"
..
:
:
.
I OFT
ORNLP
1216
:
1
.
.
i
1
.
+
.
.
.
2
.
:
:
벨
​wind
.
4.
mais
11:25 1.4 TL
MICROCOPY RESOLUTION TEST CHART
NATIONAL BUREAU OF STANDARDS -1963
TIM
MAY 5 7988
FO
OF Rey part
expressed o
to thelo rope
-LEGAL NOTICE-
The report we propered win account of Government spoonored wort. Medtkor the United
Hautes, sor the Cocaniutos, por my parton actos a bowl of the Comantistas:
A. Makes wynranty or reputation, emprened or implied, with respect to the accu-
noy, capion, or untilos a tbe information contabond to the report, or that the we
of way inkomation, appunto, method, or procoas declared to dla roport may not lafrioco
printaly orond meta; or
B. ARDI Lay Labatas with respect to the roof, or for denneue resulthy troca the
und way to formation, appunto, menthod, or procuu discloud b the report
As wund to be abort, sporton acetog og budall of the Commissioo" tacladu, my on-
plogue or catructor of the Commission, or omployers of mal contractor, to the extent that
Rock employee or cotractor of the Commis100, or employee of much contractor propers,
denumirates, or provides accuse to, may taformation pornnt to be deploymeat or contract
wil the Coomington, or hin employment with each contructor.
C
with the Union Carbide Comporation,
*Research sponsored by the U. S. Atomic Energy Commission under contract
R ibowit.
RELLEST
DISTAEUTION
i jlis
Dobrice OV FILE IN RECEWINA
OF PATENT WTEREST
NOT FØR PUBLIC REVEASE OFFICAD
EMADE. OFFICIAU ;
ED.; REPOXT
Oak Ridge, Tennessee
Oak Ridge National Laboratory
Analytical Chemistry Division
E. Ricci, R. L., Hahn, J. E. Strain, and F. F'. Dyer .
SHE ACTIVATION ANALYSIS*
Coff. 456405-45
ORAL P_1216


ARE ON EILE IN THE RECEIVING SECTION.
THE PUBLIC IS APPROVED. PROCEDURES
PATENT CLEARANCE OBTAINED. RELEASE TO
IWIAI
'anni
....
..
2
et
: INTRODUCTION
The advantages of the use of He sons in activation analysis for low-Z
elements have been pointed out by ilerkowitz and Mahony (8). Because many
reactions of the particles with low-2 elements are exoergic, and because
the Coulomo barriers of these elements are relatively low, the energy of
the She ions does not need to be hien. Conversely, if the atomic number of
the matrix is large enough, the raatrix interference becomes negligible because
the low-energy He ions cannot overcome the Coulomb barrier of the matrix
nuclei. The above authors and Demilat (2) have analyzed oxygen by the reactions
160 >Hle, p)?y and 160(*Hle, a) 28 ve per 18 : (1)
in several matrices and report limits of detection in the range of parts per
billion (0.2.0.). All hese interesting features of 3He activation analysis
led Markowitz and Mahony to suggest the possibility of constructing a small,
reasonably-priced, 8 m.e.v. cyclotron for the activation analysis.
Our work attempts to extend the study of He irradiations in several
directions: (a) Radioactivation of oxygen leading to short-lived nuclides,
as well as to 1.8-br. -OF. (b) Activation analysis for other light elements,
such as beryllium, carbon, nitrogen, and fluorine. (c)' Determination of
neutron yields for light elemento (lithium, beryllium, boron, carbon, and
oxygen) exposed to She irradiation, to investigate the applicability of the
small she cyclotron as a neutron source.
THEORY
Charged-Particle Activation
The production of radionuclide B in a thin target by the interaction of
nuclei A with a beam of charged particles can be characterized by the expression (4),
D = Inot (thin target). - *
(2)
.
pere, D is the number of disintegrations per second (dps) of nuclide B at
saturation; I, the number of charged particles per second (equivalent to
ú current) in the beam; n, the number of atoms of nuclide A per mg of target;
C, the cross section, in cm, for the nuclear reaction A + B; and t, the
thickness of the target in me/cm. In veneral, o is a function of the energy
of the charged particles; however, the energy of the beam is degraded only
slightly in a thin target so that o in Equation 2 can be considered constant.
