Nuclear excitation by electron transition in 189Os following K-shell ionization by bremsstrahlung

J. inorg nud (.'hem. Vol. 43, pp. 1%3-1966, 1981

0022-1902/81/091963~4502.00/0 Copyright © 1981 Pergamon Press Ltd.

Printed in Great Britain All rights reserved

NUCLEAR EXCITATION BY ELECTRON TRANSITION IN

18905

FOLLOWING K-SHELL IONIZATION BY BREMSSTRAHLUNG TADASHI SAITO*, ATSUSHI SHINOHARA, TAICHI MIURAt and KIYOTERU OTOZAI Department of Chemistry, Faculty of Science, Osaka University, Toyonaka, Osaka 560, Japan

(Received 11 September 1980; receivedfor publication 23 January 1981) Abstract--Nuclear excitation by electron transition (NEET) was studied on '89Os by photoionization of the atomic K shell. The radioactivity of the 6 hr isomer observed after irradiation with bremsstrahlung produced by 200 keV electrons is ascribed to the excitation of the 70 keV level induced by the KX (X = M4, Ms, MI, etc.) electronic transitions through the electromagnetic interaction between the nucleus and orbital electrons. INTRODUCTION

Nuclear excitation by electron transition (NEET) was considered by Morita[1] as a possible mechanism responsible for the deexcitation of vacancies in atomic inner shells, in addition to the two normal modes, X-ray and Auger-electron emission. This should be realized when the nuclear and electronic transitions have nearly equal energies and the same multipolarity. This situation is certainly the case for tSgOs, since, in that atom, the nuclear excitation of the 69.52 keV level proceeds by E2 as does the electronic transition between the K and M4 shells with an energy of 71.804keV. Experimental evidence for the existence of such a process was first obtained by Otozai et a/.[2, 3] who could detect the 6 hr isomer activated through the 70 keV level by bombardment of 189Os with electrons below 100keV. In their experiments, electrons were used to ionize the K shells of '89Os atoms, and hence Coulomb excitation by electron projectiles had to be taken into account as an inevitably competing mode for isomer production. The contribution of this mode was estimated to be negligibly small compared with that of NEET. On the other hand, NEET can be observed without the interference of Coulomb excitation in those experiments in which photons are used for K-shell ionization. In this study, bremsstrahlung produced by electrons were used for the irradiation of tS9Os on account of the ready availability of sutficiently high intensity. EXPERIMENTAL A metallic Os layer was prepared by an electroplating technique[4] for use as a target. Powder of osmium enriched to 95.66% in ~9Os was supplied by ORNL. Osmium dissolved in sodium hypochlorite solution was added to an electrolytic bath composed of (NH4)~HPO4 and Na2HPO4 and then plated by 0.755 mg/cm2 in thickness and 20 mm in diameter onto a 0.5 mm thick Pt disk. The target was placed at a distance of 3 mm from a Pt photon-producing converter which was bombarded with 200 keV electrons from a Cockcroft-Walton electron accelerator at the Radiation Center of Osaka Prefecture. The thickness of the converter was 106 mg/cm2, which is thicker than the maximum range of 200 keV electrons in Pt, 89 mg/cm2[5]. The irradiation of tSgos with bremsstrahlung was made under vacuum for 5 hr with a beam current of 0.2 mA. The current incident upon the converter was measured with ammeters connected to the mount of the converter. The loss in current due to backscattering

of electrons from the converter was estimated to be 48.7% by an empirical equation[6], and the corrected current was consistent with the output of the accelerator. The radioactivity induced in the target was measured with a 2zr windowless Q-gas-flow GM counter, surrounded by peripheral anticounters, with background of below 0.02 counts/s. The detection coefficient of this counter was estimated to be 0.45 for tS9mOS. The obtained decay curve, as shown in Fig. 1, was fitted to the ~89mOscomponent (tl/: = 6.0 h) and a constant background one by a least-squares method. RESULTS AND DISCUSSION

The disintegration rate of Jsg"os induced in the target was determined as Ao=0.107--0.005Bq [disintegrations/s] at the end of irradiation. The quoted uncertainty includes only the one in the fitting procedure. The NEET probability, P, is defined in this case as the probability that the 70keV nuclear level is excited per disappearance of a K hole created in the t89Os atom. The value of P can be calculated from Ao as

P= Ao/[ £~maxeK(E)Nv(E) dE nB(I - e-A')] , (I) where o,~<(E) is the K-shell-ionization cross section for Os by photons of energy E, Nv(E) the incident rate of photons with energy E, n the number of target atoms per unit area, B the branching ratio of the 70 keV level to the isomeric level lying at 30 keV, h the decay constant of the isomer, and t the irradiation time. The limits of integration of E are from EK, the K-electron binding

o 41/2 (1~

01

,,

0

*Author to whom correspondence should be addressed. *Now at Department of Chemistry, Tokyo Metropolitan University, Setagaya-ku, Tokyo 158. Japan. JINC Vol 47. No. O-a

1963

] ....

