Shani G. Radiation Dosimetry: Instrumentation and Methods
Подождите немного. Документ загружается.
220 Radiation Dosimetry: Instrumentation and Methods
79. Cunningham, J. R. and Sontag., M. C., Med. Phys.,
7, 672, 1980.
80.
Ma, C-M. and Nahum, A. E., Phys. Med. Biol., 40, 45, 1995.
81.
Almond, P. R. et al., Med. Phys., 21, 1251, 1994.
82.
AAPM Task Group 21, Med. Phys., 10, 741, 1983.
83.
Gastorf, R. et al., Med. Phys., 13, 751, 1988.
84.
Rogers, D. W. O., Med. Phys., 19, 889, 1992.
85.
Rogers, D. W. O., Med. Phys., 19, 1227, 1992.
86.
Murali, V. et al., Phys. Med. Biol., 38, 1503, 1993.
87.
Huq, M. S. et al., Med. Phys., 20, 293, 1993.
88.
Reft, C. S. et al., Med. Phys., 21, 1953, 1994.
89.
Huq, M. S. et al., Med. Phys., 22, 1333, 1995.
90.
Almond, P. R. et al., Med. Phys., 22, 1307, 1995.
91.
Reynaert, N. et al., Phys. Med. Biol., 43, 2095, 1998.
91a.
Russell, K. R. and Ahnesjo, A., Phys. Med. Biol., 41,
1007, 1996.
92.
Aukett, R. J. et al., in AAPM Proc. No. 11, Kilovolt
X-ray Beam Dosimetry for Radiotherapy and Radiobi-
ology,
1997, 196.
93.
Ma, C-M. et al., Phys. Med. Biol., 39, 1593, 1994.
94.
Wickman, G. and Nystrom, H., Phys. Med. Biol., 37,
1789, 1992.
94a.
Burlin, T. E. and Chan, E. K., Solid state and chemical
radiation dosimetry, IAEA, 393, 1967.
95.
Johansson, B. and Wickman, G., Phys. Med. Biol., 42,
133, 1997.
95a.
Greening, J. R., Phys. Med. Biol., 9, 143, 1964; 9, 566,
1965.
96.
Mie, G., Ann. Phys. Lpz., 13, 857, 1904.
97.
Wickman, G. et al., Med. Phys., 25, 900, 1998.
98.
Hubbell, J. H. and Seltzer, S. M., NISTIR Report 5632,
U.S. Department of Commerce, 1995.
99.
Dasu, A. et al., Phys. Med. Biol., 43, 21, 1998.
100.
Wickman, G. and Holmstrom, T., Med. Phys., 19, 637,
1992.
101.
Johansson, B. et al., Phys. Med. Biol., 40, 575, 1995.
102.
PTW-FREIBURG Ionization Chamber Catalog, 1999.
103.
Radiation Oncology Source Book, MED-TEC Catalog,
1999.
104.
Radiological Dosimetry, NE-Technology Precision
Dosimetry Catalog,
1999.
105.
Analog Smart Portable Technical Manual, Eberline
Model ASP-1, 1997.
106.
Wangrovich, U., in: Proc. 20th Conf. of the Nuclear
Societies in Israel,
1999, 73.
107.
Kubo, H., Med. Phys., 20, 341, 1993.
108.
Attix, F. A., Med. Phys., 20, 735, 1993.
109.
Klevenhagen, S. C., Phys. Med. Biol., 36, 239, 1991.
110.
Gross, B., Topics in Applied Physics, 33, 217, 1987.
111.
MacDonald, B. A. et al., Phys. Med. Biol., 37, 1825,
1992.
112.
MacDonald, B. A. and Fallone, B. G., Phys. Med.
Biol.,
40, 1609, 1995.
113.
Schauer, D. A. and Links, J. M., Med. Phys., 20, 1371,
1993.
114.
Klerenhagen, S. C., Phys. Med. Biol., 36, 1013, 1991.
