Home Search Collections Journals About Contact us My IOPscience

An investigation of the response of a simple design of plane-parallel chamber

This content has been downloaded from IOPscience. Please scroll down to see the full text.

1994 Phys. Med. Biol. 39 1593

(http://iopscience.iop.org/0031-9155/39/10/006)

View the table of contents for this issue, or go to the journal homepage for more

Download details:

IP Address: 131.94.16.10
This content was downloaded on 03/10/2015 at 13:05

Please note that terms and conditions apply.

Phys. Med. Biol. 39 (1994) 1593-1608. Printed in the UK

An investigation of the response of a simple design of

plane-parallel chamber

C-M Mat, R T Knight, A E Nahum and W P M Mayles

Joint Department of Physics, Institute of Cancer Research and Royal Manden Hospital,
Sutton, Surrey SM2 5PT. UK

Received 16 November 1993, in final form 9 June 1994

Abstract. This paper repofis the experimental investigation of a simple design of plane-parallel
electron chamber, which has vev thin layers af copper (0.018 or 0.035 mm) as conducting
W n a i . Measurements comparing the prototype chambers with other ionization chambers
(pTw/Markus, NACP) have been carried out, both in a %I gamma-ray beam and in high-
energy electron beams. The results show that the C, factors (proportional to the product of
water/air stopping-power ntio and perturbation factor) for converting the in-phantom air-kem-
calibrated chamber reading to the absorbed dose to water =e nearly constant for incident electron
energies between 4 and 11 MeV for prototype chambers with 0.018 mm thick copper layers
and between 4 and 15 MeV for chambers with 0.035 mm thick copper layers. Other aspees
concerning these prototype chambers, such as polarity effect. cable effect, collecting efficiency
and angular response, have also been studied and the results are presented in this paper.

1. Introduction

According to the HPA (1985) code of practice for electron-beam dosimetry in radiotherapy,
the measurement of absorbed dose to water in an electron beam is performed using ionization
chambers calibrated in terms of air kerma in a phantom exposed to a beam of 6oCogamma
rays. The chamber reading, multiplied by Nf and corrected for standard ambient conditions
and for ion recombination, is then converted to the dose to water through a conversion
factor C, (HPA 1985). For electron energies below 10 MeV the Vinten 631 plane-parallel
chamber, which was developed at the National Physical Laboratory (NPL), UK, in the
seventies (Morris and Owen 1975), was up until recently the only chamber designated.
A problem for the users of the HPA (1985) code has recently arisen due to the fact that
the Vinten 631 chamber is no longer being manufactured. The NPL has recently designed
a plane-parallel chamber based on printed-circuit-board material with very thin (0.018 or
0.035 mm) layers of copper as conducting material. Preliminary measurements indicated
that prototypes of these chambers had responses that were nearly energy independent over
a certain range of electron energies. This is in sharp contmt to the behaviour of commonly
used plane-parallel chambers but could be ideal from a practical point of view.
This paper presents the results of the experimental investigation of the prototype NPL
design of plane-parallel chamber. Section 2 describes the formalism used in this work
for the derivation of the conversion factor C., The details of the measurements, and
of the ionization chambers used, are described in sections 3 and 4. The C, factors for
converting the in-phantom-air-kerma-calibratedchamber reading to the absorbed dose to
t Current addrru: Ionizing Radiation Standards, National Research Council of Canada, Ottawa KIA OR6, Canada.

0031-9155/94/iO1593+16SI9.50 0 1994 IOP Publishing Ltd 1593

1594 C-M Ma et ul

water in electron beams are given in section 5; these have been obtained experimentally
by comparison with already-designated chambers following the procedures recommended
by the HPA (1985) code of practice for electron-beam dosimetry in radiotherapy and the
IPSM (1992) addendum to this code. Other aspects concerning the prototype chambers,
such as polarity effect, chamber cable effect, collecting efficiency and angular response, are
also described. Conclusions are given in the final section, together with suggestions for
improvements to the NPL chamber design.

