Nuclear structure of 189 Tl states studied via/EC decay and laser spectroscopy of 189m+ g Pb

Abstract

The β + /EC decay of 189m,g Pb has been studied at the ISOLDE facility using nuclear spec-troscopy and in-source laser spectroscopy. A level scheme of 189 Tl has been built from γ-γ coincidence relationships and information on the feeding of some excited levels of 189 Tl provided by the hyperfine spectra obtained from laser ionization. The half-lives of both the 13/2 + and 3/2 − 189 Pb isomers have been estimated to be T 1/2 = 50 ± 3 s and T 1/2 = 39 ± 8 s, respectively. Calculations have been performed for different oblate and prolate nuclear deformations using an axial-rotor coupled to one-quasiparticle model, a structure has been suggested for the low-lying levels of the 189 Tl nucleus.

Full text

Page 1

DOI 10.1140/epja/i2008-10693-3
Regular Article – Experimental Physics
Eur. Phys. J. A 39, 33–48 (2009)
THE EUROPEAN
PHYSICAL JOURNAL A
Nuclear structure of189Tl states studied via β+/EC decay and
laser spectroscopy of189m+gPb
J. Sauvage1,a, J. Genevey2, B. Roussi` ere1, S. Franchoo1,3,4, A.N. Andreyev5,6,7, N. Barr´ e1, J.-F. Clavelin1,
H. De Witte5, D.V. Fedorov8, V.N. Fedoseyev4, L.M. Fraile4,b, X. Grave1, G. Huber3, M. Huyse5, H.B. Jeppesen4,c,
U. K¨ oster4,9, P. Kunz3, S.R. Lesher5,d, B.A. Marsh4, I. Mukha5,e, J. Oms1, M. Seliverstov3,8, I. Stefanescu5,f,
K. Van de Vel5,g, J. Van de Walle4, P. Van Duppen5, and Yu.M. Volkov8
1Institut de Physique Nucl´ eaire, IN2P3-CNRS/Universit´ e Paris-Sud, F-91406 Orsay Cedex, France
2Laboratoire de Physique Subatomique et de Cosmologie, IN2P3-CNRS/Universit´ e Joseph Fourier, F-38026 Grenoble Cedex,
France
3Institut f¨ ur Physik, Johannes Gutenberg Universit¨ at, D-55099 Mainz, Germany
4ISOLDE, CERN, CH-1211 Gen` eve 23, Switzerland
5Instituut voor Kern- en Stralingsfysica, K.U. Leuven, B-3001 Leuven, Belgium
6Oliver Lodge Laboratory, University of Liverpool, Liverpool, L69 7ZE, UK
7TRIUMF, Vancouver BC, V6T 2A3, Canada
8Petersburg Nuclear Physics Institute, 188350, Gatchina, Russia
9Institut Laue-Langevin, 38042 Grenoble cedex 9, France
Received: 28 August 2008 / Revised: 13 November 2008
Published online: 14 January 2009 – c ? Societ` a Italiana di Fisica / Springer-Verlag 2009
Communicated by J.¨Ayst¨ o
Abstract. The β+/EC decay of
troscopy and in-source laser spectroscopy. A level scheme of
relationships and information on the feeding of some excited levels of
spectra obtained from laser ionization. The half-lives of both the 13/2+and 3/2− 189Pb isomers have been
estimated to be T1/2= 50 ± 3s and T1/2= 39 ± 8s, respectively. Calculations have been performed for
different oblate and prolate nuclear deformations using an axial-rotor coupled to one-quasiparticle model,
a structure has been suggested for the low-lying levels of the189Tl nucleus.
189m,gPb has been studied at the ISOLDE facility using nuclear spec-
189Tl has been built from γ-γ coincidence
189Tl provided by the hyperfine
PACS. 21.10.Hw Spin, parity, and isobaric spin – 23.20.Lv γ transitions and level energies – 42.62.Fi
Laser spectroscopy – 21.60.Ev Collective models
1 Introduction
Large differences in nuclear shape for the isomeric and
ground states of the neutron-deficient isotopes have been
established from the change in the mean square charge ra-
dius deduced from isotope shift measurements along the
thallium, mercury, gold, platinum and iridium isotopic
chains [1]. For the even-Z mercury and platinum nuclei
a sudden increase in the nuclear deformation appears at
ae-mail: sauvage@ipno.in2p3.fr
bPresent address: Universidad Complutense, Madrid, Spain.
cPresent address: LBNL, Berkeley, CA, USA.
dPresent address: Lawrence Livermore National Laboratory,
Livermore, USA.
ePresent address: Universidad de Sevilla, Sevilla, Spain.
fPresent address: ANL, Argonne, IL, USA.
gPresent address: VITO, IMS, Mol, Belgium.
A = 185 and the odd-A nuclei also appear more deformed
than even-even nuclei in the A < 186 region [2–4]. It has
also been shown that the nuclear deformation of the odd-A
platinum nuclei, with A < 186, depends upon the state oc-
cupied by the single neutron [4,5]. For the odd-Z nuclei,
the nuclear deformation depends on the state occupied
by the single proton resulting in different deformations
in the thallium nuclei with A < 194, a sudden increase
of the deformation for A < 187 in the gold and iridium
nuclei and a shape coexistence in the191,193Tl and186Ir
isotopes [6–11].
Nuclear shape coexistence has been found in nuclei
close to the neutron-deficient lead nuclei from nuclear
spectroscopy measurements performed using radioactive
decay and in-beam experiments [12,13]. For example, de-
coupled rotational bands built on states arising from the
h9/2 sub-shell have moments of inertia indicating well-
deformed nuclear states in the odd-A gold nuclei with

