Shani G. Radiation Dosimetry: Instrumentation and Methods

Подождите немного. Документ загружается.

230 Radiation Dosimetry: Instrumentation and Methods

for 20-MeV electron beams. Differences in the energy cor-

rection factors between rods and chips of the same thickness

were negligible. The energy correction factors changed by

up to 4% for irradiation of TLD at depths other than at d

max

for a 5-MeV mono-energetic electron beam.

The energy correction factor is deﬁned as

(4.12)

FIGURE 4.8 Quality dependence of LiF TLD rods of different diameter calculated by Monte Carlo simulations. (From Reference [5].

With permission.)

FIGURE 4.9 Variation with energy of the mass energy–absorption coefﬁcient and the mass collision stopping power ratio of water

to LiF. (From Reference [5]. With permission.)

TL Co()/D

med

Co()

TL E()D

med

E()

---------------------------------------------



Ch-04.fm Page 230 Friday, November 10, 2000 12:01 PM

Thermoluminescent Dosimetry 231

which is the ratio of the light output (TL) per unit dose

in the medium for



-rays to the light output per unit

dose in the medium for an electron beam energy (E). If

is the average dose to LiF TLD material and assuming

that is directly proportional to the light output TL(E)

at any E, then

(4.13)

The energy correction factor can be used to determine

the dose as follows:

(4.14)

The quality-dependence factor is .

Figure 4.10 shows the comparison of the energy cor-

rection factor determined by experiments and Monte Carlo

simulations. There is agreement between experiments and

Monte Carlo simulations within 1.3% .

To investigate the effect of cavity size on the energy

correction factor, Monte Carlo simulations were carried out

with rods of diameter 1, 2, and 5 mm. Figure 4.11 shows

the results of these simulations. As the thickness of the rod

is increased, there is an increase in the energy correction

factor; that is, the deviation from unity increases. There is

an increase of about 2% in the energy correction factor of

2-mm diameter rods compared with that for 1-mm diameter

at high electron energy. At low energy this difference can

be up to 5%. There is a signiﬁcant difference in the energy

correction factor curve among the 1, 2, and 5-mm diameter

rods over the entire energy range.

Thermoluminescent characteristics of LiF:Mg,Ti from

three manufacturers were compared by Fairbanks and

DeWerd. [7] All were chips of size 3 mm

 3 mm  1 mm.

Fifty TLDs of this material were obtained from each of

the three manufacturers, Solon Technologies, Inc. (Solon),

Victoreen, Inc. (Victoreen), and Teledyne Isotopes (Tele-

dyne). The TLDs were compared on the basis of their

precision, sensitivity, thermoluminescent (TL) signal as a

function of energy, and linearity of TL signal as a function

of exposure.

The Solon and Victoreen chips provided good overall

standard deviations, 7.9% and 7.0%, respectively. But the

Teledyne standard deviation was nearly twice that of the

others, 13.8%. The annealing procedure causes a further

spread in the TL signals from a given set of chips, and

because an anneal was initially performed, the Solon and

Victoreen chips may have been within the mentioned spec-

iﬁcations.

All groups showed an over-response at low energies

followed by a ﬂat response at higher energies, as shown

in Figure 4.12a. The TL signal values have been normal-

ized to remove sensitivity differences and show the cor-

respondence among the manufacturers. Figure 4.13 shows

TL signal as a function of exposure.

The dose response of LiF:Mg,Ti (TLD-100) exposed to

15 and 35 kVp (8.0  0.1 and 8.1  0.1 keV effective energy,

respectively) x-rays and



-rays has been measured by

FIGURE 4.10 A comparison of Monte Carlo simulations and experimental determination of the energy correction factor of LiF

TLDs. (From Reference [6]. With permission.)

LiF

med

E()D

LiF

E()()

med

Co()D

LiF

Co())

-----------------------------------------------------



med

E() TL E()

med

Co()

TL Co()

-----------------------



1 f

Ch-04.fm Page 231 Friday, November 10, 2000 12:01 PM

232 Radiation Dosimetry: Instrumentation and Methods

Gamboa-de Buen et al. [8] in the dose interval from

(1.25.4)  10

Gy for x-rays, and from 0.14 to 850 Gy for



-rays. In both cases the total TL signal and glow-curve peaks

3 to 9 show supralinearity. The supralinearity function f(D)

is similar for both x-ray beams, except for peak 8, where a

30% difference is observed.

