temperature dependent dielectric relaxation and electrical conductivity in single-layer zno–al2o3...
TRANSCRIPT
phys. stat. sol. (b) 244, No. 7, 2657–2665 (2007) / DOI 10.1002/pssb.200642287
© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Original
Paper
Temperature dependent dielectric relaxation
and electrical conductivity in single-layer ZnO–Al2O3
nanocomposite thin films
Soumen Das and Subhadra Chaudhuri*
Department of Materials Science, Indian Association for the Cultivation of Sciences, Jadavpur,
Calcutta 700 032, India
Received 25 May 2006, revised 1 February 2007, accepted 14 February 2007
Published online 26 March 2007
PACS 72.80.Tm, 77.22.Gm, 77.55.+f, 77.84.Lf
Single-layer ZnO–Al2O
3 nanocomposite thin films with thickness approximately 119 nm, containing ran-
domly dispersed nanocrystallites of ZnO in the Al2O
3 matrix were prepared by sol gel technique. The rela-
tive molar concentration of ZnO was 20% in the matrix. The confined nanoparticles ZnO were investi-
gated through ac impedance spectroscopy in the frequency range of 500 Hz to 5 MHz at 315 K, 325 K,
345 K, 355 K, 395 and 450 K. The dielectric relaxations in each case were found to be of Cole–Cole
type. The thermally activated relaxation phenomenon was analysed with the help of an equivalent circuit
which is proposed on the basis of the experimentally obtained complex capacitance and impedance spec-
tral studies. The temperature dependent electrical conductivity was analysed on the basis of interstitial Zn
ions and free electrons in the system. The defect concentration of the nanocomposite thin film was deter-
mined from the complex capacitance measurements.
© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1 Introduction
In the field of nanoscience, the expertise in tailoring the properties of the materials in nanodimension has
elevated to certain heights, especially for different nanostructures and nanocomposite thin films. The
films themselves could be amorphous, single-crystalline or nanocrystalline and are used for electronic
thin film devices, for wear, chemical or oxidation protection, as well as for their optical properties (e.g.,
anti-reflection) and as transparent conducting films [1]. Transparent conducting oxides play a key role in
optoelectronic devices such as flat panel devices, photovolatic and electrochromic devices [2]. Transpar-
ent conducting metal oxide ZnO is a II–VI semiconductor and crystallizes in the hexagonal wurtzite
structure having large band gap Eg of 3.37 eV at room temperature [3]. ZnO exhibits strong n-type con-
ductivity with the electrons moving in the conduction band as the charge carriers. The explicit relation-
ship between the intrinsic band gap of ZnO with the particle size is exploited in order to seek its possible
use in the optoelectronic devices. On the other hand, a decrease in the size of the nanocrystallites means
higher surface to volume ratio which can incorporate significant surface related defects, disorder and
randomness in the system.
ZnO and Al2O3 films exhibit very dissimilar physical properties, yet both are transparent in the visible
region. ZnO–Al2O3 alloy films may provide advantages in the applications in optoelectronics by allow-
ing the control of particle size, refractive index and surface roughness. Nanocrystalline ZnO exhibits
improved electrical properties from their bulk counterpart [4]. It has been shown that the dielectric re-
* Corresponding author: e-mail: [email protected], Phone: +91 033 2473 4971, Fax: +91 033 2473 2805
2658 S. Das and S. Chaudhuri: Dielectric relaxation and electrical conductivity in ZnO–Al2O
3
© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.pss-b.com
laxation phenomenon in thin film greatly affect their electrical properties [5]. In the complex-plane ana-
lytical technique the semicircular fitting of the experimental ac data suggests an appropriate equivalent
circuit representation of their dispersion characteristics. This method can highlight the relative contribu-
tion of the grain, grain boundary and the defect states to the total ac response under a given set of ex-
perimental condition [6]. The mechanism of the electronic conduction of ZnO at temperatures higher
than room temperatures has been understood as the transfer of the electrons to the conduction band when
excited from the donor levels produced by interstitial zinc atoms and oxygen vacancies [7–9] acting as
donors in ZnO. The impedance analysing technique is capable of revealing the degree of structural ho-
mogeneity and stability in composite based devices.
