Chalcogenide Letters Vol. 14, No. 5, May 2017, p.

203 - 209

PHYSICAL AGEING EFFECT IN Se75Te20In5 CHALCOGENIDE GLASS

M. M. A. IMRANa, O. A. LAFIa*, N. MEHTAb, A. F. ALSHWABKEHc,

A. A. SHAHEENd, A. ALBQOORa
a
Department of Physics, Faculty of Science, Al-Balqa Applied University, Al-Salt-
19117, Jordan
b
Glass Science Laboratory, Department of Physics, Banaras Hindu University,
Varanasi - 221005, India
c
Department of Applied Sciences, Faculty of Engineering Technology, Al-Balqa
Applied University, Amman, Jordan
d
Department of Physics, Faculty of Science, Hashemite University, Zarka, Jordan

Investigation of physical ageing effect induced by natural storage up to 11 yr of

Se75Te20In5 glass has been carried out using differential scanning calorimeter. It is
observed that10-years of natural storage increase both the glass transition temperature Tg
and the endothermic peak area A, which represents the enthalpy loss Δh of structural
relaxation. The estimated value of A of the 10-years aged glass (27.1 ± 2.1) J/g is greater
than that of as-prepared glass (11.3 ± 1.7) J/g, indicates that a significant natural ageing
effect is evident in the studied glass and a structural relaxation process towards a more
thermodynamically equilibrium state has been attained. This was also confirmed through
calculation of the enthalpy loss, during the relaxation process, from the knowledge of the
excess specific heat ΔCp. However, after one additional year of storage of the 10-years
aged glass (that is 11-years of natural storage), thermal parameters of structural relaxation
namely: Tg, Δh, ΔCp and the value of A do not significantly differ from those of the 10-
years aged sample. Thus, after 10-years of natural storage the glass reached its ageing
plateau and further additional structural relaxation of glass network towards a more
thermodynamic equilibrium state is not expected to occur. In addition, calculation of the
relaxation activation energy Et and heteropolar bond energies of the studied glass reveals
that structural relaxation occurs due to bond interchange rather than sliding of atomic
layers.

(Received April 3, 2017; Accepted May 30, 2017)

Keywords: Physical ageing; Natural storage; Glass transition temperature; Enthalpy loss;
Excess specific heat

1. Introduction

Physical ageing is one of the most important phenomena for glassy disordered solids such
as chalcogenide glasses. The promising applications of chalcogenide glasses are limited by their
structural instability, resulting from the production process, in which the liquid phase is quenched
to the ambient temperature. The process of thermal ageing of as-quenched material leads to the
achievement of the metastable state characteristic of the supercooled liquid. To avoid the changes
in physical properties caused by physical aging, a material with completely saturated ageing
should be used in chalcogenide-based devices. However, natural physical ageing requires years at
low temperatures; therefore, γ-irradiation, photo-exposure, and thermal annealing are usually
applied to accelerate the relaxation process [1-3]. The ageing process can be predicted on the basis
of thermodynamic relaxation, induced by structural rearrangements in amorphous matrices, which
is responsible for changes (as a function of time) in many physical properties of such kind of
materials. The reason is that the glassy disordered solids are conventionally obtained in a
metastable state frozen near glass transition and therefore they aspire to the extrapolated states of
the thermodynamic equilibrium of undercooled liquid. In other words, the prepared glasses kept
______________________________
*
Corresponding author: olafi66@yahoo.com
204