2. The €IPA (1985) formalism

The HPA (1985) formalism has been used in this work. According to the HF'A (1985) code
of practice for electron-beam dosimetry in radiotherapy, the following procedure should be
used for the measurements of absorbed dose to water in an electron beam using a designated
ionization chamber calibrated in terms of air kerma in a phantom exposed to a beam of
60Cogamma radiation:
(i) Intercompare the selected ionization chamber against a secondary-standard chamber
at 5 cm depth in a PMMA phantom exposed to a beam of 60Co gamma radiation. Both
chambers should be placed with the centres of their air cavities at the same depth. The
calibration factor Nf for the designated electron chamber in terms of air kerma for standard
ambient conditions is given by
N f = NsksRs/RE (1)
where R, is the reading of the secondary-standard chamber, Rf the reading of the electron
chamber, corrected for ion recombination and polarity effect, N, the calibration factor for the
secondary-standard chamber when irradiated by 2 M V radiation or 6oCogamma radiation in
air, to convert its reading to air kerma at the point in air corresponding to the centre of the
chamber, under standard ambient conditions, and k, a factor to convert the in-air calibration
of the secondary-standard chamber to an in-phantom calibration. If an NE2561 chamber is
used as the standard chamber then ks = 0.974.
(ii) For each electron-beam energy, carry out measurements on the axis of a large beam
in a water phantom to derive a depth-ionization curve for a fixed source-chamber distance
and then calculate the half-value depth (HVD) and practical range Rp of the electrons. If
a plastic instead of a water phantom is used other corrections may have to be applied to
readings of the ionization chamber and to the depths. The mean incident electron energy
EO' can then be calculated by
Eo = 2.4HVD (2)
where Eo is in MeV and HVD is in cm or g cm-' of water. The mean energy Ed of the
electrons at a depth d can be calculated by
Ed = Eo(1- d/Rp). (3)
From these values of Ed, the conversion factor C, can be derived for the designated
ionization chamber at various depths.
(iii) When the designated chamber is irradiated by an electron beam at a depth d in a
water phantom, the absorbed dose to water is given by
D, = N&E, (4)
where Re is the reading of the chamber, with its effective point of measurement at d in water
or an equivalent depth in some other medium, corrected to standard ambient conditions for
ion recombination, and for polarity effects.
 Investigation of a prototype plane-parallel chamber 1595

The C, factors in the HPA (1985) protocol are only given for two designated chambers,
the NE2571 cylindrical (Farmer) chamber for Ed > 5 MeV and the Vinten 631 plane-pdlel
>
chamber for Ed 1 MeV. The number of designated chambers has recently been increased
to include two more flat-chamber types, the NACP and Markus chambers, because of the
non-availability of the Vinten 631 chamber (IPSM 1992). The C, factors for the NE2571
chamber in a PMMA phantom have been modified in this addendum to the HPA (1985) code
of practice while those for water and polystyrene phantoms remain unchanged (IPSM 1992).
Values of conversion factors C: for other non-designated electron chambers can be
derived by intercomparison with a designated chamber in an electron beam. The non-
designated chamber should first be calibrated against a secondary-standard chamber with
the centres of the air cavities of the two chambers at 5 cm depth in a PMMA phantom exposed
to a 6oCogamma-ray beam as described above. From this measurement a calibration factor
N; can be calculated from (1) for this chamber. To derive a C: value for the non-designated
chamber it should be then intercompared in an electron beam with a designated chamber that
has previously been calibrated against a secondary-standard chamber in a MLCogamma-ray
beam with a calibration factor of N f . For the intercomparison the two chambers should be
placed in succession with their effective points of measurement at the same depth (at or
near the depth of maximum ionization) on the beam axis in a suitable phantom. The CL
value for the non-designated chamber in a particular electron beam is then given by

where C, is the conversion factor for the designated chamber, Re and RL the chamber
readings in the electron beam (corrected for ion recombination, polarity effect, temperature
and pressure variation) for the designated and non-designated chamber, respectively, and
Rca and Rho the chamber readings in a PMMA phantom irradiated by a mCo beam for the
designated and non-designated chamber, respectively.
The variation of C, with energy for a particular electron chamber depends on both the
electron-fluence-perturbation correction factor p,' and stopping-power ratio for the phantom
material to air sma,ep According to the analysis of Nahum et al (1988), the conversion
factor C, can be written as

where k& is a correction factor for a secondary-standard NE2561 chamber to account for the
lack of air equivalence of the chamber wall, electrode and build-up cap and kict a correction
factor for the effect of absorption and scattering in the chamber wall, electrode and build-up
cap at the calibration of the chamber irradiated in air by a 6oCo beam or a 2 M V x-ray
beam. g is the fraction of the energy of the charged particles lost to bremsstrahlung. The
perturbation factor, p d . corrects for the effect of non-medium-equivalent wall material
and central electrode, and for the deviation of the effective point of measurement of the
chamber from the chamber centre during the intercomparison of a secondary-standard
NE2561 chamber and an electron chamber. If the plane-parallel chambers are properly
guarded and if the phantom material is the same as the chamber-wall material, the Auence-
perturbation correction factor pg at (or close to) depths of maximum ionization will be
close to unity in high-energy electron beams (Holt ernl 1979, Mattsson et a1 1981, Ma and
Nahum 1993a) and therefore the C, factor varies only with stopping-power ratio swaIm,&.
1596 C-M Ma et a1

3. The ionization chambers

Following the procedure described in the previous section, experiments to derive the C,
factors for the prototype NPL plane-parallel chambers by comparing these chambers with
three designated plane-parallel chambers, i.e. a PIW/Markus and two NACP chambers,
have been carried out. The physical parameters of the plane-parallel chambers used in this
work are given in table 1. The geometrical details of the NPL chambers are shown in
figure 1.

Table 1. Physical characteristics of the plane-parallel chambers used in this work.

