MÖSSBAUER STUDY OF THE TIME DEPENDENCE OF THE FORMATION OF Eu2+ AFTER THE ELECTRON CAPTURE DECAY OF 153Gd In SmCl3

(1)

HAL Id: jpa-00216727

https://hal.archives-ouvertes.fr/jpa-00216727

Submitted on 1 Jan 1976

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

MÖSSBAUER STUDY OF THE TIME DEPENDENCE

OF THE FORMATION OF Eu2+ AFTER THE

ELECTRON CAPTURE DECAY OF 153Gd In SmCl3

U. Wagner, F. Wagner, J. Friedt, D. Mihaila-Tarabasanu, G. Kalvius

To cite this version:

(2)

JOURNAL DE PHYSIQUE ColIoque C6, supplkment au no 12, Tome 37, Dkcembre 1976, page C6-931

MOSSBAUER STUDY

OF

THE

TIMIE DEPENDENCE OF THE FOR~WATION

OF

Eu2+ AFTER THE ELECTRON CAPTURE DECAY

OF 53Gd

In

SnaC1,

U. WAGNER, F. E. WAGNER, J. M. FRIEDT, D. MIHAILA-TARABASANU and G. M. KALVIUS

Physik-Department, Technische Universitat Munchen, D-8046 Garching, Germany

RBsumB.

-

Les conskquences chimiques de la desintegration par capture d'electrons de 153Gd en SmC13 ont ktk Btudiks par des experiences de retrodiffusion Mossbauer avec les rayons gamma de 97 keV et 103 keV de 1 5 3Eu. Pour les deux resonances on observe un pic de 1'Eu2+ ti c6t6 du pic

dominant de l'Eu3+, mais l'intensite de la contribution de l'Euz+ dans le spectre de 97 keV est plus faible que I'on espQe a partir du spectre de 103 keV. Cela indique que 1'Eu2+ est forme dans un temps typiaue de l'ordre du temps de vie moyen du niveau de 97 keV ( 0 7 = 0,3 ns). En appli-

quant la methode des moindres carres

a

une fonction appropriee aux donnees on trouve que le temps de formation de l'Euzt doit 6tre entre les limites 0,2 ns

5

r e

5

1 ns.

Abstract.

-

The chemical consequences of the electron capture decay of 153Gd in anhydrous SmC13 have been studied by Mossbauer backscattering experiments with the 97 keV and 103 keV gamma rays in 153Eu. With both resonances the formation of E u ~ + is observed in addition to the dominant Eu3+ peak, but the intensity of the Eu2+ contribution in the 97 keV Mossbauer pattern is less than expected by comparison with the 103 keV spectrum. This indicates that the E u ~ + is formed within a typical time of the order of magnitude of the lifetime of the 97 keV level

= 0.3 ns). From least squares fits of appropriate lineshapes to the data one finds that the time

for the formation of Eu2+ must be within the limits 0.2 ns

5

r e

5

1 ns.

1. Introduction.

-

As has recently been shown [l, 21,

a substantial amount of Eu2+ is formed in addition to Eu3+ after the lS3Srn decay in nearly all anhydrous rare earth trichlorides. A strikingly exceptional beha-

viour has, however, been found for the EuCI, host. In this environment, the Eu2+ line in the Mossbauer emission pattern of the 103 keV gamma rays of 1 5 3is ~ ~

strongly broadened and much weaker than in NdCl,, SmC1, and GdCI,. This has tentatively been attributed [l, 21 to the instability of the E U ~ + species formed

after the nuclear decay in the EuCI, lattice. The parti- cularly short lifetime of E U ~ + in EuCl, can possibly be

attributed [l, 21 t o a charge relaxation, probably pro-

ceeding via a tunneling mechanism that can occur with Eu3+ ions in the lattice but is barred for other rare earths. From the Mossbauer emission spectra of (153Sm)REC13 sources one concludes [2], that the

typical lifetime of E U ~ + is of the order of several nanoseconds in EuC1, and substantially longer, or even infinite, in the other matrices. However, these data give, little clue about the timescale of the formation of Eu2 after the radioactive decay.

