Dependence of the morphology of polymer...

Dependence of the morphology of polymer dispersed liquid crystalson the UV polymerization process

S. A. Carter,a) J. D. LeGrange,b) W. White, J. Boo, and P. WiltziusBell Laboratories, Lucent Technologies, 600 Mountain Avenue, Murray Hill, New Jersey, 07974

~Received 17 October 1996; accepted for publication 15 January 1997!

Using confocal microscopy, we have studied the morphology of polymer dispersed liquid crystals~PDLC! as a function of polymer/liquid crystal composition, polymer cure temperature, andultraviolet ~UV! curing power and determined how this morphology affects the electro-opticalproperties. The PDLC morphology consists of a spongelike texture where spherically shaped liquidcrystalline domains are dispersed in a polymer matrix. These domains grow as the fraction of liquidcrystal increases and as the UV curing power decreases. We observe no significant changes indomain size with changes in the curing temperature. Instead, high-temperature cures result incoalescence and the formation of elliptical-shaped liquid crystal domains. The temperature at whichthis coalescence starts to be observed marks a threshold temperatureTth , above which the switchingproperties are strongly dependent on morphology. BelowTth the switching properties are largelyindependent of morphology. ©1997 American Institute of Physics.@S0021-8979~97!05508-4#

w

cuientelheinheteiniteuithhe

odstoneos

ecu

oaeow

usaseationasf thepa-seralLCeireomop-hedltn theap-be

lt-atndenrop-h isix,

gyle-e-

cef agylytrix,uryer-

CA

INTRODUCTION

Polymer dispersed liquid crystals~PDLCs! are promis-ing materials for reflective and projection displays, windoshutters, and holographic recording media1–4 due to theirrelatively low cost and ease of processing over other tenologies. PDLCs consist of a homogeneous mixture of liqcrystal and monomer in which phase separation has binduced by either thermal curing, UV curing, or solveevaporation. For polymer/liquid crystal compositions revant for display applications, the polymerization of tmonomer causes the liquid crystal to phase separate‘‘droplets’’ or domains separated by the polymer matrix. Tnematic texture within these domains is randomly orienwith respect to the neighboring domains such that incomlight into the cell is scattered, and the PDLC appears whWhen an electric field is applied across the cell, the liqcrystal domains align, and the PDLC is transparent ifordinary index of refraction is matched to the index of tpolymer. The voltageV and speedt at which the PDLCswitches from whiteness to transparency and the optical ctrast between the white and transparent state is controllethe size, shape, and anchoring energy of the liquid crydomains, as well as the dielectric anisotropy and viscositythe liquid crystal~LC!. While the latter properties depend othe selection of the liquid crystalline material, the formproperties depend on the liquid/crystal monomer comption and the method of polymerization.

Several investigations5,6 into the phase diagram of thPDLC mixture have shown that the phase separation ocalong a line in temperature-composition space~Fig. 1!. Be-fore polymerization the system consists of a homogenemixture of monomer and LC. The isotropic phase occursconditions above this line, and a phase separated mixturnematic LC-rich and isotropic monomer-rich regions bel

a!Currently at Physics Department, University of California, Santa Cruz,95064.

b!Electronic mail: [email protected]

5992 J. Appl. Phys. 81 (9), 1 May 1997 0021-8979/97/8

Downloaded¬09¬Oct¬2003¬to¬128.114.130.133.¬Redistribution¬subjec

h-den

-

to

dg.de

n-byalf

ri-

rs

ustof

it. When polymerization is initiated in the homogeneostate, the growth of the polymer matrix results in an increin temperature of the phase separation line. Phase separinto LC-rich domains occurs when the polymerization hprogressed far enough such that the cure temperature oPDLC mixture matches the temperature of this phase seration lineTps. At this point, nucleation of LC domains iinitiated and the LC domains start to grow. There are sevtime scales that are important for determining the PDmorphology: the time it takes for the drops to grow to thspace limited size via molecular diffusion; the time it for thdrops to coalesce via the movement of polymer matrix frbetween neighboring droplets; the time for coalesced drlets to relax into spherical domains; and the time for tpolymer matrix to sufficiently solidify to retard droplet anmatrix movement.6 Extracting these time scales is difficusince they themselves are time dependent and depend oextent of the phase separation process. In a simplifiedproach, one can assume that the rate of solidification cancontrolled by the UV curing power, with lower power resuing in larger droplets, and by the extent of polymerizationthe phase separation point, with lower compositions ahigher temperature resulting in a larger polymer matrix whphase separation is initiated and subsequently smaller dlets. In contrast, the coalescence of the LC droplets, whicdetermined in part by the viscosity of the polymer matrwill cause higher temperatures~lower viscosity! to result inmore coalescing and larger droplets. The PDLC morpholoshould be largely determined by the mechanism for nucation of the liquid crystal domains and the competition btween these time scales.