Collecting Entrance window

Air cavity electrode Guard ring Cavity
diameter diameter width depth Thickness
Model (mm) (mm) (mm) (mm) Material (mg cm-2)
NACPl 16.0 10.0 3.0 2.0 graphite Melinex 90.0
NACPZ 16.0 10.0 3.0 2.0 graphite Melinex 100.0
PTWlMarkus 6.0 5.4 0.2 2.0 graphited polyethylene 2.3
NPL (18 pm Cu) 28.0 20.0 3.0 1.5 PMMAandCu 154.0
NPL (35 pm Cu) 28.0 20.0 3.0 1.5 P M M A ~ ~ C U 184.0

Figure 1. A cross-sectional view of the NPL plane-parallel chamber with 0.018 mm thick capper
layers. The thickness of each copper layer (1, 2, 3 or 4) is 0.018 mm and the thickness of the
polyimide layer is 0.025 mm. The diameter of the collecting electrode is 20 mm and the width
of the guard ring is 3 mm. The diameter of the air cavity is 28 mm and the thickness is 1.5 mm.
The thickness of the front and back Perspex wall is 1 mm. For a 0.035 mm thick copper-layered
version the geometly is exactly the same except that the copper layers are 0.035 mm thick.
 Investigation of a prototype plane-parallel chamber 1597

There are currently two versions of this design with different thicknesses of the
conducting copper layers: (i) 0.018 mm and (ii) 0.035 mm. The thicknesses of the entrance
windows differ correspondingly between the two chamber versions. The thicknesses of the
entrance windows for the PTWMarkus and NACP chambers are based on the information
in the manufacturer’s specification sheets. One of the NACP chambers (NACPZ) has a
fairly large waterproof casing (made of PMMA). The thickness of the entrance window of
this chamber is approximately 0.1 g cm-’.

4. The measurements

Intercomparisons of the plane-parallel chambers against a secondary-standard NE2561

chamber were carried out on a Theratron 780 6oCounit. The chambers were placed with the
centres of their respective air cavities at 5 cm depth in a PMMA phantom and irradiated by a
6oCobeam with a IO x 10 cm2 field defined at 80 cm source-to-surface distance (SSD). The
chamber readings were corrected for ion recombination and the polarity effect following
the recommendations in the HPA (1985) code of practice.
The electron-beam measurements were made on an SL25 linear accelerator. The
electron-beam parameters given in table 2 were derived from depth-ionization curves
measured with a silicon diode detector (a Scanditronix field detector) in a water phantom.
The depth of maximum ionization d,,, in a PMMA phantom, for each incident electron energy,
was derived from the in-water depth-ionization measurements and scaled to account for the
difference in radiation properties between water and PMMA; the depths in PMMA in g cm-’
were obtained by multiplying the depths in water by a factor of 1.03 as recommended
by the HPA (1985) code of practice. At each energy, a 0.1 cm thick PMMA slab was
added or removed from the phantom surface to verify the location of d, in PMMA. The
mean incident electron energy EO and the mean energy at the depth of maximum ionization
E d were calculated using (2) and (3), respectively. The chambers were placed with their
effective points of measurement at the depth of maximum ionization in a PMMA phantom
and irradiated by electron beams of between 4 and 22 MeV (nominal energy). The effective
points of measurement for the plane-parallel chambers were taken to be at the centres of
the inside surfaces of the entrance windows. The chamber readings were corrected for ion
recombination, polarity effect, temperature and pressure variation as recommended by the
HPA (1985) code of practice. In all cases, the SSD was 95 cm and the field size at the
phantom surface was 10 x IO cm’.

Table 2. Pmmeters of the electron beams used in this work.

Nominal energy d, HVD RP Eo Ed
(MeV) (e ~ m - ~ ) (e c d ) (e cm-? (MeV) (MeV)
4 0.83 1.65 2.00 3.96 2.46
6 1.31 2.35 3.00 5.64 3.18
8 1.79 3.05 3.85 7.32 3.99
10 2.02 3.70 4.60 8.88 4.97
12 2.74 4.70 5.75 11.3 5.90
15 3.10 5.80 7.10 13.9 7.84
18 3.10 7.20 8.75 17.3 11.2
20 3.10 8.05 9.20 19.3 12.8
22 3.10 8.45 10.2 20.3 14.1
1598 C-M Ma et ul

5. Results and discussion

5.1. The calibration factor Nf

All the chambers were calibrated against an NE2561 secondary-standardchamber in a PMMA
phantom irradiated by a 6oCobeam under the conditions recommended by the HPA (1985)
code of practice. The Nffactors calculated according to (1) for the six plane-parallel
chambers used in this work are given in table 3.

Table 3. The calibmion hctors. Nr, for the plane-parallel chambers used in this work. The
chambers were centred at 5 cm depth in a PMMA phantom irradiated by a 6oCo beam with a
10 x 10 cm2 held and 80 cm SSD.
NPLl NPL2 NPL3
PTWlMarkus NACPl NACPZ 18 wnCu 35pmCu 35i~mCu
Nr (mGv nC-'l 503.1 151.9 157.6 33.49 30.92 29.30

The nominal volume of the sensitive air cavity of the new NPL chamber is about three
times that of the NACP chamber. The readings of the 0.018 and 0.035 mm copper-layered
NPL chambers were, however, only about 55 and 70% larger than that of the NACP chamber
respectively. This indicated a very large perturbation effect introduced by the thin copper
layers surrounding the air cavity in the new NPL chambers. The difference between the
readings of the two NACP chambers was about 3.8% in a 6oCobeam. This is thought to be
mainly due to the difference between the effective volumes of the sensitive air cavities of the
two chambers. However, small effects from the air gaps between the chamber and its PMMA
casing, both in @COand in electron beams, and uncertainty in the thickness of the entrance
window of the chamber with the PMMA casing could not be ruled out. The difference
between the readings of the two 0.035 mm copper-layered NPL prototype chambers was
about 6% both in 60Co and in electron beams. This is attributed to the difference in. the
effective collecting volume of the individual chambers.

