impedance jonscher's power law impedance -...
TRANSCRIPT
CHAPTER N
A C CONDUCTMTY AND ELECTRIC MODULUS ST'UDIES OF
SCV, SCP, SCA AND SBS SAMPLES
4.1. A.C. THEORY
4.1.1. Impedance Spectroscopy
4.1.2. Dielectric Permittivity
4.1.3. Electric Modulus
4.1.4. Jonscher's Power Law
4.2. RESULTS AND DISCUSSION
4.2.1 . Complex Impedance
4.2.2. Frequency Dependence of Conductixit).
4.2.3. Dielectric Permittivity
4.2.4. Electric Modulus
4.3. CONCLUSIONS
REFERENCES
CHAPTER IV
kc. CONDUCTIVITT AND ELECIRIC MODULUS STUDIES OF
8CV, SCP, BCA AND SBS SAMPLE8
One of the important techniques to characterize many of the electrical
and electrochemical properties of fast ionic conductors is the conductivity
spectroscopy 11-31, In general, conductivity measurements are made by
applying the d.c. bias a m the material, but it results in the polarization at
the electnxlr-electrolyte interface, which opposes the applied field and hence,the
ionic current falls with time. To overcome this problem, the four-probe or two
rrversibl~ elpcbde methods are adopted to measure d.c. conductivity of the
materials 121. later, the single frequency measurements are canied out, but
thew methods are found to be madquate for understanding the complete
electncal bhavior of the materials, since the electrochemical processes
repmnted by banous md~mdual elements (K, C & L) are frequenw dependent
16 71 I J r r~c (~ , the a L technique has t w n de\eloprd ro rneasur m w a n c e
ober a range of frequencies to estunate the exact hulk conduchwtv and
frequrnof depndrnt conducmw m FICs From the unpedanm study, one can
o h m not only the bulk conduch~ity but also gram b u n d a y effects, loruc
transport, double layer formation at the electmdc-/electrolyte mterface, etc
Hence, m m n t years, Impedana- Spectmscup\ (IS] has become a powerful
techruque for charac temg the electncal properties of the FIC rnatenals and
thev mterfaces wth electron~caly conductmg electrodes Also, the IS
measurements can pmvide the dparmc pmperties to understand the
microscopic nature of the FIC materials 12, 5,6j.
Impedance Specw'opy is also d d Immittance S p e c m p y (IS] and
it is used to measure complex impedance Z', complex admittanw Y', complex
dielectric permittivity E' and complex modulus M' I2,9]. This chapter briefly
describes the a.c, theory, measurement & analysis of impedance data, a.c
conductivity and electric modulus. It also gives the detailed discussion of
temperature dependent of conductivity, ac conductivity, dielectric permittimh
and electric modulus studies to understand ion dynamics in the glassy
systems.
4.1. kc. THEORY
4.1.1. Impedance-
The complex unpedance Z/o) is defined as ratio of voltage V(t) to current
l(t) in the time domain on appl.yng a sinuso~dal signal of IOU arnphtude acmss a
solid electrolyte
whex I Z I is magnitude and ($1 is phase angle.
Impedance or lmrmttance spectroscopy means any one of the lollowing four
quantities
Complex impedance Z' = Z' -jZ"
Camplex Admittance Y* = Y ' + J Y " =
Complex permittjvity c' = c' - jt" = 11 ,jcK'J*
Complex ~ O ~ U ~ U S M* = M '+ jMW = jCQA'
Wher j= 4- 1, 6 is the vacuum capac~tancr and t~27rf is thr angul;v frequency
The admittance and permittivity are parallel functions characteristic at low
frequencies whereas the impedance and modulus are series functions at high
frequencies ( l ,6 ] .
The complex plane is used to represent the electrical quantities of real
and imaginary parts of complex impedance (2'). complex admittance (Y'),
complex permittivity (E*) and complex modulus (M*) 15, 71. The impedance and
admittance representations for the parallel & series combinations of an id& RC
circuit elements are discussed.
In series: The complex impedance for the series combination of
rsistance R and capacitance C, is given by
Z* = Z'-IZ" 4.6
u hms Z'- H, and 4 7
Fig. 4.1.a shows the resultant unpedance for R & C m series is
rcpmnted by a vertical line parallel to the irnggaq a d s lnrerszcting the real
;ids at R. The companding admittance plot @yes the semicircle intersecting
the rPal axis at the origin and at a point l / R as shown m f i . 4.1. b.
In parallel. The complex unpxlnntr for rhc parallel combmauon of % and C, 1s gven by
I I - = - t 4 0 Z * K,
Z' = Z'- /Z"
where,
1 ~ g . 4.1. a) Z vs. Z &, b) B vs.(; for xncs and c] Z" vb. 2' 8 d] H vs.G for pmllr! combinations of RC curuits
The complex impedance for a parallel combination of R, & C, represents
a semicircle, as shown in the Fig.4.1. c, mterseaing the rPal axis at origin & at a
point R. The difference between the two intercepts gives the bulk resistance [&I of the material. The equation of a circle uilh radius R/2 and center at (R/2.0) is
obtained from equations 4.10 & 4.11.
