025 inch (1/4 mil) thick, was in the scattering chamnber at NJ of figure 1. The scattered particles which passed through both monitor counters 0, were recorded as events by a coincidence analyzer whose output was fed into a scaler, Bach (60) has given a complete description of the monitors. The main beam passed through the Mylar target N, -44

-45traversed the target a, and was stopped in the Faraday cup T. The potential of the suppressor ring S was varied and the amount of charge collected in the cup was measured in terms of t'le ratio BM of the number of counts recorded by the integrator per 64 monitor counts. The integrator was set on range B (see figure 3) for those experiments. Tre procedure was to measure B,(0) when no potential is applied to the suppressor ring, i.e,, with the ring shorted to ground. Secondly Bh(v) was measured when v volts were applied. Thirdly B13(0) was again determined, etc., alternating between readings witha no potential and with v volts. I. Preliminary 5xperiments,' great deal of difficulty was experienced in t'hat B?(O) tended to drift diring the course of an experiment. This could have been the result of part of the beam striking the monitor target N tklen failing to hit the Faraday cup target R, or vice versa. Mtiultiple scattering from the monitor was ruled out on the basis of a calculation following -iossi and Greisen. (61) A quick check with an oscilloscope showed t.;at the discrimination settings on the proportional amplifiers of the monitors were not set oni a ist ej >Ltt of the i —,crin'irtio, e. (in iole it Iiea run, experimental discrimination curves were taken to determine how to adjust the amplifier.) The principle source of difficulty lay in the geometrical arrangement. Thle lead slits L of figure 1

in front of the maonitor foil must be small compared to the area of the foil. Decreasing the vertical and horizontal slits J in front of the scattering chamber reduced the amount of extraneous scattered beam. That the main beam was actually going through target R was demonstrated by the observation of the beam pattern. It appeared as a yellow discoloration -1 inch wide and -1/2 inch high in the center of the many 1 Mylar targets t!iat were bombarded daring the course of this work. (Sirnce the slits L were most often 1/2 inch wide and 1/4 inc!h high, some idea may be gained of tahe divergence of the beam from these figures.) Bith the "background"' (counting rate of the monitor with tile Ivilar target N removed from the beam path) thus reduced to less t:han 50 monitor counts per integrator count (range B) some improvement was observed in the constancy of 3 (0). Finally the correct alignment of the monitor counters, i.e. "aiming"? t em exactly at the center of the monitor foil, provided the means of redc.cing thie drift over a period of roughly 4 hours to -~ per cent. This is illustrated by data from the final run in figure 18 which shows the fluctuations in' (0) as a function of the total number of counts recorded by the integrator. The latter increases with time, but is not proportional to it because of the variations of the cyclotron beam. The drift that is still present -.~ay be due to ti.e deterioration of the Mylar target (it becomes brittle and turns yellow under bombardment),

INTEGRATOR-MONITOR COMPARISON.030 d.028.026 0.024.022 I 0 I 34600 800 35000 200 400 600 800 36000 INTEGRATOR COUNTS ij I Comparison of the current inteprator and t montor The line inica-t4eo tT.e amount of drift (< per cent) over a 4-hour period. Thne ordinat is defined onn pao 43.

or some electronic instability of either the monitor or integrator. Altho igh range B of the integrator was not calibrated with the same care that was taken with ranges C and D (see next section), early experiments indicated thlat its inearit was as good as the ot her two higher ranges if the counting rate remains below -10 counts/min. The fluctuations of the individual values of BM(O) are hi*:her t'lan expected from an accumulation of the errors of counting statistics, timing, etc. However, by averaging oLM,(O) taken before and after By.(V) was taken, the ratio 3B(V)j /'I(1 is fairly reproducible. This is the function w:Iich is of interest in determining the effect of suppressing potential. 2. Secondar Electrons Emitted from the u In an early experiment with no target foil R in front of the Faraday cup, thie potential of the ring was set at -295 volts, and the readings taken. Next, the ring was grounded and the cup itself was biased at -200 volts by means of the R49 control on the integrator (see figure 9). In these experiments there was no detectable change in the amount of charge collected by the cup. That is, Bj(O) equaled B (V) to within the ~+ 2 per cent deviations observed for these number. B,(0) itself drifted by about 4 per cent during the two hour course of this experiment. The result is evidence that the cup was well designed insofar as its geometry is concerned —less than 2 per cent

of tlie secondary electrons can escape tiat are produced wihen the beam is stopped in the cup. 3. Secondary Electron Emitted from the Target The target R was a 0.9 rail aluminum foil placed in one of the eight positions of the probe Ihead in front of tile Faraday cup (see page 12 ). A marked effect was observed as shown in figure 19. This curve and the one of figure I were thle results of the final run lasting 14 hours. (Thale many earlier runs were much less conclusive because the techniques were still in -the process of being worked out.) The ordinate is the ratio BM(V)/BM(O) referred to above, and the abscissa is the suppressor potential, The errors are the mean deviation of at least four determinations. The figure gives the results for two different target positions. When the aluminum foil was close to the Faraday cup in position 3, the positive charge collected for a given number of monitor counts was -27 per cent higcher when a suppressing potential of more than -200 volts than when no potential was applied. When the target was moved further away making the solid angle subtended by the cup smaller, the observed effect was also smaller. Thus in position 7 the discrepency was -16 per cent. Several corrections might have been applied to obtain a more nearly correct value of BM, ut they were considered igligible. Thus the integrator drift rate (see section describing the beam integrator page 23 ) amounts to about

-5o0SUPPRESSOR RING CURVES TARGET POSITION 3 1.3 1.2 e1.0 TARGET POSITION 7 1.2 1.0 0 400 800 1200 1600 2000 SUPPRESSOR POTENTIAL (VOLTS) ligj 19. Thre eijfect of appljyiing a potential to suppress secondary electrons prodi'ced by ithe deuteron beai. The ordinate is defined oni flae 40.

+0.03 counts/min on range 3. This nearly constant correction, when compared to typical counting rates of 5 counts/min with the cyclotron beam on, is of little importance. Then there is the monitor background, mentioned above, due to extraneous scattered radiation, but this is only about 0.3 per cent of the counting rate registered with the monitor target in place. Finally there is a very small background in the monitor due to electronic noise, cosmic rays, etc. This was found negligible at about 0.01 per cent of the counting rates observed during a run. The curves in figure 19 indicate that in practice, as long as the potential on the suppressor ring is maintained below -200 volts, the effect of secondary electrons is negligible. Theoretically a slight positive slope might.)e expected out to 8(500 volts, which is the maximum -,ossible energBy of a secondary electron ejected by a 7.8 Piev deuteron. Bethe and Ashkin (62) give the formula for calculating N, the number of delta rays emitted per centimeter of path of the incident heavy particle. Applied to 7.8 Mev deuterons in 0.9 mil aluminum, the number of delta rays with energies between ";d and d-i ( kev) is dN 0.020 dl (I) where v is th1e velocity of the incident particle and c is the velocity of light. Note that the Ni will be inversely proportional to the energy of the delta rays. The energy a

-52is a function of the angle of ejection of the electron, Q: 2 2 W d 2mv2cos 2 (2) where m is the mass of thle incident particle. Taking the target in position 7 as an example, and considering the cone defined by the center of the target and the mouth of thle Faraday cup, the maximum value of Q is -25~ From equation (2) this corresponds to electrons of 7 kev energy. The total number emitted from 7 to 8.5 kev in 0.9 mil aluminum is found to be 1.3 J electrons per deuteron, by integrating and solving equation (1)'. Cornsidering the range of these electrons (33) only those produced in the last 6xlO5Scm will escape from thle aluminum. This amounts to 0.038 electrons per deuteron, which represents delta rays from zero on up to 8.5 kev energy emerging from the target. Scattering in the aluminum will further reduce the number emitted at angles from 0 - 25~. It was therefore assumed teat the plateau reached at -200 volts in figure 19 was real, and that energetic electrons passing the suppressor ring were too small in number to be observed. Primary Calibration of' the Beam Integrator The integrator was calibrated by Leeding in a known current and observing the counting rate. The current was supplied by a power supply in series with a high resistance. The connecting cable and Faraday cup were included in the calibration, as they are in parallel with the integrating

-53condensers. (T-e R-/ U cable, rated at 29.5 pyf per foot, contributes a capacitance of -2xlO-9/ f. ) These are small compared with 0.25 and 1.01tf of ranges C and D. The description of the integrator has been given in the apparatus section, pa;e 18, 1 leferring to figure S, tie condensers A.....D dischar~ze wi-en their potential V rises to -5 volts. The fluctuation in the condenser p-otential prodices variations in the calibration current. If R is tIhe total series resistance across tle condenser of capacitance C, t V = E(l-e 1C), (3) whlere E is the value of the power s:upply voltage, and t the time mneasured from end of the discharge. The value of the current i is E. (4)It may be seen from equation (4) th-at the larger E is, the less effect the variations in V will hlave on the average current. For this reason the power supply was operated at 500 volts or more. Even so, noticeaule fluctuations were observed, and the integrator was switched to range E (input shorted to ground) for the current measurement. The magnitude of the discrepency between the range E (maximum) value and the average value of the current on, say, range U can be estimated as follows. Consider the instantaneous current in the I-C circuit: substituting

equation (3) into (4) yields t i e. (5) The initial current i0 is simply io% E/R, and the average value of the current is T T -C ioRC T 1,, (1-e )- (1-e (6) where T is the tie between couts Substituting typical where T is the time between counts. Substitutin: typical experimental numbers in the equation (6) shlows that the T differs from i0 by 0 - 0.5 per cent. This constitutes the justification for using the maximum value of the current in plotting the calibration curves. The exact value of the current may be ascertained by two general rrethods: (a) the current may be observed directly with a sensitive current meter, or (b) the potential drop across a known standard resistor in series with the high resistor amav be measured. Both of thlese methods were tried in the primary calibration of the integrator, but only'the second one (employing a potentiometer) gave satisfactory results. However, the details of both will be given. 1. Direct measurement of current ~'atts (30) suggested the use of a micromicroamineter for calibration purposes. No such instrument was available in this laboratory but it was possible to borrow a sensitive portable galvanometer, (Leeds and Northrup type R No. 2500

rated sensitivity 0.0005/,a/mm). It was used in conjunction with a lamp and scale employing a 2 meter optical path. If the current sensitivity is known accurately, the galvanometer may be used as a mricroammeter. This expedient was used by Poole. (24) The sensitivity was determined according to the potential divider method given by Smith (63) and indicated schematically in figure 20A. P was a decade box (Leeds and iNorthrup N o. Q750) another decade box (General Radio type 602-J), R a wire-wound 1 per cent ITC resistor, and S a decade box (Leeds and Northrup Rio. 4775). A potentiometer (1iubicon type B tNo. 2780) was used rather than a voltmeter. current flowed through th!e potentiometer coils for at least 24 hours before use. The 10,000 ohm coil of P was used as a standard of comparison for the 10,000 ohm coil of and A. Shielded cables (h.G-u2/U) were use'd everynwhere except uAtittl the potentiometer. Thie contacts were carefully cleaned, and solder connections made with 50-50 (50 per cent lead, 50 per cent tin) solder. The galvanometer coil was warmed up with several wide deflections before a series of readings was taken. During use, the zero point shifted by about 1 per cent or less of the deflection after the passage of current. Therefore the average of' t hle zero point readings immediately before and after a deflection was assumed to be the true point. Tile results of four calibrations gave smooth curves from +250 to -250 num deflection except in the neighborhood

-56. CALIBRATION SCHEMATICS A STANDARD CELL STO RAGEE R BPOTENTIOMETER LV GALJ S STORAGE BATTERY B UNIVERSAL SHUNT HI -— MEG. FARADAY RESISTANCE CURRENT CUP POWER VTV BOX SUPP LYI N T EGRATOR C STANDARD POTENTIOM ETER STORAGE CELL BATTERY HI-MEG. STANDARD FARADAY H.V. RESISTANGE RESISTOR CURRENT CUP POWER BOX SUPPLY INTEGRATOR Fig. 20~ Schematic representation of the apparatus to A, measure the current sensitivity of the gal-vanometer; B, calibrate the current integrator by direct measurement of the current with a galvanometer; and C, calibrate the current integrator by the potentiometer method.

-57of zero, where it becomCs Cm c more sensitive. This discontinuity in the smooth curve was considered abnormal, but a more serious drawback was thle fact that successive calibrations over a period of several days gave curves showing decreasing sensitivity(all in the vicinity of 0.00035fa/mm) wh.ich was still lower than the rated sensitivity. The differences amounted to several per cent. ~ven when the galvanometer was not moved between two runs, the second curve lay about 1/3 per cent above tihe first. Aifter using the galvanonlmeter to calibrate the integrator, the current sensitivity was redetermined by 0. U. Anders, using slightly different techniques. His results, although somewlat more reproducible thanr the first cali' oration, were hihller by 5 per, cent. The non-reproducibility of this method is not attributable to temperature or humidity differences. The explanation may lie in the fact t hat no particular care was taken to assure the galvanon.rmeter was perfectly level. Other workers have observed difficulties similar to those experienced here until the instrument was leveled.(64) Two attempts were made to calibrate the integrator using the set up shown in figure 203. The universal shunt was used to extend the range of the galvanometer. The vacuum tube voltmeter V.T.V.I. was used merely to read the H.V. power supply output. The calibration curves obtained by this meth'od were not to be trusted because of tihe uncertainty in the galvanometer calibration.

2. Potent iometer method In this method the accuracy is limited due to the decrease in sensitivity brought about by the high values of the standard resistance necessary to give measurable potential drops for very low currents. Potentiometers are not designed to measure potentials in circuits containing very high resistances. A practical limit was found to be -10,000 ohms for the standard resistance. If much larger values are tried the galvanometer does not respond to slidewire changes of less than 1 per cent of the indicated potential. The particular potentiometer used also makes a difference, e.g. it was f6und that the Rubicon type B, and Leeds and Northrup type K are satisfactory, but that the Leeds and Northrup No. 8662 microvolt potentiometer is not, as it provided much less sensitivity. There are several ways to surmount this difficulty: by the use of a more sensitive galvanometer, by increasing its optical path, or by connecting a condenser in parallel with the standard resistance, etc. However, it was felt that the integrator would serve adequately for the final cross section deternminations if calibrated only for its two highest ranges C and D, corresponding to currents from 0.05 to 5 Qpa. These values provide 0.05 to 5 volt potential drops across a 10,000 ohm standard resistor which can be measured with the Rubicon potentiometer and the Leeds and Northrup type i galvanometer to good precision.

-59The schematic block diagram of tie calibration set up is shown in figure 20C. In the early runs, the Hi-Meg resistors (The Victoreen Instrument Company, Cleveland, Ohio) were out in the open, but for the final calibration tiiey were enclosed in an air' tight box with steatite selector switcjhes and two UG-496/U cable connectors. Care was talen to clean the resistors, and a bag of silida gel was placed inside the box. Ceresin wax was used to seal the box. The standard resistor at first was an IL-C 1 per cent wire-wound product, hut later a 10,000 ohm Leeds and Northrup Do. 4040 resistor, standardized by the National Bureau of Standards in 1954, was used. For the power supply, an Atomic Instrument Company model 312 High Volta-e Power Supply was used. (Earlier, the integrator control R49, figure 9, was used for this purpose, but the highest potential available was generally too low.) The vacuum tube volt meter, a Voltohmist Senior (Radio Corporation of America), measured the power supply output. A Rubicon type B potentiometer was used in conjunction with an Eplab standard cell and a Leeds and Northrup type R No. 2500 galvanometer. The results of the calibration are depicted in figure 21 for ranges C and D. Although data were obtained extending to -40 counts/min, only t-he lo-wer portion of the curves are shown since deviation from linearity is noticeable beyond 12 counts/min on range C and 6 counts/min on range D. Treating the points below these limits by the method of

CALIBRATION OF BEAM INTEGRATOR 14 / o RANGE C 0 RANGE D 13 SLOPE Kc =39+4 12 I9 cr 8LLD 8 /SLOPE = KD =9.9~O0.1 7- ~i 06 5 4 3 2 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1,0 MICROAMPERES, Fir 21i CaliL ration cur,-vos o2f t:e beam current integrator.

-61least squares gives the following values for the slopes, KC and KD for ranges C and D respectively: 37.6 + 3.6 counts/min/7a and 9.0 0 + 0.03 counts/min/M a. The intercepts were not zero, as would be expected for an ideal integrator, but were 0.30 + 0.15 and 0*045 + 0.010 counts/min for ranges C and D. This could be the drift rate encountered when no input current is fed into the integrator, except that in this case the intercept should be negative for range D to agree with observations If the intercept is assumed to be zero, KC and KD become 39.4 and 9.97 counts/min//Aa respectively, but the deviations show a trend when plotted. The curves obtained about six months previously using less care in the measurements exhibited approxi-mately the same slope: 43.4 and 10 counts/min// a for ranges C and D respectively. The following values were taken as the calibration factors for use in obtaining absolute measurements of the beam: KC = 39 + 4 counts/min//Aa KC = 99 + 0.1 counts/mrin a. These numbers represent the average of the two least squares slopes, one with the both constants determined and the other assuming the intercept equal to zero. The errors assigned are the mean deviation except when the statistical standard deviation was larger, it was chosen. lhote that

62the errors thus assigned bracket tnhe early values, so tihat during the succeeding two month period in which thne cross section bombardments were obtained, the change in calibration factor should be less than these errors. Secondary Calibrations Secondary calibrations were made every time bombard-Ments were obtained to detect any day-to-day variation in the instrument. For this purpose a more rugged apparatus was used than in the primary calibrations. The vacuum tube voltmeter and Hi-meg resistance box were used as a microammeter, and the set-up is the same as in figure 20C, except that the standard resistori and potentiometer were removed. (The readings thus obtained indicated drifts not only in the integrator but also in the V.T.V.M.Hi-meg resistance combination.) During the two month course of the bombardment work the secondary calibrations were found to show no long term drift (except perhaps at the very end) and the readings for a given V.T.V.1-..-Hi-meg resistance combination on range D always duplicated tihemselves with an average deviation of + 0.7 - 1.6 per cent. These variations might be attributed to changes in the V.T.V. r., the Hi-meg resistors, temperature or humidity. It was noted, however, that the integrator must be warmed up at least two hours before use, in order to reproduce the readings within the above limits.

-63 - B. Target Preparation in order to meas ure cross sections at well-defined deuteron energies it is essential teat the target m.aterial not degrade the incident energy significantly. The target itself mist therefore be made quite thin in order to take advantage of tlne 6.5kev/mm resolution of the focusing mnagnet. For example, ala.minlum only 0.00026 inches thick will degrade the energy of 7*. Haev deuterons by 127 kev, which represents the resolution obtained from a 3/4 inch slit at L, figure 1. ( second consideration is that the targets must be made free from thickness gradients. A uniform thin film may thlen be bombarded with a deuteron beam inhomogeneous with respect to intensity variations over the beam area, and thle cross section calculated without a knowledge of the beam intensity distribution. Thirdly, thre target mn.st be free from impurities which might give rise to interfering activities which would follow the prodict element sought in the chemical separation. This is particularly important in the present study because the (d,o) yields are typically a factor of 103 lower than those of the (d,n) and (d,p) reaction which are the major reactions induced by 7.8 vLev aeuterons. For example, the (d,n) reaction on a 0.1 per cent impurity might produce comparable numbers of radioactive atoms compared to the (d,a) product.

-64For this reason considerable time was spent in locating the highest purity target materials commercially available. In those cases in which the material was evaporated onto a substrate some additional purification may result from the process, but there is also the possibility that it may actually introduce impurities. The possibility of magnesium oxide formation in this manner is discussed below. Several general methods of producing thin films have been reviewed by Hudswell. (65) The metal evaporator described earlier on page 24 was used for this purpose, but it was soon found that this approach has its drawbacks. Data on the techniques for evaporating various metals from different types-of heaters are presented in the literature. The problem of finding a suitable substrate to which the evaporated material will adhere was the subject of many trials. Several different substrates were tried for each of the target elements. These were polystyrene, rubber hydrochloride, Mylar, Saran, cellophane and aluminum. Table I lists the four target elements encountered, some of their properties and the target preparation method.

Table I. Target materials Mg Q Ti Cd mp( C) 651 113 1800 321 bp(~C) 1110 445 >3000 767 mg/cm2/mil 4.4 5.3 11.4 22.0 pur ity --- 99.9w8 99.99, spec-pure" availability Dow Stauffer Foote Johnson, Mat they preparation evap. poured foil evap. substrate iylar polystyrene --- A1 evaporator heater IT boat --- - boat Target mounts were made to fit the slots of the probe head of the bombardment chamber and were machined from 1/16 inch aluminum stock. A hole 2 1/4 inches in diameter was cut in the mounts to match the 2 1/4 inch opening of the FAaraday cup. The substrate was glued over the hole with rubber cement. In practically all cases the target material was restricted to this 2 1/4 inch area by means of a spare target mount placed over the substrate to serve as a collimator during the deposition of the target. To prevent shadows of the edge of the spare mount from being cast in the evaporation the circular edges of the collimator were beveled. Thle details of the target preparation for each of the four elements studied will be given below.

-6!agnes ium Magnesium metal purified by sublimation was kindly furnished by Mr. A. D. Brooks of the Dow Chemical Company. This material sublimes in vacuo, and the unscreened granules of the Dow magnesium have a tendency to pop out of the heater unless the heating is done extremely slowly. At first Vycor crucibles were used to contain the magnesium, but their deep shape made them very efficient little cannons which shot larg:e particles of magnesiun right through the thin substratel Alundum crucibles were also tried before settling upon the shallow tungsten boat. W:Jith care, a full boat load could be sublimed with little loss from po,?.pin!c out. [When polystyrene was tried as a substrate, the magnesium would not stick and in addition the heat in the evaporation process is enough to melt the plastic. Magnesium adheres well to Teflon, but it can't be easily decomposed in the chemical procedure (see Chemical Separations section). mylar ( 1 mil thick) appeared to be the best backing material. In order for even deposits to result, the surface of the iMiylar had to be carefully cleaned with ether and acetone and then wiped dry to eliminate drying marks. Two or three evaporations were required to produce a film thick enoug-h to bombard. curious effect that was sometimes noted was that the surface of the evaporated magnesium was grayish and metallic looking, while at other times it had a whitish cast. There

7is some indication tihat this was due to inadvertently allowing the pressure to rise while the evaporation was in progress. Thus air or water vapor may have reacted with tihe magnesium resulting in the formation of the oxide. Ai check was -mriade on the uniformity of evaporation by cutting one of the 2 1/4 inch diameter films into eight rectangular pieces and determining their thicknesses in milligrams per square centimeter (metal plus substrate). The pieces were wei-hned on an Ainsworth type 1ODJ microbalance and their areas measured with the aid of a pocket ~riagnifier withrit a scale attached (:;dmund Scientific Cor-poration, Barring-rton, New Jersey). The standard deviation of tihe surface densities was 0.4 per cent which is of' the same magnitude as the experimental errors of thle individual measurements. Thus it is seen that individual films can be rmade extremely uniform in thickness, although these results are not directly applicable to specific targets bombarded since the films had to be cut up in order to make the measulrement. Sulfur The sulfur used was 99.98 per cent flowers purified by redistillation (66) furnished as a sample lot by H. C. Thomas of the StauiLfer Chemical Company. Sulfur could be evaporated from Vycor or alundum cruci:bles onto.iKylar substrates, but a much simpler

process was finally adopted. This amounted to pouring sulfur vapor onto the substrate. (67) Sulfur was melted in one arm of a Y-shaped tube, which was then tilted so that the liquid would run into the other arm while the vapor pours out the open mouth of the tube. Difficulty was expierienced with this element in that under bombardment the sulfur would tend to vaporize from the plastic backing,. This could be partially overcome by sandwiching the target between two layers of polystyrene, according to the following procedure. Sulfur vapor was poured onto a cleaned 0.9 mil polystyrene film and the weight of the sulfur was determined. Next acetone was sprayed over the sulfur side and another piece of polystyi-rene placed on top of the target. The sandwich was pressed togeth-ier to f"g;lue"t the sulfur firmly in place. a'hen this assembly was subjected to the beam of the cyclotron, the vaporization noted earlier was lessened. It was noted that the target films made in this way were not as uniform as those made by evaporation. The sulfur condensed in small individual grains. On standing, crystallization occurred, no matter which way they were prepared. An attempt was made to assess the evenness by scannino the target with a photomicro-densitometer, but the grain size proved to be too large for good results. There was, however, evidence for slight thickness gradients

-69over larger parts of the film. A second attempt to measure the uniformity was made oy cutting up a film and weighing the pieces. Some of the sulfur stuck to the razor as the pieces were beinrg cut, however, and the reproducibility of the surface density of individual pieces was not nearly so good as with mafgnesium films. In this case the standard deviation of 11 pieces was + 2.2 per cent. A gradient aimrounting to differences of 15 per cent may have been in evidence in going across the film. Some success was experienced with evaporating sodium s;ilfate onto iylar. This approach should yield much better targets both from the point of view of uniformity, ease of dissolution of the target, and stability under bombardment. However, the techniques of evaporating salts were not fully worked out, and the targets used in the bombardments were made of elemental sulfur. Titanium A sample lot of 99.99 per cent titanium sheet 0.G0026 inchles thick was supplied by iJ. I<. Laynor of the Foote hineral Company. A typical analysis is given in Table IIo This metal is so refractory (see Table I) that vacuum evaporation is difficult. INot only are evaporated films difficult to prepare, but too thin a target won't result in much yield of the (d,a) product. This is due to the long half-lives of the scandium isotopes involved and the cyclotron beam available. Therefore it was decided to

-70use the titanium foils as received from Foote >iineral Company. Table II. Spectrographic Analysis of Foote Titanium Si 0005/: Cr 0. 0005 Ca 0.0005 C 0.00 1 Al 0.00C1 Sn O.0001 Cu 0.001 02 0.002 i.o. CC0002 Fe 0. 003 Ti --- N2 0.002 lb i' hli ui. 0.001 D H2.<n 0. 001 ko ND HF No te: Nl)D Niot detected — Iot determined Tlie thickness and uniformity was established by measurernent of a portion of thie titanium foil (not the same portion t. at was bombarded) with a micrometer. It was found to be 0.00256 + 0.00013 inches thick, where the error is the averag-e deviation of 32 readings. Cadmium The metal used was Johnsqn, l'atthey and Company, Limited -I. S. Cadmium hl'od, greater thian 99.99 per cent pure. The spectrographic analysis is given in Table III. Pieces were cut from the rod vwith wire cutters, then cleaned with dilute 1hya9drochloric acid and distilled water. The evaporation was best carried out from a tungsten boat, although alundum crucibles also work well.

