- Email: [email protected]
Particle identification with very thin transmission detectors
NUCLEAR
INSTRUMENTS
AND
METHODS
I35
(i976)
io3-io9;
©
NORTH-HOLLAND
PUBLISHING
CO.
PARTICLE IDENTIFICATION W I T H VERY T H I N TRANSMISSION DETECTORS* D. G. P E R R Y
Brookhaven National Laboratory, Upton, New York 11973, U.S.A. and Los Alamos Sctenttfie Laboratory, Los Alamos, New Mextco 97545, U.S.A. and L.P. REMSBERG
Brookhaven National Laboratory, Upton, New York 11973, U.S A. Received 8 March 1976 A table-lookup particle ldentaficatlon scheme for A E - E counter telescopes which uses the semi-empirical range-energy data of Northchffe and Schdhng is presented. It is particularly stated for use with very thin AE detectors. Plots o f some o f the tables used are also given for use in the design of experiments.
1. Introduction The use of A E - E detector telescopes to identify charged particles emitted in nuclear reactions is based on the observation that, for a given residual energy E, the partial energy-loss AE, measured with a transmission detector depends only on the nuclear charge Z ,and the mass number A of the detected particle. It is not usually possible, however, to resolve adjacent isotopes of elements with Z > ,-~7, whde the separation ,of adjacent elements is sometimes possible up to a Z of 30-40. Many algorithms have been employed for particle identification, i.e., the conversion of a measured pair ,of energies AE and E into particle Z (and A). These mclude the original multiplier method 1,2) which is based on the Bethe rate-of-energy loss equation3), the power-law identifier introduced by Goulding et al. 4) and subsequently modified 5, 6), and a semi-empirical method described by Chuhck et al.7). They are all ,discussed in detail in the recent review article by Goul,ling and HarveyS). All of these methods are based on approximations to the range-energy relationships of the particles in the detector media, and their applicabihties are thus limited by the validities of those approximations. We have recently employed silicon A E - E detector telescopes to study the production of light fragments (6 < Z < 20) from the interaction of 28 GeV protons with gold and uranium targetsg). The AE detectors were very thin (7-11/~m) because the fragment energy spectra are broad and extend down to very low energies 'We found that none of the above-mentioned identifiW o r k performed under the auspices o f the U.S Research and Development Administration.
Energy
cation methods was satisfactory for our data. The major difficulty was with the lowest energy fragments because none of the approximations adequately describe the low energy region where the stopping powers decrease with decreasing fragment energy. We have therefore developed a "bruto-force", table-lookup procedure for particle identification which is described in the remainder of this paper. [A table-lookup procedure for the identification of one specific ion has been described by Hird and Ollerheadl°), and the conversion of the semi-empirical method of Chulick et al. 7) to a table-lookup procedure has recently been reportedi1).] We will also discuss some fundamental limitations to particle identification with very thin transmission detectors. Both our method and these limitations should apply to the thin gas ionization detectors which have been described recently12-~4). 2. Particle identification procedure The procedure is basically very simple: a twodimensional table of AE values as a function of Z and E is stored in the memory of a computer. For each event the particle Z is determined by linear interpolation of the measured AE and E values between those in the table. Since we store AE values as a function of Z and E, it is necessary to search through two adjacent columns of the table to find the AE values which bracket the measured AE. Less computer time would be required if a table of Z values (necessarily fractional) as a function of AE and E were employed because the two searches would be eliminated. This is probably a better procedure for on-line particle identification. However, for off-line data analysis we find the former procedure more convenient because it is not
104
D. G. P E R R Y
AND
necessary to change either the dimensions or the scale of the table when the AE detector thickness is changed. The table is generated from the semi-empirical range-energy tabulations of Northchffe and SchillingS5). The ranges of the various elements are converted to ranges in silicon and to a standard mass for each element, which is a smooth function of Z appro-
L. P. R E M S B E R G
priate to the experiment. For each Z and E entry in the table, the range is found by interpolation, the AE detector thickness is added, and the total energy, and thus AE, is obtained by a second interpolation; i.e., R(Z,E + AE)
R(E,Z)
=
+
t.