The corresponding complex admittance plot shorn in fig. 4.1.d gives the
straight line parallel to the irnaginq ads mtersecting the real a d s at l/R. The
~r.klionsh~p k twwn thc pariilkl and sc.nc.5 (ompnrnts is obtainrd frum the
equations 4.7,4.8,4.10 & 4.11
F g 4 2 a & b show the equnalenr cvcu~ts & the ~rnpedance plots
rspectlvely for an ]deal and ral sohd electml\?es From the fig 4 2 a, lt IS
observed a vemcal stcaght h e and a perfect srm~cucle in the unpedanw plot
for an ideal sobd electrohze From the fig 4 2 b. ~t IS observed that the inched
s t r z t h e 1s due to the presence of double layer capaatancp of elecbude
electrolyte interface. and the depressed senucucle cornponds to the parallel
combinahon of reslstanw and capautancr The angle of lnclinabon of the
s w t h e and angle of depression of the s e n u d e are due to the dsmbuted
81
la)
(b)
Fig. 4.2. Impedance spectra of a) ideal, b) real & c) polycn.;tallinc solid rl?i!roi!.rt.a
mimscopic material properties, termed as constant phase elemenr (CPE) (8.91.
The complex impedance of the real solid electrolyte is ispresented as
where 0 2 a I- 1. When a = 0, Z is frequency independent and K is the
resistance R, where as, when a = 1, Z = - jK/m, the CPE corresponds to
capacitance C. The fractional power law dependence on frequency is referred to
a s a constant-phase-angle element or non-Debye capacitor 110, 111. Fmm the
fig 4.2.b, the equivalent circuit for real solid electrolyte is given by the parallel
combination of CPE and bulk resistance in series with CPE of the interface.
Rg 4 2 c shows the unpedance spectrum for a polycnstaYlne sample
wth the two semnrcles, one due to resistance w~thm the grams of the matenals
and other due to the pamal or complete blockmg of charge camers at g m
houndanes From the fig 4 2 c the q u n almt clrcu~t for the polvcmtallme
samples IS given bv the senes of parallel combnabon of CPE & bulk reslstance,
CPE & gram boundary reslstance and the CPE of the nterface The gram
boundaries may act as a hindrance to the lomc transport or mav also
contribute to the h@ conduchon path due to the presence of large defect
density m the mterfaual w o n [7, 8, 12, 131
The complex impedance analysis has been performed using
A.Boukamp's equivalent cixuit softwan, in which non-linear least square
(NLLS) fittmg p d u r e is adopted (141. The NLLS fitting is pelformed in two
steps, i) by the pamal NILS fit, where the suitable points are selected on the
stra@t line part of the impedance data through which a tangential line is
drawn. The intersection of the line dth the real axis gives the bulk mistance.
ii) by choosing a set of three data points in the high frequency depressed
semicircle, a serniclrcle is fitted through origin and the W a l N I L 5 fit gives the
84
values for the resistance and CPE parameters. The entire impedance
distribution is simultaneously fined to the NLLS fit to obtain a set of parameters
(e.g. circuit elements) for the equivalent circuit. The most of systematic errors
are reduced and the parameters of the complex equivalent circuit are obtained.
If a varying field V(t) is applied to a matelial, then the polarization and
i n d u d charge Q are related as 1151
C) = c * I ,<, exp( /d) 4.16
where E* is the complex dielectric constant. The frequency dependent complex
dielectric permittivity is given by
(L'c) -1 4 17 (C i ) ( I ~ J O T )
uhere h and ~a arr the low and high frqurnry d~elertnc constants respecbvel\.
(1 = 2x1 is the angular fnquency, r is thc tune constant The E* is #\!en b)
C* = c8-,c,' 4.18
where E' is the relative permittivity or delrctric constant, 6" is the dielectric loss.
Separating the real and imaginary parts of the equation 4.17,
log f
Fig 4.3. Dielectric parametem a) c', b] E" & c) Ian fi as funcuon of frequenn
The above eqns. 4.19 to 4.21 are called Debye equations and 11. 4.3.a, b
& c show respectively E', E" & tan 6 as a function of frequericy for an deal
material 1151. In fig. 4.3 a, , E' is independent of frequency at low frequencies
and with increasing frequency, it passes an inflection at o = l/r and also, at
hlgher frequencies, it becomes asymptotic to %. In the fig. 4.3 b, the E" exhibits
a maximum at on = 1, where the oscillating charges are coupled with the
oscillating field & absorbs a maximum electrical energy. From the fig. 4 . 3 . ~ it is
found that tan6 exhibits a symmetric Debye behavior with a loss tangent
maximum at on =(t. / &)1/2.
Many real materials show a deviation from an ~dral equahons and exhibit
a non-Debye dielectric behavior. The non-Debpe dielectric response can be
described using Cole-Cole, Davidson-Cole and empirical expression proposed by
Iiavriliak-Negami (H-N) (16 - 181.