Ch-03.fm(part 3) Page 220 Friday, November 10, 2000 12:00 PM
221
4
Thermoluminescent Dosimetry
CONTENTS
I. Introduction ........................................................................................................................................................221
II. LiF:Mg, Ti TL Dosimeter..................................................................................................................................222
III. LiF:Mg, Cu, P Dosimeter ..................................................................................................................................252
IV. CaF:Tm(TLD 300) Dosimeter...........................................................................................................................263
V. Theoretical Aspects and Monte Carlo Calculations of TL Dosimetry .............................................................271
VI. Miscellaneous TLD............................................................................................................................................283
VII. Electron Spin Resonance Dosimetry.................................................................................................................292
References .......................................................................................................................................................................298
I. INTRODUCTION
Thermoluminescent dosimetry is used in many scientific
and applied fields such as radiation protection, radiother-
apy clinic, industry, and environmental and space research,
using many different materials. The basic demands of a
thermoluminescent dosimeter (TLD) are good reproduc-
ibility, low hygroscopicity, and high sensitivity for very
low dose measurements or good response at high doses
in radiotherapy and in mixed radiation fields. LiF is used
for dose measurements in radiotherapy since the effective
atomic number of 8.3 is close to that of water or tissue.
Lithium tetraborate is more tissue-equivalent than LiF, but
it is deliquescent (absorbs moisture from the atmosphere)
and its stored signals fade rapidly. Its use is therefore only
worthwhile for x-rays, where the closeness of its effective
atomic number of 7.3 to tissue outweighs the disadvan-
tages. Calcium sulphate has an effective atomic number
of 15.6 and is therefore much less tissue-equivalent, but
its effective atomic number is quite close to that of bone.
It is very sensitive and therefore can be used for protection
dosimetry. Calcium fluoride has an effective atomic number
of 16.9 and is also used for protection dosimetry, as it is
also very sensitive.
TLDs are relative dosimeters and therefore have to
be calibrated against absolute dosimetry systems such
as a calibrated ion chamber. A
60
Co gamma source is
generally used. Due to their small size, TLDs are con-
venient for dose-distribution measurements in medicine
and biology.
The use of TLDs in electron-beam dosimetry is inher-
ently more complicated than its use in photon dosimetry
since, for each incident electron beam energy, one obtains
a different dose response. As the electron beam penetrates
into material, it gradually loses its energy so that both
the dose and energy vary with depth in material, making
an accurate dose measurement with TLD more difficult.
Fading is an important phenomenon when dosimeters
are used for environmental or personal monitoring, which
involves reading after long periods of irradiation.
Investigation of Linear Energy Transfer (LET) depen-
dence of TL dosimeters indicates that for all commonly
used materials, the sensitivity (TL response/absorbed
dose) decreases with increasing LET. TLD-100 exhibits a
supralinear dose response above about 2 Gy.
Dosimetry of mixed neutron-gamma radiation is usu-
ally performed with a pair of and dosimeters
exhibiting various thermal neutron sensitivities. The TL
glow in the TLD is primarily generated by the
densely ionized secondary alpha particles produced in the
dosimeter via a (
n
,
)
3
H reaction. The common assump-
tion that TLD-700 (99.99% ) has a negligible thermal
neutron response relative to its photon response can lead
to errors. The TL in a TLD-700 is created by the
(
n
,
) reaction (
th
= 0.033 b) as well as the energy
transferred to dosimeter material via elastic and inelastic
scattering of the impinging neutrons.
By peak analysis of the glow curve of CaF
2
, it is
possible to separate the low and high LET components of
a mixed radiation field.
An LiF:Mg,Cu,P dosimeter is useful for low-dose mea-
surement because of its high sensitivity compared to the
TLD-100.
The use of computerized glow-curve analysis (GCA)
methods has become a normal practice. The application
of deconvolution-based GCA methods to complex curves
provides information on the parameters of each individual
peak.