These interpolations are performed with a cubic spline procedure.
10
4
I
I
I
i
_
m
~5
m z
io 3 -
-
O
la
(-
io z -O
D
In order to get the best correspondence between the experimental data and the range-energy data, the energies as measured by the AE and E detectors were corrected for energy loss in the detector dead layers and also for pulse height defect. The energy lost m the rear dead layers of the AE detector was added to the energy deposited in the E detector. It was computed by multiplying the measured AE by the ratio of the dead layer thickness to that of the AE detector and the ratio of the average stopping power of the dead layer to that of silicon. The pulse height defect in the E detector was calculated by the method of Kaufman et a1.16). Their expression includes both the Z and A of the particle. Since our experiment measures Z only, we replaced A in their expression with the same function of Z which was used to generate the table for particle identification. The energy lost in the E-detector dead layer is included in this correction. There appear to be no studies of pulse height defects m transmisslon detectors. However, Wllkins et al. 17) have separated the total pulse height defect into three components: one due to the dead layer, one due to nuclear stoppmg which dominates near the end of the range of a particle, and one which they ascribe to electronhole recombination. We have assumed that only the
io 5 io I
io 4
NO
Na
C ~/!, FNelM~IAI (/)
103
(~ io 2
io o 3
Porticle
I
I
I
5
7
9
Identif~col~on
Fig. 1. Particle ~denUficatlon spectrum o f fragments resulting
f r o m the 28 G e V p r o t o n irradiation o f u r a n m m . The zlE detector was 9 / t m thick by 8 m m diam. w~th a 6 m m d r a m . colhmator. The electronics were optimized for l o w - Z fragments
,°',o~f
III 4
I
8
I
I
12 16 Port~cle
I
I
I
I
P
20
24
28
52
56
40
Idenhf~cohon
Fig. 2. The s a m e as fig 1 except that the electronics were adjusted to cover a broad range o f fragments
PARTICLE IDENTIFICATION
105
and the particle identification is then repeated. These terms are small and the particle identification usually converges in one or two iterations. The results of this procedure are illustrated in figs. 1 and 2 in which are shown particle identification spectra from experiments with 9 #m A E detectors. The electronics were optimized for low-Z fragments for the data shown in fig. 1. The elemental separation is very good and 7Be is resolved from 9Be and 1°Be. In fig. 2 one can see elements resolved up to Z ~ 2 4 . This is about as far as one could see peaks resolved in the A E - E distributions. Thus the limiting factor was the
latter effect applies to the A E detector and have used the expression dEpHD = 3.47 X 10 -3 (Z-- 11) [ ( A E / t ) - 5.1], which was taken from Steinberg et al.lS), for the A E pulse height defect. The correction for the dead layer on the front of the A E detector has no relevance to the particle identification procedure and was made at a different stage of the data analysis. Since the pulse height defect corrections are functions of Z, the particle identification is first performed without the corrections. ']'he Z thus obtained is used to calculate the corrections
3.0 MICRONS
i0.0 MICRONS
*2-
~
:i °O
30
60
90
120
] 50
E (MEV}
°O
30
60
90
120
E (MEV)
30.0 MICRONS
8
l
i
b-I 0 O~
c~0
30
60
90
120
] SO
E {MEV]
Fig. 3. Energy deposited in 3, 10, and 30/tm AE-detectors vs that m the E-detector for the elements LI through Zr.
150
106
D.G.
P E R R Y A N D L. P. R E M S B E R G
measurements. However, the increasing deviatton of the higher-Z peaks from the correct values implies systematic errors in the Northcliffe and Schilling range tables 15). A similar effect was noted by Chulick et al. 7) whose semi-empirical particle identification method was also based on the Northchffe and Schilling data.
