Radiation Measurements 80 (2015) 29e37

Contents lists available at ScienceDirect

Radiation Measurements
journal homepage: www.elsevier.com/locate/radmeas

Effect of annealing and impurity concentration on the TL

characteristics of nanocrystalline Mn-doped CaF2
P.D. Sahare a, *, Manveer Singh a, Pratik Kumar b
a
b

Department of Physics and Astrophysics, University of Delhi, Delhi 110 007, India
Medical Physics Unit, BRAIRCH, AIIMS, Ansari Nagar, New Delhi 110029, India

h i g h l i g h t s

Nanocrystalline material CaF2:Mn is prepared by simple coprecipitation method.

The material is studied by XRD, TEM, ESR, TL and PL techniques.

High impurity concentrations give rise to clusters causing material instability.

Changes in ESR and PL and glow curve structures are studied and explained.

Better characteristics than the bulk make the nanophosphor useful for dosimetry.

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 18 October 2013
Received in revised form
14 April 2015
Accepted 3 July 2015
Available online 6 July 2015

Nanocrystalline samples of Mn-doped CaF2 were synthesized by chemical coprecipitation method. The
impurity concentration was varied in the range of 0.5e4.0 mol%. The structure of the synthesized material was conrmed using powder XRD analysis. TEM images of the nanoparticles show their size
occurring mostly in the range of 35e40 nm, with clusters of some impurity phases formed on annealing
of the material at higher temperatures. Detailed studies on TL showed that the structures of glow curves
depend on Mn concentrations and annealing temperatures. Optimization of the concentration and
annealing temperature showed that the sample (doped with 3.0 mol% and annealed at 673 K) has almost
a single dosimetric glow peak appearing at around 492 K. EPR and PL spectra were further studied to
understand the reasons for changes in the glow curve structures. All detailed studies on TL, PL and EPR
showed that the changes in glow curve structures are caused not only by the stress connected with the
difference in ionic radii of host Ca2 and the guest impurity Mn3/Mn2, but are also governed by other
reasons, like diffusion of atmospheric oxygen and formation of impurity aggregates, such as, MnO2,
Mn3O4, etc. This is true not only for nanocrystalline CaF2:Mn but could also be so for the bulk CaF2:Mn
(TLD-400) and would thus help in understanding complex glow curve structure, high fading and the loss
of reusability on annealing beyond 673 K.
2015 Elsevier Ltd. All rights reserved.

Keywords:
CaF2:Mn nanoparticles
Thermoluminescence (TL)
Concentration effects
EPR
PL

1. Introduction
TLD phosphor CaF2:Mn (TLD-400, microcrystalline powder or
hot pressed chips) shows a high sensitivity and linear response over
a wide range of radiation doses (0.5 mGye1.0 kGy)
(Thermscientic, TLD Materials, Features and Technical
Specications, 1981; Fehl et al., 1994). It is widely used for

* Corresponding author.
E-mail
addresses:
(P.D. Sahare).

pdsahare@yahoo.co.in,

http://dx.doi.org/10.1016/j.radmeas.2015.07.003
1350-4487/ 2015 Elsevier Ltd. All rights reserved.

pdsahare@physics.du.ac.in

radiation dosimetry for last four decades. However, there are some
drawbacks, such as, complicated glow curve structure, low stability
and fast fading, loss of reusability, etc. (Danilkin et al., 2008). One of
the main disadvantages of CaF2:Mn is its low stability, on annealing
and during repeated TL readouts, introducing inaccuracies in dose
measurements. It is caused by a strong dependence of TL glow
curve structures on the concentration of the impurity (Mn) and its
related redox reactions during annealing. It is believed that low
stability of the phosphor is caused by the oxidation of Mn2 to
higher charge states at higher temperatures (Danilkin et al., 2008;
Planque, 1984). However, a well dened and simple hightemperature TL peak is observed when concentration of Mn ions

30

P.D. Sahare et al. / Radiation Measurements 80 (2015) 29e37

is very close to the threshold concentration (~3.0 mol%). But even at

this concentration the sensitivity changes on annealing the sample
beyond certain temperature (~673 K). In fact, incorporating such
high concentration of the impurity forms aggregated solid solution.
High concentration of impurity leads to aggregation and formation
of impurity clusters introducing inhomogeneity in the material. The
optical properties, in such solid solutions, not only depend on the
inhomogeneity at molecular level but also on the starting materials,
method of synthesis and heat treatments (Lust, 2007).
Earlier, several workers have investigated the material
(CaF2:Mn) and tried to understand TL processes and defects
(Danilkin et al., 2008; Planque, 1984; Lust, 2007; Horowitz, 1984;
McKeever, 1985; McMasters et al., 1987; Danilkin et al., 2006;
McKeever et al., 1986). But the lack of clear knowledge about the
TL mechanism, defects formation and their instability at high
temperatures showed that the system (CaF2:Mn) is yet to be fully
understood. For example, Danilkin et al. (2006) have studied the
effect of Mn impurity concentration on TL and argued that the
changes observed in TL glow peaks could be attributed to the stress
generated in the crystal lattice because of the difference in ionic
radii of Ca2 and Mn2, i.e., on replacing Ca2 (ionic radius 1.26 )
by Mn2 (ionic radius 1.10 ), at lower concentrations but could be
compensated by incorporation of more Mn2 at higher concentrations. They further concluded that these distortions would affect
the structure of the TL glow curve by affecting the activation energies (due to changes in the stress). They suggested the role of selftrapped excitons (STE) in the energy storage mechanism. But they
could not explain satisfactorily the changes in the glow curve
structures and loss of reusability on annealing the material beyond
~673 K based on their model. This interaction model also does not
seem to be tenable as most of the defects might get annealed out
and also the stress may be reduced considerably on annealing.
Another important observation made by them was that the radiation had no effect on the amount of Mn2 after irradiating the
samples with doses up to 400 Gy as observed from the EPR studies.
If it is the case, indeed, then the phosphor should be very stable but
it has been found to be otherwise. The same group further tried to
explain the changes in the glow curve structures due to jumps of
the uorine ions into interstitials and forming the neutral F2 molecules storing the energy and releasing it on their dissociations
again changing the activation energies and the glow curve structure
(Danilkin et al., 2008). But, it seems it is not the stress or the
uorine ions only but something else that might also be contributing to it. These were some of the issues that prompted us to
revisit the Mn-doped CaF2 TLD phosphor material. The material is
studied in nanocrystalline form as the changes in optical and other
characteristics of these materials get enhanced by any small
change(s) in the matrix due to its tiny size (Sahare et al., 2012).
In the present work, we have investigated Mn-doped CaF2 in its
nanocrystalline form using XRD, TEM. EPR and PL studies have been
done to see redox reactions of impurity ions on annealing and
irradiation. TL characteristics, dose response and fading have also
been studied to nd its suitability for radiation dosimetry. TL glow
curves were also theoretically tted into different glow peaks using
Kitis' formulae (Kitis et al., 1998) to determine trapping parameters.
2. Experimental
2.1. Method of synthesis
Chemical coprecipitation method was used to synthesize
nanocrystalline CaF2. The starting materials CaCl2 (99.9%, Merck
India), NH4F (98.0%, Merck India) and MnCl2 (99.0%, CDH India) of
analytical (AR) grade were used as received without further purication. The samples were prepared taking into consideration the