5.2. The conversion factor C,

Table 4 gives the dose-conversion factors C, for the NPL prototype chambers with copper
layers of either 0.018 or 0.035 mm thickness obtained according to (5) by comparison with
the designated chambers (i.e. two NACP and one F'TW/Markus). For completeness, the C.
factors for the NACP and PTW/Marhs chambers are also given (see also figure 2). The
C, factors for the NPL chambers, derived fiom the comparative measurements with the
three reference chambers, are consistent within 1% for incident energies up to 20 MeV.
The C. factors for the NPL chambers derived from comparisons with the NACP chambers
are on average 0.3% higher than those derived from comparisons with the PTWlMarhs
chamber. The experimental uncertainty in these values was estimated to be about 1%due
to the uncertainties in determining d,,, and Ed, and due to the uncertainty in the C, values
for the designated chambers.
As can be seen from (€4, the conversion factor C, for a particular electron chamber
varies with both the electron-fluenceperturbation correction factor p: and stopping-power
ratio for water to air swafer.air.The magnitude of p : for the NPL electron chamber can be
estimated by comparisons of the readings of the NPL chamber at various depths and energies
with that of a chamber that has a p: factor close to unity, such as the NACP (Mattsson et
al 1981, Reft and Kuchnir 1991, Kuchnir and Reft 1992, Wittkamper et al 1991, Ma and
Investigation of a prototype plane-parallel chamber 1599
Table 4. The doseconversion factor, C., for the NPL prototype chamber with either 0.018
or 0.035 mm thick mpper layers obtained by eompmtive experiments with other ionization
chambers in a PMMA phantom. The C, values given here are the mean values OI C, derived from
the three designated Rat chambers (i.e. two NACP and one pTw/Markus). The COfactors for the
NPL chamber with 0.035 mm thick copper layers are the m m values for the chambers NPL2
and NPL3. The C, factors for the NACP and Markus chamber were taken from P S M (1992).
The experimental uncertainties are estimated to be about 1%.

Ed d, NPL NPL
(MeV) ( g ~ m - ~ ) (18 FmCu) (35 pmCu) NACP Markus
2.46 0.83 1.418 1.382 1.061 1.032
3.18 1.31 1.426 1.392 1.048 1.021
3.99 1.79 1.424 1.402 1.036 1.013
4.97 2.02 1.413 1.401 1.022 1.003
5.90 2.74 1.393 1.389 1.011 0.997
7.84 3.10 1.371 1.375 0.993 0.985
11.2 3.10 1.344 1.359 0.973 0.967
12.8 3.10 1.328 1.345 0.965 0.961
14.1 3.10 1.321 1.337 0.960 0.957

l'io]k
1.08
-."-.o-
_____ *
NPL(i8p Cu)
___..
NPL(35p Cu)
Markus
NACP

Ed (MeV)

Figure 2. Relative C, factors for the planepanllel chambers used in this work. The values
were tlken fmm table 4 and normalized to the values for E d = 14.4 MeV.
1600 C-M Ma et a1

Nahum 1993a). Figure 3 shows the p: factors for the prototype NPL chambers and for the
PTWNarkus chamber, obtained as the ratio of the reading of the NACP chamber to that
of the NPL and PTWMarkus chambers, normalized to the values for 20.3 MeV incident
energy (Ed = 14.1 MeV). The experimental uncertainties in the values are estimated to be
about 1%.
Compared to the NACP chamber, the 0.035 mm copper-layered NPL chamber has a very
large perturbation effect at lower electron energies. The variation of p: with energy happens
to nearly compensate the change in swater,air. As a result, the C, factor deviates by less than
1% from 1.389 for incident energies between 5 and 15 MeV for a 0.035 mm thick copper-
layered chamber while it deviates by less than 1%from 1.413 for incident energies between
4 and 11 MeV for a 0.018 mm copper-layered chamber. For higher energies p: becomes
nearly constant (figure 3(a)) and C, decreases with energy at the same rate as swurer.lirfor both
chamber versions (figure 2). These results are consistent with the Monte Carlo calculations
of Ma and Nahum (1993b). For the PTWfMarkus chamber, the p: factor decreases with
decreasing energy from unity at about 20 MeV incident energy to 0.977 at about 4 MeV
incident energy (figure 3(6)). These values are smaller than the p: factors adopted by the
Dutch code of practice (NCS 1989) but very close to those calculated from the least-squares
fit to the data in the literature (cf. Nahum 1993):~:= 1 - 0.072exp(-0.336Ed).
A possible explanation for the dose increase in an air cavity surrounded by thin layers
of copper is electron back scattering from the copper layer. As was demonstrated by Ma
and Nahum (1993b), the effect of electron back scattering from a thin layer of copper is a
function of both electron energy and the thickness of the copper layer. For a 0.018 mm thick
copper layer, the absorbed dose in a 0.1 mm thick layer of water in front of the copper layer
was increased over the value in homogeneous water by less than 4% for 4 MeV electrons
but by up to 20% for 400 keV electrons. A 0.061 mm thick copper layer increased the
absorbed dose by more than 30% at 600 keV, i.e. for thicker copper layers the maximum
dose increase occurred at higher energies. Only the dose in the water layer close to the
copper layer was elevated for 200 keV electrons. The dose to water in the water layers
0.2 mm away from the copper layer was barely affected.