AE detector and not the particle identification procedure. The peaks do not fall exactly at the correct Z-values, as they should in this procedure. This is partly due to the use of an incorrect value for the AE detector thickness in generating the table. We initially tried varying the AE detector thickness in an attempt to get all of the peaks to lie at the correct Z-values. That turned out to be impossible so we then varied the AE detector thickness to maximize the quality of the particle identification. The value finally used was typically about 5% less than that determined by e-particle energy-loss
3. Limitations to particle identification Some of the fundamental limitations of particle identification with transmission detectors are well known but can be seen quantitatively in graphs of the table used in our procedure. In fig. 3 are shown plots
10.0 MICRONS
3. O M ICRONS 150
-
,
,
--i~~
70 @
40
25
25 15 lOs
,
10
20
30
40
50
Z
°O
10
20
30
40
Z
30.0 MICRONS i
o
5
o ~E2 kJ o
°O
I0
20
30
40
50
Z Fig. 4. Energy deposited m 3, 10, and 30/zm AE detectors as a f u n c h o n of particle Z for seven values of the residual energy E
PARTICLE IDENTIFICATION
residual energy E, the particle Z is determined by the value of AE. Again one can see that for the higher Z particles AE becomes nearly independent of Z at low particle energies. This is shown more quantitatively in fig. 5 in which the slopes of the curves in fig. 4 are plotted and in fig. 6 in which the logarithmic derivatives (or AE -1 dAE/dZ) are plotted. These last two figures are useful in the design of experiments: fig. 5 illustrates the energy resolution of the AE detector required to separate adjacent elements, while fig. 6 shows the thickness uniformity required. For example, one obtains from fig. 5 that, for a fragment with Z = 20 and a residual energy of 25 MeV after passing
of AE vs E for each element from Li to Zr and for three different detector thicknesses. The lines in this figure represent the loci of ridge lines in two-parameter distributions of real A E - E detector telescope data. The basic problem is obvious: at low fragment energies the lines for the higher-Z elements become very close to each other and it is impossible to resolve adjacent elements. This is, of course, caused by the reduced effective charge of these ions which are not fully stripped. These same data are plotted in fig. 4 as A E v s Z for several values of E. These graphs illustrate the particle identification process: for a given value of the 3.0 MICRONS J
;
J
I0.0 MICRONS
8
J
107
l
)
l
2',
;2
\ 5~ N o
>~
lo /
taJ
\
is/ bA
N
\
es/
F~
40/ \
70/ 50/ 8. 8
15
24
z
32
40
30.0 MICRONS
!-i .... L 1sql I /
I4// ~
t ,o,j/ 70
1 o
"
0
~ 01
5
0
]6
~ 24
32
40
z
Fzg. 5. The denvatzves of the curves in fig. 4 as a functzon of particle Z.
;
1'.
z
,0
108
D. G. P E R R Y
AND
L. P. R E M S B E R G
3.0 MICRONS
i0.O MICRONS i
N
N
N N~ o o.
O
o
8. o
40
Z
Z
30.0 MICRONS
N Q
c.~ ~. o
Z Fig. 6. The logarithmic denvatwes of the curves m fig. 4 a s a function of particle Z. In each set the uppermost curve is for a residual energy of 5 MeV and the lowest is for 150 M e V .
through a 10/~m detector, the separation between adjacent elements is 2.2 MeV. This corresponds to 5% of the energy deposited in the AE detector, as can be read from fig. 6. From the data in fig. 2, one obtains a fwhm o f 0.6 o f a Z unit at Z = 20. This corresponds to 1.3 MeV and is primarily due to the thickness non-uniformity o f the detector. Thus the central 32 mm 2 (defined by a collimator) of this 50 m m 2 detector* had a thickness uniformity of 3% or 0.3 #m, expressed as fwhm. *
Ortec model D - 0 4 5 - 0 5 0 - 1 0 ,
References 1) R . H . Stokes, J. A. Northrop and K. Boyer, Rev. ScL Instr. 29 (1958) 61 2) W L. Bnscoe, Rev. Sci. Instr. 29 (1958) 401 3) H. Bethe, Ann. J. Phys. 5 (1930) 325. 4) F. S. Goulding, D. A. Landis, J. Cerny and R. H. Pehl, Nucl. Instr. and Meth. 31 (1964) 1. 5) G . W . Butler, A. M. Poskanzer and D . A . Landis, Nucl Instr. and Meth. 89 (1970) 189. 6) j . D . Bowman, A . M . Poskanzer, R . G . Kortehng and G. W Butler, Phys Rev. C 9 (1974) 836. 7) E. T. Chulick, J. B Natow~tz and C. Schnatterly, Nucl. Instr.
and Meth. 109 (1973) 171.