Contrastingly, in a target of thickness greater than the range, R, of the
charged particles (t > R), the observed disintegration rate, D, should be
expressed as
D = Inles dt (thick target)
(3)
where on represents the variation of cross section with thickness.
The integral of 0+at in Equation 3, which we may call the integral cross
section, depends upon properties of the target material, as well as upon the
characteristics of the nuclear reaction being studied. By using particle-
energy limits (E) for the integral cross section, we can write
E op (de/QE)DE
(4)
where og represents the variation of cross section with particle energy (the
excitation function), åt/ae is the negative reciprocal of the stopping power
of the target for the charged particles, and the interval E to o accounts for
the energy lost by the particles in travelling distance R. From Equation 4,
we see that we can calculate integral cross sections from the excitation
function for a given nuclear reaction; values of the stopping power, or of
the variation of range with energy, are known for many materials (3). We
can also measure the integral cross section of each material, for a given
reactioa, by using Equation 3..
For the purposes of activation analysis, however, we wish to circumvent
these somewhat tedious determinations. Let us define an average cross section.
Yo
CR .
Jö 07åt
J
16
Jo
PR at
=
lowlanlam
or(at/aE)E
(at/aE)ac
(5)
70
E
(6)
-
If we introduce the expression for the stopping power (4) in Equation 5, a
further simple approximation leads to:
OE ELE
JE ELE
Then, ő is practically independent of any properties of the target material,
and is constant for a given nuclear reaction and a fixed energy interval, E
to 0. Thus, õ is a satisfactory parameter for use in activation analysis.
From the definition of ö in Equation 5, Equation 3 can be rewritten as
D = Ingla dt = InGR
(7)
To determine 7 for a particular reaction in a given material, we obviously
must know the range of the charged particles in that material. The accurate
measurement of 5 for one material enables us to determine the integral cross:
section, or, for all materials.
To illustrate that ö does not vary much with changes in target material,
we calculated 7 for the reaction 160 + 3H1e - 185 for several targets, from
data given in References 8 and 3. The resulting relative values of 7, cal-
culated from Equation 5, and normalized with respect to the value for beryllium,
are plotted vs. atomic number, 2, in Figure 1. From 2 = 4 to 2 = 95, ő changes
(Figure 1)
by 8%. In fact, ö changes by only 3% for Z values from 4 to 57.
OW
.
Comparator Method
The fact that ő is approximately independent of properties of the
target material makes possible the use of the comparator technique of
activation analysis. For example, by Equation I, we can determine nye the
number of 100 nuclei per me of unknown sample, x, by comparison with a stan.
dard, s, that contains a known amount of oxygen, ne. Separate irradiations
of unknown and standard are performed so that the initial kinetic energy of
the charged particles impinging on each sample is constant. Thus, ő for
the giver reaction has the same value for both unknown and standard. Then
from Equation 7,
1 ano
n
I
R
(8)
Sss
independent of . The counting rates of -°F. in the unknown (a) and the
standard (a ) are obtained in the experiment. Use of a calibrated Faraday
cup in the irradiations determines Ix and Is. The values of Ny and Rs are
known (3). Therefore, n, is determined. Note that by choosing a standard
such that the major constituents of both standard and unknown matrix bave
similar atomic number, we can be assured that the error introduced in Equa-
tion 8 by taking Ô to be constant is small. The important conclusion to be
drawn from this equation is that the comparator method of charged particle
activation analysis can certainly be applied to many analytical problems
without our having any detailed information about ē.
Sensitivities
the number of atoms of nuclide A per mg of target is given by
n = 10-9pN /
..
(9)
where I is the weight fraction, in p.p.b., of nuclide A in the target, N
is Avogadro's number, and w is the gram-atomic weight of A, expressed in mg...
Let us now define the sensitivity,
S = D/f in dps/p.p.b.
(10)
It is clear from Equations 7, 9, and 3.0 that for a given reaction and for
fixed beam current and bombarding energy,
(11)
S = 10^'IN ÖR/W = KÖR
where K is constant; therefore, s/R is also constant. Furthermore, if we
assume that we can detect a minimum disintegration rate of 100 dps, we can
define the limit of detection, C, in p.p.b. as
C = 100/5 = 100/(KÕR)
(12)
Thus, we see that S and C both depend upon the range, R. Once ő has been
determined for a given reaction in one target matrix, S and C can be calcu-
lated for any other target matrix by simply substituting the appropriate
range-value in Equations 11 and 12. We see then that charged-particle acti-
vation analysis, used either in the determination of sensitivities or in the
comparator method, reduces to almost the simplicity of neutron activation
analysis (7).
