<<

SCf

UPPSALA UNIVERSITY INSTITUTE OF PHYSICS NUCLEAR SPIN MEASUREMENTS IN SOME NUCLIDES OF THE ALKALI ELEMENTS AND

C. Ekström, S. Ingelman, G. Wannberg and M. Skarestad

UUIP-954 March 1977

ISSN 0042-0263 NUCLEAR SPIN MEASUREMENTS IN SOME NUCLIDES OF THE ALKALI ELEMENTS RUBIDIUM AND FRANCIUM

C. Ekström ) Department of Physics, Chalmers University of Technology and University of Gothenburg, Gothenburg, Sweden. S. Ingelman and G. Wannberg Institute of Physics, University of Uppsala, Uppsala, Sweden. M. Skarestad +) E. P. Division, CERN, Geneva, Switzerland. and The ISOLDE Collaboration, CERN, Geneva, Switzerland.

ABSTRACT

The nuclear spins of several nuclides of the alkali elements rubidium and francium have been measured using atomic-beam magnetic resonance techniques. The following results have been 77 7 7>m 7 5 obtained: Rb I » V2» 'tt> I •= 0, Rb 1 = 4, »Rb I = /2' 208 209 210 211 "•mRb 1=6, Fr 1=7, Fr I = «/2, Fr 1=6, Fr 220 221 5 I = 9/2, 2i2Fr 1 = 5, 2i3Fr I = »/a» Fr 1 = 1, Fr I = /2, 222Fr 1=2. The spin valuss are discussed briefly in terms of current nuclear models.

*) Present address s E. P. Division, CERN, CH-1211 Geneva 23, Switzerland. +) Present address ; Department of Chemistry, University of Oslo, N-1000 Oslo 3, Norway. - 1 -

1. INTRODUCTION

Nuclides of the alkali elements rubidium, cesium and francium are obtained in high yields at the ISOLDE facility as spallaticn products from 600 NeV proton reactions in different targets, followed by on-line suface ionization and separation [1]. In a previous paper [2J, we reported on an on-line oven foil system which efficiently converts an ion-beam of alkalis frorr. the separator to an atomic-beam, and on the measurements of nuclear spins and moments of several cesium using ABMR techniques. The same methods have been used in the present work to measure nuclear spins in some rubidium and francium nuclides. The element francium is here for the first time subject to ABMR measurements.

2. EXPERIMENTAL PROCEDURE AND RESULTS

Referring to our work on cesium [2], we limit our presentation of the experimental techniques to details specific to the present work on rubidium and francium.

For the production of rubidium, targets of or were used. The production yields of different rubidium isotopes at the collector chamber of the isotope separator are shown in Fig. 1. In the same figure are given the corresponding data on francium yields from a target of graphite cloth impregnated with .

The main part of the experiments were performed on-line, i.e. guiding the ion-beam from the isotope separator directly to a heated foil covered with a layer of yttrium. From that surface the activity evaporated continuously in the form of free atoms. Off-line techniques were used in part of the experiments on the more long-lived isotopes.

Appropriate types of detectors were used to measure the activity passing through the atomic-beam apparatus; thin plastic scin- tillators for the detection of &*, surface barrier detectors - 2 -

for a and Ge(Li) detectors for certain Y-lines.

The simple electronic structure of the alkali elements makes them particularly suited for atomic-beam measurements [2] . The electronic splitting factor for the ground state of rubidium has been measured by Bok len et al. [3] to be gy» 2.0023319(14) . The g -factor for francium has not been determined. Spin mea- u surements performed in weak external magnetic fields are, however» not sensitive to small variations in the g -value. j According to the formula relating the nuclear spin to the resonance frequency [2j :

v 9 " J T 21 + 1 we obtain at the strongest field used in the francium measure- ments, vBB/h * 6.185 MHz, a difference of 6 kHz in resonance frequency assuming g = 2.0023 or 2.0033. This difference is smaller than the line width of the resonance signals which is about 30 kHz. In the experiments on francium we have assumed an electronic splitting factor g = 2.0028.

