A search for neutrinoless double 13 decay of 48Ca "

Physics Letters B 265 ( 1991 ) 53-56
A search for neutrinoless double 13decay of 48Ca "~
Ke You a, Yucan Zhu a, Junguang Lu a, Hanseng Sun a, Weihua Tian a, Wenheng Zhao a,
Zhipeng Zheng ~.b, Minghan Ye ~.b, Chengrui Ching b.c, Tsohsiu HO ..c, Fengzhu Cui d,
Changjiang Yu d and Guojing Jiang d
Institute oftligh Energy Physics, Academia Sinica, P.O. Box 918, Beijing 100039, China
China Center of Advanced Science and Technology (Worm Laboratoo'), P.O. Box 8730, Beijing 100080, China
Institute o.flheoretical Ph.vsics, Academia Sinica, P.O. Box 2735, Beijing 100080, China
Institute of Optics and Fine Mechanics, Academia Sinica, Changchun, China
Received 10 December 1990; revised manuscript received 5 June 1991
A search for the neutrinoless double [5decay of 4~Ca is carried out in a coal mine near Beijing. Large scintillation crystals of
natural CaF2 were used as both detector and 13source. Results obtained after a total of 7588.5 h of data taking give 9.5 × 102~ yr
(76% confidence level ) as the lower limit ofthe half-life of neutrinoless double 13decay of 48Ca.
1. Introduction
For a long time the nuclear double 13decay without
neutrino emission has attracted considerable attention as a test o f lepton n u m b e r conservation and a
crucial experiment for the study o f the properties o f
neutrinos. Recently several groups have been carrying out double 13 decay experiments on the nuclei
76Go, 828e, l°°Mo, 136Xe, and ~5°Nd.
The decay energy of 4SCa-)48Ti+2e is 4.27 MeV,
much higher than that o f other potential double 13decay nuclei. The large phase space factor therefore is
favorable for c o m p e n s a t i n g the disadvantage o f the
small matrix element. Moreover, the higher decay
energy will help to avoid a low energy background. In
1966, Ter Mateosian and G o l d h a b e r used a 48Ca enriched C a F 2 ( E u ) crystal as scintillator which contained 1 1.4 g of4SCa. A lower limit o f 2 × 102o yr for
the half-life o f neutrinoless double 13 decay o f 4SCa
was obtained [1 ]. The crystals were grown by the
Harshaw Chemical C o m p a n y and the reported energy resolution o f the crystal was rather poor [ 2 ].
Recently, in the course o f studying the scintillation
properties o f several kinds o f crystals, wc found that
Project supported by the National Natural Science Foundation of China.
the unactivated CaF2 crystals grown by the Institute
o f Optics and Fine Mechanics at Chang Chun have a
fairly good energy resolution. When quartz window
photomultipliers were used, the full energy photoelectron peak o f the 661 keV "/ray o f 137Cs was clearly
seen with an energy resolution o f about 7%/x//E
( MeV ) for small CaF2 crystals [ 3 ]. This suggests the
possibility to obtain an i m p r o v e d limit on the halflife for 48Ca neutrinoless double [3 decay by using large
natural unactivated CaF2 crystals as both the 13source
and the detector. Obviously, in comparison with a
"thin source" there is the advantage o f a large quantity o f source nuclei. Though the energy resolution o f
CaF2 crystals is much worse than that o f Ge crystals,
a larger a m o u n t o f 13 source material may somewhat
compensate for this disadvantage.
A laboratory in a coal mine near Beijing has been
set up. The height o f rock above the laboratory is 512
m, equivalent to about 1300 m water. We have run
the experiment for a total of 7588.5 h.
2. Detector
Four cylindrical CaF2 crystals are stacked together.
Each has a cylindrical part about 12.0 cm long and
17.8 cm in d i a m e t e r and a conical part 3.8 cm long
0370-2693/91/$ 03.50 © 1991 Elsevier Science Publishers B.V. All rights reserved.
Volume 265, number 1,2
and 10.0 cm in diameter at the smaller end. The
weight of each crystal is.9716.5 g, 8684.3 g, 10140.1
g, 8828.6 respectively. The total weight is 37369.5 g
which contains 43.0 g of 48Ca. Each crystal is separately canned in an oxygen-free high conductivity
copper can. Specially purified MgO powder (K < 27
ppm, U <0.013 ppm) is used as the reflective layer
between crystal and can. To reduce the possible adsorption of natural radioactive gas inside the can, we
have sealed the crystals in an atmosphere of pure argon. Photomultipliers XP-2041Q with a quartz window are" used to collect the ultra-violet scintillation
light while the natural radioactivity in glass is avoided.
Each crystal is coupled to one photomultiplier,
The energy response of the CaF2 crystals had been
calibrated with gamma radioactive sources such as
2 2 N a , 137Cs, S4Mn as well as electron beams from a
microtron for energies up to 10 MeV [4]. The CaF2
crystals have an identical linear energy response to
both electrons and y rays from 0.5 MeV to 10.0 McV.
Four our large 4aCa crystals, the energy resolution is
better than 10% at an energy E = 4 . 2 7 MeV.
Fig. 1 is the schematic drawing of the detector assembly. The crystals are surrounded by a plastic scintillator of NE110 as the anti-coincidence veto. Limited by funds, we had to use existing plastic
scintillators from our laboratory. The thickness of the
scintillator used for veto is 2.5 cm for the sides and
bottom and 5.0 cm for the top, the efficiency for the
detection of minimum ionization particles is = 100%.