following chemical reaction:

H2 oEtOH

CaCl2 2NH4 F MnCl2 0:0  4:0 mol%!CaF2 : Mn

2NH4 Cl:
Aqueous solution of NH4F (0.1 N) was added into the aqueous
solution of CaCl2 (0.1 N) drop-wise through a burette at the rate of
0.25 ml/min in the presence of ethanol (EtOH) while stirring
continuously for 2 h at room temperature for completion of the
reaction. The appropriate amount of impurity salt (MnCl2) was also
added into the CaCl2 solution for doping. All the utensils, such as,
beakers, burettes, pipettes, funnels, etc. used were made up of
Teon (PTFE) to avoid any contamination due to the formation of
reactive uoride ions. White precipitate, formed on completion of
the reaction, was separated from the solution by centrifugation at
5000 rpm and washed several times with triply distilled deionized
water and EtOH to remove traces of any unreacted ingredients. The
collected precipitates were dried at 400 K for 12 h in an inert atmosphere and annealed at different temperatures in air for further
studies.
2.2. Characterization and measurements
The Powder X-Ray Diffraction (XRD) patterns were recorded at
room temperature using a high-resolution D8 Discover X-ray
diffractometer (Bruker, Germany) equipped with a point detector
(scintillation counter). Cu-Ka1 radiation (l 1.54056 ) mono bel mirrors was used to obtain the
chromatized with a pair of Go
XRD patterns. XRD was recorded at the scan rate of 1.0 s/step with
the step of 0.013 . The high-resolution images of nanoparticles
were taken using Philips Tecnai G2 F30 TEM with 300 kV accelerating voltage.
Thermoluminescence (TL) was recorded at a constant heating
rate of 5 Ks1 using Harshaw TLD Reader 3500HT with the upper
temperature limit of 873 K. A 137Cs g-radiation was used to irradiate
the samples before TL recording. TL of the samples was taken
immediately after the irradiation (0.1 Gye100 kGy) except for the
samples used for studying fading. Almost equal amount of sample
(~5 mg) was taken for each experiment. At least three glow curves
were recorded for any sample to check reproducibility.
PL measurements of the powder samples were taken on Varian
Cary Eclipse Fluorescence Spectrophotometer using a Xenon ash
lamp as a source and a wide band PMT as a detector. Both emission
and excitation spectra were recorded with slit width at 2.5 nm. The
EPR measurements were taken at room temperature on the JOEL
JES-FE3XG X-band EPR spectrometer operating at ~9.4 GHz with a
100 kHz magnetic eld modulation. The diphenyl-picryl-hydrazyl
(DPPH, g 2.0037) was used as a standard sample for calibration.
3. Results and discussion
3.1. XRD
The XRD has been checked for the samples with different impurity concentrations in the range 0.5e4.0 mol% and annealed at
673 K. The XRD pattern of CaF2:Mn (3.0 mol%) nanoparticles is
shown in Fig. 1. All the peaks observed in the diffraction pattern are
indexed to the cubic structure (JCPDS le # 87-0971) with space
group Fm3m (225). The crystalline size of the nanocrystalline
particles, determined from the broadenings of the XRD peaks and
using Debye-Scherer's formula, was found to be about ~38 nm. The
slight relative shift of the XRD peaks due to decrease of lattice
constant is observed at increasing Mn concentration (Fig. S1,
Supplementary Material). These results are similar to those

P.D. Sahare et al. / Radiation Measurements 80 (2015) 29e37

31

Fig. 1. XRD pattern of the CaF2:Mn (3.0 mol%) sample, a) annealed at 873 K and b) 3D crystal structure of the spinel Mn3O4 phase.

reported by Danilkin et al. (2006). They supposed that the lattice

constant decreases as a result of substitution of Ca2 ions with the
smaller Mn2 ones. However, the situation changes at higher Mn
concentrations (above 3.0 mol%). When Mn concentration is too
high for the matrix to accommodate the Mn2 ions as impurity,
manganese falls out of the solid solution to form clusters or aggregates. Danilkin et al. (2006) and Lust (2007) had predicted such
a possibility but could not nd any separate impurity phase(s) or
any clusters in their materials. In our materials, we felt that there is
a possibility of formation of the Mn(OH)2/MnOOH or some other
complex phases as the materials are prepared through the mixture
of water and ethanol solution (Zhanga et al., 2004; Sharma and
Whittingham, 2001). This has indeed been observed from the
XRD patterns (Fig. 1) that there are new phases of the impurity
(Mn3O4/MnO2) formed after annealing at higher temperatures
beyond 673 K. The TEM image (Fig. 2) also has conrmed the formation of impurity clusters. These might have been formed due to
diffusion of the atmospheric oxygen (Sils et al., 2007). The diffusion
of oxygen at higher annealing temperatures is not new and is reported earlier in some TLD phosphors (Nakata et al., 1976; Lay and
Nolle, 1967; Lakshmanan et al., 2002; Sunta, 1984; Kerikmae, 2004;
Bakshi et al., 2009). The experimental results are compared with
the data available (JCPDS le # 80-0382 for tetragonal and JCPDS
le # 75-0765 for orthorhombic phases). It could be seen from Fig. 1
that some of the peaks in the XRD pattern match with the tetragonal phase (JCPDS le # 80-0382) and the others match with the
orthorhombic phase (JCPDS le # 75-0765). In fact, Mn3O4 has a
spinel structure and in this structure Mn2 ions are located in the