This present investigation explained the dielectric properties of the ZnO–Al2O3 nanocomposite thin
films with high optical transparency prepared by sol–gel technique. The relaxation phenomenon and the
impedance spectroscopy effectively reveal the degree of structural homogeneity, intergranular electrical
barriers and the role of the grain boundary defects and shallow defect states to the total ac response [10].
2 Experimental
For the preparation of the ZnO–Al2O3 nanocomposites, the sol was prepared in two parts, (i) ZnO part
and (ii) the Al2O3 part. In the nanocomposite thin film the ratio of the molar concentrations of ZnO to
Al2O3 was maintained as 20:80. All reagent grade chemicals were purchased from Merck (India) Ltd.
(i) For the preparation of the ZnO part, zinc acetate, 2-propanol and diethanolamine (DEA) were used.
The molar ratio of zinc acetate and 2-propanol was approximately 1:55. For total dissolution of the zinc
acetate, 0.006 mole of DEA was required. The zinc acetate was first mixed with 2-propanol under stir-
ring and then DEA was added to the above solution drop wise under constant stirring. Then this part was
stirred for one hour to get a transparent sol. (ii) In the preparation of Al2O3 part, aluminum nitrate
(Al(NO3)3 ·6H2O) was used as the precursor. The aqueous solution of the precursor was refluxed for one
hour and then treated with NH3 solution (25%). The white precipitate (boehmite) was dissolved in (12:1)
volume ratio of ethanol and water. 3.0 cc of acetic acid was added to it drop wise under constant stirring
and was stirred for half hour. During the process the shape and the size of the sol particle changed. At the
end, after three hours of stirring a completely transparent and viscous sol was obtained.
The separately prepared ZnO and Al2O3 sols were mixed and stirred for two hours to get the desired
sol for coating. The singly coated as prepared thin film was annealed at the precise temperature of 573 K
in air for 25 minutes. The nanocomposite thin film is a random assembly of ZnO nanoparticles dispersed
in Al2O3 matrix.
The morphology, crystallite size and crystal structure of the nanocomposite samples were determined
by transmission electron microscopy (TEM) and high-resolution TEM (HRTEM) accompanied by se-
lected area electron diffraction (SAED) using JEOL 2010 electron microscope. Atomic force microscopy
(AFM) was studied were using Nanoscope IV scanning probe microscope controller. The complex im-
pedance, admittance and capacitance data were recorded at frequencies ranging from 500 Hz to 5 MHz
using Hioki 3532-50 LCR HiTester. For the electrical measurement the thin film was deposited on an
ITO (Indium doped tin oxide) coated glass with identical post deposition conditions. Thermally depos-
ited aluminum was used as the electrode at the top surface. The electrical properties are reported after
repeated measurements of the samples in identical conditions.
3 Results and discussion
3.1 TEM & AFM studies
A drop of the above colloidal sol was carefully placed on a carbon coated copper mesh to examine the
grain size of the nanocomposite by transmission electron microscope. The mesh was annealed also at
573 K in air to restore the identical post deposition condition of the thin film. It is observed that the grain
growth likely occurs due to near boundary diffusion of the neighborhood particles. The exact determina-
phys. stat. sol. (b) 244, No. 7 (2007) 2659
www.pss-b.com © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Original
Paper
tion of the sizes of the nanoparticles from the TEM is not possible as the micrograph shows highly ag-
glomerated nanocomposite containing nanoparticles of ZnO in the Al2O3 matrix. Careful examination of
the high magnification image yield the grain size of the agglomerated grains around 10 to 20 nm as
shown in Fig. 1(a).
In Fig. 1(b), in the selected area electron diffraction (SAED) pattern spots can be observed on diffused
rings. The diffraction rings are from (100), (002) and (101) planes corresponding to the wurtzite structure
of ZnO [JCPDS File No. 35-0664]. The high resolution TEM in Fig. 1(c) shows the section of the lattice
fringes of the ZnO nanoparticle embedded in the amorphous matrix. The successive planes are deter-
mined as (002) planes with the lattice spacing of 0.258 nm.
