HAL Id: jpa-00247809
https://hal.archives-ouvertes.fr/jpa-00247809
Submitted on 1 Jan 1993
HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
Surface-polarization effect in the alignment of nematic liquid crystals
G. Barbero, A. Chuvyrov, G. Kaniadakis, E. Miraldi, M. Rastello
To cite this version:
G. Barbero, A. Chuvyrov, G. Kaniadakis, E. Miraldi, M. Rastello. Surface-polarization effect in the alignment of nematic liquid crystals. Journal de Physique II, EDP Sciences, 1993, 3 (1), pp.165-173.
�10.1051/jp2:1993118�. �jpa-00247809�
Classification
Physics
Abstracts 6 1.30Surface-polarization effect in the alignment of nematic liquid crystals
G. Barbero
('),
A. N.Chuvyrov
(Z), G. Kaniadakis(I),
E.Miraldi(I)
andM. L. Rastello
(3)
(J) Dip.
Fisica, Politecnico, C. Ducadegli
Abruzz124, l0l29Torino,Italy
(2) Bashkir StateUniversity,
Ufa, Russia(3) 1-E-N-G- Ferraris, Str. Delle Cacce91, 10135 Torino, Italy
(Received 3 March 1992, revised 7 August 1992, accepted 28 September 1992)
Abstract. The
change
in the orientation of nematicliquid crystals
due to the increase of thesample
thickness isexperimentally analysed.
The data show that thephenomenon
is a critical one,typical
of a second orderphase
transition. Thechange
in the surfacepolarization
inducedby
aselective
ion-adsorption
on the twoboundary
solid substrates isproposed
as the cause of the observed facts and discussed. The considered model is correct forangles
of tilt small with respectto the initial homeotropic
alignment,
and the theoretical curves fit both the interferometric and thevoltage-threshold
data. A few comments on the obtained results arereported.
1. Introduction.
It is well known that the mean molecular orientation of a nematic
liquid crystal,
the so-called« director » n,
depends
on both the intrinsic symmetry of thematerial,
and the structuralsymmetry
of the substrate surface.The director
profile imposed by
a solid substrate has beenanalysed by
many groups I bothexperimentally [2]
andtheoretically [3],
in order to find a connection between the observed orientation and thephysico-chemical properties
of surfaces andliquid crystal.
A
complete understanding
of theordering properties
of asurface,
due to the chemical andphysical interactions,
is far frombeing achieved,
butthey
can be described in terms of an«
anchoring
» energy.From a
phenomenological point
ofview,
theanchoring
energy is defined as a function of the director orientation close to the surfaces its minimum coincides with the mean molecular direction in the absence of bulk torques. This energydepends
on both the interaction between theliquid crystal
andsubstrates,
and the interaction among the nematic molecules themselves.In the absence of
long-range forces,
the surface energy is a localproperty
in the sense that it does notdepend
on thesample
thickness. On the contrary, it doesdepend
on the thickness when there arelong-range forces,
like electrostaticforces,
due toion-adsorption
on thesurfaces,
or viscousforces,
due to thevelocity
field in the nematic volume[4].
This energy can be
investigated by controlling
theboundary conditions,
e. g.by covering
the cell surfaces with suitableconducting
and/orinsulating
materials andby rubbing them,
as for the mostwidely
useddisplays.
166 JOURNAL DE PHYSIQUE II N°
In this paper, the average director
profile
of different nematiccells,
filledby capillarity,
has been determinedby measuring,
with standard interferometrictechniques,
theirbirefringence
for different discrete thicknesses. In this manner,
varying
thesample thickness,
one can vary its base and lateralsurfaces,
andindirectly
the concentration of theselectively
adsorbedions,
thesample
volumebeing
constant.In the
following
the averagebirefringence
will be defined asusually
:t12 n
(An )
= An
df
=
2Rt (sin~
4l)
,
(1)
-t12
~
C
~~~~/~~~~~~~~~~~~ ~
_ ,
fl '
/ t9'
,1' Wit)
°j
/
,,11,,llllll/ till/
(llllll
Illa)
llzllllllllllll (jllllll/ j
( $ jj
j
~lt)
Ii
,j
(
,
rll/ / II
llllll11(/
II film
_y
z b)
t
«lllllljll/,llllfllllll/, (
f
~/
l,/
,i ,~ ~~~~
~ 4~_?