The dlelectnr prrmltt~wty c' IS rplnted to complex ~mwdancp Z* h\
1 C* --- 4.22
l(d',Z *
o 1s the angular frequency and 6 = t / A o b 1s the capacitance of free space
In the present study, the real and unaginanr parts of the dlektnc
W m t ~ w t v are calculated uslng the above quatlons 4 23 & 4 24, pellet
dlmens~ons and the measured unpedance data
The complex electric modulus is defined by the reciprocal of the complex
permittivity [I] .
where M' is the complex modulus, E' is the dielectric permittivity, M' is the real
and M" is the imaginary parts of modulus.
The complex elmmc modulus spectrum represents the measurr of the
dlstnbubon of Ion enewes or config~ratlons m the structure and it also
descnbes the electrical mlaxab~n and nucroscop~c prupertles of lonlc glasses
(19, 20) The modulus formalism has been adopted as ~t suppresses the
p o l m t l o n effects at thr eiet trodcirlect~'~l\lc interface Hence, the con~ple-s
electric modulus M (o) \pettra rflectb the d\narmc properneb of the sample
alone
For parallel combination of RC element, the lpal and imaginary parts of
the modulus are given by
F'lg 4 4 a shows the r a l M' t s magmn M" parts of ~ ~ r n p l e \ modulus
Spectrum for the parallel ro~nbindbon of RC tn a senucurle u l t e m 11ng the d
$7
a d s at 0 and b/C. In the fig. 4.4.b, for an ideal sohd electrolyte, the modulus
M" vs. log f represents the Debyr cutve with a single relaxation time and a
madmum at or = 1. In fig. 4.4.c, the impedance Z" vs. log f give the Debye peak.
the FWHM is found to be 1.44 decade and the corresponding p value is 1,
whose peak maxima coincide the peak of M vs. log f plot.
In case of real solid electrolyte, the modulus spectra exhibit a broad and
asymmetric non-Debye nature wth distribution of relaxation times. The
variation of relaxation time is continuous and is represented by normah&
function for the distribution of r~laxation times g(rl 12 11. Then the relaxation
function ~ ( t ) in time domain 1s given by
ifthr rrl~ualion tlmr r is thrmall~. ;~rtivatcti. rhrn
w h r r F, 1s the relaxabon acuvatlon ? n e w , k 1s ~ h r HOltzmann constant. T 1s
Ihr absolutr t e m p m t u r In thc frvquenm doman, the reiaxatlon funcnon is
exprssd m electnc modulus as
log i
Fig. 4.4. a) M vs. M', b) M" vs. log f and c) Z" vs. log f plots of lded RC riernrnts
where $(t] is the decay function given by
40) = exp[- ( f l r ~ ] 4.34
p is the stretched exponent, r is the relaxation time. The b(t) is the stretched
exponential KWW relavation function and is written as sum of exponential
terms
The g, coeniclents are evaluated by hrar least square fit of the above equahons
4 35 & 4 36 The lnput parameten p, M s & r for the fit are obtained from the
measured modulus data and wen? fitted to the h W funchon followmg the
method desmkd by Moyn~han et all2 11
4.1.4. Jorucher's Power Law
A typical frequency dependrnce t:onductiviv spectrum that shown in the
fg. 4.5 evhiblt three distinguish rgime a] low frequency dispersed b) an
intermediate plateau and c) conductivity d ~ s ~ r s i o n at hlgh frequency 115, 22 -
24). The variation of conductivity in thr low frequency rrgion mav be amibuted
to the polarization effects at thr el~trodeelectmlyte interface. At very low
frequenaes, morp charge accumulation oecurs and hence, drop in conductivity.
In the intermediate frequency @on, conductivity is almost found to be
frequency independent 8s equal to dc conductivity and at the hlgh hquency
region, the cpnductivlty increases with frequency. The frequency dependence of
electrical conductivity of solid electrolytes is explained by a simple expression
given by Jonscher's power law. The power law relates the frequency dependent
conductivity or the so-called universal d.wamic response (UDR) of lonlc
conductivity and frequency by 1'25,261
XY
plateau region
___--- 8
/ polarisation region
log f
Ftg. 1 5 Schematic representauon of log conciticu\;t\ s f rquenc?
where a , is the ac conductivity, 00 is the limiting zero frequency conductiiq
(udc), A is a pre-exponential constant, o = 27rf is the angular frequency and s 1s
the power law exjmnent where 0 < s < 1.
4.2. RESULTS AND DISCUSSION
4.2.1. Compla Impedance
The real and imaginary parts of complex impedance measurements were
made as a function of Erequency as well as at low temperatures (120 to 300k)
on the pressed pellets of type
conducting elechDde (silver) / sample / conducting electrode (sllverj
for formers, mdifier to formers and dopant compositions of SCV, SCP &. SCA
glassy samples. The measurd impedance data are anal@ using the
equivalent circuit software developed by A.Boukamp. All the circuit parameters
are adjusted simultaneously, m order to fit the measured data through N U
procedure to obtain the bulk resistance and the equivalent circuit of the
materials.
Fig. 4.6 a , b & c show the real (Z') and imaginary (2'7 parts of the
impedance spectra obtained at various low temperatures (120 to 300K) for thr
hgh conducting dopant salt composition of SCV, SCP 5, SCA samples
respectively. I b m the fig. 4.6 a, b & c, it is observed that with increase in
temperature in the lower range 120K to 200K, the formation of complete
dep& semicircle occurs and on further increase above 200K, the deprpsed
~emiclrcle starts disappear with a formation of an inched straight line. F'rom
the above fgures, it is observed that the intersection of th t semicirclr with thc
00
real axis shifts towards the origin with inrrease m temperature and the
frequency of the intersection with real axis also increase with temperature.
