Optimization of holographic polymer dispersedliquid crystal for ternary monomersYH Cho,1 BK Kim1* and KS Park2

1Department of Polymer Science and Engineering, Pusan National University, Pusan 609-735, Korea2Samsung Advanced Institute of Technology, Yongin, Kyungki 449-900, Korea

Abstract: Polarized optical micrography (POM) images of gratings and UV±visible spectra of

holographic polymer dispersed liquid crystals (HPDLC) are reported for a ternary monomer system

composed of dipentaerythritol hydroxypentaacrylate/trimethylolpropanetriacrylate/N-vinylpyrroli-

done (DPHPA/TMPTA/NVP)=7/2/1 by weight. Gratings were written by irradiation with an argon ion

laser (l=488nm) at various intensities (20±200mWcmÿ2) on monomer/liquid crystal (LC) composite

®lms of various compositions (75/25, 70/30, 65/35, 62/38, 60/40). Re¯ection ef®ciency±irradiation

intensity±®lm composition relationships are obtained in three dimensional plots which show that

maximum re¯ection moves from high LC content (38%) at low irradiation intensity (20mWcmÿ2) to

low LC content (25wt%) at high irradiation intensity (200mWcmÿ2).

# 1999 Society of Chemical Industry

Keywords: holographic polymer dispersed LC; re¯ection; monomer

INTRODUCTIONA polymer/liquid crystal composite ®lm, often referred

to as polymer dispersed liquid crystal (PDLC), is a

thin composite ®lm composed of micrometre-sized

droplets of a nematic liquid crystal (LC) dispersed in a

polymer matrix,1±5 typically made of UV-curable

acrylates because of their high optical clarity and other

adjustable properties.12±16 PDLC has the potential for

a variety of electrooptic applications ranging from

directly driven windows to active matrix driven

information displays. A number of important reviews

regarding the methods of preparation, materials,

modes of operation, device applications, etc, have

become available.6±11

Recently the laser grating technique has been

applied to fabricate PDLCs with controlled architec-

tures of phase-separated LC domains.17±23 This type

of PDLC is called a holographic polymer dispersed

liquid crystal (HPDLC) in the literature.22 This new

technique combines polymerization-induced phase

separation (PIPS) and Bragg's law (eqn (1)).

L � l2 sin �

�1�

where l is the wavelength of light, L is the grating

spacing, and 2y is the interbeam angle outside the ®lm.

The grating spacing, and hence the component of

incident light which can be re¯ected by the grating

spacing, can be easily controlled by varying the

interbeam angle using the same laser wavelength.23

This periodic structure of multilayers has very

promising optical properties because only a speci®c

component of the incident light is re¯ected by the LC

layers, due to the difference in refractive indices of the

polymer and LC.

In HPDLC, light scattering is virtually minimized

because the domain sizes are of the order of nano-

metres, so that this device operates on light re¯ection

or transmittance, and its ef®ciency is electrically

controlled via the refractive index of the LC mol-

ecules.17,23

In the PIPS method, the phase separation is driven

by the polymerization reaction. The rate of polymer-

ization given below for a photoinitiated radical mech-

anism largely governs the phase separation and hence

the domain size.

Ri � 2�I0�1ÿ exp�ÿ2:3e�A�b�� �2�

Rp � kp�M� �eI0�A�bkt

� �0:5

�3�

Ri and Rp are the initiation and polymerization rates,

respectively and details of their application are avail-

able from the pioneering works by Decker.24,25 In

these eqns (2) and (3), f, e, I0 and b are, respectively,

initiation ef®ciency, molar absorptivity, incident light

intensity, and sample thickness. [M] and [A] are the

Polymer International Polym Int 48:1085±1090 (1999)

* Correspondence to: BK Kim, Department of Polymer Science and Engineering, Pusan National University, Pusan 609-735, KoreaContract/grant sponsor: Korean Ministry of Commerce, Industry and EnergyContract/grant sponsor: Korean Ministry of Science and Technology(Received 15 September 1998; revised version received 5 March 1999; accepted 13 May 1999)

# 1999 Society of Chemical Industry. Polym Int 0959±8103/99/$17.50 1085

concentrations of monomer and species which under-

go photoexcitation, kp and kt are the rate constants for

propagation and termination reactions, respectively.

With a suf®ciently rapid rate of polymerization, phase

separation cannot follow the polymerization kinetics

because the time-scale for polymerization is smaller

than that of phase separation due to the slower

diffusion of LC molecules in highly viscous media.

Therefore, monomers to be used for HPDLC should

have higher functionality than those for conventional

PDLC.

