Journal o/Thermal Analysis, Val. 51 (1998) 737-746
LIQUID CRYSTAL-POLYMER COMPOSITE MATERIALS A thennophysical and electro-optical study U. Maschke'*, F RousseP, J.-M. Buisini and X Coqueret' 'Laboratoire de Chimie Macromoleculaire, CNRS (URA N° 351), Bätiment C6, Universite des Sciences et Technologies de Lilie, F-59655 Villeneuve d'Ascq Cedex 2Laboratoire de Dynamique et Structure des Materiaux Moleculaires, Equipe de Thermophysique de la Matiere Condensee, CNRS (URA N° 80 I), Universite du Littoral MREID, F-59140 Dunkerque, France
Abstract Polymer dispersed liquid crystal (PDLC) films can be switched electrically from a lightscattering off-state to a highly transparent on-state. Thin films were prepared via a polymerization-induced phase separation process, using electron beam radiation. The liquid crystal (LC)/polymer materials were obtained from blends of an eutectic nematic mixture E7 and a polyester acrylate-based polymer precursor. The optical and electro-optical properties of the PDLC films obtained depend strongly on the LC concentration. The LC solubility limit in the polymer matrix and the fractional amount ofLC contained in the drop lets were deterrnined by means of calorimetric measurements. Keywords: liquid crystals, phase separation, polymer dispersed liquid crystals (PDLCs), polymers, solubility
Introduction Polymer dispersed liquid crystal (POLC) films commonly consist of micronsized droplets of low molecular mass liquid crystal (LC) dispersed in asolid polymer matrix [1-2]. These new materials are of considerable interest for display applications and light control devices such as optical shutters. Polymerization-induced phase separation (PIPS) initiated by electron be am (EB) radiation was used to obtain well-defined POLC films [3-5]. The use ofEB processing offers various advantages: precise control of the curing conditions, a high cure rate without thermal activation, a possible high conversion of monomers, and initiation of the polymerization at a desired temperature. Compared with the PIPs process initiated by ultraviolet light, EB cu ring has the unique advantage of not requiring the presence of a photoinitiator, which may be detrimental to the POLC film performances and to long-term ageing. *
Author for correspondence; e-mail:
[email protected]
1418-2874/98/ $ 5.00 © 1998 Akademiai Kiadä. Budapest
Akademiai Kiadä, Budapest Kluwer Academic Publishers. Dordrecht
738
MASCHKE ct al.: LIQUID CRy,sTAL-POLYMER COMPOSITES
The principle of operation is based on the electric field-controlled transmission properties of the composite materials sandwiched between two transparent conducting electrodes. Light is efficiently scattered in the initial off-state, depending on the refractive indices of both the LC and the polymer matrix. The nematic LCs used in the experiments exhibit positive dielectric anisotropy at low frequencies and therefore align with their optical axis parallel to an externally applied electrical field. In the on-state, light incident normal to the film surface essentially probes the ordinary refractiye index of the LCs. A high degree of transparency in the on-state can be achieved only if the refractive index of the polymer matrix approximately matches the ordinary refractive index. Upon remoyal of the yoltage, the drop lets return to their original random orientation that induces light scattering. In general, the PIPS process does not result in the complete separation of LC from the polymer matrix. Small amounts of LC remain dissolved in the cured polymer matrix, thereby acting as a plasticizer and reducing the glass transition temperature of the matrix. As a consequence, the refractive index of the polymer matrix will change, perturbing the refractive index matching and hence influencing the electro-optical properties. The degree of conversion ofthe initial reactive mixture has been found to be an important parameter governing the phase separation [6, 7]. Highly cured composite systems exhibit a considerably reduced solubility of LC in the matrix. The enhanced extent of phase separation will lead to maximization of the fraction of LC included in the drop lets and to an increase in the electro-optical performance. The electro-optical characteristics also depend on the size of the drop lets and their volumetric number density. The present work was focussed on the influence ofthe ratio ofthe LC content to the matrix precursors, the composition of the matrix precursors being kept constant for each of the sampies. The scattering efficiency of the PDLC films depending on the LC content was examined. The transmission properties of selected films of approximately identical thickness were investigated as a function of driving AC voltage and sampie composition. The results obtained from optical and electro-optical studies were correlated with thermodynamic properties. Smith recently introduced a calorimetric model based on differential scanning calorimetry (DSC) investigations to determine the LC solubility limit in the polymer matrix [8, 9]. The fractional amount ofLC contained in the droplets was calculated by using a simple model.