Similar behavior is observed for the low temperature dependence of the
impedance for former, modifier to formers and dopant composition respectively
of SCV, SCP & SCA glassy samples. The analysis of the spectra by the NLLS fit
yielded the values of the bulk resistance Rb from the intercept of the semidrcle
with real ads . The electrical behavior of the samples 1s represented as the
parallel combination of bulk resistance and distributed elements CPE due to
the depressed semicircle.
The bulk conductivity is calculated using the bulk resistance obtained
from the analyzed impedance data at the various temperatures (120K to 300q
and pellet dimensions for all the former, modifier to former and dopant salt
compositions of SCV, SCP & SCA glassy samples. Fig. 4.7 a, b & c show the log
(oTJ vs. 1000/T plots for all the dopant compositions of SCV, SCP & SCA glassy
samples respectively. The temperature dependence of conductidty obeyed the
Arrhenlus relation & activation ? n e w ('4 1s obtained from the slopes of the
linear fit. Fig. 4.8 a , b & c show mpectively the variahon of F_g with dopant salt
composition of all the SCV, SCP & SCA systems.
korn the fig. 4.6 a, b & c, it is observed that the depressed semicircle in
h~gh frequency w o n represents parallel combmauon of the bulk resistance
and bulk capacitance of the glassy materials (27 - 291. The center of the
depressed semicircle displaced below the real axis and it may be due to the
prrsencr of distribution of various elements at the elecb'olyteelectnxle system,
which indicates that the relaxation time r is not a single valued but distributed
cuntinuously around a mean value 11, 131. Hence, the distribution of various
elements in the system is represented by the constant phase w e element
CPE. The equivalent circuit is representd by the parallel cumbmauon of the Rb
and CPE for each glassy sample 1 I, 30.3 I ] .
Fig. 4.6. a impedance spectra for lugh conducting mmposluon of SCV glassy samples at Merent temperdtuRs
' f, *.
D 0 10 70 30 10 70
Zrd, [ohn~! (xlW'31 ,
Fig. 4.6, a Impedance spectra for high c o n d u m g co:n?oslrnn of SC'J glassy samples at different temperdturos
F!g. 4.6, b Impedance spectra for h ~ g h conductmgcaIlipos!tion of SCI'g:;is\y samplcs at dflerent ternperdtum
I:lg. 4.6, b impedance spectra for hgh conductvlg colnposiuoi> ol SCP glassy samples at diflerent temperatures
Fig 4.6, c, Impedance spectra for h~gh conducting composttion of SC, g!ass! samples at dinerent temperatures
+ 188K
\ x \;:;:
b . - - -
10 20 Z.r-1, [ohm] ( x l O A ? ) -->
Fig. 4.6. c, Impedance spectra for high conducung composition of SC.4 glassy samples at different temperatures
Fig. 4.7 Log(fl) vs. 1000/T plots of various dopant compositions of a) SCV, b) SCP and c) SCA samples
I I I
40 50 60 70
Dopant composit~on
Dopant composition
dopant composition
Fig. 4.8 Activation energy for different various dopant cornpositlons of a] S e , b) SCP and c) SCA samples
4.2.2. Frequency Dependence d Conductivity
The ac conductivity is calculated from the measured real & imaginary
parts of the impedance data and pellet dimensions for all the formers, modifier
to formers & dopant compositions of the SCV, SCP & SCA (120 to 300K) and
formers & modifier to formers compositions of SBS (300 to 473K) glassy
samples. Fig. 4.9 a, b, c & d show respectively the frequency dependence of
conductivity obtained at various temperatures for all the dopant compositions of
SCV, SCP & SCA and modifier to former compositions of SBS glassy samples.
From the fg 4.9 a, b, c & d, it is o b s e ~ e d that the conductivity spectra show the
hvO distinct regimes, within the measured frequency window, 1) plateau and 2)
dispersive regions. The plateau region corresponds to frequency independent
conductivity or known as dc conductivity. The a, is obtained on extrapolation to
loner frequencies and found to be in good ageement with the dc conductivity
o1)tained from the impedance analysis. The frequency dependence of
conductivity in the dispersive rgions for all Ihr compositions of the SCV, SCP,
SCA & SBS samples at various temperatures were analyzed using the Universal
Jonscher's power law.
The conductivity data were fitted to power law using non-linear
regression procedure. From the analysis, the fitting parameters 00, A & s were
obtained and table 4.1 gives 0, & s for the high conducting SCV, SCP, SCA &
SBS samples. Fig. 4.10 a, b, c & d respectively show the temperature
dependence of s for the dopant compositions of SCV, SCP bSCA and modifier
to former compositions of SBS samples. Fig. 4.1 1 a, b, c & d show the log (o/o,)
vs. log (a/%) plots at various temperatures, where o, is the characteristic
frequency at which om = 20~.
