Indian Journal of Pure & Applied Physics

Vol. 49, March 2011, pp. 180-185

Annealing effect on structural and electrical properties of

AgGaSe2 thin films
M R A Bhuiyan*,1, M Alauddin Al Azad & S M Firoz Hasan
*Department of Applied Physics, Electronics and Communication Engineering, Islamic University, Kushtia 7003, Bangladesh

Department of Physics, Bhawal Badre Alam Government College, Gazipur, Bangladesh

Scientific Information Division, Bangladesh Atomic Energy Commission, Dhaka 1207, Bangladesh
1

E-mail: mrab_iu@yahoo.com

Received 16 August 2010; revised 13 December 2010; accepted 4 February 2011

Structural, thermo-electrical and electrical properties of AgGaSe2 (AGS) thin films prepared in vacuum onto glass
substrates by stacked elemental layer (SEL) deposition technique have been studied. The films were annealed at 100 to
350C for 15 min. The atomic composition of the films has been measured by energy dispersive analysis of X-ray (EDAX)
method. The structural and thermo-electrical properties of the films have been ascertained by X-ray diffraction (XRD) and
the hot-probe method, respectively. The electrical properties of the films are measured by standard dc method using a liquid
nitrogen cryostat. The structural, thermo-electrical and electrical properties have been investigated as a function of different
annealing temperature. The X-ray diffraction (XRD) reveals that the films were polycrystalline in nature. The lattice
parameters, grain size, strain and dislocation densities of the films have been calculated. The thermoelectric power indicates
the presence of p-type majority carriers. The electrical conductivity of the films has been found to vary from 1.40105 to
2.18102 (-cm)1 as temperature varies between -173 and 100C. The activation energies vary from 43.60 to 94.70 meV
as annealing temperatures vary between 350 and 100C. Probable identities of the origin of AGS films have been obtained
by using these activation energies. The dominance of grain boundary effect has been ascertained by applying the Setos
model.
Keywords: AgGaSe2 thin films, Annealing temperature, Electrical conductivity, Activation energy

1 Introduction
In recent years, there has been considerable interest
in thin film ternary semiconductors for their use in
solar cells. Ternary compound of AgGaSe2 (AGS) is a
useful absorber material for the fabrication of thin
film heterojunction solar cells1-3. Electrical
conductivity is a fundamental parameter of solar
energy materials, which is optimized to apply on the
materials in the preparation of solar cells. The
conductivity is dependent on the number of charge
carriers, in turn, is affected by various sources of
scattering when they (carriers) move through the
material. Hence, precise knowledge about the mode of
transport of carriers in the material is very important
for its application as an optical absorber in the
preparation of solar cells. Many researchers4-7 have
already prepared AGS crystal and thin films by
different techniques and studied their electrical
properties. The utilization of ternary semiconductors
in photovoltaic devices has been reported in the
literature8-11. In the present paper, some aspects of the
structural and electrical measurements have been
reported. The electrical conductivity of the films has

been analyzed and some useful parameters have been

deduced utilizing the conventional models.
2 Experimental Details
2.1 Growth of the films

The films of silver, gallium and selenium were

deposited sequentially onto chemically and
ultrasonically cleaned glass substrates by thermal
evaporation of individual elemental layers to form a
stack, in vacuum (107 mbar) by using oil diffusion
pump (E306A, Edwards, UK). The thickness of the
films was 500 50 nm. The growth, compositional
and structural characteristics of AGS thin films have
been described in earlier studies12,13.
2.2 Thermo-electrical and electrical measurements

The type of electrical conduction was determined

by the hot-probe method. A small heater is attached
close to one end of the film to develop the
temperature gradient. The electrical measurement
technique of AGS thin films have been studied
earlier14.

BHUIYAN et al.: ANNEALING EFFECT ON AgGaSe2 THIN FILMS

3 Results and Discussion

To check the compositional homogeneity of the
films, EDAX data at several different locations for
each film were noted. The variation of composition
from one location to another did not exceed 2%,
which ascertains the reasonable compositional
homogeneity of the films. In our previous paper12, the
typical EDAX spectrum of AGS thin films having
stoichometry (elemental atomic composition of
25.52%, 24.54% and 49.94% for Ag, Ga and Se,
respectively) has been studied. In our observation, all
films were found to be stoichometric with different
annealing temperature.
Figure 1 shows the X-ray diffraction patterns of
stoichometric AGS films annealed at different
temperatures. The X-ray diffractogram revealed that
the films were polycrystalline in nature. Expected
peaks15,16 for AGS thin films are (112), (201), (211),
(220), (204), (301), (312). But in our work, (110),
(112), (211), (114), (220), (204), (310) peaks are
present. The intensity and position of (112) peak are
in good agreement with ASTM data for tetragonal
silver gallium selenide (JCPDS card no. 05-0724,
75-0114). The (112) peak is prominent peak for AGS
thin films and other researchers obtained (112) peak
of higher intensity15. In our case, the films annealed at
200, 250 and 350C show low-intensity (112) peak.
This peak is absent for the film annealed at 100C.