Ill II / /, / II Ill Ill IN llllllll Ill Z
C)
Fig.
I. The co-ordinate system used in text and the schematic directorprofile
: a) in presence of surfacepolarization
b) inducedby
avelocity gradient
c) in the case of a Fr£edericks transition.where R
= I
(n~/n~)~,
with n~n~ theordinary
andextraordinary
refractionindices,
and(sin~ 4l)
is the average of thesquared
sinus of theangle
between theoptical
axis and thesurface
normal,
and ( the coordinate normal to theboundary surfaces, posed
at (=
± t/2 as
reported
infigure
I.The
figure
also shows thepeculiar
differences of the directorprofile
in three conditions,Figure
la refers to the case whenonly
surface forces arepresent,
so that thealignment
isnearly uniformly
tilted from one surface to the other and 4l(()
= const.
Figure
16 shows the directorprofile
due to a kind of « memory » of thefilling
process : nearthe
boundary surfaces,
at ( = ±t/2,
thehydrodynamic
forces and theanisotropy
of theliquid crystal viscosity impose
that4l(()
reaches the maximum absolute value. Theprofile
isantisymmetrical,
that is 4l((
=
4l
(-
(),
In
figure
lc theliquid crystal undergoes
the so-called Fredericks transition in the presence ofan extemal field:
4l(()
reaches the maximum value in the middle of the cell at(
=
0. The
profile
now issymmetrical.
The three cases can be
distinguished experimentally by
interferometrictechniques
and the measurements discussed in this paper refer to the first case shown.Only
the Frdedericks transition with weakanchoring
conditions(in
the sense that 4l(±
t/2 can vary under the torqueimposed by
the extemalfield)
cangive
a result similar to theuniformly
distortedalignment
due to theonly change
in the surface energy.Indeed for this reason the same
phenomena
werediscussed,
some time ago, as« Frdedericks » effects due to
spontaneous polarization [5],
but a critical look at the accuracy of the formerexperimental set-up (mainly
the surface treatments and thesamples preparation)
induced us to
repeat
theexperiments.
With a new
experimental apparatus, designed
to match the necessarystability conditions,
and withnearly
stronganchoring
conditionsimposed by
theboundary surfaces,
asexposed
in the nextparagraph,
thisambiguity
is resolved.In section
3,
a model of surfacesanchoring
energy thatdepends
on theselectively
adsorbed ions isproposed.
Thecorresponding
surfacepolarization
is proper toexplain
the observedabrupt change
in the directoralignment,
inducedby
thesubstrates,
when a criticaI thickness of thesample
is reached. The modelpredictions
arecompared
with measurements of both meandirector
profile
and the electric field Frederickstransition,
as a function of the cell thickness. In the lastparagraph
some comments arereported
on the numerical results.2. The
experimental
results.The nematic
liquid crystal
used wasmethoxy-benzylidene-butyl-aniline (MBBA),
both pureand mixed
(3
:1)
withethyl-hydroxy-benzylidene-butylaniline (EBBA).
Two cells were
prepared
with unmixed MBBA and thecorresponding
sets ofexperimental
data
(points
and squares inFig. 2a)
refer to nematicsamples kept
betweenglass surfaces, optically
flat within 100 nm, cleaned in sulfo-cromic acid and covered withCr~,
30 nm thick.The two sets of data have been obtained in different
times,
with differentaccuracies,
but with the sameexperimental apparatus
: thecomplete matching
between the data confirms that the cell thickness is agood
control parameter for thisphenomenon,
The last set of
experimental
data(Fig, 2b)
refers to a third nematicsample
filled with mixedMBBA and EBBA and
kept
between fused quartzplates
covered withln~O
andSnO~,
The chosen surface treatments
give
in any case anhomeotropic alignment,
I-e- thestarting
director
profile
is normal to thewalls,
withnearly
stronganchoring conditions,
and in thatposition
it would continue to be while the cell thickness ischanging,
unless other causes, from inside or outside thespecimen,
are able to support a tilted directoralignment.
168 JOURNAL DE PHYSIQUE II N°
. , m m m mm . mm m
, m " "
0
30 ~
jp~j
60a)
0.12
m
~ W
'$ W
~e
C
~
W
m
0
50
jp~j
loob)
Fig.