LiF
6
LiF
7
LiF
6
L
i
6
LiF
7
L
i
7
Li
8
Ch-04.fm Page 221 Friday, November 10, 2000 12:01 PM
222
Radiation Dosimetry: Instrumentation and Methods
Knowledge of backscatter factor (BSF) for low-
energy x-ray is essential in diagnostic radiology as well
as radiotherapy dosimetry to determine the absorbed dose
at the surface of the patient.
For proton and alpha particles, a predominant contri-
bution to total stopping power comes from the electronic
stopping power due to inelastic collision with electrons.
Electron spin resonance (ESR) investigation of TL
samples can, in some cases, help identification of an ongo-
ing mechanism of luminescent centers
II. LIF:MG, TI TL DOSIMETER
LiF:Mg,Ti (TLD-100), a nearly tissue-equivalent material
of good sensitivity, is perhaps the best choice for radiation
dosimetry. It has been commonly used for medical pur-
poses because it is a nearly tissue-equivalent material
(effective atomic number of 8.2 compared to 7.4 for tis-
sue). The weakest characteristic of TLD-100 for radiother-
apy dosimetry is accuracy. The existence of fading and
sensitivity changes also complicates the situation. The TL
response of TLD-100 is not linear for large doses and its
efficiency depends on the radiation field. Studies have
shown a dependence of the TL efficiency and the supra-
linearity on the photon energy.
The energy correction factor of LiF TLD calibrated in
60
Co
-rays and used for measurements in electron beam is
rather confusing. Some show no energy dependence relative
to
60
Co
-rays, while others show correction factors in the
range 1.03–1.1.
An important use of LiF is dosimetry of mixed
neutron-gamma radiation. It is usually performed with
a pair of
6
LiF and
7
LiF dosimeters taking advantage
of the difference in cross section of the two Li iso-
topes.
Muniz et al. [1] studied LiF TLD-100 performance
for mailed dosimetry in radiotherapy, using glow-curve
analysis for TL evaluation and reusable chips. The use
of such a glow-curve analysis method permitted deter-
mination of the evolution of each peak during storage at
room temperature. Delayed TL measurements, including
controlled time delays between the steps of the TL cycle,
were employed for that determination. Storage took
place in a room whose temperature was controlled only
during working hours and set at 22°C. During the night
the temperature dropped in winter to 15°C and rose in
summer to 28°C.
A computer code developed in CIEMAT for the anal-
ysis of complex glow curves, resolving the individual
peaks, was employed in the experiment. The program fits
the experimental points using the first-order kinetics
expression for a single peak.
Figure 4.1 shows the kind of result provided by the
computer program for a typical prompt LiF TLD-100 glow
curve as produced by the preparation treatment employed.
The measurement identification and the computing time
FIGURE 4.1
An analyzed TLD-100 glow curve: squares, experimental points; continuous lines, fitted glow curve and resolved
individual peaks. The curve identification, the time required for the analysis in seconds, and the fitted peak parameters are presented.
(From Reference [1]. With permission.)
Ch-04.fm Page 222 Friday, November 10, 2000 12:01 PM
Thermoluminescent Dosimetry
223
can be seen in the figure with the fitted characteristic
parameters for each peak (2–5): the temperature of the
maximum,
T
m
; the maximum intensity,
I
m
; the activation
energy,
E
; and the frequency factor,
s
. The areas of the
different peaks, and the sums of some of them (5
4 or
5
4
3) are also given.