8 August 1991
Steel plates 2.0 cm thick and lead bricks 8-10 cm thick
are used as the hard shielding material.
CAMAC electronics are used in our on-line data
taking system coupled to an IBM P C / X T computer
and the data are registered on disk. A LED light source
is used for the routine calibration of the detectors.
Besides, the background spectrum with the anti-coincidence veto off is recorded at regular time intervals, the 4°K peak is also used for calibration.
3. Experimental results
Initial testing of the detector system had been carried
out in the above-ground laboratory of the Institute of
High Energy Physics, the detector was then moved
into the coal mine cave.
Fig. 2 shows the background spectra observed inside the cave. Spectrum ( 1 ) is obtained without hard
shielding and the anti-coincidence veto, the 4°K and
2°8T1 peaks of natural radioactivity are here seen to
be prominent.
During the first period, the experiment ran with the
first two detectors for 1700.0 h, followed by a run with
all four crystals for 5888.5 h. The counts per channel
O°o °
10 ~- ooo
103 -
,: 10 2 %
% ILl 3
on° (j°og4
10_ ~ - eea
10"2 ~.._
Fig. 1. Schematic drawing of detector assembly. A: CaF: crystal,
B: plastic scintillator, C: steel shielding, D: lead shielding.
Erie r g),(I,/,ev)
Fig. 2. Background energy spectra underground. 1. spectrum
without hard shielding and anti-coincidence veto, 2. spectrum
with hard shielding, 3. spectrum with hard shielding and anticoincidence veto.
Volume 265, number 1,2
around 4.27 MeV for all runs are shown in fig. 3. No
peak is present near 4.27 MeV. The total counts
within an energy interval (4.270+0.214) MeV are
For a rough estimate, the lower limit of the halflife of the neutrinoless 48Ca double 13decay was extracted by using the "sensitive formula" introduced
by Fiorini [ 5 ]
where A is a constant, A = ( 0 7 6 X l n 2 X 6 . 0 2 3 X
1033Xa)/(mX8760), a=0.187% is the abundance
of 4SCa, m = 78 is the molecular weight of CaF2. As
we assume that the peak of electrons has a gaussan
distribution, the area within the energy resolution
should be 0.76. MT is the total weight of all crystals
times the time of data taking. E is the FWHM of the
energy resolution of the detectors in keV, B is the
counting rate in the energy region of interest in
counts/h/keV/g. In our experiment we have
A=8.68X 10 ~4, B = 3 . 4 0 X 10 . 9 counts/h/keV/g,
M T = 2.51 X 108 h g, E=427.0 keV, then
Ti/2 >11.14X 1022 yr.
However, the same data can be treated in another
way. We have fitted the experimental data in a rather
wide energy range from 3.5 MeV to 6.0 MeV using
an exponential function. The z2/DOF(545/315) was
reasonable. From this smooth background shape the
8 August 1991
Table 1
All the experimental results concerning the 4SCa 0v2~ decay.
Experiment Quantity Duration Detector
Wu b~
present exp. 43.0
"~ Ref. [ l l .
CaF2 (Eu)
2 Xl02°
streamer chamber 2 X 102~
9.5 X 10~
h~ Ref. [61.
Table 2
The light neutrino mass (my) and the right handed current mixing parameter ~1.
r/((m,,) = 0 )
this experiment
2 Xl02°
2 ×102~
9.5X 102~
~< 8.3
~<5.09× 10 -S
~< 1.61X 10 -~
~<0.74× 10 -2
expected background counts centered at E = 4 . 2 7
MeV within 10% of the energy resolution are 355.
Therefore, the upper limit should be estimated from
the background statistical error which is quite large.
If 1.18 cr is taken the lower limit of the half-life of
48Ca is obtained to be
Tt/2>9.5× 102~ yr (76%CL) .
In table 1 all the experimental results so far obtained
concerning the 48Ca 0v213decay are summarized.
Using the experimental results, the upper limits for
the light neutrino mass ( m y ) and the right handed
current mixing parameter r/ can be calculated and
these are presented in table 2. The theoretical formulae which we used are taken from refs. [ 7,8].
. . . . . . . .
We are very grateful to the Mentougou Coal Mine
for the generous support which is very crucial to the
running of the experiment. We would thank Professor J.P. Li, J.M. Wu, S.I~. Liu and others of the Radiation Physics Department of the Institute of High
Energy Physics for their valuable help and Professor
T.P. Li, M. Wu for useful discussions.
Fig. 3. Energy spectrum near 4.27 MeV underground.
Volume 265, number 1,2
[ 1 ] E. Ter Mateosian and M. Goldhaber, Phys. Rev. 146 (1966)
[ 2 ] J. Menefee et al., IEEE trans. NS- 13 (1966) 720.
[ 31 Y.C. Zhu et al., Mod. Phys. Lett. A 1 ( 1986 ) 231.
[41W.H. Tian et al., Mod. Phys. Left. A 4 (1989) 213.
8 August 1991
[5] E. Fiorini, Proc. Intern. Symp. on Nuclear 13 decay and
neutrino (Osaka) (World Scientific, Singapore, 1986) p. I 1.
[6] R.K. Bardin, P.J. Gollon, J.D. Ullman and C.S. Wu, Nucl.
Phys. A 158 (1970) 337.
[ 7 ] C.R. Ching, T.H. Ho and X.R. Wu, Phys. Rev. C 40 ( 1989 )
[ 8 ] H.F. Wu et al., Phys. Lett. B 162 ( 1985 ) 227.