2.Average
sinus of the tiltangle (sin~
~P)
~'~ versus the thickness t forsamples
filled with pure MBBA (a), or a nematic mixture MBBA + EBBA (b). Points and squares in al refer to two sets ofmeasurements
performed
in different times but with the sameexperimental facility.
The theoreticalcurves are drawn with the value of the parameters discussed in text.
The parameter under control in every run has been the cell thickness. This can be varied
mechanically
with the aid of micrometer screws, and agreat
care was delivered in thedesign
ofthe mechanical device
appropriate
tomodify
thesample
dimension withoutmodifying
theparallelism
of theboundary
surfaces.To the same purposes the metallic materials used for
supporting
theliquid crystal cells,
and handled tochange
theirthickness,
were chosen with the smallest thermalexpansion
coefficient andkept
in a thermostatic chamber.The overall accuracy of these micrometric measurements has been of the order of
2 ~Lm for what concems the absolute
length determination,
but somewhat better if one looksat the remake of the
experiment,
aspreviously reported.
The
general properties
that the collected data show are thefollowing
ones :I)
the director orientation is normal to theboundary
surface if thesample
thickness is lower than a criticalvalue,
I.e. if t w t~;it)
for t~ t~, the director
profile
is a tilted one, the averagebirefringence (An )
# 0 anddepends
on the thickness ;iii)
for t~ co,
(An) approachs
a constant value thatdepends
on theliquid crystal
used.One can summarize in this way the three above-mentioned statements : the trend of
(sin~ 4l)
~'~, 4lbeing
the surface tiltangle,
istypical
of a second-orderphase
transition near t= t~ and shows a limit for t
~ co.
For what concems
point
I the two sets of datareferring
to unmixed MBBAgive
the samecritical
thickness,
as can be seen infigure 2a,
with a numerical value t~ m 32 ~Lm.On the contrary the critical value measured with the nematic mixture was t~ m 60 ~Lm, as can be seen in
figure 2b, showing
that also the critical thicknessdepends strongly
on theliquid crystal properties.
3. The
physical
model.As
already
mentioned in the introduction, thechange
in the directoralignment
with theliquid crystal
cell thickness could beinterpreted,
inprinciple,
in two different ways.According
to the firstpoint
ofview,
the flow of theliquid crystal during
thefilling
processmodifies the surface
anchoring
energy, andimposes
an easydirection, parallel
to the flow direction.Only
a few papers consider the influence of the flow on the nematic orientation. Some time ago, Wahl and Fischer[6]
observed that the nematic average orientationdrastically changed by
a shear
flow,
with velocities even sonegligible
as10~~ mm/day.
Note that this mechanism
implies
the surface tiltangle 4l(f)= 4l(- (),
I-e- anantisymmetric
directorprofile. Furthermore,
itimplies
also aunique flow-direction,
whereas in ourexperimental
set-up it isonly
a noiseeffect,
without any well defined direction : theflow-effects,
if any, average out to zero.Looking
for a different way toanalyse
theexperimental data, namely
the observed re-orientational
transition,
apossible explanation
could be the selective ionsadsorption
phenomenon recently
discussed[7].
According
to thispoint
of view the effectiveanchoring
energydepends
on the adsorbed ions.If the dielectric
anisotropy
of theliquid crystal
isnegative,
and the easy axis due to the surfaces treatment ishomeotropic,
ascertainly they
are in our case, theadsorption phenomena give
adestabilizing
effect on the surface orientation.In this framework let us suppose that the surface
anchoring
energy isrepresented by
:f~
= iii
cos~
4l+ y
cos~
4l,
2 4
(2)
where
cos~ (4l
is connected with the surfacepolarization. Actually
the energy term associatedto the order-electric
polarization [8]
isproportional
to(rj cos~
4l+r~)~,
where rj andr~ are the order-electrical coefficients.
Straightforward
calculations show that a termproportional
tocos~
4l should appear in the surface energy, asreported
inequation (2),
where ik is the effectiveanchoring
energy,taking
into account the « ions » effect.As it is
given by
reference[9]
:iii=w- ~~~~ ~
.~~, (3)
2
e~
where w is the
anchoring
energystrength,
A is theDebye screening length,
e is the average dielectric constant, e~ is the dielectricanisotropy
andfinally
~r= ~r
(t)
is the surfacecharge
density
due to the selective ionsadsorption.