tetrahedrally coordinated sites; and Mn3 ions are located in the

octahedrally coordinated sites. Also, the Mn3 sites (i.e., at octahedral sites) form corner-sharing tetrahedra, resulting in geometric
and magnetic frustrations. Such irreversible transformation of impurity phases not only change the amount of stress but may also
change some the Mn2 ionic states (of the impurity) into Mn3, i.e.,
Mn2
Mn3 and visa-versa (Kim et al., 2011). Initially, as Mn2
is replacing Ca2, most of the substitutional ions would be in Mn2
state but the redox reactions take place due to formation of oxide
phase(s) on annealing at higher temperatures. The formation of
such phase(s), however, would depend on the annealing temperature and the time (Sils et al., 2007).
3.2. Morphology of synthesized powder
TEM image of the Mn-doped nanocrystalline sample annealed
at 1073 K is shown in Fig. 2A. Cubic-shaped particles of the material
with clusters of the impurity phases (indicated by arrows in Fig. 2A)
are clearly seen in the magnied TEM image of the sample. Selected
area electron diffraction (SAED) pattern (shown in Fig. 2B) indicates
a good crystallinity of the annealed nanoparticles.
3.3. EPR studies
The EPR spectra of the materials were studied with X-band EPR
spectrometer at room temperature. The spectra of the samples
annealed at 673 K and having different Mn2 concentrations are
shown in Fig. 3A. Our results are similar to that reported by

Fig. 2. A) HRTEM images of the CaF2:Mn (3.0 mol%) nanocrystalline sample annealed at 1073 K B) SAED pattern of the sample. Clusters of the impurity phases could also be seen
clearly in the HRTEM image (as shown by arrows).

32

P.D. Sahare et al. / Radiation Measurements 80 (2015) 29e37

Fig. 3. A) EPR spectra of Mn2 for different concentrations of Mn in CaF2 (0.0e4.0 mol
%). B) EPR spectra of nanocrystalline CaF2:Mn (3.0 mol%) sample annealed at different
high temperatures (300e873 K) for 2 h. The spectrum for DPPH (g 2.0037) used for
calibration of the magnetic eld has also been shown in the gure.

Danilkin et al. (2006). EPR spectrum consists of six hyperne

structure (HFS) lines due to interaction with 55Mn nuclei (5/2, 100%
abundance). The ne structure due to octahedral crystal eld is not
observed in powder samples, because the structure is averaged, and
also, lines are broadened due to dipoleedipole interactions between Mn2 ions. The super-hyperne structure from 19F nuclei (1/
2, 100% abundance) is not observed due to the same reasons. As
follows from Fig. 3A, the EPR spectrum of Mn2 increases with the
concentration of manganese. This is true when Mn2 substitutes for
Ca2, but at very high concentrations Mn falls out of the lattice to
form some clusters, MnOOH, Mn(OH)2 or in some more complex
forms (McKeever et al., 1986; Chakrabarti et al., 1995; Tanner and
Pan, 2009).
The annealing temperature effect on EPR spectra is shown in
Fig. 3B (the samples with 3.0 mol% of manganese). EPR signal of
Mn2 increases with temperature increasing. Low signal of Mn2 in
as prepared samples may be explained by the fact that manganese
can be involved in some complex impurity phases due to entrapped
water and ethanol, and also, it can be present in other than
2 charge states. At annealing, the impurity incorporated in the
form of Mn(OH)2/MnOOH decomposes to release water. At the
same time, atmospheric oxygen also gets incorporated in the matrix to form MnO2 or Mn3O4 clusters. The agglomeration or clustering effect is enhanced with longer or repeated annealing (Bakshi
et al., 2009; Chakrabarti et al., 1995). Manganese coexists in clusters

and separate phases both in 2 and 3 states (see Section 3.1), and
the same could be true for the solid solution (Chakrabarti et al.,
1995). It has been shown by earlier workers (Danilkin et al., 2008,
2006; Kerikmae, 2004; Bakshi et al., 2009) that the material gets
deteriorated by conversion of Mn2 in to higher oxidation states.
However, it is not very clear about the processes involved. But from
the results presented here, it seems that it is due to formation of
new impurity phase(s) and the redox reactions taking place
thereafter. As more atmospheric oxygen gets diffused at higher
temperatures, the number and size of these clusters would go on
increasing with temperature and time. However, some of the
interstitial F ions in the matrix could still be trapped near Mn3
ions. These Mn3 ions in combination with such interstitial F ions
could thus behave effectively as Mn2 ions increasing the EPR
signal. The effect (increase in the EPR signal with annealing temperature increasing) could prominently be seen in Fig. 3B. As
mentioned earlier, the formation of these impurity phases have
already been conrmed by the XRD patterns of the samples
annealed at higher temperatures (Fig. 1). One could clearly see the
formation of different XRD peaks due to the growth of Mn3O4 impurity phase clusters inside the CaF2 crystallites. These clusters
have also been observed in the TEM images of the annealed samples (Fig. 2A).
Further similar trend has been found in the EPR spectra of the
samples with 2.0 mol% of the impurity, annealed at 673 K and
irradiated to different doses of g rays. They are similar to the spectra
shown in Fig. 3B (the spectra not given here due to space). It was
seen that the strength of EPR peaks increase with the dose
increasing. It seems that on irradiation some of the electrons from
the O2 ions from the octahedrally coordinated sites of the impurity
phase(s) or from F ions of the matrix would get trapped near Mn3
in place of F ion vacancies (generated due to displacement of the
interstitial F ions during irradiation) to effectively form more
Mn2 ions (whereby some of the Mn3 ions get converted into
Mn2 by these redox reactions). So, on irradiation these Mn3 ions
get converted into Mn2 by electron getting trapped nearby, i.e.,
form FA-type of centers (McKeever et al., 1986) or more com
plexF2H
-centers (Sils et al., 2007) or even neutral F2 molecules
(Danilkin et al., 2008). With irradiation increasing more electron
traps are formed and thus effectively more Mn2 ions are produced
which are responsible for increase in the EPR signal. Danilkin et al.
(2007) have, however, did not observe the change in the Mn2 ion
concentration on irradiation and therefore developed their own
model based on self-trapped excitons (STE). Accordingly, the lattice
gets distorted due to dissimilar radii of the Mn2 and Ca2 ions. On
irradiation, the intestinal F ions get further displaced and help in
formation of the excitons. These excitons are stable in the distorted
lattice and the excited lattice exerts more relaxations. This increases exciton bound energy at the expense of lattice relaxation
when Mn2 impurities are closer to each other (for higher concentrations). They have tried to explain the glow curve structure
and shifts in their peak temperatures with impurity concentrations.
This may be true for single crystals or even in microcrystalline
powder materials where less diffusion of oxygen may be taking
place. However, this model is not tenable in case of nanocrystalline
material where large surface area is available for diffusion of atmospheric oxygen and the Mn2 concentration changes due to
formation of the Mn3O4 phase clusters and more so with the concentration increasing.
3.4. Photoluminescence (PL) spectra
PL emission spectra of the samples with different impurity
concentrations annealed at 673 K are shown in Fig. 4
(lex 350 nm). A broad band with the maximum at 430 nm is