For studies of dynamic changes of the valence state occuring on a nanosecond timescale, the MCissbauer effect in Eu is outstandingly well-suited mainly for two resons. First of all, the widely different lifetimes of the 97 keV and the 103 keV Mossbauer levels in Is3Eu

(z,, = 0.3 ns and zIo3 = 5.6 ns [3]) provide time windows that allow one to study the time evolution of individual species without having to resort t o cumber- some coincidence techniques. In principle the 21 keV resonance of lS1Eu represents an additional, even larger time window (T,, = 14 ns [3]), but in the present work only the two resonances in 15,Eu have been studied. The second, equally important property of the mentioned Mossbauer resonances in Eu is their large change of the mean square nuclear charge radius

The 97 keV and 103 keV resonances in 15,Eu, in particular, have the largest A

<

r2

>

values known for any Mossbauer isotope [4]. Together with the reasonably large electron density difference Ap(0) between E U ~ + and Eu3+, these values of A

<

r2

>

ensure that for the 103 keV and 21 keV transitions the Eu2+ and Eu3 + peaks in the Mossbauer spectrum are

well separated, and that even with the 97 keV resonance the isomer shift between Eu2 C and Eu3 + is about

as large as the natural linewidth WO = 2 hc/E, z.

This resolution is sufficient for an analysis of the spectra in terms of Eu2+ and Eu3+ contributions, despite the short lifetime of the 97 keV level.

(3)

C6-932 U. WAGNER, F. E. WAGNER, J. M. FRIEDT, D. MIHAILA-TARABASANU AND G. M. KALVIUS

A comparison of the relative amount of Eu2 + visible

in the 97 keV and 103 keV spectra is feasible and can give information on the time that elapses between the preceeding nuclear decay and the formation of the Eu2+ species. Since the 97 keV level is populated only very weakly after the

P-

decay of 153Sm, we have performed the experiments with 153Gd as the source activity. In the electron capture decay of this isotope the 103 keV and the 97 keV gamma rays are emitted with comparable intensity. As for the

P-

decay of 153Sm [l, 21, the gamma ray spectrum of '53Gd is well-suited for scattering experiments. Apart from ensuring large resonance effects and good statistical accuracy, our choice to perform the experiments in a backscattering geometry has an additional advantage in the case of the 97 keV resonance. For this E l transition, the interference between photoeffect and internal conversion gives rise to a rather large disper- sion term in the absorption cross-section [S, 61. This results in an inherent asymmetry of the transmission Mossbauer spectra and renders their analysis in terms of a small Eu2+ contribution in the presence of a dominant Eu3+ line somewhat ambiguous. The backscattering spectra, however, are virtually unaffected by such problems [7].

2. Results and discussion.

-

Reactor produced 15,Gd with a specific activity of 14 mCi/mg was pur- chased as a hydrochloric acid solution. SmC1, was chosen as a source matrix because it yields a rather large amount of Eu2+ after the 15,sm decay [l, 21. The source was prepared by adding 100 mCi of 153Gd to a concentrated solution of 150 mg SmC1,. 6 H,O in water. After crystallisation the hydrous chloride was dehydrated by slow heating to 360 OC in a stream of dry HCl and argon. The anhydrous chloride was then transferred into the tight source holder without being exposed to moisture.

The experiments were all carried out at 4.2 K in the backscattering setup described in ref. [4]. The scattered gamma rays were counted with Ge(Li) detectors having a resolution of about 2 keV. The single channel windows were set either to the lower half of the 97 keV peak or to the upper half of the 103 keV peak in order to avoid any contribution from the unwanted resonance in the Mossbauer spectra.

Figures 1 and 2 show the spectra obtained with the 103 keV and 97 keV gamma rays and a scatterer of enriched 1 5 3 E ~ 2 ~ 3 . The 103 keV spectrum was fitted with a superposition of two Lorentzian lines corres- ponding to Eu2+ and Eu3+. The isomer shifts for these lines are S, + =

-

0.94 (2) mm/s and S, + = - 16.8 (2)

mm/s, and the relative area under the divalent line is found to be A,+/(A,+

+

A , +) = 0.22 (2).

Since the ratio of the A

<

r2

>

values for the 97 keV and the 103 keV resonance is known [g], one can use the results obtained for the 103 keV resonance to calculate the E U ~ + and Eu3+ isomer shifts for the 97 keV reso-

nance. Moreover, as long as there are no time-depen-

I I I I I I

- 2 0 -10 0 .l0 + 2 0

VELOCITY ( m r n l s )

FIG. 1.

-

Mossbauer backscattering spectrum of the 103 keV gammarays of 153Euobtained with a source of l53Gd in SmC13

and a scatterer of 1 5 3 E ~ ~ 0 3 .

I I I I I l

-60 - 3 0 0 -30 +60

VELOCITY (mmls)

FIG. 2.

