IEEE
TLa
ctionm
on
Nucteat
Science,
VoL.S-24,
No.
1,
Feb4waky
1977
"THE
SECCNDARY
SCINTILIATICN
CUTPUT
OF
XENCN
IN
A
UNIFORM
FIElD
GAS
PROPORTICNAL
SCINILLATICZN
COUNER"
C.
A.
N.
Conde,
L.
Requicha
Ferreira
and
M.
Faitima
A.
Ferreira
Departanento
de
Fisica,
University
of
Coimbra
Coinbra,
Portugal
Abstract
Data
an
the
xenon
uniform
field
gas
proportional
scintillation
counter
is
presented.
The
energy
reso-
lution
obtained
for
8.1
MeV
t--particles
was
1.2%.
Measurements
for
pressures
fran
600
to
1500
Torr
shcwed
that
the
"reduced
light
output",
i.
e.
the
num-
ber
of
photons
produced
by
a
single
electron
tra-
velling
the
unit
of
distance,
divided
by
the
gas
pres-
sure,
de
only
on
the
reduced
electric
field
but
not
on
the
gas
pressure.
An
empirical
equa-
tion
for
this
quantity
is
given.
The
role
played
by
different
mechanisms
is
discussed.
l.
Introduction
The
gas
prqportional
scintillation
counter,
a
scintillation
counter
where
the
light
is
produced
by
the
primary
ionization
electrons
while
drifting
through
a
fairly
strong
electric
field
(strong
enough
for
excitation
but
not
for
ionization
of
the
gas
mo-
lecules)
has
been
the
subject
of
recent
researches
due
to
the
good
energy
resolutions
obtained
(for
a
survey
see
Policarpo's
article
).
After
the
initial
work
with
cylindrical
field
geometries2
followed
by
the
work
with
a
spherical
anode3,
emphasis
is
being
put
recently
on
the
two
grid
uniform
field
gas
proportional
scintillation
4-9
counter
.
This
geametry
seems
to
offer
advantages
over
the
others
for
nuclear
counting
experiments
and
on
the
other
hand
allows
the
study
of
the
dependen-
ce
of
the
secondary
scintillation
upon
the
electric
field
intensity,
the
gas
camposition
and
pressure.
The
results
obtained
for
the
cylindrical
and
sphe-
rical
non-uniform
field
geametries
do
not
allow
that
study,
and
very
little
results
have
been
published
for
uniform
fields.
The
best
energy
resolution
so
far
obtained
for
charged
particles
was
1.6%
for
8.1
MeV
a-particles
8
with
an
uniform
field
argon
filled
counter.
As
xenon
is
kncwn
to
give
more
light
than
argon
a
further
im-
provament
in
the
energy
resolution
seans
possible.
These
reasons
led
us
to
carry
out
the
study
of
the
xenon
filled
uniform
field
gas
prcportional
scin-
tillation
counter.
2.Experimental
Set-up
The
experimental
systam
used
is
essentially
the
same
as
the
one
described
before
4t6t8
but
with
a
1200
-rg/an2
p-terphenyl
wavelength-shifter
deposit
and
filled
with
xenon
at
pressure
fran
600
to
1500
Torr.
Due
to
the
large
amount
of
light
produced
in
xenon,
to
avoid
saturation
and
to
increase
the
sta-
bility
of
the
EMI
9656
QR
photanultiplier10
only
the
first
four
dynodes
were
connected
and
fifth
one
was
used
as
anode.
3.Experimental
Results
a)
Energy
resolutions
Fig.l
shows
a
spectrum
for
8.1
MeV
a-particles
with
an
energy resolution
of
1.2%.
This
figure
is
the
experimentally
obtained
one
and
includes
the
fluctua-
tions
of
the
phototube
gain
and
the
straggling
in
the
1.616
mg/an2
Melinex
window.
The
straggling
contri-
bution
was
measured
to
be
0.84%
using
a
high
resolu-
tion
silicon
detector.
Therefore
the
intrinsic
reso-
lution
of
the
counter
is
below
the
1%
figure.
This
means
that
even
for
charged
particles the
gas
prcpor-
tionalscintillation
counter
can
become
campetitive
wi-
th
other
counters.
The
energy
resolution
obtained
for
the
21.988-
-22.163
keV
X-ray
group
fram
a
109Cd
source
was
7.6%
for
xenon
at
1431
Torr
and
for
voltages
of
grid
1,
V1=4
kV,
and
of
grid
2,
V2=1.2
kV.
This
figure
is
worse
than
the
one
cbtained
by
Policarpo
et
al.3(4.5%)
due
to
the
fact
that
a
single
photaoultiplier
with
no
reflector
was
used.
b)
The
secondary
scintillation
output
The
variation
of
the
intensity
of
the
secondary
scintillation
as
a
function
of
the
voltage
difference,
Vl-VV,
for
a
constant
distance
of
6mm
between
the
grids,
was
measured
for
pure
xenon
at
pressures,
p,
of
600,
900,
1200
and
1500
Torr.