PARTICLE IDENTIFICATION a) F. S. Goulding and B. G. Harvey, Ann. Rev. Nucl. Sci. 25 (1975) 167. 9) L. P. Remsberg and D. G. Perry, Phys. Rev. Lett. 35 (1975) 361. Jo) B. Hind and R . W . Ollerhead, Nucl. Instr. and Meth. 71 (1969) 231. J~) J.B. Natowltz, M . N . N a m b o o d m and E.T. Chulick, Phys. Rev. C 13 (1976) 171 J2) M. M. Fowler and R. C. Jared, Nucl. Instr. and Meth. 124 (1975) 341. Ja) H. Sann, H. Damjanschltsch, D. Hebbard, J. Junge, D.
109
Pelte, B. Povk, D. Schwalm and D. B. Tran Thoal, Nucl. Instr. and Meth. 124 (1975) 509. t,) j. Barrette, P. Braun-Munzinger and C. K. Gelbke, Nucl. Instr. and Meth. 126 (1975) 181. is) L. C. Northcliffe and R. Schllhng, Nucl. Data 7 (1970) 233. ~6) S. B. Kaufman, E. P. Steinberg, B. D Wilklns, J. Unik, A. J Gorskl and M. J. Fluss, Nucl. Instr. and Meth. 115 (1974) 47. 17) B. D. Wilkins, M.J. Fluss, S. B. Kaufman, C E. Gross and E. P. Steinberg, Nucl. Instr. and Meth. 92 (1971) 381. is) E.P. Steinberg, S.B. Kaufman, B.D. Wllkins and G . E . Gross, Nucl. Instr. and Meth. 99 (1972) 309.
INSTRUMENTS
AND
METHODS
I35
(i976)
io3-io9;
©
NORTH-HOLLAND
PUBLISHING
CO.
PARTICLE IDENTIFICATION W I T H VERY T H I N TRANSMISSION DETECTORS* D. G. P E R R Y
Brookhaven National Laboratory, Upton, New York 11973, U.S.A. and Los Alamos Sctenttfie Laboratory, Los Alamos, New Mextco 97545, U.S.A. and L.P. REMSBERG
Brookhaven National Laboratory, Upton, New York 11973, U.S A. Received 8 March 1976 A table-lookup particle ldentaficatlon scheme for A E - E counter telescopes which uses the semi-empirical range-energy data of Northchffe and Schdhng is presented. It is particularly stated for use with very thin AE detectors. Plots o f some o f the tables used are also given for use in the design of experiments.
1. Introduction The use of A E - E detector telescopes to identify charged particles emitted in nuclear reactions is based on the observation that, for a given residual energy E, the partial energy-loss AE, measured with a transmission detector depends only on the nuclear charge Z ,and the mass number A of the detected particle. It is not usually possible, however, to resolve adjacent isotopes of elements with Z > ,-~7, whde the separation ,of adjacent elements is sometimes possible up to a Z of 30-40. Many algorithms have been employed for particle identification, i.e., the conversion of a measured pair ,of energies AE and E into particle Z (and A). These mclude the original multiplier method 1,2) which is based on the Bethe rate-of-energy loss equation3), the power-law identifier introduced by Goulding et al. 4) and subsequently modified 5, 6), and a semi-empirical method described by Chuhck et al.7). They are all ,discussed in detail in the recent review article by Goul,ling and HarveyS). All of these methods are based on approximations to the range-energy relationships of the particles in the detector media, and their applicabihties are thus limited by the validities of those approximations. We have recently employed silicon A E - E detector telescopes to study the production of light fragments (6 < Z < 20) from the interaction of 28 GeV protons with gold and uranium targetsg). The AE detectors were very thin (7-11/~m) because the fragment energy spectra are broad and extend down to very low energies 'We found that none of the above-mentioned identifiW o r k performed under the auspices o f the U.S Research and Development Administration.