181

Film annealed at 300C shows (112) peak of higher

intensity. The high intensity of reflection from the
(112) plane indicates the preferred orientation17.
It shows tetragonal chalcopyrite structure with
average lattice parameters, a 6.00 and c 10.92
(5%) and c/a = 1.82. The grain size of the films
deposited at different annealing temperature was
calculated by using the well-known Scherrers
formula18 simplified by Bragg:

Bcrystallite =

0.94
B cos

(1)

where is the wavelength of Cu K radiation and B

is then full-width half-maximum (FWHM). The unit
of B should be converted into radian. Therefore,
Eq. (1) takes the form:
Bcrystallite =

50.97
B cos

(2)

The calculated values of the grain size for films

grown at different temperature are presented in
Table 1. The average grain size of the films was
calculated to be 40 nm.
In our previous paper19, variation of lattice
parameters and grain size with different annealing
temperatures has been studied. The lattice parameters
are independent unlike the grain size that varies
directly with annealing temperature. The grain size
becomes larger in the films annealed at higher
temperature. The increase of the grain size may be
due to the coalescence of small crystals. Larger grain
size minimizes the imperfect regions of the film,
which is also supported by the smaller strain and
dislocation densities.
The strain, was calculated by using the formula20:

B cos
4

(3)

Table 1 Structural parameters of AGS thin film having

different annealing temperatures

Fig. 1 XRD spectra of AGS thin films having different

annealing temperature

Annealing
temperature,
TA (C)

Lattice
Parameters
a ()
c ()

Grain
Strain, Dislocation
size
density,
104
(nm) (line2 m4) 1014
(line/m2)

100
200
250
300
350

5.93
6.00
5.99
6.01
5.99

14.85
17.10
34.20
41.04
45.90

11.32
10.91
10.89
10.93
10.89

23.20
20.13
10.11
8.38
7.50

45.34
34.20
8.64
5.93
4.74

INDIAN J PURE & APPL PHYS, VOL 49, MARCH 2011

182

The dislocation density, defined as the length of

dislocation lines per unit volume of the crystal, was
evaluated from the formula20:

1
( Bcrystallite ) 2

(4)

Strain and dislocation densities decrease as

annealing temperature increases (Fig. 2). Since
dislocation density and strain are the manifestation of
dislocation network in the films, the decrease in
dislocation density indicates the formation of highquality films21. The inhomogeneous strain component
is localized at the sub grain and sub domain level near
grain boundaries. Pertsev and Arlt22 introduced a
fictitious dislocation density to model the
inhomogeneous strain. The spontaneous strain in their
model is analogous to the mechanical plastic
deformation, for instance, in metals. This facilitates
the analysis substantially because models developed
for dislocations can be readily applied. They obtained
two components of strain/stress fields: a
homogeneous volume component extending over the
whole grain volume and an inhomogeneous
component concentrated near grain boundaries. We
argue that dislocation effects are a real origin of
inhomogeneous strains in AgGaSe2 thin films and that
large change in dislocation density and configuration
may occur. This is possibly because of the fact that
when the temperature is kept higher, the dislocation
densities get more thermal energy and have a higher
carrier concentration.
The thermoelectric power, also known as Seebeck
coefficient is a measure of the Seebeck effect. This

Fig. 2 Dependence of strain and dislocation density of AGS

thin films having different annealing temperatures

effect consists on the open-circuit voltage, V,

developed when a temperature difference, T, is
applied along the films. The phenomenon arises
because charge carriers at the highest temperature
regions of the film diffuse into suitably available
states of lower energy at the lowest temperature
regions of the films. The accumulation of charge
carriers at the cold end will set up an electric potential
difference or electric field between the ends of the
films. This electric field builds up until a state of
dynamic equilibrium is established between the
charge carriers with a natural tendency to diffuse
down the temperature gradient and those that are
forced up the temperature gradient by the electrostatic
repulsion due to excess charge at the cold end.
The thermoelectric power, S, is defined by:
S = lim

T 0

V
T

(5)

Thermoelectric power is used to interpret the

conduction mechanism in semiconductor materials.
The measurement of thermoelectric power is simple
and its sign gives vital information about the type of
electrical conduction in semiconductors whether they
are n-type or p-type. The type of electrical conduction
of the grown films having different temperatures has
been determined by hot-probe method. Figure 3
shows the variation of thermoelectric power with hot
junction temperature for different annealing
temperatures. The thermoelectric power was found to
be positive in the range 0.026-0.113 mV/C, which
indicates the presence of p-type majority carriers23,24.