AFM images of the 1.0 × 1.0 µm2 surface of the samples shown in the Fig. 2. The homogeneity and
the roughness of the surface in the thin films are evident from the images. The top surface of the film is
appeared to be quite rough which is probably due to the crystallization and grain growth in the nanocom-
posite films. Roughness is a basic parameter which indicates the deviation of a surface with respect to a
perfect plane. The surface roughness is measured in terms of the root mean square roughness r (rms)
which is defined as
2
av
1
( )
(rms) ,
N
i
i
Z Z
rN
=
-
=
Â
where Zi is the Z value of each point, Zav is the average of the Z values and N is the number of points. The
roughness (rms) of the thin film annealed at 573 K was determined to be 0.77 nm.
3.2 Dispersion characteristics
It should be noted that the diffusion of the ion occurs due to various defects in the structure of the grain
boundaries, namely dislocations. Such ionic diffusion in nanocrystalline materials is very dominant ow-
ing to (i) relaxation of the grain boundary structures, which occurs through relative grain displacements
and thus reducing the free volume of grain boundary structures and (ii) the grain boundary dislocations
Fig. 1 Transmission electron microscopic images of
the ZnO–Al2O
3 nanocomposite at 573 K, (a) the
morphology of the nanocomposite sample, (b) the
selected area electron diffraction pattern shows dif-
fraction from (100), (002) and (101) planes of ZnO,
(c) the high-resolution TEM shows (002) plane with
lattice spacing 0.258 nm of wurtzite ZnO.
Fig. 2 (online colour at: www.pss-b.com) Atomic
force microscope image of the top surface of the
nanocomposite thin film annealed at 573 K.
2660 S. Das and S. Chaudhuri: Dielectric relaxation and electrical conductivity in ZnO–Al2O
3
© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.pss-b.com
whose transformations greatly effect the diffusion process [11]. These interface and grain boundary de-
fects which may exist in nanocomposite thin films can lead to dielectric relaxation as a function of fre-
quency [6]. Such defects can be accounted by the measurements of the frequency dependent complex
parameters of the nanocomposite thin films. In case of grain boundary defects it is considered that the
grain boundary represents a resistors R, and the grain is thin insulating layer, C. Many such equivalent
R–C circuits may be thought to be present throughout the nanocomposite. The equivalent circuit analysis
can be adopted to indicate the role of grain boundary in the relaxation. Though this approach addresses
the average properties of a polycrystalline material and little information is obtained about the properties
of the individual elements, yet the measurements allow the separation of the contributions from grains,
grain boundaries and defect states. The impedance (Z) of the equivalent R–C circuit can be written as
eq eq
( ) 1/ .Z R j Cω ω= + (1)
The dispersion in the measured admittance (Y) can be defined by the following relation [6]
( )/ ( ) ( ) ( ) .Y I V G jBω ω ω ω= = + (2)
So we have
2 0 2
( ) ,G C Cω ω ω ε= =
1 0 1
( ) ,B C Cω ω ω ε= = (3)
where ω is the angular frequency; I(ω), V(ω) is the applied voltage and current, respectively; G(ω) and
B(ω) are the conductance (real admittance) and imaginary admittance, respectively; ε1 and ε2 are the real
and imaginary relative permittivities, C1, C2 are the real and imaginary parts of the capacitance. C0 is the
geometrical capacitance with vacuum and is given by C0 = ε0A/t, where ε0 is the free space permittivity,
A is the active area of 4 × 10–6 m2 and t the distance apart of electrodes, which is equal to the thickness of
the thin film (119 nm).
Spectra of ε1 and ε2 for the nanocomposite thin film were measured at different temperatures with
varying frequencies and are shown in Figs. 3 and 4.
As can be seen from the Fig. 4 that the peak shifted to higher frequency with increasing measuring
temperatures indicating the relaxation is thermally activated process. Another anomalous result observed
104 105 106 1070
5
10
15
20
25
30
35
Re[
ε]
ω (rad-s-1)
T=315 K T=325 K T=345 K T=355 K T=395 K T=450 K
Fig. 3 Frequency dependent real part of
the dielectric permittivity plot (ε1) of ZnO–
Al2O
3 thin films at different measuring
temperatures.
Fig. 4 Frequency dependent imaginary part of the
dielectric permittivity (ε2) plot of ZnO–Al
2O
3 thin
films at different measuring temperatures.