Figure 4.2 presents the results obtained in two tests
on the dose against TL dependence near the 2 Gy reference
dose using a single group of detectors. Between these tests
the group was employed several times for different types
of measurement. Figure 4.2a shows this dependence in the
range 0.5–4.0 Gy and Figure 4.2b shows the same depen-
dence but in a reduced range, 1.5–2.5 Gy. In Figure 4.2
the results are expressed as TL Gy
1
against absorbed dose
in gray instead of the more usual dose against TL plot, in
order to appreciate better the dependence on absorbed
dose. The uncertainty bars correspond to two standard
deviations (mean value) of the group results at each dose
FIGURE 4.2
TL dependence on absorbed dose, expressed as TL per unit dose against dose, (a) in the range 0.5–4.0 Gy and (b) in
a reduced range, 1.5–2.5 Gy. Two independent sets of results (open and closed symbols) are presented, obtained from the same group
of five detectors. Circles, group mean value; uncertainty bars, two standard deviations of the mean value. In each panel the fitted
dose against TL algorithm for the two sets of data is presented: in (a), quadratic dependence, and in (b), linear dependence. Each
pair of open and closed points corresponds to the same absorbed dose value. (From reference [1]. With permission.)
Ch-04.fm Page 223 Friday, November 10, 2000 12:01 PM
224
Radiation Dosimetry: Instrumentation and Methods
value. Apart from the closeness of the data obtained in the
two tests, a certain supralinearity can be appreciated in
Figure 4.2a. A higher TL Gy
1
value is obtained at 4.0 Gy
than at 0.5 Gy. Each set of data can be well-fitted by a
quadratic expression:
(4.1)
The quadratic term accounts for supralinearity, so
C
is
always negative and very small. The broken line represents
the TL Gy
1
against absorbed dose variation as deduced
from the fitted quadratic expression.
Figure 4.3 presents the TL results at 2 Gy obtained
after different storage intervals normalized to the corre-
sponding prompt values. Data are presented for two inte-
gration regions, including three peaks (3
4
5) (Figure
4.3a), or only two (4
5) (Figure 4.3b). Open symbols
refer to storage before irradiation (SB) and closed ones
refer to storage after irradiation (SA). Crosses are for the
groups irradiated in the middle of storage (M). In the
figure the different evolution of the SA and SB groups can
be appreciated. For the three-peak integration criterion,
the groups irradiated at the beginning of storage (SA)
show a lower TL signal than the groups irradiated at the
end (SB) of the same storage interval. This may be inter-
preted in terms of fading, affecting filled traps during
storage. Nevertheless, the situation is more complex as
both groups SA and SB tend to decrease when the storage
duration increases, and obviously fading cannot affect
empty traps (the SB groups). Hence, there is also evidence
FIGURE 4.3
TL evaluation during storage at room temperature, integrating under several peaks. (From Reference [1]. With permission.)
Dose Gy()ABTL CTL
2
Ch-04.fm Page 224 Friday, November 10, 2000 12:01 PM
Thermoluminescent Dosimetry
225
of trap effects. The evidence is stronger if the integration
region is limited to peaks 4 and 5 only, omitting peak 3.
In this case, the SA groups (filled traps during storage)
systematically give higher TL values than the SB groups
(empty traps during storage) and also higher values than
the reference values in the prompt glow curves. Clearly,
this behavior suggests that the dominant processes affect-
ing peaks 4 and 5 are trap processes, causing an increase
in sensitivity, and not charge leakage or fading processes.
Peak 3, lying at lower temperatures than peaks 4 and 5,
is clearly affected by fading (and also by trap effects).
A general-purpose computer code thermoluminescent
detector simulator (TLD-SIM) was developed by Branch and
Kearfott [2] to simulate the heating of any TLD type using
a variety of conventional and experimental heating methods.
In order to calculate the TL intensity, it is first neces-
sary to calculate the temperature
T
as a function of position
and time, which is given by the second order partial dif-
ferential heating equation:
(4.2)
where
is the density,
C
is the heat capacity,
is the
thermal conduction coefficient, and
S
is the heat source
term for internal heat generation. The Laplacian is numer-
ically evaluated using a second-order linear approxima-
tion, and the above equation simplifies to:
(4.3)
where
q
is the heating term for laser heating, given by
(4.4)
Here
is the absorption coefficient,
r
f
the reflectivity,
I
0
the laser power, and
z
the depth into the chip. It should
be noted that
q
0 for all other types of heating.
