- Email: [email protected]
Light conversion efficiency of small lithium scintillators for electrons, protons, deuterons and alpha particles
Nuclear Instruments and Methods in Physics Research A254 (1987) 361-366 North-Holland, Amsterdam
361
LIGHT CONVERSION EFFICIENCY OF SMALL LITHIUM SCINTILLATORS FOR ELECTRONS, PROTONS, DEUTERONS AND ALPHA PARTICLES A.W. D A L T O N Australian Atomic Energy Commission Research Establishment, Lucas Heights Research Laboratories, PMB Sutherland, N S W 2232 Australia
Received 4 July 1986
The absolute reponses of thin, cerium-activated lithium glass scintillators to electrons, protons, deuterons and alpha particles have been measured as a function of their kinetic energy (0.3-2.7 MeV), lithium concentration (2.4-8.3 wt.%) and lithium enrichment (95 wt.% 6Li to 99.99 wt.% 7Li) in the glass. Absolute calibration of the scintillation energy was derived from measurements of the energy resolution of the observed pulse height spectrum.
1. Introduction Cerium-activated lithium glass scintillators have been widely used for the detection of low energy neutrons since they were first reported [1]. The mechanism involves the nuclear reactions 6Li + n = T + 4 H e + 4.8 MeV.
(1)
7Li + n = T + 4 H e + n' - 2.82 MeV.
(2)
The E N D F / B - V cross sections of these reactions are shown in fig. 1 as a function of the incident neutron energy. Within the glass, a fraction of the combined kinetic energy of the emitted charged particles is converted into a pulse of light. A photomuitiplier (PM) coupled to the glass converts the individual scintillations into electrical pulses which are then amplified and displayed on the screen of a multichannel analyser (MCA). In the investigation of tritium breeding in fusion reactor blankets there is a need to measure the production of tritium over a wide range of neutron energies (thermal to 15 MeV). As one tritium atom is produced in each of the reactions (1) and (2), one light signal will accompany each tritium atom produced. On the basis of the cross section data given in fig. 1, it is possible that a combination of lithium glasses enriched in 6Li and 7Li could be used to measure tritium production over the entire energy range of interest in fusion blanket studies. However, signals can also arise from gamma background and from charged particles released in other neutron-induced nuclear reactions within the glass. The significance of the latter depends on the incident neu0168-9002/87/$03.50 © Elsevier Science Publishers B.V. (North-Holland Physics Publishing Division)
tron energy, the reaction Q-value and the efficiency with which the kinetic energy of the emitted charged pa/'ticles is converted into fight within the volume of the glass.
This paper reports on the experimental and analytical methods used to determine the scintillation responses of small lithium glasses as a function of lithium concentration, and enrichment for gammas, protons, deuterons and alpha particles over an energy range 0.3-2.7 MeV. These measurements were undertaken to obtain some of the data required to isolate the signals associated with the production of tritium from the total light output, and to facilitate the selection of the most suitable glasses for tritium production measurements in future fusion rector blanket experiments. The results are compared with data previously reported in the literature. l
I
I
2
~
6
I
I
8
10
~.00
I ~Li(n,n~)T
I
E - - 300
o~
W-
gQ:
200
100
12
14
ENERSY {MeV)
Fig. 1. Variation of tritium production cross sections.