Energy
cation methods was satisfactory for our data. The major difficulty was with the lowest energy fragments because none of the approximations adequately describe the low energy region where the stopping powers decrease with decreasing fragment energy. We have therefore developed a "bruto-force", table-lookup procedure for particle identification which is described in the remainder of this paper. [A table-lookup procedure for the identification of one specific ion has been described by Hird and Ollerheadl°), and the conversion of the semi-empirical method of Chulick et al. 7) to a table-lookup procedure has recently been reportedi1).] We will also discuss some fundamental limitations to particle identification with very thin transmission detectors. Both our method and these limitations should apply to the thin gas ionization detectors which have been described recently12-~4). 2. Particle identification procedure The procedure is basically very simple: a twodimensional table of AE values as a function of Z and E is stored in the memory of a computer. For each event the particle Z is determined by linear interpolation of the measured AE and E values between those in the table. Since we store AE values as a function of Z and E, it is necessary to search through two adjacent columns of the table to find the AE values which bracket the measured AE. Less computer time would be required if a table of Z values (necessarily fractional) as a function of AE and E were employed because the two searches would be eliminated. This is probably a better procedure for on-line particle identification. However, for off-line data analysis we find the former procedure more convenient because it is not
104
D. G. P E R R Y
AND
necessary to change either the dimensions or the scale of the table when the AE detector thickness is changed. The table is generated from the semi-empirical range-energy tabulations of Northchffe and SchillingS5). The ranges of the various elements are converted to ranges in silicon and to a standard mass for each element, which is a smooth function of Z appro-
L. P. R E M S B E R G
priate to the experiment. For each Z and E entry in the table, the range is found by interpolation, the AE detector thickness is added, and the total energy, and thus AE, is obtained by a second interpolation; i.e., R(Z,E + AE)
R(E,Z)
=
+
t.
These interpolations are performed with a cubic spline procedure.
10
4
I
I
I
i
_
m
~5
m z
io 3 -
-
O
la
(-
io z -O
D
In order to get the best correspondence between the experimental data and the range-energy data, the energies as measured by the AE and E detectors were corrected for energy loss in the detector dead layers and also for pulse height defect. The energy lost m the rear dead layers of the AE detector was added to the energy deposited in the E detector. It was computed by multiplying the measured AE by the ratio of the dead layer thickness to that of the AE detector and the ratio of the average stopping power of the dead layer to that of silicon. The pulse height defect in the E detector was calculated by the method of Kaufman et a1.16). Their expression includes both the Z and A of the particle. Since our experiment measures Z only, we replaced A in their expression with the same function of Z which was used to generate the table for particle identification. The energy lost in the E-detector dead layer is included in this correction. There appear to be no studies of pulse height defects m transmisslon detectors. However, Wllkins et al. 17) have separated the total pulse height defect into three components: one due to the dead layer, one due to nuclear stoppmg which dominates near the end of the range of a particle, and one which they ascribe to electronhole recombination. We have assumed that only the
io 5 io I
io 4
NO
Na
C ~/!, FNelM~IAI (/)
103
(~ io 2
io o 3
Porticle
I
I
I
5
7
9
Identif~col~on
Fig. 1. Particle ~denUficatlon spectrum o f fragments resulting
f r o m the 28 G e V p r o t o n irradiation o f u r a n m m . The zlE detector was 9 / t m thick by 8 m m diam. w~th a 6 m m d r a m . colhmator. The electronics were optimized for l o w - Z fragments
,°',o~f
III 4
I
8
I
I
12 16 Port~cle
I
I
I
I
P
20
24
28
52
56
40
Idenhf~cohon
Fig. 2. The s a m e as fig 1 except that the electronics were adjusted to cover a broad range o f fragments
PARTICLE IDENTIFICATION
105