-71It proved to be no small task to find a substrate to which cadmium would adhere in the evaporation. The plastic films tried inclJde i-rlar, Saran, polystyrrene, cellophane, rubber hydrochloride, Formvar and Zapan. Cadm;lium oxide was also tried instead of cadmium, but with no more success. Finally it was found that aluminum foil such as {Leynold's Wsrap,works, provided that its surface is first roughened with Dutch Cleanser. Targets so produced were very even and mechanically strong. Aun interesting observation was -made during this evaporation. The cadmium metal h'ad a beautiful, shiny, crystalline appearance up to the point of fusion. But just as it melted into a liquid globule, a brown scum appeared in the surface. This apparently did not evaporate with the metal but was left as a residue in the boat. Perhaps this is cadmium oxide formed from the residual gas present at pressures of 103 - 10-5mm of mercury used in the evaporation. Nlo attempt was made to evaluate the uniformity of these films.''he data on tIh-e maj-;nesium foils an&ra similar data obtained in the saeme way on mffan-anese anew indium foils made in thir3 laboratory (68) were taken as sufficient evidence for thle absence of significant thickness gradients. 1 ~~~~~~~ C)~~~~'

-72Table III. Spectrographic Analysis of Johnson, Matthey Cadmium Rods Ag <o0.001 Cu faintly visible Ca very faintly visible Fe barely visible No lines of the following elements were observed: Al, As, Au, Ba, Be, Bi, Co, Cr, Ga, Ge, Hg, In, K, Li, Mg, Mn, Mo, ha, Ni, Pb, Jib, Sb, Sn, Si, Sr, Ti, Tl, V, W, Zn, Zr. C. Chemical Separations The object of the chemical separations performed on an irradiated target is to isolate the atoms of one particular element produced by transmutation. It is not necessary to recover all of the atoms, so long as it is possible to determine the fraction recovered (chemical yield). It is important that the elemental fraction be radiochemically pure, that is, that radioactive impurities of other elements be reduced below the level of detection for the particular counting arrangement used. The disintegration rate of ttle product element may then be determined by the techniques of 4li beta counting as described in the next section.

-73One much used method of determining the chemical yield is by gravimetric procedure. A known amount (often 10 - 20mg) of the (inactive) element in question is added to a solution of the irradiated target. This is assumed to behave chemically the same as the radioactive atoms produced, and hence is known as a carrier. Carriers are also added for the various elements suspected as radiochemical impurities. Ordinary distillations, filtrations, extractions, etc. are carried out to insure interchange of the carrier with tale active species and to isolate the element of interest. The final precipitate is weighed and the chemical yield is calculated. Diesintegration rate measuremaents made with samples containing several milligrams of carrier are made more difficult by the necessity of making self absorption corrections to take into account the absorption ofmak [I rays in the sample itself (see next section on absolute beta counting). If the elemental fraction can be obtained in a solid-free or "weigh;tless" state, however, selfaosorption can be ignored. The chemical yield can be ascertained without the addition of a carrier by adding a known amnount of a radioactive tracer of the element sought to Whe target solution. This tracer may be either carr-ier free or of very high specific activity. This method is commonly employed using a-emitting tracers of the heavy elements. In this work the 3-emitting tracer was longlived, and after the short-lived species produced in the

bombardment had decayed out, the tracer was counted and related to the amount originally added. There is one note of caution that should be sounded with respect to either of these methods of estimation of chnemical yield. In both cases, complete interchange between the isotopes of the added tracer or carrier and those produced in the tarcgt must be establihled so that neasurernents on the added species are truly representative of tehae transmut-,d ato;>rs, 7 das is not neC fC riiy 3o obvious a matter as obtaining a homogenous solutions For instance, the fission yield of iodine was in error by several per cent until a satisfactory oxidation-reduction cycle had been perfected which ensured complete intercnange. (69) It is felt teat this point has been overlooked by nmany workers in the field. It was fou-nd that the purity of the water used to make up reagent solutions is a critical consideration in obtaining a final solution free from all solids. This consideration applies to all the carrier free separations performed in this research. extra precaution must be exercised in column separations as thle activity is collected over a large number of 8 ml fractions totaling in ion-exchange -500 ml eluate. From 10 - 100 ml of solution must frequently be evaporated prior to the sample preparation with t`ie result t'hat very low concentrations of impurities may result in sizable residues. The best answer found to this problem is to use "conductivity water" (- lxlO 6ohm-lcmm- ) n I n -y- n c — -; rl i -, en a- v-% A 4-~~~~~~~~~~~~

-75for making solutions, washing the colwnns, etc. Furthermore if there is a large amount of activity produced, only a fraction of the total volume need be evaporated to ijrepare a sample for counting. Thus in column separations only one 8 ml fraction under the peak of the elution curve may have enough activity to prepare a counting sample yet contain only a microgram of solid material. In these procedures the substrate upon which the target element had been placed had to be dissolved and the (d,a) product extracted. This is because nuclei recoiling from the collision with the deuteron may be ejected from the target. (Experiments demonstrating this effect will be described later.) Sodium from ~Iag nesium Targets The separation of the alkali metals from each other and from the alkaline earths has been resorted by several authors. (70 - 75) The work of Linnenbom (75) and Beukenkamp and Rienman (73) was found especially useful in developing thre procedure used here. Several preliminary experiments indicated that by using Dowex 50, a cation resin, with dilute HC1 as the eluant, the magnesium is held up by the resin so that the sodium is eluted first. Attemptingt to purify colloidal Dowex 50 by settling, washing with hydrochloric acid and sodium hydroxide solutions, etc., proved ineffective in providing a resin from which no extraneous material would be extracted in the elution

-76process. However, "CP" grades of Dowex 50 which have been screened, purified by extraction with various reagents, and analyzed, are now commercially available (Dio.iRad Laboratories, Berkeley, California). Number AG 50-X8, 100 - 200 mesh, was the resih used in this work. After the column was loaded and rinsed with conductivity water, no J urther treatment was necessary. The column was made from 22 mm i.d. 25 mm o.d. Pyrex tuAbing of cross sectional area 3.8 cm2, and was -13 inches high. A coarse sintered glass frit was fused in the bottom to support the resin column, and a buret tip and stopcock formed the outlet. The column was filled with resin to a height of about 10 inches. An automatic sample changer (University of California Radiation Laboratory, Health Chemistry design) was used to collect the eluate in -8 ml fractions in 100x15 mm test tubes (Arthur H. Thomas Company, Iho. 9446). The annotated procedure ifor magnesium targets is given in figure 22. The formlat is thle same as used by Professor Iveinke.(76) The initial sulfuric acid-hydrogen peroxide treatment results in tfhe destruction of the Mylar backing. In this case, once a homogeneous solution is obtained, the sodium tracer may be assumed to be mixed thoroughly with that produced in the bombardment. By fuming off the lar;ge excess of sulfuric acid much less magnesium is needed for neutralization, and,the capacity of the resin is not exceeded.

-77 - CHSI:~~ICAL 3EPAilR.TIONS Element separated: Sodiur Procedure by: Hall Target Material: Magnesium Time for sep'n: -8 hours Type of bbdt: 7. 8 Mev deuterons Equipment required: Yield: -5OCo 250 ml rhillips beaker Yield:'50% ion exchange column Degree of purification: -103 automatic sample changer 10Ox15 mm test tubes Advantages: carrier free (Arthur I Tomas Co,, separation for 4 P[-counting Lo. 9416) Procedure: (1) Place 1 ma con I-12S04 in 250 ml Phillips beaker and add tracer Na 2 (see remark 2 ). Introduce the 1Mg target (-20 mg) and l~ylar substrate. HIeat strongly to decompose the - ylar. (2) Cool, add several drops of 30O, H202 and reheat. Repeat until a. clear sblution is obtained. (3) Fume off excess 112S04 and add 10 ml water. Neutralize (to pH-4) by adding an excess high purity Mg. (4) iAbsorb onto 1o-?ex 50 column (H+ c.mn) at a flow rate of 2.5 sec/drop. Rinse the column with 10 ml of iater. (5) Elute with 0.5 I HC1 at 2.5 sec/drop, collecting the eluates in clean 100x15 mm test tubes. Assay in scintillation well counter to determine the elution curve. (6) Test sodium fractions for Mg+ with quinalizarin spot test (see remark 3 ), (7) Evaporate t!ue most active fractions to driyness. Take up with water and transfer to 4rr plates. Reiriiark ks: (1) General references: J. Beukenkamp and -. Rieman III, Anal. Chem. 22, 582 (50) and V. J. Linnenborn, J. Chem. is. i, 1657 (525. (2) The tracer is added for the purpose of determining the chemical yield. (3) Reference for quinalizarin test: Feigl, Qualitative inaylsisp b ty:j Tests, 3rd ed. (New York,94T), p. 172 (4) Use conductivity water to make up solutions, etc. Fig. 22. Procedure for the chemical separation of sodium from ma{jnesium targets,

The elution curve is most easily obtained by counting thle test tubes t;eemselves in a scintillation well counter. (TPie one -used in this laboratory was a N'uclear-Chicago model DS-3.) This eliminates the necessity for takinc aliquots of the fractions and making separate counting samples. The uia22-24 starts to appear in the eluate after about 500 ml of eluant has passed through. A bombard-ent to check the chemical procedure was obtained using the wrindow box probe described on page 17. hligher intensity beams by a factor of -100 are available here compared with the bombardment chamber and hence lar, er amounts -of activity can be produced. The magnesium was irradiated in an aluminum envelope (without the plastic substrate.) The purified sodium fraction was counted on the spherical 4n, counter and tlhe decay showed two components: e2 6-year 2a2 24,the 2.6-year a22 and the 15-hour a24 as expected. I\io other hal!f-lives were detected. Phosphorus from Sulfur Targets A unique application of ion exchange resins has been developed by MicIssac and Voigt (77) for the carrier free separation of phosphorous from neutron irradiated sulfur. Dowex 50 was first converted to the Fe form, and then ammonium hydroxide was passed, through the column precipitating ferric hydroxide. [adioactive phosphiorous is known to be carried by ferric hydroxide, so when an active solution

-79was passed through this column, the phosphorous was absorbed while the sulfur passed on. through as sulfate. The column was then rinsed free from sulfate with water, and the phosphorous removed by eluting with dilute sodium hydroxide solution. The effluent was channeled directly into a Dowex 50 column in the H+ form. By this means the sodium was removed by the resin, and a carrier free solution of radiophosphorous was obtained in aqueous solution. Bio*Rad resin AG 50-X7.5, mesh sized 20-50, was used. The details of' the preparation of the resin columns are given by McIsaac and Voigt. (77) Figure 23 shows the procedure as finally adopted. The initial sulfuric acid-peroxide treatment destroyed the polystyrene target support. The bromine-carbon tstrachaloride mixture dissolved the elemental sulfur and converted it to sullfate. (78) In the absolute cross section runs it is important that the mixture not be heated too strongly as phosphorus may be lost from sulfuric acid solutions at temperat ires above 150~ C.(79) The neutralization step yields a solution identical to the one with which 1~cIssac and Voigt started. The problem of complete interchange of the added P33 tracer and the p32 produced in the cyclotron deserves consideration. There are available a number of valence states and molecular species among which the p32 and P33 atoms might distribute themselves. A separate experiirent to ascertain whether or not 100 per cent interchange

-o.was accomplished did not seem to be worthwhile in the lipght of several ot'jer expsri-aental difficulties with this element. (Thle targets were not uniform, the sulfur tended to vaporize under bomrbardment, the P33 tracer was iipure, resolution of the decay curves was difficult, and the use of hot sulfuric acid may have resulted in losses of p32F ) The strong oxidizing conditions used in the dissolution step assures that the phosphorus is in the +5 state, and it wlas only assumed that interchange then occurs rapidly. Sulfur was bombarded in an aluminum envelope on the window box probe in order to test the chemical procedure. rhe target was cooled two days before processing it chemically. {io plastic film was present in this case. The decay curve shlowed the presence of only the single 14.3-day decay of P32 An aluminum absorption curve did not give any evidence for the beta ray of P33 which could possibly come from an (n,p) reaction on S33. Scandi:im from Titanium Targets Carrier free scahdium separations from calcium and titaniurim tar lets have bc er~ rerorted Jy Jile et al. (k2) and oh" (:C1).'' e;! i clea:~r i,, c;= i e- olution'i7 containi:;tracer amounts of scandium was filtered through filter paper the scandium was retained by the paper as a radiocolloid while the other activities passed on through. The scandium was removed with dilute hydrochloric acid.

MTI" KI CA L 0i." A T I ON 3 Element separated: IhospLIorus irocedure by: Hall Target Mlaterial: LSulfur Time for sep'n: -8 hours Type of bbdt: 7. 8 lev deuterons eq.ipment required: 250 ml i'hillips beaker ion exchange columns Degree of purification: -103 aultomatic sample changer 10x15 mm test tubes Advantages: carrier free (Art0lur Thoe as Cot (Arthur Ii. Thomas Co., separation for 4TT B-counting 9446)?-o. 9446) Procedure: (1) Place 2-3 ml con H2SO4 in 250 ml P'.illips beaker and add tracer P33(see remark 2). Introduce the S target (-70 ing) and the polystyrene substrate. Add several drops of 302; H4202 and heat gently to decompose the polystyrene (see reimark 3). (2) Cool arind add 25 iml of a 2:3 ilixture of Br2-CC14. Let stand 30 minubes to dissolve the sulfur. Add nore H2S044-H202 and hoat gently to cvmplete the solution. (3) INeutralize to pil-6.4 with solid Na2C03 and dilute to 500 mi. (4) Absorb onto Dowex 50 column in Fe(O H)3 form at a flow rate of 1 sec/drop. Rinse with 50 ml water. (5) Elute with 125 ml 0.125 N NaOi- at a flow rate of 1 sec/drop, cnannelling the effluent directly into a Dowex 50 columrn in H+ form. Co.lect thie eluates in Clean 100x15 mm test tubes. Assay in a scintillation well counter to determine the elution curve. (6) Test phosphorus fractions for 304' with Ba+. (7) Evaporate the most active fractions to dryness. Take up with water and transfer to 4rr plates. Remarks: (1) General reference: L. D. >;cIsaac and A. Voigt, ISC-271, (June 1952). (2) The tracer is added for the purpose of determining the chemical yield. (3) If the H2S04 solution is heated to fumes, some P04 may be lost. (4) Use conductivity water to make up solutions, etc. Firg. 23. Procedure for t; e c em ical separation of phosphorus from sulfur targets.

In this work three cycles were found adequate to give the desir ed decontamination. It wras discovered t at i,-tlat;man [o. 50 hardened filter paper is suuperior to either of the papers recommended in t1ie references. certain ainount of organic rmatter was extracted along with tle scandium when the other papers Jrere treated with hydroch!loric acid. This did not happen ii en the h-ardened -japer was used. The details of th)e adopted procedure are given in fij-zure 24. One problem that Gile, et al. (< 2) did not mention was th at of kee, jing he titaniuamS in solution at pH 8.5. The hydrous oxide tendeJ to precipitate as soon as the 1pI was raised to ~3. It vas exper imentally found that thais can be overcome by adding sufficient hydrogen peroxide which is reported to form a peroxide with titanium. (F3);ieeping thie solution dilute also helps to prevent the precipitate from salting out. dIn this case there is little caulse for concern about complete interchange, as scandium exists in only the 0 or +3 state. Thus the initial oxidation assures that all the scandium is in the same form and thorough mixing of the tracer and (d,a) product can be assumed. The chemical procedure was evaluated by a window box probe bombardm!ent of titanium. Three cycles of the procedure resulted in a sample which decayed with an initial -3.5-day half-life tailing into an i5-day component. The P-ray spectrum, as observed on the -ray survey

Tyi-,ent se~;rrate ofcandium v-: a'' e v of bbidt: 7.:. C ev,euterons qiipimnnt r eqair:ed v~iei~.~a: ~~~>250 ml i.i liiJs Seaker m icro elil'ar jLe"ree of juriiicatiln: -1i2 eG,.C 1ri;'sch funlrels,~-i t~:an o. 50 fi-1ter -dva nta es: carrier free p-a, er, pilydrion (short sepairations for 4,-Tr 53-co-nting ra pH - - e: (1) i lace 1 1 ccn 2, in 25u i ilis beaker and aud tracer Sc46 (Cee remiark 2 ). Introd'ce the Ti ar;et ( 130 ri )c a: u t ie:iUylar s i-strate Leat strongly to aco:iose ttie 10 iar. (2) Cool, ad5d everal e roa;. s of 30,3'C202 roh~ t. Les eat until a clear solution is oi3otained above the Unuttaced Ti. )t did 0 m! l 10: i:2,cc i;tanlir- 51 16 136, iTat ()) 16, 03. ieat keeping the.Ui}rep lenised until t e Ti is all dissolved. t(2~) fi.i.ute to.10 si~;':- 1 n t-r. arlize to p -,'t i(use <iiydrion i- oer ) nit> a 1:15 cixt re of 30 H d202 and.. r 0 d- eo0j::: ex cess H2202 to keep thle Ti in oTIuti l. (.5) Filter twice t.irouo:j t> l samle l~tai 1io. 50 filter pai er,-sin- scctioO-:. ndash three t iiies withl 3 _'h4CI at H. v. (6) he ieove Sc wit h several portions of; ot 3 N IC1. (7) i{Ce eat steps 4, p an. o twi e, except in final c ycle Aie C,'iouiCtiv:it-jy wat-r at p'.5 to wraso te Cc "precipitate". (:),lviruorate to di'ynneo..etro,' orlg anic!atter Orith aq re T ii ater ae transfer to 4w (1) Oertex-'1 refer ence' J:l. aile,t al., J.,ilev... _, I e, (- ( (2, Th le ura er is a _aed for toe. lrjose,of a:deter: ing the chea:lcal vJieLi (3) Use condacti:ityr later to SI;a e ip solutions, etc. i:. 24.i roc dure L'or tlie chemical sep/aration of sc-a' i::i i l,;' iL r Jb i tar i iuIf

spectrometer (8;4) was consistent with the expected radiations. The Y-ray spectrum obtained with a scintillation spectrometer (85) also showed the expected peaks at 0.9 and 1.1 Mev for Oc46. Silver from Cadmium Targets 4From a consideration of thie data presented by Hicks et al. (86) it was thought t'sat a rapid carrier-free separation would be possible for silver. They indicate th at tracer quantities of silver can be separated from cadmium, indium, palladium, etc. by eluting tracer quantities of silver on Dowex 2 with 6 - 9 NI hydrochloric acid, while the other ions remain on the column. However, it was not possible to separate Pdll and A1 ill in this way when tried in thle laboratory, which cast suspicion on the method. (IElvidently their data was obtained using a bad batch of resin ( 7) which explains why their results could not be reproduced here.) iext the procedure of Haymond et al. (a;) for the carrier free isolation of silver from palladium targets was tried. They claimed to have co-precipitated tracer silver with merc;rous chloride. To 500 ml of 0.5 N hydrochloric acid solution of the palladium tar;-aet containing rhodium and ruthenium carriers they added 0.5 ml of saturated mercurous nitrate to precipitate the chloride. However, when this procedure was tried here using irradiated cadmium plus various holdback carriers

- 3 such as Cu, Zn, Ga, In, tib, and Bi, very little silver activity coprecipitated with the mrercurous chloride. (The presence of active silver was demonstrated by precipitating silver chloride from thle mlercurous chloride supernate and following the decay of th-e silver chloride. ) It was hoped to separate the carrier mercury from the silver su.bsequently by an anion column procedure. However, repeated experiments failed to yield success in isolating t-Le radioactive silver. At one stage it was noted that wlienever palladium is present in the solution, the mercurous chloride precipitate turns black. Thus it may have ibeen that the mercurous species redLuced the palladiumi to v;e metal ih ich in turn carrie th le ilver in t-his,ro cedure..iather thlan spending i,.ore timle dve oine A c'arrier-ree separation it was decided to resort to the addition of silver carrier. A good clean, fast procedure based upon the rapid exchange of radio-silver with inactive silver chloride suspended on a platinum gauze was developed in this laboratory by Sunderman and I-einke. (E9) This, together with a more standard precipitation procedure developed 104 for use with the Ag mass assignment problem (Part II of this thesis) was used in the cross section work. The discussion of the latter procedure is given in Fart II. Here again the question of exchange of carrier and trace amounts of silver is no problem. In fact th-e very rapid rate of heterogeneous exc:aige between the solid silver chloride and Ag in solution has been the subject of

study I (O9 - 90) it has been reported that silver ions tend to absorb on tile walls ofC glass containers. (91 - 92) This effect is precluded by addition of silver carrier to th:e vessel before dissolving the target. Tihat the exchange separation does, in fact, give good decontamination from the other cadmiulm bombardment products was shown by a separate bombardment at the window box position. Cacdmium foil of 99.99 per cent purity (Belmont Smelting and Refining -iJorks, Incorporated, Brooklyn, New York) waas backed with Reynolds?;rap aluminum foil and bombarded for about 2/A a-hr. The silver was separated and its decay followed. It showed the presence of just the same components that were found in all the previous work using the well tested silver chloride precipitation procedure (Part II of this thesis).

