> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) <
1
Timing Resolution and Decay Time of LSO
Crystals Co-doped with Calcium
T. Szczęśniak, Member, IEEE, M. Moszyński, Fellow, IEEE, A. Syntfeld-KaŜuch, Member, IEEE,
Ł. Świderski, Member, IEEE, M.A. Spurrier, Member, IEEE, and C.L. Melcher, Senior Member,
IEEE
Abstract—A comparative study of five 5x5x5 mm3 LSO:Ce
samples with different co-dopings of Calcium, from 0 to 0.4% is
reported. Additionally all the results are presented in respect to
selected 10x10x5 mm3 LSO crystal tested in previous studies. The
influence of Ca co-dopand on scintillator parameters was tested
with fast timing photomultiplier Photonis XP20D0. The tests
covered measurements of light output in terms of number of
photoelectrons, measurements of time resolution in coincidence
experiments with annihilation quanta from 22Na gamma source
and decay time constants calculations on the basis of timing
spectra obtained using Thomas-Bollinger single photon method.
Also energy resolution for 662keV from 137Cs gamma source is
mentioned at the end of the paper. The results showed significant
influence of the Ca co-dopand on the properties of the LSO
scintillators. For low Ca concentrations of 0.1% photoelectron
number and energy resolution were improved up to 7800
phe/MeV and 7.3% respectively. On the other hand the shortest
decay time constants were obtained for higher concentrations of
0.3% and 0.4% of Ca giving values equal to about 30 ns. The time
resolution improvement was similar for all the samples and
calculated values for a single detector were equal to about 140ps.
Index Terms—LSO scintillator, Fast timing, Decay time, Timeof-flight PET.
I. INTRODUCTION
T
next step of improving image quality in a Positron
Emission Tomography (PET) is incorporating information
about time of flight (TOF) of annihilation quanta. Such
advancement needs a scintillator optimized for high efficiency
of 511 keV gammas detection together with best possible
timing properties. Very good time resolutions below 200 ps for
a single detector were recorded with 10x10x5 mm3 and
4x4x20 mm3 LSO crystals coupled to a Photonis XP20D0
photomultiplier [1]. Such timing features suggest that LSO
scintillator is a very good candidate for a future TOF-PET [2].
The time resolution of a scintillating crystal depends mainly
on decay time constant of a light pulse and light output and is
HE
Manuscript received June 30, 2009. This work was supported in part by
EU Structural Funds Project no POIG.01.01.02-14-012/08-00 and by the
International Atomic Energy Agency, Research Contract No. 14360.
T. Szczęśniak (e-mail: [email protected]), M. Moszyński, A.
Syntfeld-KaŜuch, and Ł. Świderski, are with The Sołtan Institute for Nuclear
Studies, PL 05-400 Otwock-Świerk, Poland.
M.A. Spurrier (e-mail: [email protected]) and C.L. Melcher are with
University of Tennessee, 301 Science and Engineering Facility, Knoxville,
TN 37996-2000 , USA
proportional to the inverse square root of the number of
photoelectrons collected in the photodetector. In the case of
LSO and other slow decaying scintillators the time resolution
also strongly depends on statistics of the photoelectrons
produced in the decay process of the light pulse [3].
Recently, a new improved LSO, co-doped with Ca was
reported [4]. The modified scintillators exhibit a faster decay
time down to 30 ns and improved light output by about 30%.
The aim of this work was to perform a comparative study of
LSO samples with different co-doping, from 0 to 0.4% of Ca
in a starting raw material. Characterization of co-doped
crystals covered the measurements of the time resolution with
annihilation quanta from 22Na gamma source, the number of
photoelectrons and the decay time constant of a light pulse.
II. EXPERIMENTAL DETAILS
A. Scintillators and Photomultiplier
The studies were carried out on five samples of nonpolished 5x5x5 mm3 LSO crystals doped with 0.1% of Cerium
and with different co-dopings of Calcium. The co-dopand
concentrations were equal to 0.0%, 0.1%, 0.2%, 0.3% and
0.4% of Calcium (in respect to Lutetium) in the melt from
which the crystals were grown. All samples were delivered
from University of Tennessee and grown using Czochralski
technique, at the same growing station, in an atmosphere
composed of nitrogen mixed with a small amount of oxygen.
The Lu2O3, SiO2, CeO2, and CaO starting materials were at
least 99.99% pure. Details can be found in [4].
As a photodetector a Photonis XP20D0 photomultiplier was
used with high blue sensitivity of 13.7 µA/lmF which
corresponds to quantum efficiency of about 34%. This type of
fast PMT is equipped with a screening grid at the anode [5]
and optimized for timing applications. The PMT was tested by
us in previous papers about timing with LSO crystals [1], [3]
Tapered voltage divider was used to assure good linearity of
the anode pulse.