Although the onset and degree of supralinearity in

the response of LiF:Mg,Ti to photons is dependent on

the batch material, the experimental conditions (heating

rate and annealing procedure), and the glow peak tem-

perature (McKeever et al. 1995), it has been established

that the degree of supralinearity increases when the pho-

ton energy increases . The supralinearity increases with

the glow peak temperature. Jain and Ganguly [9] have

reported a lower degree of supralinearity for peak 8 than

for peak 7.

The glow curves were deconvoluted into peaks 3 to

9 (see Figure 4.14a) with the Harshaw/Filtrol software

FIGURE 4.11 Energy correction factors of LiF TLD rods of different diameters calculated by Monte Carlo simulations. (From

Reference [6]. With permission.)

FIGURE 4.12 TL signal as a function of half-value layer (HVL) for each group of TLDs showing an over-response at low energies

and a ﬂat response at higher energies. (From Reference [7]. With permission.)

Ch-04.fm Page 232 Friday, November 10, 2000 12:01 PM

Thermoluminescent Dosimetry 233

that uses the Podgorsak approximation to the Randall-

Wilkins ﬁrst-order kinetics peak shape. In all deconvolu-

tions, the width of peaks 6 to 9 were kept constant (in

order to reduce the number of free parameters, since a

typical glow curve does not uniquely determine all 21

parameters which describe the 7 peaks) and the position

(temperature) and height of all peaks were adjusted in

each ﬁt of the glow curves. In the deconvolution of the

x-ray glow curves, the width of peak 4 was also ﬁxed.

Table 4.2 shows the average values and standard devia-

tions found for the deconvolution parameters. The area

under the glow curve, after background and peak 2 sub-

traction with the deconvolution software, will be referred

to as the total TL signal.

Figure 4.14a shows a glow curve after irradiation with

x-rays and its deconvolution into peaks 3 to 9. Figures 4-14b

and c show a series of glow curves for different doses after

x- and



-ray exposure, respectively. The curves have not

been normalized, and the evolution of their shape as a

function of the dose indicates a strong dose dependence

of the relative contributions from the individual peaks.

Figure 4.15 shows the TL response for the total TL signal

and individual peaks as a function of dose for both types

of radiation. In general, the response follows a linear-

supralinear-sublinear behavior.

The supralinearity function f(D), deﬁned as

(4.15)

where F(D) is the TL signal per irradiated unit mass at

dose D, and D

is a low value of the dose such that F(D

)

is linear, has been calculated for the total TL signal and for

peaks 5 to 8, and it is shown for 8-keV x-rays in Figure 4.16a

and for



-rays in Figure 4.16b.

The results of a series of experiments performed to

determine some fading characteristics of TLD-100 (Har-

shaw Chemical Co.) were presented by Vasilache et al.

The experiment was performed using two batches of

dosimeters that were irradiated with x-rays simultaneously

and with the same dose. The results indicated that, after

the low-temperature peaks had faded out, the response of

FIGURE 4.13 TL signal as a function of exposure, all three groups showing linearity between 100 mR and 100 R and supralinearity

thereafter. A linear extension is provided to make this more apparent. Nonlinearity below 100 mR may be attributed to uncertainty

in reader and chip noise. Cessation of supralinearity above 10,000 R may be the beginning of chip damage. (From Reference [7].

With permission.)

fD()

FF()D

()D

------------------------



TABLE 4.2

Glow-Curve Deconvolution Parameters for 8.1-keV

x-rays (35 kVp) and



-Rays

8.1 keV x-rays



-rays

Peak

number

Temperature

(°C)

FWHM

(°C)

Temperature

(°C)

FWHM

(°C)

3 161  6 38  4 156  2 32  2

4 188  2 20.0 186  1 21  1

5 213  2 26  2 212  1 25  1

6 237  4 40.4 234  1 40.4

7 276  2 40.4 274  2 40.4

8 305  6 32.0 307  1 32.0

9 337  2 32.0 333  3 32.0

Source: From Reference [8]. With permission.

Ch-04.fm Page 233 Friday, November 10, 2000 12:01 PM

234 Radiation Dosimetry: Instrumentation and Methods

the dosimeters exposed to light became quite stable, with

a very low fading. The signal of the dosimeters kept away

from light continued to fade to a lower level than the signal

of the dosimeters exposed to light.