P.D. Sahare et al. / Radiation Measurements 80 (2015) 29e37

Fig. 4. The emission spectra of the CaF2:Mn nanocrystalline material for different
concentration of the impurity: a) Pure (undoped), b) 0.5 mol%, c) 1.0 mol%, d) 2.0 mol%,
e) 3.0 mol%, f) 4.0 mol%. Conc. of Mn, g) theoretically tted spectrum of the spectrum c,
h) Peak 1 of the spectrum c, i) Peak 2 of the spectrum c.

observed. The intensity of this emission increases with the Mn

concentration increase up to 3 mol% and decreases abruptly at
4 mol% of Mn impurity. The observed PL seems to disappear due to
formation of complex Mn impurity clusters, falling out of CaF2
lattice (Lust, 2007; Sils et al., 2007). One of the emission spectra has
been theoretically tted to Gaussian peaks and found that it mainly
consists of two broad bands peaking at around 430 nm (2.90 eV)
and 450 nm (2.76 eV). The emission bands may be assigned to 4T2g
(4G) 4 6A1g (6S) and 4T1g (4G) 4 6A1g (6S) transitions, respectively.
However, the peak positions do not exactly match with such band
predicted by others (Zhanga et al., 2004; Bakshi et al., 2009; Kohler
et al., 1997) as the prominent peak in their materials appears at
around 495 nm. It was therefore thought it might be appearing due
to some rare earth impurity traces present in our material, especially, Eu2 as it resembles to CaF2:Eu2 spectra (Pandey et al.,
2007) but our elemental analysis done by Time of Flight Secondary Ion Mass Spectroscopy (TOF-SIMS) measurements and the data
of the starting materials provided by the manufacturers (Fig. S2 (a
and b), Supplementary Material) ruled out such possibility. Sils
et al. (2007) have predicted that the PL emission band around
495 nm (2.5 eV) is due to oxygen related defects in their undoped
samples while others have attributed it to different 4G 4 6S transitions in Mn2 in their Mn-doped materials. Thus, it is difcult to
say whether the similar emission band(s) observed could really be
attributed to Mn2 as observed by others (Zhanga et al., 2004;
Bakshi et al., 2009; Kohler et al., 1997). But we strongly feel that
they may be related to Mn2 ions or some more complex Mn and
oxygen related defects.
The effect of annealing temperature on PL was also studied. The
emission spectra (lex 350 nm) of the samples annealed at
different temperatures (not shown here due to space) has similar
behavior as in case of the impurity concentration (Fig. 5). It was
observed that the peak intensity was the maximum for the pristine
sample (i.e., 3.0 mol% Mn-doped as prepared sample) and decreases with annealing temperatures. This again could be explained
on the basis of formation of the clusters of Mn3O4 phase(s)
(whereby some of the Mn2 ions get converted into Mn3 states or
vice-versa, as discussed earlier, Mn3O4 has a spinel structure
wherein Mn could exist in 2 as well as 3 states) due to annealing
and the diffusion of atmospheric oxygen at high temperature(s).
Even if the oxidation does not take place the Mn(OH)2, as formed
initially, could get decomposed to form MnOx during the

33

Fig. 5. The emission spectra for the CaF2:Mn nanocrystalline material annealed at
different annealing temperatures: a) Pristine, b) 473 K, c) 673 K and d) 873 K. A typical
emission spectrum (for the samples annealed at 473 K) has also been theoretically
tted into two peaks appearing at around 430 nm (2.90 eV) and 450 nm (2.76 eV), e)
theoretically tted spectrum of the spectrum b, f) Peak 1 of the spectrum b, g) Peak 2 of
the spectrum b.