Fig. 4.9 Log (o) vs. log (a) plots for the hghest conducting dopant composition ofa) SCV & bj SCP samples at diRerent temperatures
a l , . 2 3 4 5 6
log (ol I
Fig. 4.9 Log (a) vs, log ((0) plots for the highest conducting c) dopant composition of SCA & dj m/f composition of SBS samples at different temperatures
Table 4.1. S, 4 8a 00 at temp (4 for high conducting SCV, SCP.
SCA and SBS samples
1 SCP / 165K / 0.643 ) 371.5 / 2.6073 x10" 1 1 A 1 l68K 1 0652 1 94.2 1.7476x107
SBS 413K 0548 62.8 1 177x10'
In the fig. 4.9 a,b,c & d of logo vs. logo plots, the frequency at which the
dispersion region starts from the dc conductivity plateau can be defied as the
characteristic kquency (%!,where the relaxation effects of the ions occur. Thls
characteristic frquency is termed as hopping rate ((+I or cross over frequency
and it occurs at a, = 20,1241. The relation between the dc conductivity and the
t~opping rate is given by a, = &, where k 1s the empmc~l constant, which
dcpends on the concentration of mobile ions and details of the conduction
mechanism. Fmm the Eg 4.9 a, b, c & d, it is observed that the dispersion
region decreases with increase in temperature, where characteristic frequency
PI, shift towards the high frequencies, and disappears on further increase of
irmperature in the measured frequency window 132,331. The ion-hopping rate
is less at low temperature a s compared to that at high temperature. It is also
found that the % is themially activated with the same e n e w a s ooT.
In the fig. 4.10 a, b, c & d, the variation of s uith temperature can be
mlated to the existence of distribution of relaxation parameters (34,351. From
the fig.4.11 a, b, c & d, the superimpossability of log (O/oO) vs, log (o/%] at
various temperatures suggests the conductivity relaxation mechanism is found
10 be temperature independent 1361. 91
Fig. 4.10 s a s a function of temperature for the vanous dopant compositions of a) SCV, b] SCP & c)SCA and m/f compositions of d] SBS samples
Fig. 4.11 Log (o/oo) vs. log (01%) plots for the hlghest conducting dopant composition of a) SCV and b) SCP samples at different temperaturrs
Fig. 4.1 1 Log (o/oo) vs. log (44) plots for the highest conducting cl dopant composition of SCA and d) m/f composition of SBS samples at different temperatures
As obsewed in the conductivity spectra, the plateau region of
conductivity may be due to transport of mobile ions in response to the applied
electric field. The long-range transport of ion result in oniy successful jumps
and their contribution yields d.c. conductivity (a,). The dispersive conductivity
region with frequency can be explained with respect to the predicted theoretical
Diffusion Controlled relaxation (DCR) model 135-37. In this model, it is
assumed that the glass forming matrix at the non-bridging oxygen (Si-O-, B-0 , V-O~, P-O- & As-O-) which is associated with a negatively charged ions to be a
site and it can be doubly occupied by the cation (Ag+). When a silver ion
diffuses to a singly occupied site, the positions of the original ion on those sites
excite and possess the higher e n e m configuration. At that time, the ions
mutually rearrange within the region to achieve a new lowest energy site result
m the former ion relaxation at that site is presumed to occur almost
instantaneously. The excited ion can relax back to its ori@ site or can diffuse
to another adjacent site. This indicates the cation hopping and diffusion
mechanism of the mobile silver ion at the local structures of the glass samples.
Thr lower s value obtained h m the plurr law r.uprssion can be correlated
with the higher successful jumps, which result in the hgher dc conductivity
obtained for the former, m/f and dopant varied compositions of the SCV, SCP &
SCA and former & m/f varied compositions of SBS samples.
4.2.3. Melectrif Permittivity Studies
From the measured 2' & 2" impedance data, E' and E" imaginaq
parts the dielectric permittivity were calculated according to the equations 4.23
& 4.24 for all the formers, m/f & dopant compositions of SCV, SCP & SCA and
formers & m/f compsitiohs of SBS systems.
Fig. 4.12 a, b, c & d show the real part of dielectric permittivity E' vs, log f
for the dopant compositions of SCV, SCP & SCA and modifier to former
Fig. 4.12 8' vs. log (freqi plots for the hlghest conducting dopant composition of a) SCV and b) SCP samples at various temperatures
Fig. 4.12 E' vs, log (fred plots for the highest conducting c) dopalt composition of SCA & d) m/f composition of SBS samples at various trmperatures
compositions of the SBS samples respectively. Fmm the fig. 4.12 a , b, c & d, it
is observed that with increase in frequenq, E' decreases and attains a constant
value at high frequencies. At high frequencies, due to the high periodic reversal
of the a.c field, there is no charge accumulation at the interface and hence the E'
remains constant, which can be explained in terms of the ion ~ s i o n
mechanism. At low frequencies, the charges get accumulated at the interfacial
region that lead to a net polarization of the ionic medium result in the fonnation
of space charge region at electrode-electrolyte interface which in turn increases
the dielectric constant E'. 124, 38-42]. In s" vs. log f, the loss peak could not be
observed in the measured frequency window and hence, the electric modulus
formalism is used for further analysis.