A. W. Dalton / Light conversion efficiency of small Li scintillators
362
/ •
2. Experimental details
Photomultiptiec tube
2.1. Procedure The glass scintillators used in this investigation (table 1) represent the full range of lithium concentrations and enrichments supplied by Nuclear Enterprises Ltd., UK; each scintillator contained about 4 wt.% of Ce20 3. All were in the form of right cylindrical sections, 12.7 mm diameter by 3 mm thick, with both faces ground flat and one face polished. Small detectors were selected to reduce the contribution from gamma interactions. For the dimensions u~ed, the maximum energy that could be deposited in the glass by a gamma was about 1.2 MeV. The scintillators were irradiated with gammas, protons, deuterons and alpha particles. Electrons with maximum energies of 0.34, 0.48, 0.64, 1.06 and 2.53 MeV were produced in the glasses by Compton interaction of the gammas emitted by a set of calibrated sources which included 22Na, Z4Na, SaMn and 137Cs; charged particles with energies of 0.8, 1.2, 1.7, 2.2 and 2.7 MeV were obtained from a 3. MeV Van de Graaff accelerator using a thin, self-supporting gold foil to backscatter particles from the beam into the scintillator. Because the range of the charged particles in solid materials is very small, they had to enter the glass scintillators via the vacuum system of the beam tube unimpeded by a window of any kind. The glasses were mounted over the collimator hole in one of the scattering ports of the target chamber (fig. 2), and vacuumsealed by an O-ring concentric to the collimator. The window of the Philips type 1911 PM tube was then sealed to a glass scintillator using high vacuum silicone grease. The PM tube had a 10-stage linear focus tube with CuBe secondary emitting surfaces, a flat end window and a 14 mm semitransparent photocathode ($11 spec-
NE901 NE902 NE903 NE904 NE905 NE906 NE907 NE908 NE909 NE912 NE913
l-[ Beam tube
~ Scatteringteg
"N~~ N
< Beam
Photomuttiptier tube
/
Fig. 2. Plan view of target chamber. tral response). The PM was coupled to a locally constructed, high stability dynode chain which operated at an overall voltage of - 1100 V supplied by an Ortec 556 high voltage unit. At this voltage, the PM has a quantum efficiency of 15%, an anode spectral sensitivity of 17 m A / W , a gain of 3 X 10 5 and a dark current of less than 0.6 hA. To screen against the influence of stray magnetic fields, the PM tube was protected by a mumetal shield. The signal from the PM was fed via an AAEC preamplifier (time constant 50 gs), a 2020 spectroscopic amplifier (time constant 500 ns) and an 8075 analogto-digital converter to a series 40 MCA system (all from Canberra Instruments). Voltage signals from a Berkeley Nucleonics Corporation Model PB-4 precision pulse generator were used to calibrate the scale of the MCA. For all measurements, a 1023 channel display was used; this allowed up to four pulse height spectra to be stored and inspected at any time during the measurements. For detailed analysis, the data were transferred to a SIRIUS microcomputer; software written by the author was used to process to archival data. 2.2. Measurement of the scintillation efficiencies
Table 1 Details of hthium glass scintillators Glass type
Scott~
Lithium content Total lithium [wt.%]
7Li/Li [wt.%]
2.40 2.26 2.41 6.60 6.23 6.63 7.50 7.08 7.54 7.80 8.30
natural 5.00 99.99 natural 5.00 99.99 natural 5.00 99.00 5.00 99.00
The relative efficiencies with which the glass scintillators converted the energy of the charged particles into light energy were determined by locating the channel in which the maximum light intensity occurred. For the heavy ions (protons, deuterons and alpha particles) this was straightforward because the spectra were Gaussian in shape. For the non-Gaussian energy distribution of the Compton electrons, the channel number corresponding to the Compton edge was taken as that at which the intensity fell to half the measured peak intensity [2-4]. Absolute values of the scintillation efficiencies were obtained from these data using a method originally reported in ref. [5] and more recently in ref. [6]. This involved measurement of the energy resolution, ~1, of a
363
A. W. Dalton / Light conversion efficiency of small Li scintillators
scintillation pulse height spectrum which has a Gaussian distribution. Breitenberger [7] has shown that for such a distribution the scintillation efficiency, S, is related to the measured spectral resolution by S=
1
R
5.545 Ep
GCg R - 1
~12
(3)
Ei '
where G is the fraction of photons produced in the glass which impinges on the cathode of the PM; C is the efficiency with which the photons are converted to electrons at the photocathode; g is the electron-optical focusing efficiency; R is the mean stage gain for n successive dynodes; Ep is the characteristic photon emission energy from the cerium activator (3.14 eV) and E i is the kinetic energy deposited in the glass during the passage of the charged particle. Of these parameters C, g and R were determined from the tube specifications and the operational voltage of the dynode chain and are given in sect. 2.1. The value of G was determined from the gamma responses in the NE907 glass both with and without reflecting alpha-alumina coated surfaces.