From Equation 2, we may obtain expressions analogous to Equations 11:
and 12, respectively, for the sensitivity, se and the detection limit, Ces
in thin target irradiations. .
EXPERIMENTAL
The irradiations were performed at the ORNL 5.5 m.e.v. Van de Graalt
accelerator. Energies up to 10 m.e.v. were attainable by acceleration of
doubly ionized the particles. The highest beam currents were about 0.05
amp. The targets were of four types: (a) thick discs of quartz (0.065 in.)
and zirconium oxide (0.02 in.) were used to determine oxygen-analysis pora-
meters, (o) thin foils of beryllium, teflon, mylar, and nylon (0.00025 -
0.00075 in.) were irradiated to measure cross sections and detection limits
for reactions of low-z elements with She sons, (c) targets of lithium bydride,
beryllium, boron, and carbon (0.02 - 0.08 in.) were irradiated with 5 and 10 .
m.e.v. She ions to measure neutron yields, and (a) discs of carbon, aluminum,
and copper (0.04 in.) were used as monitors to measure the neutrons and
gamma rays proäuced by the targets (c).
Tae nuclides produced in the irradiation with Ble ions were 21c(20.5 min),
23.0( 10.0 min), 240 (72 sec), 150 (124 sec), 177 (66 sec), ana 185 (110 min).
These nuclides are all positron emitters; then, the 0.511 m.e.v. annihilation
gamma rays were counted with a 3 in. x 3 in. NaI(12) crystal and multichannel
analyzer. Since no chemical separations were performed, decay curve analysis
of the data was made by means of Cumming's least-squares code (1) on the IBM
7090 computer. The counting rates of +40 and +'F were resolved before the
computer analysis, by using the 2.31 m.e.v. gamma ray of *4o. All disinte-
gration rates were calculated by Heath's method (5). In the neutron-yiela
determinations, a BF, counter was employed; it was calibrated with a neutron
source (isotropic) of americium and boron.
RESULTS AND DISCUSSION
Since no enriched isotopes were irradiated in our experiments, the pro- .
duction of a particular radionuclide by irradiation of an elerent may be due
to more than one nuclear reaction; for example, 12 c(He, c) *7c and 1301 Bre, an)l2c
ox 270(3ge, pm) 285, 270(>1e, 2n)28Ne-> 189, 160(3He, p) 285 and 160(3pe,m) 2816-> 185.
...
.
...............
I
.
LT
-8.
Only the last two Teactions, however, are explicitly pointed out when results
are reported. In all other cases, just the reaction thought to be the most.
important is mentioned.
The errors introduced in our experiments were estimated and are always
expressed as percent standard deviations. Jorrors attributed to experimento
involving thick and thin targets and monitors are those due to counting
statistics, decay-curve analysis, uncertainties in decay parameters, she
beam-intensity measurements and, when used in the calculations, counting
efficiencies and ranges. The largest errors, those from counting and decay-
curve analysis combined, were normally obtained from the output of the CLSQ
program (1). These were then propagated, in each case, with an overall error
of + 5%, attributed to all the other effects. In general, then, the final
standard deviation is given for each particular measurement in the tables
of results.
Thick-Target Experiments
Results for the thick targets of zirconium oxide and quartz are listed
in Table I. The detection limits, C, are interestingly very low, even when
(Table I) .
they are calculated for a minimum detectable disintegration rate of 100 dps, .
a value that is somewhat conservative. The C values are also in agreement with .
sensitivities reported by Markowitz and Mahony (8) for the reactions -60 - LOF.
It is worth pointing out that the detection limit for the reaction 100-150.
is quite close to that of the reactions producing 1°F; therefore, in many
analyses, it will be possible to measure the short-lived 250(2.07 min) rather
than 10F(110 min), with equivalent sensitivity, but with net gain in analysis
speed and efficiency.