far below glass transition temperature (Tg) lose their excess of configurational entropy, enthalpy or
free volume to reach a more favorable thermodynamic state [1]. This effect was observed
experimentally in many binary and ternary systems of chalcogenide glasses and can be evaluated
by using differential scanning calorimetry (DSC) through distinguishable changes in the glass
transition region [2–6].
The physical ageing effects are characterized by the glass transition temperature Tg and the
change in the relaxation enthalpy Δh that can be detected from the endothermic DSC curves. The
observed endothermic peak is normally associated with structural relaxation of glass network
owing to glass-to-supercooled liquid phase transition [7]. Besides, it is mentioned that the
corresponding area A under this endothermic peak in the DSC curve represents the enthalpy loss
Δh: the greater A, the greater Δh and thus the glassy material is closer to equilibrium state of
undercooled liquid [5, 8, 9].
Thermal stability represents the resistance to devitrification of glassy alloy through the
nucleation and growth processes [10]. The relaxation activation energy Et is a good indication on
thermal stability. This activation energy is the amount of energy, which is absorbed by a group of
atoms in the glassy region so that a jump from one metastable state to another is possible. This
means, this activation energy is involved in the molecular motion and rearrangements of atoms
around the glass transition temperature. When the sample is reheated in the DSC furnace, the
atoms undergo infrequent transitions between the local potential minima separated by different
energy barriers in the configuration space where each local minimum represents a different
structure. The most stable local minimum in the glassy region has lower internal energy. This
activation energy of thermal relaxation could be monitored through the knowledge of the enthalpy
loss during the relaxation process in the aged glass [11, 12].
The objective of the present work is to monitor the effect of physical ageing induced by
natural storage on the glass transition region in Se75Te20In5 glass using differential scanning
calorimeter (DSC). In this regard, the glass transition temperature, excess specific heat and
enthalpy loss of as-prepared, 10-years aged and 11-years aged Se75Te20In5 glass have been studied.
Besides, the relaxation activation energy Et was calculated from the dependence of the glass
transition temperature on heating rate and compared with bonds energies of the constituent
elements of the investigated glass.

2. Experimental details

Se75Te20In5 glass was prepared on September 2005 by melt quenching technique. High
purity (5N) In, Te, and Se in appropriate atomic weight percentage (at. wt %) proportions were
weighed into a quartz ampoules and sealed at a vacuum of 10−5Torr. The ampoules were then
heated at 900 ºC for about 15 h with continuous rotation to facilitate homogenization of the
sample. The molten sample was rapidly quenched in icecooled water to produce a glassy state. The
ingot of the so-produced glassy sample was taken out of the ampoule by breaking the ampoule and
then grinded gently in a mortar and pestle to obtain a powder form. The as-prepared sample is
divided into three parts. One part of the sample is subjected to the DSC immediately after
preparation. The second and the third parts were packed in sealed containers with ambient
atmosphere and stored under normal conditions for 10 and 11-years. The calorimetric
measurements, at different heating rates of 5, 10, 15, and 20, were carried out in a Perkin-Elmer
DSC7 calorimeter with an accuracy of ± 0.1°C. This equipment is calibrated prior to
measurements, using high purity standards Pb, Sn and In with well-known melting points. For
each heating rate, about (10 ± 0.5) mg of each of the investigated glassy samples in powder form
were sealed in a standard aluminum pans and scanned from room temperature up to 100ºC (around
the glass transition region). The operation of DSC is based on measurement of the thermal
response of an unknown specimen as compared with a standard when the two are heated
uniformly. A typical differential scanning calorimeter consists of two sealed pans; a sample pan
and an empty reference pan. These pans are often covered by or composed of aluminum, which
acts as a radiation shield. The two pans are heated, or cooled, uniformly while the heat flow
difference between the two is monitored. This can be done at a constant temperature
 205

(isothermally), but is more commonly done by changing the temperature at a constant rate (non-
isothermally) as in this work.