Page 2

34The European Physical Journal A
A < 190. This band is built on a 9/2−state located at
325keV in189Au [14,15] and on the well-deformed 5/2−
ground state in185Au [16,17]. In the even-A mercury nu-
clei with A < 190, rotational bands are built on excited 0+
states and correspond to a well-deformed nuclear shape.
The energy of the 0+deformed state decreases when N
lowers. However, the 0+deformed state remains an ex-
cited state and never becomes the ground state even at
the neutron mid-shell N = 104 [18–21] whereas the well-
deformed nuclear state 1/2−[521] is the ground state of
the odd-A185Hg nucleus. The properties observed in the
lead nuclei appear to be similar to those known in the
mercury nuclei: i) 0+states have been observed at low
energy in186,188Pb nuclei [22], ii) quasi-rotational bands,
built on 21+states, have moments of inertia indicating
the existence of a deformed nuclear shape for the excited
states in the even-A182–190Pb isotopes [23–26] and, iii)
the energy of the 21+state varies slowly against the mass,
its value is the smallest (E = 662keV) for A = 186 but
remains higher than that of the 02+and 03+states. The
186,188Pb results were recently confirmed in more com-
plete studies [27,28]. Furthermore, the first excited band
of186,188Pb has been determined to have a deformation
of β = 0.29 by lifetime measurements [29]. The question
remains whether a state corresponding to a well-deformed
nucleus could become the isomeric or ground state in the
odd-A lead isotopes close to the neutron mid-shell, in spite
of their magic proton number Z = 82.
No shape transition or coexistence has been observed
for the isomeric or ground states of the190–214Pb nuclei
from the isotope shift values measured using the atomic
beam or collinear spectroscopy [30,31]. Recently, the de-
formation investigation was extended to the182–190Pb nu-
clei using a highly sensitive in-source laser spectroscopy
technique [32]. The hyperfine spectra are obtained by
counting the radiation emitted by the nuclei of the photo-
ionized atoms as a function of the laser frequency. The
182–188Pb nuclei have short half-lives (T1/2< 25s) and
their α branching ratios (b > 5%) are high enough to be
used for the hyperfine spectrum recordings. On the other
hand, the189–190Pb nuclei have longer half-lives and small
α branching ratios (b < 1%), which made it necessary to
use the γ-emission following β+/EC decay to observe the
hyperfine spectra. As a prerequisite, γ-ray identification
has been performed since very little was known on the
β+/EC decay of189Pb [33]. In the course of this first ex-
periment, qualitative hyperfine spectra and some γ-γ-t co-
incidence events were recorded for the γ-rays emitted from
the189Pb decay. The results of this first experiment sug-
gested that a low-spin isomer exists in189Pb, however, the
observed coincidence relationships were incompatible with
the level scheme of189Tl reported in ref. [34]. A second
experiment has been performed in order to confirm the
existence of a low-spin isomer in189Pb, to determine un-
ambiguously the origin of the observed γ-rays, to establish
a low-lying level scheme of the189Tl nucleus and to obtain
more precise hyperfine spectra of the189Pb nucleus.
In this paper we present the results on the β+/EC
decay of189Pb (189Pb →189Tl) obtained during the two
experiments. The isotope shift and magnetic moment
results [35] and the results on the α decay of
(189Pb →185Hg) will be reported elsewhere. After a de-
scription of the experimental procedures in sect. 2, the ob-
tained data which allowed us to identify for the first time
the radiation emitted either from the189gPb or from the
189mPb isomer, as well as a level scheme of189Tl, will be
presented and discussed in sect. 3. Finally, the low-lying
levels of189Tl will be tentatively identified with the help
of the semi-microscopic axial-rotor coupled to one quasi-
proton model developed in the context of the Hartree-Fock
plus BCS approximation [36].
189Pb
2 Experimental procedures
Two experiments were carried out at the ISOLDE facil-
ity at CERN [37]. Radioactive lead atoms were produced
via spallation reactions by bombarding a thick uranium-
carbide, UCx, target with the 1.4GeV pulsed proton beam
delivered by the PS-Booster. The radioactive atoms re-
leased from the target effuse into the hot cavity of the
Resonance Ionization Laser Ion Source (RILIS) [38,39].
The lead atoms are then selectively ionized in a three-step
resonant ionization process. The frequency scan is per-
formed on the laser beam that produces after frequency
doubling the first excitation step (λ = 283.3nm). A laser
beam of λ = 600.2nm is used to get the second excitation
step and the ionization step is provided by laser beams of
λ = 511 and 578nm (for more details, see refs. [32,35]).
The thallium atoms are easily ionized, therefore, in spite
of the selectivity of the RILIS method the lead ion beam is
contaminated by the thallium atoms that are surface ion-
ized in the hot cavity. The produced ions are extracted by
60keV, mass-separated by the General Purpose Separator,
GPS, guided towards the counting station, collected on a
tape and then the obtained radioactive source is moved in
front of the detectors.
In the first experiment the setup consisted of a 4π-β
plastic scintillator detector and three γ detectors; one
planar Ge(HP) X-ray detector with a 0.55keV FWHM
resolution at 80keV and 0.9keV at 356keV, two coax-
ial Ge(HP) detectors with about 70% relative efficiency
and 2.5keV, 2.9keV FWHM resolution at 1.33MeV. The
ISOLDE tape transport system was used to move the ra-
dioactive source from the collecting point to the counting
setup.
To identify the γ-ray lines which belong to the189Pb
decay (189Pb →189Tl), singles γ spectra were recorded
with the laser beam tuned on a resonant frequency ex-
pected for the low-spin isomer of189Pb and then on a
resonant frequency expected for the high-spin isomer of
189Pb. Singles γ spectra were also recorded without the
laser beam to identify the γ lines due to the189Tl de-
cay (189Tl →189Hg) or due to background radiation. For
these three measurements the ion beam was collected dur-
ing a collecting time tc= 6s, starting 10s after a proton
pulse and the γ-rays from the sources were counted for a
measuring time tm= 600s.
To observe hyperfine spectra two laser frequency scans
were performed. In order to control the stability of the

Page 3

J. Sauvage et al.: Nuclear structure of189Tl states studied via β+/EC decay and laser spectroscopy of189m+gPb 35
experimental system, the laser frequency, ν, was tuned
to a reference frequency after every third-frequency step.
For every frequency step two sources were collected for
tc= 2s, starting 8s after a proton pulse. The total mea-
suring time for the two sources was tm = 223s. The
ISOLDE tape transport system was limited by the tape
length and, therefore, only 21 frequency steps, 6 of them
for the reference frequency, were taken at one time. The
scanned frequency range was from ν = 17643.03 up to
ν = 17643.63cm−1with a step δν = 0.04cm−1. The ref-
erence frequency was ν = 17643.3299cm−1.
This combined nuclear and atomic spectroscopy
method has been used for the first time to successfully
study the α decay of the185m,gPb isomers [40].
In the second experiment we used a more advanced
tape transport system [41] with precise and reliable move-
ments. Two counting stations were installed. The first sta-
tion was placed at the collecting point to observe pos-
sible γ-rays emitted by nuclei decaying with short half-
lives. It consisted of two γ detectors with Be windows, the
planar Ge(HP) X-ray detector as described above, and
one coaxial Ge(HP) detector with 18% relative efficiency
and 1.9keV FWHM resolution at 1.33MeV. The second
counting station, located at about 1m above the collect-
ing point, consisted of three detectors, one implanted Si
α detector (100mm2area and 100μm thickness) with
14keV FWHM resolution at 5.7MeV and two coaxial
Ge(HP) γ detectors one with 60% relative efficiency and
3.6keV FWHM resolution at 1.33MeV and the other with
70% relative efficiency and 4.3keV FWHM resolution at
1.33MeV. The rather poor resolution of these two de-
tectors was partly due to electronic noise created by the
laser system. In the first experiment it was shown that the
counting rate was much weakened in the β-γ coincidence
mode because the EC decay of the189Pb nucleus is not
negligible, therefore, the 4π-β detector was not installed.
To identify the γ lines belonging to the
cay, γ-γ coincidence measurements were performed for
10 hours at the 17643.20cm−1laser beam frequency for
which both the189mPb and189gPb isomers are expected
to be ionized with a good efficiency. The ion beam was
collected for tc = 300ms immediately after the proton
pulse and measured for tm = 52s after the tape move-
ment (td= 315ms).
Subsequently, hyperfine spectra were recorded using
the same procedure as before but with 38 frequency
steps, 10 of them at the reference frequency. To obtain
a more accurate frequency determination from the hy-
perfine spectra, the frequency step size was decreased to
δν = 0.02cm−1and the reference frequency was ν =
17643.20cm−1. For every frequency step only one source
was collected for tc= 2.5s immediately after the proton
pulse and measured for tm = 54s after the tape move-
ment. Four frequency scans have been executed: two us-
ing increasing frequency values and two with decreasing
frequency values.
For the two experiments a data COMET-NARVAL ac-
quisition system [42] has been used. The COMET card
serves to determine the energies and time correlations
(with a step δt = 400ps) of the detected radiations. For
189Pb de-
0
1000
With laser -189Pbg+m - a
314.1
318
386.2
398.9
463.7
428
480.3
511.0
667.4
700.4
824.5
864.5
Counts
0
2000
318
386.2
398.9
428
463.7
480.3
511.0
700.4 824.5
864.5
With laser - 189Pbm -b
0
500
1000
400600 800 1000 12001400 1600Channel
Without laser - A=189 -c
215.7
228.6
333.9
318
511.0
567.4
Fig. 1. γ spectra obtained with the laser beam tuned on the
resonant frequency expected to ionize a) the low-spin and high-
spin or b) only the high-spin189Pb isomer and c) the one ob-
tained without the laser beam. The γ-line energies are given
in keV.
each detected radiation the energy and corresponding time
are encoded and associated to create an event. Then, the
NARVAL software sends all the events and the laser fre-
quency to be recorded on disk and builds on-line control
spectra. All information being preserved on disk coinci-
dence events can be constructed off-line for any given co-
incidence window, and sorted to get γ-γ-t coincidence as
well as γ-t and ν-γ matrices.
3 Results
3.1 First experiment
Figure 1 shows the γ-ray spectra recorded with the laser
beam tuned on the resonant frequency expected for the
low-spin and high-spin
laser beam. The 215.7, 228.6, 333.9 and 567.4keV γ lines
have a stronger intensity in fig. 1c, which indicates that
they belong to the β+/EC decay of the189Tl nuclei. The
314.1, 386.2, 398.9, 428, 463.7, 480.3, 700.4 and 824.5keV
γ lines have strong intensities in figs. 1 a and b. All these
lines except the 824.5keV γ line have also been observed in
in-beam experiments [34,43,44], which strongly suggests
that they belong to the β+/EC decay of the189Pb high-
spin isomer. In fig. 1a, one 667.4keV γ line appears with
a weak intensity and is not observed in figs. 1 b and c,
therefore, it is a good candidate for the β+/EC decay of
189Pb isomers and without the