D. Absolute Beta Counting Summary of Existing Methods The problem of determining the rate at which a radioactive source decays in terms of actual disintegrations per unit time has been the subject of a great deal of study by many workers over the past decade. Manov (93) has presented a concise summary of the status of radionuclide standardization as of April 1, 1953. The various avenues of approach group themselves into three broad classifications. In the first method a measurement is made of the rate of energy released by the radioactive nuclei and the disintegration rate is calculated from this datum and a knowledge of the average energy per disintegration. Secondly, the coincidence counting rate of a sample may be related to the disintegration rate without knowing the efficiency of the detectors for the various radiations emitted. Finally, the particles ejected from the nucleus may be counted by a detector whose "counting efficiency" (defined as the number of counts recorded under particular conditions of geometry, scattering, absorption, etc. for each disintegration) is known. 1. Rate of Enerlgy Release. Myers (94) has written a review article on the subject of calorimetric methods of measuring the heat emitted from a radioactive source. The rate of evolution of heat dQ/dt, the average energy per disintegration 2, and the disintegration rate dN/dt are related as follows:

-d E dN (7) The radiation must be completely absorbed in the calorimeter. Very intense sources are required to give off enough heat to be measured accurately. Since the limit of sensitivity of most calorimeters described is 0.001 cal/hr, this means about 20 mC (4x109 disintegrations/min) of a 1 U!ev n-emitter is required for 1 per cent accuracy in determining dQ/dt. This follows from the equation given b-jy lMyers: 1 cal/hr _ 0.19614 curie (8) whe-'e E is expressed in million electron volts. The sample must decay slow enough to permit the heat measurements to be made. Thus a practical limit for the half-life is -1 day. As an example of the application of this method, the disintegration rates of J32 (95 - 97), IH3 and Au198 (97) have been determined. A method related to the calorimetric standardization has been described by Loevinger (98) and Gray (99) in which the ionization produced by 3 particles is measured in a cavity type chamber. The disintegration rate can then be expressed in terms of the ionization produced, the average energy of the ) particles, and the average energy expended in the production of one ion pair. P32 has been standardized in this way. (.99) 2. Coincidence Counting. This j ethod has been described by DunwortA,(100) Barnothy and Forro (101) and others (102-103)

who give good treatments of the procedure and the errors involved. Two counters are arranged so that the first detects only one type radiation from the source (e.g. beta rays) and the second detects only radiation of a different kind (e.g. gamma rays) or of a different energy. If the decay scheme of the nuclide in question is known (there must be at least two different radiations involved in the decay) a relation may be derived betwreen the disintegration rate, the coincidence counting rate, and the single counting rates. For instance if the nuclide decays by P- emission followed by a Y-ray transition to thle ground state, the disintegration rate is given by dN _ r v(9) dt Ly where No and Ny are the single counting rates and Ni7 is the coincidence rate. The source must be fairly intense: -105 disintegrations/min for a long-lived activity. This and certain other conditions and corrections serve to detract from the usefulness of this method. However, it is capable of high accuracy (-1 per cent) for those cases in which the requirements can be fulfilled, e.g. Co 60. (104) 3. Particle Counting with KnowunCounin Efficeny. This method merits special attention because it has been used widely. L. a. Zumwalt (96, 105 - 106) has described the manner in which the disintegration rate is derived from the

-90observation of the counting rate in end-window GeigerMueller counters. The counting rate is expressed as (c/m) = (d/m)GffAfBfHfS (10) in which (c/m) = counts/min detected by the counter; (d/m) = the actual number of disintegrations/mnin in the sample; G = thie solid angle geometry of the source and the counter. fXiJ 3= the factor for the effect of the absorption of the beta particles by the window and the air between the source and the window; f, E= the factor for the effect of air in scattering beta particles into the counter; fS r= the factor for the increase in counting rate due to backscattering from the sample mounting; fH = the factor for the scattering effect of the source support structure and housing; Ef = the factor for the effect of the mass of the source tin causing both scattering and absorption of beta particles (selfasorption and scattering). Sometimes a sample is first standardized by another means, e.g. coincidence counting, and then counted in the

-91particular detection arrangement to be used. The counting efficiency is thus determined for a particular nuclide and detection arrangement. If this is not done each of the factors in equation (10) must be separately determined. In addition account must be taken of the decay scheme, i.e., contributions to the counting rate from`-rays, conversion electrons, etc. Thus an additional factor should be included in equation (10) expressing the probability that a count is registered once one of the radiations froim the source enters the sensitive volume of the counting tube. Of course the observed counting, rate imust be corrected for background and dead tiiile losses. ilany re irements hav,e been made in thlis basic method. (107 - 121) This technique has been employed in the determination of cro;:s sections (14,111,122 - 124), specific activities (125) etc. Ti-e accuracy of such determination is limited bys thie "absolate" countin' error of 3 - 20 per cent. Anoti'er exte1;nion of particle couantin g witll definied efficiency is to -lace thte sample writhinl- the detection c lluber.!With tie active species in te':aseous form Lil by (126), Eid.rnoff (127), Bernstein and f3allentine (128) have counted C04 and H3 by incorporating them into the countinrg gas. irofessor Ht. R Crane (129) has counted C14 as caroon black in an internal counter. The source may be deosited on an electroue, pl1ced in a vacuum and the

charge carried away measured. ( 130-132 ) Nearly 4I geometry is achieved in this case but intense sources are required. A screen sall counter is capable of 5 - 10 per cent accuracy wlen the activity is placed on the wall ofi the chalmber. (135) it h the sample on the bottom of an internal flow counter, the solid geometry factor alone is 2T, but absorption and scattering problems are still present. (136-141) Thie method of "4n counting" has recently received a great deal of attention. This is really a refinement of thie particle counting techniques discussed in the last three paragraphs. An advantage is gained by the elimination of the problems of scat tering and, to a large extent, those of absorption. The principle is that the sample is surrounded by the detector's sensitive volume such that one disintegration produces only one count regardless of the decay scheme, particle scattering or secondary radiation. The method is highly sensitive, so that low intensity samples may be counted, -10 disintegrations/min being the limit of detection (for counting volume of -250 cm3). Very recently scintillation techniques have been used to obtain 4nr geometry in which the source is either sandwiched between two crystals (134,142 ), or dissolved in a liquid scintillator. (143) Much more work has been done with solid samples placed inside a gas filled chamber. It is with such an arrangement that the present work has been don@e.

_.ror tional Countin The technique of absolute D-rav counting using an internal k4 proportional counter possesses several advantages over the other mrethods. As mentioned above the effect of [-ray scattering upon the counting rate is eliminated. If the source can be obtained in a solidfree state there will be no self-absorption correction. Absorption in the source mount may be determined, and usually amounts to a small (O - 5 per cent) correction. Especially important in the study of low yield nuclear reactions, such as the (d,a) reaction, is the detection sensitivity of -10 disintegrations/min which may be obtained. The probability that a count is registered once a 5-ray enters the sensitive volume is essentially unity (to within -1 per cent) if the electric field in the chamber and the proportional amplifier are properly adjusted. Only if there is electron capture branching, (144) or if there are delayed states involved in the transition which have lifetimes -omparable to the counter resolving time (~5)sec) will there be any major discrepancy between the counting rate and disintegration rate. In the past the usual procedure in applying 4Tr counting techniques to bombardment work for the determination of absolute disintegration rates has been to employ them to calibrate an end window Gei;ger-Mueller tube. The counting of the bombardment samples is then done with the calibrated counter. In this way the same corrections for back

-94scattering, geometry, etc. must be apiolied as in Zumwalt's technique. The innovation in the present,rork was the direct appilication of 4rr countin g to bombardment work in which the decay of samiiles coritaining several components (of differing half-life, P-ray energy, etc.) was followed with the Ig counter. Thus absolute decay curves were obtained, subject only to small corrections discussed below. It was felt that the 4TT methlod iwas exploited to better advantafe in this way, by the elimination of scattering and geometry corrections. /Bince tiis research w.s unidertaken the Berkeley group (145 - 146) has reported some absolute cross sections by following the decay of C91 and Na24 in the 4 r counter, but no chemical separations were performed on the target foils and only one component decays were encountered. A description of the various 4vr internal counters which have ai,eared in the literature has already been given on pages 35 and 36, to;etIler with the references. In addition to tlhose, the University of?-ichijan (147 - 148) and the, assachusetts Institute of Technology (149) groups have reported valua-)le information. T'e work of Pate and Yaffe (150 - 152) is particularly important in connection with the specific techniques of 4Tr 5-ray counting, The particular form used in this work has been discussed on pa~:es 36 - 42 and wlas pictured in figure 15 and 16. The following sections deal w^ith the detai]ls of the procedure tiat have ieen,worked out. first lthe.methods used to

-95fabricate thin plastic films for use as source mounts will be given. The measurement or the thickness of these films and the film absorption correction will then be treated. The self absorption problem and the effect of a conducting coating of the source mount will be considered. A discussion of the amplifier gain and discrimination levels at various anode potentials will follow. The experimental estimation of the dead time of the counter, and finally the application of these techniques to tihe counting of National Bureau of Standards samples will be presented. i. Plastic Film Preparation, As described on page 36, thle activity was placed on a thin plastic film covering a 1/2 inch hole in the center of the aluminum sample plate. The active solution was pipetted onto the film and dried with the aid of an infra red lamp and hot air current from a hiair drier. The deposit was usually covered with another layer of film to prevent contamination of the counter. Several commercially available plastics were tried, including 1/4 mil Teflon, 1/4 mil rubber hydrochloride, and 1/4 mil aluminum-coated?iylar. These proved unsatisfactory for absolute counting because too large a fraction ofeRak beta particles are absorbed in this thickness of film. Fabrication of films as thin as 20,g/cm2 (several hundred fold thinner than the commercial film) was carried

out in the laboratory. Several teehniques have been worked out and reported in the literature. (149 - 153) Ki~specially helpful wIs tAle report of Fry and Overman. (149) Their method of floating a solution of the plastic on water and picking up the film thus formed on a wire frame was adopted in this work. The apparatus, sii;ilar to thaet used by professor L. L=. -iedenbeck and coworkers is illustrated in figure 25. Best results are obtained if cold fresh tp w ater is used for casting the films. Tile surface is first cl-aned -y drawing a brass bar across it. Nlext the wire frame is inmmlersed beneath the surface and suspended from the sh`laft of a 3 rpm synchronous motor. 7our drops of the pl-stic solution are pipetted onto the wiater and the solvent allo;wed to evaporate. The frame is then pulled up out of the waler vertically by means of t.lhe motor, picking up the film as shown in figure 25. The double layer thlus formed will Ihenceforth oe referred to as a sinle lacLrer. The film is transferred from the frame to a 4rr counting plate by lowering the film onto the plate. The two are forced togethler w*sith an air jet by blowing through a I lass tube, as in figure 26. The film must be wet for good adhesion. Figure 27 sh1ows howr the film is cut around the edges of the sample plate. A wire dipped in solvent, mnakes an excellent "knife ".

i~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I-i m 1'I (~~~~~ Fig, 25 Appara~us:.r Making Thin Plastic Films, ~~~~~~~~~~~ Fg~ 2~ A/o tingthe Film F g 7Ti~ig on a4 pi Countng Plae~ i ln

several plastics may be used to make th-e films. These lhave been listed by Slatis (153) along with references on their fabrication into thin films. New techniques for use with VYNS resin (a polyvinylchlorideacetate copolymer obtainable from Canadian Resins and Chemicals Limited, X ontreal, QLuebec) have been developed by Pate and Yaffe. (147) Formvar E (a polyvinyl acetal obtainable from Shawinigan Products, New York) may be made into films by i'loating a 1 per cent solution in ethylene dichloride on water. Although these films are very strong liechanically, acidic solution attacks them. For this reason Zapon films (pyroxylin lacquer obtainable from Arthur S. LaPine and Company, Chicago, Illinois) were used for routine absolute counting. A solution of Zapon, Zapon thinner, and amyl acetate in a volume ratio of either 1:1:1 (33 per cent Zapon) or 2:1:1 (50 per cent Zapon) -ras used in this case. The thickness of t.he film can be varied by using a solution of different concentration or changing the number of drops to form a film. A more practical expedient was to laminate several layers together to make thicker films. One layer was added at a time. Care must be exercised to prevent air pockets from forming between successive layers especially in the center. 2. Film Thickness Pieasurement. Pate and Yaffe ( 147 ) hNave summarized several ways of determining the film

3'9 thickness. Optical methods (33,154-156) both by interference of reflected light and by light absorption, were the subject of a few trials. Since the optical equipment was not readily available in this laboratory, this method was not exploited. The obvious procedure of weighing a given area of a uniform film was found difficult to apply due to the very small mass (as low as 20,4g) of the films. The absorption of weak f3 particles, such as from Ni63, described by Pate and Yaffe (147) is the easiest and quickest of the techniques to apply and was finally adopted for routine use. The surface density of a few films was measured graviimetrically using an Ainsworth type FDJ microbalance. The films were mounted on a weighed sample plate and the weig ght of the film found by difference. At first great difficulty was encountered establishing rest points to better than + 10,-g. Finally a procedure was evolved, based upon the work of ilillig (157) and Hull (158), wherein a weighing can be made in a period of about 40 minutes to a precision of + 3,Ag. Results of these determinations are shown in Table IV. Films belonging to the same batch number were made under identical conditions. The Ni63 D-ray transmissions of the weighed films are also presented in the table. Films 8-1 show the effect of' improved technique. Note tMhat thle surface densities of film 8-13 may be averaged to give 1S.0 + 1.5/jg/ cm2, and films 14-18 show that the surface density per layer is

19.4 + 1.3/A g/cm 2. Thus individual films may be made of the swale thickness to within 7 - C per cent as determined gravimetrically. It will be shown in the next section thlat even an 8 per cent difference introduces no appreciable error in the film absorption correction. 63 In the application of the i63 P-ray absorption technique the output of the lower half of the Borkowski 4f chamber was connected to the proportional counter rather than employing a "conventional hemispherical 2rr-proportional counter," as rate and Yaffe ( d!7) did. The i63, deposited on a one inch aluminum disk and covered with two layers of Zapon films, was placed face down on the film to be measured, and the 2rr counting rate determined. Since the beta transmission is determined over a small area in the center of the film it is not subject to peripheral irregularities and changes in support weight, and should be a much better estimate of the reproducibility of films from the same batch. The results of the Ni63 absorption measurements are presented in figure 2-. The points shown as squares were obtained from filims w —ose thticzinesses had been determ.ined gravimetrically. For instance, the thickness of film numbers S-13 were averaged to give an x-coordinate of 18.0 + 1. 5,qg/cm2 and their transmissions were averaged to give a y-coordinate of 0.830 + 0.004. The weights of films 14, 15, and 16 were the averages of two sets of weighings which took place nine months apart. The uncertainties

101Table IV. Surface Density Data for Films made from 33 per cent Zapon Solutioin Number Surface Ni63 Film Batch of JTeifht Area Density O-ray Do. Ho. Layers (/cm2) transmission i ayers ()2) (g/(cm) fractional) 1 1 1 32 5.06 6.3 2 1 1 20 5.006 4.0 3 1 1 25 5.06 4.9 4 2 1 42-72a 5.06 J.3-14.2a 5 2 1 79-89a 5.06 15.6-17.6a 6 2 1 69-90 5.06 3,3.6-17.8a 7 2 1 52-97a 5.06 10*3-19*2a 8 3 1 878 45.60 19.3 0. 32 9 3 1 661 45.60 14.5 0.836 10 3 1 867 45.60 19.0 0.827 11 3 1 772 45.60 16.9 0.632 12 3 1 $69 45.60 19.1 0*821 13 3 1 872 45.60 19.1 0*831 14 4 1 1029-1028a 45.60 22.6-22.6a 0.843+0.009 15 4 2 164S-1646a 45.60 36.2-36.2a 0. 25+0.022 16 4 3 2670-2664a 45.60 58.6-5S.5a 0.792+0.020 17 4 4 3366a 45.60 72.5a 0.760+0; 010 18 4 4 4251a 45.60 93.2a 0.7540.012 a The 5w5eight of the film plus aluminum. plate Lwas redetermlined at a later date, giving the second number.

-10-2 indicated in the y-coordinate represent the average deviation of two determinations. The points shown as circles in figure 28 correspond to the upper x-axis which displays the number of layers of films made by lamination using 50 per cent (2:1:1 volume ratio) Zapon solution. The uncertainties indicated in the y-coordinate reyresent the average deviation of at least two different films. Wiote that the films made from equal numbers of layers give the same Ni transmission to within less than 2 per cent (average deviation). The adjustment of the scales of the x-axes is a more or less arbitrary normalization of the upper, relative scale to the lower, absolute scale. Figure 28 may thus be used to obtain the approximate thickness from a measurement of the Ni63 R-transmission. 3. Film Absorption Correction. A certain fraction of tighe beta particles emitted from the source will be absorbed in the film between the source and the bottom half of the chamber. <.hether or not this fraction is significant depends on the backing thickness and the shape and endpoint of the beta spectrum. T'le'correction can be assessed in several different ways. Seliger and coworkers (17, 159) at the National Bureau of Standards have treated this subject by a consideration of the number of particles detected in the top and bottom halves of the 4rr chamber both separately and together. From the resulting analysis they have derived an expression

1.0 TRANSMISSION OF Ni""3 RADIATION (LOWER HAlF OF CY'INLRICAL 4A r PROPORTIO"NAL COUNTER) o.9 zo | r<h0 o REFERRED TO UPPER AXIS o0 REFERRED TO LOWER AXIS F<: —7-.6 1 2 3 4 5 6 7 8 LAYERS OF FILM I _ I I l l l l l l l I I I 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 FILM THICKNESS (/Lg/cm2) Fig. 28. Transmission of Ii-'3,-ray radiation in thin Zapon filIms Thue upper x-axis gi-es t.e nrumiber of la-rers of film ard the.lovrer x-axis s;io'.:s tl:e sirf ace de is ity as determined -ravimetrica! ly.

-104for the fractional absorption in the film which is related to the absolute disinterration rate. The analytical treatment is complicated by t]he many energydependent scattering and absorption considerations. Thus they had to make certain assumptions whlich cast doubt as to the validity of their treatment. Hawkins, et al. (52) at Chalk liver made the correction by placing a film on top of the source in such a matter as to sandwich the source between two identical films. Their films consisted of 100 /g/cm2 of plastic coated with 50Ag/cm2 of gold. No account was taken of tile radiation scattered from the walls of the chambers. The reduction in countingcf rate which occurs wiien the second film is placed on top of the source was considered to be the correction. This is equivalent to extrapolating a plot of the counting rate vs film thickness to zero. Cohen (45) did essentially the same thing as Hawkins except that he used muich thicker films. A 7.76 mg/cm2 thickness of aluminum was placed over the source (wh-ich itself was approximately 6 mg/cm2 thick) and the counting rate was determined. Nlext another sheet of aluminum was placed under the source and the counting rate redetermined. Assuming linearity of the resulting absorption curve the the appropriate correction was made. Smith (46) showed that this tsandwrichT' technique is unxreliarle. He states that the true correction is closer

to that obtained from a linear extrapolation of a plot of activity vs thickness of the source backing. Increasing numbers of aluminum films, each 260 Jg/cm2 thick, were used to vary the thickness of the source backing. He also gives a more complex mathematical treatment of the "sandwich" technique, which if applied should give results in agreement with the linear extrapolation technique. In determining half lives by measuring specific activity, lMacGregor (144) and Sawyer (145) corrected for film absorption at the same time as self absorption. They made up a series of samples varying both the source and surface density of the film. The specific activity was determined for each of these samples and was plotted vs the total thickness (sample plus backing). A linear extrapolation was then made to zero thickness to determine the true specific activity. iMacGreort's film varied from 15 to o5,eg/cm2a Houtermans and coworkers (43 - 44) followed a similar procedure. The source deposited on a film 30- 402 g/cm trick, was covered with aluminum foils of varying thicknesses. The resulting absorption curve was extrapolated to zero absorber. The thinnest aluminum foil they used was 240 g/cm2, however, and whether the slope of the curve remains unchanged, below this value is open to question. A consideration of tliese different methods led to the conclusion that the best way to make the source mounting correction is by the absorption curve technique. If the

-106thickness increments are of the order of 15 g/cm2 the extrapolation is only over that thickness which is the range of -2 kev electrons. (33 ) During the course of this work a paper appeared by Pate and Yaffe (147) in which they arrive at essentially the same conclusion. They present data demonstrating the validity of the absorption curve method and the erroneous results obtained when either the sandwich technique or the method of Seliger is used. The procedure used in this, work was the following. A series of sample plates was made up bearing different numbers of layers of Zapon films. Equal aliquots of the radioactive solution were pipetted onto each. After evaporation of the solution all the sardples except one were covered with an additional Zapon layer to prevent loss of activity. The one uncovered sample had only one layer of source backing. The samples were then counted several times during their decay in order to detect any changes in the absorption curve as different components decay out of a complex mixture. Decay corrections were applied (using the effective half-life during the period of measurement) to normalize each absorption curve to one particular time. Figure 29 shows an absorption curve obtained in this laboratory from P32 previously standardized at the National Bureau of Standards. Here, as in the rest of this work, the number of layers of Zapon film is plotted along the x-axis. This is justified by the fact that successive

-107layers of films can be made equal in thickness to within ~$ jper cent as discussed previously. Since the curves so obtained are very nearly flat, a fairly large error - certainly more than 8 per cent - in the value of the thickness can be tolerated. The points represent averages of two determinations taken four weeks apart. As can be seen from the figure there is no evidence of a significant slope. It will be noted that an aliquoting error is inherent in each point of the absorption curve. A separate experiment was done to evaluate the magnitude of this error. Aliquots of a Co60 solution were taken with micro pipettes which had been coated with dimethyl-dichloro silane (variously called Silgon, Drifilm, Desicote, etc.). ( 160-161 ) (These specially treated pipettes were used throughout this research.) The pipettes were rinsed with water after each filling and the FILM ABSORPTION CURVE OF NBS p32 Z | (CYLINDRICAL 4-rr PROPORTIONAL COUNTER): 3300 LLi a 3200_ o_ n i 0 0 2 3100 LUNCOVERED SAMPLE FOR DECAY AND 3100 SAAMPLE DISCRIMINATION CURVES.........1 1 _. l 3 4. I I! 2 3 4 5 6 7 8 9 LAYERS OF FILM Fig, 29. Film absorption curve of 32 standardized at the National Bureau of Standards.

10o rinse was added to the sample. The counting was done with the scintillation well counter. The results are given in Table V. The first part of the table shows that essentially all of the active solution is removed in the initial transfer. However, for the rest of the work, one or two rinses were used to be on the safe side. The second part of the table shows the reproducibility to be better than + 2 per cent, even for multiple fillings in which there was no rinse until 60 after all of the Co solution had been transferred. Table V. Aliquot Reproducibility Pipette Vol. (Y1) Vol. taken (r1l No. Rinses Activity (c/m) 100 100 2 1202 100 100 1 1201 100 100 0 1164 250 250 2 2490 250 250 1 2532 250 250 0 2531 500 500 2 4557 500 500 1 4762 500 500 0 4806 100 100, 1 722 100 100 1 713 100 100 1 688 100. 200 1 1414

-1094i. elf Absorption. Chemical procedures in which the element in question is separated in a carrier free state ipossess obvious advantages with respect to the problem of absorption of very weak X particles in the sample itself, This is a goal which is advocated for cross section imeasurements in bombardment work in conjunction with L4r P-ray counting. Even with carrier free separations there may be small amounts of foreign material, such as sodium chloride, picked up during the chemical processing. A quick check was made to find out roughly how much matter can be tolerated without absorbing an appreciable number of D-particles. Cobalt-60 was chosen for this purpose as its 0.32 iLIev 3 ray represents a lower limit to the e energies encountered in the present work. Equal aliquots 60 of Co were pipetted onto thin Zapon films. Varying amounts of a very dilute sodium chloride solution were,ipetted on top of the activity. No particular care was taken to assure an even deposit, although insulin was tried without success. After evaporation of the solution, the deposits were covered with another Zapon film and counted. The resulting self absorption curve is depicted in: igure 30. Tile errors indicated are the statistical errors of counting (standard -evit-iono. From this crude experiment it was felt that a lower tolerance limit may be set at ~l/2,g from an inspection of the curve.

-110oThis amount of inert material can be observed visually. Therefore if a visible deposit is present on a sample plate for absolute counting, one must suspect a significant self absorption correction for weak i-emitters. 5. Effect of a Conductin Coa If the plastic film used for the source mount is not conducting, there will be a small field-free space in the center of the chflalmber. Thus some very weak particles may expend their energ=y in this region without producing a pulse. Thnere are two ways of taking care of the situation: the film may be made conducting, or the size of the hole in the aluminuml sample plate may be made small enough so that the effect is negli;;ibble. SELF ABSORPTION CURVE FOR Co60 (CYLINDRICAL 4wr PROPORTIONAL COUNTER) 1900 m 1800 c 1700 CLJ F0 co 1600 o 1500 4 8 12 16 20 24 28 30 34 jg OF NaCI Fig. 3G. efi absorption curve of Co.

-111Hawkins et al. (52) state that a minimum of 25/Ag/cm2 of gold must be evaporated on thin plastic films covering a 2 inch diameter hole. M:ann and Seliger (159) at the National Bureau of Standards found that a low-field correction is necessary for Co60 but not for 32 using -0.8 inch diameter hole. However, the Bureau's standardizations are now all made with 15/g/cm2 of gold evaporated on the plastic. (159) lith a 1 1/2 inch opening, Smith (46) used 260 g/cm2 aluminum foils for strong beta rays, and thin aluminized plastic films for weak emitters. 1ieyer-Schltzmeister and Vincent (44) employed 5 - 15z/ag/cm2 of copper evaporated onto plastic films over a 0.4 inch diameter aperture. (Their purpose in using the conducting coating was other than the elimination of the low-field effect.) Pate and Yaffe ( 15T recently made a study of the minimum thickness of gold necessary for correct operation. They conclude that for Ni63 and 32 alike at least 2,Ug/cm2 gold is needed for their 1 and 2 inch apertures. Borkowski (48) states that as long as the opening in the sample plate is less than 1/2 inch no conducting coating is necessary. Since the sample plates used in this work had center holes 1/2 inch in diameter, it was thought that no conducting coating would be necessary. This point was checked in two ways. First, samples were counted that were obtained from the National Bureau of Standards where the In standardization is done with the conducting film, and excellent agreement was obtained (better than 1 per cent -

-l12the details are given later). Secondly, voltage plateau 60 curves were taken of Co (the weakest V emitter encountered in the present work) mounted both on two layers of Zapon and one layer of Lapon coated with an evaporated layer of gold. Pate and Yaffe ( 51 ) considered the existence of a voltage plateau as a criterion of the proper functioning of their 4rr counter. Although it was felt that discrimination plateaus (see next section) constitute a better measure, the data presented in figure 31 demonstrates that a voltage plateau is obtained with the cylindrical chambers for Co60 either with or without the conducting film of gold. (Pate and Yaffe's data do not extend' to anode potentials greater than 3400 volts and their sample plates had either 1 or 2 inch apertures. These differences may explain why they do not obtain voltage plateaus without the conducting film whereas such plateaus are obtained in the present work. ) Figure 31 also includes a curve of a Sr-Y90 sample for comparison, mounted on 1/4 mil Teflon, which was used as a standard throughout this work to check on the dayto-day performance of the counter-' The three curves are not normalized to the saue counting rate on the plateau because of the confusion of the individual points that would result. In the later experiments (Titanium bombardments) gold was evaporated onto Zapon films just to be on the safe side. The optimum arrangement, for the various

r2poo 12900 PLATEAU CURVES x (CYLINDRICAL 47 PROPORTIONAL COUNTER) I 1,000 x IIoX X- x -- X — X 10p000 x X SrY (TEFLON) - JULY 1955 9, 000 Co6 (ZAPON-GOLD)-JULY 1955 LU 60 --- | Co (ZAPON) - JULY 1955 E 8,000, 7,000 0 o 6,000 5,000 4,000 3,000 - 3200 3400 3600 3800 4000 42004400 3200 3400 3600 3800 4000 4200 4400 ANODE POTENTIAL (VOLTS) iMig. 31. Volt'a,&e plateau cuzrves.?ith various samples.

