ON THE USE OF LIF TLD-600 IN NEUTRON-GAMMA

MIXED FIELDS
A. Delgado

, J. L. Mun iz, J. M. Go mez Ros, A. M. Romero and R. Rodr guez

Radiation Dosimetry Unit, CIEMAT, Avda Complutense 22, 28040 Madrid, Spain
A new procedure allowing the separate estimation of neutron and gamma dose in mixed radiation elds has been developed
in our laboratory. In this communication, a description of the main features of the discrimination procedure and some
preliminary results obtained by its use are presented. The procedure is based on the signicantly different structure of the
glow curve of LiF TLD-600 produced by neutron and gamma radiation. The use of peak resolving numerical methods,
sometimes called deconvolution, for the analysis of the glow curves from controlled irradiations at absorbed doses in the
range 10300 mGy with different neutron and gamma proportions, permits to quantify the differences peak by peak, also
characterising the well-known neutron quasi-exclusive contribution to the high temperature region, above peak 5. From this
study, it was possible to propose a n/c TL factor by which the respective doses can be estimated through a simplied analysis,
not peak resolving, of the particular features of the glow curves obtained in eld measurements. A rst set of rather
satisfactory results have been obtained by irradiating TLD-600 together with TLD-700 chips using Am-Be sources with
different degree of moderation and using lead absorbers to change the gamma component. This component is directly
measured by the TLD-700 detectors, allowing the testing of the gamma estimation reached by the discrimination procedure
applied to the TLD-600 glow curve.
INTRODUCTION
6
LiF-based thermoluminescent dosemeters (TLDs)
are good thermal neutron detectors and can be a
convenient alternative for the development of dose-
meters based in the moderation method. For those
applications in which integrating dosemeters are
suitable, the comparatively small but sensitive TLDs
can offer good performance when employed in
such moderated systems. Especially, adapted TLD
procedures for mixed neutron-gamma dosimetry
have been developed allowing a reasonable discrimi-
nation capability for the two dose components
(1)
.
Some of them are based on the different sensitivity
to photons and neutrons of the low and high tem-
perature regions (HTRs) of the LiF TLD-600 glow
curve. Nevertheless, the pair method, combining
the neutron sensitive TLD-600 and the insensitive
TLD-700 forming a paired detector, is perhaps
the more common TL procedure employed for the
dosimetry of mixed elds.
CIEMAT is engaged in the construction of a
multidetector moderated dosemeter for area moni-
toring, based on a single sphere allocating inside
TLDs in various positions with different moderation
permitting to estimate a spectral correction factor
to improve the accuracy of ambient dose determina-
tions. A rst prototype has been recently assembled
and some preliminary and encouraging results have
been obtained with pairs of TLD-600/700 as ther-
mal detectors
(2)
. Attempts to use in our laboratory
the low/high temperature areas ratio to estimate the
neutron and gamma contributions from only the
TLD-600 glow curve were not successful, mainly
because of the difculty in estimating the area of
the HTR properly due to neutrons. The gamma
sensitivity in that region is not negligible and can
be different among different batches of detectors.
In addition, the HTR is prone to present signals not
caused by radiation, that are also depending on the
particular heating system employed for readout. All
these factors complicate the use of the HTR area
for dosimetry, particularly for the estimation of
low neutron doses. Nevertheless, the attractive of a
method for mixed eld dosimetry based exclusively
on a single type of detector is high, as compared to
the pair method only half of the detectors are
required.
For the CIEMAT multidetector system, the reduc-
tion in the number of detectors is a considerable
advantage, and for this reason, a new approach has
been developed in our laboratory trying to extract
separately the neutron and the gamma doses from
the TLD-600 glow curve, and not relaying exclu-
sively in the problematic HTR. The new approach
is based on the different glow curve structure pro-
duced by gamma and by neutron radiation, which
can be well appreciated with the use of peak resolv-
ing glow curve analysis methods. The two reference
structures for pure gamma and for nearly pure
neutron elds can be employed to estimate, through
a minimisation process, the relative contribution of
each kind of radiation in an n/c mixed eld of
unknown composition.
This communication describes the main features
of the new procedure, also presenting some prelim-
inary results obtained in an experiment using n/c
elds in controlled proportions. The aim was to

Corresponding author: antonio.delgado@ciemat.es

Radiation Protection Dosimetry (2007), Vol. 125, No. 14, pp. 327330 doi:10.1093/rpd/ncm210
Advance Access publication 19 June 2007
The Author 2007. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

a
t

P
u
r
d
u
e

U
n
i
v
e
r
s
i
t
y

L
i
b
r
a
r
i
e
s

A
D
M
N

o
n

J
u
n
e

3
,

2
0
1
4
h
t
t
p
:
/
/
r
p
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

check the validity of the new procedure and its range
of application in terms of gamma and neutron
absorbed doses.
EXPERIMENT
LiF TLD-600 chips (33 0.9 mm
3
) were employed
in the experiment. Occasionally, also TLD-700 chips
of the same dimensions were employed to allow an
independent estimation of the gamma component in
a mixed eld. Linear readout at 5