The
maximum
voltages
used
were
limited
by
sparking.
The
amplitude
of
the
secondary
scintillation
pulses
were
normalized
to
the
ones
fram
a
NaI
(Tl)
crystal
excited
by
0.661
MeV
y-rays.
As
the
number
of
photons
produced
in
the
crys-
tal
is
roughly
equal
to
5000,
a
rough
figure
for
the
number of
photons
reaching
the
photamultiplier
can
be
estimated.
For
8.1
MeV
a-particles
in
xenon
at
600
Torr
and
V1-V2=3000
V
that
figure
beccmes
2.9xl06
photons.
If
we
take
into
account
the
fact
that
there
is
no
reflector
in
the
counter
inner
walls,
at
most,
only
half
(or
even
less,
due
to
the
wavelength-
-shifter
effect)of
the
total
light
reaches
the
photo-
multiplier
which
means
that
the
total
number
of
pho-
tons
produced
is
about
6x106.
As
an
8.1
MeV
cxparti-
cle
produces
3.86x105
electrons
in
xenon
the
total
number
of
photons,
n,
produced
by
a
single
electron
221
Authorized licensed use limited to: Universidade de Coimbra. Downloaded on March 15,2010 at 10:10:42 EDT from IEEE Xplore. Restrictions apply.
travelling
between
the
grids
is
around
16.
Since
the-
se
figures
do
not
take into
account
the
variation
of
the
photocathode
efficiency
with
wavelength
nor
with
the
reflectivity
of the
walls,
they
are
estimates
and
should
be
used
cnly
as
a
guide.
Rather
than
plotting
the
variation
of
n
with
the
electric
field
intensity,
we
plot
in
Fig.2
the
"redu-
ced
secondary
light
output',yldefined
as
the number of
photons
produced
by
a
single
electron
travelling
the
unit
of
distance,x,
divided
by
the
gas
pressure,
2):
Y-p
a(expressed
in
photons/electron
m
orr
)
as
a
-P
11
function
of
the
reduced
electric
field:
E/p
(Volt
an7
Torr
1
).
The
results
obtained
show
that,
even
for
fields
well
above
1
Volt
am
Torr
1
there
is
no
dependence
11,13
of
the
reduced
light
output
on
the
gas
pressure
l
dn
0.06E
The
empirical
equation
p
--(-0.0074
+
0.0066
-)+Q.002
photons/electron
anm-
Tarr1
describes
fairly
well
the
behaviour
of
the
secondary
scintillation.
It
shows
that
for
reduced
fields
lower
than
about
1
Volt
anm1
-1
Torr
there
is
practically
no
secondary
light
produ-
ced;
above
this
threshold
the
variation
is
a
linear
one.
Although,
as
it
was
referred
to
before,
the
abso-
lute
number
of
photcns
is
subject
to
appreciable
er-
rors,
the
relative
figures
are
fairly
accurate;
the
relative
error
of
+
0.002
is
estimated.
4.
Discussion
of
the
Experimental
Results
The
above
pirical
equation
can
be
used
to
es-
timate
the
light
output
and
so
the
energy
resolution
of
new
designs
of
gas
Proportional
scintillation
coun-
ters
with
any
electric
field
geanetry,
at
any
pressu-
re.
For
uniform
field
counters
with
grids
at
a
dis-
tance
of
x
centimeters
and
voltages
V1
and
V2,
that
equation
leads
to
n=-0.0074
px
+
0.0066(V1-V2)
photons
which
means
that
fcr
fields
of
a
few
kilovolt
n
can
be
sufficiently
large
so
that
for
each
electron
there
are
3
or
4
photoelectrons
released
fram
the
photanul-
tiplier
cathode.
Thus,
the
contributicn
of
the
photo-
multiplier
statistics
to
the
energy
resolution
can
be
made
smaller
than
the
contribution
of
the
statistical
fluctuations
in
the
average
number,
N,
of
the
primary
i
electrons,
which
is
equal
to
(2.36F-h:
where
F,
the
Fano
factor,
is
0.17
for
xencn.
Thus,
the
limiting
energy
resolution
of
a
gas
prcportional
scintillation
counter
is
equal
to
the
one
of
a
noiseless
ionization
chamber
and
it
can
be
approched
when
the
condition
n>>l
is
verified.
As
it
has
been
discussed
before
3,12,
due
to
the
fact
that
noise
can
be
neglected
in
a
gas
prcportio-
nal
scintillation
counter
and
that
there
is
very
little
or
no
charge
multiplication,
the
energy
reso-
lution
of
this
counter
is
better
than
the
one
of
a
standard
Droportional
counter.