Fig. 3 Variation of thermoelectric power with hot

junction temperature of AGS films having different
annealing temperature

BHUIYAN et al.: ANNEALING EFFECT ON AgGaSe2 THIN FILMS

Figure 4 shows the variation of electrical

conductivity with temperature of the AGS thin films
having
various
post-deposition
annealing
temperatures in the range 100-350C for 15 min. It is
observed that the conductivity varies from 1.40105
to 2.18102 (-cm)1 as annealing temperatures vary
from 100-350C. Electrical conductivity is strongly
dependent
on
annealing
temperature.
The
conductivity increases as the annealing temperature
increases which ascertains the semiconducting nature
of the films.
The movement of charge carriers becomes easier if
the number of grain boundary (GB) decreases and
consequently the conductivity of the films increases.
Hence, scattering at the GB may make substantial
contribution to the electrical conductivity25. The
variation of logarithmic conductivity as a function of
inverse of temperature is shown in Fig. 5.
The variation shows thermally activated process
and dependence of conductivity with temperature T,
may be expressed by the equation:

E
= 0 exp

kT

183

It was found that the activation energy decreases

continuously while the electrical conductivity is
progressively increasing (Fig. 6). Considering the
similarity between CuInSe2 and AgGaSe2, the identity
of the origin of the acceptor levels may possibly be
interpreted as follows:
The films annealed at 100 and 250C have
acceptor-like levels, which may be attributed to the
Se-interstitials and Ag- and Ga-vacancies,
respectively. The probable identity of the origin of
acceptor levels for the films annealed at 200, 300 and
350C is Ga-vacancy and Ag in Ga-sites.
The average grain dimensions of these films12 were
small 40 nm. The effect of the grain boundaries is

(6)

where 0 is conductivity at infinite temperature, E

the thermal activation energy and k is the
Boltzmanns constant. The straight-line nature of
these plots suggests that GB limited conduction is the
dominant conduction mechanism26. The activation
energy which is calculated from Fig. 5, varies
between 43.60 and 94.70 meV as annealing
temperature varies between 350 and 100C as shown
in Fig. 6. Each film has single activation energy.

Fig. 4 Variation of electrical conductivity with temperature for

AGS thin films having different annealing temperature

Fig. 5 Logarithmic conductivity as a function of inverse of

temperature of AGS thin films having different annealing
temperatures

Fig. 6 Variation of activation energy with annealing

temperature for AGS thin films

184

INDIAN J PURE & APPL PHYS, VOL 49, MARCH 2011

directly, whereas strain and dislocation density vary

indirectly with temperature. The electrical
conductivity of the films varies directly with
annealing
temperature
that
ascertains
the
semiconducting nature of the material. The variation
shows thermally activated process. The probable
identity of the origin of the films annealed at 100 and
350C has acceptor-like levels, which may be
attributed to the Se-interstitials and Ag- and
Ga-vacancies, respectively. The films annealed at
200, 250 and 300C showed Ga-vacancy and Ag in
Ga-sites. The dominance of the GB effect has been
ascertained by applying the Setos model.

Fig. 7 Variation of ln (T1/2) with inverse temperature for AGS

thin films having annealing at 300C
Table 2 Probable identity of the origin of AGS thin films
Annealing Activation References Probable identity of the
temperature
energy
origin
(meV)
(C)
100
94.7
17, 28
Se-interstitials
200
61.7
26, 29
Ga-vacancy and Ag in
Ga-sites
250
60.1
26, 29
Ga-vacancy and Ag in
Ga-sites
300
52.5
26, 29
Ga-vacancy and Ag in
Ga-sites
350
43.6
17, 30, 31 Ag-and Ga-vacancies

dominant when the grain size is small. The atoms at

grain boundaries are disordered. There are a large
number of defects as a result of incomplete atomic
bonding. These results in the formation of trapping
states and charge carriers which are trapped and
immobilized. The traps become electrically charged
and give rise to potential barriers27. To verify if GB
effect is dominant, a plot of ln (T1/2) versus 103/T
(Setos model) is drawn in the temperature range
173-100C (Fig. 7). It is noticed that this plot is a
straight line, which confirms the dominance of GB
effect