5.3. Collecting efJlciency f

The collecting efficiency f is defined as f = Q'/Q, where Q is the charge created in the
sensitive air volume by the ionizing radiation and Q' the charge collected by the biased
electrode in the chamber and measured by the electrometer circuit. Figure 4 gives the
variation with applied potential of the ionization charge Q collected from a 0.018 mm
thick copper-layered NPL chamber (NPLl), normalized to the value for 300 V. It can be
seen that the difference in the charge collected for applied potentials 100-300 V is smaller
than 0.3%. In this work all the measurements with the NPL prototype chambers were made
with a potential of 250 V.

5.4. Polarity effects

Polarity effects have been checked for the three NPL planeparallel chambers, both in
and in eleccon beams, by taking additional readings with the polarizing voltage reversed
in sign. After reversing the polarizing voltage, about 10 min was allowed to elapse before
taking another reading (see section 5.3). The chamber readings with opposite polarizing
voltages were compared for each of the radiation conditions and the mean of the two values
was used if they differed. In all cases, the chamber readings with opposite polarizing
voltages differed by less than 0.3%.
Investigation of a protorype plane-parallel chnmber 1601

1.01 -
(b)

1.oo -
0.99 -
..-
" .
This. ..e
0.98 - work

----- NCS (1989)

Flt to prevlous data

0.97 -

0.96 1 I i
0 4 8 12 16 20

Figure 3. The ritio of the f l x n c e - p n u r b i u o n conecuon iaclor. p:, i o r m electron chamber

IO thnr for 3" NACP c n m x r cdcdlued 32 the ntio of ihc chamber m d m g o i [he SACP
chimbzr to th3t af thc ~.Iu:tro~~ chmber. nomllired t~ unity 31 E l = I4 I M*\' (U, for the
SPL chimbcrs with cithcr 0.018 or 0.035 mm thick ;opper $)en m d ( b ) for the PTWIMukus
chamber The \lonre Cx.o-cdc~laedp[ w h e s i n ( a ) u m tiken from $12 and hahum \ I 9 9 j b )
Thc SCS cune rn ( b ) was given by the D m h code of practice (\ICs 1989) w h i k tho solid
c u n e r w c3lcuhted from the I ~ S I - S ~ U J R I S fir to thc d m ID the litenture ,cf Nahum 1993)
Th? erpenmental uncenunties in thu results af tlus work x e estimated to be about 1 % The
staisncal unccnintier in the Xlonte Cvlo results m sholsn 3 i error bars on the curves.
1602 C-M Ma et a1

0.88 1 I 1 I
0 100 200 300

POLARISING VOLTAGE (V)

Figure 4. The variation of charge Q collected, for a given irndiation. from a 0.018 mm thick
copper-layered NPL chamber a1 2.3 cm depth in PMMA irradiated by a 11.3 MeV beam with
applied potential, normalized to the value for 300 V.

5.5. Chamber cable and stem effects

The effects of the chamber cable and the stem (i.e. the excess part of the chamber body) have
been investigated for a chamber with 0.035 mm copper layers (NPL2) in both locogamma
rays and electron beams. For the 6oComeasurements the chamber was placed at 5 cm depth
in a PMMA phantom. A 6 x 10 cm2 rectangular field was used with 80 cm SSD. The ratio
of the chamber reading with the cable in the field (figure 5(a))to that without (figure 5(b))
was 0.9978. This slight decrease in the chamber response may be due to the effect of the
chamber cable but may also be due to the relatively higher attenuation of scattered photons
by the surrounding copper layers compared with water when a larger area of copper layers
was irradiated by the beam. This is similar to the effect of the stem of a NE2571 chamber,
which consists partially of aluminium. The chamber stem increases the chamber response
in air in medium-energy x-rays due to increased photon scattering from the stem material
compared with air while it decreases the chamber response in a water phantom (Ma 1992,
Ma and Nahum 1994). Nonetheless, this 0.2% effect was very close to the uncertainty in
the measured chamber readings (reproducibility) in a %o beam (0.1-0.2%).
The chamber response was also measured for electron beams with a 3 x 10 cm2
rectangular field and 95 cm SSD at a depth in a PMMA phantom. For an electron beam of
8.88 MeV mean incident energy the ratio of the chamber response with the cable irradiated
(figure 5(a)) to that without (figure 5(b)) was 1.0038 while for 17.3 MeV mean incident
energy the response ratio was 1.0041. Again, this 0.4% effect may be partially due to the
chamber cable and partially due to the rectangular shape of the chamber body. The dose
to air in the air cavity was likely to be elevated due to excess electron scattering from the
copper layers compared with the phantom material.
 Investigation of a protorype plane-parallel chamber 1603

$$
:X
53)
Sensitive
air cavity 1 I (b)

