Dielectric characterization and ionic conductivity of α-LiIO3 crystals related to the growth conditions

Article abstract of DOI:10.1016/S0038-1098(00)00247-7

The dielectric response of α-LiIO₃ has been studied at room temperature between 20 Hz and 1 MHz with various types of electrodes and compared to the results in the literature. By changing the sample thickness, a relaxation of space charges is clearly identified and the bulk ionic conductivity is deduced from the admittance diagram. Finally a comparison is carried out between chromium doped and undoped crystals obtained from acid and neutral growth solutions.

Full text of DOI:10.1016/S0038-1098(00)00247-7

PERGAMON
Solid State Communications 115 (2000) 619–623
www.elsevier.com/locate/ssc
Dielectric characterization and ionic conductivity of a-LiIO3
crystals related to the growth conditions
a,
a
b
c
a
*
Y. Mugnier , C. Galez , J.M. Crettez , P. Bourson , J. Bouillot
a
´
LAIMAN, Universite de Savoie, 41 Av. de la plaine, BP806, 74016 Annecy cedex, France
b
c
´
LPUB, Universite de Bourgogne, 9 Av. Alain Savary, BP47870, 21078 Dijon cedex, France
´
LMOPS, Universite de Metz, CLOES-SUPELEC, 2 Rue E. Belin, 57078 Metz cedex, France
Received 20 April 2000; accepted 23 May 2000 by P. Burlet
Abstract
The dielectric response of a-LiIO3 has been studied at room temperature between 20 Hz and 1 MHz with various types of
electrodes and compared to the results in the literature. By changing the sample thickness, a relaxation of space charges is clearly
identified and the bulk ionic conductivity is deduced from the admittance diagram. Finally a comparison is carried out between
chromium doped and undoped crystals obtained from acid and neutral growth solutions. ᭧ 2000 Elsevier Science Ltd. All rights
reserved.
Keywords: D. Dielectric response; B. Crystal growth
1. Introduction
using the impedance spectroscopy technique with both
metallic and ideally polarizable insulating electrodes
(i.e. impermeable to electronic and ionic charge
carriers) on crystals grown under different conditions
with or without deliberate impurities. We have already
shown the ability of a-LiIO3 for guiding light when
implanted with protons [1] and the aim of this study
is to characterize the ionic conduction of crystal plates
in order to develop possible waveguides production by
proton exchange and/or ion diffusion methods. There-
fore, the identity of the ionic charge carriers and the
mechanism of direct conduction must be clearly identified.
Lithium iodate in the a form (a-LiIO3) is a well-known
optical material exhibiting a large transparency range and
strong piezoelectric, photorefractive and non-linear effects.
After having been used for second harmonic generation and
frequency mixing, new applications in the fields of solid
state laser technology and optical data storage have been
proposed by incorporating iron-group ions and rare earth
elements in these crystals.
Since optical properties of this non-ferroelectric material
are strongly related to the quasi-one-dimensional ionic
conduction, the response of Z-cut samples under low
frequency electric fields has been widely studied by means
of dielectric investigations, dielectric breakdown, synchrotron
radiation and conventional X-ray diffraction experiments.
Nevertheless there are many discrepancies in the literature:
results depend not only on intrinsic parameters (acidity
and inclusions of the growing solution, doping, etc.) but
also on external ones such as the amplitude of the electric
field, its time of application and the nature of the electrodes.
In this paper the dielectric behaviour is investigated by
2. Experimental details
Large a-LiIO3 single crystals were grown by slow
evaporation of slightly supersaturated aqueous solutions
(the compound a-LiIO3 was first obtained by the reaction
of Li2CO3 and HIO3 in stoichiometric amounts) at a constant
temperature of 40ЊC in neutral (pH ˆ 6) or acid (pH ˆ 2)
environment. Cr-doped crystals were grown by adding
chromium nitrate in known amounts to the solution and
setting the pH at 2. The Cr3ϩ has been found to replace a
Liϩ [2,3] and/or I5ϩ [4] ions and charge compensation
occurs through lithium vacancies.
* Corresponding author. Tel.: ϩ33-4-5009-6513; fax: ϩ33-4-
5009-6649.