Page 4

36The European Physical Journal A
the189Pb low-spin isomer. The 318 and 463.7keV γ lines
are higher in fig. 1a than in fig. 1b, which suggests that
they could be fed partly from the189Pb low-spin isomer.
We have to note that the 318keV γ line is observed in the
three γ-ray spectra implying that it is, at least, a doublet.
A 317.8keV γ line belonging to the189Tl decay has been
identified by Gowdy [45]. 317keV and 319.7keV γ lines
were observed in a decay study of the189Pb nuclei [33]
and the 319keV γ line is believed to be the γ-ray transi-
tion linking a 3/2+state to the 1/2+ground state of the
189Tl nucleus [13,33,46], which means that the 319keV
could belong to the decay of the189Pb low-spin isomer.
Therefore, the high intensity of the 318keV γ line ob-
served in fig. 1a could be due to a strong feeding of the
189Tl from the β+/EC decay of the189Pb low-spin isomer
and/or to the decay of the189Pb low-spin isomer.
The data recorded during the frequency scans were
used to obtain X-ray and γ-ray singles spectra and γ-γ-t
and ν-γ matrices. In spite of the low statistics of the γ-ray
spectra taken with the planar detector, the 318keV γ line
is clearly observed as a doublet: a 317keV and a 319keV
belonging to the decay of189Tl and189Pb nuclei, respec-
tively. Further evidence is found in the γ-γ-t matrix, where
the 319keV γ line belonging to the189Pb decay is also
clearly a doublet. The 463.7 and 428.3keV γ lines are
observed in coincident spectra for gates set on γ-rays at
about 318 and 319keV, respectively.
Although, precise hyperfine spectra were not achieved
in the first experiment because of the technical limita-
tions mentioned above, a few conclusions could be drawn.
We obtained qualitative hyperfine spectra for the primary
γ lines. The hyperfine spectra of the 667.4keV γ line have
a very poor statistics, however, they are similar to that of
the low-spin isomer of the185Pb isotope. The hyperfine
spectra of all the other γ lines are similar to the hyperfine
spectrum of the high-spin isomer of the185Pb isotope [40].
The results of this first experiment suggest that a low-
spin isomer exists in189Pb and that the 667.4keV γ line
belongs to its β+/EC decay. The 318keV γ line is a triple
line in agreement with the results reported in refs. [33,45].
On the other hand, the 318, 463.7 and 428keV γ lines have
coincidence relationships that contradict the level scheme
established from in-beam experiment results [34]. Further-
more, the hyperfine spectra of the 319keV are not similar
to that of the low-spin isomer of the185Pb isotope [40],
which raises doubts on the identification of the 319keV
γ line given in refs. [13,33,46].
3.2 Second experiment
Since no γ-ray emitted by nuclei having half-lives shorter
than 30s was observed with the collection-point counting
setup, only the results that have been obtained using the
setup installed 1m above the collection point (see sect. 2)
are presented below.
The data recorded at the fixed frequency ν
17643.20cm−1of the laser beam have been used to obtain
singles γ spectra as well as γ-γ-t and γ-t matrices. Singles
γ spectra were analysed using the GAMANAM curve-
=
0
1000
Gate on γ - 398.9 keVa
283.4
386.2
449.0
480.3
Counts
0
1000
2000
269.7
317.7
326.2
361.0
430.0
422.1
450.9
483.1
491.0
511.0
587.6
644.1
690.2748.0775.3781.6 787.0
Gate on γ - 463.7 keVb
0
500
Gate on γ - 587.6 keVc
463.7
420.2450.9429.0
0
200
400
400600 800
Channel
428.3
318.8
747.1
(386.2)
249.7
Gate on γ - 728.6 keVd
Fig. 2. Prompt coincident γ spectra for gates set on a) the
398.9, b) 463.7, c) 587.6 and d) 728.6keV γ lines. The coinci-
dent γ line indicated in parentheses in d) is due to a contri-
bution of the 730keV γ line in the 728.6keV gate. The γ-line
energies are given in keV.
fitting code [47], a modified version of the GAMANAL
code [48]. A152Eu radioactive source was used for energy
and efficiency calibrations. The energies and intensities of
the γ-rays belonging to the β+/EC decay of the189Pb
isomers are listed in table 1. Four γ-γ matrices have been
built: two for prompt coincidence events with coincidence
window times of tw= 160ns and 400ns, and two for de-
layed coincidence events with tw= 120ns shifted by 80ns
with respect to the prompt window. These matrices have
been treated using a graphic analysis software for one-
and two-dimensional histograms [49]. Coincident spectra
obtained for gates on important γ lines are displayed in
figs. 2–5 and the coincidence relationships are given in
table 1.
Four ν-γ matrices corresponding to the four frequency
scans have been built. Gates have been set on the γ lines
with intensity larger than 5% relative to the 386.2keV
γ line, the hyperfine spectra so extracted served to obtain
information on the origin of the corresponding γ-rays. The
results obtained from these four scans are very similar.
The hyperfine spectra obtained for 386.2keV gated γ lines
have the structure expected for the 13/2 high-spin189Pb
isomer, whereas the hyperfine spectra for gates set on the
667.4keV γ lines are completely different and correspond
to the structure expected for the 3/2 low-spin189Pb iso-
mer. The hyperfine spectra obtained for gates set on the
825, 657; 464, 782; 865, 821; 588, 644; 422 and 318, 428keV

Page 5

J. Sauvage et al.: Nuclear structure of189Tl states studied via β+/EC decay and laser spectroscopy of189m+gPb37
Table 1. γ-ray data for the β+/EC decay of the189Pb nucleus. γ-line intensities are given relatively to that of the 386.2keV
γ line (Iγ = 100). Parentheses mean uncertain coincidence relationship or tentative location.
E(a)
γ
(keV)
I(b)
γ
Main coincident γ-raysLocation
165(c)
166.4
194.0
217.5
249.7
269.7
283.4(c)
283.4
292.6
314.1
317.7(d)
318.8
318.8
326.2
336.3
348.5(d)
355.6
361
361(c)
365.1
372.5
386.2
∼ 0.9
∼ 2
∼ 1.7
1.7
0.6
3.4
∼ 0.4
∼ 0.9
0.9
13
11
20(10)(e)
29(10)(e)
0.9
1.0
∼ 0.8
0.6
∼ 1.7
∼ 1.7
2.3
0.5
100
1227.6 → 1062.6
1147.5 → 981.4
1614.8 → 1421.0
885.4 → 667.4
2006.4 → 1757.0
1332.3 → 1062.6
(1388.8 → 1105.3)
1829.7 → 1546.4
1757.0 → 1464.3
981.4 → 667.2
1062.6 → 744.7
599.8 → 281
318.8 → 0
1388.8 → 1062.6
1757.0 → 1421.0
1812.6 → 1464.3
1460.9 → 1105.3
1893.8 → 1532.8
1105.3 → 744.7
1032.5 → 667.4
2185.1 → 1812.6
667.2 → 281
314.1, 386.2, 398.9, 700.4, (613.6)
821.2
(667.4)
728.6, (318.8, 428.3, 747.1)
317.7, 420.2, 450.9, 463.7, 781.6
(824.5)
386.2, 398.9, 480.3
318.8, 864.5
166.4, 386.2, 427.1, (398.9, 664.5, 896.0, 978.2)
269.7, 463.7, (165.0, 562.7, 611.2)
428.3, 728.6, 821.2, 864.5, 934.1, 1015, (292.6, 784.6, 956.0)
1050.1, 1171, 1397.9
317.7, 463.7
(821.2)
318.8, 864.5
(824.5)
427.5, 824.5
463.7
667.4
(318.8, 784.6)
314.1, 398.9, 427.1, 437.8, 480.3, 613.2, 657.2, 682.1, 741.5, 885.5, (283.4, 449.0,
536.4, 560.3, 664.5, 730, 751.5, 978.2, 1094.4, 1138.0, 1197.7, 1211.4, 1507.2)
318.8, 667.4, 821.2
166.4, 283.4, 386.2, 449.0, 480.3
(269.7, 587.6)
463.7
314.1, 386.2, 700.4
361, 437.8, 824.5
(292.6, 428.3, 747.1)
318.8, 728.6, 784.6, (934.1)
587.6
317.7, 463.7, 781.6
386.2, (427.5)
391.4
398.9
420.2
422.1
427.1
427.5(c)
428.1(c)
428.3
429(c)
430.0
437.8
439.9(d)
449.0
450.9(d)
463.7
463.7
480.3
483.1
491.0
498(c)
536.4
541.8
560.3
562.7
587.6
611.2(c)
613.2(d)
644.1
657.2
664.5
2.5
11
1.3
5.0
∼ 3.5
∼ 7
∼ 3
∼ 9
∼ 1
5.4
2.6
∼ 0.9
∼ 1.4
∼ 1.7
33(3)(e)
13(3)(e)
37
4.8
0.6
∼ 0.4
2.0
0.9
2.0
1.3
7.2
∼ 0.9
∼ 1.7
5.1
10
1.9
1812.6 → 1421.0
1546.4 → 1147.5
1752.7 → 1332.3
885.4 → 463.7
1408.6 → 981.4
1532.8 → 1105.3
2185.1 → 1757.0
1028.1 → 599.8
(1761.2 → 1332.3)
1492.6 → 1062.6
1105.3 → 667.2
398.9, 480.3, (386.2)
(269.7, 463.7, 587.6)
1995.4 → 1546.4
1783.2 → 1332.3
744.7 → 281
463.7 → 0
1147.5 → 667.2
1227.6 → 744.7
1552.7 → 1062.6
(1962.2 → 1464.3)
1861.3 → 1324.8
2006.4 → 1464.3
(1227.6 → 667.2)
269.7, 317.7, 430.0, 450.9, 483.1, 587.6, 644.1, 775.3, (326.2, 361, 690.2, 787)
422.1
283.4, 386.2, 398.9, 449.0, 613.2, 682.1, (730, 785.3, 1027)
463.7
(317.7, 463.7, 781.6)
657.3, 1044.3, (318, 386.2)
(318.8, 463.7, 864.5)
(386.2), (317, 333.9, 405.8)(f)
XTl+XHg, (317.7)
420.2, 450.9, 463.7, (429)
317.7
386.2, 480.3, (700.4)
463.7
386.2, 536.4
314.1, 386.2, 700.4
1332.3 → 744.7
1761.2 → 1147.5
1388.8 → 744.7
1324.8 → 667.2
1645.9 → 981.4

