I* CARRIKE 7RSE SEPARATION 07 COBALT II. THE DISINTEGRATION 07 THE RADIOACTIVE

ISOTOPES Co54, Co57, and Co5* III. CHEMICAL AHALTSIS BT CICLOTROH ACTIVATION

DISSERTATION

Freaented In Partial Fulfillment of the Requireaenta for

tha Dagroa of Doctor of Philoaophy in the

Graduate School of Tha Ohio Stata

UnlTeraitjr

JAMES L. DICE, B. So., M. Sc.

Tha Ohio Stata Uniaereitjr

1953

Approved by*

/ / A * i,V / - u . i./r V - Adviaera 1

ACKHOULXDGHEIT

The writer wishes to express his sincere appreci­ ation to Professors J. D, and M. H. Kurbatov for suggesting this rasoareh and for their continuing advice and encouragement throughout the course of this work; to Mr* Mitsuo Sakai and Lieutenant Willis Anderson for their assistance in Spectroscopic studies; and to my wife, Marie, for her patience and encouragement* 11 TABLE OF CONTENTS

I. Carrier Free Separation of Cobalt**. 1

Separation of Carrier Free Cobalt

from a Manganese Target...... * 1

1* Introduction...... 1

2* Experimental Procedure and Results. *...... 5

3# Discussion*...... *...*•* 20

4* Summary ...... 2U

Separation of Carrier Free Cobalt

from Enriched Isotopes of Iron in

Iron Oxide •••*•..•...... 26

1* Introduction*...... 26

2* Preliminary Investigation and Results...... 28

3* The Chemical Procedure...... *...... 33

4* Summary ...... 36

II. The Disintegration of the Radio— c 4 gj active Isotopes Co , Co f

and Co'’** ...... 37

Instrumentation* ...... 37

1* Thick Lens Spectrometer...... 37

2* Solenoidal Spectrometer *••••••*.•• 53

3* Scintillation Spectrometer •••••..* 55 ill TABLE 07 CONTENTS (coat.)

JEa e a

B* Spectroscopic Studies...... 58

1. Introduction*...... 58

2. Sample Preparation • •...... 60

3. The Iaotope C o ^ ...... 64

a. The Continuous Positron Spectrum of Co^® 6 4

b. The Photoelectron Spectrum of C o 5 6 ...... 70

c* The Internal Conversion Elec­ tron Spectrumof Co56 ...... 77

d« The Gamma Ray Energies and Relative Intensities Measured with 0 Scintillation Spectro- notor...... 79

e. Summary of Results and Discussion • •..... 87

4* The Isotope C o ^ ...... 93

a* The Continuous Positron Spectrum of C o " •..*••••••••• 93

b. The Internal Conversion Electron Spectrum of Co^" .... 95

o. Summary of Results and Discussion ...... 97

5. The Isotope 98

a. The Continuous Positron Spectrum of Co^8 ...... 98

b* The Internal Conversion Electron Spectrum ofCo56,58 101

o. Summary of Results and Discussion...... 101 ir

TABLE OF CONTENTS (cont,)

Eagft

CHAPTER III. Chemical Analysis by Cyclotron

Activation...... 106

A. Introduction...... 106

B. The Beryllium—Copper Alloy...... 109

1. Experimental Procedure and Results...... 109

2. Discussion...... 113

3. S u m m a r y ...... 114

C. The Nickel Ore ...... 115

1. Experimental Procedure and Results ...... 115

2. Discussion ...... 123

3. Summary ...... 12 5

Bibliography ...... 126

Autobiography ...... 129 X, CARR HR FREE SEPARATION OF COBALT

A . CARRIER FREE SEPARATION OF COBALT FROM A MANGANESE

TARGET 1. INTRODUCTION

The bombardment of a unganese target by alpha

particles yields several cobalt Isotopes according to

the following reactions*

Mn55 (a, n) G e 58 Mn55 (a,2n) Co57

Mn55 (a,3n) Co56 In order to study the disintegration schemes of the cobalt

Isotopes formed it is necessary to separate them from the

target material* If the nuclear spectroscopic investi­ gation is to be limited to a study of the gamma emissions carrier may be added to assist in the separation. In this case the absorption effect of the residue on the energy of the gamma emission is insignificant. However, the usual conditions require a thorough investigation of both gamma and beta emissions. The presence of carrier decreases the specific activity, alters the observed energies of the beta emissions and Increases scattering especially in the low energy region of the beta spectrum*

To avoid this not only must the final sample be free of reagent Impurities which tend to accumulate during the processing, but the specific activity, the ratio of the radioactive atoms to the total isotoplc atoms present, Must be high. Therefore the study of the behavior of cobalt iaotopca in concentrations of tho order of 10“^ 1C v«i undertaken for the purpose of preparing a carrier free sample of the radioactive isotopes of cobalt from a manganese target*

After preliminary investigation of some of the more common analytical methods of separation, it was decided to attempt the preparation of a carrier free sample of the cobalt isotopes using an organic complexlng reagent*

By nature of their structure, many complexlng agents exhibit considerable selectivity in their reactions*

Alpha-nitroso beta-naphthol, one of the better known reagents in this category, was favorably considered for separating the cobalt isotopes from the manganese target material. It is best known for its separation of cobalt from large quantities of nickel. It is also applicable for quantitative separation of oobalt from mercury, anti­ mony, aluminum, sine, manganese, calcium, barium, strontium, chromium, lead, cadmium, beryllium, and arsenic (l)* Sil­ ver, tin and bismuth interfere with this reaction. Alpha- nitroso beta-naphthol exists in two Isomeric forms which react with oobalt salts to form inner complex oompounde of the following composltieht Thi purple-red oobaltie salt ii Insoluble In aeldi, but

Is successfully precipitated only from s weakly sold, aautralf or ammonlacal solution, sines tbs eobsltie sslt is derived from tbs tautomeric quinoxlme form of tbs ra- agent* Both forms exist together In equilibrium, but in strongly sold solution tbs equilibrium shifts to tbs phsnol form; tbs quinoxlme form prsdomlnates In s weakly acid solution (2). Illnskl and Knorrs (3) first observed that alpha-nitroso beta—napbthol rsaetad with solutions of eobalt salts to form a brisk rad precipitate* Tbs com­ position of tbs prseipitata formed in tbs reaction between oobalt salta and alpba-nltroso beta—napbthol corresponds to tbs formula (C^qH^0 2 M)^Go*2 H2 0 in which cobalt exists in tbs trlvalont state* The exact composition remains somewhat in doubt. The reagent commonly employed for precipitating oobalt is prepared by dissolving alpba- nltroso beta—napbthol in ethyl alcohol, acetone or $0% acetic acid* Hydrochloric acid is the usual medium In which the precipitation of cobalt la carried out, but may be replaced by nitric a d d so long as it contains no oxides of nitrogen* Tamlc and ooworkers (4.) have studied the effect of pH upon the precipitation of cobalt, iron, vanadium, palladium, and uranium* Their report is as follows s UmImI pB fll Sfllailgfl Co less than 8.74 Cu 3*96 to 13*2

To .95 to 2*00

▼ 2*05 to 3*21 Pd loss than 11*82

U 4.05 to 9.38 Thoao studies vtrt nodo uolng macro quantities of the elements in question* For application in the field of nuclear spectroscopy it was necessary to extend this investigation to quantities on the order of 10“^ M*

The effect of a number of variables on the percent of oobalt operated from manganese have been studied and a satisfactory method of separation of oarrler free cobalt from a manganese target has been developed. 2. EXPERIMENTAL PR OCX DUES AND RESULTS

a. SAMPLE PREPARATION

A sample of manganese (99*9 pareant purity) ob— taluad from Battalia Maaorial Inatltuta vaa machined to fit tha targat mount of tha cyclotron at Ohio Stata Uni­ versity, This sample, valghlng 2,34 grama, vaa bombardad

six hour# with alpha partlelaa of anarglaa from alghtaan to twenty Mar. Tha aampla vaa earafully removed and

atorad for approxlmataly thraa waaka during vhleh tlma any abort lived lmpuritlaa disintegrated, Tha raaotlona expected by bombarding tha manganasa targat with alpha partlelaa of tha energy ranga alghtaan to twenty Mar arai

(a, n) Ce58

Mn55 (g,2m) Co57

following tha waiting period in which tha ahort lived activities disintegrated, tha aampla vaa dissolved in 25 ml, of 6N hydroohloric acid and evaporated to dryness, Tha residua vaa than redissolved In 250 ml, of 3N hydrochloric acid and atorad In a 250 ml, volumetrle flask. This solu­ tion became tha parent material rrom which all experimental aaaplaa vara taken. For purposes of establishing a con­ venient oounting rata and later to aid in determining tha loss on glassware during the procedure, five 2 ml, portions of tha parent solution ware placed in individual counting dishes and slowly evaporated to dryness using a hast lamp.

These five aaaplaa yielded an average counting rata of 1440116 counts per minute.

b. GENERAL TESTING PROCEDURE KOB ESTABLISH­

ING CONDITIONS OP CARRIER PEEK SEPARATION

OP COBALT A 2 ml. sample of the solution containing tba activated targat vaa plpatted into a 50 ml. beaker and dllutad to 12 nl. with triple diatilled vater and dilute hydrochloric acid of auch atrength to give the deaired acidity to the aolution. The pH deterilnatlom vere made with a Leeda and Morthrup pH Meter. The aolution vaa then heated to approximately eighty degreee centigrade and 4 ml. of alpha—nltroao beta-naphthol aolution vere gradually added vhlle stirring. The aolution of alpha—nltroao beta-naphthol oontalned .5 grama of alpha-nltroao beta- naphthol per 100 ml. of fifty percent (by volume) acetic acid aolution. The aample containing the complexlng agent

vaa permitted to stand for a period of time (see Table I) and then extracted with three 100 ml. portions of benaene saturated vlth dilute hydrochloric acid. The three non- aqueoua phases vere combined and vaahed vlth dilute hydro­ chloric acid. The benaene phase, containing the active cobalt complex, vaa reduced in volume by evaporating the benaene and transferred to a platinum crucible. At thla point the solution vaa evaporated to dryness and the co­ balt complex eonverted to cobaltlc oxide by burhlxg in air* The oxide was then treated with dilute hydrochloric acid and transferred to a counting dish of an approximate volume of 4 ml*, 2 centimeters in diameter and 1—1/2 centi­ meters deep* This solution was evaporated to aryness*

The aqueous phase, containing the manganese chloride and any cobalt not complexed, vaa also transferred to a count­ ing dish and evaporated to dryness* The counting rates per minute for each of these two sources and the approxi­ mate counting rate per minute of the original sample were used to determine the percent of cobalt removed by the complexlng reagent and loss on glassware during the pro­ cedure* The counting was done with an end window Geiger—

Muller tube, whose window thickness was equivalent to

2*0 fflg/cn^ of aluminum, and a Potter Decade Scaler,

Model 341.

c. EFFECT OF CONTACT TIME ON PERCENT OF COBALT

EXTRACTED i? ROM A 3N HiDKUCtiLUEIC ACID SuLU-

TI UN OF THE MANGANESE TARGET

In order to establish an experimental procedure thai was both effective and rapid, the first tests were designed to find the minimum length of time required for effective separation* Two series of experiments were carried out, one in which the test aolution was made up at 3N acid concentra­ tion and a aecond in which the acidity was »5N, In each case the percent of extraction of cobalt for five time

Intervals, 1/4, 1/2, 1, 2 and 4 hours was investigated* Tha 3N teat solutions vtre prepared by taking 2 nl. of

the parent aolution and diluting to 12 nl. vltn 3N hydro—

ohloric aoia» This aolution vaa then heated to approxi­

mately 80 degrees centigrade and treated with 4 ml• of

alpha-nitroao beta-naphthol* The oomplexlng solution con­

sisted of *25 grama of complexing reagent dissolved

100 ml* of fifty percent (by volume) acetic acid solvent*

After the predetermined length of time the solution was

transferred to a separatory funnel and washed three times with 25 ml* portions of benaene saturated with dilute hydrochloric acid* Both the aqueous phase and the benaene

phase were reduced in volume and transferred to counting

dishea. Each was evaporated to dryness and counted for

three minutes* To insure reasonable accuracy and to detect

any gross errors, three samplea were prepared for each test*

A maximum of 4$ of the activity was removed by benaene ex­

traction after standing one or more hours* The results of

this experiment are shown in Table I and Figure 3.*

TABLE 1

Effect of Contact Time on Percent of Cobalt Extracted from

a 3N Hydrochloric Acid Solution of the Manganese Target*

Time Counta/Mlnute Counts/Minute Percent (hours) Benaene Phase Aqueous Phase Aotlvity Extracted

1/4 16 1380 1 . 1 1/2 56 1382 3.9 1 60 1351 4.2 2 55 1 3 4 7 3 .9 A 64 1 3 7 6 4 . A PERCENT EXTRACTION 4 2 ET F OAT )TATD RM 3 HYDRO­ 3N A FROM E)CTRACTED COBALT OF CENT IUE EFC O CNAT IE N PER­ ON TIME CONTACT OF EFFECT / FIGURE HOI AI SLTO O TE AGNS TARGET MANGANESE THE OF SOLUTION ACID CHLORIC IE (HOURS) TIME —!•- d* EFFECT OF CONTACT TIME ON PERU!NT OF COBALT

EXTRACTED FROM A *5» HYDROCHLORIC ACID SOLU­

TION OF THE MANGANESE TARGET

Although an apparent trend oould be noted from the results obtained in which the precipitation of the co­ balt complex took place in a strongly acid solution, a second aeries of tests were carried out while maintaining the acidity of the medium at *5N* Two ml* of the test solution were diluted to 12 ml. with the resulting acidity of the medium being *5N* The solution was heated to eighty degrees centigrade and treated with A ml* of the organic reagent* The complexlng solution consisting of *25 grama ofalpha-nltroso beta-naphthol dissolved in 100 ml* of fifty percent (by volume) acetic acid solvent* After the predetermined length of time the solution was transferred to a separatory funnel and washed with three 25 ml* por­ tions of benaene* Both the aqueous and benzene phases were reduced to a small volume and transferred to counting dishes*

Each was evaporated to dryness and counted for three min­ utes* The amount of cobalt extracted by bensene after one hour standing was approximately 98 percent* The results of this experiment are shown in Table II and Figure 2* PERCENT EXTRACTION 0 4 IOO 20 60 0 6 ECN O CBL ETATD RM .5N A FROM EXTRACTED COBALT OF PERCENT FIGURE FIGURE YRCLRC CD OUIN F THE OF SOLUTION ACID HYDROCHLORIC AGNS TARGET MANGANESE Z. FET F OTC TM ON TIME CONTACT OF EFFECT 4 3 2 1 - / / - IE (HOURS) TIME

12- TABLE XI

Kfr*ot of Contact Tima on Fareant of Cobalt Extracted from

a *5N Hydrochloric Acid Solution of tha Manganese Targat Tima Counta/Minuta Counta/Minuta Parcant (houra) Banaana Phaaa Aquaoua Phaaa Activity Extracted

483 819 37*1 1000 459 68*6 1 1403 33 97.8 2 1485 52 96.7 4 1410 3 99.7

a. EFFECT OP ACUITY 07 TARGSK SOLUTION ON BEN­

ZS N* EXTRACTION OP COBALT FROM A FIFTY PER­

CENT ACETIC ACID SOLUTION OF ALPHA-NITROSO

BETA-NAPHTHOL

Tha following procadura waa adopted to establiah tha optimum acidity in which the reaction of cobalt chlorida

with alpha—nitroao beta-naphthol took place* Data ware

collected from a aeries of experiments in which tha acid

concentration varied from pH 8*7 to 6N. In each teat, con­

sisting of three samples, 2 ml* of the test solution ware

diluted with acid of appropriate strength to 12 ml. Tha

sample waa heated to a pproxinately eighty degrees centi­

grade and A ml. of a fifty peroent acetic aoid solution

of alpha—nltroao beta—naphthol ware added* After one hour

tha solution was transferred to a separatory funnel and

extracted with three 25 ml* portions of benaene which had

been previously saturated with hydrochloric aoid of strength

equal to that of the test solution. The benaene phase was than, vaahad with thraa 25 il» portion* of vatar. Tha vaah vatar vaa raoo*binad with tha firat aquaoaa portion. Both tha aquaoua and nonaqaaoua phaaaa vara raducad In roluaa and tranafarrad to counting dlahaa. Thaaa vara furthar raduoad to drynaaa vlth tha aid of a heat laap ao that loaa by apattaring vould ba kapt to a mininun. Tha count— ing rata of aach phaaa vaa detarmlnad ualng an and vindov

Gaigar-Mullar Tuba and Pottar Scalar. Tha optinua acidity

In vhich tha raactlon took placa rangad fro* pH 4 to «25H.