104 105 106 107
0
5
10
Im[ ε
]
ω (rad-s-1)
T=315 KT=325 KT=345 KT=355 KT=395 KT=450 K
phys. stat. sol. (b) 244, No. 7 (2007) 2661
www.pss-b.com © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Original
Paper
in the spectra is the apparent decrease in the (ε2) value with the increase in temperatures. The analysis of
the dielectric relaxation in the ZnO–Al2O3 nanocomposite thin film will be based on this interesting
finding. As we will observe in the next section that the dielectric strength of the composite thin film also
decreases with the increase in the measurement temperature.
The frequency dependence of ε1 and ε2 concerned with a single relaxation are expressed by [12, 13]
1 s
1 sinh( ) 1 ,
π2cosh cos
2
x
x
βε ε ε ε
ββ
• •
È ˘Í ˙
= + - -Í ˙Í ˙+Î ˚
(4)
and
2 s
πsin
1 2( ) ,π2
cosh cos2
x
β
ε ε εβ
β•
È ˘Í ˙
= - Í ˙Í ˙+Î ˚
(5)
where x is given by
M
1ln and .x ωτ τ
ω
= , = (6)
In the above equation εS and ε
∝ are the dielectric constants at the very low and very high frequencies,
respectively. β = 1 – α, where α is the angle of the semicircular arc. τ is the relaxation time and ωM is the
angular frequency at the maximum loss. Therefore, Eqs. (4) and (5) can be used to initiate the fitting
process by adjusting the values of εS, ∆ε, β and ωM. The adjusted parameters are listed in Table 1.
The fittings are shown in the Figs. 3 and 4 by solid lines. The observed dispersion in the dielectric
constant in the lower frequency region for temperature 315 K which showed large relaxation, needed
separate fitting parameters and the fittings are shown by dotted lines. The increase of the dielectric con-
stant in the lower region is a result of the large contribution of the dc conductivity in the data. The tem-
perature variation of the relaxation time (τ = 1/ωc = RC, where ωc is the peak angular frequency) in log
scale plotted in Fig. 5 against 1/κBT follows the Arrhenius equation
0
B
exp ,E
T
τ
τ τ
κ
Ê ˆ= Á ˜Ë ¯ (7)
where, τ0 and ω0 are the preexponential factor and Eτ is determined from the linear plot as 0.069 eV. E
τ is
the activation energy and is thus can be interpreted as the sum of both the creation and migration free
energy of the charge carriers [14].
Table 1 Table for the different relaxation related parameters for the ZnO–Al2O
3 nanocomposite thin
films.
T (K) ωM β τ0 (s) ∆ε σdc (× 10–4)
315 2882393.5 0.90 3.47 × 10–7 16.88 ± 0.05 4.97
325 4049098.0 0.93 2.46 × 10–7 11.52 ± 0.03 3.71
345 4338367.6 0.97 2.30 × 10–7 9.16 ± 0.02 3.76
355 4550029.0 0.98 2.19 × 10–7 8.20 ± 0.04 3.43
395 4900163.4 0.85 2.04 × 10–7 3.82 ± 0.01 1.75
450 7375754.1 0.88 1.35 × 10–7 2.99 ± 0.027 1.58
2662 S. Das and S. Chaudhuri: Dielectric relaxation and electrical conductivity in ZnO–Al2O
3
© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.pss-b.com
-16.0
-15.8
-15.6
-15.4
-15.2
-15.0
24 26 28 30 32 34 36 38
-8.25
-7.50
-6.75
-6.00
-5.25
-4.50
Eτ=0.069 eV
lnσ
lnτ
1/κBT (eV)-1
Edc
=0.116 eV
Figure 6 shows the logarithmic plots of the ac conductivity versus frequency. It is evident that the
conductivity is temperature and frequency dependent. The nature of the graph is such that at a particular
temperature it first rises slowly at lower frequency and then rapidly at higher frequency region revealing
the dispersive nature of the conductivity. The conductivity is closer to each other at higher frequency
whereas at lower frequency the effect of measuring temperature is evident. The observed conductivity in
the lower frequency region can be attributed to the weakly localized carriers which drift over large dis-
tances. At higher frequency the mean displacement of these carriers is reduced to show a proximity in the
conductivity [15]. The conductivity as a function of frequency can be expressed as
n
dc( ) ,Aσ ω σ ω= + (8)
where dc
σ is the value of conductivity in the lower limiting value of ferquency as 0ω Æ and A is a
constant depending on the temperature and n ranges from 0.5 to 1.