I [ . . . . . . .

10

I .... 20

~,

,,%

3o

Time(h)

Fig. 1, Decay curve of the radioactivity induced in the target. The solid lines show the results of a least-squares fit.

1964

TADASHI SAITO et al.

energy (73.871keV for Os), to E .... the maximum energy of incident photons corresponding to the energy of primary electrons (200 keV in this case). The calculation of Nv(E) was executed on an ACOS computer at Osaka University by using the DIBREDETECTOR code developed by Nakamura et al.[7]. This code gives the spectral yield of bremsstrahlung radiated from a thick converter at a given direction with respect to the primary electron beam. In the calculation, the Pt converter was treated as a stack of slabs with a thickness of 10-3 radiation-length units (6 mg/cm2), and the degradation and scattering of incident electrons and the attenuation and buildup of bremsstrahlung passing through the converter were taken into account. The differential yield was integrated over the geometric factor between the converter and the target. For simplicity of integration, the extended beam was treated as the narrowly centralized beam and the geometric factor was replaced by the solid angle that was subtended by the target at the center of the converter, equivalent to the average one subtended at the plane of the converter. The average solid angle was calculated by a Monte Carlo method[8], provided that the converter 20ram in diameter was uniformly impinged by the electron beam. Figure 2 shows the calculated spectrum of photons incident upon the target per primary electron, N~(E)= N~(E)/Ne, Ne being the primary beam intensity. The integral in the denominator in eqn (1) was calculated numerically as

~6z

lo-2o

E ~o-21

L

I

163o

I

I

50

,

L

I tOO

,

J

m

,

I

J

*

,

~

150 ET (keV)

Fig. 2. Spectral distribution of photons incident upon the ]89Os target per primary electron, and K-shell ionizationcross sections for Os by photons. The points show the photon yields calculated by the DIBRE-DETECTOR code, and the solid squares indicate those at energies correspondingto the nuclear levels in ~89Os. In this experiment, resonance absorption of photons by the nucleus remains competing with NEET for isomer production, although Coulomb excitation by electron projectiles can be excluded inherently owing to K-shell ionization by bremsstrahlung. The radioactivity of the isomer produced through this resonant process at the end of irradiation is given by integration of the Breit-Wigner single level formula as Ao = zr2X2gnN~,(E,)(1 - e X*)For,~,,/F,

f•max

o-K(E)N.,(E) dE = Ne f72OO crK(E)N~,(E) dE, K 3.87

(2)

where E is in keV. The values of CrK were taken from Ref. [9], and the smoothed curve connecting them for interpolation is depicted also in Fig. 2 by the dashed line. The integral was thus evaluated to be 1.35x 10 -24 N, cm2/s for the conditions of this experiment. As a result, the NEET probability is obtained as P = 4.3 × 10-8 from eqn (I) with B = 1.2 × 10-3 as given in Ref. [3]. This value is smaller by a factor of four than that determined precisely in the previous experiment[3] using the electron projectiles for direct ionization of the K-shells. This disagreement is probably due to uncertainties in the irradiation procedure in this experiment. Besides, the DIBRE code may overestimate the spectral intensity of bremsstrahlung in this case of 200-keV primaries and the close configuration of converter and target, although the calculated spectra in the collimated forward direction are declared to agree well with the experimental ones in the range 0.5-1.44 MeV [7].

E

F0

Fi8 °

F

keY

eV

eV

eV

=

5/2

6.6

× 10 - 2 6

36.17

1/2

3.9

x 10 -8

69.52

3/2

3.1

x 10 -8

95.23

1

6.0

× 10 -8

4.4

Estimated

in

Ref.

[3].