As a continuation of our efforts in PDLC,26±29 we

optimized holographic PDLC with regard to ®lm

composition and irradiation intensity. Ternary mono-

mer mixtures with various monomers/LC composi-

tions have been irradiated with an argon ion

(l=488nm) laser at various intensities. Three dimen-

sional plots of re¯ection ef®ciency±laser intensity±®lm

composition are presented showing contours for

maximum re¯ections.

EXPERIMENTALMaterialsA eutectic mixture of four cyanobiphenyl and cyano-

terphenyl components with TKN=ÿ10°C, TNI=

60.5°C, ek=19.0, and e?=4.2 (E7, Merck) was used

as LC. Three types of photopolymerizable monomers

(Fig 1) (dipentaerythritol hydroxypentaacrylate

(DPHPA, functionality ( f )=5), trimethylol-

propanetriacrylate (TMPTA, f =3), and N-vinyl-

pyrrollidone (NVP, f =1) were used in appropriate

combinations to prepare the host polymers upon laser

irradiation.

DPHPA and TMPTA have very high reactivity

together with high viscosity because of their high

molecular weight, and provide the polymers with

extensive crosslinkings, whereas monofunctional NVP

simply extends the chains at a much slower rate.

However, the use of monofunctional monomers is

often essential to reduce the viscosity of LC/monomer

mixtures and to make the starting mixture homo-

geneous. Otherwise, polymerization-induced phase

separation starts with the heterogeneous reaction

mixture and the morphology of the composite ®lm

becomes out of control.

A dye, Rose Bengal (RB), was used as photoinitiator

for holographic recording with an argon ion laser,

because it displays a broad absorption spectrum with a

peak molar extension coef®cient of about

104Mÿ1cmÿ1 at about 490nm.17 To this, a millimolar

amount of N-phenylglycine (NPG) was added as

coinitiator. In this experiment 3�10ÿ6M of RB and

1.2�10ÿ4M of NPG were used.

FormulationGratings were formulated with different ®lm composi-

tions (monomers/LC) which were irradiated at various

laser intensities. Basic formulations of our ternary

DPHPA/TMPTA/NVP systems are given in Table 1.

The monomer composition was ®xed at DPHPA/

TMPTA/NVP=7/2/1 by weight, and the effects of

®lm composition and irradiation intensity were

examined.

GratingsThe holographic recording system used is schemati-

cally shown in Fig 2. An argon ion laser (l=488nm)

was used as light source. The beam passes through a

spatial ®lter, a beam expander, and is split in two

beams of identical intensity. These two beams are

subsequently passed through a collimator and only the

central portions are re¯ected from the mirrors to

impinge normally on the cell from the opposite side.

The cell was constructed by sandwiching the mono-

mers/LC between two indium±tin-oxide (ITO) coated

Figure 1. Chemical structures of photopolymerizable monomers.

Table 1. Formulation to prepare HPDLC from DPHPA, TMPTA and NVP

Monomer LC: Monomer Rose Bengal (wt%) NPG (wt%) Intensity (mWcmÿ2) Space thickness (mm)

Pentaacrylate=DPHPA 25:75 20

Triacrylate=TMPTA 30:70 50

Monofunctional=NVP 100

35:65 0.5 2.5 14

150

38:62 175

Penta-:tri-:mono-=7:2:1 40:60 200

1086 Polym Int 48:1085±1090 (1999)

YH Cho, BK Kim, KS Park

glass plates, with a gap of 14mm adjusted by a bead

spacer.30 Interference of the two beams establishes the

periodic interference pattern according to the Bragg

law; this is approximately 488nm in our case. The

laser intensity was varied from 20 to 200mW/cmÿ2,

with typical exposure times of 30±120s.

Figure 2. Experimental set-up for holographic PDLC.

Figure 3. POM micrographs of HPDLCversus film composition (monomers/LC):(a) 75/25; (b) 70/30; (c) 65/35; (d) 62/38and (e) 60/40 (I0=100mWcmÿ2).

Polym Int 48:1085±1090 (1999) 1087

Optimization of HPDLC for ternary monomers

MeasurementsThe morphology of the composite ®lm was studied by

polarized optical microscopy (POM). The re¯ection of

a speci®c wavelength by the composite ®lm was

analysed using a UV±visible spectrometer (Perkin

Elmer, Lambda 20). The re¯ection ef®ciency was

estimated from the spectral data.