Page 6

38The European Physical Journal A
Table 1. Continued.
E(a)
γ
(keV)
I(b)
γ
Main coincident γ-raysLocation
667.4
682.1
690.2
700.4
720(c)
728.6
730(c)
741.5
747.1
748(c)
751.5
775.3
781.6
784.6(c)
785.3
787(c)
811.9
821.2
824.5
848.4
860.0
863(c)
864.5
880.0
885.5
896.0
919.9
934.1
947.2
956.0
978.2(c)
981.4
1005.5
1012.8
1015(c)
1027(c)
1041.3
1044.3
1050.1
1094.4
1108.0
1117.2
1138.0
1171(c)
1179.3
1197.7
1211.4
1250.4
1253.9
1272.8
1314.4
1369.8
1397.9
1491.9
9.6
1.4
1.2
23
∼ 0.9
5.4
∼ 0.9
6.8
8.0
∼ 0.7
1.0
0.6
7.6
∼ 1.7
∼ 1.7
∼ 0.7
0.6
7.3
30
0.6
1.8
∼ 0.9
13
1.3
2.3
1.7
1.4
1.1
1.7
1.1
∼ 2.3
2.2
1.6
3.7
∼ 1.7
∼ 0.9
1.2
5.0
2.8
1.0
1.0
1.6
4.1
∼ 2.6
2.5
1.7
2.0
1.3
1.9
4.4
1.7
1.2
1.0
0.7
217.5, 365.1, (391.4, 439.9)
386.2, 480.3, 700.4
463.7, 781.6, (317.7)
166.4, 427.1, 664.5, 896.0, 978.2, (398.9, 613.2)
318.8, 864.5
249.7, 318.8, 428.3, 747.1
(386.2, 480.3)
386.2
728.6, 784.6, 934.1
(463.7)
(386.2)
(463.7)
269.7, 326.2, 430.0, (450.9, 690.2)
318.8, 428.3, 747.1, (372.5)
XTl
(463.7)
667.4 → 0
1829.7 → 1147.5
1752.7 → 1062.6
981.4 → 281
2185.1 → 1464.3
1757.0 → 1028.1
1877.8 → 1147.5
1408.6 → 667.2
1028.1 → 281
1492.6 → 744.7
1062.6 → 281
1812.6 → 1028.1
(1532.8 → 744.7)
(1959.6 → 1147.5)
1421.0 → 599.8
1105.3 → 281
318.8, 336.3, (194.0, 391.4)
355.6, 361, 427.5, (919.9, 1108.0)
XTl
(657.2, 1044.3)
(314.1, 386.2, 700.4)
292.6, 318.8, 348.5, 956.0, (372.5, 498.0, 541.8, 720)
(314.1, 386.2, 700.4)
386.2
386.2, 700.4
824.5
318.8, (428.3, 747.1)
XTl
318.8, 864.5
700.4, (314.1, 386.2)
XTl
XTl
XTl
(318.8)
(386.2, 480.3)
XTl
(536.4)
(318.8)
386.2
(824.5)
XTl
(386.2)
(318.8)
XTl
386.2
386.2
XTl
XTl
XTl
XTl
XTl, (386.2, 480.3)
(318.8)
XTl
2185.1 → 1324.8
1844.4 → 981.4
1464.3 → 599.8
1861.3 → 981.4
1552.7 → 667.2
1877.8 → 981.4
2025.2 → 1105.3
1962.2 → 1028.1
2420.3 → 1464.3
1959.6 → 981.4
1614.8 → 599.8
2174.4 → 1147.5
1324.8 → 281
(1368.9 → 318.8)
1761.2 → 667.2
2213.3 → 1105.3
(1489.8 → 318.8)
1864.9 → 667.2
1877.8 → 667.2
(1716.7 → 318.8)

Page 7

J. Sauvage et al.: Nuclear structure of189Tl states studied via β+/EC decay and laser spectroscopy of189m+gPb39
Table 1. Continued.
E(a)
γ
(keV)
I(b)
γ
Main coincident γ-raysLocation
1507.2
1523.7
1555.3
1579.5
1.8
1.9
1.8
1.7
386.2
XTl
XTl
XTl
2174.4 → 667.2
(a)ΔEγ ∼ 0.2keV for transitions with Iγ > 3. ΔEγ ∼ 0.4keV for the other transitions.
(b)ΔIγ ∼ 15%.
(c)Transition from coincident events.
(d)Transition mixed with descendant ones.
(e)Intensity determined from the hyperfine spectrum.
(f)Transition belonging to the descendant.
0
2000
4000
Gates on γ - 386+314+700 keVa
314.1
166.4
386.2 398.9
427.1
437.8
480.3
511.0
560.3613.2
657.2
664.5
741.5
Counts
0
1000
2000
500 1000
269.7
292.6*
317.7
(386.2)
422.1
428* + 430.0
463.7
483.1
491.0
511.0
587.6
644.1
690.2
821.2*
864.5*
Channel
Gates on γ - 318+464+781 keVb
Fig. 3. Prompt coincident γ spectra for sum of gates set on
a) the 386.2, 314.1 and 700.4keV γ lines b) 318, 463.7 and
781.6keV γ lines. In b) the coincident γ line given in paren-
theses is due to a contribution of the 314.1keV in the 318keV
gate, asterisks indicate the coincident γ lines due to the 318keV
γ line of the part c) of the level scheme. The γ-line energies
are given in keV.
γ lines are shown with dotted lines in figs. 6 a, b, c, d, e
and f, respectively, and are compared with the hyperfine
spectra, drawn with solid lines, which have been extracted
from the same frequency scan, for the 386 and 667keV
γ lines. We can see in fig. 6a that the hyperfine spectra for
the 825 and 657keV gates are very similar to that of the
386keV gate, which unambiguously indicates that these
two γ-rays are also emitted by levels fed from the high-
spin189Pb isomer. The same conclusion can be deduced
for the 782; 865, 821; 588, 644 and 428keV γ-rays from
figs. 6 b, c, d and f, respectively. On the other hand, the
hyperfine spectrum for the 422keV gate is similar to that
of the 667keV gate, indicating that the 422keV γ-ray is
from the low-spin189Pb isomer (see fig. 6e). The hyperfine
spectrum extracted for the 318keV γ line, shown (down
triangles) in fig. 6f, has a small bump at a frequency value
0
500
Left part
Counts
314.1
318.8
361.0386.2
428
437.8
463.7
511.0
700.4728.6
784.6
824.5
0
500
1000
Centre
0
250
500
400600800 1000
781.6
Right part
Channel
Fig. 4. Prompt coincident spectra for gates set on the left
part, centre and right part of the 428keV γ line. The γ-line
energies are given in keV.
around ν = 17643.45cm−1. This strongly suggests that
part of the 318keV γ-ray intensity could arise from the
low-spin isomer. In the same way a very small bump is ob-
served in the hyperfine spectrum obtained for the 464keV
gate, which could also correspond to a small feeding from
the low-spin isomer (see fig. 6b, down triangles).
To determine the possible admixtures in the 464 and
318keV γ lines, we have used the hyperfine spectra of
the 386 and 667keV γ lines to calculate hyperfine spec-
tra corresponding to different admixtures. The calculated
spectrum obtained if we assume that the high-spin and
low-spin isomers feed the 464keV γ line at 75% and 25%,
respectively, has the best agreement with the experimen-
tal hyperfine spectrum with an estimated uncertainty of
±5%. The experimental hyperfine spectrum of the 464keV