At this point what is intriguing is that the experimental plot reveals two distinct regions in the exponen-
tial growth of the conductivity. It is observed that in the theoretical fitting two sets of the above param-
eters were adjusted to represent the experimental data (i) below 1.0 MHz and (ii) above 1.5 MHz. In the
10 12 14 16-12
-10
-8
-6
-4
(a)
lnσ
ln ω
T=315 KT=325 KT=345 KT=355 KT=395 KT=450 K
104 105 106 1074.0x10-4
6.0x10-4
8.0x10-4
1.0x10-3
(b)
σ(o
hm-m
)-1
ω (rad-s-1)
experimental plotn=0.91n=1.15
Fig. 6 Frequency dependent conductivity of the nanocomposite samples at room temperatures. The solid
and dotted lines are theoretical fittings for (i) 1 0 MHz£ . and (ii) 1 5MHz,≥ . respectively.
Fig. 5 Plots of ln τ and ln σdc with B
1/ .Tκ The straight
lines are fittings for the Arrhenius-type temperature de-
pendence.
phys. stat. sol. (b) 244, No. 7 (2007) 2663
www.pss-b.com © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Original
Paper
former case the fittings are shown by solid lines and in the later by dotted lines. The value of n in the
lower frequency region equals to 0.91 and in the higher frequency region it is determined as 1.15. Ac-
cording to Funke [16] a value of n smaller than 1 means that the hopping motion involved is a transla-
tional motion whereas a value greater than 1 would mean that the motion involved is a localized hopping
for the carrier without leaving the neighbourhood. So, in the 2 3
ZnO Al O- nanocomposite thin film it is
obvious that in the total frequency range from 500 Hz to 5 MHz, the total ac response of the conductivity
involves both of the hopping process, i.e. the long range and the localized types. In course of the theo-
retical fitting the obtained temperature dependence of the dc
σ can be expressed as
dc
dc 0
B
exp ,E
Tσ σ
κ
-Ê ˆ= Á ˜Ë ¯ (9)
where 0
σ is the pre-exponential factor and dc
E is derived from the linear fitting as shown in Fig. 5 being
equal to 0.116 eV. The negative slopes in this case indicate that dc
E can not be interpreted as the activa-
tion energies of the system. The deviation of the relaxation time from the fitted straight line reflects the
uncertainty of the values of M
ω estimated by fitting of the model to Cole–Cole type of relaxation plots.
We also observed that the frequency dependent conductivity decreases with the increase in the meas-
uring temperatures. This decrease in the conductivity is contrary to the results obtained in various re-
search works on ZnO [17–19]. The reason for this may be two. The first is that the improper connection
of the leads with the electrodes might have yielded erratic results. But as we have repeatedly obtained
similar results using aluminum or gold as electrodes, we thought this metallic behaviour of the system
might be due the second reason. The metallic behaviour of ZnO has a precedence in the work of Miller
[20] and Natsume [21]. The decrease in the dielectric strength, real and imaginary part of the dielectric
constant and ac conductivity with the increasing measurement temperature can be best explained on the
basis of the concentration of the interstitial zinc atoms and free electrons in the nanocomposite. The
decrease in the ac conductivity can be squarely put on the decrease in the number of free electron and
that may be due to the change in the number of interstitial Zn atoms interacting to form pairs. The pair
formation of course depends upon the closeness of the neighbouring Zn atoms so that they are not too far
for appreciable interactions. At an initially higher measuring temperature when the pair breaks up, the
free electrons are captured by the interstitial Zn ions, decreasing the number of free electron density and
this leads to a decrease in the ac conductivity of the nanocomposite sample. At this point we like to stress
that the reasoning for the decrease in the ac conductivity is purely based on the reproducibility of the
measured data. And we do think that more works with better contacts and use of different materials as
electrodes is necessary to understand the implicit nature of the carrier concentration in the 2 3
ZnO Al O-
nanocomposite system. Since we have studied a single composition of the thin film, namely with relative
molar concentration of ZnO as 20% in the Al2O3 matrix, study of the other compositions with higher
concentration of ZnO in the thin film may reveal more insight into the ionic conductivity owing to better
percolation effect in the system.