-114layers is depicted in figure 32 as determined by trial and error. The gold is first evaporated onto one film covering only the central part of the sample plate. (Pate found it makes no difference to which side thie gold is applied. ) Additional layers of Zapon and the activity are then applied to the opposite side. 6. Effect of Gain, Trigger and Voltage Settings. In both of the instruments used (see the section on 4rr Counters, page 35) the amplifier gain is variable as is the trigger (discrimination) level. The latter defines the minimum pulse height necessary to trigger the scaler and hence permits one to eliminate electronic noise. Voltage plateau curves of $r-Y90 were taken covering the full ranfge of gain and trigger settings. For routine counting a point was chosen on the best plateau. For absolute Pq-ray counting Seliger (47) has shom -that it is necessary to take discrimination curves, i.e., a plot of counting rate vs trirgger setting for a given gain and voltage, to determine the position of a discrimination plateau. This is necessary for each "3 emitter measured because each one may yield characteristic curves which look quite different depending on the shape and end-point of the beta spectrum, complications from conversion electrons, etc. This necessitates taking discrimination curves at intervals during the decay of samples containing more than one nuclide, because as one component decays away the radiations coming from the sample change. Mlloreover

-115electronic changes Imay shift the plateau or ot-,erwise alter the settings for "'4bsolute" counting over a period of several months. These curves may be expected to exhibit a linear,ortion, or plateau, which may be extrapolated to zero discri mrination level to arrive at the true counting rate. The idea behind this is that in the region of the F-ray distribution in the neighborhood of zero energy, the noise level will cover up the true counts as is evidenced by a sharp rise in the counting rate near zero discrimination level. This noise is cut out by increasing the discrimination setting to a point above the noise level. But in so doing some of the weakest ~ particles will also have been lost. It is reasonable to suppose ti.at some relation exists between the pulse height and the probability of emission of beta particles with that height. It follows theat the discrimination curve will assume a definite form SAMPLE FOR 4w COUNTER ACTIVE SOURCE LAYERS OF XZAPON ___________________L "'` ALUM INUM ~~\ mSAMPLE PLATE I LAYER OF ZAPON GOLD COATING Fig. 32. Preparation of thle 4r counting plate showing the arrange':ent of th:e various layers.

DISCRIMINATION CURVES OF NBS p32 (CYLINDRICAL 47r PROPORTIONAL COUNTER) 33 -+ — GAIN 100 3900 VOLTS GAIN 75 t32.~~~, -x- GAIN 50 31- - 33 4000 VOLTS Z32 0 33::~O 3 4100 VOLTS x + + 31 I LIMITS OF COUNTING \ + 301 ERROR (STANDARD DEVIATION) 10 9 8 7 6 5 4 3 2 1 DISCRIM INATOR SETTING i -i,. 33. 9iscriflitnation curves of 32 sotandardized by the National:.-ureau of Standards. T:he limits of the counting error soown are t;wice the sta:-.dard deviation of a sinle point.

in this low energy region, and it may be assunmed to be linear to a first approximation. In practice the flat portion of the discrimination curves showed no evidence of a slope for the F emitters studied covering an energy range of 0.36 - 4.2 1Wev. An example of this is given in figure 33 which shows the curves for NBS F32 The sample selected for those measuremilents was the one mounted on 4 layers and covered with one layer of iapon (see figure 29). The experimental points have been corrected for decay. The gain numbers merely indicate the dial setting on the Nuclear-Chicago i iodel 162 scaler. The x-axis is also an arbitrary scale showing the actual dial settings the discrimination level increases to the riLat. Note tihat the disintegration rate is the same for the three voltages and the three amplifier gains shown, within the error due to counting statistics of about 1 per cent for the individual points. The limits of this error (twice the standard deviation) are indicated on the figure which applies to all of the points - this avoids the confusion which would result if tthe error of each point were plotted. This error could, of course, be reduced by increasing the number of disintegrations counted for an individual counting rate measurement, but limitations of time available for these curves dictated the duration of each nmeasurement, 7. Dead Time Correction. Another possible source of error arises from the fact that an event which produces a

pulse will render the. counter insensitive for a finite period called the dead time. In proportional counters this dead time is relatively short - typically of the order of 5/d sec - wh'ich mreans that it is possible to record very high counting rates without significant coincidence losses. Since the present work involves following the decay of samples over large variations in counting rate, this small dead tire is of particular advantage The determination of the dead time ahas been adequately treated in the literature. (162-166 ) The available methods of measurement group themselves into three classifications. 2irst, there are the paired sample techniques developed by KohrmIan (163-164 ) whaich involve counting two samples of approximately the same strength separately and then aIain simultaneously. The sum of the two separate counting rates will in general be greater than the counting rate taken together, and from this is deduced the fractional coincidence loss. Secondly, one may measure the minimum time between two discernable pulses at the various stages of the scaling circuit wit>i an oscilloscope and take the longest dead time observed to determine the coincidence correction. Thirdly, the correction can be found by tahe measurement of sources of known relative intensities, in particular the observation of the decay of a source which is decaying with a moderate half-life. Here the decay is followred for several half

-119lives to a low counting rate at which there is assumed to be no loss of counts. The theoretical curve is drawn th.rough these decay points using the known half-life and extrapolated to the high counting rates. The deviation of the observed points from the theoretical curve is interpreted as the correction term. This strictly empirical method involves no assumptions other than an accurate knowledge of the hialf-life of the activity observed,'and quite closely parallels the actual operating conditions of the counter,'Then the paired sample technique was tried in this work it invariably resulted in nieaningless negative numb ers apearing in the formula for thie dead time. An analysis of the propagation of errors indicated that for such a short dead time as -5psec, extremely accurate counting rates must be determined in order to result in an answer to which some significance can be attached. This possibly affords an explanation of the anomalous results, because the rates were determined only to within -0.5 per cent in these experiments. The 4Tr proportional counter was tested for coincidence losses'at high counting; rates by following the decay of the 5r minute In1lb. This isotope was produced by bombarding 5 mil indium foil with neutrons which are available in abundance just outside the cyclotron vacuum tank near the window box B,f igure 1. They are produced from (d,n) reactions of stray deuterons hitting the deflector,

-120paddle probe, etc. By bombarding for only about 5 minutes no observable amount of 50-day Inll4 is produced, while an adequate amount of the 54-minute Inll6 results from the 145 barn thermal neutron cross section of Inll5. The 4.5-hour Inll5m is produced only by fast neutrons, (170-172) so paraffin was placed around the indium foil to moderate thne neutrons. Actually, the first bombardment of indium foil resulted in an activity of the order of 5 hours and a long-lived component in addition to the 54-minute period. By decreasing the bombardment time from 45 to 5 minutes and increasing the thickness of paraffin surrounding the target foil from 1 to 6 inches, these two extraneous activities were not observed. (They were less by factors of at least 700 and 5,000 respectively for the 5-hour and the long-lived components. ) The results of the second bombardment are depicted in figure 34. The experimental points have been corrected for background. A single activity due to the 54-minute Inll6 is seen to remain after the short-lived impurities have died out. The decay curve is strai;ght for a factor of 104 (13 half-lives) to -10 counts/rmin (10 times less than background. ) The observed half-life agrees well with the reported values of 53.93 + 0.03 minutes (73 ) and 54.31 + 0.07 minutes. (174 The results of a least squares analysis of the decay points from t = 4.5 hours on, yields 54.26 as the half-life. (Above 4.5 hours the

-121DECAY OF IN"6 51C>5F4 (CYLINDRICAL 4.1 PROPORTIONAL COUNTER) 10 I0 - Imi- I Z t ~53.93 MIN|. 0 10 0 I 2 3 4 5 6 7 8 9 10 11 TIME (HOURS) i' 3.1 Decay cuXv of Oin' fncI i-f i,e in etermninnirg the cdead t1?iC; o. te'tO 4rW OpPttriOfall CO..L..i'r

-122points start to deviate from linearity by -0.1 per cent.) The first five points indicate the presence of an activity with a half-life of the order of three minutes. This dies down to less than one per cent of the 54 minutes activity by the time the latter is 1.3x105 cpm. These data were treated according to the method of T. F. Kohnan as outlined by Calvin, et al. (165) The provisional corrections, C', were calculated from Ne'? t -R-b Ct = (11) where R is the observed counting rate, b is the background rate, N0 is the true rate when t = 0, and N, is a tentative 0 _ value of NO. The value of the half-life was taken as 53.93 minutes for these calculations. Figure 35 shows a plot of C' vs R, and the solid line drawn through the points by means of a least squaresanalysis. Note that assuming the formulas N = R +T'R2, (12) where N is the true counting rate and _r the dead time, it follows that C = FR. (13) This is the justification for drawing a straight line through the points with a slope equal to _F. The

COINCIDENCE CORRECTION (0 YINDI CAL 47PROPORTIO NAL COUNTEI) ~0 ir.04 0 C).00 0 0 0 ~02 0 LLJ 0 1 0 0 Q cc - - 0 O 0 C 000 p 0 0CU R M0 V I. 702~ 07000 40poo 60,OOO Opoo 100 Fig - 35. co-COUNTS PER TE NU. aqOo age 5uCoincidenc (dead time) cOn SER E b ase ci up on th e decay of 1n 11 Th e s oi j n e ls th e c e a t ddtt uProvisionai cour dotted line is the t,~~~esldln ste pootoa

-124intercept of the curve at R = 0 is taken to be NO, from which the correct values of C can be calculated by the method outlined by Calvin. The curve of C vs 2. is also shown in figure 35. The results of the second bombardment of indium gives 7.1,/sec as the dead time. A third indium bombardment was followed from 1.2x106 counts/min to below background and the dead time from this was calculated to be 8.6/sec. These two values agreed well within the standard error of each determination. Tile average, 7.97/sec (corresponding to a correction of 1.2 per cent/thousand counts/min), was used to obtain a working curve of tlhe percentage correction vs the gross counting rate similar to the solid line in fig;ure 35. 8. Counting of National Bureau of Standards Sample s 32 A solution of p with 0.001 M H 3PO4 carrier was obtained from the National Bureau of Standards. It had been standardized in a 2Tr ionization chamber previously calibrated by 4n proportional counting. (49p) Micropipettos, calibrated with mercury and coated with Silgon, were used to transfer aliquots of the solution to a series of sample plates bearing varying numbers of non-conducting Zapon films (made from 33 per cent Zapon solution). Water from one rinse was added to the sample plate. Drying was accomplished in an atmosphere of ammonia to prevent the Zapon from being attacked. All the samples but one (on one layer) were

-125covered with an additional layer of Zapon. The samples were counted to give the film absorption curve already discussed (figsure 29). The discrimination curves (figure 33) taken with the sample which was deposited on four layers of Zapon have also been treated above. The decay was followTed on the four layer sample at the gain and trigger settings determined from the discrimination curves. The decay curve is presented in figure 36. The solid line was calculated assuming 1.4 per cent of P33 contaminant at the time of the standardization. Values of 1.4 and 1.9 per cent (175-176 ) have been reported for the percentage of P33 in pile-produced P32 and Jlesterrnark (177 ) reports this number varies with the amount of fast neutrons contaminating the thermal neutron flux. The present choice of 1.4 per cent was more or less arbitrary, and a variation in this number by a factor of two does not make a significant ci Lange in the following calculation. The half-lives used were 14.30 days for F32 ( 95 178 ) and 24.4 days for P33 (179 ) which were deemned to be tohe best values available in the literature. The calculated curve fits the experimental points well, and the graphical analysis leads to 159x103 disintegrations/sec/ml as the initial activity normalized to the average of the points of the film absorption curve. If the effective half-life is taken as 14.4 days as recomr-alended by Seliger, (49) the initial activity turns out to be 160xlO3 disintegrations/sec/ml. wither of these

-126-. values agree well wiith tihe Bureatus value of 160x103 * 2 suer cent disintegrations/sec/lmi

-127DECAY OF NBS P32 104 (CYLINDRICAL 47 PROPORTIONAL COUNTER) LL Fz o CALCULATED CURVE 103 L I I I I I 50 60 70 80 90 TIMF (DAYS AFTER STANDARDIZATION) ~F i 36. J:ecay curve of 32 sta;dArdized at the National iure'Uu oi.ttrlin. -r. 1e calculated curv e is based uton 1.4 per cent of P- present in the oriinal sample.

-12 8 - EXHE 2'ilJTAL RESULTS In the foregoing sections the discussion has been of a rather general nature describing the various pieces of apparatus and the techniques and procedure used in the measurement of absolute cross sections. In the material to follow the ap;plication of these experimental techniques to the specific yield-determining bombardments will be'treated. First of all the equations will be derived which are used to transform the experimental data into cross sections. A short discussion of the steps followed in the course of a bombardment will then be given. Next the data from the four different elements studied will be annotated and the values of the cross sections so obtained will be discussed in the light of the experimental errors associated with each of the component factors entering into the calculation. Finally a few additional considerations in the measurement of reaction yields will be treated. A. ierivation of Equations The cross section equations will be treated first, followed by the equations used to calculate the chemical yield by the technique of adding a radioactive tracer. Th7+e symbols used are listed at the end of this section. Cross Section The cross section of a nuclear reaction is defined as the probability that a given nuclear reaction will occur when a nucleus is exposed to a beam or flux of bombarding

-129particles. The term cross section originates from the simple picture that the reaction probability is proportional to tlhe area which the target nucleus presents to the incident beam. One assutmes that the nucleus is a small sphere suspended in a medium composed mainly of empty space. The cross section, _, multiplied by the number of nuclei per square centimeter of target, n, gives the ratio of the area of the target nuclei to the total area (including both the empty space and the target nuclei.) if I is the nunmber of bombarding Particles per second, the total number of reactions NT which occur will be given by = a InT, (14) where T is the duilration of the bombardment in seconds. Note that a may be tholught of as an experimentally determined proportionality constant with the dimensions of cm2. In order to apply equation (14) to the calculation of cross sections from experimental data it must be modified somewhat. In the first place account m-st be taken of the fact that bthe radioactive product decays at the same tiime that it is being produced. Thus dN - n - A (15) where X is the decay constant of the product nuclei. If t:ie beam does not change in intensity during the course of the bombardment, i.e. if I remains constant, equation (15)

-130may be integrated to give the number of nuclei present at any time, t, after the beginning of the bombardment: N ((1-et). (16) It is the total number of product nuclei formed NT' which is needed to substitute into equation (14). If No is the number of nuclei at the end of a bombardment of duration T seconds, N0 -o I(l-e-T) (17) This equation rmay be solved immediately for c, or combined with equation (14) to obtain N \T N ( 18) If the cyclotron beam drifts appreciably inintensity during the course of the bombardrrent, equation (17) may be modified furt.ler. As pointed out in the Beam Integrator Section, page 18, the intefrator output can be fed into a traffic counter which;,rints out the number of counts collected at equal intervals. Since equation (17) expresses the nu mber of nuclei at the end of bombardlm-ent for I constant, it may be applied to each interval of duration Ati seconds during which the beam current is assumned to be practically constant. Then

-131Ml Iicti -Xat. itt I n -f Xt. -At. NO -.-( l-e 1)e + (1-e J)e -n'f(Ii.) _...,.1..............(19) where tihe function f (I.) is defined as hai ot 3a ihie susmmation is carried out over all of the eqiual intervals, an-id ti is the time from the end of the ith interval to the end of bombardment. The term involving i is added separately because the traffic counter record cannot in general be divided into an irntegral number of equal intervals, i.e. Ati /atj, so At. is just the length of the last interval. Thus equation (19) merely adds up the riumlber of nuclei formed at the end of each interval with allowance for decay from the end of that interval to the end of bombardment. The result of rearranging equation (19) and combining with equation (14) is NoAIT NT NI T. (21) f(Iij) This reduces to equation (l ) for thle case in which I, I;= I. Eouation (21) will be used later to figure the

-132total number of nuclei of one nuclide formed relative to some other nuclide of a different half life, which in turn is needed to calculate the chemical yield by the tracer tec hnique. In order to relate 1i (or I) to the number of counts collected on the beam interrrator, the calibration curve, fig-ure 21, is consulted. The slope of this curve, K, gives the number of counts per minute equivalent to one microampe-re. For deuterons (one chlarie per particle), B. -6 I. = 60x1 -'13 (22) Kti 1.60203xlQ wh:ere Bi is the number of counts recorded by the beam integrator in the time interval &ti. The numrer of target atoms ier square centitrieter n, is obtained from the weight of the target w, the target area A, the atomic weight T`, and the (fract-ional) isotopic abundance a, as follows: wa. 6.0228xlO23 n =.103. (23) The obser-,ved absolute counting rate extrapolated to the end of boirbardment, Co, nay be converted to N0 as follows: C~ ~~~o L2.~ ~(24) where Y is the (fractional) chemical yield.

-133,Finally the cross section is derived by combining equations (14), (21), (22), (23) and (24). Thus 0. 7.3887x10'3 (25) waY'f(Bij) where - e'lti ml- i't i l-e J J i =1j K.f(Iij ) 2.667x10'9. (26) In some cases tle correction for non-cornstant beam was not made, in which case eq&uation (25) reduces to C oAMKT (7 CS 8-e r3 *7.. 387x10-38 (27) waYB ( l-e- ) Chemical Yield For thlose cases in vzTic the cheiiical yield is determined,yy adding a know.m amount of radioactive isotopic tracer, the chemical yield, Y, is the ratio of thde amount of recovered tracer to.he naount added:

-134NT2 (recovered) Y t. (28) NT2(added) The primes refer to the tiracer, and the subscript 2 refers to a particular nuclide. Sometimes the nuclide which is added as tracer will not be produced in the bombardment in sufficient quantities to be observed, but in most instances there will be a contribution to the total activity of this nuclide from the (d,ao) product of the nuclear reaction. Then it is necessary to perform a separate experiment in which the bombarded tar-et is worked up without the addition of the isotopic tracer. Thie number of atoms of the tracer isotope formed bey nuclear reaction is then rep ated to the nuimblIer of some otlhr isotopic species also produced in the reaction. Thus the ratio of the two activities, NT2 NT1 is determined,;-here subscript 1 refers to t e other activity chosen as a reference. rote that this experiment a-,iounts to,-leasuring tile relative cross sections for formation of th}e two activities. Thus 2l a

-135i1th a knowledge of. the bombardments can be carried out to determine the absolute cross section. Thlen N,2(added) atoms of tracer are added to a solution of the target, and the chemical yield becomes INT2 - NT2 Y ~t 1 (31) NT2 ( added) where NT2'NT2 + NT2(recovered), (32) that is, ZNT2 includes both the tracer and the (d,a) reaction product. In equations (29), (31) and (32) the values of NT1 are calculated from C01B NTZ1 = (33) 60 f(Bij) which is the result of substituting eqtuations (22), (24) and (2&) into equation (21) neglecting thie fractional chemical yield Y. Assuming a constant beam, the for-mula Lor iJT2 is C02T a _10.1- - __ - __ (34) 60(1-e ) 3ince component 2 is long-lived the correction for nonconstant bombardment is negligible and equation (34) is valid.

-136The maximum error introduced by applying equation (34) to calculate iT2 for the tracer may be seen from the following consideration. In effect, these equations extrapolate N0 back to an effective midpoint, T, of the bombardment, the midpoint being governed by the amount of decay occuring daring the bombardment. If <(< T, then NT # NO. This can be seen from the equation defining T: NQo - T NT IcnT. (35) From equations (35), (22) and (19) it follows that An — - (36) A f(BiJ) Thus the difference between N0 and NT will be fixed by the half-life of the isotope in question, the amount of beam hitting the target, and the time distribution of the beam. Since the tracers used have long half-lives, the error introduced by using equation (34) for NT2 and'NT2 is le ltigible. As an illustration, thie decay of 25.4-day p33 d:ring one hour (a typical bombardment time) amounts to only 0.1 per cent waich represents the maximum possible error in the calculation of NT2 and J NT2 I The error is reduced much furt per by1 the extrapolation to P via equation (3 ). The chemical yield Y is finally expressed in terms of observable quantities by substituting equations (29), (33)

-137and N(34) into (31), mriaking use of the relation CT2- X2'rNT2 (37) where ZCT2 is the total observed activity of component 2 in counts per minute. Thus TZC02 QBCo1 (1-e ) f(3i) -y i) (38) TC02 (added) (l-e ) In case the correction for non-constant bombardment is not iade for component 1, the chemical yield becomes'C02 QCol (l-e-2T (1-e-A1T'fY _ -- (39) C02(added) (1-e ) The following is a list of symbols tused in the preceding section: a = fractional isotopic abundance = area of the targjet in square centimneters B = total n-umber of counts recorded by beam integrator Bi= number of counts recorded by beam integrator during the ith interval C0= ab solute counting rate at the end of bombardment

-13eCO1 = ab;solate countinc rate of component 1 at end of bombardment C02 = absolute counting rate of component 2 at end of bombardir;ent ~CT2 - total observed activity of component 2 in count rmin = duration of thIe ith interval (sec) f(.ij) = a function of Bi and Bj defined by equation (26) f(Iij) = a function of Ii and Ij defined by equation (20) I = number of bombarding particles/sec = number of bombarding particles/sec during the ith interval Ki - slope of integrator calibration curve, i.e., count s/min/ma = decay constant = atomic weight n n number of target nuclei per square centimeter N -numb'er of nuclei at time t = number of nuclei at the end of bombarcdnent; total number of nuclei formed hT1 = total number of nuclei of component 1 formed by nuclear reaction NT2 = total number of nuclei of component 2 formed by nuclear reaction.NT2 --- total nu.mber of nuclei of component 2 formed by nuclear reaction plus recovered tracer'<T2 = total number of nuclei of tracer of component 2

-139NT2(added) = total number of nuclei of tracer of component 2 added before chemical separation NT2(recovered) = total number of nuclei of tracer of compo'ient 2 recovered in the counting sample = the ratio N T2:NT1 from a bombardment in which no tracer was added - Cross section t = time after beginninr of Oombardment t. = time from thle end of the ith interval to,Le end of bombardlment in seconds ~ = duration of bombardment in seconds = effective midpoint of bombardment defined in equation (35) w = wei'lit of the target in milligrams Y = fractional chemical yield

-1402. Bombardment Procedure Targets of magnesium, sulfur, titanium, and cadmium were prepared as indicated in the Target Preparation Section, page 63, and placed in thle slots of t-he target probe head, figure 3. X;ny of the several positions could have been used for the bombardments, but usually the target was bombarded in,position 7, leaving position 8 for a collimator. Thus the effect of secondary electron emission from the target was diminished by the geometry (see fipgure 19). The collimator in position 8 was merely an additional aluminum target support fra.:e without a film across the 2 1/4 inch hole. The 1/16 inch thickness of the alumlinum was calculated to be sufficient to stop the deuteron beam plus any secondary pJrotons prodiced in the aluminum. The current integrator,was warmed up for several hours arnd tested to see that it was flunctioning normally. A calibration check was made either before or after each bombardment follo>wing thle secondary standardization method discussed in the Beam Determination Section, page 44. The "'background" (chlarg-,e collected or lost with no input current) was also checked to make sulre it was not unduly high. XA scattering tarcget of 1//4 mil iVylar was usually kept in the bean path (N of figtmre 1) during bombardment to assist in kee'pinc tuhe deuterons focused on thie Laraday cup targ)et. The scat ered radiation froin t7iis foil was detected by the monitor counters (O of figure I) and an audible sicnal was produced as well as a deflection on a meter located at the cyclotron conutrol panel.

fter bombardent, the t-arget Jwas- dissolved and the chiemical separation of the (d,a) product performed (see Chemical ooeparation Section, page 72). The substrate was dissolved along with the target because it contained some product nuclei which had recoiled out of the target.. The final counting sample was prepared on thin Zapon films as described in the Absol:;te Beta Count1int'ection, page,:7. The decay was followed in the 4rT counter arnd the absolute value of r.'0 determined for each component of the decay curve. 2.?lagnesium Bombardments The chart of the nuclides for the magnesium region is shiown in fig;ure 37. The products of the (d,a) reaction on magnesiun include two activities in sodium, the 15hloour l':a24 and the 2.6-1rear Ia22, Sodium-24 decays with thie emission of a 1.390 VTev f-ray, followed by two Y-rays in cascaae. (1W0) Sodium-22 is a positron emitter of 0. 542 P~ev maxiil.um enerrgy, followed by a 1.277 K]ev 7 ray. (150) P`ecently electron-capture bran ching amounting to 7-11 Ijer cent has been re oorted for Nra22 (11 - 13) No ot;ier activities were observed in the deuteron bombardmi —ents carried oult in this work. Five ma-gnesiumn bomr-bardm:ents were obtained, the first of wzhich was'or the pturpose of evaluatinf the chemical rrocedure (see Chemical Separation Sect ion, page 75). The experimental data and t-he various derived quantities used in the determination of the'cross section are pre

NUCLIDE CHART OF THE MAGNESIUM REGION A123 A124 A125 Al26 127 Al28 29.13s 2. s 7.4 67s 06y 100 2.27.m 6.56rn +,y /+ Jr 106y Mg22 M 23 Mg24 M 25 M926 M 27 M9 28.13S 12S'78.60 10.11 11.29 9.5m 21 h Na20 Na21 N 22 N 23 a24 Na25.385 s 2.6 y 100 15h P' P' P+EC, Y C -, C Pfig. 37. Chart of the nuclides shoszin:: possible products of deu-teron reactions on imagna;resium.