Results obtained with the tested samples were compared
with standard LSO crystal without co-doping. Selected sample
had dimensions of 10x10x5 mm3 and it was polished on all
sides. This selected crystal was put at our disposal by Chuck
Melcher in 2003 to use in a fast timing and is characterized by
high light output and very good time resolution [1].
In all experiments crystals were wrapped with several layers
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) <
2
of white Teflon tape and coupled to a photomultiplier using
silicon grease.
B. Experimental Methods
The number of photoelectrons per energy unit (phe/MeV)
was measured for all the crystals using the Bertolaccini et al
method [6], which compares the peak position from single
photoelectrons (which determines the gain of the
photomultiplier) with the position of the 662 keV full energy
peak from a 137Cs gamma source [7].
The time resolution study was performed in coincidence
experiments with 511 keV annihilation quanta from a 22Na
gamma source. In a reference detector, a BaF2 crystal in shape
of a truncated cone (20 mm and 25 mm in diameter and 15 mm
high) was coupled to the XP20Y0Q/DA PMT. Its time
resolution was equal to 143 ± 4 ps for the 511 keV full energy
peak.
Decay time constant measurements were made by ThomasBollinger single photon method [8], [9], using 137Cs gamma
source. In this case very fast photomultiplier R5320 from
Hamamatsu, characterized by time jitter of 140ps was used as
a single photon detector. Tested crystals were wrapped with
Teflon but only on the sides leaving one surface opened to the
Hamamatsu PMT. Such configuration assured detection of
single photons from scintillator, induced by a gamma source.
C. Electronics
During the measurements of the time resolution and the
decay time, a slow-fast arrangement was used for a precise
selection of the required energy windows. The detailed
diagram of the experimental set-up is presented in Fig. 1. In
the fast signal electronic chain, using signals from the anodes,
the time spectrum of the response difference of the detectors
was recorded. In the slow signal electronic chain, a gate was
generated using dynode signals, to select the energy range of
interest.
In the case of the time resolution measurements the energy
windows were set at 511 keV full energy peaks. For the decay
time constant measurements the energy window was set at the
single photoelectron peak in Hamamatsu R5320 and at the 662
keV full energy peak in XP20D0 coupled with the tested
crystal.
Experiments were performed using a fast leading edge (LE)
discriminator, Polon 1520, and a Constant Fraction
Discriminator (CFD), Ortec 935. The time spectra were
measured with an Ortec 566 Time-to-Amplitude Converter and
recorded by a PC-based multichannel analyzer (Tukan8k) [10].
The time calibration of the Time-to-Amplitude Converter was
done using a precise Time Calibrator, Ortec 462, based on a
quartz clock.
Fig. 1. The slow-fast arrangement for timing measurements. In the fast part,
related to the anode signals, the time spectrum of the response difference of
the detectors is taken. In the slow part, formed using dynode signals, the gate
is generated, to choose the energy range of interest.
III. RESULTS AND DISCUSSION
A. Number of photoelectrons
First the number of photoelectrons was measured by means
of Bertolaccini et al method. Sample spectra of 137Cs gamma
source collected with all the tested crystals are presented in
Fig. 2. The shape of the single photoelectron spectrum is also
showed. The calculated values of photoelectrons per MeV are
collected in Table I together with result for the “selected”
10x10x5 LSO tested in [1].
The lowest photoelectron number was observed for the
standard LSO, without co-doping. This crystal gave only 5700
phe/MeV. Co-doping of 0.1% of Calcium highly improved the
light output leading to 7800 phe/MeV. Unfortunately further
co-doping gave systematically lower values of photoelectron
numbers and finally led to the result below 6000 phe/MeV for
the co-doping of 0.3% of Ca. The dependency of
photoelectron number versus Ca co-doping is presented in Fig.
3. Taking into account the results presented there the lowest
co-doping seems to be optimal, however still the “selected”
LSO without co-dopant shows better performance. It is
obvious that in the case of the “selected” LSO improved
crystal quality improves scintillation performance. On the
other hand the Ca co-dopand leads to similar improvement, so
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) <
the question arise what is easier and cheaper, better growing
techniques or standard techniques with Ca co-dopand.
Fig. 2. The sample spectra of the tested LSO crystals with various Ca codoping collected with 137Cs gamma source.
TABLE I
THE NUMBER OF PHOTOELECTRONS COLLECTED WITH THE TESTED LSO
SAMPLES AND SELECTED 10X10X5 LSO
3
B. Time Resolution
Next the time resolution was measured using experimental
set-up described in Section II. Examples of the timing spectra
are presented in Fig. 4.