E-mail address: mugnier@esia.univ-savoie.fr (Y. Mugnier).
0038-1098/00/$ - see front matter ᭧ 2000 Elsevier Science Ltd. All rights reserved.
PII: S0038-1098(00)00247-7

620
Y. Mugnier et al. / Solid State Communications 115 (2000) 619–623
3. Results
(a)
103
102
101
3.1. Determination of the best suited electrodes
A comparison of the different available methods given by
experimental setup is illustrated by Fig. 1 where the
frequency dependence of the real part of the permittivity
e 0r33 (a), the a.c. conductivity s33 (b), the dissipation factor
tan d33 (c) and the imaginary part of the permittivity e 00r33 (d)
is presented for an acid type plate of thickness 0.35 mm.
The NCM and CM with rigid electrodes give identical
responses whereas those obtained with silver paste clearly
depend on the state of the paint (“wet or dry”) and hide the
previous ones in the whole frequency range investigated. At
very low frequencies, the “wet” silver paint response shows
10-6
10-7
(b)
(c)
10-8
10-9
10-10
10-11
2.0
1.5
1.0
0.5
a
relaxation mechanism with characteristics of the
Maxwell–Wagner effect [7,8] since electrodes are not
completely free from solvent. A second relaxation peak,
still silver paste state dependent, arises at higher frequencies
(d)
103
102
101
100
but is not yet understood. Very high values of e 0r33 and e 00
r33
are observed and could be related to the “ionic nature” and
surface chemical composition of lithium iodate crystals [9],
leading to the occurrence of electrochemical reactions.
Indeed this surface decomposition has been observed by
short circuit currents with deposited Ag, C and In electrodes
and confirmed by X-ray photoelectronic spectroscopy [10].
This apparent dielectric response can be compared with
spectra obtained with silver paste or sputtered gold films
[11,12] and it gives admittance diagrams at room tempera-
ture similar to those given in Ref. [13] where deposited
layers of Ag, Pd or Li were used as contacting electrodes.
Moreover, as many dielectric materials (LiNbO3, etc.)
have identical responses with both metallic pastes and
rigid metal electrodes, it seems important to confirm that
CM and NCM reveal properties of the material itself
and not processes related to the nature of the electrodes.
In this way, insulating plates like glass, teflon, mica and
kapton were placed between the sample and rigid nickel
electrodes because they are chemically inert and present a
frequency independent response. Considering a series
combination of ideal capacitors, the permittivity can be
written:
2
3
4
5
6
Fig. 1. Dielectric response of an acid a-LiIO3 crystal
(e ˆ 0.35 mm) with different electrodes: (—) CM with rigid nickel
electrodes, (W) NCM with a 20 mm air gap and rigid nickel elec-
trodes, (O) CM with wet silver paste, (K) CM with “dry” silver
paste.
All the dielectric experiments reported here were carried
out with Z-cut plates, cut from homogeneous crystals of
good optical quality and polished with 1 mm diamond
paste. The applied electric field is set in the range 10Ϫ2
–
101 V cmϪ1 in order to be in the linear part of the current–
voltage characteristic and to avoid time dependent
decomposition (dry electrolysis), arising above a few
kV cmϪ1 [5,6].
The experimental system, an HP4284A impedance
analyser, allows the simultaneous determination of the real
partofthedielectricpermittivitye 0r33 andthedissipationfactor
tan d for frequencies ranging from 20 Hz to 1 MHz. A 4-
terminal pair measurement method is used as well as
calibrations to eliminate stray admittance and residual
impedance. The edge capacitance is also reduced by using a
guarded electrode for both the Contacting Method (CM) (with
either rigid metal electrodes of brass nickel-plated or thin film
electrodes of metallic pastes) and the Non-Contacting Method
(NCM) (here, for a constant distance between rigid elec-
trodes, measurements are made with and without the
material inserted so that contact capacitive effects due to
the sample surface roughness are reduced).
e
e0r33
ˆ
ꢀ1†
ꢀ
ꢁ
1
2ei
Sei
e0S
Ϫ
Ctot
where e (2ei) is the thickness of the sample (insulating
plates), e0 (ei) the permittivity of the vacuum (insulating
material), S the surface of the guarded electrode and Ctot
the total measured capacity. For an acid type crystal
(e ˆ 1.74 mm) we obtain a similar behaviour for e 0 with
r33
either the CM or the NCM with all ideally polarizable
electrodes (Fig. 2).