-9-
The ranges, R, for 10 m.e.v. die ions in zirconium oxide and quartz were
calculated from range-energy tables (3). Their values are 0.043 mom and 0.064 mm,
respectively. It should be noted that, although these ranges are used in
Equation 3 for thick targets, they correspond, for all practical purposes, to
thin layers; therefore, She activation analysis normally tests only a small
thickness of the sample. The ranges of Table I are actually the thicknesses
traversed by a He ion of 10 m.e.v. initial energy, until its energy is degraded
to the threshold energy of the reaction involved. For an exoergic reaction, the
range of Table I is simply the total thickness traversed by a 10 m.e.v. He
particle in the given matrix.
Results for the reactions 100 - LOF, in quartz and zirconium oxide; show
that the values s/R are the same for both targets, as predicted by Equation 11.
The linear plot of C vs 1/R is shown in Figure 2; each straight line has been
. (Figure 2) ...
calculated according to Equation 12 with the data of Table I. The dependence
S
of the limit of detection, c, on the atomic number, 2, of the matrix for the
reactions on thick targets of Table I, is shown in Figure 3, constructed from
(Figure 3;
data of Figure 2 and Reference 3. We see that, since the range increases with
atomic number, C decreases with increasine; 2.
Tain-Target Experiments
Table II presents an intercomparison of detection limits for reactions
(Table II)
of He lons with low-z elements. These results have been calculated from
excitation-function experiments, which involved irradiations of thin foils
of beryllium, mylar, teflon, and nylon. The energy intervals in the table
-10-
VILK
.
show the respective kinetic energies of the 'He ions entering and leaving
a particular target foil. The Ce values are thin-target detection :.:
limits. These results have been normalized (c) to constant thick-
ness (2 mg/cm"), so that a comparison of detection limits can be made. The
Co values are only approximate because variations of cross section with
thickness have been ignored. Table II shows that the corresponding Ch
values for 130 and OF are about one order of magnitude larger than those for
the thick targets.
Preliminary, but very interesting conclusions may be drawn from Table
II: (a) the detection limits of Bhe analyses for low-Z elements, by assay
of both short and long-lived nuclides, are very satisfactory; they range
from parts per million down to parts per billion. (b) He reactions of com-
parable sensitivities, on different elements, lead often to the same radio-
nuclide; thus, interferences may be important in matrices containing two or
more low-Z elements. (c) Only trends of variation for detection limits with
He energy may be inferred from Table II; however, it is clear from Equations
2, 3, and 12, that the C and Ce values are functions of cross section. The
last two conclusions immediately suggest the necessity for excitation curves
for all these reactions.
Neutron Production
Total neutron yields measured with the BF, counter are listed in Table III.
(Table III)
The yields obtained for lithium, beryllium, and boron compare favorably with
those obtained with electrostatic neutron-generators. These generators operate
by the reaction 3a(a, n) *He induced by ~ 150 k.e.v. deuterons on a metallic tar-
get, on which tritium gas has been adsorbed. The generator yields are normally
given for a maximum attainable deuteron current of 500-600 i amp, while the
-11..
data of Table III are calculated for 100 u amp. Moreover, the neutron output
from a tritium target decays to about 1/3 of its original value after one
hour of irradiation (7), due to depletion of tritium on its surface; the
neutron outputs of the reactions on the targets of Table III are obviously
independent of time. It must be pointed out, however, that the he-produced
neutrons are not monoenergetic as are those from neutron generators. Never-
theless, this very characteristic may be of interest to fast-neutron activa-
tion analysts, since it gives them some control (by choosing target) over
the neutron energies. In a preliminary experiment, aluminum, copper, and
carbon monitors were activated with the neutrons and garuma rays produced
"by irradiating targets of lithium hydride, beryllium, and boron with He
ions of 5 and 20 m.e.v. The nuclear reactions are 27A1(n,p)-IME, 63cu(n, 2n)62cu,
ana 12c(7,n)21c. Relative results are given in Table IV. The numbers are:
(Table IV)
court rates, normalized to the lithium results; therefore, comparisons may
be made only within each column. The results for aluminum show the same
trend as those given in Table III; we expect this agreement, since the alu-
minum threshold is low (1..91 m.e.v.). The copper-monitor results show that
lithium produces the largest amount of high energy neutrons (threshold =
21.0 m.e.v.). The threshold for the reaction 12c(n, 2n) 42c, 20.3 m.e.V.,
cannot be reached by neutrons produced by irradiating lithium, beryllium,
or boron with 10 m.e.v. She ions (8). Therefore, the carbon-monitor results
are attributed to the reaction 2c(y,n)--c, which has a threshold of 18.7 m.e.v.