Tha rasulta of thla axparlaant ara abovn In Tabla III, Flguraa 3 and 4.

lABJLfi HI Effact of Acidity of Targat Solution on Banaana Extraction of Cobalt fro* a Fifty Parcant Acatlc Acid Solution of alpha-Nltroao bata-Maphtbol

Acidity Counta/Mlnuta Counta/Mlnuta Parcant Banaana Phaaa Aquaoua Phaaa Activity Extraotad pH 8.7 1570 18 98.9 7.2 1516 38 97.5 A.O 1052 A 99.7 Horaality

.10 1508 3 100.0 •25 1368 0 100.0 .50 1A03 33 97.8 1*5 298 1118 21.0 3.0 127 1298 8.9 6.0 66 1274 4.9 PERCENT EXTRACTION IOO 80 - 80 0 4 60 0 2 lRO BETA-NAPHTHOL NlfRSO SN A 0 CTC CD OUIN F ALPHA- OF SOLUTION ACID ACETIC 50% A USING N EZN ETATO O COBALT OF EXTRACTION BENZENE ON IUE FET F H F AGT SOLUTION TARGET OF pH OF EFFECT J FIGURE ______4 8 6 4 2 I ______- V / - pH I ------I ------

1 ---

FIGURE

PERCENT EXTRACTION FET F CDT O TRE SLTO O BNEE XRCIN OF EXTRACTION BENZENE ON SOLUTION TARGET OF ACIDITY OF EFFECT 100 SN A CTC CD OUIN F LH-IRO BETA-NAPHTHOL ALPHA-NITRSO OF SOLUTION ACID ACETIC % 0 5 A USING 20 0 8 0 4 0 6 0 2 NORMALITY 3 4 5 6 COBALT

-16- f. pFFXCT OF ACIDITT OF TIEGET SOLUTION ON BEN­

ZINE EXTRACTION OF COBALT FROM AN AQUEOUS

SOLUTION OF ALPHA—NITEOSO BETA-NAPHTHOL

Samples of tha targat solution vara prepared by pipetting out 2 ml, portions from an aqueous stock solu­ tion of the target material and adjusting to the desired pH. These vere heated to eighty degrees centigrade and

1/2 ml* portions of the aqueous solution of alpha-nitroso beta-naphthol vere added to each sample. After one hour standing each solution was extracted vlth three 25*ml* portions of bensene saturated vlth dilute hydrochloric sold* The bensene phase vas washed vlth triple distilled vater* The vash vater vas recombined vlth the first aque­ ous portion. Both the aqueous and nonaqueous phases vere reduced in volume and transferred to counting dishes. After evaporating each aolution to dryness, the percent of ex­ traction of oobalt from the target material vas determined by measuring the counting rate in each of the tvo phases.

This procedure vas employed in the study of percent of complexlng of cobalt from pH 2.2 to pH 9. All points listed in the table are the average of three runs. In order to maintain a constant volume of 12 ml. in the higher pH range, a large sample of the parent solution vas evaporated to dryness and redissolved in an equal vol­ ume of triple distilled water. The acidity of this solu­ tion vas pH 6. Solutions of higher pH vere obtained by -17-

the addition of .01 N ammonium hjdroxlda. Tha atxlaua

amount of activity waa axtractad when the acidity of the

target aolution vaa maintained between pH 4.7 and 6.2.

The reaulta of this experiment are shown In Table IV and

Figure 5. XAgfcl ,-IY Effect of Acidity of Solution on the Benzene extraction of

Cobalt from an Aqueous Solution of Alpha-ftitroso Beta-

Naphthol.

pH Counts/Minute Counts/Minute Percent Bgflaaav Pbftgt Activity Extracted

9.0 71 1004 6.7 7.4 399 676 37.1 6.2 1411 28 93.4 4.7 1017 29 97.3 4.1 496 577 46.2 3.7 545 738 42.5 3.6 296 1020 22.6 3.5 223 1061 17.4 3.0 53 1211 4.4 2.2 21 1068 1.9

g. EFFECT OF THE QUASTlTl UF ALPHA-NlTRUSO Be TA­

NA PH THOL ON THE BENZENE EXTRACTION OF COBALT

FROM THE TARGET SOLUTION

Samples of the target aolution were prepared by pipetting 2 ml. portions from the stock aolution, diluting to 12 ml. and adjusting the pH between 5 and 6. These were heated to eighty degrees centigrade. Four ml. of an aqueous solution of alpha-nitroao beta-naphthol were added to a teat aeries consisting of three samples. One and one-half ml. were added to a second and third test series. PERCENT EXTRACTION IOO 0 4 20 60 0 8 - IUE" FET F H F OUIN N BENZENE ON SOLUTION OF pH OF EFFECT FIGURE5" F LH-IRO BETA-NAPHTHOL. ALPHA-NITRSO OF XRCIN F OAT SN A AUOS SOLUTION AQUEOUS AN USING COBALT OF EXTRACTION 2 / ~ 4 8 - pH 6 8 10

-19-

After o d « hour standing, eseh solution vas aztraoted with three 25 al« portions of bsnsans saturated vlth dilute hy- droohlorlo aoid* The bensene phase vas vashed vlth triple distilled vater* The vash water was recombined with the first aqueous phase. Both the aqueous and aonaqueous phases were reduoed in volume and transferred to counting dishes* After evaporating eaoh of the phases to dryness the percent of extraction of cobalt from the target mater­ ial for 4, 1 and 1/2 ml* respectively of complexlng re­ agent was calculated from the counting rate per minute measured in the aqueous and nonaqueous phases* More than

98$ of the activity was removed when as little as 1/2 ml* of an aqueous solution at alpha—nltroao beta-naphthol vas added to the solution*( The results of this experiment are shown in Table V* lAfil* Y Effect of the Quantity of Alpha—Nltroao Beta-Naphthol on the Bensene Extraction of Cobalt from the Target

Solution

Amount of Counts/Minute Counts/Minute Percent Complexlng Bensene Phase Aqueous Phase Aotivlty Extracted Salirtlaa - - ;______1/2 ml* 1411 28 98*4 1 ml* 14 7 6 28 98*5 4 ml* 1238 42 96*7 -20-

3* DISCUSSIOB

As described In the previous section two series of toots vere oorriod out, on# in which tho tost solution vos aado up at 3H hydrochloric acid concentration, and a second in vhich tho acidity was *5H. It is notod in tho study of Fig* 1 that complexing in a 6H hydrochloric acid medium is Tory alight. The trend indicates maximum com— ploxing after one hour contaot between tho reagent and the solution* A study of Fig* 2 immediately reveals that a very satisfactory extraction can be expected if the target solution and complexing reagent are permitted'to remain mixed for a period of one hour. It is also observed that periods of longer standing do not indicate any marked

Increase in complexing*

The purpose of using an extraction method rather than filtration for the removal of the cobalt salt of alpha- nltroso beta-naphthol from the aqueous solution is two fold*

The precipitate formed in the reaction of quantities of cobalt of the order of 1 0 “ ^ m with alpha-nitroso beta- naphthol is so small that much of the activity could readily escape through the paper if filtered* Secondly, the ashing of the filter paper, although considered ash- less by macro standards, contributes to a slaeable residue in terms of nuclear chemistry. To avoid these difficulties two organic solvents were studied* Xylene vas first selected beceuse it is insoluble in water, thus it would -21- be •xptottd to givt a cleaner separation, In addition it ia laaa volatile than bensene haTing a boiling point of

138 dagraaa eantigrada aa conparad to 80 dagraaa oentl— grada for bansana« Thia hlghar boiling point aakas it a slightly more daalrabla laboratory reagent. However, it failed to extract tba complex from the aqueous phase, Aa a result the eaphasia vaa shifted to the use of benzene.

Although slightly soluble in water, J08 grama per 100 ml, of water at 22 degrees centigrade, it proved to be the most satisfactory solvent. Thus the problems associated with filtration were avoided by employing a solvent ex­ traction method.

In an effort to establish a satisfactory time inter­ val with respect to both maximum effective eomplexing and efficiency of time utilized, it became apparent that the acid concentration of the test samples was also a major factor in governing the degree of extraction that would take place, A comparison of tne percent of activity re­ moved at all intervals between l/A and A hours showed that the efficiency is much higher in the test samples of

•5N acidity than in 3H acid concentration. Confirmation of this trend was obtained by conducting a series of ex­ periments in which the acid concentration was varied from pH 8,7 to 6N while the volume of the sample was held eon- stant and the period of eomplexing maintained at one hour.

In each ease A ml, of alpha-nitroso beta-naphthol were -22- added to tho toot solution. Flguroo 3 *nd 4 show tho ef­ fect of oeid concentration. Tho eomplexing proeooo lo quite offielont in tho rango between pH 8.7 and .5H with

tho optimum oonditlono botwoon .1011 and «2$H. It ia notod

that tho ability of tho eomplexing agont to rooet with co­ balt doereaaoa rapidly whon the acidity of the sanplo

oxoooda »5K with only five percent of tho activity ex­

tracted at 61*

Tho problem of removing oxceea eomplexing reagent

waa of auffieiont importance to encourage further investi­

gation. The beat meana of accomplishing this waa to limit

the initial concentration to a minimum. With this in

mind, an aqueous solution waa prepared. This waa a satu­

rated solution of the eomplexing agent in wets'*, concen­

tration approximately thirty milligrams per 100 ml. of

water. A trial run showed that the efficiency of complex-

ing and extraction remain high, 97 to 98 percent at pH 5. However, when a single test was made in .5N hydrochloric

acid medium, the percent of activity extracted decreased

greatly becoming 2.6 percent. In a similar acldlo con­

dition using fifty percent acetic acid solvent for the

eomplexing agent approximately 98 percent of the activity

was extracted. This would indicate that the optimum

acidity for extraction may have shifted toward lower acid

concentration if water is to be employed as a solvent

for the eomplexing agent. The encouraging results of this - 2 3 - phase or tfaa experiment indicated that the amount of tha eomplexing agent oould be reduced to an almost invlsable quantity and thua practically eliminate tha problem of dlapoaing of tha excess. A aarias of axtraotlona vaa per­ formed in which tha acidity of tha sample aolutlon waa varied. Each sample waa traatad with l/2 ml. of a watar aolutlon of alpha-nitroso beta-naphthol. Figura 4 ahowa that highly afflolant eomplexing and aaparatlon could ba axpactad in tha pH ranga fro* 4.7 to 6.2. Although tha aaount of activity raaovad by tha aquaoua aolutlon of tha ooaplax, approxiaataly 99 parcant at optlaua eondltlona aa ooaparad to aora than 99 parcant whan using a fifty par­ cant aeatlo acid aolutlon of tha coaplaxing raagantf waa quita aatlafactory9 tha acidity of tha aaapla waa a auch aora critical factor. However, there waa no difficulty in controlling tha pH within thla range. The use of such a saall quantity of tha coaplaxing reagent so simplified tha oxidation process that it was possible to convert tha cobalt to tha oxide and remove the excess alpha—nitroso beta-naphthol by burning in a pyrex beaker rather than a platinum crucible. Tha activity could ba removed with considerably more ease from the former* 4. SUMMARY

The effects of several variables on the eomplexing of cobalt In concentrations of 10“^ In the presence of a manganese target material have been studied. They are summarized as follows:

One hour is sufficient time for the radioactive cobalt chloride to react with the conplexing agent, alpha-nitroso beta-naphthol.

The cobalt salt of alpha—nltroso beta-naphthol is best separated from the target material by usihg benzene in a solvent extraction method.

When using a fifty percent acetic acid solution of alpha-nitroso beta—naphthol as the eomplexing reagent, the optimum acidity falls between .0001 h and .25 N for maximum separation of cobalt.

When using an aqueous solutionof alpha—nitroso beta- naphthol as the eomplexing reagent, the optimum acidity falls between pH 4*7 and pH 6.2 for maximum separation of cobalt.

Based upon these conclusions a procedure for the separation of cobalt produced in a manganese target has been developed.

Place the target material in a small beaker and treat with several small portions of equilibrium hydrochloric acid until the sample has been completely converted to the chloride. Evaporate the solution to dryness and redissolve in a quantity of triple distilled water such that the final solution is approximately *iM in salt content*

Adjust the acidity to pH 5*5 using resaturated hydrochloric acid. Heat to eighty degrees centigrade and add 1/2 ml. of an aqueous solution of alpha-nitroso teta-naphthol.

Stir constantly during the addition of the eomplexing reagent. Allow one hour for eomplexing. Transfer the sample solution to a separatory funnel and extract with three portions of benzene saturated with dilute hydro­ chloric acid. The volume of the benzene should exceed the volume of the solution by a factor of two to insure good extraction. Wash the benzene phase with triple distilled water. Transfer the benzene phase to a small beaicer and evaporate to dryness. After the oenzene has beeh driven off, heat gently over an open flame to burn off excess eomplexing reagent and convert the cobalt salt of alpha-nitroso beta-naphthol to cobaltic oxide.

Dissolve in dilute hydrochloric acid and transfer to the spectrometer mount for study. - 26— B. CARRIBE FREE SEPARATION OF COBALT FROM ENRICHED

ISOTOPES OF IRON IN FERRIC OXIDE 1. INTRODUCTION

The study of tbs disintegration scheme of co­ balt Isotopes was attack from a second approach. In this phase of the work enriched samples of iron oxide were bombarded with 20 Mev molecular protons in the 60—inoh

Cyclotron at the University of California. As discussed in the section pertaining to the Separation of Carrier Free Cobalt from a Manganese Target, the chemical problems of sample preparation associated with nuclear spectroscopy were again to be solved. If the radioactive cobalt was to be investigated for gamma emissions only, carrier cobalt could be added to aid in the chemical purification. In

addition the presence of small reagent Imparities would not Interfere with the sample preparation. However, as in the previous work, it was necessary to study both posi­ tron and gamma emissions in an attempt to obtain a complete picture of the decay scheme.