-143 - sented in Table VI for the last four bomr)ardments. The follolVing additional factors are needed for the calculation of the cross section: T'i = 2/+.32 LI - 24,32 9.'+ 0. 1 counts,/i-:in/J a 26 a (Ig )..1129 a g724 ) 0.760'he errors shown are standard deviations. Bombardments 2 anc 3 were exploratory in nature out are included for thie sake of completeness. The following paragraphs deal,with the distinctive features of the individual bombardiment s. T ar vet,rearation 1The preparation of the maglesium target has been uiiscussed on page 66. In all but one of the targets, the miagnesiumr covered a circular area just matching the 2 1/4 inclh hole in thLe aluminum target support. Tarfjet i!o. 3 covered a rectangular area lard er than the circular opening.'h,1is target (and to a lesser extent, target io. 5) was also peculiar in that it had the whitish cast referred to before on page 66. Target No. 3 was weighed before and after bombardm-ent to estimate the amount of material lost from magnesium targets either in the cyclotron or in transit. This djd not yield conclusive information, as thte wei-'t was 9.6796 g before compared to 9.7804 g aiter. The gain in weighit may have been due to adsorption of moisture onto the plastic substrate.

-144Table VI, Sumnmary of Bombardnient Data for Mg,(d, ),Na Bombardment No. 2 3 4 5 Suppressor ring(volts) floating 1110 1113 1110 D. C. supply (,xa) 5 5 10 Slits J(in. ) vertical 0.250 0. 500 0. 500 horizontal C.250 1,000 1.000 1.000 Slits L(in.) vertical none 0,25 0.25 0.25 horizontal none O 50 0*50 0. 50 Target i(mils of M.ylar) 0.50 0.25 0.25 0.25 Target Ai position 1 7 7 7 w(Img) 27.1 21o.8 16.0 18.5 A( cm2) 25.7 40.2 25.7 25.7 substrate Teflon ylar Mylar Myl ar Integrator range C D D D D(counts) 3 50O: 2.0 16420 10 154.0 f(B 4) (counts/sec) 0.001017 0.001D9 0.001916 13 T(sec) 3600 5160 7 117 4614 Mumber of -,apon layers 1 4 4 4 on counting sample,apon soln. used ( ) 33 50 50 50 O'1(iLa24) (counts/min) 6YI` 1+366 162130+1050 40398+175 35585+916 0C02(1a22)(counts/min) 4"J3.7+1.5 1646+6 300.5+2.3 546.4+4.2 10. 3- 10.73 ~a22 tracer added (,!) 0 10 0 30 C02(added) (for 10 I 1) 0 1352+4 0 21,22+_1.9 Y 0.341 0.0435

-145 Chemical Procedure The chiemical procedure'has been presented in figure 22. A noticeaole residue was present in the counting sample prepared for 4T1 counting in boimbardments 2 and 3, so doubly distilled water was used to make up the eluting agent in Oombardment 4, while the use of conductivity water (lxlO0-6 ohmm cm 1) reduced the residue even more in bombardment 5. A quick estimate of thie amount of residue (from l10 ml of eluate) present in the counting sample of bombardment 4 showed it to be below the detection limit (0.0001 g ) of the alinsworth type DLB balance. The chemical yield was determined in bombardment 3 and 5 by adding tracer!a22 produced in bombardm.;ent 1. The stock solution was stored in a polyethylene bottle, and 10Al1 aliquots were transferred to the Phillips beaker just prior to dissolving thle target in bombardllent 5, and about an hour before in ombrardment 3. After the chemical separation was under way, I.0 1 aliquots were transferred to 4n, counting plates to de22 ieri.ine the amount of 1,22 added as tracer. One 10/1A1 sample zias prepared in this way in bombardment 3 and three 10,1 samples in bombardmrent 5. The alifquoting error was assessed 22 in the latter case from the average of the three Na counting rates. The 4rr plates used in each case were prepared identically to the ones used for the samnple of the sodium separated from the tarcget.

-146Aobssolute DBe-'a ountig Discrimination curves were taken for the decay samples of bombardlnents 4 and 5. Curves obtained at various times after bombardment corresponding to 3 and 100 per cent of la22 in the mixture of Na2 and Na24 are presented in figures 38 and 39 respectively. azch point has been corrected for the small amount of decay of the Na22- Na 24 mixture that occurs durin;g the time in which the data were obtained. It can be seen that the discrimination plateau exhibits no evidence for a slope within the experimental error due to the counting statistics. The limits of this error are indicated at the bottom of the graphs; the length of the line represents twice the standard deviation of a single point. As seen from 22 24 figure 38 and 39, altering tihe Naa N Ira2 ratio has no effect on shiftinrg the plateau. Thiis the error in the absolute counting rate due to inaccuracies in the gain, voltage, and discriminator settings is represented by the standard deviation of + 0.7?4 per cent obtained from the average deviation in the points on the plateau of figure 38. The effect of absorption of weak beta particles in the Lapon film was studied by a series of samples mrade with different numbers of backin;- films.'Three typical absorption curves, taken fronm bombardment 4 are shown in figure 40, Each point has been corrected for decay which Occured during the time ticat the six samples were being counted. The points exhibit a scatter about a line drawn arbitrarily through the points and show no evidence for a slope. The errors indicated in

-147DISiCRIMINATION CURVES OF SODIUM + (CYLINDRICAL 41r PROPORTIONAL COUNTER) — + —GAIN 100 -- GAIN 75 9800' 3800 VOLTS -X- GAIN 50 0 + 9600- - I x a 9800 | 3900 VOLTS,9600 _ t- x + x; o 9400 - L_ X 9800 u 9800 94000 VOLTS LU + 9600 + 9400 I LIMITS OF COUNTING ERROR (STANDARD DEVIATION) I I I I I I I I I I 10 9 8 7 6 5 4 3 2 DISCRIMINATOR SETTING g. 3ci. Discrinmination curves of a mixture of 3 eer cent Na22 and 97 oer cent Na24 taken from magnesium bombardm-ent 4. The limits of counting error shown are twice tie standard deviation of a sin,!le,oint.

DISCRIMINATION CURVES OF SODIUM (CYLINDRICAL 4Th PROPORTIONAL COUNTER) -+ -- GAIN 100 -o- GAIN 75 1550 + t\ 3800 VOLTS -x- GAIN 50 1500 + ______ O + 1450 + 4000 VOLTS 1550 1500 1450 I LIMITS OF COUNTING ERROR (STANDARD DEVIATION). I, I I I I, I I! 10 9 8 7 6 5 4 3 2 DISCRIMINATOR SETTING Fig. 39.'iscrilmination curves of 100 per cent NJa22 taken from magnesium bombaordment 3. The limits of the counting error s:,own are twice':he st'-;'ncrd dvi-tion of a single point.

FILM ABSORPTION CURVE OF SODIUM (CYLINDRICAL 4T PROPORTIONAL COUNTER) 7500 1.7% Na22 17000 _ 16500 LLI z 1500 aW 20% Na22 1460 20% Na FD 1420 0 0 310 22 100% Na22 300 290 L I UNCOVERED SAMPLE 1 2 3 4 5 6 7 8 9 10 LAYERS OF FILM Fig. 40. Film absorption curve of various mixtures of Na22 and Na24 taken from magnesium bombardment 4. The errors slhowm are standard deviations due to countingr statistics.

*T' -i9.C-:,,T- i q..o, T T o T.l,-.,'eTU,,Tt h- ~: tl TO-TU pDoq.a; ot; eq $suGuodLUoo 4+ei JllotL-Qc: pu e Z9 9 +ozS- paTOSeJ um sLu,,;'L wOJT P8TJtUg ufpOs JO WI'.Z Tpo o7 K-I (SdfOH) 3WII OZZ 00Z 081 091 ol OZI 001 08, 09 Ob OZ OZDN 0 00 0 o \ z01 0 I! I I I I I ON o I 01 0 01 ONN (W3INfOO'VNOll!AOdO~d.u. -VDIdtC1 NIH- %DY)' ~Yniaos ~o-0~'[

fi/jure 40 are standard deviations of the individual counts. Th'e points, correstpodiing to different samples, do not fall on the curve because of the aliquoting error which has already been shown to be of the order of 2 per cent (see Table V). Figure 40 also shows that changing the Na22- ha24 ratio produces no noticeable effect. The average of the points representing the covered samples was taken for each of the absorption curves, the average deviation in the points computed, and the corres^,onding standard deviation was calculated to be 0.43, 1.15, and 1.23 per cent, which is of the same order as thle counting' error.!nalysis of eav Crves It was felt that tle decay data warranted treatment by thle ki-ethod of least squares, since the two nuclides encountered in the magnesiumn bombardments possess vastly different half-lives (15 hours compared 2.6 years), and can be produced in sufficient quantity to make it possible to obtain good counting statistics. A typical decay curve obtained from bombardment 5 is pictured in figure 41 showing the resolution into thwe two components. T'ie procedure followed for the least squares analysis was to set up "machine equations' (1..4) that facilitate tile compatation of th.e sums by use of a calculator. Computational errors were avoided by obtaining agree ment between two persons as to thie valle of a summation before accepting it.;Similar checking went into the conversion of calendar time to the number of hours after the end of bombardnent. It wras

-152decided to weig;it each of t?~e points eqiually in spite of tlie fact that some were obviously better tilan others. Half-Life of Na24. The gross decay curve was thus resolved by first finding the least s-uares line of best fit for the 2.6-year ~a22 component and thien subtracting it from the gross culrve to pjroduce the points through which a line of best fit was computed byJ tie same method as flor the 15-!hour 1Naa Jince the long-lived activity could not be followed for a time lor, enotugh to determine its half-life experimentally, the value of 2.60 years obtained by Laslett (1:35) was assumned for the purposes of this calculation. Laslett did not quote an error, but one mnay assu-ne the uncertainty to be -0.4 per cent. This will not affect.he present decay curve analyhsis 22 24 because of the relaitively large differertce in the ia22 and a4 half-lives. Tihe short-lived colmponent, how.rever,.....ib.ited an exponential decay over factors of 150-4000, so tiat trhe uiaLf-lilfe could be aeterm-ined,y least s i. ares. Thle decay c:irves were first analyzed graphically to 22 determine.which points sholild be included in tile Na22 analysis. 22 24 In'able VII the half-life and intercept of the Na and Na24 obtained b)-/ the methrod of least squares are gLiven together with the graphical resollution for comparison. In bombard.ment 2 t&he Lapon film broke after most of t.he ia24 had deca-yed, so thle sam.ple was repaired by sandwiching it between two additioria l Japon fil-ms. lpparertly somne activity was lost in this operation as the Na24 decay }oints deviated rm-arkedly from linearity after this happened and were

therefore neglected in calculatin'g t-he half-life and intercept of the!ha24 component for bombardment 2. Table VII.'Analysis of Decay Curves of Sodium Bbdt.:!ethod of Period of CO Lko. analysis Half-life observation (counts/min) 2 1. s. 2.60ayr 0. 64 yr 483.7+1.5 Cgraph. " 485 3 1. s. 0.21 yr 1646+6 jia22 graph. 1650 4 1. tt 0.18 yr 300.5+2.3 graph. 298 5 1. s., 0.16 yr 546.4+4*.2 graph. 545 2 1. s. 14. -13+.025hr 100hr 6*,C 8 1+366 graph. 15.1b 66,000 3 1. s. 14. 910+.025 144hr 162,130+1050 1a24 graph. 15 1b 157,000 4 1. S. 15.002+. 017 166hr 40,398+175 graph. 15*1 b 40,000 5 1. s* 14.946+ 101 146hr 35,8$5+916 graph. 15. lb 36,000 a See reference (185) b See reference (l16)

-154The resolved lines, computed by least squares analysis have been depicted in figure 41 for bombarC lient 5. They are seen to pass through the points, as do for the points obtained by graphical resolution. The weighted average of tiLe Ia24 half-life is compared with the values present in the literature in Table VIII. The value obtained in this work compares favorably with those of Solomon (158) and Tobailem (193) which seem to be the best values available. Table VIII: 3urmrnary of Half-Lives of Na24 t 1/2(Ho-irs) Authiors Reference ervation Observation 14.97 +0.01 Loclkett and Th'omas 187 5 Half-Lives 15.06+0. 04 Sreb 1P6 15.04+O. 06a 3olomon 1` 8 13 Half-Lives 15.10+0.04 Cobble and Attleberry 19 15.1 +. 1 1Tilson and B'ishop 190O 2 Ialf-Lives (recalculated by Oreb) 15.0 +L.0_b oinclair and HIolloway 191 4l S +-1~ Van Voorhis 1'32 10 Half-Lives 14.90+0.05 Tobailem 193 (Differential "Method) 14.93+0.04a This work 7-12 Half-Lives a Standard Deviation b Lrobtale EZrror

-155Cross Section for Na 22 and i, a24 As mentioned above, bombardments 2 anld 3 were exploratory in nature, and the measurernent procedures had not been developed to their fullest, whereas bombardments 4 and 5 were performed with much more refined techniques. Bombardments 2 and 4 were used to determine relative values, and 3 and 5 were used to determine absolute values of the cross section. In bombardment 2 the traffic counter was not used with the integrator and hence f( ii) could not be calculated. The residue on the I4 plates was relatively larvae which could have resulted in an unknown amount of self absorption. In bombardment 3 one circumstance which may have resulted in too low a value for e was th-at the target was suspected of having some oxide impurity. Another unknown error acting in the same direction was the possible self absorption due to the residue on the l4 plate. There were three sources of error in this bombardment acting to give too high a value of a. First, the magnesium target extended beyond the area defined by the 2 1/4 inch collimator (spare target mount), and the possibility exists that a contribution to the yield of Na24 bt rai ~2gn 24 was made by the reaction g24(n,p) ia24 resulting from neutrons produced in the collimator. Secondly, by adding thie tracer an hour before dissolution of the target, some 1a22 may have exchanged with the sodium of the Pyrex beaker, Thirdly, the aliquots made for determining C02(added) were talken several hours after adding the tracer to the P'hil].ips

-156beaker. Since the volume of the Na solution was only -2 nil, evaporation from the solution may have produced a significant carnge in concentration which would have resulted in too hligh a valu.e of C02 (added), 1-Toreover only one saliple of the tracer was prepared to determine C02(added). In bombardment 5 the target!may hnave had a slight amount of oxide impurity, and the aliquote of the tracer were taklen several.houlrs after adding the tracer to the ihillil:s beaker. T' e results of bol-mbard%- ents 4 and 5 were obtained sing Im.ch better techniques than bombardm' ents 2 and 3, as seen Yfrom the al)bove considerations, and t;-erefore tlhe cross sectioans are based upon bombardments 4 and 5 only. Tie values are 2 (hia2 ) 0.151 + C.006 barns at 7.8 + 0.1 Ilev r(:a24) * C.O'4 + 0.004 barns at 7.$ + 0.1 tev, calculated withl the aid of equ~ations (25) and (3cJ). The cross seclion for tihe formation of 4Na22 has been corrected for electron-capture branching as follows. The branching ratio has been reported variously as 7.1,er cent,(l?]l) 11.0 per cent,(1l2) and 6.5 per cent. (183) 2ssumninr a value of 8 1,er cent,,and the counting efficiency of thre 4r counter' to be, per cent for electron-captture events, L;]ie overall counting; ei'iciency of the 4Tr counter Jherr ard i ilter (iS2) ueterinined experimentally for the 3orkowsTki 4r counter, and was ised to calcilate the cross 22 section for Na quoted above.

-157The errors in t-he cross sections quoted above are based upon the errors associated with eaclh of the factors of equations (25) arid (3b), and may be considered to be an estimate of thle standard deviation. The component errors which were assuined in the propa>,ation are listed in Table IA. Table I!. Lropa.ation of Errors in Cross Jection for:>g(d, a Na Estimated Standard Dieviation (per cent) Fa c for Bombardment Bombardment No. 4 No, 5,02 0.77 0.77 O 0.01 0.01 (l-e' X 2') 0D.19 0.19 COl 0.43 2.55 B 0.03 0.03 f(8 ij) 0.03 0.03;, eqn. 29) 0.91 02 (added) 0.09 A G.417.02 1.01 26.05 a(i:g 24.01 a(h. 24 c(:a2 ), eqn.(25) 4.04 cT(a 22), eqn.(30) 4.15

-158Lxcitation functions of tie kIg(d, a) 1ia reactions have been previously reported by Irvine and C'larke (194) and 2artell and Softky.(195).f ieading the cross section values at 7. 8 Mev from their published graphs, 0.092 and 0.147 barns were obtained for Na24 and Na22 yields from Irvine and Clarke's 22 work, and 0.009 barns was obtained for the a22 yield from B3artell and Softky's results. The results of the present investigation are seen to duplicate those of Irvine and Clarke quite closely. i0. Sulfur Bombardments The chart of the nuclides for the sulfur region is shown in figure 42. There is only one (d, a) reaction product of' sulfur t, at is long-lived enough to observe convenienrtly. Thiis s tile 14.3-day p32 which decays by the ermission of a 1.701 rHev f ray. (180) It was the only activity observed in the bombardmrIents of sulfur.?'ive sulfur bombnardments were obtained, the first of w;hich wag for thne purpose of evaluating the chemical procedure (see Chemical Separations Section, page 78). bombardments 2 and 5 were worthless, as the sulfur was vaporized'by the cyclotron beam. Thie experimental data and derived quantities for bomnibardments 3 and 4 are presented in Table f. The following additional factors are needed for calculation of th-ie cross section: i~ = 32 066 ~D = 9.9 + 0.1 counts/min/ a a(S ) 0.04215

-159NUCLIDE CHART OF THE SULFUR REGION l 33 Cl 34 Cl33 cl36 Cf37 cl38 Cl39 2.8s 33.2 m 75.4 4.4xlOy 24.6 3729m 55.5m _ _ + I+,y, / -.r -; S 31 s32 S33 34 S35 S36 S 37 3.18s 95.018 0.750 4.215 87.1 d 0.017 5.04m!P+ __- __ _ p 28 p 29 p30 p 31 p32 p33 p 34.2 s 4.57s 2.5rm I00 14.3 d 24.4d 12.4s Fig. 42. Chart of the nuclides showing possible products of deuteron reactions on sulfur.

-160The errors given are estimated standard deviations. The distinctive features of oomnbardments 3 and 4 are treated below. Taret Preparation The preparation of the sulfur target has been described on paole 67. Target io. 4 was weighed before and after bombardment and it was found to ilave lost 8.g2 ng, or 11 per cent, during thle irradiation. Inspection revealed that with-lin the area struck oy tiLe beam there appeared to be less sulfur than wtas originally present. Thus it seemed that sealing the target between two layers of polystyrene was not 100 per cent effective in preventing vaporization of tile sulfur by the Deam. Chemical?rocedure The chemical p;rocedure has been presented as fig-ure 23. Difficulty wias experienced in dissolving target No. 4. it was finally found necessary to heat the mixture strongly for a few minutes after 7 hours of less vigorous treatment. Since sulfuric acid was present the possibility existed for some phosphor-:s to escape before tl-e tracer had the ojiportunity to complete the interchtange wfith; the (d, ct) product. The tracer used for c"hemical yield determination in 33:onimibardnrrent 4 -was the 24.4-daa P33 This was obtained by allowing the l14.3-day F32 to decay from several old bottles 32 of high specific activity P (Carblide and Carbon rhemicals Co., Oak?idge National Laboratory, Ojak iidse, Tennessee, Catalog io. t:-32-i'-11). Two 100, samples of the tracer

-161Tahbe X. Su.mmriary of "Ionmbardinent "jata for S(d,u)P 3ombardrmient i;o. 3 4 Suppressor ring (volts) floating 1110 D. C. supply (Aa) 25 10 Slits J (in.) vertical 0.500 horizontal 1.000 Slits L (in.) vertical none 0.25 horizontal none 0. 50 Taret i (mils of Niylar) 0*25 Target iR position 3 7 w(mg) 74.1 73.7 A(cm2) ) 25.7 25.7 substrate polystyrene polystyrene Integrator ranLe D D,( cnts) 3241+ 64 ( sec) 6060 4''79 i.umber of "Zapon layers 5 2 aapon soln. used (;) 50 50 001 (32 ) (cnts/mirn) 40,000 9700+ 200 C2( p33) (cnts/min) 0 1750+ 100:33 tracer added (Al) 0 750 C02 (added) (for 100/Al) 0 1173+ 33 Y 0.20

-162DECAY OF PHOSPHOROUS (CYLINDRICAL 4W PROPORTIONAL COUNTER) 104 LLJ 0 14.3 DAY o3 I0 20 30 40 50 60 70 80 90 I00 I10 TIME (DAYS) Fig. 43. Decay curve of phosphorus separated from sulfur showing the decay of 14. 3-day p32. The data were taken ftrom sulfur bombardment 3.

-163were made to determine the wamount of tracer added, one at the same time and one 30 minutes after adding 750 /1 of the tracer to the Phillips beaker. The 4rn counting plates used for the tracer and for the separated phosphorus were prepared under identical conditions. An insignificant amount of P33 is produced in the bombardment from secondary neutrons as indicated by the absence of a significant amount of activity with a half-life greater than 14.3 days. This is shomwn in figure 43, taken from sulfur bombardment iDo. 3 in which no P33 tracer was added. An upper limit of 0.6 per cent of the initial activity could possioly be attributed to P33. Absolute Beta Counti in Discrimination curves had previously been taken of p32 wren the work with the n'hational Bureau of Standards P32 sample was done (see page 114), so only enough curves were taken with the (d, a) product to determine the voltage, gain, and discriminator settings to count the sample. These curves are shown in figure 44 and correspond to 77 per cent P32 in the mixture of t}32 and P33 Note that the characteristics of the counter have changed since the work was done on the S33 sample so that the discrimination plateau has shifted towards a higher bias potential. The magnitude of the error due to a possible slope should not be affected, so that the limits of this error can be taken to be the same as previously determined.

-i 64DISCRIMINATION CURVES OF PHOSPHOROUS (CYLINDPICAL 4m PROPORTIONAL COUNTER) + GAIN 100 o GAIN 75 3900 VOLTS 3500 w 3400 t.... z 3300 LLU H-z i \ 4000 VOLTS o 3500 3400 + o 3300 I LIMITS OF COUNTING ERROR (STANDARD DEVIATION) I I I I I I I I I 10 9 8 7 6 5 4 3 2 1 DISCRIMINATOR SETTING Fi;. 44. Jiscri:;-i'an~tion curves of a mixture of 77 per cent F32 and 23 per cent }p33 taken fromn sulfur oomllardrnent 4.'ihe li:its of counting error si-~own are tvsric ce te:tndard devii-.'ion of a sinrle point.