Oscillations observed at low frequencies for the NCM are
due to the small detected values inherent to the method

Y. Mugnier et al. / Solid State Communications 115 (2000) 619–623
621
2
2
1
2
3
4
5
6
Fig. 2. e 0 dispersion of an acid a-LiIO3 crystal (e ˆ 1.74 mm)
r33
with different electrodes: (—) CM with rigid nickel electrodes,
(W) NCM with a 60 mm air gap and rigid nickel electrodes, (B)
NCM with two 70 mm insulating glass plates and rigid nickel
electrodes, (O) CM with wet silver paste.
≅
3
4
5
6
which are in the limit of the impedance analyser measure-
ment range.
Fig. 3. Dielectric response of an acid a-LiIO3 crystal (e ˆ 1.52 mm):
(W) NCM with a 30 mm air gap and rigid nickel electrodes.
3.2. Characterization of the dielectric response
Afterwards, several equivalent representations exist for
analysis:
The continuous decrease of e 0r33 and the appearance of a
wide loss peak on e 00 (Fig. 3) correspond to a relaxation
r33
mechanism. From an electrical point of view, the response
can be modelled by a series RS–CS circuit in parallel with a
• Cole–Cole plots are well suited for true dipolar response
since eS and e directly appear as the extremities of a
∞
frequency-independent capacity
C
∞
representing any
semicircle.
physical process with a faster response. The resistance RS
is assigned to the macroscopic displacement of electric
charges in the bulk and CS to the space charge accumulation
so that the complex admittance Y(v) becomes:
• The impedance diagram allows, especially in the field of
electrochemistry when interphases of different conduc-
tivities are permeable to ions, to separate processes
related to the bulk from those of the interfaces.
• Here, for completely blocking electrodes, the admittance
diagram is used (Fig. 4) for the pH ˆ 2 plate studied
ivCS
Yꢀv† ˆ ivC∞ ϩ
above. A vertical straight line corresponding to C and
∞
a semicircle associated with the relaxation mechanism
are observed. The two parts are well described if the
ratio es=e∞ is large [7] and according to Eq. (2) the
high frequency limit of the real part of Y(v) tends
towards the direct conduction sc.
1 ϩ ivRSCS
ivCS ϩ v2R2SCS2=RS
ˆ ivC∞ ϩ
with
(2)
1 ϩ v2R2SCS2
As often observed in experimental responses the circular
arc is tilted by an angle a from the real axis and the
explanation of this inclination is founded on the assumption
of a weak dispersion of the local value of conductivity
around the average value sc [16,p. 163]. So the experimen-
tal data are fitted with the circle of radius A centred in
(A cos(a), ϪA sin(a)) from which we can deduce sc ˆ
2A cosꢀa† ˆ ꢀ0:80 ^ 0:02† 10Ϫ5 VϪ1 cmϪ1 (Fig. 4).
e0eSS
e0e∞S
e
1 e
CS ˆ
; C
ˆ
; RS ˆ
sc
∞
e
S
where eS (e ) is the static (infinite) permittivity and sc the
∞
direct current conductivity.
The relaxation peak in a log–log scale (Fig. 3) is then
fitted by the commonly used empirical law [7,p. 192]. For
an acid a-LiIO3 crystal of thickness 1.52 mm, linear fits
below and above the relaxation frequency give, respectively,
m Ϸ 0:65 and n Ϫ 1 Ϸ Ϫ0:8: Since the absolute slopes
m and ꢀn Ϫ 1† are lower than 1 this relaxation of
ionic conductivity deviates from an ideal Debye response
[14,15].