temperature range 320e550  C and further higher order of more

complex phases at higher temperatures (Augustin et al., 2015). The
Mn-impurity thus could be in the 2 as well as in 3 or even in
4 states but as it gets agglomerated to form clusters, i.e., the energy may be released non-radiatively. The PL intensity thus gets
diminished due to formation of these clusters (Lust, 2007; Sils et al.,
2007). This has also been conrmed by change of the color of the
sample on annealing at higher temperatures (Fig. S3:
Supplementary Material).
3.5. Thermoluminescence (TL)
3.5.1. Optimization of the impurity concentration and annealing
temperature
The effect of impurity concentrations on the structure of the TL
glow curves and sensitivity of the TLD phosphor material depends
not only on the form in which the starting compound of the impurity is taken but also synthesis route which may determine the
form in which it has entered in the matrix (i.e. the ionic state of the
dopant in the host materials) (Danilkin et al., 2006; Pandey et al.,
2007). Similarly, the effects of annealing temperatures on the
structure of the TL glow curves not only depend on redox reactions
taking place during annealing/readouts but also depend on certain
other phenomena, such as, aggregation of impurities, the diffusion
of atmospheric oxygen during annealing and so on (Lakshmanan
et al., 2002; Sunta, 1984; Kerikmae, 2004; Bakshi et al., 2009;
Kim et al., 2011; Azorin et al., 1993). The effect of impurity concentration (0.5e4.0 mol%) on the TL glow curves of the samples
annealed at 673 K is shown in Fig. 6. It is observed that the samples
with ~3.0 mol% impurity concentration give the maximum TL
sensitivity. It is also observed from this gure that the glow curves
of the samples having impurity concentrations other than 3.0 mol%
(that of lower as well as higher concentrations) are more complicated and consists of more than one peak but for samples having
~3.0 mol% there is apparently only a single peak. It seems that
3.0 mol% impurity ions get well incorporated (dissolved) in the CaF2
host lattice during synthesis. The samples having lower concentrations (lower than 3.0 mol%) on irradiation displace some of the
F ions to interstitial positions disturbing the interactions between
Mn2 ions and also creating distortions in the lattice leading to
formation of more complex defects. This gives rise to more

34

P.D. Sahare et al. / Radiation Measurements 80 (2015) 29e37

Fig. 6. TL glow curves for different concentrations of the impurity: a) Pure (undoped),
b) 0.5 mol%, c) 1.0 mol%, d) 2.0 mol%, e) 3.0 mol%, f) 4.0 mol%. The ordinate needs to be
multiplied by the number (3) near the curve (curve e) to get the relative intensity.
The samples were irradiated to 100 Gy of g rays dose from 137Cs with the dose rate of
0.1 Gy/s.

complicated glow curves (Danilkin et al., 2007). For the samples

having 3.0 mol% apparently it looks like a single glow curve,
however, it consists of three highly overlapping peaks with their
activation energies very close to each other. But, in the solid solutions beyond this concentration, besides this, the impurity falls out
of the matrix to form some well-dened clusters (phases) resulting
again in more complicated glow curves and the concentration
quenching subsequently.
The effect of annealing (in the range of 473e873 K in air for 2 h)
on the samples (0.5e4.0 mol%) are also studied. However, the data
only for the impurity 2.0 mol% is given in Fig. 7 due to space. It was
observed from these studies that the glow peak structures (of all
the samples except the sample having 3.0 mol%) change on
annealing at different temperatures and the glow peaks of these
samples (for all the impurity concentrations) annealed beyond
873 K also consists of only one peak at around 480 K. There is also
an appreciable quenching of the TL intensity for the samples
annealed at higher temperatures. As discussed earlier, this could be

Fig. 7. TL glow curves of the nanocrystalline CaF2:Mn (2.0 mol%) for different
annealing temperatures: a) Pristine, b) 473 K, c) 673 K, d) 873 K and e) 1073 K.
Deconvoluted different peaks by CGCD method (Peaks P1eP3) of the glow curve for the
sample annealed at 673 K along with the corresponding theoretically tted curve
(curve f) is also shown. The samples were irradiated to 100 Gy of g rays dose from 137Cs
with the dose rate of 0.1 Gy/s.

attributed to the diffusion of atmospheric oxygen at higher temperatures rstly forming oxide impurity phases, such as MnOx and
Mn3O4 and at still higher temperatures there could be formation of
some other polymorphs of the manganese oxide due to diffusion of
more oxygen in the matrix. Formation of clusters also leads to
quenching (Danilkin et al., 2006; Sils et al., 2007; Bakshi et al.,
2009). The TL emission spectra of the samples annealed at
different temperatures were also studied by recording the monochromatic glow curves using narrow-band interference lters and
3D isomeric plots were constructed to nd out whether there is any
change in the TL glow curve/emission spectra. Such 3D plots
(isomeric plot of TL emission spectra) for as prepared sample and
other samples annealed at different temperatures are as shown in
Fig. S4 (Supplementary Material). A broad TL emission peak(s) with
maximum in the range of 410e490 nm was/were observed. This
may be attributed to the transfer of electronehole recombination
energy to nearby Mn2/Mn3 and relaxation of (Mn2)*/(Mn3)* (in
excited state) to Mn2/Mn3. It may be noted here that the TL
emission spectra are simpler (consisting of a main peak at around
495 nm) for the sample annealed at 400  C. However, it is less
intense, broader and more complicated for the as prepared sample
and the sample annealed at 800  C. In case of as prepared sample it
could be due to formation of the Mn(OH)2/MnOOH or some other
complex phases and at higher temperature due to formation of
impurity MnOx or Mn3O4 clusters.
3.5.2. Glow curves, dose response and TL mechanism
The TL glow curve structure is also found to change with the
impurity concentration (Fig. 6) and on annealing (473e1073 K) as
shown in Fig. 7. The TL glow curve of the nanocrystalline nominally
pure (without doping) CaF2 material is also given in Fig. 6 for
comparison. It could be observed from these gures that all the
samples annealed below the temperatures 673 K and having concentrations in the range 0.5e4.0 mol% (except for 3.0 mol%), there
are at least three peaks, i.e., main peak appearing at around 492 K
with shoulders on both sides at around 430 and 550 K. It could also
be clearly seen that the shoulders diminish at the impurity concentration ~3.0 mol% (i.e., the intensity of the peaks on both sides
decrease with the impurity concentrations increasing) and only a
single peak at around 480 K could be observed.
According to Lust (Lust, 2007) and Danilkin et al. (2006), the
distances between Mn2eF are smaller than the distances between Ca2eF due to difference in ionic radii (ionic radii of Mn2
and Ca2 are 1.10 and 1.26 , respectively), hence, the Mn2 substitution for Ca2 causes stress and distortions in the CaF2 lattice,
i.e., Ca2eF distances are decreased and Mn2eF distances are
increased against their usual values. These lattice deformations also
limit the solubility (intake within the tolerable limits) of the impurity and more impurity than this limit may precipitate the impurity ions to form clusters. The limit here seems to be ~3.0 mol%.
Also, the stress developed in the lattice affects the activation energies of the traps involved in TL and change the glow curve
structure (Lust, 2007; Danilkin et al., 2006). Therefore, at low impurity concentrations as well as at higher concentrations the glow
curves are more complicated but at around 3.0 mol% the glow curve
consists of only a single peak when the concentration is enough to
get it accommodated without much stress. As per these considerations, initially when the Ca2 is replaced by Mn2 the strain is
developed in the lattice at lower concentrations but as the impurity
concentration increases more and more Mn2 ions go to interstitial
cites as they fall out of the lattice. The anions (i.e., F ions) also go to
interstitial cites for charge compensation. This may also compensate for the strain. At very high concentrations, they may form
aggregates and clusters with non-uniform distribution and again
generate strain (Danilkin et al., 2006). However, formation of more