In this article we attempt to deconvolute the importanthat these effects have in determining the morphology oPDLC. We study compositions with foam-type morpholo~;80% LC! where the light scattering occurs primariamong droplets, rather than that between LC and maleading to a reduction in scattering at larger angles. Omethodology is to fix the monomer polymer viscosity bfixing temperature, and then control the extent of polym

1(9)/5992/8/$10.00 © 1997 American Institute of Physics

t¬to¬AIP¬license¬or¬copyright,¬see¬http://ojps.aip.org/japo/japcr.jsp

tioInththmrgthexni, iceerusloti

in

e

inadde

-waththno--hadbn

singellsm-10–

sehe-mer

m-ofhegughe

cell

-

rd-

al

n-s ingcesicaleri-

ect

ame

--heso-oflyislar

-

enm-toe-loseb-ano-gthall

estoioneratiioes

ization at the phase boundary and the rate of polymerizaby varying composition and UV cure power, respectively.addition, we change the polymer matrix/LC viscosity andextent of polymerization at the phase boundary by fixingcomposition and curing power and curing at different teperatures. In the former experiments, we observe lachanges in the morphology which can be explained bysimplistic model described above. However, contrary topectations, we observe that the drop size does not sigcantly change with increasing cure temperature. Insteadcreasing cure temperatures cause the droplets to coalesform elliptical shaped domains which become larger sphcal domains at sufficiently high temperatures. We discpossible mechanisms for the formation of these morphogies and compare the morphology to the electro-opproperties.7

EXPERIMENTAL METHODS

The PDLC cells were assembled in a clean room uslarge glass plates, 4203300 mm2, coated with indium–tin–oxide ~ITO! and separated by 8mm spherical spacers madof plastic. For the microscopy measurements, a 175-mmthick ITO-coated glass plate was used as the top plate sthe confocal microscope used for the measurements h200 mm focal range. The two plates were sealed unvacuum using a UV curable glue.7 The resulting cell gap was13.064.0mm due to the flexibility of the top plate. Measurements of cell gap over a single cell showed that the gapuniform to within 15%. Repeated measurements showedthe morphology was not affected by the variances ingap.8 The PDLC mixture consists of an acrylate based momer mixture~Merck PN393! and a superfluorinated LC mixture ~Merck TL213! where the LC fraction was varied between 76% and 82%. The 82% mixture had an average pseparation temperature of 15 °C which decreased withcreasing LC fraction as shown in Fig. 1. A fluorescent pro~Exciton Corporation!, pyromethene 580, of weight fractio

FIG. 1. Binary phase diagram for TL205/PN393 LC/polymer mixturwhere I is isotropic andN is nematic. The coexistence curves priorphotopolymerization are indicated by solid black lines. As polymerizatproceeds, the coexistence curves move to higher temperatures and lowfraction.Tps is the temperature at which the phase separation into nemrich LC droplets and isotropic-rich polymer occurs. The thatched regTth is the threshold cure temperature at which the droplets start to coal

J. Appl. Phys., Vol. 81, No. 9, 1 May 1997


n

ee-ee-fi-n-to

i-s-c

g

cear

sate-

see-e

in the range of 1024–1023 was added to this mixture. Thiprobe dissolved preferentially into the monomer enablgood optical contrast between the LC and polymer. The cwere filled by capillary forces and then UV cured on a teperature controlled plate at temperatures in the range of48 °C. The UV lamp power was varied from 22mW/cm2 to16 mW/cm2 using neutral density filters. In all cases, phaseparation occurred within 20 s of UV exposure; nonetless, the cells were cured for 400 s to ensure that the polymatrix was completely cross linked.