4.2.4. Electric Modulus
The electric modulus data is cdculated from the real & imaginary parts
of the measured impedance data and pcllet dimensions respectively using the
rquauons 4.27 & 4.28 in section 4.1.3 for all thr formers, m/f & dopant
compositions of SCV, SCP & SCA and formen & m/f compositions of the SBS
samples. The relaxation behavior is ana?vzed using the complex electric
modulus (M' = M' + j M ) formalism and the decay function 4 given by
Kohlrausch-Wk-watts ( k v . Fig 4.13 a, b, c Pu, d respechvely show the
unaginary part of electric modulus M vs. log f for the high conducting dopant
compositions of SCV, SCP & SCA samples and high conducting m/f
compositions of SBS samples measured at various temperatures.
R-om the fig 4.13 a, b, c & d, it is obsenred that the shape of the curves
are asymmetric with a long tad extending in the higher fquency region
exhibiting non-Debye behavior and it also o b s e ~ e d that the peak relaxation
frequency, f,, shfts towards the hlgher h q u e n q region with temperature.
The M curves a s a function of frequency werp fitted to equation 4.3 1 using the
9 5
procedure of Moynihan et al. by a least s q u m iterative software developed m
Fortran 77 as discussed in section 4.1.3 for all the dopant compositions of SCV,
SCP, SCA & m/f composition of SBS samples 12 11. The initial parameters M,, r
and fl are used in the fitting pmedure to obtain the best fit. The M, value are
obtained from the hlgh frequency limit of the M' vs, log f plots. The r & P are the
inverse of the peak relaxation frequency at M", and the full width half
maximum (FWHM) of the M vs. log f c u ~ e s respectively. The continuous lines
represent the fitted values of M , whereas the symbols represent the
experimental data. The value of P is graphically tuned such that the simulated
modulus curve could be superimposed on the experimentally obtained modulus
curves. Fig 4.14 a, b, c & d shows the temperature dependence of value for all
dopant compositions of SCV, SCP & SCA & rn/f compositions of SBS samples.
Fig 4.15 a, b, c & d show the log r vs. 1/T plots of all the dopant SCV, SCP &
SCA and modifier to former compositions of SBS samples. The log T vs. 1/T
plots are fitted to
by the linear least square fit and the relaxation activation energy is obtained
from the slope of the fit. The obtained Er, b, f,,% and !j values are gven in table
4.2 for the high conducting composition SCV, SCP, SCA & SBS systems. Fig
4.16 a, b, c & d show the normalized M"/M"t, vs. log f/f- obtairied at
different temperatures respectively for all the dopant compositions of SCV, SCP
& SCA and m/f compositions of SBS samples.
According to the obtained results, the formers, m&er to formers &
dopant compositions of SCV, SCP & SCA and former & m/f compositions of
SBS samples, the M curves are related to the energy dissipation in the
hversible conduction p m s s and they exhibit a non-exponential character of
decay p m s s .
pig. 4.13 M vs. log jfreq) plots for the highest conducting dopant composition of a) SCV and b) SCP samples at different temperatures
4 - + 1 2 3 4 5
log ifrrq)
Fig. 4.13 M" vs. log (freq) plots for the highest conducting c) dopant compsiuon of SCA & d) m/f composition of SBS samples at different temperatures
Rg. 4.14 P a s a function of temperature for the various dopant compositicr~s of a) SCV, b) SCP & c)SCA & m/f co~npositions of dj SBS samples
Pig. 4.15 Log (7) vs. 1000/T plots for the various dopant compositions of aj SCV, b) SCP & c) SCA & m/f compositions old) SBS samples
Table 4.2. P & f- at temp(K) and Er & bfor high conducting SCV, SCP,
SCA and SBS samples
-- / ~ i g h I T e m ~ / I Relaxation 1 1
cond&ng System
SCV
The broad non-Debye modulus spectra were interpreted in terms of
distribution of relaxaflon times r, which in turn were related to the distributions
of energy baniers Ea [15, 361. The M" vs. log f curves for all the samples
showed a deviation over the simulated data at higher frequencies much larger
than the peak frequency f-, whjch could be explained by considering the
contribution of asyrnmetnc double well potential (ADWP) confguration to the
conductivity 133, 43, 441. Fmm the above results, it could k concluded that
the relaxation time r is not a single valued, but is distributed over a mean value
140, 451.
1 ,
From the fig. 4.13 a, b, c & d, with increase in temperature the
broadness of the curves remain almost same, where P remains almost constant
would imply that all elements of the distributions have about the same
activation energy. However, the shift in the fm, with temperature could be
explained based on the distribution of attempt frequencies for the barner
crossover or a distribution of jump or ll@t distances following the crossover.