F

~ 10 - 2 0 3.4

a)

(3)

where X is the wavelength of photons corresponding to the resonant energy Er divided by 2zr, g is the statistical factor, Fo, F~,, and F are the radiative width to the ground state, the partial width to the isomeric level and the total width of the resonant level, respectively, and others are the same for eqn (1). Numerical values for the nuclear parameters required for eqn (3) are listed in Table 1 for all levels below 200 keV. Those were evaluated from a compilation by Lewis[10] and the.tables of internal conversion coefficients by Rfsel et al.[ll]. Values for F~so were estimated in Ref. [3] for the 70 keV level, or by the Weisskopf formula for those levels at 36 and 95 keV with factors of 100 for E4 enhancement and M3 hindrance, respectively. However, the indirect feeding to the isomeric state via the 70 keV level is found to account for F~s,, of the 95 keV level compared with the direct feeding. As is evident from eqn (3) and Table 1, isomer production by 3,-resonance absorption is predominantly due to the 70 and 95 keV levels. The values of N~, ( = NJN~) at each resonant energy were calculated by the DIBRE code as shown in Fig. 2 by the solid

Table I. Numerical values for the nuclear parameters in eqn (3)

30.80

10-22 200

x i0 -I0 x i0 - I 0

a)

2.1×10

-20

8.6xi0

-7

2.8×10

-7

2.0x10

-6

NEET in '89OsfollowingK-shellphotoionization squares. The contribution of this competing mechanism is estimated to be smaller by two orders of magnitude than that observed in the experiment. The reason why 200keV was chosen as the primary beam energy is to avoid explicitly the appreciable contribution of the 219 keV level which has much larger Fifo than the 70 and 95 keV levels, though electrons with higher energies are more efficient in producing bremsstrahlung. Consequently, the observed isomer can be wholly ascribed to the excitation of the 70 keV level by NEET following K-shell photoionization. Hereafter, a more reliable value P = 1.7× 10 7 obtained in the previous experiment[3] is used for further discussion. In addition to the KM4 electronic transition which has been exclusively considered in Ref. [3], the KMs, KN4 and KN5 transitions should participate in the NEET process in '89Os as the atomic counterparts, since these transitions and their nuclear counterpart also have nearly equal energies and the common multipolarity E2. If it is permitted to neglect the mutual interference of these atomic counterparts, the NEET probability for each combination can be simply related to the Coulomb interaction energy between the nucleus and orbital electron, E', as

1965

principal quantum numbers of the initial and final electronic states, respectively, and f is the correction factor due to the collective character of nuclear transitions. Numerical values of the parameters in eqn (4) and the calculated P's are listed in Table 2. Evidently the total NEET probability becomes about twice as large as that estimated only for KM4. In recent years, it has been pointed out that the magnetic interactions between the nucleus and orbital electrons should make a non-vanishing contribution to NEET in addition to the Coulomb interaction, naturally under the circumstances in which the NEET conditions are satisfied. Compared with the KM4 transition, a smaller A-value of 1.30 keV is attained by considering the KMI electronic transition as a substitute of the atomic counterpart. This transition had been disregarded because of its multipolarity of M1, although the 70 keV nuclear transition is an admixture of M1 and E2. The Ml-type interaction of such a combination should, however, be taken into account, since both transitions also satisfy the NEET conditions. Restricting the ns,2--, lsln electronic transitions in atoms with the atomic number Z, Morita[15] has estimated the energy of this M1 interaction between the nucleus and orbital electron as

P = (1 + F21F~)(E'/A) 2,

(4) Ed41 = - 8~BI~N (Z/ao)3/n 3/2,

where F, and F2 are the total widths of the initial and final atomic levels participating in NEET, respectively, and A is the energy difference between the atomic and nuclear transitions, .%~- EA - E~ = (E, - E2)- EN. For Os, the value of F, was estimated by the semiempirical K-width expression of Leisi et al.[12] as F I = F ( K ) = 42.6 eV, and those of F2 were taken from the calculated ones by McGuire[13, 14]. A crude estimate of E' for El transitions is expressed by eqn (6) in Ref. [3] for given A, Z. l, ~ [=-(n,+n2)/2] and f. where nm and n2 ire the

where tzB and ~ are the Bohr and nuclear magnetons, respectively, and ao is the Bohr radius. For KMt and other candidates, KN, and KO,, values for Eg,, were estimated by eqn (5). On the contrary, we cannot estimate Egu for KM4 and KN4 which may also contribute by MI. The width of the Oi subshell was extrapolated from those of 58 multiplets in the rare earths calculated by McGuire[16]. For these M1 transitions, P'8 are also calculated by eqn (4) as listed in Table 3. After 8ubtrac-

Table 2. Numerical values for the parameters in eqn (4) and the calculated NEET probability P for the E2 transitions T r a n s i t i o n a)

KM 4

a)

and b)

~/keV

F2/eV

E ' / e V b)

2.32

4.18

-0.60

pb)

7.35

×

10 -8

KM 5

2.39

2.35

-0.60

6.63

× 10 -8

KN 4

4.0~

8.38

-0.16

1.85

× 10 -9

KN 5

4.08

8.13

-0.16

1.82

× 10 -9

For

all F1 =

Provided

transitions, F(K)

E 1 = 73.871

= 42.6

eV.