If
the
distance
between
the
grids
is
increased
by
Ax
(an)
keeping
the
voltage
and
the
gas
pressure
constant,
the
number
of
photons
is
increased
by:
An=-0.0074
p
Ax
Therefore,
if
the
grids
are
not
exactly
parallel
and
their
average
distance,
x,
has
fluctuations
of
the
or-
der
of
Ax,
the
contribution
of
these
to
the
energy
re-
solution
is
of
the
order
of:
Ax
An"
1
X
n
0.0066
(V1-V2)
1-
0.0074
px
Which
means
that
for
the
conditions
of
the
spectrum
of
Fig.l,
if
for
example
the
fluctuations
of
the
dis-
tance
are
of
the
order
of
2.5%
the
energy
resolution
is
deteriorated
by
1%.
For
the
cases
where
the
numeri-
cal
values
are
such
that
V1-V2
t
px,
instead
of
having
a
decreased
contribution
we
have
an
increased
cne:the
2.5%
fluctuations
lead
to
a
deterioration
of
the
order
of
25%.
The
reduced
light
output,
y,
as
defined
before
should
be
in
Drinciple
equal
to
the
photon
production
coefficient
of
Massey
14,
a
,
divided
by
the
gas
pressure,
E.
However
as
at
doesn't
take
into
account
the
wavelength-shifter
photon
conversion,
we
must
wri-
te:
a
ph
~ya
-
SC
p
where
atSC
is
the
efficiency
coefficient
for
photon
conversion.
Let
us
assume
that
the
photon
production
is
due
to
the
two
body
electron-atan
collisions
(the
three
and
more
body
processes
being
absent).
Then,
as
the
energy
acquired
between
two
collisions
depends
only
on
E/p,
and
the
number
of
photons
depn
only
on
the
number
of
collisions
(for
the
same
energy
acquired
between
collisions)
the
reduced
secondary
light
output
as
a
function
of
the
reduced
electric
field
11
should
be
independent
on
the
gas
pressure
.
As
said
before
our
experimental
results
(Fig.
2)
confirm
this
prediction.
This
is
in
disagreement
with
Szymanski
and
Herman's
results13.
The
discrepancy
might
result
fran
the
fact
that
in
their
experiments
the
nuclear
radiation
interacted
with
the
gas
in
the
light
production
region.
Thus
primary
electrons
produced
at
different
points
travelled
along
different
distances
and
produced
different
amounts
of
light;
as
the
gas
pressure
is
increased,
the
radiation
is
more
absorbed
and
the
electron
paths
get
longer.
Our
experimental
set-up
doesn't
have
this
inconvenient.
The
referred
to
above
independence
on
the
gas
pressure
shcws
that
indeed
the
excitation
of
xenon
atans
by
electrons
is
due
mainly
to
two
body
processes.
But
as
the
mean
free
path
of
electrons
in
xenon
at
atmospheric
pressure
is
at
least
1.6
x
10
4
cm
(the
maximum
of
the
total
cross-section
is
around
5
x
10
16
an2)
and
the
threshold
for
secondary
light
222
Authorized licensed use limited to: Universidade de Coimbra. Downloaded on March 15,2010 at 10:10:42 EDT from IEEE Xplore. Restrictions apply.
production
is
around
800
V/an,
an
electron
acquires
between
collisions
at
least
an
energy
of
0.13
eV.
This
implies
a
very
low
prcbability
for
an
electron
to
reach,
in
a
single
step
process,
the
energy
of
the
first
metastable
state
of
xenon
at
around
8
eV.
However,
a
lower
cross-section
and
the
fact that
an
electron
loses
very
little
energy
in
elastic
collisions,
make
possible
the
direct
excitation
of
xenon
atoms
by
electrons.
Let
us
now
consider
other
theories
for
the
production
of
secondary
light.
As
it
has
been
shown
before4
due
to
the
fact
that
the
secondary
scintillation
is
indeed
produced
between
the
grids,
the
theory
of
Braglia
et
al.
5
doesn't
seem
to
explain
it
for
the
electric
field
intensities
used
in
gas
proportional
scintillation
counters.
The
bremsstrahlung
theories
(for
a
discussion
see
reference
11)
cannot
be
ruled
out.
However
they
might
have
difficulty
in
explaining
the
threshold
for
light
production
at
about
1
Volt
amn1
Torr
1.
The
fact
that
an
electron
needs
to
travel
across
a
difference
of
potential
of
200
Volts
to
produce
one
photon,
means
that
the
efficiency
for
conversion
of
electrical
into
optical
energy
it
is
of
the
order
of
2
or
3%
for
the
reduced
fields
used.
This
figure
is
of
significance
for
laser
work.
As
the
energy
resolution
of
gas
prcportional
scintillation
counters
is
improved
the
sccpe
for
applications
increases.