C/s up to 350

C
was always employed. Detectors were reused simply
after readout without any external annealing, show-
ing very good stability and reproducibility.
TLD detectors were irradiated in terms of absorbed
dose for gamma rays in the range between 1 and
200 mGy, and for neutrons between 10 mGy and
up to 2.5 Gy. For neutron irradiation, a highly mod-
erated (parafn) Am-Be source was employed and
the thermal neutron doses estimated by comparison
with the readings of an instrument LB 6411 calibrated
in terms of Ambient Dose at PTB. The Am-Be source
produced the irradiation position a weak gamma dose
rate of the order of 1.5 mGy/h together with a neu-
tron dose rate estimated as 100 mGy/h. Additional
gamma irradiations were imparted exposing the neu-
tron irradiated detectors to one of the 137 Cs beams
of the CIEMAT Secondary Standard laboratory.
In this way, mixed elds of different n/c proportions
were simulated, presenting ratios for the neutron to
gamma-induced TL signals within the range 1/50
to 10. Both neutron and gamma irradiations in this
demonstration exercise, have been made in very spe-
cic and somewhat simplied conditions respect to
practical situations in which the backscattered radia-
tion may vary depending on the geometry. This point
should be taken into account when applying the
method to routine dosimetry.
A peak resolving glow curve analysis code using
rst order kinetics expression and a Levenberg-
Marquadt minimisation algorithm has been employed
for the analysis of the TL curves obtained in the
experiment. This method allows study the evolution
of the peak distribution as a function of the n/c
relative contributions and is the basis of the discri-
mination procedure. The glow curve analysis code
was developed in our laboratories and has been in
very satisfactory use for more than a decade. Its per-
formance was checked in an intercomparison of
glow curve analysis methods with excellent results
(3)
.
As indicated above, the n/c discrimination is based
on the different and characteristic glow curve struc-
ture produced by pure gamma and pure (really nearly
pure) neutron irradiation. The glow curve obtained
in a mixed eld is then analysed in terms of these
two distributions through a minimisation procedure
in which the free parameter is just the proportion of
the two reference curves. The minimisation is on the
differences between the experimentally determined
glow curve and the synthetically produced, adding
the two reference glow curves and varying their rela-
tive proportion in an iterative way. In all the cases, the
iteration process has found convergence and produ-
cing rather satisfactory estimation of the neutron and
gamma contributions, and this in a broad range of
the relative proportions in different mixed radiation
elds.
RESULTS AND DISCUSSION
The rst step in the experiment consisted in the
determination and the analysis of the reference
glow curves to determine their respective structure.
The determination of the peak distribution for
the gamma curve was straightforwardly achieved.
It was obtained simply irradiating the TLD-600
detectors in the 137 Cs beams and analysing the cor-
responding glow curves. For neutrons, the process is
somewhat more complicated due to the gamma con-
tamination of the AmBe source. In order to estimate
the gamma dose contribution, some TLD-700 detec-
tors were exposed to the AmBe at the irradiation
position, a xed position employed for the whole
experiment. The gamma contribution was estimated
as 1.5 mGy/h, i.e. a weak contribution nearly 1/10
of the neutron induced TL signals. It was accounted
for when analysing the glow curves of the TLD 600
detectors irradiated with the AmBe source to obtain
the peak distribution of the reference glow curve for
neutrons.
Figure 1 presents the reference glow curves after
20 mGy (137 Cs), curve (a); and after 20 min expo-
sure to the AmBe source (approx. 20 mGy), curve
(b). The glow curve corresponding to the weak
gamma contribution in curve (b) (0.5 mGy) is also
presented for comparison. The already known dif-
ferences between the gamma and neutron induced
glow curves
(4)
can be appreciated in Figure 1: An
increased sensitivity of the HT region to neutrons
and a substantially decreased peak 4 in the neutron
curve compared to the gamma curve. In addition,
the peak 3 relative intensity is systematically higher
for neutrons than for gamma irradiation.
The dose estimator employed in the experiment
was the added area of the tted peaks 3, 4 and 5,
P
345
, and for every curve, each peak, i, is charac-
terised by the proportion of its area, R
i
, in relation to
the P
345
area of that curve: R
i
P
i
/P
345
, where
i represents either peak 3, 4 or 5. The stability of
the R
i
values for the reference curves, gamma and
neutrons, were checked by repeated irradiation and
TL measurements, and found satisfactory and con-
venient for the purpose of our work. For gamma
curves, ten successive measurements at 20 mGy pro-
duce consistent R
i
values within 12%, while for
neutrons they lay within 35%. The reproducibility
A. DELGADO ET AL.
328