3. Results and discussion
Table 2 Response of lithium glass scintillators to electrons Glass type
Scintillation efficiency
NE901 NE902 NE903 NE904 NE905 NE906 NE907 NE908 NE909 NE912 NE913
[eV/photon]
[%]
165 145 164 161 206 218 265 312 290 307 384
1.90 2.17 1.91 1.95 1.52 1.44 1.18 1.01 1.08 1.02 0.82
the shape of their responses versus energy was the same for all glasses; the mean normalised curves are illustrated in fig. 4. This normalisation indicated that the responses for all three particles conformed to the same general shape which is approximated by a least-squares binomial fit of the form L, = 33.5 E,(1 + 0.35 E i ) / ( k r ) ,
3.1. Variation o f scintillation response with particle energy
The relative variation of light output with the electron energy absorbed in the NE902 scintillator, which is typical of the responses obtained for all glasses, is shown in fig. 3. Because these relationships are linear, it is possible to determine an energy-independent scintillation efficiency for each of the glasses from a linear least-squares fit to their measured responses; the results are listed in table 2, and fitting errors were less than 1%. For the heavier charged particles, the light output was found to increase nonlinearly with increase in their energy, the relative variation depending on both glass composition and particle type. Normalisation of the measured data for each charged particle showed that
(4)
where k and r are the normalisation factors for particle and glass type, respectively. Values of k and r are listed in table 3 for all glasses; the response of the NEg02 glass to protons is shown in fig. 5 (estimated errors are about 8%). From table 3 it can be seen that the responses varied inversely with the ionisation density of the particle. This effect, which is observed in all types of scintillators [8], is attributed to "ionisation quenching" which results from the higher specific energy losses for heavier particles and higher electronic charges. The nonlinear variation with energy observed in our investigation contrasts with the linear dependence assumed for the ex-
1050
Key to S~mbois
I
I
I
I
2/_
;;% :___-_
600 500
60O
IK 3oo
~50
--
--
.
s" . S ~ /
100 0
J 0
I 0.2
t
I
0.t~
0b
I Err0~s'7 0.8
1s o
1
IONISINfi ENERfY (MeVl Fig. 3. V a r i a t i o n o f l i g h t o u t p u t w i t h e l e c t r o n energy.
1.2
0.5
1
1.5 2 PARTICLEENERGY[HeV)
25
3
Fig. 4. M e a n c h a r g e d p a r t i c l e responses o f all l i t h i u m glasses.
A. W. Dalton / Light conversion efficiency of small Li scintillators
364
trapolation of response data in several previous investigations [6,9,10]; this is discussed in sect. 3.3 below.
Table 3 Normalisation factors for scintillation responses Glass type
Normalisation factors Electrons
Protons
Deuterons
Alphas
NE901 NE902 NE903 NE904 NE905 NE906 NE907 NE908 NE909 NE912 NE913
1.14 1.00 1.13 1.11 1.42 1.50 1.73 1.32 2.00 2.12 2.65
1.10 1.00 1.07 1.14 1.41 1.49 1.72 1.42 .1.99 2.13 2.39
1.16 1.00 1.21 1.11 1.38 1.44 1.83 1.34 1.85 2.20 2.51
1.12 1.00 1.13 1.08 1.53 1.56 2.10 1.42 2.50 2.55 2.94
Particle
1.00
2.10
2.80
9.50
210
I
I
I
I
3.2. Variation of scintillation response with lithium content The relative variations of scintillation responses of the glasses with lithium content are shown in fig. 7 for electrons, and fig. 6 for the heavier charged particles. The present results indicate that for each type of charged particle the amplitude of the light output decreases with increase in lithium content. Within the statistical errors of the measurements, the variations of the normalised scintillation responses are the same for aU particles, including electrons. Although significant differences were observed between glasses containing different lithium enrichments (but with the same total lithium content) no systematic variation in light output with lithium enrichment could be established. To convert the relative response measurements to the absolute values given in table 2 and figs. 5, 6 and 7, the channel scale of the MCA display was calibrated in terms of light energy using the measured peak and resolution of the pulse height spectrum produced by 2.7 MeV protons in the NE902 glass. The operational characteristics of the detector system and the transmission losses from the glass to the photomultiplier, which are also required for the calibration, are reported in ref. [11].
I
150 >
12o ~
90
60
3.3. Resume of previous investigations
30
0 /
i
I
05
1
[
I
i
1.5 2 PARTICLE ENERGY(HeV)
25
Although lithium glass scintillators have been in use. for more than 25 years, little attempt has been made to determine the efficiencies with which they convert the
Fig. 5. Absolute response of NE902 glass to protons.
120
--
I
I
I
I
I
I
I
100
n
~
ao
~
60
~ _z
20
I
I
I
[
I
I
1
1
2
3
6
S
6
7
LITHIUM CONTENT(wt %1 Fig. 6. Variation of light output with lithium content (protons).