Moreover the disturbances observed around 300 kHz are
due to the piezoelectricity of a-LiIO3. Being aware that the
number of measurements per decade ( Ϸ 20 pts) is low,
these disturbances appear in Fig. 5 below 200 kHz for the
sample of thickness 6.26 mm. They are not observed for the
plate of thickness 0.35 mm, being probably above 1 MHz,
because the thicker the sample is the lower the resonance

622
Y. Mugnier et al. / Solid State Communications 115 (2000) 619–623
electrochemical double layer structure, the excess of
ω ε0ε∞
charges induces a variation of the average electrostatic
potential. The value of this potential which finally limits
the accumulation of charges must not depend on the thickness
of the sample.
0.50
0.25
So, the apparition of space charges was checked by
measuring the capacity of samples with thickness varying
between 0.35 and 6.26 mm, i.e. in a ratio 18. As the
measured capacity remains constant for all plates, the
apparent static permittivity increases with the thickness as
it appears in Fig. 5 below the relaxation. For e 00r33, we
observe a shift of the relaxation peak towards the high
frequencies when the thickness decreases. This remains
coherent with the displacement of a same charge carrier of
constant mobility for the acid type crystals.
ω
σC
0.00
0.25
0.50
0.75
α
(A.cos(α),-A.sin(α))
- 0.25
On the admittance diagram represented in Fig. 6, the
semicircle for the sample of thickness 6.26 mm (i.e. with
the greatest apparent static permittivity) is more “traversed
with increasing frequency before the vertical spur makes a
significant contribution” [7,p. 72]. But for all the acid type
plates, the direct conductivity deduced from the admittance
diagram when the frequency tends towards the infinite is
-3
-1
σ33
Ω -1
Fig. 4. Admittance diagram of an acid a-LiIO3 crystal
(e ˆ 1.52 mm): (W) CM with rigid nickel electrodes, (—) (A cos (a),
ϪA sin(a)) circle of radius A.
constant around 10Ϫ5
V
Ϫ1 cmϪ1 and this also confirms a
relaxation of ionic conductivity due to the formation of
space charges.
frequency is. Moreover we have checked, by using the finite
element modelling software ANSYS, that the frequency
range around 300 kHz corresponds to the principal modes
of vibrations for a parallelepiped of 1–2 mm in thickness.
3.4. Influence of the growth conditions
Being assured to observe the dielectric properties of the
material itself, a first comparison is made for crystals
obtained with different growth conditions. The responses
of a neutral type plate (e ˆ 1.03 mm) and of a chromium
doped sample (0.05 mol%, in a growing solution pH ˆ 2,
3.3. Influence of sample thickness
For ionic conductors the displacement of electric charge
carriers and their accumulation in the vicinity of the surface
create a zone of space charges. In this zone, also called
3
2
2
1
0
-1
2
3
4
5
6
Fig. 5. Dielectric response of acid a-LiIO3 crystals of various thicknesses: CM with rigid nickel electrodes. (W) e ˆ 6.26 mm, (A) e ˆ 1.74 mm,
(X) e ˆ 0.35 mm.

Y. Mugnier et al. / Solid State Communications 115 (2000) 619–623
623
Identification, concentration and activation energies of
the charge carriers are now investigated by studying thermal
properties of the dielectric response. A vacancy mechanism
seems to confirm the room temperature relaxation of space
charges.
ω ε0ε∞
In addition, we have shown in the present study that,
depending on the growth conditions, ionic conductivity
can be already reduced at room temperature. Now, the
strong increase of the ultraviolet photorefractive effect
below 200 K [17,18] has been connected with the freezing
of the ionic conduction. If they remain photoconductive,
neutral a-LiIO3 crystals could be of great interest for ultra-
violet photorefraction.
ω
Acknowledgements
The authors are grateful to M. Maglione (Dijon) for
fruitful discussions.
-3
-1
σ33
Ω -1
Fig. 6. Admittance diagram for different a-LiIO3 crystals: CM with
rigid nickel electrodes. (W) e ˆ 6.26 mm, pH ˆ 2, (A) e ˆ 1.74 mm,
pH ˆ 2, (B) e ˆ 2.83 mm, pH ˆ 2, Cr3ϩ doped (0.05% mol.), (X)
e ˆ 1.03 mm, pH ˆ 6.
References
[1] C. Rosso, P. Moretti, J. Mugnier, D. Barbier, Y. Teisseyre, J.
Bouillot, Phys. Status Solidi (a) 139 (1993) K137.
