COSMIC BACKGROUND REDUCTION IN THE RADIOCARBON MEASUREMENTS
BY LIQUID SCINTILLATION SPECTROMETRY AT THE UNDERGROUND
LABORATORY OF GRAN SASSO
Wolfango Plastino1 • Lauri Kaihola2 • Paolo Bartolomei3 • Francesco Bella4
ABSTRACT. Radiocarbon measurements by two 1220 Quantulus™ ultra low background liquid scintillation spectrometers
were performed at the underground laboratory of Gran Sasso and the Radiocarbon Laboratory of E.N.E.A.-Bologna to study
the efficiency and background variations related to measurement sites. The same configuration setup, i.e. the same center of
gravity of the 14C spectrum (SQP(I) = 410 ± 1) was obtained in both instruments. Many different background and modern
standards with pure analytical benzene were used and spectra for 40 one-hour periods were obtained. The data indicates a
background reduction of approximately 65% between the surface and underground laboratories, with no differences in the
efficiency. Recording similar efficiencies in both spectrometers is probably due to fairly identical photomultiplier characteristics. The cosmic noise reduction observed at the laboratory of Gran Sasso makes it possible to perform high precision 14C
measurements and to extend for these idealized samples the present maximum dating limit from 58,000 BP to 62,000 BP
(5 mL, 3 days counting).
INTRODUCTION
The underground laboratory of Gran Sasso is located in the central Apennines (Italy) at 963 m above
the sea level. The maximum thickness of the rock overburden is 1400 m, corresponding to 3800 m
of water. The rock has a density of 2.71 g/cm³, a mean nuclear charge number (Z) equal to 9.4 and
a low rate of natural radioactivity. The underground laboratory is characterized by a cosmic-ray flux
that is extremely reduced compared with the external flux; in fact, only neutrinos and high energy
muons can penetrate through the thick rock. Due to their low interaction probability, incoming neutrinos can filter through the rock independently of their energy, but of the 100 muons per square
meter per second that arrive on the earth’s surface, only those that have a sufficiently high energy
(>1.4 TeV) can reach the underground laboratory. The resulting muon flux in the underground laboratory is 1 per square meter per hour (Bettini 2000).
These characteristics are optimal to investigate the contribution of the environmental radiation to the
residual background signal of the 1220 Quantulus™.
In the first step in our investigations, we performed measurements using idealized samples to study
the efficiency and background variations between the surface laboratory located at E.N.E.A.-Bologna and the underground laboratory of Gran Sasso.
RESULTS
For the background, two sets of three teflon vials with benzene volumes ranging from 1 mL to 5 mL
filled with pure analytical benzene were used. For the modern standards two sets of three teflon vials
with the same benzene volumes filled with pure analytical benzene enriched with radiocarbon to
give the same activity as the standard sucrose ANU (sucrose ANU/modern = 1.0866) were used. The
scintillation cocktail comprised 15 mg butyl-PBD/mL benzene (Gupta and Polach 1985). For each
of these standards, spectra were obtained for 40 one-hour periods.
1 Department
of Physics, University of Roma Tre, via della Vasca Navale, 84, I-00146 Roma, Italy, and I.N.F.N.,
Section of Rome III, via della Vasca Navale, 84, I-00146 Roma, Italy. Email: plastino@fis.uniroma3.it.
2 PerkinElmer Life Sciences, Wallac Oy, P.O.B. 10, FIN-20101 Turku, Finland
3 E.N.E.A., Radiocarbon Laboratory, via dei Colli, 16, I-40136 Bologna, Italy
4 Department of Physics, University of Roma Tre, via della Vasca Navale, 84, I-00146 Roma, Italy
© 2001 by the Arizona Board of Regents on behalf of the University of Arizona
RADIOCARBON, Vol 43, Nr 2A, 2001, p 157–161
Proceedings of the 17th International 14C Conference, edited by I Carmi and E Boaretto
157
158
W Plastino et al.
Cylindrical teflon-S vials designed by ISTA Ltd (Faenza, Italy) with Delrin cap sealed with epoxy
resin have been used. The vials characteristics are: height of 50 mm, external diameter 27 mm, thickness of the bottom teflon base 12 mm, capacity of 9 mL.