~rdepends
on thethickness,
as discussed in170 JOURNAL DE PHYSIQUE II N°
reference
[9],
and the second term inequation (2)
can beguessed
as a termcorrecting
the usualRapini-Papoular' expression
and connected to the order-electricpolarization [8].
The nematic orientation in the cell can be obtained
minimizing
the total energy of thesample,
defined as the sum of the bulk and surface contributions. The bulk term is due to the elastic deformation of thenematic,
and it isproportional
to the square of thegradient
of the average molecular orientation. In thehypothesis
of a uniform orientation of the nematic in thebulk,
this term vanishes and the orientation near the surfaces is deducedby minimizing only
the surface contribution
f~. Simple
calculationsgive
anhomeotropic profile
stable ifik ~ y, that is for :
«
(t)
~ «(tc)
=l~ )~j
~(4)
~a Y
Equation (4)
defines the critical thickness t~. for t ~ t~, that is if~r(t)
~~r(t~),
the stable nematic orientation is characterizedby
a tiltangle given by
:sin~
4l= a
i~r~(t> ~r~(t~)1
,
(5)
where
e~j
A"
(2 ye~l'
~~~Of course the tilted orientation is stable
only
if iii~
0,
thatimplies
~r ~ ~r*
=
fi, (7)
~a(
otherwise the stable orientation is the
planar
one, with 4l=
ar/2,
not observed in our case.Note that
according
to this mechanism thedependence
of(An )
versus t is due to achange
of the surface energy. Moreprecisely
thephenomenological
parameterscharacterizing f~ depend
on t.
Consequently
also the casy direction deduced withdf~/d4l
=0, depends
on t.Hence the
equilibrium
conditionstating
that the total torque on the surface has to be zero,gives
a surface tiltangle
thatdepends
on the thickness.From
equation (5)
one can see, for t 5S t~.sin4lmp /~, (8)
where
» =
la ~i~ ,~)"~ (9)
Equation (8)
shows that thehomeotropic
~ distorted transition is of the second
order,
asexperimentally
observed.According
to the modelpresented
in reference[9]
~r(t
isgiven by
:"~~~
"~2
~~/r
'
~~~~
with r =
~r~/p~,
where p~ is the ions concentration in thebulk,
~r~ is theasymptotic
value of the surfacecharge density,
and both concentrations refer to an infinitesample.
By substituting equation (10)
into(4)
one obtains :sin~
4l=
awl ~~~~~~
~
~~~~~~
~
l. (ll)
(
+ t/r) (
+t~/r)
It is
possible
to estimate the order ofmagnitude
of the parametersappearing
in the model.For what concems the surface parameters
[8, 10]
one haswm10~~erg/cm~
and ym
10~ ~ =
10~~ erg/cm~
for the electrical[9,
it ones one has :e~(
m0.5,
e m5,
A m 0. I=
I ~Lm, ~r~ m 10~ ~
C/cm~,
p~m 10~ ~
= 10~ ~
C/cm~.
Consequently
: r=
~r~/p~
m10 ÷10~
~Lm,a =
e~(
Al(2 ye~)
m
10~
=
10~ (cm~/C)~, awl
m10~
=
10~.
By using equations (4)
and(10)
one can calculate the critical thickness t~.tc
12 e~(w
y)
"~
t~ + 2
r e~ y
From this
equation
andusing
the values of thephysical
parameters listedabove,
one has t~ m(IN)
r, that is of the correct order ofmagnitude.
In
figures
2a and b two theoretical curves,computed
withequation (11),
and thecorresponding
sets ofexperimental data, (sin~ 4l)"~
versus t, arereported.
It is
straightforward
to see that the agreement of the theoretical curves with theexperimental points
isfairly good only
near the critical thickness. This is due to the very naive modelproposed
for the ionadsorption,
describedby equation (10).
As is wellknown,
thisexpression
is similar to the
Langmuir isotherm,
andconsequently
it can run wellonly
in the limit of small~r values. A more
general expression
can be obtainedonly by taking
into account theelectrostatic
repulsion
between the adsorbedions, neglected
in thisanalysis.
The model used leads in fact to a
polynomial expression
that containsonly
threeparameters, namely
one is the critical thickness and the other two are related to the ion concentrations in the bulk and on the surfaces.By
means of the above-mentionedmodel,
nevertheless, one can deduce the criticalvoltage
U for the Fr£edericks transition as a function of the thickness of the
sample.