To analyze the evolution of each glow peak, the glow

curves were deconvoluted using an improved method of

gradients. To ﬁt the experimental curves, the well-known

Podgorsak approximation of the Randall-Wilkins equation

was used:

(4.16)

along with the relationship between



and T

(4.17)

FIGURE 4.14 (a) TLD-100 glow curves after irradiation with x-rays at 1.8  10

Gy. Dotted curves show the deconvolution into

peaks 3 to 9. Peak 2 and the high-temperature emission (b) were subtracted from the measured glow curve by the deconvolution

software. (b)TLD-100 glow curves after irradiation with 8-keV x-rays for different doses. In increasing order: 1.5

 10

, 7.4  10

2.3

 10

, 4.5  10

, 8.9  10

, 1.8  10

, 2.7  10

, and 5.4  10

Gy. (c) TLD-100 glow curves after irradiation with



-rays

at different doses. In increasing order: 1.7

 10

, 4.3  10

, 9.2  10

, 1.7  10

, 4.3  10

, and 8.5  10

Gy. (From Reference [8].

With permission.)

FIGURE 4.15 Response curves of TLD-100 exposed to (a) 8-keV (35-kVp) x-rays and (b)



-rays, for the total TL signal and

for peaks 3 to 9. The dotted lines correspond to linear responses. The measurements have been arbitrarily displaced for clarity. (From

Reference [8]. With permission.)

IT() I





------------------





------------------





expexp







()exp

Ch-04.fm Page 234 Friday, November 10, 2000 12:01 PM

Thermoluminescent Dosimetry 235

where I(T) is the peak intensity at the temperature T, I

is the intensity of the maximum, T

is the temperature of

the maximum,



is the heating rate, and s is the attempt

to escape frequency of the electrons in the traps. The full

width at half-maximum of the peak is given by:

(4.18)

where E is the depth of the energy trap and k is the Boltzman

constant.

The results of the ﬁrst experiment are presented in

Figure 4.17. The best ﬁt led to the following two functions

for the fading factor: [10]

FIGURE 4.16 Supralinearity function f(D) of the total TL signal and peaks 5 to 8 of TLD-100 exposed to (a) 8 -keV (35-kVp)

x-rays and (b)



-rays. Symbols show the measurements and curves as a guide to the eye. (From Reference [8]. With permission.)



()E f



() 0.06



0.19

1.191

Ch-04.fm Page 235 Friday, November 10, 2000 12:01 PM

236 Radiation Dosimetry: Instrumentation and Methods

for the dosimeters that were exposed to light, and

for the dosimeters that were kept in the dark. In both cases,

the fading factor is given by f

(



)  I(



)/I(



 0), where



) is the integral TL emission at the time



The second experiment (UV post-irradiation) led to a

drastic change in the shape of the TL glow curve after the

exposure to UV light, by comparison with the glow curve

of the TLD-100 detectors irradiated with x-rays only

(Figure 4.18a and b). As can be seen, peaks 4 and 5 are

now resolved and, furthermore, peak 5 now shows a higher

fading than peak 4.

Pre-irradiation annealing treatment at 150°C reduces

the sensitivities of peak 4 (197°C) and peak 5 (209°C),

whereas the intensity of peak 2 (108°C) is enhanced. The

TL intensities of peak 7 (271°C) and peak 10 (441°C)

remain practically unaffected (Bhatt et al.). [11]

Figure 4.19 shows an example of a deconvoluted curve.

The area under each peak of the deconvoluted curves is

being used for the calculation of sensitization factors and

intensities.

Figure 4.20 shows a typical glow curve for virgin as

well as for 150°C, 6-h pre-irradiation annealed samples.

For 150°C, 6-h pretreated samples, the intensity of peak

2 increases by a factor of 1.82, whereas the intensity of

TL peak 5 is reduced by a factor of 36. Another TL peak

at 235°C (peak 6) appears due to this temperature treat-

ment. TL peaks 7 and 10 are affected only slightly by this

treatment.

It is seen that for the 150°C pre-annealing treatment,

both peaks 4 and 5 decrease, possibly due to precipitation

of phase 6 LiF:MgF

. It may be noted that optical bleach-

ing and thermal annealing investigations have shown con-

vincingly the interrelationship between TL peaks 4 and 5

and the 310-nm absorption band. Thus, the two peaks are

expected to behave in a similar fashion during the pre-

irradiation annealing treatment.