A literature survey of methods of quantitative pro­ cedures on the macro scale was conducted as a preliminary step in the preparation of a carrier free cobalt sample, the latter being produced by proton bombardment of an iron target. A very satisfactory method consists of pre­ cipitating the iron at approximately pH 3.5 leaving the cobalt In the supernatant. However, in the presence of - 2 7 - « conelderable quantity of iron, tha volumlnoue natura of

tha praelpitata raquiraa that tha volume of tha aolutlon

nust bp rathar large. Thia floooulant praolpltata adaorba

a conaidarabla quantity of tha cobalt, thareby reaulting

In a loaa of aotivity. Tha introduction of incraaaad anounta of Inpuritiaa fron tha raagenta and lneouplata precipitation of the hydroua ferric oxide, even though

the latter la conaiderad inaolubla by ordinary quanti­

tative etandarda conalderably incraaaa the alia and weight

of the final cample* There ia alao tha problem of vola- tlslng large quantltiaa of aiaonlua aalta. A aecond

method conaiderad waa baaed upon a aolvent extraction

procedure. Ferric chloride la much more aoluble in ethar,

ethyl (5) or laopropyl (6),^han in water. finder conditiona

of controlled acidity greater than nln*y-nlne percent of

the ferric chloride can ba extracted from tha aqueoue

phaae by a aingle aeparation. No cobalt ia extraoted by

the ether. Further atudy revealed that laopropyl ether would probably ba auparlor to ethyl ether for extraction

of iron. Tha firat reaeon being that the efficiency of the extraction ia greater, ninety-nine percent at optimum acidity for ethyl ether aa compared to ninety—nine and nine—

tentha for laopropyl ether. Although thia difference

could be eaelly overoome by aeveral additional extraetlona, there la alwaya loaa of oobalt reaulting from ita trane—

fer into the other phaae beoauae of the aolublllty of water In ether* Secondly, the acidity ia a vary critical factor in tha procedure amploying athyl athar for extraction, baing 6#2N for aaxiaua attraction, vharaaa axcallant ex­ traction by iaopropyl athar can ba expected between 7*75N and 8*25N* The control of tha acidity of tha aolutlon la further complicated by tha rather high eolubility of athyl athar, 7*5 grama par 100 ml* of water at 22 dagraaa centi­ grade* Tha eolubility of laopropyl athar la *2 grama par

100 ml* of water at 20 dagraaa centigrade* A third raaaon which encouraged tha uaa of laopropyl athar rather than athyl athar, although not affooting tha eompleteneae of extraction ia tha lower volatility of tha laopropyl other*

It haa a boiling point of 67*5 dagraaa centigrade aa com­ pared to 34*6 dagraaa centigrade for athyl other* Thua isopropyl athar oan ba handled more easily in tha laboratory* 2* PRELIMINARY INVESTIGATION AND RESULTS

Tha first approach to tha development of a pro­ cedure for tha separation of carrier free cobalt from a proton bombarded iron oxide target waa baaed on a solvent extraction method using laopropyl ether*

For preliminary studies, a sample of iron powder waa bombarded at The Ohio State University Cyclotron* The sample, weighing 250 milligrams, waa removed from the tar­ get holder, placed in a small beaker, and dissolved in 6H hydrochloric acid* After heating the solution, l/2 ml* of 14K nitric acid waa added to oxidise the iron to the + 3 •tat*. Thia step being necessary •• ferrous chlorida la

Insoluble (7) in athar. Tha aolutlon waa evaporated to dgness. It waa traatad two additional times In a similar manner and takan to drfnaas to inaura complete removal of tha oxidising agant. Kltric acid intarfaraa with tha ax* traction, reacting with tha athar. Tha residua was than takan up in 50 ml. of 8 N hydrochloric acid and extracted with three 100 ml. portions of redistilled laopropyl athar saturated with 8N hydrochloric acid. Tha iaoporpyl athar, commercial grade, waa redistilled over ferrous sulfate to insure complete removal of any peroxides formed in tha laopropyl ethar while in storage. Only that fraction be­ tween sixty-seven and sixty-seven and one-half degrees centrigrada was usad in all extractions. Tha aqueous phase was heated to drive off any dissolved ether and than traatad with a small portion of 14 N nitric acid to insura that all remaining iron was in the + 3 state. Again tb* solution was reduced to dryness, traatad three times with

12 N hydrochloric acid and finally extracted with isopropyl ether. On evaporation of tha aqueous phasa to dryness in preparation for mpeotrometer mounting, a white residue appeared. Gentle heating volatlsed much of it, but varying amounts of solid remained turning black on stronger heating.

This portion remained in the aqueous phajse with the cobalt activity even after continued extraction. It was soluble in dilute hydrochloric acid. A series of tests in which separate combinations of tbs reagents used, hydrochloric acid and iaopropyl ether, hydrochloric acid and nitric acid, and nitric acid and Isopropyl ether indicated that a reaction waa taking place between the oxidising agent, nitric acid, and the Isopropyl ether. Apparently a alight amount of nitric acid remained in the aqueous phase in

spite of repeated attempts to remove it with hydrochloric acid. A search was made for a more suitable oxidizing agent. Hydrogen peroxide was capable of oxidising the iron to the + 3 state but contained considerable amounts of stabilising material which in turn contributed even greater residue to the final sample. This was abandoned in favor of bromine water. Bromine water proved to be the most acceptable oxidising agent, being capable of oxidising the iron to the + 3 state, the excess easily removed, and contributing no residue to the final product. During the series of tests established for the purpose of detecting the source of the large residue, it was also noted that a slight residue remained in the beaker which had contained only hydrochloric acid and isopropyl ether. Further in­ vestigation revealed that the ethyl ether-hydrochlorle acid mixture showed even slighter residue than the hydro— chloric acid—isopropyl ether mixture on evaporation to dryness. Furthermore the residue could be kept to a minimum if both the ethyl ether and solution were ohilled to approximately five degrees centigrade before the extraction prootn* Not only did tho ohillod conditions

of tho liquldo roduco tho apparent reaction between tho

tvo, but it waa further noted that ferric chloride la much loaa aoluble in cold aqueoua aolutlon thua enhancing the extraction of the iron. However, aa ethyl ether la

not aa efficient an extractor aa laoppopyl ether, aeveral

additional extractlona were necessary. The residue waa

accumulative during the aeveral extractions and the final

residue waa again beyond the limits of that deaired for

beta spectroscopy. The ultimate procedure adopted for

preparing a aatiafactory sample waa a combination of the

above procedures. Since ferric oxide waa to be the chemi­

cal compound of the enriched isotope bombarded, a sample

of reagent grade ferric oxide waa aelected for the study

of the procedure. Two hundred and fifty milligrams of

ferric oxide were placed in a 100 ml. beaker and treated with 10 ml. of 6 N hydrochloric acid. A l/2 ml. quantity

of bromine water waa added to lnaure all the iron had been

oxidised to the ferric state. After repeating the above

procedure three times, the solution waa evaporated to dry­

ness and rediaaolved in 25 ml. of 6 N hydrochloric acid.

As a final cheek to insure complete oxidation of the iron

to the ferrle atate, a drop of the teat solution waa diluted to 1 ml. and placed on a spot plate. To thia waa added

1 drop of .01 M potassium ferrloyanlde solution. A blue color indicates a positive test for ferrous ion. The -32- oolor range* from blua to blue-green, tha latter pradomi­ nating if tha farroua ion i# prasant in very small quan­ tities. 1 negative test resulted. It being apparent that all of the iron was in the + 3 state, both the solution and ethyl ether were ehilled to approximately five degrees centigrade in preparation for extraction. The oold solu­ tion, 25 ml. in volume, was transferred to a chilled separatory funnel followed by the addition of 5 0 ml. of chilled ether. The two phases were shaken and separated, the aqueous phase oontalning more than ninety-five percent of the activity and some ferric chloride. This solution was stirred vigorously permitting thm highly volatile ethyl ether to escape, then heated reducing the volume to approximately 1 ml. This was diluted to 20 ml. with triple distilled water and the acidity adjusted to pH 1*5. The solution was heated and dilute ammonium hydroxide added until pH 3.8 was obtained. Hydrous ferric oxide began to appear at pH 2.5* It was again heated and permitted to stand one hour before filtering. On evaporation of the filtrate to dryness and volatisatlon of the ammonium chlor­ ide, the presence of iron was still visible. Further study suggested longer standing following the precipitation of the iron or precipitation of the iron at a higher pH, approximately 8*5 to 9« The latter proved to be the most successful although considerable loss of activity was en­ countered due to adsorption of the cobalt by the hydrou* -33- ferrlo oxide in the beeio medium* The filtrate which shoved presence of iron on evaporation to dryness was redissolved in 1 to 2 ll, of dilute hydrochloric acid.

Two drops of bromine water were added and the solution evaporated to dryness. It was redissolved in 1 to 2 ml* of dilute hydrochloric acid then made basic with ammonium hydroxide* Following gentle heating, the sample was per­ mitted to stand for ten hours* At the end of this period a small quantity of hydrous ferric oxide was observed*

The solution was filtered, evaporated to dryness, and the ammonium chloride volatlsed* No visible residue could be observed*

3* CHEMICAL PROCEDURE

The following procedure was developed from the information gathered in a series of experiments that have been discussed in the preceding section* This was the procedure used for preparing samples of cobalt from ferrlo oxide targets later used in nuclear spectroscopic studies*

The ferric oxide target, weighing 230 milligrams, was dissolved in approximately 25 ml, of 6 N hydrochloric acid* After the material had gone into solution, a 1 ml* portion of bromine water was added and the solution evapo­ rated to dryness* The residue was treated twice more in a similar manner. This was followed by the addition of

10 ml* of 12 N hydrochloric acid and evaporated to dryness*

It is essential that the oxidising agent be completely removed* The residue was redissolved in 6 I hydrochloric •eld and taatsd for tha presence of ferrous ion using *01 M

potassium ferrleyanlde* A daap blua eolor la a positive

tast (8) for larga quantities of tha farrous ion, blue- green for smallar amounts* If a negative tast rasulted,

both the ethyl ether and target solution vara chiliad to

approximately five degrees centigrade* Using a 125 ml*

separatory funnel, the solution vas extracted once with

50 ml* of ethyl ether. The aqueous phase vas stirred

vigorously to remove all tha ether* Tha volume vas re­

duced, by heating, to approximately 1 ml* It vas then

diluted to 25 ml. vith triple distilled vatar and adjusted to pH 1*5* After heating, *01 N ammonium hydroxide vas

added until the pH approached 3*8* Again it vas heated

gently to aid in coagulating the hydrous ferric oxide,

then permitted to stand for tvo hours* Tha solution vas

filtered and vashed vith hot triple distilled vater* Tha

filtrate vas evaporated to dryness by gentle heating,

folloved by stronger heating vhich volatised the ammonium

chloride. The residue vas redissolved in 1 to 2 ml. of

dilute hydrochloric acid and traatad vith 1 to 2 drops

of bromine vatar* The solution vas again evaporated to

dryness and treated vith a fev drops of 12 N hydrochloric

acid to remove the oxidising agent* It vas evaporated to

dryness and the residue dissolved in 1 to 2 ml* of dilute hydrochloric acid* The solution vas made basic, pH 8 to

9, by adding ammonium hydroxide, and stored for a long period, approximately ten hours, It was then filtered, the precipitate washed vith 1 ml. of dilute ammonium hydroxide and the filtrate evaporated to dryness. The ammonium chloride was volatlsed, The final product con­ sisted of an invisible quantity of radioactive cobalt which could be transferred to the spectrometer mount using dilute hydrochloric acid. - 3 6 - 4, SUMMARY Proa the foregoing dlseuvslon, It ean be concluded

that a combination of solvent extraction, baaed upon the

greater solubility of ferric chloride In ethyl ether

than in vater, and precipitation of Iron aa hydroua fer­

ric oxide ualng ammonium hydroxide produces the moat

desirable sample of cobalt.

An oxidising agent la necessary to insure that all

of the iron has been oonverted to the ferric state. How­

ever, the oxidising agent must be completely removed

before the extraction with ether. Bromine water was the

most suitable oxidising agent tested.

Ethyl ether waa apparently more satisfactory than

Isopropyl ether as it left practically no residue other

than the ferric chloride that had not been extracted.

Iron not removed by extraction by ether could be

precipitated as hydrous ferric oxide at pH 3«S* Under

these conditions there was minimum adsorption of the active cobalt,

Pinal precipitation of hydrous ferric oxide must be

carried out in a very small volume, 1 to 2 ml, at pH 8

to 9 with filtration coming after several hours standing.

The cobalt sample, suitable for spectroscopic studies, was obtained by combining the techniques of solvent ex­ traction and precipitation of hydrous ferric oxide. II. THE DISINTEGRATION OF THE RADIOACTIVE ISOTOPES . 56 57 ^ 5S Co , Co , and Co A. INSTRUMENTATION

1. THICK LENS SPECTROMETER

a* INTRODUCTION

A magnetic Ians apaotronetar (9) vas uaad to study tha beta spectrin* of each of tha three isotopes of cobalt. In addition to the beta spectrum, an internal conversion electron spectrum, and a photoelectron speo- trua produced by gamma rays using a seventy milligram uranium radiator were also studied on this instrument.

This particular spectrometer was selected for these studies because of its upper energy limit of approximately 5 Mev and high percent transmission, approximately 1—1/2 per­ cent. The spectrometer consists of a brass tube thirty inches long and approximately ten inches in diameter*

The chamber Is equipped at one end with a detector, a

Victorean Thyrode Geiger—Muller Tube having a window thick- ness of 1*9 mg/on. The source mounting device is lo­ cated at the other end. Torridal coils are used to control the magnetic field along the chamber. Beta particle of various momente are focused at the detector end by changing the magnetic field strength. Contained within the chamber ia a ring focusing baffle system for increas­ ing the resolution of the instrument* b. XXPERIMENTAL PROCEDURE AMD RESULTS

Btfora undertaking tha actual atudlaa of tha

samples of the cobelt isotopea, several calibration

measurements were u d e ualng radioactive isotopes vhoae

spectra were well known. In this manner it waa poaaible

to eatabllah the charaoterlstioa of the apectroaster, 137 The first study was made using Ca . An aluminum cap

with a centered hole of 5/16th inch in diameter served aa

the spectrometer mount. The opening of the cap waa covered

with a film of rubber hydrochloride weighing .55 mg/cm .

The sample of C s ^ 7 waa mounted on the rubber hydro­

chloride. A similar procedure waa followed in the pre­

paration of all samples. The 625 kev internal conversion

electron peak waa used as a calibration point. From the

data. Figure 6, the calibration constant was determined

as 6 2 6 gauas-centimetera per ampere. Thia ia in very close

agreement with the previous work (9) in which the constant

was determined to be 625 gauss—centimeters per ampere.