3. Results and discussion

The enthalpy Δh recovered as a result of structural relaxation during physical ageing is

dependent on the ageing time. The kinetics for the enthalpy relaxation have been extensively
studied in amorphous materials [5-9, 13]. The enthalpy recovery increases with ageing time, but
approaches a plateau when a sufficient time is given. The time rate of increased enthalpy increase
(dΔh/dt) gradually decreases with ageing, because the mobility of polymer chains is reduced as the
systemreaches the metastable equilibrium state [14]. The glass transition temperature Tg (the
temperature at which the endothermic peak (enthalpy recovery) appears on the DSC thermogram)
increases with ageing time. It has been previously reported that Tg increased in a linear fashion,
relative to the logarithmic values of ageing time [5].
Typical DSC endothermic curves of as-prepared Se75Te20In5 glass at the heating rates of 5,
10, 15, and 20K/min are shown in Fig.1, as an example. Similar thermograms (not shown in the
text) were recorded for the 10-years aged and 11-years aged Se90In4Sn6 glass. Depending on these
figures, the glass transition temperatures (Tg) have been deduced from the intersection of the two
linear portions adjoining the transition elbow in the DSC trace in the endothermic direction [15].
In addition, DSC data for three Se75Te20In5 samples, one just after preparation, the second after 10-
years of natural storage, and the third after 11-years of natural storage, are shown in Fig. 2 at a
heating rate of 20K/min. The increase in Tg and endothermic peak area in the vicinity of the glass
to undercooled liquid transition testifies to the significant natural physical ageing effect in this
glass. As mentioned before, the endothermic peak area represents the enthalpy losses occurred
during prolonged natural storage of the samples. The two parameters Tg and A are determined from
the DSC endothermic peaks that are shown in Fig. 2 for as-prepared, 10-years aged and 11-years
aged Se75Te20In5 glass at a heating rate of 20K/min and listed in Table1. It is clear from this table
that both Tg and the area A under the endothermic peak increase after 10-years of natural storage.
After one additional year (11-years of natural storage), the value of A do not differ significantly
from that of the 10-years of natural storage sample. From this result, one can say that additional
structural relaxation of glass network towards thermodynamic equilibrium of super-cooled liquid
after this period of time is not expected and the glass reached its ageing plateau. Thus, the 10-years
aged sample is in a glassy state that is as close to the thermodynamic equilibrium states of
supercooled liquid as possible under the natural conditions.

5K/min

10K/min

15K/min
DSC Signal

20K/min

Endo

30 40 50 60 70 80 90 100
o
T ( C)

Fig.1. DSC endothermic curves, obtained at heating rates of 5, 10, 15 and

20K/min for as-prepared Se75Te20In5 glass
206

11 years-Aged

10 years-Aged

DSC Signal
As-Prepared

Endo

30 40 50 60 70 80 90 100
o
T ( C)

Fig.2. DSC endothermic curves, obtained at a heating rate of 20K/min, of as-prepared,

10-years aged and 11-years aged Se75Te20In5 glass

Table 1: Glass transition temperature Tg, and the area under endothermic peak A (converted into unit
of energy) of Se75Te20In5 glass with different characteristics at a heating rate of 20K/min

Sample characteristic ↓ Tg (oC) A (J / g )

As-prepared 60.8 11.3 ± 1.7

10-years aged 66.5 27.1 ± 2.1
11-years aged 67.1 28.2 ± 3.2

Physical aging is a phenomenon that is kinetically and thermodynamically controlled, and

is generally thought to occur in a nonlinear and nonexponential fashion [13, 14]. Nonlinearity is
resultant from the dependence of relaxation time (τ) on both the temperature (T) and the structure
of glassy matrices. Among the expressions widely used to describe this phenomenon is the Tool-
Narayanaswamy-Moynihan equation, which is applicable at a temperature near the glass-transition
temperature [16-18]:

τ = τoexp(–Et/RT)exp(– c(h – he)) (1)

where τo and c are constants, R is the universal gas constant (= 8.314 JK-1mol-1), ∆h= h-he is the
enthalpy loss during the relaxation process and Et is the relaxation activation energy. It is
mentioned earlier that the endothermic peak area represents the enthalpy loss ∆h during thermal
relaxation. However, in the present case since natural storage time is so long therefore one can
calculate the value of ∆h from the knowledge of the specific heat difference ∆Cp=Cpe- Cpg versus
temperature curve, through the use of the well known formula ∆h = ∆Cp [Tcg– Ta] as described in
[19] and the references therein. Here Cpe is the equilibrium specific heat obtained at a temperature
after the glassy region is completed where the glass attains equilibrium liquid like structure and
Cpg is the specific heat of the material just before jumping at Tg takes place. The temperature
difference [Tcg– Ta] is considered as the difference between the glass transition temperature
obtained when the glass is cooled through the transition region at a rate of 20K/min and the ageing
temperature Ta (which was taken as room temperature = 298 K). The variation of specific heat as a
function of temperature for as-prepared, 10-years aged and 11-years aged Se75Te20In5 glass at a
heating rate of 20K/min is shown in Fig. 3. It is clear that Cp is weakly temperature dependent
except in the glassy region where it exhibits a sudden jump at Tg which is attributed to additional
degrees of freedom that are available for the atoms to formulate complete short range order in the
liquid state [20]. The Cp against temperature curves were used to calculate ∆Cp and ∆h values,
which are listed in Table 2, for the studied glass. The obtained values of ∆h indicate that drastic
relaxation occurs due to natural storage and that the aged glass reached almost equilibrium state.
 207

The observed increase in the values of ∆h with ageing time at ageing temperature below the glass
transition is an indication of the fact that atomic rearrangement becomes more cooperative in
relaxation process in which more enthalpy is released with an increase in ageing time [19].
2.0
As-prepared
1.8 10-years aged
11-years aged
1.6

1.4

Cp(J /g C)
o
1.2

1.0

0.8

0.6

30 40 50 60 70 80 90 100
o
Temperature ( C)

Fig. 3. The variation of specific heat as a function of temperature for as-prepared, 10-years
aged and 11-years aged Se75Te20In5 glass

Table 2: Specific heat difference ∆Cp and the enthalpy loss during natural storage of Se75Te20In5
glass at room temperature (at a heating rate of 20 K/min)

Sample characteristic ↓ ∆Cp (J / g oC ) ∆h (J / g )

As-prepared 0.31 10.9 ± 1.1

10-years aged 0.62 25.4 ± 1.4
11-years aged 0.65 26.8 ± 1.7

Depending on equ.(1), Moynihan and co-workers [17] derived the following relation, from
which Et can be calculated from the variation of Tg with the heating rate β:

d(ln β) / d(1/Tg) = –Et /R (2)

Fig.4 shows plots of lnβ vs. 103/Tg for all samples. The values of the relaxation activation
energy Et are calculated from the slopes of the resulting straight lines and found to be 2.33, 2.24,
and 2.22 eV for as-prepared, 10-years aged, and 11-years aged glass, respectively. It is clear that Et
decreases after 10 years of natural storage which is an indication that the state of equilibrium is
almost reached. In addition, a comparison between the estimated value of Et for 11 years of natural
storage and that of 10 years of natural storage reveals that no noticeable change in the stability of
the studied glass. This agrees fairly well with the above result obtained from the relaxation
enthalpy (endothermic peak area) calculations.
208
3.2

3.0

2.8

2.6

2.4

ln
2.2

2.0

1.8 As-Prepared
10-years aged
1.6 11-years aged

1.4
2.94 2.95 2.96 2.97 2.98 2.99 3.00 3.01 3.02 3.03 3.04 3.05
3 -1
10 /Tg(K )

Fig.4. Plot of lnβ vs. 103/Tg for as-prepared, 10-years aged and 11-years aged Se75Te20In5 glass.