The glow-curve characteristics of a single a crystal

LiF:Mg,Ti, grown by the Bridgman method, were studied

by Weizman et al. [12] as a function of Ti concentration

(3–14 ppm Ti). The glow curves were deconvoluted into

component glow peaks using mixed-order (MO) model

kinetics. The TL sensitivities and activation energies of

the glow peaks were studied in both post-irradiation

annealed and unannealed samples. In glow curves follow-

ing a 165°C

15-min anneal to remove peaks 2–4, the acti-

vation energy of peak 5 was observed to increase from

1.65  0.1 eV at the lowest Ti concentrations to 1.9 

0.1 eV from 8–14 ppm Ti concentration. Deconvolution

of the glow curve is shown in Figure 4.21.

The sensitivity of the various peaks (2–5) as a function

of Ti concentration is illustrated in Figure 4.22. The de-

pendence of the TL sensitivity of peak 5 on the square of

the Ti concentration would seem to indicate that the

TC/LC complex may be the result of the incorporation of

two Ti(OH)

or O complexes coupled independently to

two of the dipoles participating in the formation of the

Mg-Li

vac

trimer.

Thermoluminescence and optical absorption of



-ray-

and



-particle-irradiated LiF:Mg,Ti single crystals have

been studied by Bos et al. [13] The optical absorption

(OA) bands were measured during a heating regime which

FIGURE 4.17 The fading of the TL signal for TLD-100 detectors irradiated with bremsstrahlung x-rays. The continuous lines represent

the best ﬁt for the two sets of experimental points. Line

A (), exposed to light, (ax

)  C, error: 7.200  10

3

, a  6.151  10

2

, b

 1.939  10

1

, C  1.191. Line B (), stored in darkness, a exp(cx

d

), error: 7.492  10

3

, a  1.162, b  7.326  10

3

, c 

5.312  10

1

. (From Reference [10]. With permission.)



() 1.16 0.007



0.53

()exp

Ch-04.fm Page 236 Friday, November 10, 2000 12:01 PM

Thermoluminescent Dosimetry 237

mimics the glow-curve readout. This technique yields, in

principle, a more reliable correlation between glow peaks

and OA bands than the conventional procedure of sample-

annealing using the l

-l

stop

technique. Computerized anal-

ysis was used to resolve the OA spectra into Gaussian bands.

In f-irradiated samples, apart from the known 5.5-eV band,

a previously unnoticed 4.5-eV band has been recognized

which may be caused by defects that are active as com-

petitors in the recombination stage.

Optical absorption bands induced by ionizing radiation

arise due to changes in the trapping stage only, while a TL

glow curve is the result of both trapping and recombination

processes. Thermal treatments drastically alter the absorp-

tion bands. A commonly used technique to study the effects

of thermal treatments is to subject the sample to a pulse-

annealing sequence; i.e., the sample is heated to one tem-

perature, then cooled and measured, subsequently heated

a second time to a higher temperature, and so on. The record-

ing of the TL glow curve, however, is done by heating the

sample just one time without quenches and reheating in

between. In LiF:Mg,Ti there are many strongly overlap-

ping OA bands.

Figure 4.23 shows two thermoluminescence glow

curves of LiF:Mg,Ti after irradiation at room temperature

with 4.5-MeV and 30-keV



particles. The glow curves

(above 200°C) show great similarity, from which it is

concluded that in both cases the same trapping centers are

involved. This is plausible since the stopping powers do

not vary very much (7.4  eV for 4.5-MeV



particles and 5.8  eV for 30-keV



particles).

Figure 4.24 shows the optical absorption spectra of

LiF:Mg,Ti at different temperatures following irradiation

with gamma-rays and 30-keV



particles. The two radi-

ation qualities produce different spectra. At 50°C the OA

band at 5.0 eV (F center) is recognized. It is seen that

after



irradiation, the maximum of this band shifts to a

FIGURE 4.18 Glow curves for detectors irradiated only with x-rays (a) and exposed to UV after irradiation with x-rays (b). Both

detectors were read immediately after irradiation. The continuous line represents the best ﬁt obtained by computerized glow-curve

deconvolution. (From Reference [10]. With permission.)