P.D. Sahare et al. / Radiation Measurements 80 (2015) 29e37

complex defects like [MnF6]4 or [MnF8]6 clusters at higher concentrations cannot be ruled out. Generally, defects involved in
thermoluminescence processes of Mn-doped CaF2 material are
complexes of manganese and the radiation induced defects
(McKeever et al., 1986; Mathur et al., 1986; Allen and McKeever,
1990). However, some of the Mn2 ions oxidized to higher manganese states (i.e., Mn3 ionic state) due to incorporation of the
oxygen no more remains a luminescence activator (Lust, 2007) and
formation of clusters also adds to it. This leads to the quenching of
TL and loss of reusability. This is more so, in case of nanoparticles
where more surface area is available for diffusion. Preparing the
material in a reducing atmosphere does not improve the TL characteristics. There are also reports that the Mn2 could get reduced
to Mn or neutral Mn by CO, while preparing the material in a
graphite crucible (Lust, 2007). But this did not help improving the
TL characteristics.
TL studies of CaF2:Mn samples (annealed in the temperature
range of 473e1073 K in air for 2 h) have also shown that the glow
curve structure changes with annealing temperatures (Fig. 7). It
could be observed (Fig. 7) that a high-temperature peak appears in
case of the as prepared samples in the temperature range
665e675 K which later vanishes in case of the samples annealed
beyond 473 K. This shows that the high temperature peak might be
related to some complex defects formed with the trapped water
and ethanol molecules during synthesis. These molecules might
form some complexes responsible for such deep traps or luminescence centers (LC). However, it is difcult to assign a particular
TL glow peak to a particular kind of impurity/defect in a complex
system, such as, CaF2:Mn, where different kinds of defects/radicals
are formed due to ionization on irradiation. But as the high temperature peak (in the range 665e675) vanishes in case of the
samples annealed beyond 473 K, either the complexes formed due
to ethanol or water molecules might be getting decomposed or the
trapped water molecules might get released relaxing the crystal
lattice. This would help in getting the energy levels reorganized and
subsequently changing the traps/LC levels. At the same time formation of the manganese oxide clusters due to diffusion of the
oxygen kills the prospecting luminescence centers because of formation of Mn3 ions and subsequently quenching occurs.
The glow curve structures of the annealed samples are similar to
that reported by Danilkin et al. (2006), with some change in the
peak temperatures that may be due to difference in particle size/
shape. Further, they have studied the effect of impurity concentrations up to 2.76 mol% only and the effect at higher concentrations was not studied in their samples. At higher concentrations the
impurity gets precipitated out and forms aggregates of the impurity
phase(s) quenching the TL intensity. However, we could not
observe a signicant change of the peak positions in a glow curve
with irradiation dose increasing (not shown here due to paucity of
space).
The dose response of the synthesized nanomaterials (CaF2:Mn
(3.0 mol%) and annealed at 673 K) was plotted by taking integral
area under the glow curve and peak height of the dosimetry peak as
well and the resulting response curve is as shown in Fig. 8. TL
response shows the linearity up to 10 kGy (i.e., it is sublinear up to
1.0 kGy and becomes linear before it saturates). Thus nanocrystalline Mn2 doped CaF2 applicable for the detection of wide range
doses (0.1 Gye10 kGy) of radiation while TLD-400 saturates around
100 Gy. For comparison of the TL sensitivity of the synthesized
nanocrystalline CaF2:Mn (3.0 mol%) phosphor (after annealing at
673 K) with the known standard phosphors, such as, TLD-100
(LiF:Mg,Ti) and TLD-400 (CaF2:Mn), they were exposed to 1.0 Gy
of g rays from a 137Cs source simultaneously. All the phosphors
were given required annealing treatment before irradiation. TLD400 (Harshaw) phosphor was found more sensitive (around

35

Fig. 8. TL response of the nanocrystalline CaF2:Mn (3.0 mol%) annealed sample (673 K)
for different doses of g rays: Peak height vs. dose (curve a), Integral peak area vs. dose
(curve b).

twice) to the synthesized nanocrystalline CaF2:Mn2 phosphor,

while it is around ve times more sensitive to TLD-100 (Harshaw)
for 1.0 Gy dose, as shown in Fig. 9.

3.5.3. Glow curve convolution and deconvolution (GCCD)

To understand the TL processes in the present CaF2:Mn nanocrystalline system, the glow curves were further deconvoluted into
simple peaks by the computerized glow curve deconvolution
(CGCD) method (detail could be found in our earlier communication (Sahare et al., 2012)) and trapping parameters were determined from these glow peaks. The single TL glow curve composed
of the several shoulder peaks which are overlapped caused the
asymmetric traps distribution within forbidden gap of the phosphors. Therefore, the TL glow curve deconvoluted into several
peaks using the function of general order kinetics as given by Kitis
et al. (1998). The activation energy, E or the trap depth needed to
free the trapped electrons can be calculated pretty accurately using
Chen's set of empirical formulae (Chen, 1969; Chen and Kirsh,
1981). The value of gure of merit (FOM) should not exceed 2.5%
for the good tting (Puchalska and Bilski, 2006).
The kinetic trapping parameters of the deconvoluted peaks of

Fig. 9. Comparison of the nanocrystalline CaF2:Mn (3.0 mol%) annealed sample (673 K)
with other commercially available TLD phosphors: a) Nanocrystalline CaF2:Mn (TLD
Phosphor under investigation) b) LiF:Mg,Ti (Harshaw TLD-100 chips), c) CaF2:Mn
(Hawshaw TLD-400 chips). The ordinate needs to be multiplied by the respective
numbers near the curves to get the relative TL intensity.