RESULTS AND DISCUSSIONPolarized optical microscopy (POM) images of our

gratings for various ®lm compositions are shown in Fig

3. The width variation of the grating is very small

throughout the ®lm compositions. In general, spatial

variation of the grating width is observed due to

inhomogeneity of the laser spot. This can be mini-

mized by using a more square beam pro®le, which is

obtained by expanding and collimating the Gaussian

beams and passing only the central portion through an

aperture in contact with the front plate of the

sample.19 This procedure was used in our experi-

ments. With the increase of LC content in the ®lm, the

thickness of the LC lamella (dark area) increases while

the Bragg spacing remains constant.

UV±visible spectra of the ®lms are shown in Fig 4.

Two peaks are observed at about 480 and 580nm,

which correspond to the re¯ection by the holographic

grating and absorption by the dye Rose Bengal,

respectively. The Bragg spacing is slightly smaller than

the incident laser wavelength (488nm), presumably

Figure 4. Irradiation intensity dependence of UV–visible spectra of HPDLC films (monomers/LC): (a) 75/25; (b) 70/30; (c) 65/35; (d) 62/38 and (e) 60/40.

1088 Polym Int 48:1085±1090 (1999)

YH Cho, BK Kim, KS Park

because of the shrinkage of the mixture upon

polymerization. As mentioned above, the 480nm peak

will be approximated as re¯ection by the gratings

because the scattering is small, with nanometre sized

domains. At low LC content (monomer/LC=75/25),

the re¯ection intensity is maximum for high irradiation

power, and at high LC content (60/40), the peak

intensity is maximum for low irradiation power; at

intermediate composition (62/38), the maximum

re¯ection is obtained for intermediate irradiation

power.

In HPDLC, the peak intensity should depend on the

perfectness of holographic gratings. Obviously, more

perfect gratings give higher peak intensity. Then, what

is governing the perfectness of the grating? The answer

should be the proper LC±polymer phase separation,

where nanometre sized LC domains imbedded in

polymer layers are separated by almost LC-free

polymer layers. Phase separation in the polymerizing

system is regarded as a liquid±liquid demixing process

where spinodal decomposition prevails.31,32 Elemen-

tary Flory±Huggins theory is often used to obtain the

interaction energy which increases with the progress of

the polymerization reaction.

Following Tanaka et al,22 small LC droplets of high

density give higher re¯ection ef®ciency. Our results

indicate that an optimum polymer±LC phase separa-

tion may exist. According to the reaction kinetics of

photoinitiated radical polymerization described

earlier, it has been noted that both Rp and Ri increase

with irradiation intensity I0, although the effect of I0 on

Ri is more pronounced. Rp also increases linearly with

monomer concentration, which corresponds to the

monomer content of the composite ®lm. Therefore,

high monomer content directly gives a high reaction

rate and hence augments the crosslinking density of

host polymers.

The polymerization rate is highest when the ®lms of

highest monomer content are irradiated at the highest

laser intensity; conversely, it is lowest when the ®lms of

lowest monomer content are irradiated at the lowest

laser intensity. When the polymerization rate is too

fast, the rate of phase separation cannot follow the rate

of network formation and LC domains are entrapped

within the polymer nets leading to imperfect gratings.

Also, coalescence of LC domains into larger ones

becomes less plausible with a highly viscous host

polymer matrix, and this also retards phase separation.

However, when the polymerization rate is too low,

phase separation cannot take place thermodynami-

cally. Therefore, there exists an optimum monomer

content for the desired maximum re¯ection, depend-

ing on the irradiation intensity.

The re¯ection ef®ciency±irradiation power relation-

ship is shown in Fig 5 for various LCs. One observes a

monotonic increase, then an asymptotic increase, and

®nally a maximum when the LC content increases

from 25 to 40wt%. It seems that some monomers

necessitate more powerful irradiation.

The same data were replotted for re¯ection

ef®ciency±®lm composition in Fig 6. Regardless of

the irradiation power, the re¯ection ef®ciency shows a

maximum for a given LC content, its value increasing

as the irradiation power increased. However, the LC

content of maximum re¯ection decreased with in-

creasing irradiation power.

A three-dimensional representation of the re¯ection

ef®ciency±irradiation intensity±®lm composition rela-

tionships is shown in Fig 7. This plot shows that the

maximum re¯ection depends on ®lm composition and

laser intensity, and the optimum set of conditions is

much narrower than with binary monomers.33 The

maximum re¯ection moves from high LC content

(38wt%) at low irradiation intensity (20mWcmÿ2) to

low LC content (25wt%) at higher irradiation

intensity (200mWcmÿ2).