The PDLC morphology was determined using a comercial fluorescent confocal microscope with a resolution0.4 mm. A 60 Hz voltage source was used to switch to ttransparent~on! state of the PDLC cell, sufficiently reducinthe light scattering so, that three-dimensional scans throthe entire thickness of the PDLC cell could be obtained. Wobserved that the PDLC morphology does not depend onthickness, except at distances roughly within 1–2mm of thetwo plates~Fig. 2 and Fig. 6 below!. Measurements, therefore, were typically obtained at a depth of 3–4mm below thetop plate. Switching voltage, switching speed, and standaized luminance~the reflectivity of white light normalized to a40% brightness standard! were measured using a normviewing angle display measurement system~Autronic-Melchers DMS-703! on both the microscopy cells and stadard thick glass cells prepared under the same conditionorder to verify consistency.7 In some cases the switchinvoltages and luminance had to be adjusted for the differenin gap size. The fluorescent probe did not affect the electrproperties in the range of concentrations used in the expment. At concentrations near 1023, a slight orange coloringof the PDLC cells occurred, making it necessary to corrthe measured values of luminance~by an offset,15%! inorder to match the standardized cell prepared at the sconditions without the probe.

RESULTS

In Fig. 3 the morphology of PDLCs cured at room temperature and at 16 mW/cm2 lamp power is shown as a function of LC fraction between 76% and 82%. Below 76% tLC domains are no longer clearly discernible within the relution of the microscope. The morphology consistssponge- or foamlike texture where the LCs form closepacked spherical domains within the polymer matrix. Thmorphology has recently been observed for simimixtures.6

Alternative PDLC systems with lower LC weight fraction, however, exhibit a more dropletlike morphology.9 At76% the domains have sufficient polymer matrix betwedroplets to form nondistorted spherical shapes with a diaeter of 0.6560.10mm. These domains grow in a size up1.860.2 mm at 82%, at which point the spherical shape bcomes distorted to a polygon/hexagon shape due to the cpacking of the domains. At this fraction we can clearly oserve that each domain is surrounded in the plane byaverage of five to six other domains. Above 82%, the dmains become sufficiently large compared to the wavelenof light that the visible scattering cross section is too smfor display applications.

,

LCc-nce.

5993Carter et al.


f thChenin

ing

sing

FIG. 2. Confocal visible microscopy scan through the entire thickness o10 mm PDLC cell in order from top to bottom of ITO-covered glass/PDLinterface, 3mm, 6 mm deep, and PDLC/ITO-covered glass interface. Tmorphology away from the interface is only weakly dependent on scandepth. The size of the picture is 32332 mm2.

5994 J. Appl. Phys., Vol. 81, No. 9, 1 May 1997


e

g

FIG. 3. Dependence of the PDLC morphology on composition at curconditions of 24 °C and 16 mW/cm2. Compositions from top to bottom are76%, 78%, 80%, and 82% LC. The droplets increase in size with increaLC fraction.

Carter et al.


th

veeninthtrs

eafoto

ifanechsinx

i-

soinoouefrosata

erohoinpe

enernohi

caet

ermanit

rea-

e at8,rgeren-

The average droplet~domain! diameter, determined byanalysis of a two-dimensional fast Fourier transform ofimage, is plotted versus composition~in Fig. 6!. The increasein droplet size with increasing LC fraction has been obserpreviously6,9 and can be explained by changes in the extof polymerization at the onset of phase separation. Withcreasing fraction, the PDLC mixture moves closer tophase boundary resulting in a less developed polymer mawhen phase separation occurs and more time for the LCform large domains. The phase separation temperatureTpsvaries from25 °C for 76% to 15 °C for 82%. Since the curtemperature is at 24 °C, the 82% sample is only 9 °C awfrom the phase transition point in comparison to 29 °Cthe 76% sample. The factor of 3 difference is also the facobserved in the difference in their droplet size~Fig. 6!, sug-gesting similar linear diffusion rates for the two systems.

To further test this model, we cured the polymer at dferent temperatures while keeping the composition constFor the 76% samples we observe a change in droplet sizless than 0.07mm for samples cured at 12 and 20 °C, a musmaller change than the 30% decrease in the dropletexpected from the above model in which the effect ofcreasing temperature is to cause a larger polymer matribe formed atTps resulting in less time for domains to form~smaller drops!. Similar results are observed for compostions 78% through 82%~summarized in Fig. 6!. These re-sults imply that the extent of polymerization at the phaseparation point cannot completely account for the morphogy changes. An alternative explanation would be thatcreasing temperature also causes a decrease in the viscof the polymer matrix, and therefore an increase in the mlecular diffusion rate and the coalescence rate. This woresult in the formation of larger droplets, offsetting thsmaller droplets expected due to the increased distancethe phase separation point. We note, however, that this ioverly simplistic picture, and that there are several other ftors, such as surface tension, LC anchoring energy, andnucleation mechanism itself which need to be taken intocount for a complete understanding of the morphology.