The broadness of the M" vs. log f curves is interpreted in terms of the
distribution of relaxation times for distinguishable physical processes. Fig 4.14
a , b, c & d showed that the P value is almost constant with temperature
b 1 I
1.5523~10-'~ / activation energy (E)
SCP 1 165K 1
! I
70 1 0.232i0.01 1 1.642 x10-l3
showing that the relaxation processes are temperature independent. The 97
168K ( 0.68
0.54
0.6
SBS t'"" 413K
5.89
44 0.3145i0.02
0.54 i x . 8 0.6212i0.03 1 9.562 x10-l3
0.260610.02 1 7.585 x10-13 I
Fig. 4.16 Mn/M" max vs. log (f/f-) plots for the highest conducting dopant composition of a) SCV and b) SCP samples at various temperatures
Fig. 4.16 M"/M" , vs. log (f/f-) plots for the highest conducting c] dopant composition of SCA and d) m/f composition of SBS samples at various temperatures
activation energies for the relaxation Er obtained from the slopes of the log r vs.
1000/T plot in fig. 4.14 a, b, c & d are slightly higher than thermal activation
energy, Ea. In the fig.4.lb a, b, c & d, the supeninpossability of normaked of
M"/MWmax vs. log (f/f-) spectra revealed that the distribution of relaxation time
is temperature independent.
Fig. 4.17a, b, c & d show the Z" & Mot vs. log f plots obtained at two
different temperatures for the high conducting compositions of SCV, SCP &
SCA and m/f compositions of SBS glassy samples respectively. From the fg
4.17 a, b, c & d, the observed broadening of Z" and M" spectra is due to the
existence of a distribution of relaxation timrs of the intrinsically dispersive
elements of the materials. The peak madma of Z" and M" spectra do not
coincide, a s there exist a dispersion effect due to the distribution of relaxation
times [46,47].
Fig 4.18 a, b, c PL d show the M" & log o vs. log f plots obtained at two
diiferent temperatures for the h~gh conductmg composition of SCV, SCP & SCA
and m/ f compositions of SBS glassy samples respectively. From the fig 4.18 a,
b, c & d, it is observed that the f,, from the M vs. log f modulus curve and the
% fmm log o spectra occur at the same frequency and with increase in
temperature both f,, & 4 shifts to higher frequencies. The hopping rate (+
with respect to the conductivity spectra and also the f- in the M" plot could be
a sensitive function of ion concentration and also may k due to some other
material property 1331. As observed from the impedance, power law and
modulus analysis respective&, the frequen~y at which the intersection occurs
with real axis, the frequeng at which relaxation effects hgin and the peak
relaxation frequency of M"-, shift towards the higher frequencies with increase
in temperature. From these observations, the conduction mechanism by which
the Ag ions preclude the matrix are comlated with ion hopping diffusion [48].
Fig. 4.17 Nonnalised M & Z" vs. log (freq) plots for the hghest conducting dopant composition of a) SCV & b) SCP samples at various temperatures
Fig. 4.17 NormaJkd M" & Z" vs. log (freq) plots for the highest conducting c) dopant composition of SCA & d) m/f ratio of SBS samples at vanous temperatures
pig. 4.18 M & log (0) vs. log (freqi plots for the highest conducting dopant composition of a) SCV & b) SCP samples at various temperatures
Fig. 4.18 M &, log (a) vs, log (freqj plots for the highest conducting c] dopant composition of SCA & d) m/f ratio of SBS samples at various trmperatures
4.3. CONCLUSIONS
The impedance measurements were made for all the formers, m/f &,
dopant compositions of the quaternary SCV, SCP & SCA samples at low
temperature (120 to 300K) in the frequency mnge of 40Hz to 100kHz. The a.c.
conductivity, dielectric permittivity and electric modulus were calculated from
the measured real & imaginary parts of the impedance data and pellet
dimensions, using their corresponding inter related formalisms. In the analyzed
impedance spectra, with increase in temperature, the intersection of the
semicircle with the real axis shills towards the ongin, in turn Increase in
conductivity. 'Rum the Arrhenius linear least square fit of the log (a] vs. 1/T
plot, the activation energy of the mobile charge carriers were calculated.
The conductivity with frequency dependence spectra were fitted to the
Jonscher's power law expression and obtained the fit parameters 00, s & A. The
obtained 00 is in good agreement with the ode. The conductivity spectral results
are explained with the eldsting theoretical Dflusion Contmlled Relaxation (DCR)
model. The modulus spectra were fitted to the KWW decay function and
obtained the stretched exponent B, the relaxation time r and the shape of the
spectra is non-Debye conlimed that the samples are ionic in nature. For
various temperaturs, the p value is found to be almost sarne. The
superimpossability of the n o m a h d M" spectra at different temperatures were
ascribed to temperature independent mechanisms of relaxation. The
conductivity and the irnaginaq pa^? of modulus M with a function of frequency
compared & idenhlied that the peak frequency f,, & the o, occurs at the sarne
frequency, which determines the material response in the vicinity of the
conductivity.relaxation time. The analysis by impedance, power law & modulus
formalisms suggest that the Ag' are the mobile cations & the mechanism by
which they migrate are correlated with the DCR model to e x p h the Ion
hopping diffusion of @ ions in the glassy matrix.
