`Linearly Polarized UV Exposure
`M. Nishikawa*, B. Taheri, and J. L. West
`Kent State University, Liquid Crystal Institute, Kent, Ohio, 44242, USA
`*JSR Co., Yokkaichi Research Laboratories, 100, Kawajiri-cho, Yokkaichi, Mie, 510-0871, Japan
`
`Abstract
`We designed polyimide films for LC alignment
`using polarized UV exposure. Unidirectional LC
`alignment with any desired pretilt angles from 0 up to
`90 ° were obtained using a single oblique polarized UV
`exposure. Results suggest that photo-decomposition of
`polyimide produces anisotropic van der Waals forces
`which aligns the LC.
`
`Introduction
`LCDs are widely used because of their low power
`consumption, thin profile, and full color capability. To
`fabricate LCDs, unidirectional LC alignment with
`controlled pretilt angle is typically produced using surface
`rubbing techniques [1]. However, rubbing may cause
`static charge, dust, or scratches which
`lowers
`the
`production yield of LCDs.
` LC alignment using polarized UV irradiation [2] is a
`strong candidate for overcoming the above problems.
`However, generation of pretilt angle using this technique
`is not easy because conventional photo-alignment
`materials, poly(vinyl cinnamate) [3] and polyimides (PIs)
`[4-7], result
`in LC alignment perpendicular to
`the
`polarized UV axis. Some methods have been proposed to
`generate pretilt angles: (1) double polarized UV exposure
`with different polarization and irradiation angles [6-8], (2)
`combined polarized UV exposure with both p-wave and s-
`wave [9], (3) oblique polarized UV exposure on
`poly(coumarin) [10,11]. Double UV exposure is too
`complicated for mass production, and combined UV
`exposure with both p- and s-wave decreases
`the
`polarization efficiency of the polarizer. Poly(coumarin)
`aligns the LC parallel to the polarized UV direction and
`results in any desired pretilt angles by adjusting oblique
`irradiation angles. However, the long term reliability of
`LCDs with UV cured poly(coumarin) alignment film is
`uncertain due to the unreacted coumarin group which is
`thermally unstable. In addition, the glass transition
`temperature of poly(coumarin) is lower than that of the PI
`used for conventional LCD application, also making it
`less thermally stable.
`In this paper, we present results on newly synthesized
`PIs which produce unidirectional LC alignment and
`controllable pretilt angles using a single oblique polarized
`UV
`exposure
`technique.
` The mechanism
`and
`characteristics of the LC alignment are also discussed with
`respect to the chemical structures of the PIs used.
`
`Results and Discussion
`1. Design of Polyimides
`Polyimides which align LC parallel to the polarized
`UV axis are desirable since pretilt angles can be obtained
`with only a single exposure.
` We first obtained
`homogeneous LC alignment on various PI
`films
`containing a cyclobutane moiety such as tetracarboxylic
`dianhydride, Fig. 1. These PI films photo-decompose
`upon UV exposure [12]. The LC alignment produced by
`various PIs cured at 250 C for 1 hr and either rubbed or
`normally exposed with polarized UV are summarized in
`Table 1. For UV exposure, we used a 450 W-Xe lamp as
`the UV source, and a surface film polarizer (ORIEL,
`model 27320) whose effective range is between 230 nm to
`770 nm. In most cases, PI shows LC alignment parallel to
`the rubbing direction and LC alignment perpendicular to
`the UV polarization. This is consistent with results
`reported previously [4-7]. PI-4, however, aligns LC
`perpendicular to the rubbing direction and parallel to the
`UV polarization.
`Figure 2 shows the relationship between UV exposure
`time and the dichroic ratio of the LC cell. LC cells for
`dichroic ratio measurement were fabricated using two
`polarized UV exposed substrates with parallel polarization
`axis. Dichroic LC, ZLI-2293 + 0.5% M-618 (l max=550
`nm), was filled into LC cells, and the dichroic ratio was
`measured using one polarizer and UV-Vis analyzer. The
`dichroic ratios of
`the LC cells
`initially
`increase
`logarithmically with UV dosage, and approach a constant
`value at higher dosage. To elucidate the LC alignment
`
`N
`
`O
`
`O
`
`O
`
`O
`
`N
`
`R
`
`H
`
`CH2
`PI-2
`
`H
`
`R
`
`O
`PI-1
`
`PI-3
`
`PI-4
`Fig 1 Chemical structures of PIs
`
`Table 1 LC alignment on PIs
`LC Alignment Direction
`
`- R -
`
`PI-1
`PI-2
`PI-3
`PI-4
`
`Rubbing
`
`Polarized UV
`
`//
`//
`//
`
`^^^^
`
`//
`
`ISSN0098-0966X/98/2901-$1.00 + .00 (c) 1998 SID
`
`
`
`
`
`
`Page 1 of 4
`
`Tianma Exhibit 1036
`
`
`
`ratioofLCcell
`Dichroic
`
`s
`2 1
`102050100
`.1135!