Experimental Materials The LC mixture E7 (Merck Ltd, GB) was used during this work; it exhibits a nematic phase at room temperature, which easily forms a glassy nematic upon 1. Thernwl Anal., 51, 1998
MASCHKE ct al.: LIQUID CRYSTAL-POLYMER COMPOSITES
739
cooling, and crystallization does not readily occur upon rewarming. The glass gradually becomes a fluid nematic at Tg=-62°C with ~Cp=0.48 J (g °Cr l . The nematic - isotropie transition of E7 occurs at T NI=59°C with LVlN1 =4.4 J g -I. E7 is characterized by the following refractive indices at T=20°C: n o=1.5183 and ne=1.7378 (1-.=632.8 nm). The prepolymer chosen consists of an aromatic polyester acrylate composition (Rahn AG, Switzerland) blended with tripropyleneglycoldiacrylate (UCB, Belgium). The refractive index of the prepolymer in the cured state in the absence of E7 is n p = 1.5120 (A=632.8 nm).
Preparation of PDLC films (IOO-x) weight percent (wt%) ofthe prepolymer (x=lO, 20, ... ,90) andx wt% of the liquid crystals were mixed together at room temperature for several hours. The thickness and the uniform application of the reactive initial mixtures on glass plates coated with a thin transparent layer of conducting indium-tin-oxide (ITO) (Balzers, Liechtenstein) was controlled by using a wire-wound rod as a bar-coater of 25 Ilm (Braive, Belgium). SampIes used for DSC measurements could be obtained by uniform application of the reactive mixtures on aluminium sheets as supports, using a bar-coater of 75 Ilm. For each composition, several sampIes were prepared and exposed to the EB radiation to eure the polymerizable mixture.
Electron beam curing The EB generator used to prepare the PDLC sampIes by a PIPS process was an Electrocurtain Model CB 150 (Energy Sciences Inc.) with an operating high voltage of 175 kY. The sampIe supports were placed in a tray, which was passed under an inert atmosphere to the accelerated electron curtain on a conveyor belt. The applied dose of 60 kGy was achieved by using a beam current of 4 mA and a conveyor speed of 0.22 m S-I. These values were not changed during the experiments, in order to apply the same curing conditions each time. The applied dose was delivered uniformly in the depth of the sampIe. No temperature control was performed during the irradiation process. Film thicknesses were measured with a micrometer calliper (Mitutoyo; uncertainty: ±1 Ilm). To minimize errors in the determination of film thickness, 18 different sites on each plate before and after composite film preparation were taken into account.
Electro-optical measurements The electro-optical experiments were performed at room temperature by measuring the transmission of unpolarized HeNe laser light at a wavelength of A=632.8 nm. The glass plates with the PDLC films were placed normal to the laser beam. The distance between the sampIe cell and the detector (silicon photoJ. Thermal Anal., 51, 1998
740
MASCHKE et al.: LIQUID CRYSTAL-POLYMER COMPOSITES
diode) was approximately 30 cm. The collection angle of the transmitted intensity was about ±2°, so that principally forward scattering was detected. The intensity of transmitted light was recorded on a micro-computer, using an interface card (DAS 1600). The transmission values were corrected for the loss of transparency wh ich results from the reversible darkening of the glass plates upon EB irradiation. The scattering properties ofthe cured PDLC films in the initial off-state were measured at different sites on each glass plate, allowing averaging of the obtained transmission values. For electro-optical measurements, the composite film of uniform thickness (l5±1 ~m) was sandwiched by another ITO-coated glass substrate and an extern al electric field was applied across the PDLC film. The output of a frequency generator was amplified and used to drive the shutter device. Starting from the electrical off-state, the applied sinusoidal voltage of frequency 145 Hz was increased continuously up to a desired maximum value Vrnax • Subsequently, it was decreased in the same way. The whole scan up and down ramp was usually performed during 120 s. An additional measuring time of 60 s allowed the relaxation behavior of the transmittance to be followed in the off-state. The same procedure was repeated twice (rarnps 2, 3).
Thermophysical characterization The DSC measurements were carried out on a Seiko DSC 220C equipped with a liquid nitrogen system allowing cooling experiments. Tbe DSC cell was purged with 50 ml min -I of nitrogen. Rates of 10°C min -I (heating) and 30°C min -I (cooling) were used in the temperature range from -120°C to + 100°C. The program consisted first in cooling the sampie, followed by several heating and cooling cyc1es. Data analysis was carried out on the second heating ramp.