Besides
the
ones
described
before:
low
energy
X-ray
spectrametry,
fast
coincidence
experiments
with
X-rays,
internal
gas
counting
(e.g.
carbon-14,
tritium
and
alpha
particles),
etc.,
applications
to
low
energy
proton
spectrametry
seem
practicable.
Ackncwledgements
This
work
has
been
supported
by
Instituto
Nacional
de
Investigacao
Cientifica
under
project
CF/2.
References
1-
A.
J.
P.
L.
Policarpo,
"The
Gas
Proportional
Scin-
tillation
Counter"
in
Space
Science
Instrumenta-
tion,
D.
Reidel
Publishing
Carpany,
Dordrecht
-
-
Holland
(1976)
(to
be
published).
2-
C.
A.
N.
Conde
and
A.
J.
P.
L.
Policarpo,
Nucl.
Instr.
Meth.
55
(1967)
105.
3-
A.
J.
P.
L.
Policarpo,
M.
A.
F.
Alves,
M.
C.
M.
dos
Santos
and
M.
J.
T.
Carvalho,
Nucl.
Instr.
Meth.
102
(1972)
337.
4-
M.
Fatima
A.
Ferreira,
Calia
A.
Sousa,
Cidilio
Cruz,
Lucilia
P.
de
Brito,
Rui
F.
Marques
and
C.
A.
N.
Conde,
"Estudo
de
um
contador
de
cintila-
cao
de
resolucao
elevada
can
argon
puro
e
campo
electrico
uniforme",
Cammunication
presented
to
the
"XXXI
Congresso
Luso-Espanhol
para
o
Progresso
das
Ciencias",
Cadiz-April
1974.
5-
H.
E.
Palmer
and
L.
A.
Braby,
Nucl.
Instr.
Meth.
116
(1974)
587.
6-
C.
A.
N.
Conde,
M.
C.
M.
Santos,
M.
Fatima
A.
Ferreira
and
Celia
A.
Sousa,
IEEE
Trans.
Nucl.
Sci.
NS-22,
n9
1
(1975)
104.
7-
H.
E.
Palmer,
I
Trans.
Nucl.
Sci.
NS-22,
n9
1
(1975)
100.
8
-
C.
A.
N.
Conde
and
M.
Fatima
A.
Ferreira,
Nucl.
Instr.
Meth.
124
(1975)
307.
9
-
R.
D.
Andresen,
L.
Karlsson
ard
B.
G.
Taylor,
IEEE
Trans.
Nucl.
Sci.
NS-23,
nQ
1
(1976)
473.
10
-
C.
E.
Cohn,
aE>E
Trans.
Nucl.
Sci.
NS-21,
n9
1
(1974)
146.
11
-
Yu.
A.
Butikov,
B.
A.
Dolgoshein,
V.
N.
lebedenko,
A.
M.
Rogozhin
and
B.
U.
Rodimnov,
Soviet
Physics
JT
30
(1970)
24.
12
-
C.
A.
N.
Conde,
in
Proc.
of
the
NATO
Summer
School
on
Material
Characterization
Using
Ion
Beams,
held
in
Aleria
(Corsica)
29
Aug.
-12
Sept.
1976
(to
be
published
by
Plenum
Press).
13
-
A.
Szymanski
and
J.
A.
Herman,
J.
Chimie
Phys.
60
(1963)
379.
14
-
H.
S.
Massey,
"Electronic
and
Ionic
Impact
Phenanena",
2nd
editim,
Vol.
II,
pg.
904
-
-
Ccford
University
Press
(1969).
15
-
G.
L.
Braglia,
G.
M.
Munari
and
G.
Manbriani,
Nuovo
Cimento
43B
(1966)
130.
XENON
UNIFORM
FIELD
GAS
PROPORTIONAL
SCINTILLATION
COUNTER
500_
o(
-
Particle
Spectrum
PX
:1403
Torr
V.
4400
V
V0
.1200
V
8.1
MeV
400
-J
z
300
-
z
w
"z-2
00
-R=11.2%
0
100
860
880
900
920
940
CHANNELS
Fig.
1
-
Alpha
particle
spectrum
for
a
xenon
uniform
field
gas
prcportional
scintillation
counter.
223
Authorized licensed use limited to: Universidade de Coimbra. Downloaded on March 15,2010 at 10:10:42 EDT from IEEE Xplore. Restrictions apply.
IL.
0
I-
c
ol
0
%A
0
-E
clx
-1
CL
Fig.
2
-
iced
y
light
aotput
for
a
xenan
uniform
field
gas
prcorticnal
scintillation
counter
with
p-terphenyl
wavelength-shifter.
224
Authorized licensed use limited to: Universidade de Coimbra. Downloaded on March 15,2010 at 10:10:42 EDT from IEEE Xplore. Restrictions apply.