Table 1 The count rate of the set A of teflon vials with benzene volumes of 1, 3, and 5 mL related
to surface and underground (italic) laboratories. The labels L and H indicate background and modern standards, respectively.
Sample
Count
rate
(cpm)
Count
error
(cpm)
Modern
activity
(dpm)
Modern
activity
error
(dpm)
Eff
(%)
FM
fM
Tmax
(BP)
L1A
H1A
0.278
12.949
0.022
0.148
8.853
0.088
80.93
25,531.503
17
48,200
L1A
H1A
0.059
12.282
0.010
0.144
8.540
0.094
76.76
99,734.000
35
54,000
L3A
H3A
0.398
39.140
0.026
0.257
27.068
0.161
83.28
17,419.104
43
55,900
L3A
H3A
0.150
38.235
0.016
0.254
26.609
0.166
81.35
44,052.587
69
59,600
L5A
H5A
0.655
65.206
0.033
0.332
45.101
0.209
83.60
10,676.865
57
58,000
L5A
H5A
0.235
63.874
0.020
0.328
44.464
0.215
81.89
28,580.867
92
61,900
Table 2 The count rate of the set B of teflon vials with benzene volumes of 1, 3, and 5 mL related
to surface and underground (italic) laboratories. The labels L and H indicate background and modern standards, respectively.
Sample
Count
rate
(cpm)
Count
error
(cpm)
Modern
activity
(dpm)
Modern
activity
error
(dpm)
Eff
(%)
FM
fM
Tmax
(BP)
L1B
H1B
0.273
12.646
0.021
0.146
8.645
0.087
79.03
22,856.284
17
48,100
L1B
H1B
0.096
12.020
0.013
0.142
8.331
0.090
75.13
58,659.673
27
51,900
L3B
H3B
0.442
38.463
0.027
0.255
25.565
0.159
81.84
15,152.806
40
55,300
L3B
H3B
0.157
38.181
0.016
0.254
26.567
0.166
81.24
42,038.809
67
59,400
L5B
H5B
0.590
64.470
0.32
0.330
44.628
0.208
82.65
11,570.346
59
58,300
L5B
H5B
0.230
64.176
0.020
0.329
44.678
0.216
82.28
29,488.245
94
62,100
Cosmic Background Reduction
159
The count rate of the two sets (A, B) of teflon vials with benzene volumes of 1, 3 and 5 mL related
to surface and underground (italic) laboratories is shown in Tables 1 and 2. The data are related to
optimized soft window in channels 5–540 and PAC equal to 200. Labels L and H indicate background and modern standards, respectively. Also the modern activity, the efficiency (Eff), the figure
of merit (FM), the factor of merit for 14C dating (fM) and the maximum determinable age (Tmax) for
a 2-σ detection criterion and a counting time of 4742 min are shown in Tables 1 and 2 (page 158).
Figure 1 shows the anticoincidence count rate of guard with sample events recorded at surface (a)
and underground (b) laboratories. Also, Figure 2 shows the coincidence count rate of guard with
sample events recorded at surface (a) and underground (b) laboratories.
a)
CPM/Ch
5
4
3
2
1
0
1
101
201
301
401
501
601
701
801
901
1001
Channel
b)
CPM/Ch
5
4
3
2
1
0
1
101
201
301
401
501
601
701
801
901
1001
Channel
Figure 1 The anticoincidence count rate of guard with sample events recorded at the surface (a) and the underground (b) laboratories
DISCUSSION
From the spectra shown in Figure 1, it may be observed that the cosmic muon flux is missing at the
underground laboratory (0.071 cpm) compared to the surface one (414 cpm). The anticoincident
Compton continuum spectrum shows 669 and 161 cpm at the surface and the underground laboratories, respectively. This portion of guard spectrum is related to the flux of the environmental gamma
photons and secondary cosmic particles. The remaining Compton signal in Gran Sasso is due to the
guard phototubes, which are fully embedded, in the scintillating cocktail filling the guard (Figure 3).