-165104 DECAY OF PHOSPHOROUS (CYLINDRICAL 4'" PROPORTIONAL COUNTER) 10,000 8,000 6,000 4,000 LU 0 2, 000 K 0 0.2 0.4 0.6 0.8 1.0 LU._| e( X2- XI)t z o-:D 0 0 \,R,,,p32 33 10 20 30 40 50 60 70 80 90 100 TIME (DAYS) Fig. 45. Decay curve of phosphlorus separated from sulfur. The resolution is sio~n in the insert,'he slope and intercept ~jivin j the initial activities o t>le P3 and P33 respectively. The data were tken fr:rom sulfur bombardmentre 4,

Similarly, th&e film absorption problem nad been investinated before (see page 1C2). Analfsis of 0Iecay Curves The decay curve of the phosphorus sarnple from bombard1ment 4 is given in figure 45t Since the two component halflives are so close together as to make ordinary graphical analysis extremely difficult, the data was treated by a met. od indicated by`:Vorthing and Geffner (lI) and applied by Biller (197). The gross decay curve given by -X t -T C = C0e + C02e (40) may be rearranged as follows: ce2t (h(2-X)t Ce Co01e + C02 (41) The left hand mrember of equation (41) is plotted argainst (X2- X)t e to give a linear curve whose slope and inte-cept are respectively C01 and C02. The insert of figure 45 S.1ows such a plot when applie 3 and t3 as components 1 and 2 toi a an' P as components 1 and of the gross decay curve, e iving thie values apaearing in Table X. The decay of the F33 tracer is shown in figure 6. Although mnost of the e32 had decayed, there is some longerlived activity than the 24.4-day P 33. Unfortunately t.he tracer was not purified'before use in bombardment 4, so for the purposes of calculation it w-as assumed that the contamlinant was C7.1-day;35 w,.ilch is also produced when sulfur is

irradiated with pile neutrons. The decay curve was resolved into t.he 24.L-day and 17.1-day components by application of equation (41). This is shown in tlie insert of figure 46 35 wnere the intercept is thie initial activity of S5 and the 13 t slope is the initial activity of tie P',-C02(added), as given in Table (. In this graph t-le points corresponding to both samples are plotted, illustrating the aliquoting error. Cross S.ection for P32 i,,h accuracy of t-lese experiments is in doubt due to the several unknow;n errors associated (,ith;il thle sulfur oa om-ardci e nt s. As mentioned above, bombardments 2 and 5 yielded no res.lts, and bombardment 3 was used to obtain relative ~;ields (in t:.is case,'.t = O). Bombardment 4 was thle only onh- in which the absolute cross section of P32 was measured, and unfortunately it was impossible to obtain another bombardment to determine the cr'oss section more accurately. The loss of sulfur from the target when it was struck by the beam may have amounted to -10 per cent, and possibly resulted in an error in c of several ti1les 10 per cent. T.is would'e tLhe case if the sulf'ur was lost from tie one-half square inch 33 area that was hlit by the beam. Likewise any P33 tracer lost by the s- lfuric acid treatment of t-e target wo4ld also lead to too low a res;ult. Thie variation in target thickness was estini-ated at ~2 per cent, but may have been larger. The radioactive i:-iurity in tl-e P33 tracer andi the ifficulty in the resolution of thLe 3 2-33 decay curve led to sizable errors, Jut certainily sm-iall cowmpared to t -e above unknown errors.

-16~_ DECAY OF P33 TRACER (CYLINDRICAL 4T PROPORTIONAL COUNTER) 1600 - 1200 - 800 - Lw O 400.2.4.6.8 1.0 cr (X -,)t L 103 e 2I F- L SAMPLE 1 OD 1 <o SAMPLE 2 0 33 +35 I 2 0 I I f I I I;;;L 10 20 30 40 50 60 70 80 90 100 TIME (DAYS) Fig;.'r(). 3ecay curve of 013 tracer. The resolution into the P33 and assunid l,3 co^-mponcits is soworn in the insert, the slope and inltercept Jivinn thLe initial activities of p37 and S35 respecti vely.'Te data wiere takenr from sulfur bomvbardment 4.

-169The cross section, calculated with the aid of equations (27) and (30) is ((P32) = 0.3 + 0.2 barn at 7.7 + 0.1 Lev. The uncertainties introduced by target vaporization and chemical yield determination were so large as to outweigh tie errors inherent in the individual factors as outlined for magnesium, page 157. 2urthermore there was no way to assess the magnitude of these large errors, so that the error in the cross section quoted above is merely the writer's surmise of the standard deviation. 2. Titanium Bombar-dment s The chart of the nuclides for thle titanium region is sh-own in figure 47. Three of thze (d, a) reaction products of titanium possess half-lives of t}he same order of magnitude: 2.4-day Sc44, 3.44-day Sc47, and 44-hour`c48. The 2.4 day 44 isomer decays by internal transition to 3.9-h:our Sc which in timrn decays with the emission of a 1.4i ev C ray to stablE Ca44. Scandium 47 and 46l decay by the emission of 0.61 and 0. 64!ttoev f rays respectively accompanied by Y radiation. The fourth (d, c) product is Sc46. The 19.5 second isomer decays to the 85-day ground state of Sc46 which enits a Y ray and a 0.36 Kiev P ray. (160) Five bombardmients of titanium were obtained, the first of which was for the purpose of evaluating the chemical procedure (see Chxemical Separations Section, page 80). The experimental data and derived quantities for the other

NUCLIDE CHART OF THE TITANIUM REGION V 47 V 48 IV49 V 50 V51 v 52 V53 32 rn 16.3 d 635 d 0.24 99.76 3.75 m 23 h __+ p,ECr EC,Y /. r;r Ti44 i Ti45 Ti46 Ti47 T 48 | 49 Ti50 Ti51 >23y 3.09 h 7.95 7.75 73.45 5.51 5.34 5.8m SC40 SC41 SC43 44 45 S46 S47 48 49 Sc Sc Sc Sc Sc Sc Sc Sc Sc.3 s.873 s 3.9 h 2.4 dc3.9h 100 19.5sI 85d 3.44d 44h 57m l_____r I____ 19,Y II rITIC ____.r 1T3,& r I7 fTr )W lFi:;3. 47. Chart of' the nuclides show;ing possible pr'oiucts of deuteron reactions on titaniumi.

Table rI.,Summrary of ombardment Data for Ti(d,mc)Sc Domibardment Nio. 2 3 4 5 Suppressor ring (volts) 1100 1105 1095 1090. CC.. suply ( a) 2.5 48 30 38 )ulits J(in.) vertical 0.250 none none none horizontal 1.000 none none none Slits L(in.) vertical 0.25 0.25 none 0.25 horizontal 0.50 0.50 none 0.50 Target i(mils of Ifylar) 0.25 none 0.25 0.25 Taret [t position 8 7 7 7 w/Ai (mg/cm2) 29 29 29 29 " su s t,rate lar It rar Integrator raiarFe B D D D 3( crts ) 1726 17 - 240 210 T(sec) 35340 4573 13051 6809 ilumnlber o n' Japon la::ers 4 2 2 2 -,ajon soln. used (7) 50 50 50 50 C01(c44-47-4) (cnts/min) 253+45 1%9+100 574+10 707+15 Half-life of Sc44-47-4 (days) 2.63 2.91 22 2. 82 C02(3c46j) (cnts/min) 80+10 510+15 244+6 277+4 10.09 7.7a.49 c46 tracer added (01) 0 10 20 C02(added) 0 0 15:?1+8 3055+_34 Y O ^)0. 0_t60).oo9 0.022;t.004

-172four bombardments are presernted in Table -I:. In addition tie:Follolwing factors are needed for the cross section calculation: 1 47.90 K. = 9*9 + 0.1 counts/min/ya a(Ti4 )= 0.73145 The errors given are estimated standard deviations. The distinctive features of the individual bombardn-fents are treated below. Beam Determination it was noted th-at in "bombardment 4 the traffic counter occasionally introduced extraneous char-es into the current integrator. The ch-eck on the calibration factor showed that, irt ay have been high by about C25 per cent for bombardments 4 and 5. Taet rearation The titanium foil used for the targets has been described on page 69.'The metal was cut into pieces sligh-tly larfTer than the opening, in the target support which was specially made from 1/1Jt inch aluminum. The opening was a rectangular hole mleasuring 1/2 x 1 inchl machined in tlhe center of t he tar-,et supiport. The tar:tet w.ras Scotch-taped to thie back of the suOpport, so t'.at the rectangrular openinr d.efined the area o-! the titaniumr hit by the deuterons. In t+he absolute determirnations, C0.001 inch.:'lyar films were placed belhind the foil to serve as ca;tciers for recoil in th-e same way as the target substrates served for the other target elements.

-173 - Chemical Procedure The chemical procedure of figure 24 was followed except for minor changes. In bombardient 3, the scandium was carried through four cycles instead of t ie usual three. Conductivity water uasn't used for the preparation of reagents until bombardi-mlents 4 and 5, so that more solid matter was present on tle 4Tr samples in the earlier runs. The tracer used was high specific activity 85-day Sc46 (Carbiie and Carbon Chemicals Co., Oak lidge R'ational Iaboratory, Oak Jidge, Tennessee, Catalogue i;o. Sc-46-P). The solution of Sc40 used in tile chemical yield determination was processed throuCL two c-cles of the chemical separations procedure used in the titaniumq bocmbaricdents. About 20 ml of the purified sc46 solution was stored in a polyethene bottle. Three samples of the Sc45 solution for thLe determination of the amount of tracer added were prepared at the same time as an equal aliquot was added to tl:e Phillips beaker. In bombardment 4 this was done after a homogeneous solution was obtained, and in bombard;ent 5 before the target was dissolved. The 4n plates used for the tracer were identical bo those used for the clemically separated product. A..solute Beta Countint Discrimination curves for scandium are shown in figures 4!J and 49 corresponding to 70 and 100 per cent'5-day e46 respectively in the c44- Sc46- c47 oc4 mixture. rlThe data were taken on t he sample from t itanium bombardment 4 and were corrected for decay. The standard deviation

DISCRIMINATION CURVES OF SCANDIUM (CYLINDRICAL 4r PROPORTIONAL COUNTER) — + -- GAIN 100 ao -GAIN 75 360 - 3900 VOLTS - GAIN 50 350 5 340 t + + D 330 36O 350 _ 340 JERROR (STANDARD DEVIATION) 4"? a;: i ci ra ixt r t* \7e r f 0 e citinrtror c ad J'I iL24 C, k4 ta.tiren fri:i titani um rk- n 4h T!e i00its o n li.eror so; r twic. L- o

' -7 5DISCRIMINATION CURVES OF SCANDIUM (CYLINDRICAL 47 PROPORTIONAL COUNTER) + GAIN 100 o GAIN 75 155 x GAIN 50 3900 VOLTS 4000 VOLTS F-155:150 X, \ \ 4100 VOLTS 145 140 LIMIT OF COUNTING ERROR (STANDARD DEVIATION I I I I I I I 10 9 8 7 6 5 4 3 2 1 DISCRIMINATOR SETTING t'i:,. 49. Dis crirmination curvre~ of 100 pjer cent Sc]'6 taken from titaniu:.~ bo.;.ib;,rdil-;-nt 4. T,1e li. i'- s o! "- he'ountirl"' error s;'o-.:n.re tsiice thie ot; —nd...rd dev-i'- ion o.!,.;.~in..;le ipoint.

-176ABSORPTION CURVE OF sCANDIUM (CYLINDRICAL 4' PROPORTIONAL COUNTER) 46 380 - r 50 % Sc 370 ii 360,_230 82% SC46 220 o 2z 0 C 150 (3 \~ 1\00% SC4 1450 140 Manua 135 130 1345671 90 - 12 13 2 EQUIVALENT LAYS F FILM ElTiK300 400 500 FILM THICKNESS (p.9/cm) o);c~? 5Filmlabsol ptlon curvo of VariOUS! of-l.a o,0 aen from titan Li nT:Qrj eflum 5.!,j C deviationls ue to counting satl errors shownf are standard

-177corresponding to t e average deviation of the points on the discrimination plateau of figure 49 was found to be 2.0 per cent. This is larger th-an trhe error diue to counting statistics, -but it is reasonable considerir:nr that these data wrere collected over a 7eriod of eight days. The film ab'sorption cui.rves are presented in figure 50. The sample plates were prepared as illustrated in figure 32 and consisted in different numbers of lsaers of Zapon coated with gold. h'ihe thickness of th'fe films were determined by the _ 63 i3 Pn-ray a xsorption m-ethod discussed on pa:-e 100. The data w"ere taken from bombardm- ent 5 and each point was corrected for decay. -iA thoug-,h the curve does exhibit a definite slope, the decay curves were taken of sanIples mounted on only two layers of Zapon, so thte film absorption correction amounts to only about 1 per cent. A-nalLsis of Jeca Curves typical decay curve is shown in fiLgure 51 for titanium 0boflbardlnmenit 4..then the (-'5-day 3c4 is graplhically resolved from the (ross culrve, the resulting points define a strailht line wl~hose slope corresponds to a half-life of 2.o3 - 2.91 days. This is, no douibt, th:e compound decay curve of the 2.4-day isomer of. 5c44m, the 3.44-day jc47 and the 2.4-day *c44. If one assiiumes th'at thiese isotopes are produced in eqLual yield, th'at thre electron-capture branching of the 3.9-hho r'c44 is negligible, and that the counting efficiency ifor the i-ternal transition of Jc44m is zero, the calculated decay cuirve is practically a straight line corresponding

17'C m, DECAY OF SCANDIUM (CYLINDRICAL 47 PROPORTIONAL COUNTER) 0D 00 xa \~~~o ~O SC46 x4 00 0 Lo 4 150 300 30 io 00 2.82 DAY200 0 IF-' —'I I0 0 50 2 4 6 8 10 12 1 0 3 0 40 50 _ ETIME (DAYS) Fi. 51.:jecay curve of sc dium separated from titanium resolved into t!e 5-day Sc40 and a 2.62-day component, Ti'c cata viee ta1o k:en from titanimuli bomioard.ment 4.

CALCULATED DECAY OF Sc 44. 47,48 >o Sc44+ S47+ Sc48 (2.5 DAYS) SC_ 4 SC47 SC44 2 3 4 5 6 7 8 9 i0 I 1 12 13 TIME (DAYS) Fig. 52. Calculated decay curve of Jc44, 47 cand c48, assumirig equal reaction yields, negliible electron-caplure branching and zero countin -; e fficiency for interval trasition of.ic

to a 2.6-day half-life. This is illustrated in figure 52, starting 2 days after bombard:ernt to allow for the growth of tie 3..-houlr positron emittilng daug;hter of 3c4m. These assumptions are not completely valid, but it can be seen that the experimental decay curve i.s not in conflict withl expectations, and furthermore thiat it is practically irlpossible to resolve the 2.63-2.Q1-day slope inlto its comrponent s. Cross Section for Sc46 BombDardment 2 was obtained during the exploratory stage of the work. It resulted in a sample containing a sizable residue. The counting? rate was very low, so that the resolution of tae viecay curve was not very accurate. Therefore the results of this bombardinent were rejected in tihe calculation of cross sections. Bombardment 3 determined relative yields wihile bombardment 4 a-id 5 determined absolute cross sections. Jince part of the titanium tarcet Jfoils were behind t ue collimator in each case, thle possibility existed t;hat somne scandiuimi activities cou!ld nhave come from (n, p) reactions from fast seconcd-ary neitromns produced in thae collimator. This resulted in an unknorn syst;ernatic error. Anotler error associated with borbardmients 4 and 5 is in thue dete rmuination of, t!e nurrmber ofL incident deutuerons.,s iierftiouled above, thie iu teserator recorded a sm-all but unknoJn &:.ount of' beam in co;u~oard.u.ent 4. correction of 0.25 per cent wvas considered ne ~ligiule in the interator calibration

factor for bombardment 4 and 5, as determined by the calibration check. Because of the uncertainties mentioned above and the difficulltv in resolving the decay curves, the correction for non-constant bombardment was considered negligible. The expjerimental resolution of the decay curve of bombardtclent 3 into the`5-day Sc46 and the shorter component was the basis for calciulating, i. e., the number of atoms of Sc46 produced by the (d,o) reaction was referred to the total number of atoms of the 2.4-day, 3.44-day, and 44-hour activities. The standard deviations in the cross section of each of' the bombardments, estimated as for the?;g(d, a)ha yields, were smaller thlan thle spread in thie valAues from thie two separate determinations. Jince there was insufficient reason to reject eituier one of the bombardments 4 or 5, ioth results are presenteJl below. 7 e cross sections thus obtained wit! ti.e aid of equ-ations (27) and (32) are Bbdt 4: ($c4')-.O0O1; +.00004 barn at 7.0 + O.b iKev Bbdt 5: c( c4 ) -.00070 +.00012 barn at 7.0 + O0.-!,lev where the errors are estior-.ts s Xf the standard deviation of bo;mbardmn-entts 4 and 5 based upon the iJropa -ation of the errors in the ind ividual factors as indicated in the magnesium bombarldme nts. F. Cadmium B3 cimbardment s The chart of the nuclides for tle cadlminum region is shown in fiL:-ure 53. Of t-ie several possible (d, a) products

NUCLIDE CHART OF THE CADMIUM REGION In'07 In'08 In'09 In'lO In''l In''2 In''3 In114 Inl'5 in''6 In''7 In''8 In19 30m 50m 4.3h 5.oh 66m 2.8d 21m 14.5m 104m4.23 49d 72s4.5h957 54m13s 1.1h 4.5hm 17.m 1P 3+, Y,EC,Y EC,ITPEC EC, YE', I T E IT I;IT TIre B 9 - Cd'05 Cd'06 Cd'07 Cd'08 Cd'09 Cdl"0 Cd"' Cd "2 Cdl3 Cd 114 Cd"'5 Cd 6 Cd"7 57m 1.215 6.7h 0.875 470d 1 12.39 49.0812.75 24.07 5.1 l12.26 28.86 43d 53h 7.58 3.0h S50m +: m/. S IBEC,Y,EC,Y EC IT IT r IT Ag'03 Ag'04 Ag05 Ag106 Ag107 Ag108 A 109 Ag I' Ag'II Ag 112 Ag113 Ag 114 Ag 115 | 1.I h 1.2h 27m 40d 24m 8.2d 44s51.35 2.3m 9s48.65 270d 24s 7.6d 3.2h 5.3h 2m 20m P Y /3EC/ I+Y EC, EC) IT /EC, L T I e- r f I I Fig. 53, Chart of the nuclides showin po: sibloe -ro ucts of deutCeron reactions on c1dmLiui-.

-l 3only four wrere seen in thie bombardmiie:-ts with 7. r8 Nev deuterons. These were thle 27-minute and"l.2-houP' Ag;, isomers, the 7. 6-day g a:-d the 3. 2-hour Ag112 all ~ and Y emitters. The maximum P-ray energies of Ag1 are 1.04 and 1+.2 >ev respectively. (180) Ten bombardments of cadmlium metal were obtained, most o0 ui1;Lich were for the purpose of developing the chemical Lirocedure (see Chemical Separations Section, page'4). Ou-ly in bomobardment 9 was it possible to det ermine absolute cross sections, with thle chemical yield determined gravim etrically. The data and derived quantities from this,ombardlment are presented in Table iHII. T'he errors are estimates of the standard deviat ioins The following parajra1 js deal *with th-e distinlctive features of thne Ibomb )ardimrent. Tar.=et Preparation This h-as b-en adequately discussed on page 70. Chiemical Procedure Ce c:iemical procedtjure'ias also been discussed previously on nase $1. Using SJuniernan and.:einke's iprocedure (89) in bomlbalrdmrent 9 the active silver did not completely exchange with the inactive silver chloride on The platinum g2auze, becau.se the gauze was dried in a 1100 oven prior to use. This has the effect of deactivating the silver chloride surf'ace. Therefore the target sol-ution was sc-ve nged later wiith -10 mg of inactive silver carrier ~sing thle recipitation frocedure (Part II of th.esis). The chemical yield was final.ly

Table fI. SuE;luary of oib ardiment Diata for Cd(d,)}Ag Bombardment go.* 9 Suppressor ring (volts) 1110 iD. C. supply (.a) 40 Slits J (in.) nQne Slits L (in.) vertical 0.25 horizontal O. 50 rar;,et? (mils of Olylar) 0.25 Tar et 2i position 7 J(mr ) 47.9 4(cm ) 25.7 substrate? eynolds I;rap Intecrator range D B lcnts) 279 T( sec) 4991:i:umber of Zap on layers 4 Ja-on soln. used (') 50 COl (Agt0) 61 00+ 500 C(02 ( g112) 20000+ 300 1113 ( g111) 372 + 8 Y O. 54+ * 02 112.41 KD (cnt s/,min/ a) 9*.9+0.1 106 0.01215 a2(Cdll11) 0.2',6 a3 (Cd113) 0.1226

calculated from t e wrei-, it of'iAg1 recovered. iA solute Beta Counting Discrimi-ation curves of thie silver obtained at a time corres-ponding to 3 per cent of Ag in the 1g104 - g 112 A!12 rmixture are shown in figure 54+, each point being corrected for decay. Once again there is no evidence for a slope in thle plateau, and t'he average deviation of the points on the plateau corresponded was 0.85 per cent standard deviation. iho measurement was made of' the extent of self ablsorption, (nd it was felt'haat this unknown error was far creater thlan the fiLm absorption correction, and hence the latter correction was not ascertained. nal-s is of becEa Curve The decay cur ve is shown in figure 55. The resolut ion was dol-le graphically. T! ree components were observed, the 27-mSinute AglO4 haviinc decayed before t.le counting was started.,iI0d 111 112 Cross Section for'"1.2't hr 104, 1; anld and Thi:e uncertai.nties associated wit-. bombardment 9 include tihe unknotwn error of self absorption. The decay curve resolution was handicapped because of the three components; two have comparable hialf-lives. The correction for nonconstant bombardment was not made. Unfortunately, additional c,-clotron time was not available to obtain more bombardments of cadmium for tile iurpose of measuring absolute cross sections.

DISCRIMINATION CURVES OF SILVER (CYLINDRICAL 4f PROPORTIONAL COUNTER) 13,200 — +-GAIN 100 3800 VOLTS GAN 75 13,000 x GAIN 50 12,800 + 13,200 3900 VOLTS, 13,000 6 z 12,800 rr ~ X\ \4000 VOLTS CL x 13,200 o 13,000 + 127800 T LIMIT OF COUNTING ERROR I (STANDARD DEVIATION) 13, 200 \4100 VOLTS 13,000 12,800 I0 9 8 7 6 5 4 3 2 1 DISCRIMINATOR SETTING Fig. 54. i)iscriur:ina ion curves of, r mixture of 3 per cent A-,111 and 97 per cent,.igll taken from cadrmiumr bombardment 9. The lii its of counting error sho;,n are t;Jice t:ie standard deviLtion of a sinJge o oint.

-1s7 - DECAY OF SILVER (CYLINDRICAL 4 PROPORTIONAL COUNTER) 0102 Ag 0 0 12 o10 Ag 0 L0L0 o 0 0 0LI..,I.I I I I o 10 %3 10 2.0 30 40 50 60 0-N0 TIME (DAYS) Cj)'-into,1:it.101,4u.l4, 3.2- lour.'1an7.6*rl 1omene2nts T0e uata rfere takern s ro. oadiniui boilbardment 90 10 2 Trhe cata',,ere taken'r'om cad:miumn bombardment 9.

T;ce values cac ilate-i,ithl t ie aid of equations (27) a?-d (39)i are (i.2hg'l104) O. 0017 + O.OOC2 barn at 7.8 +0.1 Mev l2;0120049 0.00002 barn at 7.8 +0.1 Mev cdj1) 0.0&t4 + 0.000C4 biarn at 7. +o0.1 iev The errors are biased upon the propag;-ation of the individual errors in each of t le factors, as indicated for t'e icmagrnesium boilibard.ct~enlts, and!hence represent an estimate of' the standard deviat ion. C. iLoss of;iecoil iiuclei.:len a tar-get nucleu.Ls is struck y a deuteron and undergoes a nuclear reaction, the product nucleus recoils wi;tih a certain aliount of kinetic enlergy.:;ome nu clei will thus be ejected from t.e target itself, and if some provision is not made to catch th;Cese recoil nuclei, tlhe ailparent cross section will be too low.- Fung (198) has studied the dependence of the recoil loss from t-1in (0.00025-0.001 inch) aluminum foils as a function of incident particle energy. He used protons, deuterons annd alpha particles, and found losses ranging from 10 to 30 per cent for energies of 40-380,iev. In the present -wvork it wras necessary to determine the thickness of target backing which will stop all recoil nuclei. (Fung reported that little or no nuclei are caught by catcher foils placed in front of the target.) An experiment

was performed in which cadmium metal was bombarded with Mylar films serving as recoil catchers in front and behind the cadmium foil' The cadmium was evaporated onto a 65/9g/cm2 Zapon film coated with aluminum; The cadmium was 97,/g/cm2 in thickness. One 0.00025 inch Mylar film was placed in front of two 0.00025 inch films followed by one 0*001 inch film placed behind the target. The catcher films were counted after bombardment with a Geiger counter. All the films except the second film in back exhibited the same decay curve. This one film showed the presence of a lO0-hour period in addition to the 10-minute N13 The amount of activity was too small to make any more accurate deterrmination of the half-lives. It was concluded that the range of the recoil nuclei is such that they were stopped in the second 0.00025 inch film. Therefore all the targets used in the absolute cross section bombardments were all made with 0*001 inch thick substrates to be sure of stopping all the recoils. He Energy of the Deuteron Beam The energy of the deuterons from the University of Michigan cyclotron has been measured by D. R. Bach. (60) On two separate occasions he found it to be 7.89 and 7.74 iev. Thus a day-to-day variation of 0.15 Mev may be expected. The dispersion of trhe focusing magnet,was 6.5 kev/min at 7.8 Mev, i.e., a point source of deuterons of various energies was spread out at the focal plane in a band of a widthL corresponding to the initial energy spread. For the 1/2 inch wide slit used in this work the energy

-190spread is`3 key for a point source, and in practice it is greater due to the finite slit at the window box. The MIylar scattering target (N of figure 1) was usually kept in the beam path during bombardment. The degradation in energy of the deuteron beam in passing through this 0.00025 film is negligible in comparison with the above energy spreads. The targets bombarded in te is work degrade the energy by approximately 0.08, 0.21, 1.6 and 0.08 Mev in the cases of riaagnesium, sulfur, titanium and cadriium respectively. (199) Thus the cross sections reported for these elements may be said to have been obtained at 7*.70.1, 7.70.1 70., and 7.d.0.) 1Iev respectively

'?e acc.iracy of tlie ireaos;.rement of cross sections by..lie proceduires evolve d in th'-ils work is depcendent on several factors. Larg1e syrstematic errors due to lack of suppression of secondary electrons in the beam determiniation and loss of recoil nuclei from tie target itself have been successfully overcome in thie present work. The problem of fast secondary neutrons was eliminated by -preparing targets of just the riT lit area to mnatch the col-li iator a perture. -.ince no observalble (0.6 per cent of i-32) 33 w's produced in t'he sulflr bo2rbar-a./ents fromi a possible (n, p) reaction, it was co ncluded t'at stray neti;ronls conta-minating the Ieulteron beanm did not -present a.proble m wh. en the tar,-et area does not overlap thze collimator. c ne: -iajor source of error arose from thle difficultis i.e resclvin7 the decay curve of t-le (d, a).roduct. T his has oe. ov,,er'com;e in tho v' i or.t iL c se i -tui Iiu U' from0 r.,ai:nesinum o-mbardments ( 15 —' o.r and 2.6-year components) yr tihe statistical anablosis of' a larg;e nuir er of decay points. AT e method of least sq-iares was applied, but thiis mietlhod was foulnd to be extre,-iev -tie consumnin'. It is suggested thiata this nt.etlod ibe coded iror use w ith an elec tronic co-mputer:suLch s i (i: -it:toic o.iyut;r).: ti nce the initial codin. were Acco;m lis Ce&, for a -:;eneral decay curve conl- istinl of n oo-ones of m u;-k:corn hrilaf- ives, the resol. ior. of complex decay curves could be acconmplisied in a rminimurn amount of time.