Spin searches in the rubidium and francium nuclides were made according to the equation above. The final results, shown in Table 1, were established by the observation of resonance signals in at least two different external magnetic fields, for rubidium u_B/h = 1.079, 1.370, 2.964 and 4.029 MHz and for francium 4.482 and 6.185 MHz, and by the absence of resonance signals at the frequencies corresponding to other spin values. 1 Spin values in the range /2 - 'V: were checked in the odd-A isotopes and 0 - 7 in the doubly-odd ones. In Figs. 2 - 4, we exhibit decay curves of some resonance signals obtained in rubidium and francium. Of particular interest are those shown in Fig. 3. In an off-line experiment, a sample of 7*Rb was evaporated from the atomic-beam oven. By applying the frequencies corresponding to the spin values I « 0 and I * 4 in exposing two collectors alternatively, it was possible to make an isomer - 3 - separation. The two decay curves correspond to the 17 min ground state and the 6 min isomeric state, respectively. particular attention was paid to the ground state of 7*Rb, in which resonance signals were observed at the 1=0 frequencies. Such signals may also be obtained in the case of a non-zero spin and an extremely small hyperfine structure separation Av. A nucleus with 1=1 would give three resonance signals separated by 2/j«Av. in order to check the possibility of a non-zero spin we made a careful frequency scan in a field of u^B/h = 1.370 MHz, from 700 kHz below to 200 kHz above the spin 1=0 frequency. Only one resonance signal was observed, having a line width of 40 kHz. Assuming this line to be composed of three unresolved lines h?ving widths of 30 kHz t the width of the calibration line at this field, we get a hyperfine structure separation of 8 kHz. This is an upper limit in the case of a non-zero spin, since spin values larger than 1=1 would imply a larger number of resonances and thus a smaller hyperfine structure separation. The corresponding upper limit of the nuclear magnetic moment may be calculated to be 3*10~6n.m. We have not taken into account any broadening of the observed resonance line due to field inhomo- geneities, which would be three times larger in the case 1=0 than for 1=1, thus making the assumption of a non-zero spin in 7eRb even more improbable.

3. DISCUSSION 3.1 The rubidium isotopes The neutron-deficient rubidium isotopes belong to a region where both the protons and the neutrons range between the magic numbers 28 and 50. Nolte et al. [4] have made a systematic study of the first excited 2+ states in doubly-even nuclei belonging to this region. The shell closures at the magic numbers are clearly identified through the high excitation energies. In addition to these shell closures, closed subshells are observed at neutron number N = 38 in nuclei close to Z • 28, and at proton number - 4 -

Z » 40 in nuclei close to N = 50. The effect of the subshell closures is» however» completely lost far front Z = 28 and N = 50 respectively» and maximum deformation is obtained at the centre of the region enclosed by 28 = ( N,Z ) =50. The lightest rubidium 7 isotopes investigated in this work, "''••'•m» »m)# are located close to the centre of this region. Here, rotational bands based 1 on the orbitals [301 /i], [301 V2] and [303 V2]» and perturbed rotational bands associated with the low-spin Nilsson orbitals originating from the g9 . shell model state have been identified in several nuclides, both odd-proton and odd-neutron ones. The different orbitals may be found in the energy level diagram for odd protons shown in Fig. 5. These levels have been calculated within the Nilsson model employing the shell parameters tc = 0.071 and |i - 0.300. The levels occupied by the 37th proton of rubidium at a deformation of about e = 0.2 are drawn in heavy lines. It is 5 79 seen that the measured nuclear spin I = /2 in Rb may be inter- 5 3 77 s preted as [303 V2] or /2[440 V2] and I = /2in Rb as [301 /2] A preliminary result from a magnetic moment measurement in 7sRb V= 3.37(6) n.m.indicates that the ground state should be classi- 5 5 fied by the orbital /2[440 V2] rather than [303 /2] • The theoretical moments of these states are 3 n.m. and 1 n.m., respectively. The nuclear spins of the ground and isomeric states of 78Rb, 1=0 and 1=4, may be obtained by coupling the neutron 5 77 5 orbital /2[.44O Vz] of Kr to the proton orbitals /2[440 V2] and [301 V2] . Further experimental data, particularly on the level structure in 77Rb and 79Rb would be of great interest. A rotational structure would support the assignments proposed here.