The resolution waa found to be 3.5 percent. However, upon further study of the Ca1^7 spectrum in the peak region,

it was noted that the peak lackad symmetry. The counting rate appeared to be excessive on the low energy aide. This characteristic indicated the presence of scattering in the low energy region of the apeotrometer.

The second standard used waa F^^* A aample waa ob­ tained from Oak Ridge National Laboratories and mounted 20001—

COUNTS / M IN U TE 1600 1200 0 0 4 0 0 8 - FIGURE FIGURE “ - - b OVRIN LCR S F s137 C OF NS ELECTRO CONVERSION OETOEE RAIG (VOLTS) READING POTENTIOMETER 28 .9 0 3 .29 8 .2 5% RESOLUTION % .5 3 RGNL AFE SYSTEM BAFFLE ORIGINAL EIE SAE F SPECTRUM OF SHAPE DESIRED S I.I860 6 8 I . I * H 8 0 .2 1 L 8 3 3 & S3 - * H CONVERSION K • p 1 6 2 6 280

3 Y- -3 - 4 0 - on • rubber hydrochloride backing similar to that previous­ ly desoribed. Tha sampla used Tor tha study waa vary thin, JJ o on tha ordar of 75 4|gma/om • Figure 7, tha Farm! Plot of tha Beta Spectrum of P ^ , reveals that tha points bagln to deviate from a straight llna plot in the vicinity of

320 kev. Tha Fermi Plot of this Isotope has bean found to give a straight line as low as 300 kev. (10) This curvature, shown in Fig* 7, is further confirmation of poasible scattering in tha low energy region of the spec­ trum, To insure that tha sampla thickness was not the major factor contributing to scattering, three successive

samples of similar area but of decreasing counting rata and therefore decreasing thickness were studied. 411 showed the same trend.

Tha third calibration study was made using Co^®«

This isotope has two conversion electron peaks at 1.19 Hav and 1.33 Mev respectively. An analysis of Figure 6 shows the two conversion peaks and gives a calibration constant 137 in agreement with that determined using Cs . Spurious counts were observed between the upper energy limit of the beta particle and the conversion electron peaks. They appeared to originate from some gamma ray effect. It was also noted that a high background was present resulting from the lack of a gamma ray absorber. This would account for the small extent to which the conversion peaks rise above it. - V7 -

ORIGINAL BAFFLE SYSTEM30

20

BEGINNING OF SCATTERING ~ 8 2 0 Kev

END POINT ENERGY y— 1.70 Mev

1.0 2.0 3.0 4.0 5.0 ENERGY

FIGURE 7 FERMI PLOT OF P 32 COUNTS / M IN UTE - C 0 4 - 0 0 3 SOOT 0 0 2 100" < 0 IUE 8 FIGURE -

. 0.3 0.2 I I OETOEE READING POTENTIOMETER EA PCRM F Co60 OF SPECTRUM BETA -# ------OPRSN EWE TE W BFL SYSTEMS ------BAFFLE TWO THE BETWEEN COMPARISON ------UEIPSD OMLZD CURVE NORMALIZED SUPERIMPOSED * OIID AFE YTM AA OMLZD TO NORMALIZED DATA SYSTEM BAFFLE MODIFIED RGNL AFE YTM DATA SYSTEM BAFFLE ORIGINAL OMN EK HEIGHT PEAK COMMON IVOLTS) .4 0 AKRUD EFFECT ► BACKGROUND AKRUD EFFECT BACKGROUND RM AM A EFFECT RAY GAMMA SCATTERING FROM IN DECREASE RGNL BAFFLE ORIGINAL MDFE BAFFLE ^MODIFIED

▲ diagram of tha baffle system Installed at the time

of these studies Is shown in Figure 9*

A new baffle system waa constructed in an effort to eliminate the scattering effects as reflected in Figs. 6,

7, and 8 and to reduce the high background aa shown in

Fig. 8. A diagram of the new baffle system is shown in

Figure 10. Comparing this with Fig. 9, it can be seen

that the changes includet

1. Increasing the number of baffles in the region

of the deteotor.

2. Placing a block of lead 9 inches long and 3

inches in diameter in the center of the baffle

system.

3* Installing an aluminum discriminator (not shown

in the diagram)

Vith the new baffle system installed, the calibra­

tion studies were repeated. Figure 11 is a study of the

conversion electron peak of Ca^^* It can be seen that

the peak shows greater symmetry. Scattering on the low energy side appears to be greatly reduced if not entirely

eliminated. The distortion on the upper energy side is

due to the contribution from the L conversion electrons.

The resolution was leas than that with the first baffle,

4.25$ compared to 3«5$» thua the I* peak was not resolved as clearly as that shown in Fig. 10. The sacrifice in resolving power waa made in order to obtain high trana- ALUMINUM

T— r r “ B— D— □— 0— D— 0---T U V - B A F F L E

SOURCE DETE TOR D

n n n - j o d o a.

36*

FIG URE $ ORIGINAL BAFFLE SYSTEM. ------ALUMINUM ------BRASS ------LEAD

T1 BAFFLE-r-*. fl ^ v s a * i

\ U oUQ OETE CTOR “ urce raji._jirTEA7 a it ■ ui_ "-IB iP n 11 D n cl

36"

FIGURE 10 MODIFIED BAFFLE SYSTEM MODIFIED BAFFLE SYSTEM

7 0 0 0 H/a * 3381. e ■ 1 2 2 3 2 7 6 5

6 0 0 0

5 0 0 0

4.25% RESOLUTION

3 0 0 0 o

2000

lOOO —

.2 5 .2 6 .27 .2 8 .2 9 .3 0 POTENTIOMETER READING (VOLTS)

F IG U R E It CONVERSION ELECTRONS OF C*'37 mission for photoelectron peak studios of Co56.

Tho study of the charsctoristics of tho spootromotor with the now bsfflo system was oontlnuod using F ^ . Pig- 32 ure 12 Is a Fermi Plot of the Beta Speotrum of P * It is apparent that the plot follows a straight line down to approximately 4 0 0 kev.

The third well known Isotope used In studying the change of the characteristics of the spectrometer through the modification of the baffle system was These data are shown In Figure 13* By normalising the data obtained with the modified baffle system Installed to the data taken with the original baffle system In the in­ strument, It can be seen in Fig. 8 that the background has been reduced by a factor of four. In addition the analysis of the data to the left of the conversion peaks shows a decrease in spurious counts. This is shown by superimposing Curve B on Curve This is in agreement with the results obtained when using C s * ^ and P ^ as calibration sources*

The modified baffle system as installed for the studies of the cobalt isotopes had a resolving power of

4*25 percent and approximately 3 percent transmission.

That of the original baffle system; 3.5 percent resolu­ tion and 1-1/2 percent transmission. Several additional changes are now in progress to further Improve the re­ solution and transmission. “ '/s-

30 MODIFIED BAFFLE SYSTEM

20

BEGINNING OF SCATTERING * 410 Kev

_N

END POINT ENERGY r ~ 1.70 Mev

1.0 2.0 3.0 5.0 ENERGY (m0cz )

FIGURE 12 FERMI PLOT OF P32 COUNTS /M IN U T E 0 0 6 — 0 0 4 0 0 3 200 0 0 5 IOO 0.1 IUE( BT SETU O Co6° OF SPECTRUM BETA (3 FIGURE OETOEE RAIG (VOLTS) READING POTENTIOMETER 0.2 OIID AFE SYSTEM BAFFLE MODIFIED .3 0 .4 0 .5 0 -50-

O. DISCUSSION

The calibretion of the spectrometer vae carried out for two reasons. The first and most obvious waa to establish a calibration constant. Good agreement with previous work (9) was obtained. The second reason was to obtain a general picture of the characteristics of the instrument. A careful analysis of the data revealed the presence of exoessive scattering in the low energy range. This was indicated in Fig. 6 by the nonsymmetrio shape of the Cs-**^ peak. Had this distortion appeared on the upper energy side of the spectrum, it could have been explained as the influence of the L conversion elec­ trons and poor resolution. However, the distortion is evident on the low energy side. The dashed line in Fig.

6 shows the desired shape in this region. A study of

Fig. 11 shows that this distortion has been removed from the low energy side. The L peak was no longer apparent because the resolution was not as good as witn the previ­ ous baffle. Its Influence is evident from the broadening of the peak on the high energy side.

Jensen and coworkers (10) have reported that the 32 continuous spectrum of P will give a straight line

Fermi plot as low as 300 kev. This departure from the straight line at 800 to 900 kev (Fig. 7) once again indicated scattering. Using the modified baffle system, the data indicated a reduotlon in scattering and was in -51- better agreement with the literature ae points on the

Fermi plot did not tend to deviate from the straight line

until in the region of 4.00 kev, (Fig. 12). It is obvious

that the presence of such scattering eould easily be

interpreted as a low energy component in the study of a

lesser known isotope.

was used to study the effect of gamma rays on

the beta spectrum. The presence of the gamma rays con­

tributed to high background and scattering in the spectro­ meter. It was noted that both these effects were in excess with the original baffle system installed. However, by replacing it with the new baffle system, these effects were greatly decreased. The decrease in background was primarily due to the addition of the lead block to the center portion of the baffle system. Scattering was re­ duced as a result of the change in the arrangement of the baffles. For comparing the two baffle systems the data was reproduced in Fig. 8 by normalizing the conversion electron peaks.

d. S UMMA.fi Y

Calibration studies of the magnetic lens spectro­ meter were made to check the calibration constant and investigate the characteristics of the spectrometer.

Samples of well known isotopes, C * ^ 7, and Co^° were used. In each case considerable scattering was found.

The low energy side of the Cs1^7 peak was distorted. The -52- Farai plot of bagan to depart froa a atraight line

In the vicinity of 800 kev* The effect of the gaaaa raya of Co^O vaa quite pronounced both In contributing to high background and scattering*

The baffle aystea vaa modified to attempt to correct thla situation* Several baffles were added to the aystea in the vicinity of the detector* A lead block was added to the center of the baffle aystea* A discriminator vaa

Installed*

These changes improved the characteristics of the spectrometer* Scattering was noticeably reduced* The Cs3-^ peak was more symmetric* The Feral plot of remained a straight line to 4.00 kev* Background froa gamma rays was greatly reduced* 2. SOLANOIDAL SPECTROMETER

The solenoidal spectrometer was also used for atudies involving tha various cobalt isotopes. It differs from the thick lens type spectrometer primarily in the dis­ tribution of the magnetic field strength along its axil,

being of uniform strength from source to detector. One

reason for studies on this Instrument was to make a com­

parison of the data with that obtained from the magnetic

thick lens spectrometer. In addition, the portion of

the instrument containing the detector was designed to

accomodate either a commercial tube or a thin window tube.

A glass system had previously been designed (11) for fill­

ing the thin window tube. The window was made of e ither

rubber hydrochloride wieghlng .55 mg/cm or xapon weigh­

ing .14 mg/cm2. The thin window made possible studies

of the beta spectra of the cobalt Isotopes to a much lower

energy region. Here again careful calibration studies were made. The information gained from C a ^ ^ and Co^^

indicated very slight scattering and minimum "gamma effect"

A study of a Fermi plot of P revealed the possibility

of scattering in the low energy region. If present, it

was very slight and will be discussed later in connection 56 with a possible new beta component of Co . A discrimina­

tor was also constructed and installed (12) during the

period of the cobalt studies. Its influence will be dis­

cussed in a later section as it affected the spectra of -54- the cobalt isotope*•

Current control end resolution vere two other factors that United the extent of sone of the Inrestlgations9 notably Co^« 3. SCINTILLATION SPECTROMETER

The scintillation spectrometer was used exclusively to study the gemma ray spectra of the cohalt isotopes.

It consisted of a thallium activated sodium iodide crystal mounted on a 5819 photomultiplier tube. This in turn was connected to a linear amplifier, pulse height selector, and a scaler. It is possible to detect both the gamma ray energy and r elative or absolute intensity depehdlng on whether a standard is available. In this work the gamma ray energies and their relative intensities associated with the cobalt isotopes were determined in appropriate oases.

The instrument was calibrated with Ca^*^, Mn-^,

C o ^ (using a well established gamma ray), Zn^, C o ^ , and Ce^^, and Ha2^. Figure 14 is a plot of the first calibration curve. As can be seen this was obviously non-linear. It was later learned that the voltage setting on the photomultiplier tube was too high and was respon­ sible for such irregular results. Following the adjust­ ment of the voltage, a linear plot was obtained over a wide energy range. This is shown in Figure 15.

The data compiled on the gamma rays of the cobalt

Isotopes from the studies on this instrument were used to supplement the findings by other methods. GURE I BRATI UV OF NTI ON IO T A L IL T IN C S F O CURVE N IO T A R IB L A C IH E R U IG F ENERGY (M EV) 0.6 0 0.4 1.0 1.4 . 8 O - XESV VLAE ON VOLTAGE EXCESSIVE PLI UBE B TU R IE L IP T L U M O T O H P R E T E M O R T C E P S 20 HI ON N IO T A IL IH N N A 0 4 AL IA T N E T O P

r 0 6 0 8

144 IOO .4 0 0.8 ENERGY (M EV) 20 2.8 4 Z 1.6 O GURE E R U IG F HOT TPLER T E B TU R LIE LTIP U M TO O PH OR VLAE ON VOLTAGE L A RM NO IS BRATI F NTI ON IO T A L IL T IN C S OF E V R U C N IO T A R IB L A C 20 ROMETER E T E M O TR C E P S 0 4 AL IA T N E T O P - V 5 -

Ce 0 6 0 8 Ce 144

IOO -58-

B. SPECTROSCOPIC STUDIES

1. INTRODUCTION mf. C

possibility of C o ^ with a half life of 70 to 80 days.

Cook and MoDaniel (14), using absorption and coincidence

methods, reported that the disintegration took place by

K capture and a simple positron emission with an upper

energy limit of 1.23 Mev. The average energy for the

several gamma rays present was 1.74 Mev. Elliott and

Deutsch (15), using a magnetic lens spectrometer and

coincidence methods, reported a positron spectrum con­

sisting mainly of a single group of maximum energy of

1.5 Mev. Six gamma rays were found ranging in energy

from .845 Mev to 3.25 Mev. Cheng (11) reported two

positron spectra having been observed with end point

energies of 1.5 Mev and .995 Mev respectively.

Livlngood and Seaborg (13) also investigated C o ^ , They reported that it disintegrates by the emission of 57 a .26 Mev positron to an excited state of Fe • Later

studies by Plesset (16) and Elliott and Deutsch (15) show

that the excited state of Fe57 M kes a transition to the

ground state by two ways, radiation of a 119 kev and 1 4

kev gamma ray, and radiation of a 133 kev gamma ray. Re­

cent investigations by Cheng (11) report the positron •mission to hato an upper energy limit of 320 kev. The

119 kev, 133 kev, and a very soft (less than 18 kev) gamma rays were observed. The K/L ratio of the 119 kev gamma rays to the 133 kev gamma rays were established to be about 6.3 and 5*2 respectively.

Livlngood and Seaborg (13) found that Co decays by positron emission to 7e5®. Using absorption methods, they reported a maximum beta energy of .5 Mev and a gamm* ray energy of .8 Mev. Deutsch and Elliott (17) continued the s tudy of this Isotope by means of a magnetic lens spectrometer and coincidence techniques. They found that it decays by K capture and positron emission and

14*5% respectively (18)) to an excited state of Fe^® .803

Mev above the ground level which in turn decays by emission of a single gamma ray. The maximum energy of the posi­ tron emitted is . 4 8 Mev.