Results and discussion Optical properties The influence of the ratio of the LC content to the polymer matrix preCursors on the transmission in the initial off-state (To ) is presented in Fig. 1. In the LC weight percentage range from 0
MASCHKE ct al.: LIQUID CRYSTAL-POLYMER COMPOSITES
741
100 80
~ c
0
"u;
60
I/)
"E
I/)
c
~
40 20
00
10
20
30
40
50
60
70
80
90
100
Liquid Crystal concentration (weight-%)
Fig. 1 Transmission in the initial off-state of EB-cured PDLC films as a function of LC concentration
The limit of miscibility of the starting reactive mixture before curing was reached at 70 wt% ofLC. Blends including more than 70 wt% ofLC did not form homogeneous solutions at room temperature and were not used for electro-optical measurements.
Electro-optical behaviour Figure 2 illustrates the electro-optical response for the first voltage application (ramp 1 up), when the LC content was varied. The transmission values of all sampies including 30 wt% E7 did not change even with the highest voltage (Vmax =120 V) applied to the films. The corresponding data were therefore omitted from Fig. 2. At a LC concentration of 40 wt%, upon application of a voltage ramp with a maximum of 120 V, the transmission values changed by only a few per cent. Starting from a high off-state transmission of 50%, the threshold voltage VIO (the voltage required for 10% of the maximum transmission value) was estimated to be as high as 60 V. Constant transmission values characterizing the on-state (T lOo ) were not obtained. The electro-optical characteristics changed drastically when the LC content was changed from 40 to 70 wt%: increasing the LC concentration in this range considerably increased the off-state scattering (as already shown in Fig. 1). Simultaneously, the threshold and the saturation voltage (V90 , voltage required for 90% ofthe maximum transmission value) were reduced substantially. The electro-optical curves for sampies containing 50 to 70 wt% were characterized by high transmissions in the on-state. Composite films containing 70 wt% ofLC exhibited strong scattering in the off-state, low threshold and saturation voltages (V IO =3.6 V; V90=21 V), and a high contrast ratio (by ca1culating TlOoITo).
1. Thermal Anal", 51, 1998
MASCHKE ct al.: LIQUID CRYSTAL-POLYMER COMPOSITES
742
100 90 80 70
~ ~ c
50
.!!.!
50
0 'üj
E (/) c
~
40
40
-·- ;W: E7 50wt% E7 r o 50wt% E7
*
30 20
•
10 10
20
30
40
50 50 70 80 Applied Voltage (V)
-1
70wt% E7
90
100 110
120
Fig. 2 Effect of the LC concentration on the electro-optical response for the first voltage application (ramp I up, A.=632.8 nm, v=145 Hz)
Thermodynamic properties The phase diagram of the LC/polymer eomposite materials derived from thermal DSC seans is shown in Fig. 3. The glass transitions of the LC (TgLC ) and of the polymer matrix (TgMatrix), and also the nematie - isotropie transition (TNI ) are plotted as a funetion of LC eoneentration. The deerease in TgMatrix on inerease of the LC eontent from 0 wt% up to 20 wt% ean be explained by the plastieizing effeet ofthe LC. Above 20 wt% ofLC, the glass transition temperature ofthe polymer matrix does not ehange further, indieating the limited solubility of the LC in the matrix. At around 30 wt% LC, phase separation oeeurs and thus the LC glass transition and the nematie - isotropie transition ean be deteeted. The TgLC values remain eonstant in the LC eoneentration range from 30 to 100 wt%, indieating that the phase-separated LC is essentially pure. These observations agree well with the optieal studies in terms of the LC solubility limit. A signifieant reduetion of the transmitted intensity was found on inerease of the LC eoneentrations above 20 wt%. The phase separation leads to the formation ofLC miero-droplets and thus to inereased seattering. In eontrast with the small eoneentration dependenee of the glass transition values, the nematie - isotropie transition temperatures were strongly influeneed by inereasing LC eoneentration. The dependenee of the T NI values on the LC eontent is qualitatively similar to a phase diagram eharaeterized by an upper eritieal solution temperature (UCST) [11]. A maximum of TN1 was found for sampies eontaining 50 wt% LC. Figure 3 also shows clear evidenee of the existenee of a