Page 8

40The European Physical Journal A
0
500
Gate on γ - 422.1 keV
463.7
511.0
Counts
0
200
5001000
217.5
(314.1)
365.1
(386.2)
511.0
(700.4)
Channel
Gate on γ - 667.4 keV
Fig. 5. Prompt coincident spectra for gates set on the 422.1
and 667.4keV γ lines. Coincident γ lines given in parenthe-
ses are due to a contribution of the 664.5keV γ line in the
667.4keV gate. The γ-line energies are given in keV.
γ line, as a solid line, and the best calculated hyperfine
spectrum, as circles linked with a dotted line, are shown
in fig. 7. The case of the 318keV γ line is more compli-
cated since an important contribution is due to the decay
of the189Tl nucleus. To simulate this contribution due to
both the direct production of189Tl in the target and the
decay of the189Pb isomers we have used the hyperfine
spectrum of the 228keV gate. None of the calculated hy-
perfine spectra reproduces exactly the experimental spec-
trum obtained for the 318keV gate. The latter, shown
with a solid line, is compared in fig. 7 to the hyperfine
spectra calculated assuming the following contributions:
i) 40% from the decay of189Tl, 35% from the low-spin
189Pb isomer and 25% from the high-spin189Pb isomer
(open circles) and ii) 40% from189Tl, 20% from low-spin
189Pb isomer and 40% from the high-spin189Pb isomer
(up triangles). The contribution due to the189Tl decay
in the two calculated spectra is shown at the bottom.
The structure of this contribution is due to the part of
the189Tl coming from the189Pb decay, it can hence de-
pend not only upon the percentage of the production by
decay but also on the relative feeding from the low-spin
and high-spin Pb isomers. The 228 and 317keV γ lines
belonging to the189Tl decay can have different feeding
modes, which could explain why it is not possible to get
a very good agreement between the experimental hyper-
fine spectrum of the 318keV and the calculated spectrum
whatever the admixtures assumed. Therefore, using a con-
servative estimate for the uncertainties, we obtain for the
admixture in the 318 keV: 40±10% from189Tl, 28±10%
from the low-spin189Pb isomer and 32 ± 10% from the
high-spin189Pb isomer. Nevertheless, the intensities that
are deduced from these results for the 464 and 318keV,
have been corrected for taking into account the different
intensities observed in the γ singles spectra obtained dur-
ing the coincidence measurement and the frequency scan
and then, they have been reported in table 1. It is worth
noting that the main intensity differences are observed for
the γ-rays of the189Tl decay, the 228keV γ-line relative
intensity measured during the coincidence measurement is
only 36% of that measured during the frequency scan.
3.2.1 Level scheme
The partial level scheme of189Tl was built from the γ-ray
energies, intensities and coincidence relationships reported
in table 1, as shown in fig. 8. The hyperfine spectra indi-
cate very clearly that the three parts a, b and c, placed
on the left-hand side of fig. 8, are fed by the 13/2+high-
spin189Pb isomer, which means that the lowest-energy
states of every parts have spin values I ≥ 7/2 (see be-
low). Therefore, the a, b and c parts cannot be based on
the 1/2+ground state but on the 9/2−isomeric state of
the189Tl nucleus. Some weak links observed between the
states of these different parts give support to their place-
ment. Part d on the right-hand side of fig. 8 is fed from
the low-spin189Pb isomer. No link could be found between
the part d and the three other parts of the level scheme,
therefore, we have located the first state of the parts a,
b and c at 281keV according to the energy suggested by
Coenen et al. [46] from the energies of α-rays emitted by
the193Bi isomers, the uncertainty of this energy is esti-
mated to be ±7keV. It is worth noting that we show in
this work that the 258keV value proposed for the energy
of the189Tl isomer in ref. [34] is wrong since the level
scheme built in ref. [34] has to be modified (see also below
the description of part b).
Part a of the level scheme consists of 23 levels, the
667.2, 981.4, 1147.5, 1408.6, 1546.4 1829.7 and 1995.4keV
levels correspond to states already observed in in-beam
experiments [34,43,50], examples of coincident spectra for
gates set on the 398.9keV γ line and the sum of the gates
set on the 386, 314 and 700keV γ lines are displayed
in figs. 2a and 3a, respectively. The 1645.9keV level is
due to the 664.5keV γ line observed in coincidence with
the 314.1, 386.2 and 700.4keV γ lines (see table 1 and
fig. 3a). This confirms the 664.5keV γ-ray position in
the level scheme given in refs. [34,43,44]. The 863.0keV
that de-excites the 1844.4keV level confirms the results re-
ported in refs. [34,44]. The 657.2keV in coincidence with
the 386.2keV γ line, de-excites the 1324.8keV level. This
transition was only observed by Porquet et al. [43]. The
1861.3keV level decays to both the 1324.8 and 981.4keV
states, which allowed us to confirm the position of the
657.2keV γ line on the 667.2keV state. An intense γ
line has been observed at 824.5keV; this line was never
mentioned in in-beam works but was seen in a decay
work [33]. However, its hyperfine spectrum clearly indi-
cates that it belongs to the high-spin isomer decay (see
fig. 6a). Moreover, it appears with the 437.8keV γ line in
the coincidence spectra obtained for gates set on the left
part and at the centre of the 428keV γ line, as can be seen
in fig. 4. Therefore, the 824.5keV γ-ray was put directly

Page 9

J. Sauvage et al.: Nuclear structure of189Tl states studied via β+/EC decay and laser spectroscopy of189m+gPb41
Fig. 6. Comparison of the hyperfine spectra obtained for gates set on the main γ lines (open triangles) with those for gates
on the 386.2 (full circles linked with solid lines) and 667.4keV (asterisks linked with solid lines) γ lines that correspond to the
high-spin and low-spin189Pb isomers, respectively. All the hyperfine spectra have been obtained for gates set on the same ν-γ
matrix but the intensities of those of the 386.2 and 667.4keV gates have been divided by different factors to make the figures
clearer (386.2keV: N/4 in a), b) and f), N/6 in c) and N/10 in d) and e); 667.4keV: N/2 in b), c), d) and e)). Frequency values
in cm−1correspond to the laser beam of the first excitation step before doubling.