Fig. 4. The sample timing spectra of the tested LSO crystals with 0.0% and
0.4% Ca co-doping collected in coincidence experiments with 22Na gamma
source.
The spectrum on the left shows the standard LSO crystal
without co-doping. The measured time resolution was equal to
224 ps what gave 173 ps after subtracting the reference BaF2
detector. This value was the worst measured one during the
study.
On the other hand, the best time resolution was obtained
with the highest co-doping of Calcium of 0.4%. Collected
spectrum is presented on the right in Fig. 4 and shows
coincidence time resolution below 200 ps what corresponds to
136 ps for a single detector.
All the timing results together with data for the selected
10x10x5 mm3 LSO crystal are presented in Table II. In the
second column measured values for two detectors in
coincidence are presented. In the third column the values for a
single crystal are given after subtracting the reference detector.
TABLE II
THE TIME RESOLUTION OF THE TESTED LSO SAMPLES AND SELECTED 10X10X5
LSO
Fig. 3. Dependence of the photoelectron number versus Ca co-doping
measured with the tested samples and selected LSO.
As it can be easily seen in all cases of co-doping the time
resolution is improved. However, the difference between 0.1%
and 0.4% of Calcium is small and within the error range. Plot
of a single detector time resolution versus Ca co-dopand is
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) <
4
presented in Fig. 5. In general, crystals co-doped by Ca led to
about 140 ps of timing resolution. In contrast to measurements
of photoelectron number the time resolution of 166ps for the
selected 10x10x5 LSO was overcome by the tested samples.
Fig. 7. Time spectra of single photon distribution induced by
source in LSO sample co-doped with 0.3% of Calcium.
Fig. 5. Dependence of the single detector time resolution versus Ca codoping measured with the tested samples and selected LSO.
137
Cs gamma
TABLE II
THE TIME RESOLUTION OF THE TESTED LSO SAMPLES AND SELECTED 10X10X5
LSO
In the last column of Table II the measured time resolution
for a single detector is normalized to the number of
photoelectrons for 511keV, presented in 4th column and to the
excess noise factor (ENF = 1.1) calculated using FWHM of
the single photoelectron spectrum. Now, improvement of the
calculated normalized time resolution can be observed for
successive co-doping. Such behavior suggests differences in
the decay time constants of the tested samples, as reported in
[4].
C. Decay Time Constants
Finally the decay time constants were measured using
Thomas-Bollinger single photon method. Examples of the
collected spectra for 0% and 0.3 % of Ca co-dopand together
with fits of single exponential decay are presented in Fig. 6
and 7. All the results are presented in Table III. The measured
quantities agree well with those given in [4].
Fig. 6. Time spectra of single photon distribution induced by
source in LSO sample co-doped with 0.0% of Calcium.
Both crystals without Ca co-doping demonstrate the longest
decay times up to 42.4 ns for the selected LSO and 39.5 ns for
the first LSO sample. The samples with the highest co-doping
of 0.3% and 0.4% showed the shortest decays of around 30 ns.
Dependence of a decay time versus Ca co-doping is presented
in Fig. 8. Successive co-doping of the LSO led to about 25%
improvement of decay time constant in comparison to the
standard LSO characterized by values of 40 ns. However, the
question is it the limit of improvement remains open.
137
Cs gamma
Fig. 8. Dependence of the decay time constants versus Ca co-doping
measured with the tested samples and selected LSO.
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) <
5
In our previous work [3], dependence of time resolution on
decay time was showed for a wide set of scintillators. A similar
plot of a normalized time resolution (from Table II) versus sqrt
of the decay time is presented in Fig. 9. Linear fit with a slope
of 1.4 is also presented.
In the last column of Table III the normalized time
resolution from Table II (last column) was divided by a square
root of the measured decay time constants. The value obtained
for all the tested samples is equal to 1.4 and reflects influence
of the decay time constant on final timing performance of the
detector.
Fig. 10. Dependence of the energy resolution for 662keV from 137Cs gamma
source versus Ca co-doping measured with the tested samples and selected
LSO.
IV. CONCLUSIONS
Fig. 9. Dependence of the time resolution normalized to the number of
photoelectrons versus square root of the measured decay time constant.
Unfortunately the value for the selected LSO is higher and
inconsistent with the data obtained for Ca co-doped samples. It
is hard to explain such behavior but it can be caused by
different conditions of crystal growth. The selected LSO was
produced in 2003 so it is possible that growing conditions
influenced its properties. Moreover an unpolished surface of
the tested co-doped crystals may be taken into account. The
selected LSO crystal was highly polished.