- chamher cable /

Figure 5. A schematic illustration of the geometry set-up for the measurement of the chamber-
cable effects. The chamber was placed at a depth in a. phantom irradiated by either 6oCo or
electron beams of a rectangular field. In one case the chamber cable was irradiated ( 0 ) while
in the other the cable was excluded from the field (b).

As the chamber response is likely to decrease in 6oCo beams and increase in electron
beams, an overall cable- and stem-effect correction factor smaller than unity should be
applied if a chamber is calibrated in a phantom in a 6oCobeam and subsequently used at
a depth in a phantom in electron beams (HPA 1985, IPSM 1992). For a commonly used
10 x 10 cmz field, however, the overall correction required is estimated to be less than 0.5%
for the chamber investigated (NPLZ). Similar corrections should apply to the other prototype
NPL chambers. It should be noted, however, that this correction has automatically been
included in the dose conversion factor C, if C, is determined by comparative measurements
with a designated chamber.

5.6. Other effects

Investigations were canied out into the chamber response with the high-tension-(HT)
electrode side of the chamber (figure 1) facing either toward or away from the radiation
source. For %o gamma rays, the chamber reading with the €IT-electrode side facing
the radiation sonrce was between 1.1 and 1.8% lower compared to the reading with the
collecting-electrode side facing the source for the three prototype chambers. The difference
in chamber reading between the two set-ups decreased to about 0.5% in electron beams.
As the NPL chambers are nearly symmetrical in both wall thickness and material except
that a narrow ring of the copper layer has been removed between the collecting electrode
and the guard ring, the difference in chamber response between the two set-ups could only
be a result of the difference in the effect of the copper layers. Compared to PMMA, copper
causes more photon attenuation, if the copper layer is in front of the air cavity, and more
electron back scattering, if the copper layer is behind the air cavity. The overall effect,
therefore, was a slight elevation of the chamber response when the collecting-electrode side
1604 C-M Ma et a1

of the chamber was facing the radiation source (in this case, the ratio of the area of the
copper layer in front of the air cavity to that behind the air cavity is about 0.95). In this
work all the measuremen= with the new NFL chambers were made with the same set-up,
i.e. the collecting-electrode side of the chamber facing the radiation source.

5.7. Angular-response measurement

Experiments comparing the angular response of an NPL chamber with 0.018 mm thick
copper layers (NPLl) with that of an NACP chamber (NACP2) have been carried out.
The experimental set-ups were similar to those for the measurement of dose equivalent at
various depths for electrons incident from different angles (ICRU 1992). The chambers were
imbedded in a 20 x 20 x 13 cm3 PMMAphantom with their front faces at the same level as
the phantom surface. The effective point of measurement was therefore about 0.154 g
below the phantom surface for the NPL chamber and about 0.1 g cm-' below for the NACP
chamber. The chambers were irradiated by electron beams with a 20 x 20 cm2 field and
95 cm SSD. The results are summarized in figure 6 for three energies, 4, 8 and 12 MeV
(nominal energy), and incident angles between 0 and 75".

1.4

CI
5 1.3 -0- NACP4MeV
n
da --t NACP 8MEV
--D- NACP12MeV
U 1.2
W
m NPL 4 MeV
I
4
:: 1.1
NPL 8 MeV

W NPL 12 MeV
2
+
5W 1.0
U

0.9
0 20 40 60 80

INCIDENT ANGLE (degree)

Figure 6. Variation with incident angle of the response of a prototype NPL chamber with
0.018 mm chick copper layers and of a NACP chamber in electron beams of 4, 8 and 12 MeV
nominal energy (see- table Z), divided by the values for 0 degree (i.e. the axis of the beam is
perpendicular to the front face of the chamber). The detzils of the geomeuy are described in
the text.

It can be seen that, for 4 MeV electron beams, the chamber readings increase with
incident angle and reach maxima at around 45". The chamber readings decrease with
incident angle for angles greater than 60". For 12 MeV electrons the chamber readings
increase more at small angles and do not drop off until about 75". These angular response
curves are similar to the results of Cross in ICRU Report 47 (ICRU 1992) although
the thickness of the material in front of the dose-scoring region in his calculations was
 Investigation of a prototype plane-parallel chamber 1605

0.07 g cm-’. Compared to the NACP chamber, the NPL chamber gives slightly larger
signals at small incident angles but much smaller signals at large angles. This is probably
partially due to the slight difference between the thickness of entrance window in the two
chambers. However, the thin copper layers surrounding the air cavity in the NPL chamber
may also play a role in its angular response.