Page 10

42The European Physical Journal A
Fig. 7. Comparison of the hyperfine spectra observed for the
464 and 318keV gates (solid lines) with the hyperfine spec-
tra calculated assuming that the high-spin (HS) and low-spin
(LS) isomers feed: i) the 464keV γ line with 75% and 25%,
respectively (full circles) and ii) the 318keV γ line with 25%
and 35% (open circles) or with 40% and 20% (full triangles),
respectively, with in both cases a contribution of 40% of the
189Tl decay (D). The hyperfine spectrum of the 228keV gate
that has served to simulate the contribution of the189Tl de-
cay in the 318keV γ line is shown at the bottom of the figure.
Frequency values in cm−1correspond to the laser beam of the
first excitation step before doubling.
on the 281keV level. Thus, an 1105.3keV level was es-
tablished as well as the 1460.9, 1532.8, 1893.8, 2025.2 and
2213.3keV levels since they decay towards the 1105.3keV
state. Six other levels have been built from the coinci-
dences observed with the 386, 480 and/or 700keV γ lines.
Part b of the level scheme includes the 463.7, 317.7,
781.6, 269.7, 587.6, 430.0 and 450.9keV γ lines that were
also observed by Riedinger et al. [34] but we established
a completely different level scheme from our coincidence
relationships. From the level scheme of ref. [34] the 781.6,
450.9, 291 and 395keV γ lines should be the main γ lines
observed in coincidence with the 463.7keV γ-ray. One can
see in the coincidence spectrum of the 463.7keV gate,
shown in fig. 2b, that the 317.7keV is much higher than
the 450.9keV γ line and that neither the 781.6keV nor
the 291 and 395keV γ lines are observed. In the same
way the 450.9keV should be the main γ line observed in
coincidence with the 587.6keV γ-ray. In the coincidence
spectrum of the 587.6keV gate the 450.9keV is very small
compared to the 463.7keV γ line (see fig. 2c). Thus, our re-
sults are in contradiction with the level scheme of ref. [34]
and consequently indicate that the 258keV value proposed
for the energy of the189Tl high-spin isomer is wrong.
The hyperfine spectra shown in figs. 6 b and d indicate
that the levels de-excited by the 587.6, 644.1 and 781.6keV
γ-rays are only fed by the high-spin189Pb isomer whereas
the level de-excited by the 463.7keV γ-ray should be fed
at 25 ± 5% by the low-spin isomer. In addition to this,
the 422.1keV γ line which must de-excite a level only fed
by the low-spin isomer (see fig. 6e), is in coincidence with
the 463.7keV γ-ray. The sum of the energies of these two
coincident γ-rays is equal to 885.8keV, this energy sum is
very close to the energy of the cascade 217.5-667.4 equal to
884.9keV. These facts strongly suggest that the 463.7keV
could be a doublet. For these reasons, the 422.1-463.7keV
cascade has been placed in part d of the level scheme. The
relative intensity of this 463.7keV transition is 13±3 from
the hyperfine spectrum (see table 1).
The 361keV is in coincidence with both the 427.5 and
463.7keV γ lines, and is about twice higher in the coin-
cidence spectrum of the 427.5keV gate than in that of
the 463.7keV gate (see figs. 2b and 4a). This means that
the 361keV is very likely a double line: one placed above
the 427.5keV transition and the other one under it. This
latter is one of the links found between parts a and b of
the level scheme. The other γ-rays that are possible links
could not be confirmed since no γ transition was observed
above the levels that these links de-excite. These links are
seen as very weak peaks in the coincidence spectra: the
560.3keV is in coincidence with the 386.2keV γ line (see
table 1 and fig. 3a), the 787keV with the 463.7keV γ line
(fig. 2b), the 429keV with the 587.6keV γ line (fig. 2c)
and the 491.0keV with the 463.7 and 781.6keV γ lines
(see table 1, figs. 2b and 3b).
As for part c of the level scheme, the hyperfine spectra
indicate that the levels de-excited by the 428.3 and 747.1,
821.2, 864.5 and 728.6keV γ lines are fed by the high-
spin189Pb isomer. On the other hand, the hyperfine spec-
trum of the 318.8keV gate reveals that the contributions
of the low-spin and high-spin isomer decays are about the
same (see fig. 7). The 599.8keV level that is linked by the
428.3keV γ-ray to the 1028.1keV level directly fed by the
13/2 high-spin isomer decay, is probably not fed directly
by the low-spin isomer decay since its spin value is at least
7/2. This led us to conclude that the 318.8keV γ line is
a doublet. For these reasons, part of the 318.8keV γ line
has been attributed to part d of the level scheme. The rel-
ative intensity of this 318.8keV γ line has been estimated
at 29 ± 10 from the hyperfine spectrum.
The 657.2 and 1044.3keV γ lines, de-exciting the level
at 1324.8keV, displayed in part a of the level scheme, have
been observed in coincidence with the 860.0keV γ line.
This gives a level located at 2184.8keV, an energy very
close to the energy of a level established at 2185.1keV in
the part c of the level scheme. Thus, the 860.0keV γ-ray is

Page 11

J. Sauvage et al.: Nuclear structure of189Tl states studied via β+/EC decay and laser spectroscopy of189m+gPb43
0
599.8
318.8
667.4
463.7
1028.1
885.4
1032.5
1421.0
1464.3
1757.0
1489.8
1368.9
1/2+
1614.8
1962.2
1812.6
2006.4
1716.7
2185.1
2420.3
1783.2
1752.7
1492.6
1388.8
1332.3
1227.6
1062.6
744.7
281281
667.2
981.4
1105.3
1147.5
1324.8
1408.6
1460.9
1532.8
1546.4
1552.7
1645.9
1761.2
1829.7
1844.4
1861.3
1864.9
1877.8
1893.8
1959.6
1995.4
2025.2
2174.4
2213.3
9/2-
11/2-
13/2-
13/2+
15/2-
13/2-
17/2-
17/2+
17/2-
17/2+
15/2+
318.8(20)
318.8(29)463.7(13)
667.4(9.6)
428.3((9)747.1(8)
217.5(1.7)422.1(5.0)365.1(2.3)
821.2(7.3)
864.5(13)194.0(1.7)
1015(1.7)
1050.1(2.8)
292.6(0.9)336.3(1.0)728.6(5.4)
1171(2.6)
348.5(0.8)391.4(2.5)784.6(1.7)498(0.4)
934.1(1.1)
1397.9(1)
249.7(0.6)541.8(0.9)
372.5(0.5)428.1(3)860.0(1.8)720(0.9)
956.0(1.1)
450.9(1.7)
690.2(1.2)420.2(1.3)
748(0.7)430.0(5.4)
644.1(5.1)326.2(0.9)
587.6(7.2)269.7(3.4)
483.1(4.8)165(0.9)
781.6(7.6)317.7(11)
463.7(33)
1108.0(1.0)
1507.2(1.8)1027(0.9)
919.9(1.4)
449.0(1.4)
978.2(2.3)811.9(0.6)
361(1.7)
1211.4(2.0)896.0(1.7) 730(0.9)
1197.7(1.7)
880.0(1.3)536.4(2.0)
863(0.9)
682.1(1.4)283.4(0.9)
1094.4(1.0)613.2(1.7)
664.5(1.9)
885.5(2.3)
398.9(11)
427.5(7)
355.6(0.6)
741.5(6.8)427.1(3.5)
1044.3(5.0)657.2(10)
480.3(36.5)166.4(2)
824.5(30)437.8(2.6)
700.4(23)314.1(13)
386.2(100)
361(1.7)
787(0.7)
491.0(0.6)
429(1)
283.4(0.4)
560.3(2.0)
189Tl
abc d
Fig. 8. Partial level scheme of the189Tl nucleus obtained from the β+/EC decay of the189Pb isomers. The energy of the 9/2−
isomer is 281 ± 7keV taken from ref. [46] (see text). The reported spin and parity values have been taken from refs. [34,43].
The part d of the level scheme, fed by the low-spin189Pb isomer, is tentative.
the only possible link we found between part a and part c
of the level scheme.
Part d of the level scheme is built around the
667.4-217.5keV cascade indisputably fed from the low-
spin isomer of the189Pb nucleus. One can see in fig. 5
that the 667.4keV is only in coincidence with the 217.5
and 365.1keV γ lines. The parts of the 318.8 and 463.7keV
double γ lines that de-excite the levels fed from the low-
spin isomer have been placed on the ground state. Unfor-
tunately, their position could not be confirmed from other
coincidence relationships since no γ-ray is observed above
the 318.8 and 885.4keV levels, therefore, their placement
in the level scheme should be considered tentative.
Most of the γ lines identified as belonging to the
β+/EC decay of the189Pb isomers have been placed in
the level scheme shown in fig. 8, which corresponds to 93%
of the observed γ intensity. Given that i) no line with en-
ergy lower than 165keV has been observed, ii) no level has
lifetime long enough to observe delayed transitions in the
matrices of delayed coincidence events, which eliminates
the high multipolarities (> M2 and M2 for transition en-
ergy < 500keV) for the observed transitions and iii) most
of the transitions have an unknown multipolarity, a first
qualitative level intensity balance has been performed us-
ing the γ line intensities. In table 2 the γ intensity bal-
ances have been listed for the levels that are fed at more
than 5% of the 386.2keV γ intensity or those that have
a spin value attributed from in-beam works [34,43]. One
can see that the γ balance value is positive for all levels
except for the 599.8keV one. If we assume an M1 mul-
tipolarity for the 318.8keV transition that de-excites the
599.8keV level, its total intensity is then equal to 26.6
whereas the total level feeding is equal to 31.8 ± 1.1, the
error is estimated from the E1, E2 or M1 possible mul-
tipolarity of the four transitions that feed the level. This
indicates an underestimation of this 318.8keV transition
intensity, the γ intensity should be at least 24 instead of
20 which is in agreement with the estimated error bar of
the 318.8keV line intensity (20 ± 10) reported in table 1.
One can see from table 2 that the states with spin value
11/2, 13/2 or 15/2 are obviously the most fed levels from
the β+/EC decay of the 13/2+isomer. It is clear that, in