The object of the present work is to resolve the con­ flicting data b^ additional investigation of the recent findings using enriched isotopes of iron and to extend the information concerning the decay schemes of these Isotopes• «60« 2* SAMPLE PREPARATION

Samplea of the oobalt iaotopea uaed for the apectro- soopio atudlea to be deaorlbed in the following aeetiona were the producte of p,n reactiona with iron oxide*

The radioactive iaotope, Co^6 was produced by boa* barding an enriched aaaple of The enriolia<* iao— tope was obtained from the Oak Ridge Rational Laboratory*

The aaaa and apeotrographic analyala are given in Table VI and Table VII. Theae analyaea were made at the Oak Ridge National Laboratory* XAfifci YI Maaa Analyala of Enriched Ve^O^

Iaotope Atom Feroent Error

54 0*050 ±0.005 56 99*34 +0.02 57 0*094 +0.006 53 0.012 +0.007

XAM 111 56 Speotrographlc Analyala of Enriched * + 2 3

Element Feroent

Ag 0.04 Cd 0*15 Mg 0*02 SI 0*05 Sn 0.03

(Impurltlea other than thoae H a t e d were not deteeted)

The target waa under bombardment for alx hours and reoeived a total of 28 mieroampere houra of molecular protona (Hj)* Thia repreaenta fifty-aix microampere -61- hours of atoalo protons of approxiaately 10 Mar* Tha

Boabardaant vas aada In tha 60—Inch Cyclotron at tha Uni­ versity of California* 97 Tha radioactive laotopa Co*' was producad by tha proton boabardaant of an anriehad saapla of Fa|^0^. This anriehad laotopa vas also obtained froa tha Oak Ridge National

Laboratory* Tha aass and spectrographie analysis la given in Tables VIII and IX* These analyses ware aada at tha Oak Ridge National Laboratory*

Mass Analysis of Enriched Isotope Atoa Percent Error

54 0.537 ±0.009 56 45.69 ±0.02 57 53*53 ±0.02 58 0*252 ±0.005 IABML Spaotrographio Analysis of Enriched Fa|^0^

Elaaant Percent

A1 *L 0*08 Bi 0.15 Cr L 0.15 Cu 0.31 Na L 0.15 Hi L 0.08 Pt 0.31 Si 0.04 Ti 0.04 *L Oataotad but lass than H a l t of deteraination. Ia— purities other than those listed vara not detected. -62-

The target was under boabardaant for tbraa hours and received a total of 14.0 aloroaapare hours of aolaoular protons. This raprasants 28 aieroaapara hours of atonic protons of approximately 10 Mar. Tha bombardnent was also made in tha 60-ineh Cyclotron at tha University of

California. Tha activity of tha sample was low as com­ pared to that obtained from enriched F e ^ *

Tha radioactive isotope Co58 vas produced by tha proton bombardment of an enriched sample of Fa|^0^. Tha source of this isotope was also Oak Ridge National Labora­ tories. The mass analysis is given in Table X. This analysis was made at tha Oak Ridge National Laboratory. TABLB X

Maas Analysis of Enriched Fe|^0^

Isotope Atom Percent Error

54 2.5 ±0.3 56 70.4 ±0.5 57 3.5 +0.3 58 23.4 +0.5 The target was under bombardment for three hours and received a total of 14 microampere hours of molecular protons. This represents 28 microampere hours of atomic protons of approximately 10 Mev. Similar to the two bom­ bardments described above, this was also made in the 60—

Inch Cyclotron at the University of California.

The sample of C o ^ prepared for photoelectron spec­ trum studies was not separated from the target material but transferred directly froa the shipping oontsiner to

a spectrometer lount designed for this type study*

The samples of radioactive isotopes prepared for

continuous beta spectrum studies vere subjected to the

chemical procedure described in Section I—B* An alumi­

num cap with a centered hole 5/l6th inches in diameter was

the mounting base used in the thick lens and solenoidal

spectrometers* The opening was covered with either rubber

hydrochloride or zapon film weighing 550 Mg/om^ _ 30 J^g/cm^

respectively* Radioactive solutions of the cobalt Isotopes were evaporated on this very thin plastic backing* 3. THI ISOTOP* Co56 a. THE CONTINUOUS POSITRON SPECTRUM OP Co56

A sample of C o ^ separated froa a targat of

farria oxide, anriehad In Fe^, by tha chemical procedure daaeribad in Chaptar I—B, waa mounted on a aapon film weighing 30 Ag/om2. Tha flln waa attached to an aluminum

cap which waa mounted in tha Solenoidal Spectrometer.

The spectrometer waa equipped with a thin window tuba for atudy in tha low energy region. Tha tuba window waa made of 4 layara of sapon, weighing 140 Mg/cm2 and re­

inforced by a nylon grid. Tha tuba waa filled with a mixture of argon and ethyl alcohol vapor to a pressure

of 7 centimeters. The gas and vapor were mixed in a ratio of 9 to 1 respectively.

As mentioned in Section A of this chapter, a care­

ful calibration atudy of the Solenoidal Spectrometer was made using P^2 » Co^® and Ce^*^. There was some slight indication of scattering that could not be attributed

to source thickness as a very careful study was made In which samples of decreasing thickness were measured. This tendency is shown in the Fermi plot of F^2, Fig. 16, in which the points deviate from a straight line in the vicin­ ity of 800 kev. Rather than attempt to construct a new baffle system as in the case of the thick lens spectro­ meter, the points were corrected so as to fall on a straight line. A {lot of the correction factors and the 1.0 2 .0 3l0 4 .0 a o ENERGY (m0c2 UNITS)

FIGURE FERMI PLOT OF BETA SPECTRUM OF P52 CALIBRATION OF SOLENOIDAL SPECTROMETER -66- resultlng curve are shova In Figure 17, This c u t t # was used to obtain oorraotion factors which ware applied latar in tha Fermi plot of tha eontinuoua beta spectrum of Co56#

A Fermi plot of tha continuous beta spectrum of C o ^ is shown In Figure 18* Both the original data and the corrected data are shown* It can be seen from the plot that the energy of the main component is approximately

1.48 Mar* A second component is also observed, even after correcting the points for scattering. This has an end point energy of *44 Mev, In order to insure that this second component was not the result of scattering caused by source thickness, less than 200 ng/cm^ in the most intense sample, the intensity of the source was re­ duced by one—third and one—sixth while maintaining a con­ stant area* The thickness of these samples was on the order of 90 Mg/om^ and 45 ng/cm^, The data of the three samples were normalised as shown in Fig, 18, The low energy component was found In all three studies. Thus scattering froa source thickness was eliminated as being responsible for the component. The momentum energy dis­ tribution plot, Figure 19, shows the main component to be present to about 95»7Jt and the low energy component as

4«3£, The contribution of the low energy component can be considered only approximate since the Farml plot of the positron spectrum of the cobalt isotope was correeted CORRECTION FACTOR .4 0 06 8 0 1.0 1.0 OAL SPECTROMETER E T E M O R T C E P S L A IO O N E L O S GURE E R U IG F 17 BRATI E R O F VE R U C N IO T A R IB L A C 2.0 -47- NRGY m0 UNI ) S IT N U * 0c (m Y G ENER 3 j 0

0 5 O UNCORRECTED POINTS O CORRECTED POINTS

.44 Mev 1.48 Mev

1.0 1.5 2.0 2.5 3.0 35 ENERGY (m0c2 UNITS)

FIGURE Id FERMI PLOT OF BETA SPECTRUM OF Co56, DATA FROM SOLENOIDAL SPECTROMETER 200 -69-

100

0 — 0 ACTUAL DISTRIBUTION 0 - 0 CORRECTED DISTRIBUTION X— X DISTRIBUTION IF NO 2ND COMPONENT

AMPERES 100

FIGURE 19 MOMENTUM DISTRIBUTION CURVES OF C

However* the data obtained before tha modification of the baffle system Indicate the presence of a second component. The analysis of this curve shows a low energy component with an end point energy of .918 Mev. It is observed that the plot obtained following the modification of the baffle system gives a straight line as far as the study was carried into the low energy region. The oomponent shown apparently resulted from scattering associated with the inatrument.

b. PhUTOELECTECK SPECTRUM OF Co56

The photoelectron spectrum of C o ^ was measured 4 0 □ ------□ ORIGINAL BAFFLE SYSTEM INSTALLED

MODIFIED BAFFLE SYSTEM INSTALLED

3 0

20

916 Kev 1.46 Mev

4 0 0 6 0 0 0 1000 1200 1400 160080 ENERGY (KEV)

FIGURE 2 0 FERMI PLOT OF BETA SPECTRA OF Co (UPPER ENERGY REGION ONLY). DATA FROM THICK LENS SPECTROMETER. -72-

ln th» thlek ltn i cp»otr*m t*r using a urtnina radiator

vaighing 70 a|/«a^« Tha spostronatar vaa oallbratad

using tha 667 k#T oonrarsion alaotron paak of Ca^^,

Fig. 11; tha photoalaotron apaotrum of Co*® ualng a

70 ng/en2 urahlvua radiator, Fig. 21; and tha annihilation

paak of C©5*. Tabla XI ahows tha calibration eonatant

obtalnad by aararal nathoda. Fotantioaatar raadlnga vara

takan at tha paak, tha intaraaetion of tha high anargy

alda of tha paak and tha basa lino (20), and tha intar—

aaetion of tvo straight linaa aupsrimpoasd on tha aidaa of tha paak*

TABLE IT

C alibration Constant for tho Thiek Lana Spaetronatar

Paak of Curva High Inargy Sido Intarsaotlon and Baaa Lino o f L i n o s P o t . C o n s t . P o t * C o n s t * P o t . C o n s t . K d f . i d * .

Cs ( 6 6 3 * 6 ) . 2 2 1 1 2 0 5 . 2 9 1 1 1 6 a • 2 6 0 5 1 2 0 7

A n n i h i l a ­ t i o n p a a k o f C © 56 .205 1217 .210 1 1 9 0 * 2 0 5 1 2 1 7

C o 6 0 1 . 1 7 2 . 3 9 7 1 2 4 3 . 4 1 7 1 1 6 5 . 3 9 9 1 2 4 0

l * 3 3 3 k • 4 4 3 * 4 6 2 1 1 9 1 . 4 4 6 1 2 3 6

1 * 3 3 3 1 • 4 9 2 1 1 6 3 . 4 7 9 1 2 1 6

A r a . C a l * C o n s t a n t 1 2 2 2 1 1 6 3 1 2 2 4

Tho oomplots photoolootron spsetrnm of Co** la shown in Figaro 22. Soron ganna rays (niao poaka) can ba ob— 0 0 0 7 COUNTS/MINUTE 0 0 0 4 0 0 0 5 0 0 0 6 0 0 0 3 lOOO 2000 RADIATOR) GURE E R U IG F HC LN SE R / UM IU N A R U ? M ./C G M 0 7 ( M TRU SPEC LENS THICK .03 21 HTEETO SETU Co F O SPECTRUM PHOTOELECTRON OETOEE RAIG (VOLTS) READING POTENTIOMETER - 3 7 - 04 .0 .3 o C NE) E CL IN L Mov 1.333 -L 1.172 Mov

J05

eo COUNTS/ MINUTE 1500 2000 1000 50C- IUE22 POOLCRN PCRM F o6 70 M.C2 RNU RDAO) TIK ES SPECTROMETER. LENS THICK RADIATOR); URANIUM MG./CM2 0 (7 Co56 OF SPECTRUM PHOTOELECTRON 2 2 FIGURE - .511 .20 .30 85 K LINE) {K .835 80 L LINE) (L .840 .40 POTENTIOMETER (VOLTS) READING 1.22 LINE) (K 5 .0 .70 .60 .50 .76 2.31 .80 2.61 .90 1.00 ■3.25 1.10 -75- serred; *511 Her the positron annihilation radiation,

•835 Mar (K line), .840 Mar (1* line), 1.22 Mar (K line),

1.24 Mar (L lina), 1.74 Mar, 2*31 Mar, 2.61 Mar, and 3.25 Mar. Tha energies of tha gamma raja vara datarmlnad by tha methods Identical with those described for analysing Co*^, taking the potentiometer reading at the peak, tha

Intersection of tha high energy aide of the paak and tha base line, and the Intersection of tha two straight lines superimposed on the sides of tha peak. Examples of these three methods are shown In Figure 23* The relatire intensities of the raspectlra gamma rays ware determined by analysing tha photoelectron peaks and correcting for the momentum selection width of tha spectrometer and tha photoalectron cross section (21). The areas of four peaks, those of the 1.74 Mar, 2*31 Mer, 2.61 Mar and the

3*25 Mer gamma rays, were reduced by a factor of 20% (22) to correct for tha contribution of the shells lying be­ yond tha K shell. For instance it was not possible to separata tha K and L peaks of tha 1*74 Mar gamma ray, therefore the area of this peak was reduced by 20%, so that in computing its realtlre intensity, a more accurate ralua would be obtained. Table XII shows the energy and relatire intensity of the gamma rays found in Co^« COUNTS/MINUTE 0 00 1 000 - 0 0 9 GOO 0 0 5 .50 7 G/M UAIM AITR TIK ES SPECTROMETER LENS THICK RADIATOR) URANIUM MG./CM, (70 IUE2 PRIN F H POOLCRN PCRM F Co96 OF SPECTRUM PHOTOELECTRON THE OF PORTION 23 FIGURE - .4 M«v 1.74 OETOEE RAIG (VOLTS) READING POTENTIOMETER .65 .1 Mev 2.31 80 .8

TABLE XII 56 Analjsla of tha Photoalaetron Spectrum of Co*7

▲ t o * Gaaaa Snergj aa Determined front Relative Gaaaa Raj Raj Intersec- Intaraac- Intensitj Energj Paak tion of tion of Baaa Lina two linaa ___

3*251 3.237 3.257 3.258 .29 2 . 6 0 6 2.595 2 . 6 1 6 .17 2.313 2.294 2.331 2.313 .19 1.758 1.740 1.788 1.746 .29 1*246 L 1.237 1*267 1.233 1.225 X 1 * 2 2 2 1.229 1.224 .80 Q.843 L 0 . 8 4 0 2 0.8499 0.8386 O o H 0.855 X 0.8327 0 . 8 3 5 6 0.8375 .

0 . THE INTERNAL CONVERSION ELECTRON SPECTRUM

Tha conversion eleotron paaka of tha 0,84 Mar

and 1 * 2 4 Maw gaaaa raja wara aaaeurad in tha thiok lana

apaotroaatar• Thla atudj ia shown in Figura 24* Using

tha aaaa sample, but reversing tha field, tha positron

spectrum was also measured in order to compare the gamma

raj intensltj to the poaitron intensitj of the sample.

Assuming that the radiations of tha 0,84 Mev and 1*24

Mev gaaaa rajs are electric quadropole, E2, tha conversion

coefficients from Rose's Table are found to be 2*7 x 10“^

and 1*07 x 1 0 “^ respectivelj. Using this information,

the intensitj ratio of tha 0*84 Mev gaaaa to tha 1.24 Mev

gamma to the positron spectrum vaa found to be Is . 8 6 s *2 1 .