,4.14/. Dalton / Light conversion efficiency of small Li scintillators I
I
t
3OO
> ~
2OO
z
100
o 00
,
I
2.0
,
I
,
60
I
6.0
L
I
8.0
CONCENTRATION (wt %1
Fig. 7. Variation of light output with lithium content (electrons).
energy of individual charged particles into scintillation energy. However, several measurements of lithium glass responses to monoenergetic alpha particle sources have been reported [6,9,10]. Relative to electrons of the same energy, the scintillations produced by the 5.15 MeV alpha particles from a 239pu source were found to be smaller by a factor of 4.3 [9] and 5.1 [10]. For a range of glasses similar to that used here, the mean response to 5.48 MeV 2 4 1 A n l alpha particles was found to be less than that for electrons by a factor of 12.5 [6]; a mean ratio of about 10 was observed in the present investigations (table 3). Many investigations [5,6,9-15] have shown that the scintillation produced in the glass by the combined kinetic energy of the triton (2.73 MeV) and alpha particle (2.05 MeV), emitted in the 6Li(n,a)T reaction, is equivalent to that of an electron of about 1.5-1.6 MeV. From linear extrapolations of their alpha particle measurements [6,9,10] to the energy of the alphas in the latter reaction, the scintillation response of the 2.73 MeV triton was estimated to be less than that of electrons by factors of 2, 2.4 and 2.3, respectively. Absolute responses to thermal neutrons have been determined [6,15] for a range of lithium glasses similar to that used here. Combining these data with those given above provides estimates of the absolute scintillation efficiencies for electrons ranging from a maximum of 1.53% (NE902) to a minimum of 0.27% (NE912) and for alpha particles from 0.15% to 0.02%, respectively. No other data, relative or absolute, were found in the literature relating the light responses of the glasses, to electrons, protons, deuterons, tritons or alpha partitles, with particle kinetic energy.
365
4. Conclusions The light signals produced in glass scintillators by energetic charged particles have been shown to vary with the lithium content of the glass and the energy, mass and electronic charge of the detected particle. The relationship between the amphtude of the light signals and the kinetic energy transferred to the glass by electrons is linear whereas for the much heavier protons, deuterons and alphas it is nonlinear; the shapes of the nonlinear curves are the same for all of the heavy particles and for all the glasses studied. For all the charged particles studied, the light output of the scintillators decreased with increase in the lithium content in the glass; within the statistical error of the measurements the shape of this relationship appeared to be the same for all particles including electrons. For all glasses, the spread in the normalisation factors for the different charged particles increased with the increase in lithium content. The scintillation efficiency of all the glasses was greatest for electrons and least for alpha particles. For kinetic energy transfers of 1 MeV, the response of the glasses to electrons exceeded that due to protons, deuterons and alpha particles by factors 2.1, 2.8 and 9.5, respectively.
Acknowledgements
Thanks are due to Mr. R.J. Blevins for assistance with the experimental work, to Mr. W.J. Crawford for the design and construction of the target chamber, and to Dr. J.H. Boldeman and Mr. J.P. Fallon for help in operating the Van de Graaff accelerator.
References
[1] R.J. Ginther, IEEE Trans. Nucl. Sci. NS-7 (1960) 28. [2] L.E. Flynn, L.E. Glendenin, E.P. Steinbert and P.M. Wright, Nucl. Instr. and Meth. 27 (1964) 13. [3] V.V. Verbinski, W.R. Burrus, T.A. Love, W. Zobel, N.W. Hill and R. Textor, Nuct. Instr. and Meth. 65 (1968) 8. [4] H.H. Knox and T.G. Miller, Nucl. Instr. and Meth. 101 (1972) 519. [5] L.M. Bollinger, G.E. Thomas and R.J. Ginther, Nucl. Instr. and Meth. 17 (1962) 97. [6] A.R. Spowart, Nucl. Instr. and Meth. 135 (1976) 441. [7] E. Breitenberger, Progress in Nuclear Physics, ed., O. Prisch (Pergamon, London and New York, 1955). [8] J.B. Birks, The Theory and Practice of Scintillation Counting (Pergamon, London, 1964). [9] D.G. Anderson, J. Dracass, T.P. Hanagan and E.H. Noe, Proe. Syrup. on Photoelectronic Image Devices, Imperial College, London, England (1961) p. 429. [10] F.W.K. Firk, G.G. Slaughter and R.J. Ginther, Nucl. Instr. and Meth. 13 (1961) 313.
366
A. W. Dalton / Light conversion efficiency of small Li scintillators
[11] A.W. Dalton, AAEC E-Report, in preparation. [12] L.A. Wraight, D.H.C. Harris and P.A.Egelstaff, Nucl. Instr. and Meth. 33 (1965) 181. [13] A.R. Spowart, Nucl. Instr. and Meth. 82 (1970) 1.