Page 12

44The European Physical Journal A
Table 2. Intensity balances, logft values, and spin and parity values for the most fed levels of the parts a, b and c of the level
scheme, the 599.8 level and those that have spin and parity values assigned from in-beam experiments [34,43].
Intensity balance for high-spin states
Feeding Decaying
Iγ
Iγ
Iγ
3120
−11
80.2 10019.8 19.1 ? 37.6
14.23621.8 22.6 ? 26.2
8.2178.8
12.4 18.66.2
10.434.323.9 22.9 ? 25.4
16.53922.5 21.0 ? 22.5
7.77.7
3.81511.2 11.2 ? 11.7
410.66.6
6.46.4
10.3 10.3 10.5 ? 10.9
5 138
6.16.1
1.77.76
2.310.48.1
1.9 1.9
2.3 2.3
0.90.9
1.41.4
6.26.2
Elevel
keV
599.8
667.2
981.4
1028.1
1062.6
1105.3
1147.5 13/2+
1227.6
1324.8
1332.3
1388.8
1408.6
1464.3
1492.6
1532.8
1546.4 15/2+
1645.9
1829.7 17/2+
1844.4
1995.4 17/2+
2185.1
Iπ
BalanceFeeding
Itot(%)
–
11.0 ± 3.6
9.5 ± 0.7
4.0 ± 0.1
2.3 ± 1.5
9.4 ± 0.5
8.4 ± 0.3
3.5 ± 0.5
4.4 ± 0.1
2.9 ± 0.5
2.6 ± 0.2
4.1 ± 0.1
3.0 ± 0.2
2.5 ± 0.2
3.2 ± 0.2
3.7 ± 0.3
0.74
0.99 ± 0.06
0.35
0.62
2.5 ± 0.1
logf0tlogf1tIπ
refs. [34,42]Itot
11/2−
13/2−
6.04 ? 6.34 8.0 ? 8.3
6.1
6.5
6.5 ? 7.1
6.1
6.1
6.5
6.4
6.5
6.6
6.3
6.5
6.5
6.4
6.4
7.0
6.8
7.3
7.0
6.3
8.0
8.410.0 ? 10.5
2.3 ? 9.8
11/2±,13/2±,15/2±
8.4 ? 9.0
8.0
8.0
8.4
8.2
8.4
8.4
8.2
8.4
8.4
8.3
8.2
8.9
8.7
9.1
8.8
8.1
11/2±,13/2±,15/2±
7.8 ? 10.2 11/2±,13/2±,15/2±
13/2−
6.3 ? 8.9
6.4 ? 7.2
11/2±,13/2±,15/2±
11/2±,13/2±,15/2±
15/2−
7.3 ? 8.3
6.2 ? 6.9
7.8 ? 8.8
8.7 ? 10.4
1.9
2.4 ? 2.7
0.9
1.6
6.2 ? 6.8
11/2±,13/2±,15/2±
11/2±,13/2±,15/2±
11/2±,13/2±,15/2±
17/2−
17/2−
11/2±,13/2±,15/2±
this decay work, we can observe the states with spin value
≤ 17/2 only. The 281keV 9/2−level cannot be directly
fed from the 13/2+isomer, therefore, it is possible to esti-
mate the number of disintegrations of the high-spin189Pb
isomer by summing up the intensities of the 8 transitions
feeding the 281keV level. Doing so we obtain 247 ± 12,
where the error is due to the unknown transition multi-
polarities assumed to be only E2 or M1 since a positive
parity is not probable for the low-lying levels under con-
sideration (see further).
To determine the feedings to the states fed from the
189Pb high-spin isomer we have used the transition mul-
tipolarities and spin values known from in-beam exper-
iments [34,43] and assumed an E2 or M1 mutipolar-
ity for the 317.7, 318.8, 428.3, 437.8, 463.7, 747.1, 781.6,
824.5keV transitions and an E1, E2 or M1 multipolarity
for the other transitions. We used the internal conversion
coefficients of ref. [51]. The sum of the calculated feedings
is equal to 258±30, the error bar corresponds to only the
unknown transition multipolarities. It is in good agree-
ment with the number of disintegrations given above. The
logf0t and logf1t values, for the most fed states (≥ 2.3%)
and for those with assigned spin and parity values, have
been calculated using the logft tables [52] with the ob-
tained feedings, QEC= 6.76MeV [53,54] and T1/2= 51s
(ref. [53]) and are reported in table 2. In spite of the miss-
ing multipolarities, the β+/EC level feedings larger than
2.4% have rather small relative uncertainties (< 17%) ex-
cept for the 667.2keV level feeding (33%). The calculated
logf0t and logf1t values are smaller than 8.5 except for
the 1062.6keV and 17/2 levels, which means, from the
Raman-Gove rules [55], that the β decay to these lev-
els corresponds to allowed or first-forbidden non-unique
transitions. Therefore, the spin value of these states is
I = 11/2, 13/2 or 15/2 with a positive or negative parity
(see the last column of table 2).
3.2.2 Half-lives
The measuring time tm = 52s was too short to provide
accurate decay half-life determination of the189Pb iso-
mers. However, estimations were extracted from the γ-t
matrix built from the singles γ data recorded during the
γ-γ coincidence measurement. The time spectra obtained
for gates set on the 386, 480, 667, 700 and 865keV γ-rays
were analysed by the conventional least-square method, χ2
method, and the results were confirmed by fits performed
using the Awaya method [56]. Then, the half-lives were
corrected for the dead time of the acquisition system that
is measured for every acquisition channel, the dead time
due to pile-up in the amplifier is not taken into account:
correction of 6.2±1.0% for the half-life of the high-spin iso-
mer from the 386.2keV γ-ray and 5±0.8% for that of the
low-spin isomer from the 667.4keV γ-ray. The results are
the following: T1/2= 49.7±0.1, 46.9±0.2, 54.8±0.5s from
the 386, 480 and 865keV γ-rays, the given error bars are
only due to statistical uncertainties. The Compton back-
ground represents only 24% of the total intensity for the
386keV gate, whereas it is 63% for the 865keV gate result-
ing in different uncertainties, which explains the different
T1/2values. From these results, the half-life of the 13/2+