CONCLUSIONSThe re¯ection ef®ciency of holographic polymer

dispersed liquid crystals (HPDLCs) for ternary

monomers has been studied as a function of ®lm

composition and irradiation the intensity. Regardless

of the irradiation power, the re¯ection ef®ciency

showed a maximum with regard to the LC content;

the maximum value generally increased with increased

irradiation power. In contrast, higher irradiation

power gave maximum ef®ciency at lower LC content,

Figure 5. Reflection efficiency versus laser intensity of HPDLC films forvarious LC contents.

Polym Int 48:1085±1090 (1999) 1089

Optimization of HPDLC for ternary monomers

implying that optimum irradiation power should

depend on the amount of monomers to be cured.

ACKNOWLEDGEMENTSThe research is in the program of G7 project which has

been supported by the Korean Ministry of Commerce,

Industry and Energy, and the Korean Ministry of

Science and Technology. The ®nancial support is

gratefully acknowledged.

REFERENCES1 Zumer S and Doane JW, Phys Rev A 3373 (1986).

2 Drzaic PS and Gonzales AM, Appl Phys Lett. 73:1332 (1993).

3 Fergason JL, SID Technical Digest 16:68 (1985).

4 Montgomery GP, Proc SPIE 242(9):1080 (1989).

5 Miyamoto Y, Kikuchi H, Morimura Y and Kajiyama T, New

Polym Mater 2:1 (1990).

6 Doane JW, in Liquid Crystal±Applications and Usage, vol 1, p 362,

World Scienti®c, Jersey (1990).

7 Crawford GP, Doane JW and Zumer S, in Handbook of Liquid

Crystal Research, Ed by Collings PJ and Patel JS, p 347, Oxford

University Press, Oxford (1997).

8 Drzaic P, Mol Cryst Liq Cryst 154:289 (1988).

9 Kitzerow HS, Liq Cryst 16(1):1 (1994).

10 West JL, Mol Cryst Liq Cryst 157:427 (1988).

11 Drzaic PS, Liquid Crystal Dispersion, World Scienti®c, Singapore

(1995).

12 Nolan P, Tillin M and Coates D, Mol Cryst Liq Cryst Lett,

8(6):129 (1992).

13 Fujisawa T, Ogawa H and Muruyama K, Jpn Display 690 (1989).

14 Ono H and Kawatsuki N, Jpn J Appl Phys Part 1 33:6268 (1994).

15 Smith G, Mol Cryst Liq Cryst 180:201 (1990).

16 Cho YH and Kim BK, J Polym Sci Polym Phys 36:1393 (1998).

17 Bunning TJ, Natarajan LV, Tondiglia VP, Dougherty G and

Sutherland RL, J Polym Sci Polym Phys 35:2825 (1997).

18 Sutherland RL, Tondiglia VP, Natarajan LV, Bunning TJ and

Adams WW, Appl Phys Lett 64:1074 (1994).

19 Zhang J, Carlen CR, Palmer S and Sponsler MB, J Am Chem Soc

116:7055 (1994).

20 Rosa MEC, Tondiglia VP and Natarajan LV, J. Appl Polym Sci

68:523 (1998).

21 Murai H, Gotoh T, Suzuki M, Hasegawa E and Mizoguchi K,

SPIE, 1665:230 (1992).

22 Tanaka K, Kato K, Date M and Sakai S, SID 95 Digest, 267

(1995).

23 Date M, Naito N, Tanaka K, Kato K and Sakai S, Asia Display

95:603 (1995).

24 Decker C, Polym Photochem 3:131 (1983).

25 Decker C, J Coating Technol 56(713):29 (1984).

26 Kim BK and Ok YS, J Polym Sci Polym Phys 32:561 (1994).

27 Kim BK and Kim SH, J Polym Sci Polym Phys 36:55 (1998).

28 Kim BK, Ok YS and Choi CH, J Polym Sci Polym Phys 33:707

(1995).

29 Kim BK, Kim SH and Choi CH, Mol Cryst Liq Cryst 261:605

(1995).

30 Kim BK, Hong EY and Ok YS, Korea Polym J 5(2) 77 (1997).

31 Lin JC and Taylor PL, Mol Cryst Liq Cryst 237:25 (1993).

32 Kim JY and Palffy-Muhoray P, Mol Cryst Liq Cryst 203:93

(1991).

33 Kim BK and Jeon YC, Mol Cryst Liq Cryst in press.

Figure 7. Reflection-efficiency–LC-content–laser-intensity relationships ofHPDLC films (NVP/TMPTA/DPHPA=1/2/7).

Figure 6. Reflection efficiency versus LC content of HPDLC films irradiatedat various laser intensities.

1090 Polym Int 48:1085±1090 (1999)

YH Cho, BK Kim, KS Park