and the particle identification is then repeated. These terms are small and the particle identification usually converges in one or two iterations. The results of this procedure are illustrated in figs. 1 and 2 in which are shown particle identification spectra from experiments with 9 #m A E detectors. The electronics were optimized for low-Z fragments for the data shown in fig. 1. The elemental separation is very good and 7Be is resolved from 9Be and 1°Be. In fig. 2 one can see elements resolved up to Z ~ 2 4 . This is about as far as one could see peaks resolved in the A E - E distributions. Thus the limiting factor was the
latter effect applies to the A E detector and have used the expression dEpHD = 3.47 X 10 -3 (Z-- 11) [ ( A E / t ) - 5.1], which was taken from Steinberg et al.lS), for the A E pulse height defect. The correction for the dead layer on the front of the A E detector has no relevance to the particle identification procedure and was made at a different stage of the data analysis. Since the pulse height defect corrections are functions of Z, the particle identification is first performed without the corrections. ']'he Z thus obtained is used to calculate the corrections
3.0 MICRONS
i0.0 MICRONS
*2-
~
:i °O
30
60
90
120
] 50
E (MEV}
°O
30
60
90
120
E (MEV)
30.0 MICRONS
8
l
i
b-I 0 O~
c~0
30
60
90
120
] SO
E {MEV]
Fig. 3. Energy deposited in 3, 10, and 30/tm AE-detectors vs that m the E-detector for the elements LI through Zr.
150
106
D.G.
P E R R Y A N D L. P. R E M S B E R G
measurements. However, the increasing deviatton of the higher-Z peaks from the correct values implies systematic errors in the Northcliffe and Schilling range tables 15). A similar effect was noted by Chulick et al. 7) whose semi-empirical particle identification method was also based on the Northchffe and Schilling data.
AE detector and not the particle identification procedure. The peaks do not fall exactly at the correct Z-values, as they should in this procedure. This is partly due to the use of an incorrect value for the AE detector thickness in generating the table. We initially tried varying the AE detector thickness in an attempt to get all of the peaks to lie at the correct Z-values. That turned out to be impossible so we then varied the AE detector thickness to maximize the quality of the particle identification. The value finally used was typically about 5% less than that determined by e-particle energy-loss
3. Limitations to particle identification Some of the fundamental limitations of particle identification with transmission detectors are well known but can be seen quantitatively in graphs of the table used in our procedure. In fig. 3 are shown plots
10.0 MICRONS
3. O M ICRONS 150
-
,
,
--i~~
70 @
40
25
25 15 lOs
,
10
20
30
40
50
Z
°O
10
20
30
40
Z
30.0 MICRONS i
o
5
o ~E2 kJ o
°O
I0
20
30
40
50
Z Fig. 4. Energy deposited m 3, 10, and 30/zm AE detectors as a f u n c h o n of particle Z for seven values of the residual energy E
PARTICLE IDENTIFICATION
residual energy E, the particle Z is determined by the value of AE. Again one can see that for the higher Z particles AE becomes nearly independent of Z at low particle energies. This is shown more quantitatively in fig. 5 in which the slopes of the curves in fig. 4 are plotted and in fig. 6 in which the logarithmic derivatives (or AE -1 dAE/dZ) are plotted. These last two figures are useful in the design of experiments: fig. 5 illustrates the energy resolution of the AE detector required to separate adjacent elements, while fig. 6 shows the thickness uniformity required. For example, one obtains from fig. 5 that, for a fragment with Z = 20 and a residual energy of 25 MeV after passing
of AE vs E for each element from Li to Zr and for three different detector thicknesses. The lines in this figure represent the loci of ridge lines in two-parameter distributions of real A E - E detector telescope data. The basic problem is obvious: at low fragment energies the lines for the higher-Z elements become very close to each other and it is impossible to resolve adjacent elements. This is, of course, caused by the reduced effective charge of these ions which are not fully stripped. These same data are plotted in fig. 4 as A E v s Z for several values of E. These graphs illustrate the particle identification process: for a given value of the 3.0 MICRONS J
;
J
I0.0 MICRONS
8
J
107
l
)
l
2',
;2
\ 5~ N o
>~
lo /
taJ
\
is/ bA
N
\
es/
F~
40/ \
70/ 50/ 8. 8
15
24
z
32
40
30.0 MICRONS
!-i .... L 1sql I /
I4// ~
t ,o,j/ 70
1 o
"
0
~ 01
5
0
]6
~ 24
32
40
z
Fzg. 5. The denvatzves of the curves in fig. 4 as a functzon of particle Z.