These are summarised in Table XIII. .32 .34 .36 .38 40 .42 .44 .46 .48 POTENTIOMETER READING (VOLTS)

F IG U R E D CONVERSION ELECTRON PEAKS OF Co5 6 , DATA FROM THICK LENS SPECTROMETER -79- TABLE XIII

Intensity of t h o Internal Conversion Electrons of C#56

Gaaaa Ray Relative Intensity True Area Ratio

*84.® 167 62 x 10* 1 1.24 57*2 53.4 x 1 0 * . 8 6 Co56 125840 12.5 x 104- .21

d. THE GAMMA. RAl ENERGIES A*D tu£LATl¥E INTENSITIES MEASURED WITH A SCINTILLATION SPECTROMETER

Tho scintillation spoctroaater was usod to con­ tinue tha studies of the gaaaa ray speotrua of C o ^ , Before aaking a study of the C o ^ isotope, the lnstruaant was calibrated using Cs^7^ Zn^, and C a ^ ^ . A typical pulse height distribution curve,

Cs^^, is shown in Figure 25. A calibration curve is shown in Figure 2 6 . In this rigure the Energy Scale is at the right of the graph. Ca*’^, Ce^^, and Z n ^ were used to obtain the Energy vs. Potential Plot. Gaaaa ray intensities were first measured by a Geiger—Muller Counter, The affloianey of the counter was determined with the aid of an Efficiency (£) vs. Energy (Mev) Plot (23).

The gaaaa ray speotrua was then measured on the scin­ tillation spectroaeter• The gaaaa ray intensity of each peak was determined as the product of the peak height, h, and the half width, fl/2. From this information it was COUNTS /M IN U T E IOO 140 120 20 0 4 0 6 0 8 O I E R U FIG F 37; N 4; AV * 2 * V A ; 4 IN A G ; 7 '3 « C OF E V R U C - US' 20 HEI STRI ON IO T U IB R T IS D T H IG E -H E S L U P - 0 0 - OTNTAL TIA TEN PO 0 8 0 6 0 4

COUNTS/MINUTE 200 IOO 500 20 NRY OTO) NTI ON SPECT . R E T E M O TR C E P S N IO T A L IL T IN C S PORTION) ENERGY GURE26 GMA A SE RUM O C56 (LOW 6 Co5 OF M U TR SPEC RAY GAMMA 6 2 E R U IG F 0 4 *8/’ 0 6 ENTAL TIA N TE O P 845 Mev 5 4 .8 0 8 O D UJ ui o >

120 -82- possible to obtain tha relative counting efficiency of

tha aeintlllatlon eountar by tha following foraula * d . illhz ■ In which

m relative counting afficiancy of the scintilla­ tion counter £ ■ G a u a ray afficiency of the Geiger Muller Tuba h ■ Peak height

Pl/2 * Halt width N m Counting rata of sample measured by a Geiger Muller Tube

Tha relatire counting afficl enoy of the spectro­

meter was determined using C s ^ ^ f M n ^ , Z n ^ , and *

It varied from .36 at 667 kev to .104 *t 1*33 Mar. A

summary of the results is shown in Table XIV, Figure 27*

Counting Afficiency of Scintillation Spectrometer

Isotope Energy V £ h ^1/2 *s Ar c.1^ 667 kev 93.8 • 285 15.5 7.6 . 3 6 0 2.51 0.70

1ta54 8 3 5 kev 3 0 . 0 .370 2.3 7*0 . 2 0 0 1*37 0.69

Z« 6 5 1*12 Mev 756*5 .530 35 4.0 .098 0.78 0.79

Co6 0 1*17 Mev 1.33 Mev 4007 .62 1 6 0 4.2 .104 .61 0.58

In order to determine the detection efficiency in the vicinity of 1*7 Mev, a comparison was made between the relative counting efficiency of the aeintlllatlon 300 0 .3

EFFICIENCY 200 lOO 500. 5 .7 0 0 .5 0 IUE EETO EFCEC CURVE EFFICIENCY DETECTION 7 2 FIGURE 1.00 NR ( ) V E (M Y ENERG 83- HTEETO SPECTRUM PHOTOELECTRON NTI ON C UNTER CO N IO T A L IL T IN C S

L751 0 1.5 5 2 - 84- oouattr and tha photoelectron cross ssetion for lodlns

(23). It vas noted, as shown in Table X7, that the was constant within the range of experimental error of

15% with the exception of ^^ f o r Co6®, This results from its two gamma rays of 1,17 Mev and 1,33 Mev which cannot be distinguished on the scintillation counter. Having established the constant forwas possible to de­ termine the detection efficiency of the scintillation spectrometer for the 1,73 Mev gamma ray and thereby es­ tablish its relative intensity. Recently Shemr. and

Gerhart (24) arrived at the same conclusion regarding the relative efficiency-of the two processes in this energy range,

Following the calibration of the Instrument for energy and establishing a method for determining the relative intensities of several of the gamma rays, a sample of

Co56 was mounted for study. Six gamma rays were detected with energies of ,845 Mev, 1,25 Mev, 1.77 Mev, 2.10 Mev,

2,55 Mev, and 3,07 Mev, These results are in close agree­ ment with those found in the photoelectron spectrum. The gamma speotrum of C o ^ is shown in Figure 28. Figures

26 and 29 show portions of this spectrum. These enlarged plots aided in determining the relative Intensities of the ,845* 1,25 and 1,70 Mev gamma rays. The relative in­ tensities of these gamma rays were found to be in the ratio of 1 to ,69 to ,27, This too is in close agreement COUNTS/SEC. (LOG SCALE) ioor 50 100 0 5 O NRY ON) CNILTON PCRMETER SPECTROM N SCINTILLATIO ) N IO G E R ENERGY F IG U R E 2 8 G A M M A RAY S P E C TR U M O F Co56 (UPPER Co56 F O M U TR C E P S RAY A M M A G 8 2 E R U IG F I ______.0 05 Mev 5 .0 ± 1.20 - T . 8 ' I ______AL IA T N E T O P .7 0 Mev .07 ± 1.77 2 . IO IO . 2 ± v e M 7 0 05 ± . Mev 7 .0 ± 5 .0 2 I ______07 OMev M IO ± 7 .0 3

COUNTS/MINUTE 700 400 0 0 5 0 0 6 - C 0 2 100 0 2 GURE £9 GMA A SE RUM OF Co56 F O M U TR SPEC RAY GAMMA 9 £ E R U IG F 0 4 NTI ON SETOMETER M SPECTRO N IO T A L IL T IN C S 06- 6 -0 0 6 POTENTIAL 0 8 .3 Mev 1.73 IOO 120 -87-

wlth that found In tho photooloctron speotrua* A summary

of tho results Is found in Table IY.

TABLE XY

Relative Intensities of Gaaaa Rays of C o ^

Determined on the Scintillation Spectroaeter

Gamma Ray Area Detection Intensity Relative —— ——— — — Efficiency ______Intensity

.845 249 1.33 1 8 8 1 1.23 84 .65 129 .69 1.73 19 .38 5 0 .27

e. SUMMARY OF RESULTS AND DISCUSSION

The positron spectrum of Co56 appears to have

two components witn end point energies of *4 4 Mev anc*

1.50 Mev respectively* This latter value is slightly

higher than that shown in the Ferai plot as it was cor- 32 rected using the well known end point energy of F 9 1.70 Mev* It is in good agreement with that of Elliott

and Deutsch (l6 ). The second component has an end point

energy of about • u Mev and Its intensity is about 4.3^

of the total spectrum* The *995 Mev positron component reported by Cheng (11) is apparently the result of ex­

cessive scattering* An analysis of the Feral plot of

C o ^ (Fig* 20), obtained before the baffle system was modi*

fled indicates the presence of this component* It was

not observed when the new baffle system was installed*

The study of the low energy portion of the spectrum

of C o ^ was made possible by the use of the enriched -Sa­ ls o tope F e ^ * Until thie time the presence of the C o ^ end Co^& isotopes end their accompanying positron emissions with energies in the region of 300 kev to 4 0 0 kev made it impossible to study the low energy portion of C o ^ with­ out interference from them* In this work it is felt that this sample contains only C o ^ since a thorough investi­ gation was made for evidence of Co^? and Co^®* A study of the low energy portion of the spectrum on the solenoidal spectrometer, equipped with a thin window tube, 140 Mg/cm , did not reveal the characteristic conversion electrons of Co5? at 119 ker and 133 kev, respectively* Also a study in the region -of 800 kev did not show the presenoe of the 805 kev gamma ray associated with Co^®* The low energy positron component with an end point energy of

•44 Mev is shown in Fig* 18*

The gamma rays of Co56 were measured by three methods, the photoelectron spectrum using a uranium radiator, internal conversion electrons, and the scintillation spectro­ meter* The relative intensities were also measured by appropriate methods* Table XVI is a summary of the re­ sults and a comparison with that of Deutsch et al (15)* - 8 9 - XU 56 Oanu Raja of Co Photoeleetron Coavarsloe Scintillation Measurement Electrons Spectrometer Deutseh Energy Intanaity Energy Int. Energy Int. Energy Int.

3.25 .3 3.07 3.25 .2 2 . 6 1 .2 2.55 2.55 .2 2 . 3 1 .2 2 . 1 0 2 . 0 1 .1 1.76 .3 1.77 .27 1.74 .2 1.25 1.23 .6 1 . 2 4 . * 6 1 . 2 3 .69 1 . 2 6 .5 . 6 4 . 8 4 1 . 0 .848 1 . 0 . 8 4 5 2.0 .845 1 . 0 P+ .21

Tha raaulta obtained ara in Tory good agreaaant vHh tha work of Deutseh at al (15) with tha exception of tha following points;

1. Tha discrepancy between tha values of 2.31 May and 2 . 0 1 Mar for one of tha gaauaa rays is outside tha range of experimental error. This is shown in Fig. 23. The

2.01 May gaaaa ray ia not observed. The data obtained from tha acintillation spectrometer can only be considered approximate in this region bacauae of difficulty of in­ terpretation.

2. The ratio of tha relative intensity of tha 1.24

Mev gamma ray to the .84 gamma ray is much larger than that estimated by Deutseh (15).

T^a pey'tftel decay scheme of Co^ 6 i* shown in Fig­ ure 30 . C o ^ decays by tha amission of two positron groups having end point energies of 1.50 May and .44 Mev respec­ tively. Tha 1.50 Mev positron group decays to a second ENERGY 4.60

3.82

log ft = 8.3 ft = 8.4 9.56 3.15 ft(wf-l) - 06 -

2.08 + 4

.8 4 + 2

56 FIGURE JO PARTIAL DECAY SCHEME OF Co excited level 2,08 Mev above the ground level which in turn decays to the ground state by tha emission of two successive g a u i rays of 1,24 Hoy and .84 Mev respectively, Thia ia in agraaaent with praYioua authora (11,15), Using

1*50 Mev aa tha end point energy of tha poaltron emitted and a half life of 76 days, a log ft value of 8*0 ia ob­ tained from Moaskowskl's Table (25)* Thia log ft value la tha maximum value aa 1 0 0 percent branching ratio la aaaumed. According to Mayer and Nordheim (2 6 ) the log ft value of 8 , 0 would make the tranaition firat forbidden, 56 From the ahell model developed by Mayer (27), Co la

*7 / 2 "* P3 / 2 *nd ®Pin 5 according to empirical rulea of Nordheim (26), It ia reasonable to assume the parity la odd, since as Indicated in Fig, 30, the firat and second levels of the daughter, F e ^ , can be assumed to be -*-2 and

+ 4 respectively according to the empirical rule of Gold- haber (28). This assumption ia also consistent with the intensity ratios of the 1,25 Mev and ,84 Mev gamma rays determined from the Internal conversion peak measurements,

Fig, 25, assuming they have £-2 characteristics, because this ratio is in good agreement with the results from photoelectron peak and scintillation counter measurements.

Therefore the decay characteristics of the positron of C#56 may be AI*1, Tee. Thus according to Nordheim's (29)

Classification, the log ft value lies in the region of about 6,5 and the transition is first forbidden. The possibility of 1*2, lee for vhiob tbs log ft Talus is abovs 8 is eliminated because the Fermi plot is linear and the spin assignments. If the ground state of is assumed to be —2 , the high energy component of the spectrum will have the characteristic I*2, las. In this case the log ft value can be calculated from the formula log ft (W0 2 -l) which gives 9.56. This is in good agreement with Nordheim1a (29) observation.

On the basis of an energy balance within the pro­ posed scheme, it appears that the . 4 4 hev £+ decays to a 3rd excited level 3.15 Mev above the ground level which in turn decays to the first excited state by emission of a 2.31 Mev gamma ray, and finally to the ground state by the emission of a .84 gamma ray. Since the analysis of the beta spectrum did not reveal the presence of the .99 Mev it can be assumed that the

1.74 Mev gamma ray is preceded by K capture only and decays to the second excited state 2.08 Mev above the ground level.

No attempt was made to lit the remainder of the gamma rays into the decay scheme as the only basis for assignment was that of the relative intensity. At the time of this writing gamma-gamma coincidence studies were being made which will aid in the final establish­ ment of the decay scheme. 4, THE ISOTOPE Ce5 7 a* THE CONTINUOUS POSITRON SPECTRUM

The enriched sample of Fe*7 used for producing

Co^ 7 by proton bombardment did not bar# auoh a prepon— daranea of tha laotopa in quaation aa did tha aampla of tha anriohad C o ^ * In this case F e ^ and 7a^ 7 vara found in proportiona of 45*692 and 53*532 respectively. Aa a result it was poaaibla that the cobalt iaotopes could make approximately equal contribution to the beta spectrum, if the cross section for the p—n reaction was of the same order of magnitude* In the event the two iaotopes emitted positrons of approximately the same end point energy, the analysis of the Fermi plot for the individual Isotopes would beoome extremely difficult and the results unreli­ able* A study of the data indicated that such a condition existed* A Fermi plot of the beta speotrum of Co^ » ^ 7 sample is shown in Figure 31* Two samples of equal area were prepared, one with an intensity six times that of the other* The purpose of making studies of two auoh samples was to attempt to deteot the presence of scattering from source thickness* If source thickness was a contributing factor to the curvature in the Fermi plot, the two curves eould not be superimposed when they were normalised* It can be seen in Figure 31 that when normalised, the two curves resulting from the studies of the respective samples, one six times as intense aa the other, are INTENSITY RATIO ■OCo56’57 6 -XCo96*57 I ♦ Co56 NORMALIZED TO Co5*® 7

44 1.46 M av

20 2 .5 5 0 3 .5 ENERG Y (m«p2 UN ITS)

FIGURE 3/ FERMI PLOT OF BETA SPECTRUM OF Co56*57 DATA FROM SOLEN01 DAL SPECTROMETER - 9 5 - identical within the rang* of experimental error* la diacuaaed previously, a conaldarable contribution vaa expected from C o ' * Thia ia indicated by the end point of the high energy eoaponent| 1.4-3 Mer from C o ^ . In an attempt to analyse the Feral plot of the continuous positron spectrum of Co57 in the presence of Co-^, the latter was normalized to that obtained from the Co^ * ^ 7 sample. By subtracting these two normalised curves, the influence of C o ^ should be eliminated leaving only the contribution of Co-*7 . Unfortunately, the Fermi plots of the two cobalt isotopes, Co5& and Co^7 , were so nearly identical that the resulting analysis could not be con­ sidered very reliable. However, there appeared to be a low energy component associated with C o ^ of approximately