Our resolution at 76% is too low to directly observcoalescence; however, at 82% composition, the larger dlet size permits a more detailed study of changes in morpogy as shown in Fig. 4. We can identify three regionstemperature phase space delineated by a threshold temtureTth ~Fig. 1!. At low temperatures (Tps, T , Tth) repre-sented by the top panel of Fig. 4, we observe no coalescand a uniform sponge-type texture forms. As before, vlittle change in the domain size occurs. In the second pathe aggregation or coalescence of the droplets can beserved at 38 °C. We define the temperature range at wthis coalescence starts to be observed asTth . ForT ; Tth , thepolymer matrix freezes before the coalesced dropletsform spherical droplets, leading to elliptical shaped dropland subsequently large moments of inertia~Fig. 6!. At 48 °C,T @ Tth , the matrix viscosity is sufficiently reduced that thcoalesced droplets have sufficient time to relax and folarger, more spherical, droplets, as seen in the bottom pof Fig. 4. This same effect is observed at 78% and 80% w



e

dt-eixto

yrr

-t.of

ize-to

el--sity-ld

manc-hec-

p-l-

ra-

ceyelb-ch

ns

elh

Tth decreasing with lower LC fractions. This effect is modifficult to observe for the smaller droplet size, and our mesurements indicate that the elliptical droplets form at 30 °C,

FIG. 4. Dependence of the PDLC morphology on UV cure temperatur82% LC and 16 mW/cm2. Cure temperatures from top to bottom are 24, 3and 48 °C. Coalescence of the droplets starts to occur near 38 °C with ladroplets forming at higher temperatures. Droplet size is only weakly depdent on UV cure temperature below 40 °C.

5995Carter et al.


timiorizua

fthretiorrer

ethel

tkbyth

isiist tleoa23e

f

rthege

p,

theethna

lec-ill

the

ouresee of

eden-tgedga

note

thatllstslC,and

itedinge.letsi-ng

ndnalse

rerveedthec-ug-of

lte.c-e

Tth,36 °C for 80% and 24 °C, Tth,31 °C for 78%,roughly 25 °C aboveTps ~Fig. 1!.

Last, we vary the speed at which the phase separaline is approached by varying the UV power at fixed teperature and composition. The results for 80% compositcured at 24 °C, are shown in Fig. 5~a!, where the cure powehas been changed by a factor of 1000. The droplet sshown in Fig. 6, increased substantially with decreasing cpower below 8 mW/cm2. The average time for phase sepration increased from a few seconds for 16 mW/cm2 expo-sure to above 10 s for 23mW/cm2 exposure. The extent othe polymer matrix at phase separation is not affected bypolymerization rate; nonetheless, the lower UV cure ratesults in a slower freezing of the PDLC once phase separais induced, allowing more time for the LC to form largedomains as observed. If we repeat the experiment in thegime whereT . Tth , then the coalesced drops form largspherical drops as expected.

Typically, the morphology that subsequently forms dpends on the difference between the diffusion rate andpolymerization rate, unless the diffusion is limited by the cdimensions. In the bottom section of Fig. 5~a!, a vertical scanthrough the entire 13mm PDLC shows that at the lowesexposure the cell is still two to three LC droplet layers thicThe molecular diffusion has not been significantly limitedboundary effects, and the droplet size is determined bydiffusion rate, and the shape of the droplets~distorted-pentagon and hexagon! is determined by the packing. Thisnot the case for the 82% composition cured at 24 °Cshown in Fig. 5~b! where the scale on the bottom figure2.5 times the top scale. Here, the vertical scan shows thalayer is less than one droplet thick. As such, the dropgrow to the size of the gap at which point they start to clesce. The ‘‘donut-type’’ structure that occurs for themW/cm2 cure presumably results from trapped polymer btween coalesced neighboring droplets.