D. Energy Resolution
This paper is focused on timing properties of the LSO codoped with Calcium. However, since energy spectra with 137Cs
gamma source were collected during the study (see Fig. 2) also
values of energy resolution will be presented.
Fig. 10 shows plot of energy resolution for 662keV from
137
Cs gamma source versus Ca co-doping. The observed trend
is similar to the one presented in Fig. 3 in the case of
photoelectron number. For small amount of co-doping a big
improvement is observed from 9.8% to 7.3% but further codoping leads to deterioration of energy resolution up to 8.3%.
It follows well results of the study presented in [11].
Detailed study of non-proportionality, energy resolution and
afterglow of the crystals tested here can be found in [11].
The results presented in this study showed that Calcium codoping in LSO crystals has significant influence on the
properties of these scintillators. The light output is improved
with small amount of Ca co-dopand and above 0.1% it drops
to values similar to standard LSO, however all the tested
samples were less bright than the ‘selected’ LSO from 2003.
The time resolution is improved with Ca co-doping to around
140 ps for a single detector with the best value for 0.4% of Ca.
The decay time constant is improved to around 30 ns for codopings of 0.3% and 0.4% of Ca and no additional
components are observed.
Of course presented study was performed only on one set of
samples and more tests have to be done to fully confirm all the
properties reported here.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
M. Moszyński, M. Kapusta, A. Nassalski, T. Szczęśniak, D. Wolski,
L. Eriksson, C. L. Melcher, “New prospects for Time-of-Flight PET
with LSO scintillators”, IEEE Trans. Nucl. Sci., vol. 53, no. 5, pp.
2484-2488, Oct. 2006.
W. W. Moses, S. E. Derenzo, “Prospects for Time-of-Flight PET using
LSO scintillator”, IEEE Trans. Nucl. Sci., vol. 46, no. 3, pp. 474-478,
Jun 1999.
T. Szczęśniak, M. Moszyński, Ł. Świderski, A. Nassalski, P. Lavoute,
and M. Kapusta, “Fast Photomultipliers for TOF PET”, IEEE Trans.
Nucl. Sci., vol. 56, no 1, pp. 173-181, Feb. 2009.
M. A. Spurrier, P. Szupryczynski, K. Yang, A. A. Carey, and C. L.
Melcher, "Effects of Ca2+ Co-Doping on the Scintillation Properties of
LSO:Ce," IEEE Trans. Nucl. Sci., vol. 55, no. 3, pp. 1178-1182, Jun.
2008.
M. Moszyński, M. Kapusta, D. Wolski, M. Balcerzyk, S.O. Flyckt, P.
Lavoute, C. Marmonier, H. Mach, “New fast photomultipliers with a
screening grid at the anode”, IEEE Trans. Nucl. Sci., vol. 51, no. 4, pp.
1701- 1706, Aug. 2004.
M. Bertolaccini, S. Cova, C. Bussolati, “A technique for absolute
measurements of the effective photoelectron yield in scintillation
counters”, Proc. Nucl. Electr. Symp. Versailles, France, 1968.
M. Moszyński, M. Kapusta, M. Mayhugh, D. Wolski, S.O. Flyckt,
“Absolute light output of scintillators”, IEEE Trans. Nucl. Sci., vol. 44,
no. 3, pp. 1052–1061, Jun. 1997.
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) <
[8]
L.M. Bollinger and G.E. Thomas, “Measurement of the time
dependence of scintillation intensity by a delayed coincidence method”,
Rev. Sci. Instrum., vol. 32, pp. 1044-1050, 1961.
[9] M. Moszyński and B.Bengston, “Light pulse shapes from plastic
scintillators”, Nucl. Instrum. Meth., vol. 142, pp. 417-434, 1977.
[10] Z. Guzik, S. Borsuk, K. Traczyk, M. Płomiński, “TUKAN-an 8K pulse
height analyzer and multi-channel scaler with a PCI or a USB
interface”, IEEE Trans. Nucl. Sci., vol. 53, no. 1, pp. 231–235, Feb.
2006.
[11] A. Syntfeld-KaŜuch, M. Moszyński, Ł. Świderski, T. Szczęśniak, A.
Nassalski, C.L. Melcher, M.A. Spurrier, „Energy Resolution of Calcium
Co-doped LSO:Ce Scintillators”, Nuclear Science Symposium
Conference Record, NSS-MIC 2008, Dresden 19-25 Oct. 2008, pp:
2744-2750, Oct. 2008, IEEE Trans. Nucl. Sci., in press.
6