Once the temperature is determined for each volume
element, the glow intensity is then calculated by the first
order Randall-Wilkins model: [3]
(4.5)
Here
S
m
is the frequency factor of oscillation of the par-
ticular trap,
E
m
is the trap depth for that trap,
D
(
x
,
y
,
z
) is
the depth-dose profile, and
k
is the Boltzmann constant.
Trap data for LiF is given in Table 4.1. [2]
C
Tx,y,z,t()
t
----------------------------
2
Tx,y,z,t() S
T
C
-------
t
2
T
qx,y,z,t()
C
------------------------
t
qx,y,z,t()
1 r
f
()I
0
e
z
Ix,y,z,t()s
m
Dx,y,z()
E
m
kT x,y,z,t()
----------------------------
exp
s
m
t
E
m
kT x,y,z,t()
----------------------------
td
expexp
TABLE 4.1
Values for Input Trap and Thermophysical Parameters
Parameter
Value Units
Density 2.4 1/cm
Heat capacity 0.431 cal/(gK)
Thermal conduction coefficient 0.007 cal/(s cm K)
Absorption coefficient 4.0 1/cm
Reflectivity 0.0 none
Free convection heat transfer 2.80
10
–4
cal/(s cm
2
K)
coefficient
Ambient temperature 300 K
Chip thickness 0.38 cm
Boltzmann constant 8.62
10
–5
eV/K
Energy depth of 2nd LiF trap 1.13 eV
Energy depth of 3rd LiF trap 1.23 eV
Energy depth of 4th LiF trap 1.54 eV
Energy depth of 5th LiF trap 2.17 eV
Frequency factor of 2nd LIF trap 6.10
10
11
1/s
Frequency factor of 3rd LIF trap 4.00
10
11
1/s
Frequency factor of 4th LiF trap 7.30
10
13
1/s
Frequency factor of 5th LIF trap 4.00
10
19
1/s
Source:
From Reference [2]. With permission.
Ch-04.fm Page 225 Friday, November 10, 2000 12:01 PM
226
Radiation Dosimetry: Instrumentation and Methods
LiF:Mg,Ti (TLD-100) extruded ribbons and crystals
were exposed by Carrillo et al. [4] to monoenergetic pho-
tons of 275-2550 eV energy to determine their potential
usefulness as radiation dosimeters for radiobiology exper-
iments. The undesirable effect of air annealing increased
with decreasing photon penetration in the dosimeter.
Under certain experimental conditions, UV radiation pro-
duced anomalous bleaching of high-temperature traps.
The crystals and the chips presented a supralinear
response. Supralinearity factors were determined to be on
the order of 1.5 for crystals and 1.7 for chips.
Extruded chips were 3.2
3.2
0.9 mm in size.
Crystal samples had varying sizes as a result of the cleav-
ing procedure; however, they were cleaved to be as similar
in size as possible to the chips. The entire crystal was
annealed at 400°C in air for 1 h, after which time the
crystal was removed from the oven and placed on an
aluminum block to reach room temperature. Then the
crystal was placed in an 80°C oven for 24 h. After this
the crystal was cleaved.
Figure 4.4 presents glow curves for a TLD-100 crys-
tal exposed to 193 nJ of
137
Cs
-rays and to a monoen-
ergetic photon beam of 275 eV energy. For
137
Cs this
exposure (about 1 Roentgen) results in an absorbed dose
in the crystal of 8.14 mGy. Glow curves for 2550, 1500,
1200, 900, 730, 600, 500, and 400 eV exposures with
193 nJ of total energy deposited had the same shape as
the one for 275 eV. The glow curves were normalized to
the same height for the main peak at 200°C. The total
energy deposited in each case was also 193 nJ. Photons
of 275 and 730 eV energy are the least penetrating ones
in LiF.
FIGURE 4.4
Glow curves of LiF crystals exposed to
137
Cs
-rays (a) and to 275-eV photons (b). Glow curves following 2550–400 eV
irradiations were similar to those shown for 275 eV. Data are normalized at the dosimetric peak at 200°C. (From Reference [4].