DISCUSSION

Several studies have shown that the morphology oPDLC strongly affects the electro-optic properties.6,7,9 Ex-pressions for such effects can be derived from equationslating the elastic energy density for the LC orientation tofrequency-dependent electric-field energy density insiduniform dielectric sphere, yielding a relation for the voltarequired to turn the cell ‘‘on’’~transparent! Von and the timeit takes to turn the cell offtoff of

Von;d/R~L221!1/2~4pK/De!1/2 ~1!

and

toff;gR2/@K~L221!#, ~2!

whereR is the average droplet radius,L 5 R2/R1 whereR2

andR2 are the major and minor radii of an elliptical drod is the cell gap,g is the rotational viscosity, andK andDeare the elastic constant and the dielectric anisotropy ofLC, respectively.10,11DecreasingR causes an increase in thsurface anchoring energy of the LC such that more voltagneeded to switch the LC. When this voltage is removed,same energy will cause the LC to switch back to its origi



on-n,

e,re-

e-n

e-

-el

.

e

s

hets-

-

a

e-ea

e

isisl

state faster. Elliptical droplets (L . 1) switch at higher volt-ages because they favor a preferred orientation that the etric field must overcome, but by the same argument they walso switch back to their original state more quickly oncefield is removed.

We have measured the electro-optic properties ofcells to understand the effect the morphology has on thproperties in light of the above theories. The dependencthe switching voltage on the inverse droplet radius 1/R as afunction of changing UV power and composition is graphin Fig. 7. For higher compositions and voltages, the depdence is linear as expected from Eq.~1! ~assuming constanL!; however, we clearly observe that the switching voltadoes not depend onR over certain ranges of composition antemperature. In addition,Von does not depend on UV curinpower, even thoughR increases significantly. There existsminimumVon, 5.060.4 V for a set gap size of 8.5mm and7.060.4 V for a set gap size of 11mm, which occurs at lowenough temperatures regardless of the drop size. Wethat when the droplets are spherical, i.e.,L 5 1, Eq.~1! ap-proaches zero; our morphology measurements indicatethe droplets remain very close to spherical in PDLC ceexhibiting minimum switching voltage. This result suggesthat a complete equation forVon would require an additionaterm that must depend on the intrinsic properties of the Lsuch as the dielectric anisotropy and the elastic constant,not on the PDLC morphology. The slight increase inVon thatoccurs for the larger droplets~small 1/R) is probably due tothe distortion away from sphericity (L . 1) that occurs in thepacking of the larger droplets.

The dependence ontoff should be similar toVon. Figure7 shows changes intoff vsR

2 for changes in UV power andcomposition. We observe linear dependence over a limcomposition range, but at higher compositions the switchtime drops off in agreement with the slight rise in voltagThis is presumably due again to the distortion of the dropfrom sphericity (L . 1) which occurs at the higher compostions. As before, UV power does not effect the switchivoltage even thoughR is changing significantly. A possibleexplanation here would be that whenT , Tth , L ; 1 and Eq.~2! becomes undefined (toff approaches infinity! and anothereffect must limit the switching time. This effect must depeon the intrinsic properties of the LC, such as the rotatioviscosity and elastic constant. The functional form of theadditive terms is currently under further investigation.

The luminance~LUM ! shows a fair amount of scatteover the samples studied. Nonetheless, we can still obs~Fig. 7! that the optical scattering in the off state is reducas the droplet size becomes significantly larger thanwavelength of visible light. The peak in the luminance ocurs for samples with composition around 76%–78% sgesting that the optimum drop size for back scatteringvisible light is around 0.860.2 mm, in agreement with pre-vious findings.12 As expected, lower UV cure powers resuin a decrease in the luminance due to the larger drop siz

The droplet radiusR remains roughly constant as a funtion of cure temperature~Fig. 6! so the dependence of th

Carter et al.


FIG. 5. Dependence of the PDLC morphology on UV cure power showing the increase is droplet size with decreasing UV cure power.~a! Changes inmorphology for 80% composition and 24 °C cure temperature, with UV cure powers from top to bottom of 16, 2, and 0.02 mW/cm2. ~b! Changes inmorphology for 82% composition and 24 °C cure temperature, with UV cure powers of 16 mW/cm2 ~top! and 0.02 mW/cm2 ~bottom!. Bottom slice is a crosssection of the cell showing that the droplet size is limited by the cell gap dimensions.

thlip

o ence

switching voltage, switching speed, and luminance ontemperature is determined primarily by changes in the elticity L due to the changes in the onset of coalescenceTth . InFig. 8 we graph the electro-optic properties as a function



e-

f

temperature for 80% composition cured at 16 mW/cm2, forwhichTth starts around 30 °C. The changes int andV occurslightly beforeTth , which may be due to a resolution limit inour optical observation of the coalescence. Since coalesc

5997Carter et al.