The isotope 81>Rb is located relatively close to the N * 50 shell closure. The measured isomeric state spin 1-6 may accordingly be given the shell model assignment 6~( wg, , vd, , ) . It '2

3.2 The francium isotopes

The light francium isotopes studied in this work, 20lFr - 2l3Fr, appear in the corner Z > 82, N < 126, close to the shell closures. The shell model proton state h» , here accounts for several ground states in the Z * 83 and Z * 85 isotopes. - 5 -

2 9 211 213 9 The isotopes ° ' ' Fr, with nuclear ground state spins I = /2, may consequently be associated with the same assignment.

The shell model states available for neutrons just below the N = 126 shell closure are pt , and f5 . . The measured nuclear ground state / 2 / 2 20B 21 spins in the doubly-odd francium isotopes ' o»2i2Fr are readily accounted for by coupling these neutron states to the h9 , proton / 2 state. In 212Pr we have the configuration 5~( who . vsi , ), the /2 /2 spin value being predicted by the coupling rules of ref. [5].

Taking into account the successive filling of the f5. neutron shell, the nuclear spin I = 7 of 20iFr and I = 6 of 209Fr are both explained by the configuration ( nh» , vfs , ) and the /2 /2 coupling rules [5J. One may note that the spin sequence observed in the corresponding doubly-odd isotopes of bismuth [6,7] is somewhat different from that in francium. It would therefore be of interest to make spin measurements in the interjacent astatine in order to see whether trends may be observed. The francium isotopes in the mass range A = 214 - 219 are too short-lived to be studied with the present atomic-beam technique. A possible exception is 219Fr, but here the production yield has so far been too low. The half-lives and yields of the heavier francium isotopes again become favourable. The spin measurements 5 in 22«'22i'222pr nave given the result 1=1, /2 and 2, respectively. These isotopes are located in a transition region between the spherical nuclei close to Z = 82, N = 126 and the strongly deformed actinides. It is thus very difficult to find simple configuration assignments accounting for the measured nuclear spins. The proposed negative parity of the 221Fr ground state [8] implies an association with the h9 , shell model state. Also in the case / 2 of a moderate ruclear deformation, it seems plausible to assume 5 221 that the main component of the I * /2 ground state of Fr comes from the h, . shell model state. The present direct measurement of the nuclear ground state spin of 220Fr, I • 1, confirms the conclusion drawn by Briand et al. [9] - 6 - from spectroscopic arguments. The positive parity [9] of the state suggests a negative parity neutron coupled to the odd proton. Awaiting further experimental data in this region, we refrain from speculating on the configurations giving rise to the measured nuclear spins 1*1 and I « 2 in the doubly-odd francium isotopes a20Fr and 22*Fr.

ACKNOWLEDGEMENTS The authors would like to express their gratitude to Prof. I. Lindgren, Goteborg, and the members of the ISOLDE Collaboration for generous support and helpful discussions. This work has been supported financially by the Swedish Council for Atomic Research. - 7 -

REFERENCES

[lj H.L. Ravn, L.C. Carraz, J. Denimal, E. Kugler, M. Skarestad, S. Sundell and L. Westgaard, Nucl. Instr. 139 (1976) 267. [2] C. Ekström» S. Ingelman, G. Wannberg and M. Skarestad, to be published. [3] K.D. Boklen, w. Dankwort, E. Pitz and S. Penselin, Z. Physik 200 (1967) 467. [4] E. Nolte, Y. Shida, W. Kutschera, R. Prestele and H. Morinaga, Z. Physik 26JB (1974) 267. [5] M.H. Brennan and A.M. Bernstein, Phys. Rev. 120 (1960) 927. [6] CM. Johansson and I.P.K. Lindgren, Nucl. Phys. 9_ (1958/9) 44, [7] S. Axensten, CM. Johansson and I. Lindgren, Ark. Fys. 15 (1959) 463. [8] Nuclear Data Sheets 10 (1973) 625. [9] J.P. Briand, P. Chevallier and A. Touati, J. de Physique 32 (1971) 101. - 8 -

Table 1. Summary of results from the spin measurements in rubidium and francium nuclides.