To this aim4
0
lo t
jp~j
32Fig.
3. The theoretical curve and theexperimental
data collected for the threshold voltage in the Frdedericks transition U versus thesample
thickness t. Thesamples
used were the same as infigure
2a.172 jOURNAL DE PHYSIQUE II N°
measurements of the threshold
voltage
U for the Frdedericks transition have beenperformed,
and the uniform
homeotropic alignment
of the director forsamples
with t~ t~ was confirmed.
In
samples
filled with pure MBBA U wasfalling
withincreasing
thickness of thecrystal,
tilla
complete vanishing
was reached for t m t~,supporting
the existence of a permanent tilt of thedirector,
as in theoptical
measurements.By minimizing
the total free energy of thesample,
obtainedby summing
up the bulk term(including
the elastic term and the extemal fieldinteraction)
and the surface one, as shown in reference[12],
one obtains that the criticalvoltage depends
on the elastic(k)
and electric(e~) properties
of thenematic,
the effective surface energy(iP)
and thesample
thickness(t).
Itsanalytical expression
is~t= ~tg (i~ ), (12)
" k UC 2 UC
where U~ = ar
kleo
e~ is the criticalvoltage
in the stronganchoring
case, and U is the one in the weakanchoring
case.By using
alsoequations (3)
and(10),
it ispossible
to calculateU/U~
versus t.The
experimental
resultsconceming
the criticalvoltage
and the theoretical curve derived from thephysical
models arereported
infigure
3. The agreement betweenpoints
and curve isobtained with the same values of the
parameters
used above.4. Conclusions.
The influence of the thickness of a
sample
on the average orientation of a nematiccrystal
hasbeen considered. The
experimental
data show that thereorienting phenomenon
connected with the thickness of thesample
is a critical one. Moreprecisely
it isexperimentally proved
that for t ~ t~, where t~ is a criticalthickness,
thesamples
are in ahomeotropic
orientation. For t m t~ the average orientationchanges according
to(t t~)~'~, typical
of a second orderphase
transition.
The
experimental
data areanalyzed by using
asimple model,
in which the surface energydepends strongly
on the adsorbed ions.If one assumes in the
expression
of the surface energy a termproportional
tocos~ 4l,
connected with the order-electricpolarization,
a transition from thehomeotropic
to the tilted directorconfiguration
must beexpected
if one suggests that the selective ionsadsorption
process follows the
Langmuir
law for theperfect
gases.In the same frame of the
Langmuir-adsorption
law the thresholdvoltage
for the Frdedericks transition has beenanalyzed
: the theoreticalpredictions
are ingood
agreement with theexperimental
data for t< t~.
In contrast, the theoretical
(sin~ 4l)~'~
curves and theexperimental
data are ingood
agreement
only
for t 5St~. This fact can beinterpreted
in the sense that theLangmuir
law must be consideredonly
as a very crudeapproximation,
validjust
in thebeginning
of thetransition,
for tm t~. Work is in progress to
improve
the theoretical model.References
[1] JEROME B.,
Rep. Frog.
Phys. 54 (1991) 391.[2] See for instance COGNARD J., Mol. Cryst. Liq.
Cryst. Supp.
1(1987) 1.[3] YOKOYAMA H., Mol.
Cryst. Liq. Cryst.
165 (1988) 265.[4] LESLIE F. M., Arch. Rat. Mech. Anal. 28 (1968) 265.
[5] CHUVYROV A. N., Sov.
Phys. Crystallogr.
25 (1980) 188.[6] WAHL J., FISCHER F., Mol. Cryst. Liq. Cryst. 22 (1973) 359.
[7] YOKOYAMA H., KOBAYASHI S., KAMEI H., J. Appt. Phys. 56 (1984) 2645.
[8] BARBERO G., DURAND G., J. Phys. France 47 (1986) 2129.
[9] BARBERO G., DURAND G., J. Appt.
Phys.
67 (1990) 2678.[10] BLINOV L. M., KABAENKOV A. Yu., SONIN A. A., Liq. Cryst. 5 (1989) 645.
[11] THURSTON R. N., CHENG J., MAYER R. B., BOYD G. D., J.
Appt. Phys.
56 (1984) 263.[12] RAPINI A., PAPOULAR M., J.