With permission.)
Ch-04.fm Page 226 Friday, November 10, 2000 12:01 PM
Thermoluminescent Dosimetry 227
Figure 4.5 shows the TL response as a function of
total energy deposited in the He-annealed crystals for the
photon energies used. The TL response in both cases
corresponds to the integrated area of the glow curve
between 130°C and 215°C, while the total energy depos-
ited was calculated by using the photodiode measure-
ment. The integrated TL response was a linear function
of energy deposited for 275, 730, 900, 1200, 1500, and
2550 eV photons. However, for exposures with photon
beams of 400, 500, and 600 eV, a “saturation-like” effect
is noticed.
The effect of the annealing atmosphere was found to
be crucial for reproducible TL output and glow curves.
Three annealing treatments were tested by Carrillo et al.
Figure 4.6 shows the glow curves of crystals annealed in
helium and in air for 1 and 23 h, when exposed to 275-eV
photons. The adverse effect of an air anneal is clearly seen
as an overall reduction in the height of the peaks, mainly
the dosimetric 200°C peak. For crystals annealed in air
for 23 h, the reduction is even more pronounced. Similar
adverse behavior for the air-annealed crystals and chips
was noticed for all other photon energies.
The quality dependence of LiF TLD in megavoltage
photon beams with qualities from
60
Co
-rays to 25-MV
x-rays has been studied experimentally by Mobit et al.
[5] against ion chamber measurements and theoretically
by Monte Carlo simulation using the EGS4 Monte Carlo
code system. The experimental findings are that the
energy dependence of 1-mm thick TLD-100 (micro-rods
and chips) on average decreases slowly from 1.0 for
60
Co
FIGURE 4.5 Integrated TL response of TLD-100 crystals against energy deposited. Irradiation energy as marked. (a) employed the
mylar or Be window, (b) employed a thin Al window. (From Reference [4]. With permission.)
Ch-04.fm Page 227 Friday, November 10, 2000 12:01 PM
228 Radiation Dosimetry: Instrumentation and Methods
-rays to 0.989 1.3% for 6-MV x-rays (
0.685) and to 0.974 1.3% for 25-MV x-rays(
0.800) relative to
60
Co
-rays. The Monte Carlo results
vary from 0.991 0.9% for 6-MV x-rays to 0.978
0.8% for 25-MV x-rays. Differences between chips and
micro-rods were negligible, and there was no difference
in the energy dependence between TLDs irradiated in
water or Perspex (PMMA).
When a detector is placed in a medium exposed to
ionizing radiation in order to measure dose, it forms a
cavity in that medium. The cavity is generally of different
atomic number and density from the medium. Cavity the-
ory gives the relation between the dose absorbed in the
medium (
D
med
) and the average absorbed dose in the detec-
tor or cavity ( ):
(4.6)
where
f
med,cav
is a factor that varies with energy, radiation
type, medium, size, and composition of the cavity. For a
cavity that is small compared with the range of the elec-
trons incident on the cavity in electron and photon beams,
the Bragg-Gray relation applies:
(4.7)
where
S
med,cav
is the average mass stopping power ratio of
the medium to the cavity. For a cavity that is large com-
pared to the range of electrons incident on it in a photon
beam, the dose in the medium can be obtained from the
mass energy–absorption coefficient ratio of the medium
to the cavity material:
(4.8)
where (
en
) is the ratio of the mass energy–absorption
coefficients, medium to cavity, averaged over the photon
energy fluence spectrum present in the medium. This
expression completely neglects any perturbation effects or
interface effects that may occur by the introduction of the
detector material into the uniform medium.
Burlin proposed a general cavity theory for photons for
all sizes which approaches the Spencer-Attix theory in the
small-size limit and the ratio of the mass energy–absorption
coefficient for very large cavities. According to this theory,
(4.9)
where d is a weighting factor which gives the contri-
bution to the total dose of medium-generated elec-
trons, and (1 d ) is the contribution to the total dose
from electrons generated by photon interaction in the
cavity.