Figure 7 shows plots of the real and imaginary capacitances measured at different temperatures.
Cole–Cole empirical relationship [12] which describes the semicircular relaxation as given by
highfrequency (1 )* ,
1 ( ) i
i
hC C
j
ε
ωτ-
È ˘= + Í ˙+Î ˚
(10)
where ih ranges between zero and 1,
iε denotes the chord length and τ denotes the time constant (whe-
re peak( ) =1ωτ ). The complex capacitance *C is expressed as 1 2
*C C jC= - . The experimental data do
in fact lie on the circular arc centered at a point on and above the real axis, which corresponds to the
appearance of a depression angle 0θ π and indicates a Cole–Cole type relaxation phenomenon. The
area of the semicircle decrease and dielectric strength s
(∆ )ε ε ε•
= - with the increase of the measure-
ment temperatures. The intersection of the solid line on the real axis in the lower frequency region may
be ascribed to be the contribution of the grain which form a thin insulating layer . The total grain
2664 S. Das and S. Chaudhuri: Dielectric relaxation and electrical conductivity in ZnO–Al2O
3
© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.pss-b.com
0.0 4.0x10-9 8.0x10-90.0
2.0x10-9
4.0x10-9
Cg
Rg
RL
Im[C
] (F
)
Re[C] (F)
T=315 K T=325 K T=345 K T=355 K T=395 K T=450 K
capacitance g
C may be defined as g g 0 /C A dε ε= , where gε is the relative dielectric permittivity of the
grain region. The experimental data in the lower frequency region show an asymptotic increase which is
attributed to the contribution from the external load which can be represented as L
R . Therefore, on the
basis of the complex impedance, dielectric permittivity and capacitance the ac response of the
2 3ZnO Al O- nanocomposite thin film can be represented by a circuit as shown at the inset of Fig. 7. The
equivalent circuit has been modeled by a parallel RC element resembling the characteristics of the sam-
ple i.e. it is consisted of a frequency independent capacitance g
( )C and resistance due to grain g
( )R in
series with an external load resistance L
( )R . Thus, we have
2 2 2
g g
1 2 2
g L
1 C RC
C R
ω
ω
+
= ,
2 2 2
g g
2 2 2 2
L g g g
11
(1 )
C RC
R C R R
ω
ω ω
+È ˘= Í ˙
+ +Î ˚ . (11)
It is clear that 1
C varies with the applied frequency and is equal to infinity at zero frequency and only g
C
at higher frequency. The capacitive dispersion as a function of frequency can be used to determine the
defect density contribution on the relaxation. The applied ac voltage facilitate the defect level to move up
or down or in course to cross the Fermi level. At this point a change of the charge of the defect take place
which in turn causes dispersion in the complex capacitance. Therefore, the defect density can be deter-
mined from knowing the complex capacitance ( *)C . The defect density can be calculated from the rela-
tion 2/C qA=D [22, 23], where q is the elementary charge. The defect density of the
2 3ZnO Al O- is
determined to be in the order of 15 2 110 m V .
- - At this point, once we relate the defect concentration and
the imaginary capacitance 2
( )C of the system, it is interesting to note from Fig. 6 that the defect concen-
tration decreases significantly with increasing measuring temperatures, though the reduction is more in
the lower frequency region. The observed reduction can be assessed in the following way. It is predicted
[24] that the relaxation strength has a linear dependence on the defect concentration and the square of the
dipole shape factor based on the concept that each point defect (e.g. oxygen vacancy) creates an elastic
or electric dipole. On the other hand according to Norman [25] oxygen adsorption on the surface of poly-
crystalline ZnO can lower the oxygen deficiency by more than 30%. So, we can conclude that the initial
measuring temperatures facilitate the adsorption of oxygen, thus reducing the defect densities, dielectric
strength, and area under the curve in the plane of complex capacitance. Accepting that oxygen adsorption
is a possible mechanism for the oxidation of the surface of the grains, the intrinsic donor species at that
grain boundaries, oxygen vacancies or zinc interstitial, will be annihilated. Since donors are the major
defects contributing to the n-type character of the ZnO conductivity, any decrease of these defects will
turn the composite a resistive one, a case which we observe with increasing temperatures as described in
Fig. 6.