REFERENCES
1. J .Ross Macdonald (ed.1, Tmpedance Spectroscopy', John Wiley & Sons, New York (1987)
2. J. F. Mccann, S.P.S. Badwal J.Elect&em.Soc., 129, No.3, (1982), 551
3. B. Roling J . Non-Cryst. Solids, 244, (1999),34
4. M. Le Stanguennec, S.R. Elliott Solid State lonics, 73, (19941, 199
5. A.Hooper 'Application of ax . measuvment and analysis techniques to materials research', AERE-R9757, (1980)
6. S.P.S.Badwal 'Solid State Ionic Devices', Chowdari. B.V.R. and RadhakrishnaS. (eds.), World Scientific, Singapore, (1988)
7. ,J.E.Rauerlr J. Phys. Chem. Sohds, 30 (1969) 2657
8. W.1 Archer, R.D Armstrong 'Electrochemistry', Chemical Soc. Specialist Periodical Reports, 6 (1978) 98
9. J . Ross Macdonald Electrochimica. Acta., vol. 35, no. 10, (19901, 1483
10. H.Engstrom, J.B.Bates, J.C.Wang Solid State Cornrnun. 35 (1980) 543
11. A.K.Jonscher a) Phys. Status. Solidi (a) 32 (1975) 665 b) J.Mater. Sci. 13 (1978) 553
12. J.C. Wang, J.B.Bates Solid State lonics, 18-19 (1986) 224
13. R.D. Armstrong, M.F. Bell, A.A.MetcaJfe 'Electrochemistry',Chernical Soc. Specialist Periodical Reports, 7 (19801 157
14. Bernard A.Boukarnp 'Equivalent Circuit', University of Twente, (1 9891, Reports No. U 128/88 / CT 112/89
15. L.L.Hench, J.K. West 'Principles of Electronic Ceramics',
John Wiley 8a Sons, lnc. Singapore. (1990)
16. K.S.Cole, R.H.Cole J. Chem. Phys., 19, (1951), 1484
17. D.W.Davidson, R.H.Cole J . Chem. Phys., 19, (1951),1484
18. S. Hawiliak, S.Negami Polymer 8, (1%7), 16 1
19. P.B. Macedo, C.T. Moynihan, R.Bosz Phis. Chern. Glasses, 13, (19721, 17 1
20. V. Provenzano, L.P.Boesch, V. Voltem, C.T.Mopihan, P.B.Macedo J. Am. Cream. Soc. 55, (19721,492
21. C.T. Moynihan, L.P. Boesch, N.L.Lalxrge f'hys. Chem. Glasses. 14, b (19731, 122
22. A.K.Joncher Phys.Thin Films, 11 (1980) 231
24. M.D.1ngra.m Phys. Chern. glasses, 18 (1987],2 15
25. A.K.Joncher Nature, 267 (1977) 673
20. A.K.Joncher Dielectric Relaxation in Solids', Chesla dielectric press, London, (1983)
27. R.J.Granl, M.D.lngram, L.D.S Turner and C.A Vincent J . Phy. Chem., 82 (1978) 2838
28. C. Chiodelli and A.Mag~stris Solid State lonics, 18 & 19 (1986) 356
29. A k h Doi Solid State Ionics, 40 & 4 1 (1990) 262
30. Bernard A. Boukamp Solid state ionics, 18 & 19 (1986) 136
3 1. Bernard A. Boukamp, J.R. Macdonald Solid state ionics, 74 ( 1995) 165
32. H. Jain, 0.Kanert 'Proceedings of the XI1 1nt.Coni. on Defects in Insulating Materials'. (FA.) 0. Kanert, & J.M.Spaeth vol. 1 (19931 World Scientific Co., 274
33. H. Jain, C.H.Hsieh .J. Non-Cryst. Solids, 172.174 (1994) 1408
34. H.Jain, J.N.Mundy Sohd State Ionics, 91 (1987) 3 15
35. A.H.Verhoef, H.W.den Hartog Solid State lonics, 68 (1994) 305
XI, K.J. Rao, N. Haskaran, P.A. Ramakrishnan, B.G. R a ~ i , A. Karthikeyan Chem. Mater., 10, (1998), 3109
37. S.R. Elliott, A.P. Owens Philos. Magn. 60,6 (1989) 777
38. C.Liu, C.A. Angell J . Non-Cryst. Solids, 83 (1986) 162
39. D.L. Sidebottom, P.F.Green, R.K.Bmw J. Non-Cryst. Solids, 183 (1995) 15 1
40. K.L.Nagi, J.N.Mundy, H. Jain, 0. lianert, G.Balzer Jollenkck Ph.ys. Rev., B 39 (1984) 6169
4 1. S.W. Martin, C.A.Angell J . Non-Cryst. Solids, 83 (1986) 185
42. F.S. Howell, R.A. Bose, P.B. Macedo, C.T.Moynhan J.Phys.Chern., 78 (1974) 639
102
43. K. L. Ngai, R. W. Rendell, H. Jain Phys. Rev., B 30 (1984) 2133
44. K. L. Ngai, R. W. Rendell Handbook of conducting polymer vol. 2(ed). T.A. Skotheirn, Dekker. New York, (1986) 967
45. S.W. Martin, C.A.Angell J. Non-Cryst. Solids, 83 (1986) 185
46. D.P. Almond, A. K. West Solid State lonics, 11 (1983) 57
47. D.P. Almond, A. R. West Solid State Ionics, 9 & 10 (1983) 277
48. J.C. D ~ r e Phys. Lett. 108A no. 9 (1985) 457