The quality dependence factor or response ( ) is
defined as
(4.10)
FIGURE 4.6 Glow curves of He- and air-annealed LiF crystals exposed to 275 eV photons. (From Reference [4]. With permission.)
TPR
10
20
TPR
10
20
D
cav
D
med
D
cav
f
med cav,
f
med,cav
s
med cav,
f
med,cav
en
--------
med cav,
f
med cav,
ds
med cav,
1 d()
en
--------
med cav,
F
co
X
F
Co
X
TL X()D
med
X()
TL Co()D
med
Co()
---------------------------------------------
Ch-04.fm Page 228 Friday, November 10, 2000 12:01 PM
Thermoluminescent Dosimetry 229
where TL(X)D
med
(X) is the light output (TL) per unit
dose in a medium for the beam quality X of interest.
TL(Co)D
med
(Co) is the light output per unit dose in the
same medium for
60
Co
-rays. If D
LiF
is the dose to TLD
material, then assuming D
LiF
is directly proportional to the
light output TL(X) at any X, one can write
(4.11)
The assumption of proportionality between the light out-
put and the dose to the TLD ignores the possibility of sensi-
tivity changes in the TLD which depend on interactions
between trapping centers. Thus, supralinearity and LET
effects will not be accounted for. However, for megavoltage
doses under 1 Gy, these are unlikely to be a problem. The
energy correction factor to correct for the under-response
of a photon beam (
X) relative to
60
Co
-rays is
where are the quality dependence factor of
chips and micro-rods, respectively, and is the aver-
age energy correction factor of chips and micro-rods of
the same thickness.
Figure 4.7 shows the comparison between Monte Carlo
simulation and experimental determination of the quality
dependence factor . The agreement is within 0.7%.
Comparisons have been made for micro-rods, with water
as the irradiation medium. In all cases, the Monte Carlo
results show a lower deviation from unity of the energy
dependence factor compared to the experimental values.
The size of the cavity (thickness of TLD) is critical to
whether it behaves as a small cavity, an intermediate, or even
a large cavity. To investigate the effect of cavity size on the
energy correction factor, simulations were carried out by
Mobit et al. with micro-rods of diameter 1, 2, and 5 mm.
Figure 4.8 shows the results of such simulations. There is
no significant difference between 1-mm and 2-mm diameter
micro-rods. The deviation from unity is, however, smaller
with 5-mm diameter micro-rods. The results seem to suggest
that thicker LiF TLD are better for photon dosimetry than
1-mm diameter LiF TLD, provided that high spatial resolu-
tion is not required, e.g., at the edges of the field.
The mass collision stopping power ratio and the mass
energy–absorption coefficient ratio are shown in Figure 4.9.
The energy correction factor of LiF TLDs in mega-
voltage electron beams was calculated by Mobit et al.
[6] The experiments show that the energy correction
factor of 1-mm thick TLD-100 has an average for both
rods and chips which varies from 1.036
1.3% (1
SD) for 4-MeV electron beams to 1.021 1.3% (1
SD) for 20-MeV electron beams for measurement per-
formed at d
max
in PMMA (Perspex). The results of the
Monte Carlo simulations were within 0.6% of the
experimental results and ranged from 1.041 0.9%
(1 SD) for 2-MeV electrons to 1.028 0.8% (1 SD)
FIGURE 4.7 Comparison of Monte Carlo simulations and experimental determination of the quality dependence factor for LiF TLD.
(From Reference [5]. With permission.)
F
Co
X
D
med
D
LiF
()
Co
D
med
D
LiF
()
X
--------------------------------
f
Co
X
1F
Co
X
F
X
chip Co
F
X
rod Co
,
f
X
ave Co
F
Co
X
Ch-04.fm Page 229 Friday, November 10, 2000 12:01 PM