-192The present irteo:rator is th"e ot;er,major source of error. This instrumen-t was only linear at the very low countin- rates and t:e cal ibration was only trusted to 1 per cent. A new unit is under construction, but no data 2ave been obtained with it as yet. It is expected thiat with its completion 0.1 per cent accuracy7 can be achieved in the i;neas lr:onierent of t.e cyTclotron )beam.he ietal evaporator is very useful for obtainingi t hin, even tarr ets. It could be improved if a larger diffusion Opump on the mnetal eva,,orator would make it possible to o erate, at loz!er pressures (10-6 em rercury).rit'out t0e!o.sibility of introduction of air and otier gases by small leaks and outa'ss-ing of tlhe heater assel:b ly- Thius the forl. ation of oxides of thr metal to be evaporated would be alleviated. Lwo adbitiorna1 pieces of work could be done on the 4n. count in; procedu re. Tle -ef'ect of a cond; ctin;ng coating on th-le t in plastic filmrs'sed in 4trr counting coutlld be more precisely' evaluated as t ere is still disa:reement on this "oint in the literat;ure. (51, 4S) This could be done by Ipreparing two series of samples of eqial aliqu.-ots of a weak 2 emitter on plastic films of varying tlhickness. If or-e series ILad a condt:ctin; layer and t o-1e other did not, any discrepencv betwt;een thle two albsolute countingr rates 4.rould be evidence for t;me effect.,An implrovement could also be made in the i::easolrement of thle th-iclkness of tlahe,old plus a-lon films 3by securing more absolute (gravimetrically

-193deterniiLne~d) loiL-ts on thle.i[63 -ray transmission curve, fi~l;re 27. -The milaximnumi accuracy of the met-:ods developed in this -.work is represented by the Ig(d, La)Na cross sections in w.,tich 4 per cent accuracy (standard deviation) was achieved. 2vyen this errcr could have bee~h reduced if more points had been obtained on thie decay curve. The accuracy of the other cross section measureiments suffered from experimental difficulties which could be overcome as outlined below. The accuracy of the P(d, u)S cross section could be ir.-,roved if a tu.^in tar-,et of a salt suich as Na2SO 0 could be Prepared, an irvesti-ration of the exchange of phosphorous tracer and (d, o) prod ct were made, and enoup-: 33 tracer ao.rta) ir ann- pu-rified.,~ i" ter sa, estion is to set up a countin- apparatus near the cyclotron and atte-:ut to imeasure th.e yield of the 2.55-,inute 13 3 relative to`132 by countinu;p the enrjer,etic (3.5 iyev) r rays without performing c'lemical scearat ion, TI-,e Pi(d, u) reactio-ns could be studied to better- adva'ta;!ie if separated isotopes were obtained and tec hniques dcev eloped for their ifabrication in;to tliin films witl-out any loss. Thus tI.e dIecay curves could be analyvzed and 44 47 48 th'-e yields of,4c,c47, and 3c48 be determined..;s far as tle Cd(d, o)Ag cross sections are concerned, ti;e 1.rin-ciple exten:sion of th.e present work would be to devnelp a carrier-free se —aration of thie silver, perhaps sing an ion-excnlange y rocduree.

-194Finally, wh'en eno rgh accurate (c, u) reaction cross sections are obtained to, ive a general picture of the syestematics of the reaction, a correlation with the existing theories snould be attempted, e.g., the Bohr tieory of the compound nucleus. The variation of the cross section as a function of atomic number should re-:ilect the coulomb barrier. If absolute excitation functions wvere obtained, one could conmpare the cross sections at an ener?]y ccrrespondingr to a certain iercentas e above the threshold. Schott and K;ein he (200) retiort equ.al yields of (d, cL) prodclCts irom even-even zirconiun target isotopes, while the results of t le cadmium bombardlments in tthis o ork indcicate t:a t the yrields are not equal in thie case otf even-odd 3 and the even-even Cd target isotopes. i —us it would be interesting to note whether or not this odd-e en effect is fortuituous. if a trenid does exist the nagnitude of this effect could be investig~rrated.

PART II: BS ASSIII OF A9 1 INTRODUCTION A survey of the nuclide chart was made very early in thiis work to choose target elements best suited for study of the (d, a) reaction. One of the considerations in the choice was the possibility of obtaining new nuclear data by studying isotopes which heretofore had not been studied to any great extent. Thus information on the characterization of isotopes and absolute cross sections could be obtained at the same time. Cadmium was one of the first elements to be bombarded. An inspection of the nuclide chart for the cadmium region figure 53 reveals that several isotopes of silver may result from the Cd(d, a)Ag reactions. The 16-minute and 1.2-hour activities were of interest because little work had been done to assign mass numbers to these periods, and the (d, a) reaction would result in the production of Ag104. The opportunity thus presented itself to observe the periods associated with Agl04. The.radioisotope AglO4 was first reported by Enns (2C1) as a product of a (p, n) reaction on palladium. He observed the 73-minute and a 16-minute activity which he assigned to Ag102 and Ag104 respectively. Lindner and Perlman (20;2 reported a 70-minute activity in a silver fraction separated from the spallation products of antimony bombarded with 190 Mev deuterons, but they were unable to make any mass assignment of this activity. Johnson (203) observed a -1954

-19627 + 1-minute positron activity as a daughter of Cd104. Bendel and co-workers (2Q ) bombarded palladium with 12.Mev deuterons and found a 1.1-hour activity in the silver fraction separated from the target. Recently Haldar and Wiig (205) bombarded silver with high energy protons and purified the silver radiochemically. They noted a 25-minute and a 1.1-hour activity and attributed the former to Ag106 and/or Agl04. The silver fraction was "tmilked" for palladium daughter activities, and they observed the 17-day Pd103 using an x-ray proportional counter. The yield of this daughter activity decreased with a half-life corresponding to the 1.1-hour period of silver, and hence 103 they assigned the 1.1-hour activity to Ag. Additional radiation characterizations of Ag104 have been made by Cassatt (85) using silver produced by the (d, a) reaction on cadmium. EXiERINEiNTAL i;ROCEDURES A- Target In some of the work reported here natural cadmium oxide (Baker and Adams reagent grade) powder was used for the target material. It was wrapped in an envelope made of 1.5 mil aluminum (Reynolds Metal Co., Richmond, Virginia) and placed on the window box probe head (figure 7) for bombardment. Cadmium oxide electromagnetically enriched in Cd106 (Carbide and Carbon Chemical Co., Oak Ridge National Laboratory, Oak Ridge, Tennessee, Series AC Sample 85(a) ) was bombarded in a similar manner,

-197The mass analysis of these oxides are given in Table XII. Table XIII. Mass Analysis of Cadmium Oxide Target Materials Cadmium Isotope Atom per cent Natural Enriched 106 1.215 19.94 108 0.* 75 0.65 110 12.39 7.21 111 12'.75 39.16 112 24.07 11.85 113 12.26 4.14 114 2. 86 15.25 116 7.58 1.*80 In other cases cadmium foil of unknown purity was rolled to 1-5 mils in thickness for bombardment on the window box probe. B; Chemical Separations Since cadmium is a bi-product of zinc ores or from copper and lead refining, these elements were suspected as impurities in the development of the chemical procedure. Aluminum was also present from the envelope containing the bombarded cadmium. Considering the activities produced in these metals encountered from (d, n), (d, p), and (d, a) reactions there is the possibility of finding interfering activities of the following elements present

in the target solution: copper, zinc, gallium, silver, cadmium, indium, thallium, lead and bismuth, The chemical separation must be fast enough to permit the observation of the 27-minute period, and one technique was found especially helpful in this respect. Filter tubes (WilkensAnderson Company, Chicago, Illinois, Catalogue No. 8480B) fitted with glass wool plugs for filters make it possible for a precipitate to be filtered, washed aid dissolved in a matter of -2 minutes. The procedure adopted is shown in figure 56. To check the purity of the AgC1 precipitate, the following procedure was tried. A portion of the final AgC1 precipitate was subjected to further purification as follows. The precipitate was dissolved in 3 N HC1, Ag2S was precipitated with H2S, the precipitate washed and dissolved in 1 ml. of concentrated HNO3. The solution was diluted ca four fold and AgCl was reprecipitated with 1 ml. of 6 N HCl. A sample was prepared for counting and the remaining AgCl was subjected to a further repurification step. The chloride was converted to the oxide by digesting for 10 minutes with 2 ml of 6 N NaOH. This was dissolved in one drop of concentrated HNO3, diluted ca four fold with water, and AgCL precipitated as before with five drops of 1 N HC1, and a sample prepared ~for counting. Thus three separate samples were counted representing three different stages of chemical purification, and the resulting decay curves were compared. If any radioactive contaminant were present in the first silver sample (or in the silver sample repurified by the sulfide

-199C l?,~ICA=L SE iARATIONS Element separated: Silver Procedure by: Hall Target Material: Cadmium Time for sep'n: 15 min Type of bbdt: 7.8 14ev deuterons Equipment required:.Yield: -- 701,'oBarber filter tubes and Yi14eld: suction bulbs(WilkensDegree of purification: -104 Anderson Co., cat. no. Advantages: Rapid separation 84t0B and 1686) water bath micro bell jar Filtration apparatus (Tracer Lab, Boston, Lass, ) Test tubes, beakers, etc. Procedure: (1) Place 1 ml con 1H1NO3 in 100 ml high form beaker containinr 20 mg o? A;+ and 10 mg each of T1+, Pd+, Bi+++ Cu, Zn, Ga, In+++ carriers. Introduce the Cd target (-50 mg) anjd dissolve. (2) Dilute ca fourfold with hot water, add 2 drops 0.l~C aeresol solution and 5 drops 1 N'IC1. (3) Adjust (C1-) to 0.0025 Ni and digest 1 min. (4) Filter t.rough glass wool using filter tube and micro [Oell jar. IWash twice wit'h hot 0.1 N HN03. (5) Dissolve precipitate off of filter by recycling 4 drops of 6 N NH40H, catching tihe solution in a 5 inch test tube. (6) Add 2.5 mg Fe+++, filter and wash thie precipitate twice with 6 I,,iI. (7) To the filtrate from (6) add 3 rg eac. of the original carriers, substituiting Cd++ for Ag+. (8) Add Ii103 and 2 drops 1 N NH4C1 to precipitate AgC1 and recycle if desired. (9)'ilter the final AgC1 precipitate on a 1 inch diamieter filter paper using the filtration apparatus, vwash with. 1 Li I403, acetone, and ether. Remiarks: General references: li:. W. IJeinke, Chemnical P'rocedures Used ih Bombardment Wdork at Berkeley, AECD-2738(UCRLi-432) (August 1949). Fi;i." 56. i'rocedure L'or ti'e c —,emnical separation of silver from cadmium targets.

-200step) it is to be expected that further purification involving a different chemical method would result in the alteration of the ratio of the impurity to the silver. If this were the case the decay curves of the samples would not be identical. Since it was found that the three samples do, in fact, exhibit identical decays, it is concluded that the first silver sample (before repurification) is radiochemically pure. C; Counting The samples were counted using two "pancake" endwindow Geiger tubes (Anton Electronic Laboratories, Inc., Brooklyn, New York, Type 1001HOO) placed above and below the sample. The outputs of both tubes were connected in parallel and fed into a scale of 64 counting circuit (Nuclear-.hicago, Chicago, Illinois, Model 165). The sample was sandwiched between two copper absorbers of such a thickness as to stop the 1.04 Mev D rays of Ag A thin end-window chlorine quenched Geiger tube (Amperex Electronic Corporation, Hicksville, New York, Type 100C) mounted in a standard lead housing was also used in conjunction with a similar scaler. No absorber was used over the sample counted with this apparatus. A survey beta-ray spectrometer (84) was used to distinguish between positrons and negatrons and to obtain rough beta spectra.

-201 EXIERIMENTAL RESULTS Five bombardments were obtained for the mass assignment work. The first 2 were of natural cadmium foil, numbers 3 and 4 were natural cadmium oxide, and number 5 was of cadmium oxide enriched in Cd106 Bombardments 1, 2 and 3 were primarily used to evaluate the chemical procedure and develop the techniques, but some yield data were obtained from bombardment 1. The experimental results are summarized in Table XIV for bombardment 1, 4 and 5. The yields reported have been normalized to the 3.2-hour Agill acting as an internal standard. Numbers are given for the two different counting arrangements used. The decay curves were quite similar to those obtained in Part I of this thesis, and the curve presented previously in figure 55 is typical except that the 27-minute silver activity does not appear there. In the mass assignment work, samples were counted as soon as 18 minutes after bombardment. This permitted the observation of the 27-minute period. Table XIV. Relative Yields of Silver Isotope Silver Bbdt. 1 (Cd foil) Bbdt. 4 (CdO) Bbdt. 5 (Cd1060) mass No absorber INfo abs. Cu abs. No abs. Cu abs. 104(27m) 1. 833 1.347 33 4 41.8 104(68m) 0.493 0.300 1,3C6a 6.36 39.2 111 1 1 1 1 112 0.0178 0.020 o0.126 aHalf-life found to be 66 mmn.

-202The decay curves were resolved graphically with the assumption that the components were the 7.6-day Aglll 06), the 3.2-hour Ag112 207), the 27-minute Ag104(2C3), and an isomer of Ag104 of unknown half-life. Thus the initial activity and slope of the unknown component was found by successive resolutions of the gross curve until three lines were obtained, possessing the assumed slopes. The increase in the relative yield of the 27-minute and the 68-minute components which occured when the enriched cadmium was bombarded shows that these periods arise from the Cd106 target isotope. In fact, the relative increases in yield approximates the value calculated if one assumes that the thin target formula is valid. The comparison of relative yields from enriched and natural cadmium targets is shown in Table XV. Table XV: Ratio of the Relative Yields from Enriched (Bbdt.5) and Normal Cadmium Targets (Bbdts. 1 and 4) Silver Bbdt i (Cd foil) Bbdt, 4 (CdO) mass Theoretical no absorber no abs. Cu abs. 104(27m) 30.9 18.2 31.0 104(68m) 30.9 12.9 21*2 30.0 111 1 1 1 1 112 0. 635 0.71 0*. 63 For products of Cd106, the ratio of the relative yield from the enriched target to the relative yield from the natural

.203target was calculated to be 30.9. Similarly the yield of reaction products from Cd113 was expected to decrease by a factor of 0.635. These theoretical values are duplicated in the table as well as can be expected. Additional experiments using the beta-ray survey spectrometer indicated that the 62-minute activity decays by positron emission. The maximum energy of these positrons must be greater than about 1 Mev. DISCUSSION The assignment of the 68-minute and 27-minute activities of silver to mass number 104 was based upon the increased yield of these activities when cadmium enriched in Cd106 was bombarded with 7.8 Mev deuterons. The silver was produced by the (d, a) reaction. The energy of the deuterons was low enough so that no (d, an) reaction products were observed. Thus the 5.3-hour Ag113 was not present in the silver fraction. Cd116 is 7.58 per cent abundant and a 5.3-hour activity amounting to only 2 per cent of the initial activity of the 3.2-hour Ag112 or 8 per cent of the initial activity of 68-minute Ag104 would have been detected in the decay curve. Therefore the (d, an) reaction must'be only about 0.002-0*05 times less probable as the (d,a) reaction. Other work in this laboratory (200) has shown that no (d, an) products are observed when yttrium is bombarded with 7.8 Mev deuterons. An attempt was made to show specifically that no Ag103 was produced by the (d, an)

-204reaction in which the 17-day Pd103 daughter was milked from the silver fraction. No activity was observed in the palladium fraction, when counted with a Nal scintillation crystal, although a lower detection limit could not be set. The 68-minute activity is therefore Ag104, an isomer of the 27-minute Agl04. It seems that the 1.1-hour (66-minute) activity attributed by Halder and Wiig to gl103 on the basis of "milking" experiments is due to a different nuclide from the one observed in this work, so tihat two similar half-lives belong to adjacent isotopes Ag103 and Ag104.

ACi1JOLE',E ENTS The author is truly grateful to Professor W. W. Meinke for the valuable aid, cooperation and guidance that he rendered thoughout the course of this work. The suggestions and advice of Professor W. C. Parkinson, Professor P. V. C. Hough, and Professor E. F. Westrum are sincerely appreciated. The willing assistance of Harvey M. NIye in obtaining bombardments, of Walter E. Barrett, Almon G. Turner,Jr., Jerry L. Kochanny, Jr., and Mrs. Rosemary S. Maddock, in various phases of the problem was gratifying to the author. Thanks also go to Wayne A. Cassatt, Jr., for the many helpful discussions and to irs. Jacqueline Hall, Larry Hall, and Mrs. Oswald Anders for their painstaking help in preparing the manuscript. The author is indebted to the U.S. Atomic Energy Commission for partial support of this research, to the memory of Florence Fenwick for the receipt of a Memorial Fellowship, and'to the Regents of the University of Michigan for the two grants-in-aid received while engaged in this work. The help of the Isotopes Division of the U.S. Atomic Energy Commission for supplying both enriched and radioactive isotopes for use in this work is acknowledged. 204a

AFPOI{ DXI I Instructions for the Use of the Ti~etal Evaporator Refer to figures 10-14 for the valve numbers, pictures of the heater assembly, etc. To Turn hurij2s On 1. CLOSE valves 2 and 3 and open valve 1. 2. Turn water to diffusion pump ON. 3. Turn the fore pump ON. 4. Turn the diffusion pump ON. 5..ifter the oil is hot, feel the intake of pump to see if cooling coils are operating properly. CAIUTIrCO: Do not tighten sylplhon valves too tight — just a sixteenth of a turn after the valve has seated itself is sufficient. To install Filament and Catcl'er ( Substrate 1. Clamp the filament (Helix, basket, or foil or alundum crucible) between electrodes. 2. Check for short or open circuit. 3. Pilace the metal to be evaporated on filament. 4. Screw aluminum "ringstand" rod in spare hole in base plate and JUST BAI'LY tighten the hex lock nut. 5* Place substrate on aluminum frame and adjust its position on the l'ringstand"' rod. -205

-206To Evacul-ate Bell Jar'fait until diffusion pumnp has been hot for about fifteen minutes, then 1. CLOJSE valve 1 (3 is already closed). 2. CLOQ.E valve 4 (Hoke valve). 3. CFPEN valve 2 for only five minutes at a time. 4. Turn UN vacuum gauge and adjust heater current to 70 ma. 5. CL03E valve 2 and wait thirty seconds: then 6. OPJN valve 1 for one minute. 7. CLOLS valve 1 (3 is already closed). o. C0, valve 2 fcr an additional period of five minutes only. i.epeat steps 5, 6, 7, $ until thermocouple gauge reads 1.5 mv. 9. OPEN valve I and continue pumping until thermo-2 coulple gau indicates -5x10 mm. 10*. 0PiN valve 3 and LCLEt valve 2. To Use Cold Trap;ait until pumps,have evacuated to abou!t 104mm Hg, then 1. Aill dewar abouit 2/3 full and bring it up under the cold trap so that only the bottom quarter inch is ijmersed. 2.;After about two minutes raise the trap ii'hler. 3* In remioving the trap, allow it to reach OOC before opening the system to the atmosphere.

-207To Open IL Bell Jar 1. Remove cold trap and allow to warm up with valves 1 and 2 open and 3 closed (while pumping on the system). 2. CLOSE valve 2. 3. 0CiEN valve 4 (Hoke valve). 4. Remove bell jar and set it on a soft towel. To Turn Pump Off 1*. Make sure that valves 2 and 3 are closed. 2. Turn diffusion pump OFF and await until oil reservoir is cool to the touch. 3. Turn fore pump OFF. Use of Glass Bell Jar Haridle the bell jar with dry hands, always keeping at least two finers under the rim so it won't slip. Set the bell jar on a towel to prevent damage to rim and gasket. Always use wire bell jar guard. Coat the inside of the bell jar with a thin layer of stopcock grease for the purpose of reducing the amount of metal aticking to the glass. Thlis permits observation throughout the evaporation process, and the metal which did stick to th'e bell jar is easier to remove.

APPFEiL IX II Itaintenance of the 4nT Counting ChamberSee section on 4rt Proportional Counters, page 35, for a description of the chambers, and figures 15 and 16 for pictures of them. Center'Jire epacement: cndrical hamber 1. Grasp a one foot length of 0.002 inch stainless steel wire about 1/4 inch from its end using lucite forceps. 2. Clean this end in stainless steel soldering flux (Joseph iiyerson and )on, Inc., Detroit, Michigan). 3. Drop a. small globule of molten solder onto an asbestos board and immirediately plunge the end of the wire into it and let it solidify. 4* Thread the wire through the lucite insulator so that the sphere of solder catches in the recess. 5. Clean the wire as described below. 6. Replace the lucite insulator bearing the cleaned center wire in the chamber by screwing it in loosely (leave it out about- one turn). 7. d-emove the old solder from the hollow tube of the kovar seal with a needle point file, and clean the entire seal with water and dry. d. Thread the center wire through the hollow tube in the kovar seal and affix the seal to the chamber. 9. Thin the external connecting wire and insert it in the hollow tube alongside the 0.002 inch wire. -20 -

-20910. Stretch the center wire taut and solder the two wires in place. 11. Pull on the 0.002 inch wire to see if the connection is good. 12. Cut off the excess wire as close as possible to the solder joint. 13. Apply a globule of solder to the Joint to cover the end of the 0.002 inch wire. 14. Turn the lucite insulator inside the chamber until the wire is taut. Center Wiire Replacement: Spherical Chamber (53) I, Pulsh the slug about 1/2 inch out of its teflon insulator. 2. i[emove the old solder from the end of the slug with a needle point file. 3. Clean out the radial hole (in the side of the slug near the end). 4. L-oop tLhe 0.001 inch platinum wire throtgh the radial hole, out the end, and back in again. 5. Place thle end of the slug on a hot soldering iron, radial hole down. 6.!,kake a fillet of rosin core solder to secure the loop. 7. Pull on the ends of the 0.001 inch wire to see if the connection is good. Oo Cut off the excess wire and file smooth. 9. Clean the loops as described below.

-21lOCleaning and l'djustment of ~ Counter Chambers The following remarks are partially the results of experience and partially a verification of some suggestions made by Mr. W. H. Bradley, President of Nuclear Measurements "or:oration ( 53 ) and suggestions contained in the NMC Instruction LManual. If the counter is contaminated or if a voltage plateau can not be obtained (with the methane flowing at 2-3 bubbles/ sec) it may suffice to brush out t'le chamber with a camels hair brush to remove dust particles, etc., taking care not to touch the center wires. If, however, the chamber does not present a clean shiny surface after the brushing, it should be disassernbled and washed t-oroughly with distilled water, and dried in an oven (preferably at about 750C since a hotter oven will melt the Apiezon 7f Iax which seals the gas inlet tubes to the chamber). The O-rings are treated in a like manner. A more drastic cleaning is necessary only if the chamber is coated with a brown layer of oxidation products and/or if trouble is still experienced obtaining a plateau. They should then be placed on a lathe and the outer surface removed using in succession coarse carborundumr paper (3SiC-Lo. 1I) followed by the finer Si`-Jio. 400 and lastly SiC-Nio. 600, so theat the surface is made smooth and shiny. The chamber is then washled thoroughly with Tide solution, rinsed with distilled water and dried in an oven as described above.