`Polan'mdUVeIposuetirne(rnin)
`
`F‘ Zltelatronshrp‘' betweenUVarposuretimeanrl
`lg
`dichroicrztmsot'I£oc1lsonvariousl'Itilus
`
`mechanism, we chose two PI materials which give DC
`alignment perpendicular (PI-1) and parallel (PI-4) to the
`Uvpolarizatim
`UV absorptim spectral changes in PI films before and
`alter pohrized UV exposure were monitored. PI films on
`qmrtz substrates were prepared and exposed for 2 hrs
`with measured norrnalto the sur'face_ Figure 3 shows UV
`absorptbn spectra of PI fikns before and alter pohrized
`UV ecposure. Afier UV exposure, both PI filns showed
`decreases in the broad absorptim around 250 nm, whkh
`can be attrbuted to 1I:—1I:“ transitions of benzene rings in
`P], and ircreases in the broad absorption above 290 nm
`We have not yet determined the photochemical changes in
`P1 filns, but it is clear that the broad absorption above
`290 nm are generated by the decomposition of PI whkh
`corresponds to the broad absorptbn arormd 250 nm
`These phenomena have been previausly reported on PI
`material wih ditferent chernital structure [4,5].
`Figures 4 and 5 show the dizhrok UV absorptim
`spectra of rubbed PI films and pohrized UV exposed PI
`filrrs measured parallel (Apm) and perpendcuhr (A...) to
`the rubbing direction and the exposed UV pohrizatim,
`respectively. Dichrot UV absorbance was measured
`using a surface film polarizer. In the case of rubbiig (Fig
`4), PI-1 shows a positive dichroi: spectrum, and PI-4
`shows a negative dichroic spectrum. The behavbrs using
`pohrized UV exposure '5 opposite to those produced by
`rubbing (Fig. 5). Furthermore, it shoukl be noted that the
`
`Absorbancc(au)
`
`2002flI2A02(01H3tH320.H03‘O3H-III
`W-velmsfi (IIIII)
`Fig 3 UVabsorptionspec1raofPIfilrnshefore
`andaflerUVurposure
`
`0.04
`
`.6B
`
`DtdirolcUVdnorbmee(nu) 6e.23
`
`‘u'.§OO220MO2‘tI1N3I0330.N03flI3fi-COO
`Wavelengthflun)
`Fig 4 DidrroicUVabsorptionspectra(Apan-Aper)o
`rubbedPIfiln1smeasuredrelativetottierubbing
`direction
`on
`
`DichmicUVtsorbmoe(Iu)
`
`3
`
`znzzonozuullwoszouosasuwo
`WIvelargth(mn)
`Fig 5 DichroicUVabsorptionspectr'a(Apara-Aper)ot
`UVexposued_PIfilmsmnsuredrelativetothe
`exposedpolarrzaIronofUV
`
`subtractim spectra of UV absorptbn spectra above 290
`nm show no dichroism (Fig.
`5).
`Birefringence
`measurement of PI films afier linearly polarized UV
`exposure ako showed that PI-1 had the opti: axis
`perpend'mulartotheUVpolarization, andPI—4hadthe
`parallel opfr axis.
`Changes in PI films before and after polarized UV
`exposure were also monitored rising FT-IR PI films on
`silicone wafers were normally exposed to pohrized UV
`for 2 hr.
`In both PI films, peak intensities at 1380 cmil
`attributed to V(irnide C-N-C) decrease and those at 1720
`cm’! attributed to v(C=0) increase afier UV exposure.
`These results suggest
`that
`imiie rings
`in PIs were
`decomposed by UV exposure‘ and a new band with IR
`absorption at around 1720 cm" was generated by the by-
`product afier the decomposiim of PI
`We calculated the conformafnns of PI-l and PI-4
`
`using molecular mechanks method (Fig. 6) [13]. The
`
`‘** %£;";_r§'~“&
`
`sans. .¢L......\&.»—~
`(b)H—4
`(I)PI-I
`Fig 6 Pleorrfirruntialurlamduéngnnlecularnmdraniamethod
`
`Page 2 of 4
`
`|SSN0098-0966XI98/2901—$1.00 + .00 (c) 1998 SID
`
`
`
`90
`II
`C 70
`
`3‘
`
`.5 C0
`.1:
`
`gso
`=‘=‘- an
`3 .0
`III
`III
`
`I
`20
`-M
`60
`U
`100
`MolarfiactiorrofF4ii:nirreX(molar%)
`of
`‘I:
`r§;".':’E.'.’.T'3fi5aa.§s."?.‘§..:
`ar'posuretum;40mrn)
`
`10
`
`F‘
`'9
`
`lat
`°‘p'i'('.,";".’sso,
`
`3. Behavior: ofPretiltxinglm
`Figure 11 shows the relatbnship between polarized
`UV exposure time and pretilt angles of 112 cells at
`difierent exposure angles. Ptetilt angles
`increase
`with pohrized UV exposure angle (0) and show
`xinmm values for exposure time. Figure 12 shows the
`rehtbnshj) between pohrized UV exposure angle and
`pretilt angls of IE cells. Pretilt anghs increase with
`pohrized UV exposure angle (0).