The studied glass is a chalcogen rich alloy in which the structural configuration consists of
rings or chains held together by Van der Waals interaction [21] with bonds have energies in the
range 0.12-0.40 eV. However, the bond energies between constituent atoms for the glass under
investigation were calculated following the procedures reported in [12] and the references therein.
These bond energies found to be 1.76, 1.71, and 1.39 eV for Se-In, Se-Te, and Te-In, respectively.
Comparing the obtained values of the relaxation activation energy Et with Van der Waals energies,
one can say that the relaxation is not possible to occur due to sliding of the atomic layers.
However, the values of Et for as-prepared, 10-year aged, and 11-years aged are higher than the
above calculated heteropolar bonds energies. Hence, it is reasonable to believe that the relaxation
process in the sample under study is due to bonds interchange.

4. Conclusions

Physical ageing effect and thermal stability of Se75Te20In5 glass have been studied using
differential scanning calorimeter (DSC) and the following conclusions may drawn:
Firstly, the estimated enthalpy loss during relaxation process measured from the
endothermic peak area agrees fairly well with the calculated value for the studied glass with
different characteristics.
Secondly, the obtained values of the thermal parameters Tg, ∆h, and ∆Cp, indicate that the
process of relaxation is terminated in almost 10-years of natural storage.
Finally, the calculations of the relaxation activation energy and bonds energies of the
constituent elements of the studied glass reveal that the structural relaxation occurs due to breaking
of interchain bonds which must reform after completion of relaxation process.

References

[1] O. Shpotyuk, M. Hyla, V. Boyko , R. Golovchak, Physica B 403,3830 (2008).

[2] M.K. Vanitha, M.V. Hanumantha Rao, S. Asokan, K. Ramesh, J. Phys. Chem.
Solids,74, 804 (2013).
[3] R. Golovchak, A. Kozdras, S. Kozyukhin, O. Shpotyuk, Nucl. Inst. Meth. Phys.
Res. B 267, 2958 (2009).
[4] J.M. Saiter, M. Arnoult, J. Grenet, Physica B 355, 370 (2005).
[5] O. Shpotyuk, A. Kozdras,V. Balitska , R. Golovchak. J. Non-Cryst. Solids 437, 1 (2016).
[6] Omar A. Lafi, Mousa M.A. Imran, Hassan K. Juwhari, Ma'rouf K. Abdullah,
Radiation Physics and Chemistry 112,1 (2015).
[7] S. L. Ratushnyak, N. O. Gonchukov, Glass Physics and Chemistry,
33(2), 130 (2007).
[8] I.M. Hodge, J. Non-Cryst. Solids 169, 211 (1994).
 209

[9] R.Ya. Golovchaka, A. Kozdrasb, O.I. Shpotyuk, Solid State Communications

145, 423 (2008).
[10] N. Mehta, R.S. Tiwari, A. Kumar, Materials Research Bulletin 41,1664 (2006).
[11] M.M.A. Imran, N.S. Saxena, D. Bhandari, M. Husain, Phys. Stat. Sol. (a)
181, 357 (2000).
[12] O.A. Lafi, M.M.A. Imran, M.K. Abdullah, Mater. Chem. Phys. 108, 109 (2008).
[13] J. M. Hutchinson, Prog. Poly. Sci. 20,703 (1995).
[14] J. M. Hutchinson, S. Smith, B. Horne, G. M. Gourlay, Macromolecules 32,5046 (1999).
[15] O.A. Lafi, M.M.A. Imran, J. Alloys Comp. 509, 5090 (2011).
[16] A. Q. Tool, J. Am. Ceram. Soc. 29, 240 (1946).
[17] O.S. Narayanaswamy, J. Am. Ceram. Soc. 54, 491 (1971).
[18] C.T. Moynihan, A,J. Easteal, J. Wilder, J. Tucker, J. Phys. Chem. 78,2673 (1974).
[19] M.M.A. Imran, I. F. Al-Hamarneh, M. I. Awadallah, M. I. Al-Ewaisi, Physica B
403, 2639 (2008).
[20] M. M.A. Imran, D. Bhandari, N.S. Saxena, Mat. Sci. Eng. A 292,56 (2000).
[21] S.R. Elliott. Physics of Amorphous Materials. Longman, London. 1990.