/ = 1

(E'

that

~ f,

keV,

f N = 69.52

keV,

P ~ f2).

Table 3. Numerical values for the parameters in eqn (4) and the calculated NEET probability P for the MI transitions Transition

115)

A/keY

F2/eV

E'/eV

P

KM 1

1.30

20.4

-0.13

1.48

× 10 -8

KN 1

3.70

15.7

-0.09

7.57

× I0 -I0

KO 1

4.27

3

-0.06

2.26

× 10 -10

1966

TADASHI SAITO et al.

ting the contribution of MI which amounts to about 11% of that of E2, we get f = 1.0 from the experimental results obtained in Ref. [3]. As seen from Tables 2 and 3, the KM4 and KM5 transitions contribute predominantly to NEET in 189Os, and the KM, transition is next to them. For the resultant atomic systems including the dominant KM4 and KM5 transitions, the mixing angles 0 given by eqn (5) in Ref. [3] are then determined to be tan 0 = - 2 . 6 × 10 -4 and - 2 . 5 × 10 4, respectively. The energy shifts ~ given by eqn (7) in Ref. [3] are 1.6 × 10 4 and 1.5 x 10-~ eV respectively for these KM4 and KM5 transitions. It is natural that these results are about a half of those obtained in Ref. [3], since the interference between the participants in NEET is not taken into account. For other transitions, 0 and ~ are about an order of magnitude smaller than those for two dominant transitions. Finally, it can be concluded that NEET in ~89Os was observed in this experiment using photons for K-shell ionization, with exclusion of Coulomb excitation by electrons and with superiority over 3,-resonance absorption. The results obtained may offer additional verification of NEET the first stage of which is a purely atomic process. Acknowledgements--The authors are greatly indebted to Drs. S. Okabe, M. Kitagawa, Y. Shono and T. Oka of the Radiation Center of Osaka Prefecture for their making the accelerator

available for this study. They are grateful to Prof. T. Nakamura of Institute for Nuclear Study, University of Tokyo, and Dr. K. Shin of Department of Nuclear Engineering, Kyoto University, for offering the DIBRE code. They wish to appreciate the helpful suggestions by Prof. M. Morita of Department of Physics, Osaka University. REFERENCES

1. M. Morita, Prog. Theor. Phys. 49, 1574(1973). 2. K. Otozai, R. Arakawa and M. Morita, Prog Theor. Phys. 50, 1771 (1973). 3. K. Otozai, R. Arakawa dnd T. Saito, NucL Phys. A297, 97 (1978). 4. R. Arakawa, T. Saito and K. Otozai, Nucl. lnstrum. Methods 131, 369 (1975). 5. L. Pages, E. Bertel, H. Joffre and L. Sklavenitis, At. Data 4, 1 (1972). 6. T. Tabata, R. Ito and S. Okabe, Nucl. lnstrum. Methods 94, 509 (1971). 7. T. Nakamura, M. Takemura, H. Hirayama and T. Hyodo, 3.. Appl. Phys. 43, 5189 (1972). 8. L. Wielopolski,Nucl. Instrum. Methods 143, 577 (1977). 9. WM. J. Veigele, At. Data Tables 5, 51 (1973). 10. M. B. Lewis, Nucl. Data Sheets 12, 397 (1974). 11. F. R6sel, H. M. Fries, K. Alder and H. C. Pauli, At. Data Nucl. Data Tables 21, 291 (1978). 12. H. J. Leisi, J. H. Brunner, C. F. Perdrisat and P. Scherrer, Heir. Phys. Acta 34, 161 (1961). 13. E. J. McGuire, Phys. Rev. A 5, 1043 (1972). 14. E. J. McGuire, Phys. Rev. A 9, 1840(1974). 15. M. Morita, private communication. 16. E. J. McGuire, Phys. Rev. A 10, 13 (1974).