5.8. Depth-ionization measurement

Measurements of the depth-ionization curve for electron beams have been made with an
NACP chamber and two NPL prototype chambers, one with 0.018 and one with 0.035 mm
thick copper layers. Figure 7 gives the depth-ionization curves obtained using ionization
chambers in a PMMA phantom irradiated by a 5.64 MeV (incident energy) elecfron beam
with a 10 x 10 cm2 field defined at the phantom surface and 95 cm SSD. The one-standard-
deviation uncertainty in the measured ionization is estimated to be less than 1%.

”* 1 NACP

0 10 20 30

DEPTH IN PMMA (mm)

Figure 7. The relative depth-ioniution a m e s for5.64 MeV (incident energy) electrons incident
on a PMMA phantom obtained with three ianivtion chambers, i.e. NACPZ NPLl and NPLZ.
The field sire was 10 x IO cmz defined at 95 cm SSD.

It can be seen that the depths of maximum ionization measured using the three chambers
are consistent (within 0.5 mm). Compared to the values at the depth of maximum ionization,
the NPL chamber with 0.018 mm thick copper layers gives lower signals in the depth-
ionization build-up region but higher signals in the depth-ionization fall-off region than
that of the NACP chamber. The response of the NPL chamber with 0.035 mm copper
layers is even lower in the depth-ionization build-up region but becomes closer to that
1606 C-M Ma et a1

of the NACP chamber in the depth-ionization fall-off region. As the NPL chamber with
0.018 mm copper layers overestimates the ionization at large phantom depths, the depth-
ionization curve obtained using this chamber is not suitable for the dctcrminntion of the
electxon mean energies at various depths through the EO = 2.4 HVD relation (2).
5.9. Variation of C, with depth
Comparisons between the C, factor for the NACP chamber and that for the NPL chambers at
different depths but for the same incident energy have been made. Following the HPA (1985)
formalism, the mean energy Ed at a depth d can be calculated from (3). The C, factors
for the NACP chamber can then be derived from these E d values. The C, values for the
NPL chambers can then be calculated from (5). Figure 8 gives the C, factors for the NPL
chambers with either 0.018 or 0.035 mm copper layers as a function of phantom depth in
PMMA, normalized to the values for 11 mm depth, for a 5.64 MeV electron beam with a
10 x 10 cm2 field and 95 cm SSD. The C, factors for the NACP chamber increase nearly
linearly with depth. For a 0.018 mm copper-layered NPL chamber, the C, factor varies by
less than 1% between 1 and 13 mm depth while for a 0.035 mm thick copper-layered NPL
chamber, it varies by less than 1.3% between 6 and 16 mm depth. This indicates that for
the NPL chambers the C, factors become less sensitive to the change in electron energy
and therefore less sensitive to the change in the depth of the chamber.

0
a
U.
z
0
In
U
w
0.95 - I
1
I
5 6'. I
z ----t NACP I
0 I
0 I
w
- ----O---,NPL (18~) ' 9 :
--*-.
v) 0.90
0
0 NPL ( 3 5 ~ ) '.*. I
I
I

0.85 -r 1 I 1
*.'

1
+ I

Figure 8. C, factors for three ionization chambers at different depths in PMMA irradiated by a
5.64 MeV electron beam with a 10 x 10 em2 field and 95 cm SSD. normalized to the values
for 11 mm depth. For the NACP chamber, the dosexonversion factor was taken from the
lPSM (1992) addendum. The dose-conversion factors for the NPL chambers were calculated
according to (5) by comparisons with the NACP chamber.
 Investigation of a prototype plane-parallel chamber 1607

6. Summary and conclusions

We have investigated experimentally the prototype NPL design of plane-parallel chambers.

The C, factors for the NPL chambers have been derived by comparative measurements with
three designated planeparallel chambers, i.e. two NACP and one PrWIMarkus, following
the procedures recommended by the HPA (1985) code of practice. Other aspects related
to the prototype NPL chambers have also been investigated, including polarity effects,
chamber-cable effects and chamber angular response. The experimental results are consistent
with the Monte Carlo calculations of Ma and Nahum (1993b). The conclusions of this
investigation are as follows.