[14] G.L. Jensen and J.B. Czirr, Nucl. Instr. and Meth. 205 (1983) 461. [15] A.R. Spowart, Nucl. Instr. and Meth. 75 (1969) 35.
361
LIGHT CONVERSION EFFICIENCY OF SMALL LITHIUM SCINTILLATORS FOR ELECTRONS, PROTONS, DEUTERONS AND ALPHA PARTICLES A.W. D A L T O N Australian Atomic Energy Commission Research Establishment, Lucas Heights Research Laboratories, PMB Sutherland, N S W 2232 Australia
Received 4 July 1986
The absolute reponses of thin, cerium-activated lithium glass scintillators to electrons, protons, deuterons and alpha particles have been measured as a function of their kinetic energy (0.3-2.7 MeV), lithium concentration (2.4-8.3 wt.%) and lithium enrichment (95 wt.% 6Li to 99.99 wt.% 7Li) in the glass. Absolute calibration of the scintillation energy was derived from measurements of the energy resolution of the observed pulse height spectrum.
1. Introduction Cerium-activated lithium glass scintillators have been widely used for the detection of low energy neutrons since they were first reported [1]. The mechanism involves the nuclear reactions 6Li + n = T + 4 H e + 4.8 MeV.
(1)
7Li + n = T + 4 H e + n' - 2.82 MeV.
(2)
The E N D F / B - V cross sections of these reactions are shown in fig. 1 as a function of the incident neutron energy. Within the glass, a fraction of the combined kinetic energy of the emitted charged particles is converted into a pulse of light. A photomuitiplier (PM) coupled to the glass converts the individual scintillations into electrical pulses which are then amplified and displayed on the screen of a multichannel analyser (MCA). In the investigation of tritium breeding in fusion reactor blankets there is a need to measure the production of tritium over a wide range of neutron energies (thermal to 15 MeV). As one tritium atom is produced in each of the reactions (1) and (2), one light signal will accompany each tritium atom produced. On the basis of the cross section data given in fig. 1, it is possible that a combination of lithium glasses enriched in 6Li and 7Li could be used to measure tritium production over the entire energy range of interest in fusion blanket studies. However, signals can also arise from gamma background and from charged particles released in other neutron-induced nuclear reactions within the glass. The significance of the latter depends on the incident neu0168-9002/87/$03.50 © Elsevier Science Publishers B.V. (North-Holland Physics Publishing Division)
tron energy, the reaction Q-value and the efficiency with which the kinetic energy of the emitted charged pa/'ticles is converted into fight within the volume of the glass.
This paper reports on the experimental and analytical methods used to determine the scintillation responses of small lithium glasses as a function of lithium concentration, and enrichment for gammas, protons, deuterons and alpha particles over an energy range 0.3-2.7 MeV. These measurements were undertaken to obtain some of the data required to isolate the signals associated with the production of tritium from the total light output, and to facilitate the selection of the most suitable glasses for tritium production measurements in future fusion rector blanket experiments. The results are compared with data previously reported in the literature. l
I
I
2
~
6
I
I
8
10
~.00
I ~Li(n,n~)T
I
E - - 300
o~
W-
gQ:
200
100
12
14
ENERSY {MeV)
Fig. 1. Variation of tritium production cross sections.
A. W. Dalton / Light conversion efficiency of small Li scintillators
362
/ •
2. Experimental details
Photomultiptiec tube
2.1. Procedure The glass scintillators used in this investigation (table 1) represent the full range of lithium concentrations and enrichments supplied by Nuclear Enterprises Ltd., UK; each scintillator contained about 4 wt.