.44 Mev. This la shown in Fig. 31*

b* THE CONVERSION ELECTRON SPECTRUM OF Co57

The Internal Conversion Electron Spectrum of

Co*’7 is shown in Figure 32. The peaks from the 119 kev and 133 kev gamma rays are easily distinguishable but the resolution of the instrument is not sufficient to enable each to be clearly defined. As discussed in a previous section the current control was not sufficiently sensitive to obtain data of the accuracy demanded by this partieular aspect of the problem* COUNTS / MINUTE 20 4 0 - 0 4 0 5 0 6 BO 0 3 0 7 .20 GURE E R U IG F 2 3 25 .2 ON EETON EK O Co57 OF PEAKS N ELECTRO N IO S R E V N O C AA RM OLNOI SPECTROMETER L A ID O LEN SO FROM DATA ?6- 6 -? 5 AMPERES E R E P M A I5 X 30 .3 1 Kev 119 .35 3 Kev 133 0 4

-97-

o* SUMMARI Ur RESULTS AwD DISCUSSION

ExperlMQtil conditions, including instrument 56 C7 limitations and the prasenca of Co along vith C o , limited the extent of the investigation that could be carried out. The positron, reported by Llvingood (13) to have an end point energy of .280 Mev and by Cheng (11) to have an end point energy of *320 Mev, was detected and its end point energy appeared to be Mev. This is somewhat higher than previously reported. A new approach for obtaining the isotope without interference from C o ^ or must be made, possibly a d,n reaction on the enriched isotope of F e ^ * -98-

5. THE ISOTOPE Co58

a. THE CONTINUOUS POSITRON SPECTRUM OF Co58

Similar to the situation encountered with the

enriched Isotope of Fe^?, the sample enriched in Fe^8 also

contained a considerable quantity of Fe56, 2 3 *4 % Fe^ 8

and 70.4% Fe56. Although the percentage of the respec­

tive isotopes does not Indicate an exceedingly large quan­

tity of Fe^8 , it was far superior to the natural occurring

mixture of the iron Isotopes, 91% Fe8^ and *31% Fe^8 re—

spectively. As previously discussed, the sample was

bombarded with 10 Mev protons snd the resulting cobalt

isotopes separated ffom the target material by t he chemical

procedure described in Seotion I. The Fermi plot of the

positron spectrum obtained from the solenoidal spectro­ meter is shown in Figure 33. In a manner similar to that described for Co->7, the Fermi plot of C o ^ f Fig. 16, was normalized to the Fermi plot of The presence

of the high energy component of Co^ 6 with its end point en­ ergy at 1.4-& Mev is observed. In the lower energy region, a positron component appears which is not present in the

C o ^ plot. The analysis of these data shows this to be a component with an end point energy of .45 Mev. The momentum energy distribution is shown in Figure 34. A similar study of the Co^ * ^ 8 isotopes was made using the thick lens spectrometer. As done in the previous oaae, the Fermi plot of Co-^, Fig. 18 , was normalised to that - 1 1 -

3 0

20

1.48 .4 5 Mev Mev

1.5 2.0 2 .5 3 .0 ENERGY (mGc2 UNITS)

FIGURE 3 3 FERMI PLOT OF BETA SPECTRUM OF Co56*58 DATA FROM SOLENOIDAL SPECTROMETER ~ / o o - 3 5 0 T

3 0 0 -

2 5 0 -

200 N A

150

100

5 0

AMPERES

FIGUREMOMENTUM ENERGY DISTRIBUTION CURVE OF C o 5 6 * 5 0 of C o ^ # 5 8 f After subtraction of the two curves, the lower energy component assigned to Co^S was observed to

have an upper energy limit of *49 Mev. This is shown

in Figure 35* b. THE CONVERSION ELECTRON SPECTRUM OF Co56»58

The thick lens spectrometer was used to study

the conversion electron peak of the .805 Mev gamma ray

of Co^® and the *84.0 Mev gamma ray of C o ^ » These peaks

are ahown in Figure 36. By resolving the two peaks, measuring the portions contributed by the .84 Mev gamma

ray of C o ^ and the .81 Mev gamma ray of C o ^ , and assum­

ing the character of the gamma ray to be E 2

(a « 3*1 x 10-* from Rose's Table (30)), the relative Intensity is estimated to be 54*6 x 104a Comparing this

value with that of the positron spectrum of Fig* 34 which is 7*84 x 1 0 ^, the K/p+ ratio of Co^® is determined as followsi

E/6♦ - 1 ~ P* - £*. - 1 - 54»8 -1-6.0 7.8

This value is in very good agreement with that of Elliott and Deutsoh (17).

o. SUMMARY OF RESULTS AftD DISCUSSION

The first excited state of was determined to be +2. This was established from the intensity of the

*81 Mev gamma ray meaaured by conversion electron studies Co56 OB □---□ Co®6» 58

0.6

02

1.0 1.5 ZO 2 5 3 0 3.5 ENERGY (m0c2 UNITS)

FIGURE 3 S FERMI PLOT OF BETA SPECTRA OF C o 9* 5*, DATA FROM THICK LENS SPECTROMETER COUNTS / MINUTE 0 8 0 5 0 6 0 7 32.33 2 .3 IUE OVRIN LCRN EK OF PEAKS ELECTRON CONVERSION FIGURE o65, AA RM HI ES SPECTROMETER LENS K IC TH FROM DATA Co56*58, 82 Mv C 84 Mv f Co of Mev 4 .8 Co f o Mev 2 .8 OMETER (VOLTS) R E T E M IO T N E T O P 103 - 34. 5 .3 4 .3 36 .3 5658 The positron with an end point energy of 4.80 kev reported by Deutech (17) end by Cheng (11) wee confirmed. The

K/£+ wee found to be 6.0. Thie ie in very good agreement with that of previous workers (17,11). The decay scheme ie shown in Figure 37. 2 3 2 4 2 .2 9 9

.8 0 5

FIGURE J7 PROPOSED DECAY SCHEME OF Co** —106— ill. CHEMICAL ANALYSIS Hi CYCLOTRON ACTIVATION

A. INTRODUCTION

Activation analysis ia • relatively new technique

in the field of analytical chemistry. The ability to

detect quantitlea of elements too minute to be discovered

by atandard analytical methods and freedom of interfer­

ence from elements not being measured have contributed

greatly to its development as an analytical tool.

The application of activation analysis to difficult

analytical problems began shortly after the discovery of

artificial radioactivity. Hevesy and Levi (31*32) first

uaed this method for the determination of trace rare

earths In rare earth mixtures. Shortly thereafter several

other problems of similar nature were solved employing

methods of activation analysis. Minute quantities of

gallium were detected in iron (3 3 )* copper in silver (3 4 )*

and trace elements in biological tissue (3 3 ).

Boyd (36), and Taylor and Harris {37) have recently

presented articles which serve as a guide in activation

analysis procedures.

The sample may be activated by bombarding with either

neutrons or charged particles. A cyclotron, or some simi­

lar charged particle accelerator immediately available

permits studies of radionuclides of relatively short life

and eases the problems of shcedullng bombardment time and

transportation of active lAotopes. This technique* cyclo- -107-

tron activation, is in many cases considerably more diffi­

cult, Decause of competing nuclear reactions, than neutron

pile activation*

Activation analysis may be either qualitative or

quantitative* If it is to be qualitative, the procedure

consists of studying the half life, and the type and

energy of radiation emitted by the radionuclide* Quan­

titative studies are carried out in one of two methods; monitoring the neutron flux, or by comparative measure­

ments* If tne neutron flux Is known, the weight W of

tne element present is given by the equation:

AM W " 6702xl()53fS^- where

A m the activity in disintegrations per second

M « chemical atomic weight of the element in question

f » neutron flux per square centimeter per second

S « the saturation factor 1— or the ratio of the a m o u n t of activity produced in time t to that produced in infinite time

the activation cross section (probabilityJ for the reaction in square centimeters per atom

It is to be noted that in this equation, the final re­ sult is complicated by the monitoring of the neutron flux, the assumption of constant rate of production of the radioactivity, and the relationship between the ob­ served counting rate and the disintegrations per second* -106-

Quantltatlve studies by comparative Measurements eliminate these problems oy bonbaraing a standard and an unknown under identical conditions and comparing the two counting rates corrected by the simple ratio

Total Activity from ll.mant I In Unknown . Hi 6 8 Of X la. VnEaW fl Total Activity from Mass of X in Standard Element X in Standard The purpose of the work discussed in the following sections is to expand the application of cyclotron acti­ vation analysis• One phase of the work was devoted to determining the presence of beryllium in a beryllium- copper alloy. The second phase was that of detecting nickel in an ore. - 1 0 9 -

B* THE BERYLLIUM-COPPER ALLOT

1. EXPERIMENTAL PROCEDURE AND RESULTS

A sample of beryllium metal was bombarded In the cy­ clotron of The Ohio State University for onehour with

15 Mev alpha particles, A portion of the target was placed in a counting box equipped with an end window Oeiger Muller

Tube and a Potter Decade Scaler, The decay of the acti­ vated sample was followed. It was expected that the bom­ bardment would yield an c,2 n type reaction with as the product. The half life of this Isotope is 2 0 minutes.

Figure 38 is a plot of the rate of decay showing a half life slightly greater than 19 minutes, A study of the energy of the emitted particle was made using aluminum absorbers. This is shown In Figure 39. The end point energy being approximately 1 Mev*

This bombardment was followed by a second activation in which a beryllium—copper alloy containing approximately one percent beryllium was bombarded with 15 Mev alpha particles for one hour. The activated sample was removed from the target mount and a portion of it placed in a counting box equipped similar to that described previously.

The counting rates were recorded with no absorber, a 111 2 mg/cm aluminum absorber capable of absorbing all beta energies less than ,4 Mev, and a 4.00 mg/cm aluminum ab­ sorber capable of absorbing all beta energies less than

1 Mev, Figure 40 shows the characteristic decay rate, COUNTS / MINUTE 400 0 0 7 0 0 5 - 0 0 6 0 0 2 GURE30 DCY UV O C* RDCD Y ALPHA BY PRODUCED * C OF CURVE DECAY 0 3 E R U IG F - 4 2 NT F RYLUM METAL M M YLLIU ER B OF T EN M D R A B M O B IE MI ) S E T U IN (M TIME - o / / - 6 4 UTE NUTES, C , S E T U IN M 6 9 120

COUNTS/MINUTE 10,000 OOO- O O IO IOO 0 5 O - Y LH B AD N OF EYLU METAL. BERYLLIUM F O ENT BARDM M BO ALPHA BY D E C U D O R P FIGURE FIGURE 9 3 200 EN D P O IN T E N E R G Y OF POSITRON O F F O POSITRON OF Y G R E N E T IN O P D EN AG I AL NUM

* C IOOO

5 0 0 0

3 0 0 0

ui »- 3 Z

h- Z 2000 - 3 s 21 Minutes, C o o

0 20 6 0 IOO 140 1 8 0 TIME (MINUTES)

FIGURE VO DECAY CURVE OF C ~ PRODUCED BY ALPHA BOMBARD­ MENT OF Cu-Be ALLOY, ANALYSIS OF DIFFERENCE IN COUNTING RATE WITH .4 MEV ALUMINUM ABSORBER AND I MEV ALUMINUM ABSORBER INSERTED BETWEEN SAMPLE AND TUBE 21 minutes, to be expected of the product, C ^ , formed by the bombardment of beryllium by alpha particles*

2. DISCUSSION

The purpose of selecting a simple alloy for study in the development of an activation analysis procedure was that it should present relatively easily recognizable characteristics of the radionuclide formed from the bom­ bardment. These should also be rather free of numerous otaer interfering activities. After careful study of

Sullivan's Chart of Nuclear Species (38), it was decided to attempt a one hour alpha bombardment of a beryllium- copper alloy. The reaction expected from the alpha bombardment of beryllium was a,2 n giving C ^ with a 2 0 * 5 minute half life and a .99 Mev fl + . Other isotopes that could have been formed have a very short half life,

19.1 seconds or are stable, C ^ and C ^ . The radioactive isotopes formed from the alpha bombardment of copper would be either Ga^ having a 6 8 minute half life and a 1.9 Mev £+ or G a ^ having a 9.4 hour half life and a 3*5 Mev £ + .

These activities could easily be distinguished from that of C H . As discussed previously, the sample was bombarded and a study of the activity was made in which aluminum absorbers of 1 1 1 mg/cm^ and 4 0 0 mg/c* 2 were inserted be­ tween the counting tube and the source. The absorbers were capable of absorbing beta particles of energies less than *4 Mo t and 1*0 Mev respectively. Since the upper energy limit of tne beta particle of was *99 Mev, Interference from other activities of greater energy could be eliminated by subtraction of the counting rate measured through a 2 4,00 mg/cm aluminum absorber from that measured without an absorber. The use of the 111 mg/om^ absorber was to ab­ sorb beta particles of energies less than ,4> Mev in the event they were present and interfered with the identifica­ tion of the radionuclide the product of the a,2 n reaction on beryllium* 3* SUMMARY

The characteristic half life of C*-1* and the end point energy of the emitted particle formed by the alpha bom­ bardment of oeryllium in a beryllium—copper alloy were detected. Fig. 38 shows the presence of the 20*5 minute component belonging to C ^ . Fig. 39 is a plot of an aluminum absorption study showing the end point energy,

.99 Mev, of the positron given off by . These results, obtained when beryllium metal was bombarded with alpha particles, established the fact that the desired reaction

Be9 (a,2n) C1 1 was produced. Bombardment of the beryl­ lium—copper alloy containing one percent beryllium was then attempted. Figure 4-0 is a plot of this study showing the contribution from beryllium, twenty—one minute com­ ponent of C11. Thus it was possible to detect the pre­ sence of beryllium in the beryllium-coppvr alloy* - 1 1 5 -

G. THE NICKEL ORE

1, EXPERIMENTAL PROCEDURE AND RESULTS

A sample of hlckel ore was obtained from the Pacific

Nickel Company, Vancouver, British Columbia, The ore was prepared for analysis by crushing it in a steel mortar using a steel pestle*. One half of the entire sample available, approximately two kilograms was used in order to obtain as representative a sample as possible. The ore was ground until the entire sample was of such partiole slse as to pass through a # 1 0 0 mesh screen. Particular care was exercised In keeping the loss of the dust particles to a minimum. The final product was thoroughly mixed and quartered when taking a sample for study. Quartering consists of placing the material in question on a sheet of weighing paper and dividing it into four sections.

Two alternate sections are mixed and placed on a second sheet of weighing paper and quartered a second time. The process is repeated until a sample of suitable size is ob­ tained.