-

Mossbauer backscattering spectrum of the 97 keV gamma rays of 153Eu obtained at 4.2 K with a source of 153Gd

in SmCls and a scatterer of 153Eu203. The dashed line is the best fit of a curve with fixed isomer shifts, linewidths and relative line intensities for the EuZ+ and Eu3+ species, as described in

the text. The fully drawn curve takes the formation of E u ~ +

on a timescale comparable to the lifetime of the 97 keV level into account.

dent effects, the ratio of the areas under the divalent and the trivalent Mossbauer line in the 97 keV spectrum should be identical with the ratio obtained from the 103 keV spectra. This would not be true if the ratio f,

,h,

of the recoilfree fractions of EU'+ and Eu3+ in SmCl, were substantially different for the 97 keV and the 103 keV resonances. A substantial difference in this ratio seems improbable, however, in view of the small difference between the gamma ray energies. The assumption that

f,

+lf, + does not depend

strongly on the gamma ray energy is supported by a comparison of the relative amounts of Eu2

'

and Eu3 +

found in SmCl, with the 103 keV resonance after the 153Sm decay (A,+/(A,+

+

A3+) = 0.109 (6) [l, 21)

(4)

TIME DEPENDENCE OF THE FORMATION OF E u ~ + AFTER THE ELECTRON CAPTURE DECAY C6-933

It would thus seem justified to fit the 97 keV spectrum with A, +/(A,+

+

A, +) and the positions of the Eu2+ and Eu3+ lines restrained to the values derived from the 103 keV measurement. The Iinewidth in such a fit can also be restrained to an independently measured value, since there should be virtually no broadening of the 97 keV resonance due to unresolved hyperfine splittings. A measurement with a source of 15,Gd in Sm203 and the 1 5 3 E ~ , ~ 3 scatterer used in all other experiments yielded a line width of W = 20.0 (2) mm/s. Together with the other restrained parameters, this value was used to obtain the fit shown as a dashed curve in figure 2.

Obviously the 97 keV spectrum differs considerably from the best fit of this expected shape, in the sense that the Eu2+ contribution is too weak and too close to the dominant Eu3+ peak. We interpret this as evi- dence for the formation of the EuZ+ species from a preceeding state of Eu with a typical transition time comparable to the lifetime of the 97 keV level.

To describe the lineshape expected for this time- dependent case, we make use of an expression [I01 based on the superoperator formalism [l l]. Since none of our spectra shows any trace of a Mossbauer emission from E U ~ + species, we must assume that, within

less than a few times 10-" S after the nuclear decay,

Eu3+ has becomes the highest charge state present. Then a certain fraction of the Eu3+ will transform to Eu2+, while the remainder does not change its oxi- dation state and gives a Lorentzian contribution at the Eu3+ position that shows no time-dependence. The fraction undergoing the transformation to Eu2+ must at least be equal to the amount of EuZf seen in the 103 keV spectrum, but it may be larger if some of the Eu2+ decays back to Eu3' via an additional fast reaction mechanism.

With the A, +/(A2 +

+

A, +) vaiue from the 103 keV spectrum used for the fraction that undergoes the transformation, a fit of the formulae given in ref. [l01 results in a considerably better agreement with the experimental curve and yields a typical time of

r e x 0.2 ns for the electronic transformation leading to the formation of Eu2+. The value of ze returned by such a fit depends sensitively on the fraction of Eu3+ that is assumed to undergo the transformation, although the quality of the fit is, within wide limits, quite insensitive to the value assumed for this fraction. The best fit (fully drawn curve in figure 2) is actually obtained if one assumes half of the area under the spectrum to arise from the transforming species. This results in 2, x 0.6 ns. Thus from the present data we can only give a range of 0.2 ns 5 ze

5

1 ns for the decay time with which Eu2+ is formed.

Experiments with a second source prepared in the same way as the first one yielded a slightly smaller ratio of A, +/(A, +

+

A , +) in the 103 keV spectrum

(A, ,/(A, +

+

A, +) = 0.17 (2)), but all other conclu-

sions drawn from these experiments agree with those described above.

3. Conclusions.

-

The data reported above reveal the time-dependence of the formation of Eu2+ but they only allow one to estimate the order of magnitude of the lifetime of the parent states, presumably EU,+ species, from which the EU'+ is formed. The evalua-

tion of the data would have become even more involved if allowance had been made for the possibility that at least part of the Eu2+ transforms back to Eu3 + on a

timescale comparable to that of its formation. In this case the formalism used to describe the spectral shape [l01 would have to be extended to take this second decay into account. It is felt that more data will be needed before such a refined evaluation may become reasonable. Further experiments with other matrices are presently being performed, and the reliability of the prediction of the 97 keV Mossbauer spectra from 103 keV data will be increased by a new determination of the ratio of the A

<

r 2

>

values for the 97 keV and 103 keV transitions.