1

Ch-04.fm Page 237 Friday, November 10, 2000 12:01 PM

238 Radiation Dosimetry: Instrumentation and Methods

FIGURE 4.19 Typical deconvoluted TL glow curve for LiF TLD-100, annealed at 400°C for 1 h, followed by 150°C, 6-h treatment and

irradiated with 1 kGy at 77 K. The peak positions are numbered according to the usual nomenclature. (From Reference [11]. With permission.)

FIGURE 4.20 Typical TL glow curves for LiF TLD-100. Virgin ( — ) and 400°C for 1 h, followed by 150°C, 6-h pre-annealing



-). The numbers 10

7

, 3  10

8

, etc. refer to the current ranges of the DC ampliﬁer. (From Reference [11]. With permission.)

Ch-04.fm Page 238 Friday, November 10, 2000 12:01 PM

Thermoluminescent Dosimetry 239

lower energy. After



irradiation, the 4.50-and 5.54-eV

OA bands are not present; their absence suggests that

they are not active during x-irradiation and, therefore,

one can expect a linear behavior of OA bands as a func-

tion of dose.

Figure 4.25 displays the glow curves plotted for LiF

and its mixtures with Li

(nonluminescent material).

The shape of the curve obtained for the composite mate-

rial applied for dosimetry coincides with that observed

for the thermoluminescent phosphor. However, the ther-

moluminescence output (i.e., the values of thermolumi-

nescence signal per unit of dose) appear to be somewhat

lower for mixtures due to the presence of additives to

phosphor.

Various LiF thermoluminescence dosimeters have been

investigated by Osvay and Deme [15] for selective assess-

ment of low and high LET radiation for space dosimetry

purposes. The aim of the investigations was to compare the

selectivity of the glow peaks of the Harshaw TLD-100, the

Polish LiF:Mg,Ti (type MTS-N) and the Austrian LiF-F

dosimeters using alpha, gamma, and thermalized neutron

radiations. Comparing TL responses using glow-curve

analysis of the peaks 4–5 (180–210°C) and on the high-

temperature peaks 6–7 (240–270°C) generated by high

LET radiation, it was found that separation of radiations

with different linear energy transfer (LET) was the most

promising in the case of Polish LiF:Mg,Ti (type MTS-N)

dosimeters.

Figure 4.26 shows the glow curves of TLD-100, LiF

(MTS-N) and LiF-F after 6-mGy gamma irradiation.

The gamma sensitivities were calculated from the mean

value of the TL responses. The sensitivity of LiF (MTS-

N) was found to be two and a half times higher compared

with TLD-100, and LiF-F was found to have nearly the

same sensitivity as TLD-100. The standard deviation of

dose readings for gamma-rays was about 5%.

The glow curves of TLD-100, LiF (MTS-N), and LiF-

F dosimeters irradiated by 24-mGy (volume-averaged)

alpha dose are shown in Figure 4.27. The glow curves

demonstrate the gamma sensitivity and the sensitivity to

5-MeV alpha particles of the LiF dosimeters investigated.

Comparing the TL responses, the following can be stated:

1. The glow peak of the gamma-irradiated LiF-F

single crystal dosimeter was found at a higher

temperature compared to TLD-100 and MTS-N

LiF dosimeters; the glow peak temperature of

LiF-F is about 240°C (Figure 4.26).

2. The effect of high LET alpha irradiation is con-

siderable for each type of dosimeter investi-

gated, but the qualitative change on the structure

of the glow curve was found to be the most

signiﬁcant for the LiF (MTS-N) dosimeter.

3. The LiF-F crystal does not show an explicitly

high temperature peak using alpha irradiation

in the dose range 1–40 mGy.

Comparing the glow curves of Figure 4.28, the fol-

lowing can be stated. First, the main peak (peaks 4–5) is

shifted to lower temperature values for an initial irradia-

tion made by neutrons in comparison with the reversed

condition, i.e., with an initial gamma exposure. Second, in

addition to this effect, there is a reduction in TL response

for an initial irradiation with neutrons.

FIGURE 4.21 Deconvoluted glow curve of LiF:Mg,Ti containing 80 ppm Mg and 10 ppm Ti. Analysis of this particular glow curve

resulted in the following activation energies: peak 2 (1.3 eV), peak 3 (1.1 eV), peak 4 (1.6 eV), and peak 5 (1.4 eV). (From Reference [12].

With permission.)

Ch-04.fm Page 239 Friday, November 10, 2000 12:01 PM