[2] M. Krupski, A.A. Mirzakhanyan, W. Hilczer, E.G. Sharoyan,
Phys. Status Solidi (b) 133 (1986) 355.
e ˆ 2.83 mm) are compared on the admittance diagram
(Fig. 6).
[3] W.C. Zheng, Phys. Status Solidi (b) 153 (1989) K103.
[4] K.T. Han, J. Kim, Solid State Commun. 106 (6) (1997) 467.
[5] L.I. Isaenko, I.F. Kanaev, Malinovskii, V.I. Tyurikov, Sov.
Phys. Solid State 30 (2) (1988) 199.
[6] O.G. Volkh, I.A. Velichko, L.A.A. Laz’ko, Sov. Phys.
Crystallogr. 20 (2) (1975) 263.
[7] A.K. Jonscher, Dielectric Relaxation in Solids, Chelsea
Dielectric Press, London, 1983.
[8] N. Bogris, J. Grammatikakis, A.N. Papathanassiou, Phys. Rev.
B 58 (16) (1998) 10 319.
[9] V.V. Azarov, L.V. Atrshchenko, V.T. Gritsyna, T.I. Korneeva,
V.F. Pershin, Yu.V. Sidorenko, N.V. Khodeeva, Sov. Phys.
Tech. Phys. 29 (12) (1984) 1399.
[10] A.A. Blistanov, N.S. Koslova, V.V. Geras’kin, Ferroelectrics
198 (1997) 61.
If the static permittivity of doped crystals behaves,
on dielectric spectra (not given here) like that of the
acid type plates, the direct conductivity sc ˆ (1.70 ^
0.02) × 10Ϫ5
V
Ϫ1 cmϪ1 is practically doubled. Thus, the
vacancies incorporated during the growth seem to present
a determinant contribution to the conduction mechanism of
the samples pH ˆ 2.
More striking is the neutral a-LiIO3 crystal behaviour
which also presents a relaxation of ionic conductivity
(inset of Fig. 6) but the absorption peak is shifted by a factor
approximately 30 towards the low frequencies i.e. appears
around 1 kHz. On the admittance diagram the direct ionic
conductivity varies with the same ratio and is measured at
(0.019 ^ 0.001) × 10Ϫ5
V .
Ϫ1 cmϪ1
[11] L.G. Jacobsohn, P. Lunkenheimer, F. Laeri, U. Vietze, A.
Loidl, Phys. Status Solidi (b) 198 (1996) 871.
[12] Y. Mugnier, C. Galez, J.L. Jorda, P. Bourson, J.M. Crettez, A.
Righi, R.L. Moreira, Radiat. Eff. Defects Solids 150 (1999)
333.
4. Conclusions
After having demonstrated dubious dielectric behaviour
when some metallic pastes are applied on lithium iodate
Z-cut plates, we have used impedance spectroscopy to perform
fast non-destructive characterizations of crystals grown in
different environments. With the use of chemically inert,
ideally polarizable electrodes, the low frequency dielectric
response observed is to be dominated by the formation of a
zone of space charges. The true value of the direct conduc-
tivity, i.e. corresponding to macroscopic displacements of
ionic charges in phase with the applied electric field, are
then deduced from the admittance diagrams.
[13] B.V. Shchepetil’nikov, A.I. Baranov, L.A. Shuvalov, Sov.
Phys. Solid State 29 (3) (1987) 450.
[14] A.A. Abramovich, A.Sh. Akramov, A.E. Aliev, L.N. Fershtat,
Sov. Phys. Solid State 29 (8) (1987) 1426.
[15] A.E. Aliev, A.Sh. Akramov, L.N. Fershtat, P.K. Khabibullaev,
Phys. Status Solidi (a) 108 (1988) 189.
[16] C. Deportes, M. Duclot, P. Fabry, J. Fouletier et al, Electro-
chimie des solides, Presses Universitaires de Grenoble, 1994.
[17] J. Xu, X. Yue, R.A. Rupp, Phys. Rev. B 54 (23) (1996)
16 618.
[18] J. Xu, H. Kabelka, R.A. Rupp, F. Laeri, U. Vietze, Phys. Rev.
B 57 (16) (1998) 9581.