% of Ce20 3. All were in the form of right cylindrical sections, 12.7 mm diameter by 3 mm thick, with both faces ground flat and one face polished. Small detectors were selected to reduce the contribution from gamma interactions. For the dimensions u~ed, the maximum energy that could be deposited in the glass by a gamma was about 1.2 MeV. The scintillators were irradiated with gammas, protons, deuterons and alpha particles. Electrons with maximum energies of 0.34, 0.48, 0.64, 1.06 and 2.53 MeV were produced in the glasses by Compton interaction of the gammas emitted by a set of calibrated sources which included 22Na, Z4Na, SaMn and 137Cs; charged particles with energies of 0.8, 1.2, 1.7, 2.2 and 2.7 MeV were obtained from a 3. MeV Van de Graaff accelerator using a thin, self-supporting gold foil to backscatter particles from the beam into the scintillator. Because the range of the charged particles in solid materials is very small, they had to enter the glass scintillators via the vacuum system of the beam tube unimpeded by a window of any kind. The glasses were mounted over the collimator hole in one of the scattering ports of the target chamber (fig. 2), and vacuumsealed by an O-ring concentric to the collimator. The window of the Philips type 1911 PM tube was then sealed to a glass scintillator using high vacuum silicone grease. The PM tube had a 10-stage linear focus tube with CuBe secondary emitting surfaces, a flat end window and a 14 mm semitransparent photocathode ($11 spec-
NE901 NE902 NE903 NE904 NE905 NE906 NE907 NE908 NE909 NE912 NE913
l-[ Beam tube
~ Scatteringteg
"N~~ N
< Beam
Photomuttiptier tube
/
Fig. 2. Plan view of target chamber. tral response). The PM was coupled to a locally constructed, high stability dynode chain which operated at an overall voltage of - 1100 V supplied by an Ortec 556 high voltage unit. At this voltage, the PM has a quantum efficiency of 15%, an anode spectral sensitivity of 17 m A / W , a gain of 3 X 10 5 and a dark current of less than 0.6 hA. To screen against the influence of stray magnetic fields, the PM tube was protected by a mumetal shield. The signal from the PM was fed via an AAEC preamplifier (time constant 50 gs), a 2020 spectroscopic amplifier (time constant 500 ns) and an 8075 analogto-digital converter to a series 40 MCA system (all from Canberra Instruments). Voltage signals from a Berkeley Nucleonics Corporation Model PB-4 precision pulse generator were used to calibrate the scale of the MCA. For all measurements, a 1023 channel display was used; this allowed up to four pulse height spectra to be stored and inspected at any time during the measurements. For detailed analysis, the data were transferred to a SIRIUS microcomputer; software written by the author was used to process to archival data. 2.2. Measurement of the scintillation efficiencies
Table 1 Details of hthium glass scintillators Glass type
Scott~
Lithium content Total lithium [wt.%]
7Li/Li [wt.%]
2.40 2.26 2.41 6.60 6.23 6.63 7.50 7.08 7.54 7.80 8.30
natural 5.00 99.99 natural 5.00 99.99 natural 5.00 99.00 5.00 99.00
The relative efficiencies with which the glass scintillators converted the energy of the charged particles into light energy were determined by locating the channel in which the maximum light intensity occurred. For the heavy ions (protons, deuterons and alpha particles) this was straightforward because the spectra were Gaussian in shape. For the non-Gaussian energy distribution of the Compton electrons, the channel number corresponding to the Compton edge was taken as that at which the intensity fell to half the measured peak intensity [2-4]. Absolute values of the scintillation efficiencies were obtained from these data using a method originally reported in ref. [5] and more recently in ref. [6]. This involved measurement of the energy resolution, ~1, of a
363
A. W. Dalton / Light conversion efficiency of small Li scintillators
scintillation pulse height spectrum which has a Gaussian distribution. Breitenberger [7] has shown that for such a distribution the scintillation efficiency, S, is related to the measured spectral resolution by S=
1
R
5.545 Ep
GCg R - 1
~12
(3)
Ei '
where G is the fraction of photons produced in the glass which impinges on the cathode of the PM; C is the efficiency with which the photons are converted to electrons at the photocathode; g is the electron-optical focusing efficiency; R is the mean stage gain for n successive dynodes; Ep is the characteristic photon emission energy from the cerium activator (3.14 eV) and E i is the kinetic energy deposited in the glass during the passage of the charged particle. Of these parameters C, g and R were determined from the tube specifications and the operational voltage of the dynode chain and are given in sect. 2.1. The value of G was determined from the gamma responses in the NE907 glass both with and without reflecting alpha-alumina coated surfaces.