Measured Isotope Half-life spin I

77Rb 3.8 min V, 7lRb 17.7 min 0 a)

78mRb 6.0 min 4

7 »Rb 23.0 min V2 8l>mRb 21 min 6

2 0 8j.r 59 sec 7

2 0»pr 54 sec V,

2J0pr 3.18 min 6

2 1lpr 3.08 min V,

2J2pr 19.3 min 5

2 13Fr 34.7 sec V, 220Fr 27.5 sec 1

2 2 1pr 4.8 min V,

2 2 2Fr 14.8 min 2

a) The resonances observed in 7lRb may also be due

to a non-zero spin in case the magnetic moment

is smaller than 3*10~* n.m. - 9 -

FIGURE CAPTIONS

Fig. 1 : Experimental yields of rubidium and francium nuclides from a 50 g/cm2 target of niobium powder and a target of graphite cloth impregnated with 25 g uranium, respectively. The proton beam intensity was in both cases 1 pA. Dots indicate measurements using a Faraday cup. Squares indicate in the case of rubidium beta- counting and in francium alpha-counting. Our spin measurements are marked by crosses at the bottom of the figure, and previously known spins by circles. Note that the present results on francium represent the first ones from direct spin measurements in this element.

Fig. 2 : Decay of 77Rb sample obtained at the frequency corre- sponding to the ( 2, -1 ) «—f { 2, -2 ) transition for 3 Rb I = /2 in weak external field. The spin I = V? is further confirmed by the fact, that no other resonances were induced by applying frequencies which correspond to spin values in the range V2 ~ li/z*

Fig. 3 : Isomer separated samples of 7*Rb. A half-life of 17 min was found in the decay of the sample exposed at the frequency according to I = 0, while that corresponding to I • 4 yielded the sample of half-life 6 min. This attributes spin I « 0 to the ground state and spin 1=4 to the 6 min isomeric state of 78Rb,

Fig. 4 : The o-decays of samples of 210Fr and211Fr being exposed

at frequencies corresponding to spins 1 = 6 and I » V2/ respectively. - 10 -

Fig. 5 : Energy level diagram for odd protons, relevant to the rubidium region. It is based on calculations performed within the Nilsson model using the shell parameters K • 0.071 and p » 0.300. Full lines represent levels with positive parity, dashed lines negative parity. The portion of the levels around e * 0.2, available to the 37th proton of rubidium is drawn in heavy lines. A similar diagram is valid for the odd neutrons in this region. T '

SATURATION ACTIVITY (OPS) *o3S3*35d3 o <*•»«# • id • -4 m m

• •

oo oo o g: V o o o 0 o O s I 4 I I I to (fl i§ m 31

• - • •

M - 12 -

III I

500 m

77Rb 1« 3/2

200 -

>

100 •

50

T ( '\.

3.8 min i- 1.24 b N

20 m

1 t ft i iiiiiiiiii 10 min

Fig. 2 - 13 -

20 -

10 •

50 min

Fig. 3 INTENSITY, ARB. UNITS

i» o - 15 -

E f»i*>o 404 9/2

413 7/2

t 5.1 »9/2 431 1/2

301 1/2 5.0 303 5/2 422 5/2

4.9 . 301 3/2

4.8 431 3/2

4.7

312 3/2 4.6 440 1/2 _ P3/2 310 1/2 303 7/2 4.5

321 1/2 PROTONS

010 0.20 0.30

Fig. 5 ADDENDUM to

NUCLEAR SPIN MEASUREMENTS IN SOME NUCLIDES OP THE ALKALI ELEMENTS RUBIDIUM AND FRANCIUM

In the paper above we gave a preliminary result on the magnetic moment of 7SRb, y = 3.37(6). This value has since been confirmed by further experiments. From spin measurements in the ground and isomeric states of'7iRb, having spin 1 = 0 and I > 4, we assigned the neutron orbital Vs [440 V2] and the proton orbitals 5/2[440 Vi] and [301 i/i\, respectively, to these states. As the ground state has spin 1=0, hyperfine structure measurements were undertaken in the 1=4 isomeric state of 78Rb only. The results, based on three experiments performed in external magnetic fields of strengths up to 11 mT, are:

Hfs Dipole Magnetic Nuclide 71 separation constant moment AV MHz a MHz ]i- n.m. H .14 5378(25) 1195(6) 2.56(3) 78 85 +) A hyperfine anomaly A • ± 1 % is assumed.

Calculating the magnetic moment from the neutron orbital V2C440 V*] and the proton orbital [301 i/i] , one obtains the value 2.6 n.m., in agreement with the measured value.