J. Therl1Ull Anal., 51, 1998
743
MASCHKE ct al.: LIQUID CRYSTAL-POLYMER COMPOSITES
60 40
E l!,>
~GI
0-
E ~
20 0 -20
1-:
-40
•
Tg E7 TgMatrix T N1
-60 -80 0
10
20
30
40
50
60
70
80
90
100
Liquid Crystal content (weight-%)
Fig. 3 Phase diagram of LC/polymer composite materials prepared by a polymerization-induced phase separation process using EB radiation
pure nematie phase for LC eoneentrations above 70 wt%, as already found in other PDLC systems [12]. One of the major parameters governing the properties of PDLC materials is the phase separation proeess. Smith has shown [8, 9] that the LC solubility limit, A, in the polymer matrix ean be related to the nematie - isotropie enthalpy m N1 and'to the inerease in speeifie heat eapaeity at TgLC : P(x) _ _~_H--,-N=I(,--,X), - ~HNI(LC)'
or
P(X) _ _~_C_pr...:(x---,)_
(1)
~Cp(LC)
P(x) represents the ratio of the nematie - isotropie transition enthalpy (or the heat eapaeity inerement) for a LC / polymer eomposite material to the equivalent value for the pure LC. This expression is based on the following assumptions: a) the LC in the droplets exhibits the same thermophysieal behaviour as in the bulk state; b) the amount of LC dissolved in the polymer matrix is eonstant for LC eoneentrations (weight pereentage) x2A and does not eontribute to m NI or to ilCp ; e) the TN1 values as a funetion of x remain unehanged, so that m N1 is not influeneed by the presenee of matrix moleeules dissolved in the LC droplets; d) the densities of the polymer preeursors and the LC are approximately equal. The effeets of the LC eoneentration on m N1 and on ilCp are presented in Fig. 4. Both quantities inerease linearly with x, validating the model given in Eq. (1). The LC solubility limitA was determined by linear regressions ofthe experimental data sets in Fig. 4, followed by ealculating the x-axis intereepts. In the ease of mNJ, the value of A was 28 wt%, whereas in the ease of ilCp , a value of 15 wt% was found. The lower value of A from ilCp than from m N1 ean be ex-
1. Thermal Anal.. 51. 1998
MASCHKE et al.: LIQUID CRYSTAL-POLYMER COMPOSITES
744
0.6
7 (*l
6
0.5
P
~CPE7* -
,
0.4
I
,
~
~HNI
.-
5
-
-
*
.r
,----~-
~ 0.3 [;j
1>
:I:
;!;
c:::
3 ~
Q.
U
4
0.2 ,
0.1 , '.
0.0
0
10
20
.
)
2
,,
80 30 40 50 60 70 Liquid Crystal content (weight-%)
90
0 100
Fig. 4 Influenee of the LC eontent on the nematie - isotropie transition enthalpy the heat eapaeity ~Cp
MINI
and on
plained by the lower solubility of the LC in the matrix at the LC glass transition temperature TgLC =-60°C than at the nematic - isotropie transition temperature TN1 =+52°C [13-15]. These results were confirmed by the optical studies, where a value of A=20 wt% was found at the intermediate temperature of 20°e. tMlN1 and L\Cp can be used to calculated
J (IOO J
a exp =m~C - = ( 1 +mp mLC
mLC
(x) = -
(x)
(2)
x
where m~c represents the mass of LC included in the droplets, while mp and m LC are the masses of the polymer matrix and of the LC in the sampie, respectively. Combining Eqs (1) and (2) yields 100
acalc
= ( ---;-
Y A100x-A- A 1
(x
~A)
(3)
Figure 5 illustrates the dependence of (l on the LC concentration. The points represent
J. Thermal Anal.. 51. 1998
MASCHKE et al.: LIQUID CRYSTAL-POLYMER COMPOSITES
ö f/)
~
~
u
1,0
0,8
g
0,7
-0
~ro
CQ)
f/)
Q) f/)
ro
.s::
C-
ec
*
0,6
/
,,
0,3 0,2 0,1
~
0,0 0
-_..