36

P.D. Sahare et al. / Radiation Measurements 80 (2015) 29e37

the entire glow curves (of all the samples with different impurity
concentrations and annealed at different high temperatures) have
been determined. The kinetic trapping parameters are tabulated in
Table 1. The data for the other glow curves is not shown here due to
space. From the data one could see that the trapping parameters,
especially, activation energies and the frequency factors vary on
annealing the samples at different temperatures indicating that the
energy levels get reorganized with annealing as well as with the
impurity concentration indicating the complexity of the problem.
However, for Mn concentration around 3.0 mol% the peak shape
does not change with annealing temperatures and the material
having impurity concentration around this is the most suitable for
the dosimetry applications.
3.6. Fading of nanocrystalline CaF2:Mn TLD phosphor
Fading (loss of TL signal from the phosphor with storage time
after exposure) is an important characteristic of the materials for
the dosimetric application. CaF2:Mn (TLD-400) found to have very
fast fading, i.e., very fast fading of about 10% in the rst few hours,
with further decreases at the rate of 1% per day is reported and is
not good from the application point of view (Pandey et al., 2007). In
the present study also done an experiment to observed the fading
characteristics of the TL signal of nanocrystalline CaF2:Mn. Twelve
pallets (samples) of nanocrystalline CaF2:Mn2 phosphor were
exposed to 10 Gy to g rays of 137Cs source and TL is taken for
different time intervals, i.e., after 3 h, 5 day, 20 day, 60 day, 125 day,
190 days. The results are shown in Fig. 10 as the mean integral of TL
glow curve for each group of 3 samples. It can be seen that
maximum fading of ~15% for a period of a few months after
exposure. Nanocrystalline CaF2:Mn2 thus shows the better storage
capacity of the radiation signal than reported in literature (Azorin
et al., 1993).
4. Conclusions

Fig. 10. Fading curve (Plot of the TL intensity vs. Storage time) of the nanocrystalline
CaF2:Mn (3.0 mol%) annealed sample (673 K). The samples were irradiated to 10 Gy of
g rays dose from 137Cs with the dose rate of 0.1 Gy/s.

2.

3.

1. Cubic shaped Mn2 doped nanocrystalline CaF2 are prepared

successfully via chemical coprecipitation route. The materials
are characterized by XRD, TEM, EPR and the luminescence

Table 1
Trapping parameters of the nanocrystalline CaF2:Mn (2.0 mol%) and annealed at
different temperatures in the range of (300 Ke1073 K). All the peaks were deconvoluted by CGCD method. The temperatures mentioned in the table are the temperatures at which the material was annealed.
Sample
Pristine
1
2
3
473 K
1
2
3
4
673 K
1
2
3
873 K
1
2
3
1073 K
1
2
3

Tmax. (K)

E (eV)

S (s1)

426
469
493

1.07
1.25
0.47

1.52  1012
8.73  1012
6.52  103

1.35
1.35
1.54

386
458
485
550

1.36
0.6
1.59
0.97

2.92
6.21
1.30
1.37

1017
105
1016
108

1.74
1.54
1.05
1.54

458
484
503

0.77
1.92
0.39

5.76  107
4.49  1019
671

1.98
1.98
1.35

428
473
583

0.77
0.68
0.95

2.66  108
2.85  106
2.50  107

1.74
1.74
1.54

374
410
475

0.86
0.49
0.49

1.35  1011
166  105
1.77  104

1.35
1.54
1.74

4.

FOM
0.5

0.5





1.5

5.
0.2

0.4

studies are done. XRD patterns of the annealed samples

revealed that oxide phases of manganese, i.e., Mn3O4 has been
formed in clusters due to diffusion of atmospheric oxygen on
annealing to high temperatures in air. The HRTEM images have
also shown such clusters in the doped and annealed samples.
The EPR studies with the impurity concentration, annealing and
irradiation have also been done. The EPR spectra of the samples
having different impurity concentration, annealed at different
temperatures and irradiated for different doses of g rays from
137
Cs source showed increase in the strength of the EPR spectra.
This also supports the formation of the MnOx, Mn3O4 or more
complex impurity phases.
The changes taking place inside the material are responsible for
complicating glow curve structures, quenching the TL and
making the phosphor material unstable on annealing at higher
temperatures. The stress/strain due to inhomogeneity and
cluster formation in the material also adds to such complications. This also explains the quenching of the PL intensity due to
impurity concentration and annealing. However, the material
could still be used with certain precautions, especially, with
annealing/taking TL readouts under inert atmosphere below
673 K.
We have tried to understand the complexity of the glow curves
and the changes in glow curve structures on varying the impurity concentration and on annealing the material at high
temperatures and showed that it is not the stress/strain generated due to difference between the ionic sizes of the host and
the guest ions but the redox reactions taking place on diffusion
of the atmospheric oxygen inside the material and the formation of aggregated clusters are also responsible for the complicated glow curve structures and the loss of reusability. The
change in color of the material on annealing at high temperatures also supports these ndings.
Synthesized nanocrystalline CaF2:Mn2 (3.0 mol%) shows better
TL characteristics (except its little less sensitivity than the
microcrystalline commercially available CaF2:Mn2, TLD-400),
such as, good sensitivity (around half that of TLD 400 and ve
times more than that of LiF:Mg,Ti, TLD-100), good linearity over
a very wide range of radiation doses (0.1e104 Gy), easy method
of synthesis, low fading, (around 15% in a 3 months as compared
to 30% in 30 days in case of TLD 400), etc. The sensitivity though
somewhat less than that of the commercially available TLD 400,