-211it is very important that thie chambers never be cleaned or wiped with any aromatic organic agent or carbon tetrachloride, since tile counting rate would be seriously affected for a long period of time due to the continual emission of vapors of the agent which are capable of being absorbed onto the metal and O-ring surface. In cleaning the center wires it is best to remove thoem from the chamber. First a solution of lacquer thinner is used, followed tby a Tide solution and finally distilled water. To do this, a drop of liquid is forced partially out of the end of a medicine dropper, and thne drop is traced over the wire taking care not to contaminate the insulator with any of the liquid. The drop is allowed to fall on a piece of absorbent paper and the process repeated several times with a new drop of the same solution. The electrodes are finally dried thoroughly. Several other adjustmlents and corrections were made to the spherical counters. Platinum wire electrodes 0.001 inch in diameter (Sigmund Cohn, New York, No. 479) %were substituted for the stainless steel electrodes originally used. The original insulators for the electrodes were replaced with teflon insulators. Dow Corning /02 high vacutun oil was used as the bubblin, fluid. Finally the chamber was sealed ti'Ay so t iat it did not leak. The comnbination of the above alterations was found to result in plateaus -300 volts in length obtained with a gas flow rate of -3 bu:jbles/sec.

BIB L:.. ( )'t _'-'R' Y lo J. D. Luntz, Nucleonics 13, No. 2, 7 (1955). 2. J. D. Luntz, liucleonics 13, Lo. 6, 80 (1955). 3 0 F.. Frost and J. E. Putnam, University of California, aHadiation Laboratory Unclassified Report UC:IL-2672 (P1ovember 1954). 4. S. T. Clarke and J.!. Irvine, Jr., Phys. levO 6_, 231 (1944). 5. D. J. Hlughes et al., U.S. Atomic Energy Cormmission, Unclassified Document AE CU-2040 (May 1952). 6. R. F. Taschek, -private colmmunication to i. f;. Nleinke (January 1955). 7. G. A. Jarvis et al., Phys. Rev. 79, 929 (1950). 81 v? Ii. i. McMillan and. D. Miller, Phys. Rev. 73, to ( 194). 9. Ai. A. Glass, University of California Hiadiation Laboratory Unclassiiied Report UCRL-2560 (April 1954). 10.. A. H. Fanof sky and R. Phillips, Phys. Rev. 74, 1732 (1948). 11i C. P. Baker et al., U.S. Atomic Enersy Commission Declassified Document ASCi-21!!9 (LlAC-S.541) (June 1943). 12. J. i' Btaser et al.,.lelv. Fhys. Acta 24, 3 (151). — 13*.. La. amodt, V. IPterson, and R. PFiillips, khis. Rev. 88, 73Q (1952). 1L. i. H. Bura us et al., Uriversity of California;{adia-tion,5aoratory Unclassified Report UCHlL-4292 (February 195&4). l50 P. Kafalas, r.D. Thesis, T;)assachusetts Institute of Technology (iAugust 1954). 16. C,. V ileikowsky and Rt. T. auli, irkiv f. Fvsik 4, 299 (1952). 17. B3. Cork, L. Johnston, and C. >.ichman, Phys. iev. 79, 71 (1950). -212

13. E. L. Kel ly and'i, Segre, Fhys. Rev. 7, 999 (1'49)) 19. i:1;-. [3avison, Fhys. Rev. 75, 757 (1949). 20. 3. Cork,?-ys. Rev. g0, 321 (1950). 21. I1. r. &ittin s, Jr., Xev. Sci. Instr. 20, 325 (lrT49c, 22. I. A,.. Lewis and A,. C. Clark, J. Sci. Inst. 26, ~0 (1949). 23. C. J. I3rown, Los'llamos Scientific Laboratory 9eclassified feport L.-1059 (July 1949). 24, P. J. I-oole, J. 3ci. instr. 26, 113 (1949). 250. K;.:,T. Harils,!,;ound Laboratory Unclassified Keport XLK-630 (December 1951). 246. iA. Jtebbins II, I. S. Thesis, Purdue University ('ruust 1951). 27*. I.. Sihenk and iA.. Canfora, U.S. Pat. 2,Q549,022 (A`pril 19 51). 22. i,'. A. Higinbotham and o. Aankowitz, Xtve. S ci. Inst. 22, 688 (1q51). 29. G. I. B. Bouricius and F. C. Shoemaker, I ev. Sci. Instr. 22, 1l3 (1951). 30.?. J. Ziatts, {ev. Sci. Instr. 22, 356 (1951). 31.;.?i. 3uechner,,. i.. Dvans, and i!. S. iivingstonn, tU.3. tomic Energy Commission Unclassified Document AZCU-2 616 (l ay 1952P.ay 1953). 32. i. ilirnbaum et al., University of California liadiation-ab r atory Unclassified Report UCiJL-27 56 (November 1954). 33* a. 0. Lane and i. J. Zaffarano, Phys. Rev. 94, 960 (1954). 34. I. ~. D. Lewis and 3. Collinge, liev. bci. Instr. 24, 1113 (1953). 35. J:. CD ing, PhD). Thesis, University of Iichigan (1953). 36. J. E. [Aose,?-ev. Cci. Instr. 8, 130 (1937).

-21437. F. u.;. iurie, X:ev. Sci. Instr. 1., 485 (194k). 3 J. J. on. u. iu;ond, Rev. Sci. Instr. 6, 285 (1935). 39. -i. Kiuka, private communication (1954). 40* L, C. Olsen, C,. rmnith, and I. C. Crittenden, Jr., J. pp. Ihys. ppl.16, 425 (194. 41;. K. aldttell, J. Appl. Phys. 12, 779 (1941). 42. 0. Haxel and A. Sa. il outermLns,,. Physik 124, 705 (1948). 43. F. S. Houtermans, L. iieyer-Schfltzmeister and,. Vincent, Z. Physik 134, 1 (19523. 44. L. f'zeyer-Sctftzmeister and D. iI. Vincent, 2. Physik 134, 9 (1952). 45. i. Cohen Ann. il-y1sil (12th series) 7, n185 (1952. 46.. B. Smithi, British 5.tornic Energy Research Establishment Unclassified Teport;STSi —Ti.1210 (June 19 53). 47. I -. H. Seli-er and L. Cavallo, J. 1iesearch National r.,reau of Standards 47, 41 (1051). 48. C. J. Borkowski and`F. H-. Handley, Oak SIidr-e Lational Laboratory Urclassified.e1 ort C;:',L-1056 (April 1 51). 49*1. H. H. I:eeler and Ai d ciw we;-)~1, 1 ucleonics 12, Io. 7, 54 (1954). 50. L. Cavallo >rivate com..unication to.i. Meinke fAugust 1952). 51 B, D. Pate and L. Yaffe, Can. J. hemi. 33 510 (1955). 52. 5.i C. Hawkinrss,;.. F. r.erritt, and J. I. Craven, Proceedings of a Symposi um "iecent Developments and Techniques in the iaintenance of Standardst" held at tie 1National i-hysical Laboratory (Lay 1951). 53.:;7. HI, Cradley, private comimnunication (June 1954). 54.'. E. eible, private communication to. &. Cassatt (August 1953). 55. V. G. Hexroth, Los:Alamnos Scientific Laboratory Unclassified leport LA-1239 (;ay 1951).

56. E. L. Kemp, Los Alanos Scientific Laboratory Unclassified heport LA-1207 (July 1950). 57..4. i:. Cassatt,Jr., private communication (June 1953). 5c. a.:. Cassatt,Jr., private communication to W. iW. Neinke (March 1955). 59. M. Trott, private communication (August 1953). 60. D. R.:3ach, Ph.D. Thesis, University of Michigan (September 1955). 61. B. Rossi and K. Greisen,Revs. Modern Phys. 1_, 240 (1941). 62. HI. A. Bethe and J. Ashkin, in ExPerimnental Nuclear _hys cs, E. Seg-re, editor (John"-ieaandd Sons, InC,.?ew York, 1953),p. 251. 63. A. Ad. Smith, Electrical IMeasurements in T1eore and %ppic atlo~i (rccraw -Hi1 Book -Company,.nc, ~ew York, 194F-), Chap. III. 64* E. F. ~.estrum, private comrimunication (1955). 65. F. Hudswell and C. J..Mandleberg, British Atomic Energ-y Fesearch ZEstablislhment Unclassified Report ER2-C/R-~:61 (Mrarch 1951). 66. H. U. Thiomas, private communication to 7..J leinke (October 1953). 67. E. N ielsen and A.?J/einstein, Brookhaven lational Laboratory Unclassified leport BDL-259 (S-19) (Septemnr!er 1953). 6:. I. B. Ackerman, private communication (February 1955). 69. Lo E. Glendenin et al., Radiochemical Studies: The Fission ProduFcts -(cGraw-Hil' Book C., Tnc., iEew York, 171) National Nuclear Ener-ry Series, Plutonium l-roject itecord, Vol. 9, p. l129. 70. d. Cohn and IIU 7. Kohn, J. Am. Chen. Soc. 70, 1Q86 (194`!-). 71. R. Bouchez and G. Kayas, Compt. rend. 228, 1222 (1949). 72. q'g L, Arons and A. K. Solomon, U. S, Atomic Energy Commission Unclassified Document AECU-2661 (1952), 73. J. Beukenkamp and o. Rieman, III, Anal. Chem. 22, 582 (1950).

-21674s.. G. Kayas, Compt. rend. 225, 1002 (1949). 75. V.'J. Linnenbom, J. Chem. Phys. 20, 1657 (1952). 760.. 1,ieinke,,J. 3 Atomic nergy Commission Declassified Document AECD-2738 (UCRL-432) (August 1949). 77. L. u. Y McIssac and. Voigt, Iowa State College Unclassified Report ISC-271 (June 1952). 75. *.J. Scott, Standard l; etlhods of Chemical Analysis (D. Van No trand Cb, nc. — ewYor, 193),. 931. 79* 7*. F. Hillebrand and G. 5. F.F Lundell J. Am. ChIem. oc,.42, 2609 (1920), 80. J. a. Haskins et al., Prhys. Rev.?8, 876 (1952). 81. IT s; Duval and:. I. Kurbatov, J. iA. Chem. Soc. 75, 2246 (1953). 52. J. D. Gile, -,i. Garrison, and J. G. lamilton, J. Chrem. -hys. 15, 1685 (1950). c3. N. V. Sidgwfick, The Chemical Elements and Their Compounds (Oxford University Press,lon, P-077 F84. 7. 7. IJeinke, 7. A. Cassatt,Jr. and K. L. Hall, U. S. Atomic Energy Commission Unclassified Report AECU-2944 (June 1954). 55. *7..* Cassatt,Jr., TPh.D. Thesis, Unive rsity of Miichigan (September 1954). see also U.. Atomic Energy Commission Unclassified Report AZCU-2955. s6. Hi. G0. -icks et al., California Research and Development Company?-eport LR2L-65 (Decemnber 1953). 87. H. G. Hicks, private com.,urnication to 7.r ~ 1.einke (June 1955). L1t. 11. I. 1Haymond et al., J. Chem. hys. 15, 391 (1950). 89. D. i' S inderman and 7.1. 7:. einke, Science 121, 777 (1955S). 90. A. Langer, 5. C:iem. Phys. 10, 321 (1942). 91. G. lRouser and ~. F, Hahn, J. lAm. Chem. Joc. 74, 2398 (1952).

920. A S iewton, reported in "C"herical pirocedures Used In Bombardunent Work at Berkeley," by ~,. i,. Meinke, i. 3o. Atomic ~,nergy Commission Declassified Document ECSD-2738 (UCRL-432) (August 1949). 93. G. u i ianov, Preliminary ieport -o., 13, Nuclear Science Series, FNational -e~search Council (July 1953). 94. 0. E. ~iyers,:ucleonics 5, o. 5, 37 (1949). 9>. J.. D, ayly, Ca.) J Research. 2. A, 520 (195C). -96. L. ii. Zumwalt et al., 3cience 107, 47 (1948). 97. C. V. Cannon, reported in Conference on Absolute i Counting, -Preliminary Report No. $, 8 i<,clear cScience S3eries, National Research Council (October 19350)......:;. Loevingee,;ev.:ici. istr. 24, 907 (1953). 99. 1b.. Gra7, Britt. i. adiology 22, 677 (1949;. 1CO. J, tV. Dunworth, iev. Sci. Instr. 11, 167 (1940). 101. J. Barnothy arid 1Sk. Forro,;i.ev. ci. Instr. 22, 415 (1951). 102. UF. Deutscli, reported in oonference on i.isolute, Counting, ~ireli2m inary Aeport ro. 8, NSuclear Science Series, s. ational'Research Council (( ctob er 1'950). 1030.. S-leuler and r.i J. Goldsmith, Zperimental Nucleonics (ainehart and Co., Inc., -ewkV 32 Yor, 5T 104..,.. ann, iational Aesearch Co ncil of Canada;eport C ifif-40 (ca. 19 50 ) 105. I. A. urmnwalt,. 5. Atoric Jnersy Co.n.is'sion ieclassified D oc um-en:t o 9-13r46 (December 1947). 106.. Zuwat,. oAt lic n ner, Cornnmission r classif ied Doclment i,,U- 5, 7 (;e teme r 1 49 ) 10{7.. Jaji, -. a;aktisaka, and A tiyake, J( *lays. Soc Japan:, 550 9 (19>3). 108.,. chwFe'itzer r-nd -t.. t nein, -ucleonics 7, No'. 3, 05 (190).C 109. G. -R. Jalton, J. S. Thompson, and I. F. Croall, British:itomic EJnergy:esearch ]3stablishment Unclassified N~eport 1A3-,E:/2-1136 (]!arch 1953 j.

-21'llOo Ps I Ucpson i tcieonSics i 12, t_, 49 (1T53)* 11o1.,. C. ia-pas, t. 2. tornmc WEnergy Co-l-missi on bi-classif ied d oculment i-.3Ui-2C06 (Septe-mbor 1953). 112v. }. 8. Sjnc^weitzer and I. *'ell s, J. hys. (hem. 36, 541-4 (1952). t 113 *...ieed,Jr., private communication (January 1954). 114. 1.',t 3. L;yon and 3. 4- Kteynolds, Oak l ide'National iLab oratory.-cnlciassifieMd Ri-1e ort QOiKl L-955 (Januar r 1 1).!11. $ > I.rc 1, b, elv' IUhys.::;ta 26,.663 (1 53). 16. 3. 1. Gleason, J j. Tayl and D. B..'aboern, Nucleonics t,Lo.O 5, 12 (a1y 171). 1170 3. S. 2iudstamr, ucleonics 13, 1o. 2, 64 (1935). 1l1;).'. L. losen and,. i)avis, Jr., U.... A-tomic Energy Cormmission Ce clissified Cji ocument nAt 2D-31h4 Ol Veem,oe r f3 ) 11O.':it. U. Baker an-d.:a tz, ixucleonics 11, o. 2, 12 4 (1953;* 120. 0. F. rurtt, xtucleonics 5, o. 2, 2~ (1949). 121.. J o i. ervik anLd C..tevenson,'ucleonics 10, o. 3, 1: (1352).,.. _, u. iAlemiia arniu.. l an son, ahys. Iev. 73, 106 + (.1941) 123. A,. _. 3atze1, * r. T. Crane anr G..'" lley/ Plays. _iev. 71 93) (193 ). 124. E. 3.~ vervik, University of California 1ladiation Laboratory r ciass iied i e[dort UC r L-2 542 (ril 1954). 125.,. i. B. elaney, hy.S Reev. 21, 158 (1951). 12. 0.;.. LibUby, etSensitive Radiation detection'Techniques for Tritium 4aturail Rad-ioactives, and namrima Radiation" ATI-133901 (December 1951). 127. -i. L iidinoff, reported in Conference on Absolute 3 Counting, Treliminary;ie port'o C, U* cl ear,Jcience.eries,?VNational flAesearclh Co.ncil (Ucto'ber 13 C0) 12.... Bernstein ard (1. Bailentine, ev. ci. ~nstr. 1, 1 W (19':0).

12. Ii.. C rane, Lucleonics 9, N:o. 6, 16 (1951). 130.?. K. Clark, Rev. Sci. Instr. 21, 753 (1950). 131. G. Failla and {'.!I.; ossi, U.. iat. 2,659i826 (;'ove m'-er 1953). 132. 7. Gross,. eutsc l, and G. Failla, in Annual Rleport on -esearc'ih roject, NY-&45A82 (April 1-95+4) 133. -. io. uttle,Jr., i. 3. A itomic lnergy Commission Unclassified lejort IECU-2524 (2August 1952). 134*:. L. Graf, C. A.. Comar,and 1. D..hitney, iucleonics 9,?io. 4, 22 (October 1051). 135.?. L. Curtis, i. L. Clark,and J. J. Dauby, y ound Laboratory irnclassified Report uiL>-O1 (Deceribr 1'952) 136. 7. L. Curtis, I ond Laboratory Unclassified Report 1iJ-603 (Se9tember 1951). 137.. C. Conway and J. (:. Rasmussen, University of California Radiation Laboratory Unclassified`ceport UCiR-2075 (January 1953). 13. J. A. iader, A P. H-lageeand L. -. etter, ucleonics 12, o, 20 (1:54). 139 ~.. i. Ieliger, I.ys.:ev.,-';, e 408 (1952). 140. -i, A,. Wlartin, B.,.iven, and 1. 9. Taschek, Phys. 1ev. 93, 1%)3 3! o54). 141. 8. 1i. Aeler, J. Aci. Instr. 30, 2 6 (1953). 142*.'. C. Conway and J.. Rasmussen, University of California adia ion Laboratory Report U3C;L-2076 (January 19 33). 143. T r Lindner and i orne, Cali'fornia Research and;Jevelopment Company;ieport?TA-22 (January 1953). 144.:I. H. Iac]-regore,:h.. Thesis, University of ilichigan (1953). 145. G. A..: awyer T esies, un iversit y of;: icig an (1950) 146. iassac husetts Instit-ite of'Tec'lnolojy Progress i'Aeport L:0o, 29 ( - iLay 1953). 147 L.'. Pate ani. Yaff~, Can., C,:em. 33, 15 (15).

-22014S. C. ). iPate and L. YaKfe, Can. J. Chen. 33, 929 (1955). 14D. 1_. *.. Fry and.* v. Overman, U. 3 totormic.nr gy Cowrmission KJeclassified Document I'CD-i 1 00 (January 194L) 150. J. Drown et al, LCv ev. ci. Instr. 19, 118 (194-) 151. J. Backus, rhys. dev. 6(, 59 (1945). 152.0 Pi iLanger and 2.. Cook, Rev. Dci. instr. 19, 257 (1943). 153.'i. ilatis, Beta an"d amma -a pectroscoy, K. Sie,,ahn, editor T - l orth o11od i —u. co o., rn t JstIerdarm, 1955),;'. 260, 151.o;i. C rown, Iil 1n i. *ID illo.L iito', v. A ci. i-nstr_1,:20 ( 194 ) 1i5. l. o oten, ev,:ci. Inst r. 16 26 (1945).. 16. T. Harris and J.. easley, J. k0pt. Soo. r. 42, 634 (1952). 157. 1. T2. 1;- THesil, h'nive; sity of i;:iciigan (1953). u,15.. S. H ull,.. rtomic nerrg Comraission Declassified )ocument AE: -: i ll (Rarch?94 i). 159.. 2. -;ann and i; 1. l r.;esearch lational 3ur. St aindar ds 50, 197 (1953). O.. 15 L. Du; an and K.'L. Smith, Science 116, 305 (1952). 1L31.. T. Gilbert, Jr, r cience 4, 637 3(1951). 162. D. A1l1iday, intrcudctory luclear hvrsics (John Ailey 2~3,:F1'C U~ os, +Inc, e ov r /....... 133. 1 ohn * hys. 1 ev. 65, 63 (1944). 164. T. P. T o unan, The "rans;ranimn c,Le-ients:;esearch'apers ca ilr7I'ok i ric*. 7T rw, 1ork949) i'ation.l l uclear cnerjD3r <eries, ilutonium Pro ject -ecord, Vol, 143, p. 1655. l 5. M.;:alv n e) t al., Isotouiic -Carbon (To o,lcy aand Dons, Kc. iC rk T T eW rnKix IIO!I;...' aker,?ound L.aboratory T nclaSsified Report i:L-65c (I arcih 1I52). 1 - 7:7.J1t 1., 1:, uJ*lr tn. 1.1it * cit., p. 320.

-221165. e o J. Xain ate r,.... ~ i, 7CIeo^nitrs iJo 2, 60 (1947), 169... 5 _ S inwate an 2.;. tt, iucleonics 2, io. 1, 42 (1 47). 170.. ~ 7. ito e!1 and _. 7. Laner, i-hys. tev. 53, 5 (1732)5 171. ~.I zilard &?'iT.:L. C:allmers, Yatire 135 98 (1935). 172: l di 2t al. P ro. o io.(odn 1.49, 522 (15/ E —O 173.. I. il ver, hyN. no.v. 76, 3j9 (1949). 174. 3 u dussoll and. t:attenber>; 7 * 7. tric -ner y (CorrminliCon Jecliassif ie.Iocumre nt A:'iD-2926 P(ctobcr 1945) 1'75. 7. i. Jensen et al., ihvs.:v. 5, 112 (1952). 176.. 1.. 5 hlo 1'3.!oltzman, an. Y Fan, iys Rev.' —' ( "'51). 177. 0:[. esteriark, irh s. ev. -, 5 73 (19 2) 17.".c. viautJi O t v met 15, 213 (193t). 1 It. L1bek?. [. ilsen and 2., - ieisn, in vs. dev. 1. 5.a.'o ) 11ader, I ierla, draa d;. 7. Jeaors,evs io rn i'hys. 25, 469 (1953 ) 1- 1.,. I'. iornvak and T Coor, i hv$ xev. 92, 675 (1953). i1(2". t'. Qierr. - 1 i'ller, -hhs. vyr,; 93, 1076 (1954).::L.; }:ie rpra an,:..'':. h 2 93, ~ ~ 1? * (. arpak, 7,. phs. et radium 16, 62-S (155). 1:5. L.. LasIet;, ~ cs c':,. 75,.5 I >r2..'4~' 1 6. 0. t. 3reb, 4hys. dcv. 1, 469 (1 1).:7 ~ ~ Loc.kett and,.. Thomas, 1Suleonics 11, SO 3, 14 ( 1953 ).. t4 i 1 v 7 0 (lOiC

-222189. J. J. Cobble and W. 7I. Atteberry, Phys. Rev. 80, 917 (1950). 190. R. Wilson and G. R. Bishop, Proc. Phys. S-oc (London) 62, 457 (1949). 191.'1.-K. Sinclair and A,. F. Holloway, Nature 167, 365 (1951). 192. 35.. Van Voorhis, Phys. Rev. 49, e9 (1936). 193. J. Tobailem, J.- phys. et radium'16, 48 (1955). 194.' J. W. Irving, Jr. and E. T. Clarke, J. Chemar Phys, 16, 686-w90 (1948). 195.- F. 0. 3artell and 5. Softky, Phys. Rev. 4, 463 (1951). 196. A. G. W'Jorthing and J. Geffnert Treatment of Experimental 1Data'(John Wiley and Sons, Inc,,New 1943) Chap. II. 197. w. F. Biller, Ph.D. Thesis, University of California Radiation Laboratory Unclassified Report UCRL-2067 ( December 1952 ) 198. S. C. Fung, rh.D. Thesis, University of California iRadiation Laboratory Unclassified Report UCRL-1465 (September 1951). 199. WJ A. Aron, Bo G. Hoffman, and FP C, Williams, Atomic Energy Commission Unclassified Document AECU-663 (UCRL-121) (May 1951). 200. G. L. Schott and W1. 7. Mleinke, Phys. Rev, 8_, 1156 (1953). 201. T. Enns, Phys. Rev. 56, 872 (1939). 202* M. Lindner and I. Perlman, Phys. Rev. 78, 499 (1950). 203. F. A. Johnson, Proc. Roy. Soc. (Canada) 46, 135A (1952)c 204. 71. L. Bendel et al., Phys. Rev. 0, 888 (1953). 205. B. C. Halder and E. O. Wiig, Phys Revo 91, 1713 (1954). 206. E. P. Steinberg and L. E. Glendenin, Radiochemical Studies: The Fission Products (McGrawHHill B6ok Co. 30 c.7 - eW YT rk, I95)7,'National Nuclear Energy Series, Plutonium Project Record, Vol. 9, 207. K. Shure, Ph.D, Thesis, Massachusetts Institute of Tech-nology (MIay 1951).