lei

th

ee

verherropsAl-de-

ploma

erm

s-

s

leads to an increase in the elliptical nature of the drop~Fig. 6!, L increases resulting in the expected increaseVon predicted by Eq.~1!, and the expected decrease intoff ,predicted by Eq.~2!. Von remains constant belowTth as ex-pected. The mechanism for the weak decrease oftoff belowthe peak is unknown, but could be due to changes inelastic constant or anchoring energy with temperature.13 Thefastest speeds are observed for small coalesced drops~smallR, largeL! as predicted by Eq.~2!. The observed peak in thluminance has been attributed previously to an increas

FIG. 6. From top to bottom, plots show the dependence of the LC drodiameter on cell depth, composition, UV cure temperature for different cpositions, and light intensity for a fixed composition and temperaturecure. The bottom graph shows the dependence of the moment of in~ellipticity of the droplets! on the UV cure temperature which results frothe coalescence of droplets for two different compositions.



tn

e

in

the backscattering due to changes in the drop size;12 how-ever, we observe negligible changes in the droplet size othis temperature range. The falloff in the luminance at higtemperatures could be due to the coalescence of the dwhich leads to less efficient scattering between droplets.ternatively, there must be some additional temperature

et-ttia

FIG. 7. Summary of the dependence of the electro-optic propertiesVon ,switching voltage;toff , the response time for switching between the tranparent and scattering states; and LUMoff , the luminance of the scatteringstate, on the LC droplet radius.

FIG. 8. Dependence ofVon , toff , and LUMoff on the UV cure temperaturefor 80% LC and 16 mW/cm2 UV cure power. The change in propertiearound 30 °C marks the threshold temperatureTth , at which droplet coales-cence starts to occur.

Carter et al.


ahe

opmge

ly-pring

cklur

polzeeiz.

R.

nd

hi

W.

ato,

pendent scattering, possibly due to changes in the LCchoring or dielectric constant, which determines tluminance.

CONCLUSIONS

We have studied the morphology and electro-optic prerties as a function of varying composition, UV cure teperature, and UV cure power. We find that for the spontype PDLC ~;80% LC fraction!, the morphology isdetermined primarily be the rate of solidification of the pomer matrix and the viscosity of this matrix once phase seration has been initiated, with the latter property being pmarily determined by the curing temperature. If this curitemperature is below the temperatureTth , then the morphol-ogy is determined by the nucleation mechanism and molelar diffusion, and the electro-optic properties are only weaaffected by the changes in morphology. However, if the ctemperature is increased aboveTth , the LC domains start tocoalesce resulting in substantial changes in the electro-oproperties. These results suggest that for low switching vages and high luminance, PDLC cells should be polymeriunder conditions which are below the coalescence tempture. The fastest switching times are obtained by polymertion at temperatures above the coalescence temperature



n-

---

a--

u-ye

tict-dra-a-

ACKNOWLEDGMENT

We acknowledge very useful discussions with K.Amundson.

1J. L. Fergason, Soc. Information Display Digest Tech. Papers16, 68~1985!.

2J. W. Doane, N. A. Vaz, and B. G. Wu, Appl. Phys. Lett.48, 269~1986!.3P. S. Drzaic, J. Appl. Phys.60, 2142~1986!.4R. L. Sutherland, V. P. Tondiglia, L. V. Natarajan, T. J. Bunning, aW.-W. Adams, Appl. Phys. Lett.64, 1074~1994!.

5T. Hirai, S. Niiyama, H. Kumai and T. Gunjima, Rep. Res. Lab. AsaGlass Co.40, 285 ~1990!.

6K. R. Amundson, A. van Blaaderen, and P. Wiltzius, Phys. Rev. E55,1646 ~1997!.

7J. D. LeGrange, S. A. Carter, M. Fuentes, J. Boo, A. E. Freeny, andCleveland, and T. M. Miller, J. Appl. Phys.81, 5984~1997!.

8This result is consistent with the findings of R. Yamaguchi and S. SJpn. J. Phys.33, 4007~1994!.

9A. J. Lovinger, K. R. Amundson, and D. D. Davis, Chem. Mater.6, 1726~1994!.

10B.-G. Wu, J. H. Erdmann, and J. W. Doane, Liq. Cryst.5, 1453~1989!.11P. S. Drzaic and A. Muller, Liq. Cryst.5, 1467~1989!.12P. Wiltzius, K. R. Amundson, and P.-Y. A. Wong~unpublished results,1994!.

13C. A. McFarland, J. L. Koenig, and J. L. West, Appl. Spectrosc.47, 598~1993!.

5999Carter et al.


Dependence of the morphology of polymer...

Documents

Transcript of Dependence of the morphology of polymer...