At present, no detailed interpretation of the mechanism leading to the formation of Eu2 + can be given.

Any such model must account for the fact that the relative amounts of Eu2+ formed after the electron capture decay of 15,Gd and after the

P-

decay of 15,Sm [l, 21 are not radically different. This conclusion is supported by preliminary measurements with the 103 keV gamma rays emitted after the decay of 153Gd in NdCI,. In this case the relative intensity of the divalent Mossbauer line is even slightly smaller than that previously [l, 21 observed after the 15,Srn decay in the same matrix. The mechanism of the EU" for-

mation must therefore be largelyindependent of preceeding electronic shake-up or shake-off processes and Auger cascades. This situation differs from that encountered in oxalates, where a substantially higher amount of Eu2+ has been observed after the electron capture decay of 1 5 1 ~ d than after the

P-

decay of 151Sm 112, 91. This behaviour has been explained 1121 in terms of a reduction involving radicals that are formed radiolytically from the oxalate anions.

In the case of the anhydrous rare earth chorides, the absence of a Mossbauer emission attributable to Eu4+ shows that the decaying atom acquires the 4 f 6 electron configuration of Eu3+ within substantially less than 100 ps. This is indeed expected in view of the high electron affinity of E U ~ + [13]. The problem is to explain why, in about 15

%

of all decays, Eu2+ is formed although Eu3+ is expected to be the stable valence state of Eu in RECI, matrices, and why this formation of EU'+ occurs with a relatively long transition time z, of somewhat less than a nanosecond. To both questions no definite answers can be given at present.

(5)

C6-934 U. WAGNER, F. E. WAGNER, J. M. FRIEDT, D. MIHAILA-TARABASANU AND G. M. KALVIUS

that is possibly associated with a hole in the valence band states. Such a state can only be formed from an Eu3" state (4f6) with a sufficiently high excitation energy. In this model it is conceivable that the fraction of atoms that acquire the 4f7 configuration depends mainly on the fraction of atoms formed with a sufficiently high excitation in the transition from E U ~ '

-

or even higher charge states - to the 4f6 configuration of Eu3+. This fraction will not depend very strongly on the nature of the nuclear decay. The observed time dependence of the formation of the

4f7 configuration can be associated with the lifetimes of the excited Eu3+ states. The Eu2+ (4f7) observed in the 103 keV Mossbauer spectra, however, must be considered as a metastable species that will eventually decay back to E U ~ ' with a 4f6 configuration. This

latter process seems to be of importance at least for the Mossbauer emission spectra in the EuCI, lattice [l, 21.

Acknowledgment. - We wish to thank Dr. H. Mic- klitz and Dr. F. J. Litterst for stimulating discussions.

References

[l] MIHAILA-TARABASANU, D., WAGNER, U., WAGNER, F. E. [6] GOLDWIRE, H. C. and HANNON, J. F., to be published. KA1-VIUs, G. M.s in : Proceedings of ?he Internu- _{[7] WAGNER,}_{F. E.,}_J._{Physique, Colloq.}₃₇_{(1976) C6-673.}

tional Conference on Mossbauer Spectroscopy, Cracow, _STEICHELE,_{E., Z. phys.}₂₀₁₍₁₉₆₇₎_331.

Poland 1975. Eds. A. Z. Hrynkiewicz and J. A. Sawi-

cki, Vol. 1, p. 291. [9] PROBST, Chr., KIENLE, P., LUCHER, K., WAGNER, F. E.

121 MIHAILA-TARABASANU, D., WAGNER, U., WAGNER, F. E. and ZAHN, U,, Radiochim. Acta 18 (1972) 19.

-

and KALVIUS, F . M., to be published. [l01 AFANAS'EV, A. M. and GOROBCHENKO, V. D., Phys. Star.

[3] STEVENS, J. G. and STEVENS, V. E., MiiSSbauer Efect Data Sot. (b) 76 (1976) 465.

Index Covering the 1974 Literature. (IFIIPIenum Data [I 11 AFANAS'EV, A. M. and GOROBCHENKO, V. D., ~ h y s . Stat.

Company, New York) 1975. Sol. (b)

_.

, 73 (1976) 73.

.

[41 KALvrvs9 G. M. and G . K,9 Nuczear Data and _{[l21 GLENTWORTH,}_{P., NICHOLS,}_{A. L., LARGE,}_{N. R. and BUL-}

Nuclear Data Tables 14 (1975) 639. _LOCK,_R._{J., J.}_{Chem. Soc. (Dalton) (1973) 2364.}

[51 HENNMG, W., BAEHRE, G. and KIENLE, P.. Phys. Left.