The coincidence count rate of the guard with sample events shows at the surface laboratory a cosmic
muon flux of 10.9 cpm and Compton continuum spectrum of 1.06 cpm while at the underground laboratory we have recorded 0.0 cpm and 0.063 cpm, respectively. The coincident guard events are
active background reducing events. The data indicates a background reduction of approximately
65% at the underground laboratory compared to the surface one, with no differences in the efficiency
160
W Plastino et al.
a)
CPM/Ch
2
1
0
1
101
201
301
401
501
601
701
801
901
1001
Channel
b)
CPM/Ch
2
1
0
1
101
201
301
401
501
601
701
801
901
1001
Channel
Figure 2 The coincidence count rate of guard with sample events recorded at the surface (a) and the
underground (b) laboratories. To a better graphical representation the CPM/Ch of surface (a) and underground (b) laboratories are multiplied by factors 10E+01 and 10E+03, respectively.
Copper
Lead
Scintillator
G
G
S
S
Figure 3 Schematic diagram of 1220 Quantulus™ lead shield
assembly and phototubes. The labels S and G indicate sample and
guard phototubes, respectively.
particularly for the 3 and 5 mL benzene volumes of set B. This background reduction results from
the low environmental radiation flux in Gran Sasso.
The vials are fully symmetric with no extra reflectors to cut off the light path from one phototube to
another and no metallic cap on top (Shibata et al. 1983; Polach et al. 1988; Kaihola et al. 1992).
Epoxy resin is used to seal the cap to minimize benzene loss. The surface activator and the epoxy
resin may emit light when hardening with the background signal increasing. High bias threshold was
Cosmic Background Reduction
161
selected in the measurements to cut off low energy emission signal and therefore interference from
epoxy resin does not appear in the results.
CONCLUSION
The first step of measurements carried out in the surface and underground laboratories, with the
1220 Quantulus™, have shown a background reduction of 65% in the underground laboratory.
Thus, the characteristics of the underground laboratory of Gran Sasso make it possible to perform
high precision radiocarbon measurements, to improve fM by 60% and to extend for these idealized
samples the dating limit from 58,000 BP to 62,000 BP (5 mL, 3 days counting). However, experiments on real material have emphasized that the problems in extending the limit of detection back
in time, have much more to do with benzene synthesis and contaminants introduced through lithium,
surface reaction catalysts and impurities than the counter background level (Radnell and Muller
1980; Lowe 1989; McCormac et al. 1993).
Coincidence count rate of guard and sample events is almost negligible in Gran Sasso indicating that
the contribution from the external radiation has vanished. Vial design does not affect the background
in Gran Sasso, but is very important on the surface. The residual background is only dependent on
the sample volume in Gran Sasso, approaching nil at zero sample volume. This means that the sample acts as a target of photons from the small inherent residual radioactivity of sample phototubes.
ACKNOWLEDGMENT
We wish to thank Prof Alessandro Bettini, Director of Underground Laboratory of Gran Sasso, for
his kind collaboration.
REFERENCES
Bettini A. 2000. A vision becomes reality. The Gran
Sasso laboratory 1979–1999. Laboratori Nazionali
del Gran Sasso – I.N.F.N.: 7–33.
Gupta SK, Polach HA. 1985. Radiocarbon dating practices at ANU. Canberra, Radiocarbon Laboratory,
Research School of Pacific Studies, ANU. p 72–5.
Kaihola L, Kojola H, Heinonen H. 1992 A minivial for
small-sample C-14 dating. Radiocarbon 34(3):402–5.
Lowe DC. 1989. Problems associated with the use of coal
as a source of 14C-free background material. Radiocarbon 31(2):117–20.
McCormac FG, Kalin RM, Long A. 1993. Radiocarbon
dating beyond 50,000 years by liquid scintillation
counting. In: Noakes JE, Schönhofer F, Polach HA,
editors. Liquid Scintillation Spectrometry 1992. Tucson: Radiocarbon. p 125–33.
Polach H, Calf G, Harkness D, Hogg A, Kaihola L, Robertson S. 1988. Performance of new technology liquid
scintillation counters for 14C dating. Nuclear Geophysics 2(2):75–9.
Radnell CJ, Muller AB. 1980. Memory effects in the production of benzene for radiocarbon dating, Radiocarbon 22(2):479–86.
Shibata S, Kawano E, Nakabayashi T, Kawamura S.
1983. The liquid scintillation counting of low-level tritium activity with teflon vial. Annual Report of the Radiation Center of Osaka Prefecture 24:67–70.