3. Results and discussion
Table 2 Response of lithium glass scintillators to electrons Glass type
Scintillation efficiency
NE901 NE902 NE903 NE904 NE905 NE906 NE907 NE908 NE909 NE912 NE913
[eV/photon]
[%]
165 145 164 161 206 218 265 312 290 307 384
1.90 2.17 1.91 1.95 1.52 1.44 1.18 1.01 1.08 1.02 0.82
the shape of their responses versus energy was the same for all glasses; the mean normalised curves are illustrated in fig. 4. This normalisation indicated that the responses for all three particles conformed to the same general shape which is approximated by a least-squares binomial fit of the form L, = 33.5 E,(1 + 0.35 E i ) / ( k r ) ,
3.1. Variation o f scintillation response with particle energy
The relative variation of light output with the electron energy absorbed in the NE902 scintillator, which is typical of the responses obtained for all glasses, is shown in fig. 3. Because these relationships are linear, it is possible to determine an energy-independent scintillation efficiency for each of the glasses from a linear least-squares fit to their measured responses; the results are listed in table 2, and fitting errors were less than 1%. For the heavier charged particles, the light output was found to increase nonlinearly with increase in their energy, the relative variation depending on both glass composition and particle type. Normalisation of the measured data for each charged particle showed that
(4)
where k and r are the normalisation factors for particle and glass type, respectively. Values of k and r are listed in table 3 for all glasses; the response of the NEg02 glass to protons is shown in fig. 5 (estimated errors are about 8%). From table 3 it can be seen that the responses varied inversely with the ionisation density of the particle. This effect, which is observed in all types of scintillators [8], is attributed to "ionisation quenching" which results from the higher specific energy losses for heavier particles and higher electronic charges. The nonlinear variation with energy observed in our investigation contrasts with the linear dependence assumed for the ex-
1050
Key to S~mbois
I
I
I
I
2/_
;;% :___-_
600 500
60O
IK 3oo
~50
--
--
.
s" . S ~ /
100 0
J 0
I 0.2
t
I
0.t~
0b
I Err0~s'7 0.8
1s o
1
IONISINfi ENERfY (MeVl Fig. 3. V a r i a t i o n o f l i g h t o u t p u t w i t h e l e c t r o n energy.
1.2
0.5
1
1.5 2 PARTICLEENERGY[HeV)
25
3
Fig. 4. M e a n c h a r g e d p a r t i c l e responses o f all l i t h i u m glasses.
A. W. Dalton / Light conversion efficiency of small Li scintillators
364
trapolation of response data in several previous investigations [6,9,10]; this is discussed in sect. 3.3 below.
Table 3 Normalisation factors for scintillation responses Glass type
Normalisation factors Electrons
Protons
Deuterons
Alphas
NE901 NE902 NE903 NE904 NE905 NE906 NE907 NE908 NE909 NE912 NE913
1.14 1.00 1.13 1.11 1.42 1.50 1.73 1.32 2.00 2.12 2.65
1.10 1.00 1.07 1.14 1.41 1.49 1.72 1.42 .1.99 2.13 2.39
1.16 1.00 1.21 1.11 1.38 1.44 1.83 1.34 1.85 2.20 2.51
1.12 1.00 1.13 1.08 1.53 1.56 2.10 1.42 2.50 2.55 2.94
Particle
1.00
2.10
2.80
9.50
210
I
I
I
I
3.2. Variation of scintillation response with lithium content The relative variations of scintillation responses of the glasses with lithium content are shown in fig. 7 for electrons, and fig. 6 for the heavier charged particles. The present results indicate that for each type of charged particle the amplitude of the light output decreases with increase in lithium content. Within the statistical errors of the measurements, the variations of the normalised scintillation responses are the same for aU particles, including electrons. Although significant differences were observed between glasses containing different lithium enrichments (but with the same total lithium content) no systematic variation in light output with lithium enrichment could be established. To convert the relative response measurements to the absolute values given in table 2 and figs. 5, 6 and 7, the channel scale of the MCA display was calibrated in terms of light energy using the measured peak and resolution of the pulse height spectrum produced by 2.7 MeV protons in the NE902 glass. The operational characteristics of the detector system and the transmission losses from the glass to the photomultiplier, which are also required for the calibration, are reported in ref. [11].
I
150 >
12o ~
90
60
3.3. Resume of previous investigations
30
0 /
i
I
05
1
[
I
i
1.5 2 PARTICLE ENERGY(HeV)
25
Although lithium glass scintillators have been in use. for more than 25 years, little attempt has been made to determine the efficiencies with which they convert the
Fig. 5. Absolute response of NE902 glass to protons.
120
--
I
I
I
I
I
I
I
100
n
~
ao
~
60
~ _z
20
I
I
I
[
I
I
1
1
2
3
6
S
6
7
LITHIUM CONTENT(wt %1 Fig. 6. Variation of light output with lithium content (protons).