;
1'.
z
,0
108
D. G. P E R R Y
AND
L. P. R E M S B E R G
3.0 MICRONS
i0.O MICRONS i
N
N
N N~ o o.
O
o
8. o
40
Z
Z
30.0 MICRONS
N Q
c.~ ~. o
Z Fig. 6. The logarithmic denvatwes of the curves m fig. 4 a s a function of particle Z. In each set the uppermost curve is for a residual energy of 5 MeV and the lowest is for 150 M e V .
through a 10/~m detector, the separation between adjacent elements is 2.2 MeV. This corresponds to 5% of the energy deposited in the AE detector, as can be read from fig. 6. From the data in fig. 2, one obtains a fwhm o f 0.6 o f a Z unit at Z = 20. This corresponds to 1.3 MeV and is primarily due to the thickness non-uniformity o f the detector. Thus the central 32 mm 2 (defined by a collimator) of this 50 m m 2 detector* had a thickness uniformity of 3% or 0.3 #m, expressed as fwhm. *
Ortec model D - 0 4 5 - 0 5 0 - 1 0 ,
References 1) R . H . Stokes, J. A. Northrop and K. Boyer, Rev. ScL Instr. 29 (1958) 61 2) W L. Bnscoe, Rev. Sci. Instr. 29 (1958) 401 3) H. Bethe, Ann. J. Phys. 5 (1930) 325. 4) F. S. Goulding, D. A. Landis, J. Cerny and R. H. Pehl, Nucl. Instr. and Meth. 31 (1964) 1. 5) G . W . Butler, A. M. Poskanzer and D . A . Landis, Nucl Instr. and Meth. 89 (1970) 189. 6) j . D . Bowman, A . M . Poskanzer, R . G . Kortehng and G. W Butler, Phys Rev. C 9 (1974) 836. 7) E. T. Chulick, J. B Natow~tz and C. Schnatterly, Nucl. Instr.
and Meth. 109 (1973) 171.
PARTICLE IDENTIFICATION a) F. S. Goulding and B. G. Harvey, Ann. Rev. Nucl. Sci. 25 (1975) 167. 9) L. P. Remsberg and D. G. Perry, Phys. Rev. Lett. 35 (1975) 361. Jo) B. Hind and R . W . Ollerhead, Nucl. Instr. and Meth. 71 (1969) 231. J~) J.B. Natowltz, M . N . N a m b o o d m and E.T. Chulick, Phys. Rev. C 13 (1976) 171 J2) M. M. Fowler and R. C. Jared, Nucl. Instr. and Meth. 124 (1975) 341. Ja) H. Sann, H. Damjanschltsch, D. Hebbard, J. Junge, D.
109
Pelte, B. Povk, D. Schwalm and D. B. Tran Thoal, Nucl. Instr. and Meth. 124 (1975) 509. t,) j. Barrette, P. Braun-Munzinger and C. K. Gelbke, Nucl. Instr. and Meth. 126 (1975) 181. is) L. C. Northcliffe and R. Schllhng, Nucl. Data 7 (1970) 233. ~6) S. B. Kaufman, E. P. Steinberg, B. D Wilklns, J. Unik, A. J Gorskl and M. J. Fluss, Nucl. Instr. and Meth. 115 (1974) 47. 17) B. D. Wilkins, M.J. Fluss, S. B. Kaufman, C E. Gross and E. P. Steinberg, Nucl. Instr. and Meth. 92 (1971) 381. is) E.P. Steinberg, S.B. Kaufman, B.D. Wllkins and G . E . Gross, Nucl. Instr. and Meth. 99 (1972) 309.