,
.--
,
0,4
0
-
'*
0,5
.'Ö LL
*
0,9
-0
':;
745
I
*
_~--,
a~1oN UI!)(pr-l
cx calcCp
Ct..llpCp
I
,, 10
20
30
40
50
60
70
80
90
100
Liquid Crystal content (weight-%)
Fig. 5 Dependence of the phase separated LC fraction on the LC content. The points represent u exp values determined for each composition by applying Eg. (2), whereas the curves were ca1culated by using the A values from Fig. 4 and Eg. (3)
LC concentration range where the LC is entrapped in the droplets. Within these limits, the maximum values of Ucalc were 0.92 (determined by using ~Cp) and 0.83 (using MINI), observed for 70 wt% E7. Higher LC contents lead to macroscopic phase separation; the corresponding U values in this case represent the total LC fraction included in the micro- and macroseparated phases.
Conclusions The relationship between the optical, electro-optical and thermodynamic properties of LC / polymer composite materials has been studied. Various blends made of a reactive mixture of acrylate derivatives and of LC E7 were efficiently cured by EB radiation, using a polymerization-induced phase separation process. The electro-optical properties ofthe composite films obtained are strongly related to the LC concentration. The off-state scattering was highly increased by changing the LC content from 20 to 70 wt%. The electro-optical curves for sampIes of 50 to 70 wt% E7 can be characterized by high transmission values in the on-state. Increasing the LC content from 40 to 70 wt% leads to substantiaIly reduced threshold and saturation voltages. An extensive thermophysical investigation of the cured materials was carried out. Calorimetric measurements of ~Cp (at T=TgLc) and MINI (at T=TNI ) were used to evaluate the LC solubility limitA in the composite materials. The values found (i.e. 15 and 28 wt%, respectively) agree weIl with the LC solubility limit determined by means of optical studies. Significant reductions in the transmitted intensity occurred at LC concentrations J. Thermnl Anal., 51, 1998
746
MASCHKE ct al.: LIQUID CRYSTAL-POLYMER COMPOSITES
between 20 and 30 wt%. The LC fraction included in the droplets, a, was calculated from the ßCp and !:JIN1 values. The sampies including 70 wt% of E7, which exhibit the best electro-optical performance, can be characterized by a maximum value ofa=0.85. Further analysis is in progress to correlate the results obtained from this study with the characterization of the droplet morphology.
* * * The authors gratefully acknowledge the support ofthe C.N.R.S., the Region Nord-Pas de Calais and the Ministere de l'Enseignernent Superieur et de la Recherche. The authors are indebted to N. Gogibus, B. Duponchel and V. Allouchery for their assistance, and to Dr. Hugelin for providing reactive prepolyrner sarnples.
References 1 J. W. Doane in Liquid Crystals-Applieations and Uses, Ed. B. Bahadur, World Seientific, Singapore 1990, Vol. I, Chap. 14, pp. 361-395. 2 P. S. Drzaie in Liquid Crystal Dispersions, World Seientifie, Singapore 1995. 3 N. A. Vaz, G. W. Srnith and G. P. Montgornery Jr, Mol. Cryst. Liq. Cryst., 197 (1991) 83. 4 U. Masehke, X. Coqueret and C. Loueheux, J. Appl. Polyrn. Sei., 56 (1995) 1547. 5 U. Masehke, J.-M. Gloaguen, J.-D. Turgis and X. Coqueret, Mol. Cryst. Liq. Cryst., 282 (1996) 407. 6 G. W. Srnith, Mol. Cryst. Liq. Cryst., 241 (1994) 77. 7 F. Roussel, J.-M. Buisine, U. Masehke and X. Coqueret, Liq. Cryst., 24 (1998) 555. 8 G. W. Srnith and N. A. Vaz, Liq. Cryst., 3 (1988) 543. 9 G. W. Srnith, Mol. Cryst. Liq. Cryst., 180B (1990) 201. 10 J. L. West, Mol. Cryst. Liq. Cryst., Ine. Nonlin. Opt., 157 (1988) 427. 11 T. Nishi, J. Maerornol. Sei. Phys., B 17, 3 (1980) 517. 12 C. Shen and T. Kyu, J. Chern. Phys., 102 (1995) 556. 13 G. W. Srnith, G. M. Ventouris and J. L. West, Mol. Cryst. Liq. Cryst., 213 (1992) 11. 14 F. Roussel, PhD Thesis, Universite du Littoral, Dunkerque (Franee) 1996. 15 F. Roussel, J.-M. Buisine, U. Masehke, X. Coqueret, Mol. Cryst. Liq. Cryst., 299 (1997) 321.
1. Therl1Ull Anal.. 51. 1998