Page 13

J. Sauvage et al.: Nuclear structure of189Tl states studied via β+/EC decay and laser spectroscopy of189m+gPb 45
Fig. 9. Energy evolution of the first two members of every high-spin coupled states (I > 5/2) as a function of the deformation
parameter β (β < 0 for oblate deformation, β > 0 for prolate deformation).
isomer of189Pb was estimated to be T1/2= 50 ± 3s, in
agreement with the value T1/2= 51 ± 3s reported in the
literature for189Pb [53]. Therefore, our result shows that
this T1/2value has to be attributed to the 13/2+isomer
and not to the low-spin isomer of189Pb. The longer value
T1/2= 63s found for the 700keV gate could be due to
the presence of a weak-intensity line with an energy close
to 700keV and belonging to a189Pb descendant or to the
background noise. The half-life of the 3/2−isomer of189Pb
has been estimated to be T1/2= 39±8s from the 667keV
gate, the error bar is mainly due to the uncertainty on the
Compton background subtraction that represents 71% of
the total intensity for the 667keV gate.
4 Discussion
To identify the low-lying levels established in189Tl we
have used the semi-microscopic axial-rotor coupled to one-
quasiparticle model described in ref. [36]. The quasiparti-
cle wave functions, in the188,190Pb neighboring cores, have
been obtained from HF plus BCS calculations [57–59] us-
ing the Skyrme III force [60] and the usual pairing inter-
action with constant matrix elements (G0n = 14.6 and
G0p= 13.3 for Pb) and standard cut-off. The moments of
inertia used in the calculations with the188Pb and190Pb
cores have been deduced from the energies of the states
built on the first 2+levels of the188Pb and190Pb nuclei lo-
cated at 724 and 774keV, respectively; the energies of the
0+states of these bands have been estimated by extrap-
olation of the variable moments of inertia of the excited
bands.
Deformation parameters, β, for the 9/2−isomer of the
189Tl nucleus, have been deduced from laser spectroscopy
measurements: ?β2?1/2= 0.181 from the isotope shift
value assuming β = 0 for207Tl and β = −0.153 from
the intrinsic quadrupole moment, Q0, deduced from the
measured spectroscopic quadrupole moment, Qs, assum-
ing strong coupling model and K = I = 9/2 (ref. [7]).
The existence of a small component (few %) K = 7/2 in
the wave function of the 9/2 state would be enough to
explain the smaller β value deduced from Qs. In this mass
region, it is well known that the nuclear deformation can
depend on the state occupied by the coupled proton [6–11].
Thus, in191Tl, the β value obtained for the 9/2−isomer
is ?β2?1/2= 0.16 from the isotope shift value (β = −0.142
from the Qs value) whereas for the 1/2+ground state
?β2?1/2= 0.112 from the isotope shift value [8]. This led
to calculate every coupled states for different deformations
by constraining the quadrupole moment of the core in the
HF plus BCS treatment. The energies and wave functions
of the quasiparticle states in the188,190Pb cores have been
determined for the following deformation parameters: β =
−0.21,−0.18,−0.15,−0.12,+0.12,+0.15,+0.18,+0.21. It
is worth noting that the results obtained using the two
cores,188,190Pb, are similar. The energy evolution of the
two first states of every coupled states as a function of the
β value, obtained using the190Pb quasiparticle wave func-
tions, is shown in fig. 9 for high-spin states (I > 5/2) and
in fig. 10 for low-spin states (I < 7/2). In these figures the
particle states are labelled with the quantum numbers,
I K[NnzΛ], of the main component of their wave func-
tion (I is the angular momentum of the state, K is the
projection of I on the symmetry axis of the nucleus and
N, nz, Λ are the Nilsson asymptotic quantum numbers).
We can note that for the rotational bands built on the
1/2−[521] state corresponding to an oblate nuclear shape
and on the 1/2−[541] and 1/2+[640] states corresponding
to a prolate nuclear shape, the bandheads have I = 11/2,
I = 9/2 and I = 13/2+, respectively indicating a decou-
pling of the single-particle and core movements due to the
Coriolis force.
4.1 High-spin states
The ΔI = 1 structure of the rotational band built on the
9/2−isomeric state is well reproduced with calculations
performed assuming an oblate nuclear shape (negative β
values). Moreover, the 9/2−[505] state is predicted to be
the first high-spin level for oblate shape with β ≥ −0.18
(see fig. 9), which is in agreement with the deformation de-
termined from the laser spectroscopy studies [7,8]. As seen
in fig. 9 several other coupled states are expected to be
lying at low energy in189Tl, 7/2−[514] and 13/2+[606]

Page 14

46The European Physical Journal A
Fig. 10. Energy evolution of the first two members of every low-spin coupled states (I < 7/2) as a function of the deformation
parameter β (β < 0 for oblate deformation, β > 0 for prolate deformation).
states with oblate nuclear shape and 9/2 1/2−[541] (aris-
ing from the h9/2sub-shell) and 11/2−[505] states with
prolate nuclear shape.
The 1147.5, 1546.4 and 1829.7keV levels observed in
this work had already been identified as the 13/2+, 15/2+
and 17/2+members of the 13/2+[606] band from previous
in-beam works [43,50]. Shown in fig. 9 the energy location
of these states is well reproduced for an oblate deformation
with β ∼ −0.17.
The 7/2−[514] state is predicted to be located above
the 9/2−[505] state and below the 13/2+[606] state for
an oblate deformation with β ≥ −0.18 and it could
be the established level at 744.7keV. Furthermore, the
317.7-269.7-450.9 γ cascade observed in the present work
and in in-beam work [34] could link the first members of
the 7/2−[514] rotational band, which leads us to assign
the spin and parity values Iπ= 9/2−, 11/2−and 13/2−
to the 1062.6, 1332.3 and 1783.2keV levels, respectively.
The weak feeding of the 744.7 and 1062.6keV levels and
the possible Iπvalues of the 1332.3keV level deduced from
the logft values are in favour of these assignments.
The 1105.3keV level, with Iπ= 11/2±, 13/2±or
15/2±, decays to the 9/2−level at 281keV and the 11/2−
level at 667.2keV. It decays also towards the 744.7keV
level proposed above as the 7/2−[514] state. It is likely
an 11/2−level, therefore, it could be the 11/2−[505] state
predicted in the calculations performed assuming prolate
deformations (see fig. 9). The 824.5keV γ line that de-
excites the 1105.3keV level, has never been observed in
in-beam experiments, which lends support to this last pro-
posal. The 13/2−member of the 11/2−[505] band should
be the 1460.9keV rather than the 1532.8keV level because
of the possible link of the 1532.8keV level with the (7/2−)
744.7keV state.
The 9/2 1/2−[541] state with a prolate nuclear shape
is predicted to be located at a smaller energy than the
11/2−[505] state for β < 0.21 and the rotational band
built on it has a ΔI = 2 decoupled structure. The
599.8keV level could be assigned the 9/2 1/2−[541] state;
the 1028.1 level decays towards the 599.8keV level and to
the 9/2−level at 281keV would then be the 13/2−state
of the decoupled band. This hypothesis is in agreement
with the fact that the 599.8keV level is not directly fed
by the Pb β+/EC decay and with the possible Iπvalues
of the 1028.1 level (see table 2). Anyway, the 599.8keV
level cannot be the 7/2−[514] state because this hypothe-
sis would lead us to attribute Iπ= 9/2−to the 1028.1keV
level, which would be in contradiction with its possible Iπ
values. Furthermore, the K = 1/2 property of the 599.8
and 1028.1keV levels can explain the absence of an ob-
served link between these levels and the parts a and b of
the level scheme. These linking transitions are indeed hin-
dered due to their ΔK ≥ 3 character. The first level which
could be the 17/2−member of the decoupled band is the
one located at 1757.0keV.
To illustrate the quality of the agreement between ex-
perimental and theoretical results in the frame of the pro-
posed state identification, the experimental levels identi-
fied above are compared with the theoretical states pre-
dicted from the190Pb core in fig. 11 for oblate and pro-
late nuclear states. We can see that the agreement is
qualitatively good for the oblate 9/2−[505], 7/2−[514],
13/2+[606] and prolate 11/2−[505] states.
4.2 Low-spin states
From isotope shift measurements the deformation parame-
ters of the 1/2+ground states of193Tl and191Tl have been
determined to be ?β2?1/2= 0.099 and 0.112, respectively.
So, a deformation ?β2?1/2∼ 0.12 is expected for the189Tl
ground state. The comparison of the part d of the level
scheme and the theoretical predictions shown in fig. 10
for |β| ∼ 0.12 suggests that the low-spin part of189Tl
might rather correspond to a prolate nuclear shape. The

Page 15

J. Sauvage et al.: Nuclear structure of189Tl states studied via β+/EC decay and laser spectroscopy of189m+gPb47
Fig. 11. Comparison, in the framework of the proposed identifications, of the experimental levels with the theoretical predictions
obtained using the190Pb core with β = −0.18 and β = −0.15 for the oblate nuclear shape (left panel) and β = +0.12 and
β = +0.15 for the prolate nuclear shape (right panel).
Fig. 12. Levels of the189Tl level scheme for which a structure has been proposed.
experimental level density observed below 1MeV is in-
deed better reproduced assuming a prolate nuclear shape
than assuming an oblate nuclear shape. The 1/2+ground
state would have then as main component of the wave
function, the 1/2+[400] state, and the 318.8, 463.7 and
667.4keV levels would have the 3/2 1/2+[400], 3/2+[402]
and 3/2−[532] states as main components of their wave
functions, respectively. However, given the scarcity of the
experimental information in part d of the level scheme, it
must be considered as tentative.
Finally, to summarize we show in fig. 12 the states
of the189Tl level scheme for which a structure has been
proposed. They are drawn as quasirotational bands la-
belled with the quantum numbers of the main component
of the wave function describing the bandhead and the cor-
responding shape of the core. The theoretical approach
allowed us to propose a structure to 20 levels among the
36 ones established below 1.8MeV.
5 Conclusions
A level scheme of the
lished from the β+/EC decay study of the
mers using both nuclear spectroscopy and in-source laser
spectroscopy experiments. The hyperfine spectra obtained
from gates set on main γ-rays have provided information
on the feeding origin of the levels the γ-rays de-excite.
For example, they clearly indicate that most of the lev-
els shown in the parts a, b and c of the level scheme are
purely fed from the189Pb high-spin isomer. Thus, the hy-
perfine spectra strongly suggest that parts a, b and c of the
189Tl nucleus has been estab-
189Pb iso-