As mentioned previously one aim of this work was to expand the application of the cyclotron as a tool in activation analysis. Therefore In developing the procedure for handling the nickel ore by activation analysis, it was necessary to examine it by several established quan­ titative methods. In order to have an idea of the numer­ ous elements that might possibly interfere with the - 1 1 6 -

detection of the nickel, a spectrographic analysis of

the ore was made on a Three Meter Grating Spectrograph

using spectr©graphically pure carbon electrodes. In

this study the trace elements copper, chromium, nickel,

boron, cobalt, gallium, zinc, lead and vanadiun were

found. In addition the more common elements found in most

ores; iron, silicon, silver, magnesium, calcium, aluminum

and manganese, were present. The existence of copper,

silver, cobalt, zinc, and lead was further confirmed by

spot tests (39), Standard quantitative methods were em­

ployed to determine the amount of nickel and copper

present. Following this analysis, a thorough investiga­

tion of the possible nuclear reactions that could be ex­ pected was made based on the data taken from Sullivan's

Trilinear Chart of Nuclear Species (38). The Bummary of this study was used as the baBis for establishing the conditions which would permit the detection of the characteristic half lives of the elements in question#

Before bombarding the ore, a sample was placed in a counting box equipped with an end window Geiger Muller

Tube and Potter Decade Scaler, No activity could be de­ tected in the sample. A second sample was placed in a con­ tinuously operating cloud chamber. No alpha tracks were observed# To insure the cloud chamber was in working order, a known alpha emitter was simultaneously placed in the field# -117-

Approxlmately . 1 2 grams of the powdered ore were

pressed In an aluminum target holder and bombarded thirty

minutes with 7 Mev protons at The Ohio State University

Cyclotron. The sample was removed from the cyclotron and

placed in a counting box* The counting rate per minute

was measured through aluminum absorbers weighing 8 3 0 ag/ca^,

1100 mg/cm2j and 1650 mg/cm^. These absorbers were of

sufficient thickness to absorb all beta emissions of

energies less than 1.8 Mev, 2*3 Mev and 3*3 Mev respec­

tively. Figure 41 shows the comparison of the three decay

curves obtained.

To study the decay rate of a beta emission within a

particular energy range, for example 1.8 Mev to 2.3 Mev,

a subtraction was made of the counting rates as measured

through an aluminum absorber capable of absorbing all

beta energies less than 2.3 Mev and that measured through

an absorber capable of absorbing all beta emissions with

energies less than 1*8 Mev. A second study was carried out

in which the decay rate of the radionuclide was deter­

mined by plotting the difference between the counting

rate through an aluminum absorber capable of absorbing

all beta particles of less than 2.3 Mev and the counting

rate through an aluminum absorber capable of absorbing all beta particles less than 3.3 Mev. In both studies

subtractions had to be made at simultaneous time intervals*

Figure 42 Is a plot of the difference in counting rate =0 0 4 3=30 o 26

I V/ NI OR BMADD NUT ITH W S TE U IN M 0 3 BOMBARDED RE O L E K IC N / V E R U FIG RATE (COUNTS/MINUTE) 0 0 5 - OFF CYCLOTRON I O O -IO 30. 0 :3 3 0 0 0 5 - MEV PROTONS V E M 7 -K 2.3 2.3 -K O 1.8 ■O ' //0 ' 3.3 3.3 00 1 20 2 10 0 :0 5 0 5 0 4 0 3 0 2 ME IM T M ev Al Al ev M Mev Al Al Mev Mev Mev I BOBR N POSITION IN ABSORBER AI ABSORBER ABSORBER BOBR N POSmON IN ABSORBER IN IN POSITION

— //?- 5 0 0 0

------COMPONENT ASSIGNED TO Cu63

co

1000 T ~ * 38 X ui MINUTES K

K

5 0 0

3 30 AO 50 4:00 10 20 30 40 50 500 10 TIME

FIGURED ANALYSIS OF DIFFERENCE OF COUNTING RATE WITH 1.8 MEV ALUMINUM ABSORBER (FIG >/) AND 2 2 MEV ALUMINUM ABSORBER IFIG.47 ) INSERTED BETWEEN SAMPLE AND COUNTER. - 1 2 0 - between the emissions passing through an aluminum ab­ sorber capable of absorbing all beta particles with energies less than 1*8 Mev and those emissions passing through an aluminum absorber capable of absorbing all beta particles with energies less than 2.3 Mev. This shows the characteristic decay curve of Zn resulting £ q from a p,n reaction with Cu •

Figure 43 is a plot of the difference in counting rate through an aluminum absorber capable of absorbing all beta particles of energies less than 2.3 Mev and the counting rate through ar aluminum absorber capable of absorbing all beta particles of energies less than p. 3 6> ? Mev. This shows the characteristic decay curve of Cu resulting from a p,n reaction with Ifi^*

Figure 44 is a plot of the analysis of the differ­ ence in counting rate through an aluminum absorber capable of absorbing all beta particles of energies less than

1.8 Mev and the counting rate through an aluminum absorber capable of absorbing ail beta particles of energies less than 3.3 Mev. Tnis plot snows tne contribution from ootn nicKei and copper. CYC LOTRON CYC

F - OFF COUNTING RATE (COUNTS/MINUTE) - 0 0 5 0 5 LMNM BOBR FG4) NETD BETWEEN INSERTED (FIG.47) ABSORBER ALUMINUM IH E AUIU ASRE (I /) N 33 MEV M 3 3 AND RATE ) COUNTER V/ (FIG COUNTING AND OF ABSORBER PLE ALUMINUM DIFFERENCE SAM OF MEV ANALYSIS 3 2 WITH 3 ^ FIGURE =0 O 0 2 IO 0 0 4 0 5 0 4 3=30 -- 1. MINUTES 10.5 * T-i- ~/Zi~ OPNN ASGE T Ni TO ASSIGNED COMPONENT TIME O 10 0 0 & 0 5 AO 0 3

62

NSERT EWEN AMPE N CUTN TUBE COUNTING AND PLE M SA EEN BETW D TE R E S IN I W ANALSS DIF N COUNTI AE ITH W RATE ABSORBER M U G IN IN M T LU N A U O C V E M IN 3.3 E C N E R AND IFFE D ABSORBER F O INUM LYSIS A ALUM N A EV M LB W E R U FIG Q 0 0 5 COUNTS/MINUTE >- X X K Y LOTRON CYC OFF 3 *3 3 100 10 :1 4 0 5 : 3 0 3 : 3 I ______-f2Z~ TIME ,» IO M in u te s , , s te u in M IO ,» =40 Mi es, 3 * n Z , s te u in M 0 4 = i T 0 5 : 4 0 3 : 4 Cu*e

- 1 2 3

2, DISCUSSION

The nickel ore wee analysed by spectrographic analy— sla in order to aid in developing conditions necessary for eliminating those activities emitted that would interfere with the detection of the radionuclide formed by bombarding nickel. The amount of nickel was deter­ mined by standard analytical methods using dimethyl glyoxime. The analysis showed 1,38 percent nickel in the ore, A study of the spectrographic analysis combined with Sullivan's Trillnear Chart of Nuclear Species ^38) was made in oraer to establish conditions necessary ior the detection of the elements in question. From this information it was decided that a proton bombardment, in hopes of getting a p,n reaction from nickel, would pro­ duce the least number of competing nuclear reactions and minimum interference from other elements in the ore.

It was first hoped that the nickel could be Identified by the measurement of the positron emissions of energies of 1,8 and 3*3 Mev respectively from twenty-four and six- tenth* minute Cu*5^. To accomplish this aluminum absorbers of such thickness to absorb all beta emissions less than

1,8 Mev a net 3.3 Mev respectively were used. The charac­ teristic decay curve could not be located. It was decided that interference of long lived activities might have mask the desired activity. Thus the bombardment time was re­ duced from one and one-half hours to one-half hour. As -124- can be seen from Fig. 43# th® 24*6 minute activity of C u ^ was not found but a 10.5 minute activity of Cu # th® pro— 62 duct of a p,n reaction on Ni was detected.

It was also possible to detect th® presence of copper from the characteristic decay curve of Zn4*3 with a 3 8 minute half life resulting from the nuclear reaction:

Cu63 (p,n)Zn6 3

Seventy—five hundredths percent copper was detected in the ore by standard analytical methods.

An analysis of the difference of the counting rate using aluminum absorbers capable of absorbing beta particles of energies less than 1.8 Mev and those of energies less than 3*3 Mev is shown in Fig, 44* A study of this data shows contributions from both nickel and copper. The component showing the forty minute naif life can be as­ signed to the presence of copper according to the reaction:

Cu6 3 (p,n)Zn6 3

Zn^ 3 has a half life of thirty eight minutes. The forty minute component is within the range of experimental error.

The component showing the ten minute half life can be o 'Signed to the presence of nickel according to the re­ action:

Hi62 (p#n)Cu62 -125-

3, SUMMARY

Nickel and copper were detected in nickel ore in the presence of iron, silicon, magnesium, zinc, lead, and vanadium employing cyclotron activation and identify­ ing the characteristic decay rate, A study of Figs,

4 2 , 4 3 and 4 4 shows that it is quite feasible to extend the application of activation analysis to ores containing nickel and copper even if there are a large number of accompanying elements present in the sample. The success of the analysis depends upon the proper selection of bambardment time, bombarding particle, and aluminum absorbers. The conditions ior this particular study in wnicn the elements in question could be detected without chemical treatment of the ore were:

1* Select high energy (7 Mev) protons as the

bombarding particle

2, Control the bombarding time to thirty minutes

3» Follow the decay of the activated sample using

aluminum absorbers of sufficient thickness to

absorb all beta emissions of energies less

than 1,8 Mev, 2,3 Mev, and 3,3 Mev respectively. -126-

BIBLIOGRAPHY

(1) I. Kolthoff and E. Sandell, "Textbook of Quanti­ tative Inorganic Analysis", New York: Macmillan Co.,

1943, p. 32. (2) P. Welcher, "Organic Analytical Reagents", New

York, D. Van i.orstrand Co., 1947. Vol. Ill, p. 318.

(3) Ilinski and Knorre, Ber., 18 (1885) p. 699-704*

(4 ) K. Tamic, H. Hosimya and T. Ikeda, J. Chem.

Soc., 61 U940) p. 269-276.

(^>) , Swift, J. Am. Chem. Soc., (1924) Vol. 4 6 , p. 2375.

(6 ) G. Dodson, G. Forsey and E. Swift, d. Am.

Cham. Soc., 58 (1936) p. 2573.

(7) E. Swift, J. Am. Cham. Soc., (1924) Vol. 4 6 p. 2378.

(8 ; I. Kolthoff and E. Sandell, "Textbook of Quanti­ tative Inorganic Analysis", New York: Macmillan Co., 1948,

P. 321.

(9) L. A. Kilay, Ph.D. Dissertation, The Ohio State

University (1952).

(10) E. Jensen, R. Nicholas, S. Clement and A. Rohm,

Phys. Rev., 85, 1 (1952) p. 112.

(11) L. Cheng, Ph.D. Dissertation, The Onio State

University (1952).

(12) K. Haskins, Ph.D. Dissertation, The Ohio State

University (1952). -127-

(13) J* Livingood and G. Seaborg, Phy. ReT.f 60

(1941) P. 913* (14) C. Cook, P. McDaniel, Phy. Rev., 62 (1942) p. 612.

(15) I* • G. Elliott and M. Deutsch, Phy. Rev., 6 4 (1943) p. 321.

(16) G . jrlesset, Phy. Rev., 62 (1942) p. 181.

(17) M. Deutsch and L. G. Elliott, Phy. Rev., 65

(1*4) p. 211.

(18) W. Good, D. reaslee and M, Deutsch, Phy. Rev.,

69 (1946), p. 313.

(19) G. Owens, H. Primakoff, Rev. of Sci. Inst., 21

(1950) p. 447.

(20) K. Siagbahn, Fhysica XVIII (1952) p. 1043*

(21) C. Davisson and R. Evans, Rev. of Mod. Phys.,

24 (1952) p. 97.

(22) Rutherford, Chadwick, Ellis, "Radiations from

Radioactive Substances", p. 4 6 4 *

(23) E. Bleuler and G. Goldsmith "axpuriffientai,

Nucleonics", Rinehart and Go., Inc., (1952).

(2 4 ) R. Sherr and J. Gerhart, Phy. Rev,, 91 (1953) p. 909.

(25) S. Moszkowski, Phy. Rev., 82 (1951) p. 35.

(2 6 ) M. Mayer, S. Moszkowski and L. Nordhelm, R ev. of Mod. Phy., 23 (1951), p. 315.

(27) M. Mayer, Phy. Rev., 78 (1950) p. 16. - 1 2 8 -

(28) M. Goldhaber, A. Sunyar, Phy. Rev., 83 (1951) p. 906.

(29) L. Nordnelm, Rev. of Mod. Phy., 23 (1951) p. 322.

(3 0 ) M. Rose, G. Goertzel and C. Perry, ORnL-1023

1951.

(31) 0* HeveBy, H. Levi, Kgl. Danske Vldenskab,

Selskab Matb-fys. Medd, 14 Ho. 5 1,1936).

(32) Ibid.. 15 Ko. 11 U938).

(33) 0. Seaborg and J. Livingood, J. Am. Chem. So©.,

6 0 (1938) p. 1764.

(34) 0. King and W. Henderson, i’hys. Rev., 5 6 (1939) p. 1169.

(35) 0. Tobias and R. Dunn, Atomic Energy Commission

AECD (1948) p. 2296.

(3 6 ) C. Boyd, Analytical Chemistry, 21 (1949) p* 335.

(37) T. Taylor and W. Havens, Jr., Hucleonics 6 ,

No. 4, 54 (1950).

(38) W. Sullivan "Trilinear Chart of Nuclear Species”

New York, John Wiley and Son*, Inc., 1949.

(39) Fiegl, "Qualitative Analysis by Spot Tests",

New York, Elsevier Publishing Co., 3rd Ed. (1946). - 1 2 9 -

AUTOBIOGRAPHY

I, James L. Dick, was born in Harrisburg, Pennsyl­ vania, July 8, 1921. I received my elementary education at New Florence, Pennsylvania and Indiana, Pennaylvaniaj my secondary education at New Florence High School, New

Florence, Pennsylvania. My undergraduate training was taken at State Teachers College, Indiana, Pennsylvania, from which I received the degree of Bachelor of Science in Education in 1942. I held the position of instructor for one year at Homer City High School, Homer City,

Pennsylvania, teaching mathematics and science. I joined the United States Air Force in 1943, receiving a com­ mission of second lieutenant and areonautical rating of navigator the latter part of the same year. Following a period of eighteen months serving as a navigation in­ structor, I entered meteorological training at Uhanute

Field, Illinois. After successfully completing the course

I served as a weather forecaster with the United States

Air Force until the date of my discharge, September, 1946.

I immediately entered the University of Pittsburgh,

Pittsburgh, Pennsylvania and received a Master of Science

Degree in Chemistry in June, 1948, After holding the position of research chemist at the Solvay Process Com­ pany, Syracuse, New York, for six montha# I was granted a Regular Commission in the United States Air Force in

November, 1948. Since this date I have served continu— ously with the United States Air Force which Included a combat tour during the Korean Affair* I was selected for graduate training in the field of nuclear chemistry at The Ohio State University by the United States Air

Force Institute of Technology in June, 1951*