P.D. Sahare et al. / Radiation Measurements 80 (2015) 29e37

it could still be considered as very good as compared to any

other nanocrystalline TLD phosphor. It is still around ve times
more than commercially available TLD-100 (Harshaw) phosphor. All these characteristics of the nanocrystalline CaF2:Mn2
(3.0 mol%) TLD phosphor makes it a suitable candidate for the
radiation dosimetry.
Acknowledgments
We are thankful to the University of Delhi for partial nancial
assistance through R & D grants (File # RC/2014/6820). The nancial assistance by the Inter-University Accelerator Center (IUAC),
New Delhi under UFUP project (File # UFR- 54306) is also gratefully
acknowledged. The author (MS) is thankful to the University Grant
Commission (UGC) for Rajiv Gandhi National fellowship (RGNF).
We are also thankful to the anonymous reviewers for their critical
comments and suggestions which helped in improving the quality
of the paper.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.radmeas.2015.07.003.
References
Allen, P., McKeever, S.W.S., 1990. Radiat. Prot. Dosim. 33, 19e22.
Augustin, M., Fenske, D., Bardenhagen, I., Westphal, A., Knipper, M., Plaggenborg, T.,
Kolny-Olesiak, J., Parisi, J., 2015. Beilstein J. Nanotechnol. 6, 47e59.
Azorin, J., Furetta, C., Scacco, A., 1993. Phys. Stat. Sol. 138, 9e46.
Bakshi, A.K., Dhabekar, B., Rawat, N.S., Singh, S.G., Joshi, V.J., Kumar, V., 2009. Nucl.
Instr. Methods Phys. Res. B 267, 548e553.
Chakrabarti, K., Sharma, J., Mathur, V.K., Barkyoum, J.H., 1995. Phys. Rev. B 51,
16541e16548.
Chen, R., 1969. J. Appl. Phys. 40, 570e584.
Chen, R., Kirsh, Y., 1981. Analysis of Thermally Stimulated Processes. Pergamon
Press, Oxford, p. 163.
ndar, H., Must, M., 2006. Radiat.
Danilkin, M., Lust, A., Kerikm
ae, M., Seeman, V., Ma

37

Meas. 41, 677e681.

Danilkin, M., Kirillov, A., Klimonsky, S., Kuznetsov, V., Lust, A., M
andar, H.,
Nikiforov, V., Ratas, A., Ruchkin, A., Seeman, V., 2007. Radiat. Meas. 42,
594e596.
e, M., 2008. Radiat. Meas. 43,
Danilkin, M., Lust, A., Ratas, A., Seeman, V., Kerikma
300e302.
Fehl, D., Muron, D.I., Suijka, R.R., Vehar, D.W., Lorence, L.I., Westfall, R.L., Jones, S.C.,
Sweet, I.A., Braunlich, P., 1994. Rev. Sci. Instrum. 65, 3243e3251.
Horowitz, Y.S. (Ed.), 1984. Thermoluminescence and Thermoluminescent Dosimetry, Vol. 1. CRC Press, Boca Raton.
Kerikmae, M., 2004. Ph. D. Thesis: Some Luminescent Materials for Dosimetric
Applications and Physical Research. Tartu University Press, ISBN 9985-56-8958. ISSN: 1406e0299.
Kim, M., Chen, X.M., Wang, X., Nelson, C.S., Budakian, R., Abbamonte, P., Cooper, S.L.,
2011. Phys. Rev. B 84, 174424e174435.
Kitis, G., Gomez-Ros, J.M., Tuyn, J.W.N., 1998. J. Phys. D. Appl. Phys. 31, 2636e2641.
Kohler, T., Armbruster, T., Libowitzky, E., 1997. J. Solid State Chem. 133, 486e500.
Lakshmanan, A.R., Jose, M.T., Ponnusamy, V., Kumar, P.R.V., 2002. J. Phys. D. Appl.
Phys. 35, 386e396.
Lay, F.M., Nolle, A.W., 1967. Phys. Rev. 163, 266e275.
Lust, A., 2007. Ph. D. Thesis: Charge State of Dopants and Ordered Clusters Formation in CaF2:Mn and CaF2:Eu Luminophors. Tartu University Press, ISBN 9789949-11-6-621.
Mathur, V.K., Abundi, R.J., Brown, M.D., McKeever, S.W.S., 1986. Radiat. Eff. Defects
Solids 99, 9e14.
McKeever, S.W.S., 1985. Thermoluminescence of Solids. Cambridge University Press,
Cambridge, p. 153.
McKeever, S.W.S., Jassemnejad, B., Landreth, J.F., Brown, M.D., 1986. J. Appl. Phys. 60,
1124e1130.
McMasters, D.W., Jassemnejad, B., McKeever, S.W.S., 1987. J. Phys. D. Appl. Phys. 20,
1182e1190.
Nakata, R., Kohnom, K., Sumita, M., Higuc, E., 1976. J. Phys. Soc. Jpn. 41, 470e474.
Pandey, C., Dhopte, S.M., Muthal, P.L., Kondawar, V.K., Moharil, S.V., 2007. Radiat. Eff.
Defects Solids 162, 651e658.
Planque, E.G.de, 1984. DOE Report EML-418, USA.
Puchalska, M., Bilski, P., 2006. Radiat. Meas. 41, 659e664.
Sahare, P.D., Bakare, J.S., Dhole, S.D., Kumar, P., 2012. Radiat. Meas. 47, 1083e1091.
Sharma, P.K., Whittingham, M.S., 2001. Mater. Lett. 48, 319e323.
Sils, J., Radzhabov, E., Reichling, M., 2007. J. Phys. Chem. Solids 68, 420e425.
Sunta, C.M., 1984. Radiat. Prot. Dosim. 8, 25e44.
Tanner, P.A., Pan, Z., 2009. Inorg. Chem. 48, 11142e11146.
Thermscientic, TLD Materials, Features and Technical Specications, 1981. www.
thermoscientic.com.
Zhanga, W., Yanga, Z., Liu, Y., Tang, S., Han, X., Chen, M., 2004. J. Cryst. Growth 263,
394e399.