(i) The dose conversion factors C, for the new NPL chambers are nearly constant
over a certain range of electron energies depending on the thickness of the copper layers
used. Energyindependent C, factors are ideal from a practical point of view; for the NPL
prototype chambers, this constancy is obtained at the expense of having a large and rapidly
changing perturbation effect, which happens to almost compensate the variation of the
stopping-power ratio for water to air. The new chamber design deviates from the traditional
concept for chamber design, i.e. the preference to have p: N 1 and to tolerate the variation
of C, with energy due to the energy dependence of the stopping-power ratio.
(ii) All three prototype NEL
' chambers have a negligible polarity effect. However, there
is about a 1% variation in the chamber reading depending on which side of the chamber
faces the radiation source. This was attributed to the effect of photon attenuation and
electron scattering due to the slight difference in the area of the copper layer between the
m-electrode side and the collecting-electrode side of the chamber.
(iii) The prototype NPL chambers have a very high collecting efficiency, similar to
the NACP and PTW/Markus chambers. The change in the charge collected between 100
and 300 V for these three chambers studied was smaller than 0.3%. For the subsequent
experiments in this study the polarizing voltage used was 250 V.
(iv) The chamber-cable effect has been studied for the prototype NPL chambers. A
small (less than 0.5%) correction is required if the chamber is calibrated in phantom in a
6oCobeam and then used in phantom in an electron beam. This effect may be partially due
to the chamber cable and partially due to the increased attenuation of the scattered photons
in a 6oCo beam and the increased electron-scattering effect both in a photon beam and in
an elecwon beam because of the rectangular shape of the chamber body (i.e. the copper
layers). This effect may thus partially be eliminated by changing the chamber shape from a
rectangle to either a square or to a circle, or by removing much of the surrounding copper.
(v) The prototype NF'L chambers show a slightly higher response to electrons incident at
small angles compared to the NACP chamber but show a much lower response to electrons
incident at large incident angles. Caution must therefore be exercised in the measurement
of dose equivalent at different depths for electrons incident at different angles (ICRU 1992)
with these chambers.
(vi) Compared to an NACP chamber, the NPL chambers give smaller signals at small
depths in a phantom irradiated by electron beams but higher signals at depths beyond the
depth of maximum ionization. Due to the relatively smaller variation in C, for a certain
range of electron energies the C, factors at different depths in a phantom irradiated by
electrons of lower energies (e.g. 6 MeV) are less sensitive to change in depth around d,,,
than those for the NACP chamber. Strictly speaking, however, the C, values given in
table 4 are only valid for use at (or close to) the depth of maximum ionization.
1608 C-M Ma et a1

Acknowledgments

We would l i e to thank D Bums and R Sanders of the National Physical Laboratory,

Teddington, for helpful discussions and comments, and for supplying the NACP and NPL
prototype chambers. One of us (CMM) was funded by the NF’L through an extnmural
research agreement. AEN acknowledges the financial support of the Cancer Research
Campaign.

References
HOB J G, Buffa A, Perry D J, Ma I C and MacDonald I C 1979 Absorbed dose measurements using pwdlel-plate
polystyrene ionisation chambers in polystyrene phantoms Int. 3. Radial. Oncol. Bid. Phys. 5 2031-9
HPA 1985 Code of practice for electron beam dosimetry in radiotherapy Phys. Med. Biol 30 1169-94
ICRU 1992 Mwuremenr of dose equivalents from extemal photon and electron radiations ICRU Reporf 47
(Berhseda. M D lntemational Commission on Radiation Units and Measurements)
IPSM 1992 Addendum to the code of practice for electron beam dosimetry in radiotherapy (1985): interim
recommendations of the IPSM electron dosimetry working party Phys. Med B i d . 37 1477-83
Kuchnir F T and Reli C S 1992 Experimental values for P ~ Iand, Pwty.r~ for five parallel-plate ion chambers-?.
new analysis of previously published data Med. Phys. 19 367-367
Ma C-M 1992 Monte Carlo simulation of dosimeter response using transputers PhD Thesis University of London
(Institute of Cancer Research); Royal Marsden Hospirnl Physics Department IntemaI Report ICR-PHYS-Il92
Ma C-M and Nahum A E 1993a P l a n e - p d e l ionisation chambers in electron beams: Monte Carlo findings on
the perturbation factor Pmc. Int. Symp. on Measurement krsurmce in Dosimetry IAEA-SM-330/71 (Vienna:
I B A ) pp 481-93
- 1993b Investigation of the new prototype NF’L design of plane-parallel chamber Pmc. Inf. Symp. on
Meuumment Awurunce in Dosimefry IAEASM-33W4 (Vienna: M A ) p p 495-504
- 1994 Stem effect corrections for NE2561 and NE2571 chambers in medium-enerm . photon
. beams Phys.
Med B i d . submitted
Mausson L 0.Johansson K A and Svensson H 1981 Calibration and use of plane-parallel ionisation chambers for
[he dctcminuion of absorbed dose in electron beams A m Radio. On& 20385-99
Morris W T and O w n B 1975 An i o n h i o n c h p b e r for thcnpy-level d o s i m q of electron b c m Phys. Mrd
B i d 20 718-27
Nahum A E 1993 Pcnurbmon factors in dosimetry Ph)s. Med Biol. submiucd
Nahum A E,Thwaitcs D I md Andrea P 1988 Andyrir of the revised HPA docimevy prolocols P h y Med. B i d
33 923-38
SCS 1989 Code of pracocc for rhe dosrmetry of highenergy elecuon b e a n s Nederlondqe Commirse Voor
Stroltngdusir,nerrie ( N U ) Repurr 5 (Bilthovsn: NCS)
Ref1 C S and Kuchnir F T 1991 Measurement of the r e p l x x ” correction factor for pmllel-pl~techambers i n
electron fields .$fed Phys. 18 123743
Wmhmpcr F W, Thiercns H. Van der Plactsen A. De Wager C and Xlijnheer B 1991 Perturbation cormtion
factors for some ionization chambers commonly applied In elecmn b e m Phyr. Ned. BioL 36 1639-52