Gamma rays from (He,y) and high-cross-section (3He, cy) reactions may easily
reach this energy, since the Q's for these reactions on lithium, berylliw, and
boron ere generally larger than * 10 m.e.V., reaching the value of + 26.3 m.e.v.
i
.
..
-12-
for "Belhe, y)-c. (The excitation curve for the reaction (y,n) on copper is
less than 2 tenth (9,6) that for the (n, 2n) reaction.) 'The neutron and
zemma-ray yield results raise the interesting possibility of using the He
cyclotron as a source of neutrons and high-energy samma rays for fast-
neutron and photoactivation analyses, or for other purposes.
.
4.-.
'
-13-
REFERENCES
1. Cumming, J. B., in "Applications of Computers to Nuclear and Radio-
chemistry," ed. G. D. O'Kelley, Natl. Acad. Sci. - Natl. Research
Council, nucl. Sci, Ser. NAS-NS 3107, 25 (1963).
Demildt, A. C., Anal. Chem. 35, 1228 (1963).
Dernilat, A. Co, Lawrence Radiation Laboratory Rept. UCRL- 10647 (1963).
Friedlander, G., Kennedy, J. W., and Miller, J. M., "Nuclear and Radio-
chemistry," 2nd. ed., J. Wiley & Sons, Inc., New York, 1964.
5. Feath, R. L., U. S. Atomic Energy Commission Rept. IDO-16880 (1964).
6. Hughes, D. J., and Schwartz, p. B., Brookhaven Nati. Laboratory Rept.
BNL-325 (1958).
7. Lyon, W. S., Jr. (ed.), "Guide to Activation Analysis," D. Van Nostrand,
Co., Inc., Princeton, 1964.
8. Markowitz, S. S., and Mahony, J. D., Anal. Chem. 34, 329 (1962).
9. Montalbetti, R., Katz, I., and Goldemberg, J., Phys. Rev. 91, 659 (1953).
UNCLASSIFIED
ORNL.-DWG. 64.-104
RELATIVE ő
FIG.1
-
0.85.20

40
60
do
100
-
-
-
--
-
.
.
.
-
-
.
.
.
:
M
.-
.
-
.-
entiende
recientes de la cuisine sans
UNCLASSIFIED
ORNL-DWG. 64-40464
R, mg/cm2
14.0 0.75 0.5 0.4. 0.3 0.25
TT -
A/R , cm²/mg

(X
C, p. p. b. .
(X 10-3)
ㅜ ​'
.
|
양 ​'ade dos abs ods olo ole 'ond4
0 30 28 1 18 2
|
.
AR, cm² /mg
R, mg/cm2 ..
FIG. 2.
UNCLASSIFIED
ORNL-DWG. 64-10463

-
6, p.p.b.
!ㅜㅜ
​- 110 (X 10-1)
150
- 140 (X 10-3) 1
TTT
-
187
Lidille
© 20 40 60 80 100
.... .....
.
:
' fia
F16,3
.....
......
Table I. Sensitivities and Detection Limits for activation
Analysis of Trick Targets Containing Oxygen (20 m.e.v. ile sons)"
.
natungunnaranemomme
Range
(R)
Detection
Limit
(c)
P.13.b.
H. e.V
Estimated
Standard
Deviation
Cris
...
Target
zroz:
.............................leg............ apne...........
mg/cm²
0.7
24.3
Sensitivity Per Unit Range Sensitivity
(S/R)
(s)
(m3/
m
dps(p.p.b.)+(11:32/cm) -?dns/y.m.b.
a ps/n.n.bom
0.06;5
0.0472
0.803
19.5 .
2120
....
14
.
Reaction_ _n.e.v.
160(3tie,an)240 -8.3
160(3xe, 0,250 +1.9
160(>1e, 20) 120 -5.3 .
16031ie,p)281 .+2.07
8160 ( 3xe, n) 18ves?87 3.0 )
Zroa
14
12.9
5.12
53.8 :
0.144.