Fig. 7 Cole–Cole type relaxation plot of complex
capacitance at different measuring temperatures. The
equivalent circuit is shown at the inset.
phys. stat. sol. (b) 244, No. 7 (2007) 2665
www.pss-b.com © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Original
Paper
4 Conclusion
In summary, the physical properties of the capped ZnO nanoparticles in the alumina matrix in the thin
film form is successfully investigated by dielectric relaxation phenomenon. The ac response of the con-
ductivity, capacitance and relative permittivity in the frequency range of 500 Hz to 5 MHz was investi-
gated to represent the system by an equivalent circuit consisting of a parallel RC component. The defect
concentration of the nanocomposite thin film was determined with the knowledge of frequency depend-
ent complex capacitance which arisen out due to the grains and insulating layers in between. Significant
reduction in the defect concentration was observed with increasing measuring temperatures.
Acknowledgements This work is performed in the Nanoscience & technology initiative programme of Depart-
ment of Science & Technology (DST), Govt. of India. The authors like to express their sincere gratitude for financial
assistance from DST during the tenure of the programme. One of the authors (S.D.) would like to pledge gratitude to
Mr. Kamalakanta Das, Technical Officer, IACS for useful technical help.
References
[1] B. Ismail, M. Abaab, and B. Rezig, Thin Solid Films 383, 92 (2001).
[2] D. S. Ginley and C. Bright, MRS Bull. 25, 15 (2000).
[3] C. Bundesmann, M. Schubert, D. Spemann, T. Butz, M. Lorenz, E. M. Kaidashev, M. Grundmann, N. Ashke-
nov, H. Neumann, and G. Wagner, Appl. Phys. Lett. 81, 2376 (2002).
[4] L. F. Dong, Z. I. Cui, and Z. K. Zhang, J. Nanostruct. Mater. 8, 815 (1997).
[5] Y. Fukuda, K. Aoki, K. Numata, and A. Nishimura, Jpn. J. Appl. Phys. 33, 5255 (1994).
[6] M. A. Alim, M. A. Seitz, and R. W. Hirthe, J. Appl. Phys. 63, 2337 (1988).
[7] K. I. Hagemark, J. Solid State Chem. 16, 293 (1976).
[8] Y. Natsume, H. Sakata, and T. Hirayama, phys. stat. sol. (a) 148, 485 (1995).
[9] E. Ziegler, A. Heirich, H. Opperman, and G. Stover, phys. stat. sol. (a) 6, 635 (1981).
[10] S. Ezhilvalavan, M. S. Tsai, and T. Y. Tseng, J. Phys. D, Appl. Phys. 33, 1137 (2000).
[11] A. A. Nazarov, A. E. Romanov, and R. Z. Valiev, Nanostruct. Mater. 4, 93 (1994).
[12] R. H. Cole and K. S. Cole, J. Chem. Phys. 9, 341 (1941).
[13] C. Ang, Z. Yu, and L. E. Cross, Phys. Rev. B 62, 228 (2000).
[14] D. Damjanovic, Rep. Prog. Phys. 61, 1267 (1998).
[15] G. C. Psarras, E. Manolakaki, and G. M. Tsangaris, Composites A 34, 1187 (2003).
[16] K. Funke, Prog. Solid State Chem. 22, 111 (1993).
[17] Ce-Wen Nan, A. Tschöpe, S. Holten, H. Kliem, and R. Birringer, J. Appl. Phys. 85, 7735 (1999).
[18] Y. Igasaki and H. Saito, J. Appl. Phys. 69, 2190 (1991).
[19] G. W. Tomlins, J. L. Routbort, and T. O. Mason, J. Appl. Phys. 87, 117 (2000).
[20] P. H. Miller, Phys. Rev. 60, 890 (1941).
[21] Y. Natsume and H. Sakata, J. Mater. Sci., Mater. Electron. 12, 87 (2001).
[22] S. Ezhilvalavan and T. Y. Tseng, Appl. Phys. Lett. 74, 2477 (1999).
[23] M. S. Tsai and T. Y. Tseng, Mater. Chem. Phys. 57, 47 (1998).
[24] A. S. Nowick and B. S. Berry, Anelastic Relaxation in Crystalline Solids (Academic Press, New York, 1972).
[25] V. J. Norman, Aust. J. Chem. 20, 85 (1967).