,4.14/. Dalton / Light conversion efficiency of small Li scintillators I
I
t
3OO
> ~
2OO
z
100
o 00
,
I
2.0
,
I
,
60
I
6.0
L
I
8.0
CONCENTRATION (wt %1
Fig. 7. Variation of light output with lithium content (electrons).
energy of individual charged particles into scintillation energy. However, several measurements of lithium glass responses to monoenergetic alpha particle sources have been reported [6,9,10]. Relative to electrons of the same energy, the scintillations produced by the 5.15 MeV alpha particles from a 239pu source were found to be smaller by a factor of 4.3 [9] and 5.1 [10]. For a range of glasses similar to that used here, the mean response to 5.48 MeV 2 4 1 A n l alpha particles was found to be less than that for electrons by a factor of 12.5 [6]; a mean ratio of about 10 was observed in the present investigations (table 3). Many investigations [5,6,9-15] have shown that the scintillation produced in the glass by the combined kinetic energy of the triton (2.73 MeV) and alpha particle (2.05 MeV), emitted in the 6Li(n,a)T reaction, is equivalent to that of an electron of about 1.5-1.6 MeV. From linear extrapolations of their alpha particle measurements [6,9,10] to the energy of the alphas in the latter reaction, the scintillation response of the 2.73 MeV triton was estimated to be less than that of electrons by factors of 2, 2.4 and 2.3, respectively. Absolute responses to thermal neutrons have been determined [6,15] for a range of lithium glasses similar to that used here. Combining these data with those given above provides estimates of the absolute scintillation efficiencies for electrons ranging from a maximum of 1.53% (NE902) to a minimum of 0.27% (NE912) and for alpha particles from 0.15% to 0.02%, respectively. No other data, relative or absolute, were found in the literature relating the light responses of the glasses, to electrons, protons, deuterons, tritons or alpha partitles, with particle kinetic energy.
365
4. Conclusions The light signals produced in glass scintillators by energetic charged particles have been shown to vary with the lithium content of the glass and the energy, mass and electronic charge of the detected particle. The relationship between the amphtude of the light signals and the kinetic energy transferred to the glass by electrons is linear whereas for the much heavier protons, deuterons and alphas it is nonlinear; the shapes of the nonlinear curves are the same for all of the heavy particles and for all the glasses studied. For all the charged particles studied, the light output of the scintillators decreased with increase in the lithium content in the glass; within the statistical error of the measurements the shape of this relationship appeared to be the same for all particles including electrons. For all glasses, the spread in the normalisation factors for the different charged particles increased with the increase in lithium content. The scintillation efficiency of all the glasses was greatest for electrons and least for alpha particles. For kinetic energy transfers of 1 MeV, the response of the glasses to electrons exceeded that due to protons, deuterons and alpha particles by factors 2.1, 2.8 and 9.5, respectively.
Acknowledgements
Thanks are due to Mr. R.J. Blevins for assistance with the experimental work, to Mr. W.J. Crawford for the design and construction of the target chamber, and to Dr. J.H. Boldeman and Mr. J.P. Fallon for help in operating the Van de Graaff accelerator.
References
[1] R.J. Ginther, IEEE Trans. Nucl. Sci. NS-7 (1960) 28. [2] L.E. Flynn, L.E. Glendenin, E.P. Steinbert and P.M. Wright, Nucl. Instr. and Meth. 27 (1964) 13. [3] V.V. Verbinski, W.R. Burrus, T.A. Love, W. Zobel, N.W. Hill and R. Textor, Nuct. Instr. and Meth. 65 (1968) 8. [4] H.H. Knox and T.G. Miller, Nucl. Instr. and Meth. 101 (1972) 519. [5] L.M. Bollinger, G.E. Thomas and R.J. Ginther, Nucl. Instr. and Meth. 17 (1962) 97. [6] A.R. Spowart, Nucl. Instr. and Meth. 135 (1976) 441. [7] E. Breitenberger, Progress in Nuclear Physics, ed., O. Prisch (Pergamon, London and New York, 1955). [8] J.B. Birks, The Theory and Practice of Scintillation Counting (Pergamon, London, 1964). [9] D.G. Anderson, J. Dracass, T.P. Hanagan and E.H. Noe, Proe. Syrup. on Photoelectronic Image Devices, Imperial College, London, England (1961) p. 429. [10] F.W.K. Firk, G.G. Slaughter and R.J. Ginther, Nucl. Instr. and Meth. 13 (1961) 313.
366
A. W. Dalton / Light conversion efficiency of small Li scintillators
[11] A.W. Dalton, AAEC E-Report, in preparation. [12] L.A. Wraight, D.H.C. Harris and P.A.Egelstaff, Nucl. Instr. and Meth. 33 (1965) 181. [13] A.R. Spowart, Nucl. Instr. and Meth. 82 (1970) 1.
[14] G.L. Jensen and J.B. Czirr, Nucl. Instr. and Meth. 205 (1983) 461. [15] A.R. Spowart, Nucl. Instr. and Meth. 75 (1969) 35.