4 Conclusions
The films were reasonably homogeneous in
composition and stoichometric as evident from the
EDAX data at various locations on the films. The
films have been found to be polycrystalline in nature.
The lattice parameters are independent of annealing
temperature of the films. The grain size varies

Acknowledgement
One of the authors (M R A Bhuiyan) is grateful to
the University Grants Commission, Bangladesh for
granting him PhD fellowship for carrying out this
work. He is also grateful to the authority of Atomic
Energy Centre, Dhaka (AECD) for allowing him to
utilize its laboratory facilities. The authors are
thankful to Mr Syed Ayubur Rahman, Mrs Jahanara
Parvin and Mr Albert Joydhar for their assistance in
preparing the samples.
References
1 Murthy Y S, Uthanna S, Naidu B S & Reddy P J, Solid State
Comm, 79 (1991) 277.
2 Murthy Y S, Hussain O M, Naidu B S & Reddy P J, Mater
Lett, 10 (1991) 504.
3 Murthy Y S, Naidu B S & Reddy P J, Mater Sci Eng: B, 8
(1991) 175.
4 Zhao B J, Zhu S F, Fu S S, Li Q F, Jin Y R & Li Z H, Mater
Res Bull, 35 (2000) 1525.
5 Brisson O, Simonnet A, Darriet B & Launay J C, J Crys
Growth, 193 (1998) 597.
6 Zhu S F, Zhao B J, Liu J, Li Z H & Li W T, Mater Chem
Phys, 50 (1997) 94.
7 Patel S M & Kapale V G, Mater Lett, 4 (1986) 145.
8 Chaure N B, Young Y, Samantilleke A P & Dharmadasa I
M, Sol Energy Mater Sol Cells, 81 (2004) 125.
9 Schunemann P G, Setzler S D, Pollak T M, Ohmer M C,
Goldstein J T & Zelmon D E, J Crys Growth, 211 (2000)
242.
10 Rusu M, Gashin P & Simashkevich A, Sol Energy Mater Sol
Cells, 70 (2001) 175.
11 Murthy Y S, Hussain O M, Uthanna S, Naidu B S & Reddy
P J, Phys Lett A, 152 (1991) 311.
12 Bhuiyan M R A & Firoz Hasan S M, J Phys D: Appl Phys,
39 (2006) 4935.
13 Bhuiyan M R A & Firoz Hasan S M, Sol Energy Mater Sol
Cells, 91 (2007) 148.
14 Bhuiyan M R A, Das D K & Firoz Hasan S M, Indian J Pure
& Appl Phys, 47 (2009) 787.
15 Soliman H S, J Phys D: Appl Phys, 28 (1995) 764.
16 Murthy Y S, Hussain O M, Naidu B S & Reddy P J, 21
PVSC, 1 (1990) 487.

BHUIYAN et al.: ANNEALING EFFECT ON AgGaSe2 THIN FILMS

17 Tembhurkar Y D & Hirde J P, Thin Solid Films, 215 (1992)
65.
18 Suryanarayana C & Norton M G, X-ray diffraction, (Plenum
Press, New York), 1998, p.212.
19 Bhuiyan M R A, Rahman M K & Firoz Hasan S M, J Phys
D: Appl Phys, 41 (2008) 235108.
20 Lalita S, Sathyamorthy R, Senthilarasu S, Subbarayan A &
Natarajan K, Sol Energy Mater Sol Cells, 82 (2004) 187.
21 Pal U, Samantha D, Ghorai S & Chaudhuri A K, J Appl
Phys, 74 (1993) 6368.
22 Pertsev N A & Arlt G, Sov Phys Solid State, 33 (1991) 1738.
23 Ravinder D & Reddy P V B, J Magnetism Mag Mate, 263
(2003) 1273.

185

24 Misbah-ul- Islam, Abbas T & Chaudhry M A, Mater lett, 53

(2002) 30.
25 Kerkache L, Layadi A, Dogheche E & Remiens D, J Phys D:
Appl Phys, 39 (2006) 184.
26 Wasim S M, Sol Cells, 16 (1986) 289.
27 Seto J Y W, J Appl Phys, 46 (1975) 5247.
28 Masse G, J Phys Chem Sol, 45 (1984) 1091.
29 Neumann H, Nowak E & Kuhn G, Crys Res Tech, 16 (1981)
1369.
30 Migliorato P, Shay J L, Kasper H M & Wagner S, J Appl
Phys, 46 (1975) 1777.
31 Rincon C, Gonzalez J & Sanchez P G, J Appl Phys, 54
(1983) 6634.