a
t

P
u
r
d
u
e

U
n
i
v
e
r
s
i
t
y

L
i
b
r
a
r
i
e
s

A
D
M
N

o
n

J
u
n
e

3
,

2
0
1
4
h
t
t
p
:
/
/
r
p
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

for peaks 3 and 5 is better than for peak 4, for
neutrons particularly. Similar gures were obtained
for different groups of detectors. For the experiment,
batches of 25 detectors, selected according to batch
homogeneity (within 5%), were used.
Once the reference glow curves were established
and the corresponding R
ni
and R
ci
values were pro-
perly determined, they were introduced in the
programme performing the analysis of the curves
produced by mixed elds in terms of the two refer-
ence curves. A rst test of the performance of the
programme was made irradiating groups of ve
detectors to a constant neutron dose (35 mGy) and
different gamma doses between 10 and 100 mGy.
The test was designed to check: (1) the linearity of
the estimated mixed eld gamma dose, (2) the coher-
ence between these gamma estimation and the results
obtained from pure gamma irradiations at the same
dose values, and (3) the due constancy of the neutron
contribution estimated for the groups with different
gamma doses.
Figure 2 presents the results of the test. The good
linearity of the estimated gamma component, the
good agreement between these estimations and the
pure gamma data for each dose value and the good
constancy of the neutron component estimation for
the ve groups of detectors, can be appreciated.
Similar tests at different neutron doses (up to
200 mGy) produced results well in line with those
in Figure 2a.
Figure 3 presents the results of a different test
performed irradiating with Am-Be source for differ-
ent durations (from 20 min to 24 h) with the detec-
tors always placed at the same point respect to the
source. As the gamma dose rate produced by the
source at this point was independently determined
by LiF TLD-700 (1.5 mGy/h), the gamma contri-
bution for the different exposure times was known
and can be compared to those estimated by the
Figure 1. TLD-600 glow curves produced by (a) 20 mGy
gamma irradiation, (b) 35 mGy thermal neutron irra-
diation. The individual peaks resolved by the analysis are
presented. In curve (b) the weak gamma component
produced by the neutron source is also shown. In the
inset, the ratios of area of the individual peaks and the
total area (added area of peaks 3, 4 and 5) are indicated.
Figure 2. Variation of the estimated gamma (~) and
neutron (&) contributions in TLD-600 glow curves as a
function of the additional gamma dose. For comparison,
the measured gamma contribution by TLD-700 is also
indicated (!).
Figure 3. Dependence of the estimated gamma (~) and
neutron (&) TLD-600 components with the duration of the
neutron irradiation. The gamma contribution measured by
TLD-700 is also shown (!).
ON THE USE OF LIF TLD-600 IN NEUTRON-GAMMA MIXED FIELDS
329

a
t

P
u
r
d
u
e

U
n
i
v
e
r
s
i
t
y

L
i
b
r
a
r
i
e
s

A
D
M
N

o
n

J
u
n
e

3
,

2
0
1
4
h
t
t
p
:
/
/
r
p
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

analysis procedure. Within contrast to the previous
tests, now the neutron contribution is not constant
but is directly proportional to the exposure time, and
this proportion should be observed in the estimated
neutron contribution. In the Figure 2b, it can be
observed that all these predictions are well fullled.
The agreement between the true (TLD-700) gamma
doses and the estimated ones for the different expo-
sure times are always acceptable (within 1020%),
and at the same time, the estimated neutron contri-
bution exhibit a very good linearity with the time of
exposure, with regression coefcient of R0.9996
and a very small intercept. To be noted, the linear
response of TLD-600 for neutron doses from35 mGy
up to high doses, which in this test reached values of
the order of 2.5 Gy.
CONCLUSION
The satisfactory results obtained in the different tests
applied so far, support the correctness of the idea
behind the method herein described for the ana-
lysis of the TLD-600 glow curves from mixed elds.
That is, these glow curves are additive with respect
to the neutron and gamma components and their
respective structure, taken as reference, can be
employed for the analysis of the mixed eld glow
curves, separating the two contributions and per-
mitting to quantify separately the corresponding
absorbed doses. New tests are under way and some
others are being planned for the near future, includ-
ing the irradiation in other facilities with different
n/c elds, looking for the improvement of the neu-
tron dose estimations and also the renement of the
analysis procedure for mixed elds TLD-600 glow
curves.
REFERENCES
1. Burgkhardt, B. and Schwartz, W. Evaluation techniques
for different TL albedo dosemeters using automated read-
out. Radiat. Prot. Dosim. 17, 131134 (1986).
2. Mun iz, J. L., Vicente, M. C., Gonzalez, E. M.,
Romero, A. M., Embid, M. and Delgado, A. A new
area multidetector dosemeter for mixed n-gamma elds.
Radiat. Prot. Dosim. 110, 243248 (2004).
3. Bos, A. J. J., Piters, T. M., Go mez Ros J. M. and
Delgado, A. An intercomparison of glow curve analysis
programs: I. Synthetic Glow Curves. Radiat. Prot.
Dosim. 47, 473477 (1993).
4. Youssian, D. and Horowitz, Y. S. Estimation of gamma
dose in neutron dosimetry using peak 4 to peak 5 ratios in
LiF:Mg,Ti (TLD-100/600). Radiat. Prot. Dosim. 77,
151158 (1998).
A. DELGADO ET AL.
330

a
t

P
u
r
d
u
e

U
n
i
v
e
r
s
i
t
y

L
i
b
r
a
r
i
e
s

A
D
M
N

o
n

J
u
n
e

3
,

2
0
1
4
h
t
t
p
:
/
/
r
p
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m