..........
leto
1.75"
2.35
14
................
.. .. ...
... .
Sio,
17.0.
.
1.72
3.4:1
of 5
.* -.
Svalues calculated for an irradiation interval of one hall-life and a
Sple-beam current of 200 u anp.
.
These errors apply to the respective values of S/R, S, and C.
Table II. Detection Limits for Activation
Analysis of Thin Foils Containing Low-Z Elements
Detection
. Limit
Normalized
Detection
Limit
Estimated
Standard
Energy
Interval
m.e.v.
(C)
Reaction
2.9.b.
Deviation
p.p.b.
49.8
† 5%
50.6
87.8
14
4.4
- 3.7
88.2
117
31.2
48.0.
14
$ 5%
20.9
159
14
207
1950
812
14
-
1410
931
--
17
-
-
772
772
AI-
1
+27%
1120
Be(She, n) 220 5.0 - 3.6
20.0 - 9.3
120(3#e, )22c
9.5 - 9.1
120(3 le, pn 3230 9.5 - 9.1
12cf3le, n) 740 4.4 - 3.7
9.5 - 9.1
2410( 388, c1231 2.8 - 0
24( 33e, y)277 2.8 -0
1601378,Q) 250 4.4 - 3.7
9.3 - 8.5
1.60(3xle, pn) 275 9.3 - 8.5
160(38e,p) 285 2 4.4 - 3.7
4260 (3Ele, n) 28:12-289.7 - 9.3
1979( 3ue, on) ?
9.5 - 9.1
197( 3le, aj28 4.1 - 3.2
9.5 - 9.2
dm
A
2120
359
93.1
17
6%
tor
A
323%
**,
. *
172
+36%
149
81,5 (8.8)
151
84.9
28.6 (4.1)
..
1+
+ 5%.
207
68.5
:
$ 5%
+37%
4.1 -
3.2
.
2410
3620
3590
1+
2370
2420
$ 5%
1580
14. 1+
108
72.4
avalues calculated for an irradiation interval of one half-11fe and a 3he-beam current
of 100 4 amp.
" ...mors apply only to the respective values of Cp
For the purpose of comparison the thick target detection limits for 10 m.e.v.
.. particles are given. They have been calculated from Figure 2 for mylar (R = 14.17 mg/cm). .
:-)
Table III. Heutron Yields izom. Sze-Induced Reactions
in 10-2 TCK 12.cets
Neutro: Yielda
sec
Target
10 ..e.v. He Isins
..liüium (LH)
beryllium
5 .e.v. e Ions'
4,10 x 2020
5.59 x 2010
.
2.84 * 2020
.
..
:
:
Boron
· carbon
oxygen (Zroz)
:
3.78. X 1012:.
2.07 x 2021
3.66 x 2020
2.28 x 2010
3.90 x 209
6.49 x 208
100
e calculated for a Bue-beam current of 200 y'amp. The standard
deviations of these results are 1, 3%; they include variations
.:
od duplicate experirrenésoro
te experirents, propagated with long-counter cali-
bration errors.
1
Halle IV. bezu:14s 01.
0itos experients
Tasick
3e-mergy
Relative
Relative Iveutron Yields "Gamma-ray Yields
Above Threshold for "Above Threshold.
Aluminum and Copper For Carbon
Target
m.c.v.
1.itiiium (115) 20
beryllium
Al
cu
1.00 : 1.00
1.00
(1.3** 204, (2.3 x 204) e (5.5 x 102,
1.420 .. 0.77 0.26
0.90 :.0.92 4.10
. 0.18 : : 0.01 :
0.16
0.05 - ..
bozon
beryllium
.
.5.
5
boron
0.05
Amhe estimated standard deviations of these results are < 20%.
Prick targets of carbon and oxygen (2:0,) were also irradiated
viti 10 m.e.v. Dhe ions. The resulting counting rates were
much smaller than those for the targets shown in the Table when
irradiated under similar conditions.
Ernese numbers indicate experimental disintegration rates, in
aps/(u amp Bye), obtained during irradiations on . IlH
C
iframe
that
in the
more time itine
r ari
e
to
END

--
-
-
-
.
.
--
TAA
*
DATE FILMED
5 / 9 /66




.
.-,
.
.
.
,
.