`
`ofLC(')Iv){N
`
`5
`
`Pruiltande
`
`O
`
`80
`CO
`40
`20
`PolariudUVelposuretirne(min)
`
`100
`
` E E
`
` 0.0
`
`10
`
`60
`50
`-IO
`30
`20
`PoIl'medUVuposIletime(rnirr)
`Fig 8 RelIim§ripbdweenpolarizniUVqpos|le
`tirneand1IetiltanghsforwrinuUV
`exposureargls
`
`10
`
`optmal ax's of PI-1 is aligned along the polymer
`backbone. On the otha hand, the fluorene unit in PI-4
`which has large biefringence is aligned perpendicular to
`the axis ofthe in chain
`
`2. Genemtion ofPretilt Anglw
`PI-4 was irradiated withpolarized UV at a rmmber of
`incident angles (9) as shown in Fig 7. Figure 8 shows the
`rehtionshp between pohrized UV exposure time and
`pretilt gles of the LC for various iradhtirn angles.
`Pretilt angles increase with UV exposure time and/or with
`increase of the UV exposure angle (0). However, the
`pretilt angbs obtained are still bebw 1 °. High pretilt
`angles over 1 ° areneededto preveri IC alignment from
`reverse tilt disclinatiors during volage applkation
`To increase the pretilt angles, we synthesized Pls with
`fhrorine atoms by the introduction of fluorine containing
`diimine (F—d'nmine) as shown ii Fig 9.
`It has already
`been reported that F-diamine shows high pretilt [13].
`Figure 10 shows
`the relationshj) between chemical
`structures of PB and pretilt angles of III. The pretilt
`angles of DC gadually increase with the mole fiactbn of
`F-dinmine in PI. We chose the polyimide containing 30
`molar % of Fdhrnine (PI—30F) to analyze the behaviors
`more in detail
`
`Fig 11
`
`enposrgetrrnearllrletillanglcsofll-3tl-‘
`fi1rvziousUVexposueanglcs
`
`5I0
`
`5 I
`
`n
`
`
`
`PmihangleofLC(°)
`
`<1(7)
`Fig 9 Chemicalst1uctureofPIwithflnorineatorrs
`
`0
`
`so
`40
`2o_
`Pohnnduvuposnre algae (’)
`
`so
`
`larizedUV
`Fig 12 Relatbdrip between
`e ofPI-30F
`angle and p_ret:rlt
`(exposure time ; 20 min)
`
`Page 3 of 4
`
`|SSN0098-0966XI98/2901—$1.00 + .00 (c) 1998 SID
`
`
`
`100.0
`995
`
`g 99.0
`.9 935
`
`E 5
`
`'
`2- 97.5
`g, 97.0
`‘E 96.596.0
`955
`95.0
`
`0
`
`
`
`6.
`-I0
`20
`PolIn'zdUVuposuretirr:(rnin)
`
`8.
`
`difihuitsurfaee
`Fig 15 VHRsofI£oel.Is
`;zLI-4792,
`treatedPI_sIbstntes
`fiarmpenod; l67msec)
`
`'
`
`lowest value probably due to the surfiice contamination by
`fi'om ribbing cbth VI-IRS of DC cells using
`UV exposed PI in nitrogen show higher VHR than that of
`UV exposed PI in air, which E the same as those using
`conventimal PI materhls for active trix use [15].
`
`Conclusion
`
`IC alignment
`We successfully obtained
`withanydesiredpretiltangles l'rom0upto90°usinga
`single obli1ue linearly pohrized UV exposure.
`PI
`materials also were thermally stable up to 200 °C and
`excellent VI-IR, comparable to conventional PIs for active
`matrix use. These PI films will not only eliminate the
`problems caused by rubbing but also greatly simplify the
`productim of rnuli-doin, wide viewing angle DCDS.
`
`Acknowledgements
`We acknowledge Drs. L C. Chien. T. Kosa. and X. D.
`Wang of Kent State University for terial synthesis.
`birefringence measurenrnt. and useful discussion. We
`also tlnnk Drs. N. Bessho and Y. Matsuki of JSR Co..
`Ltd. for their support in this research. Research was
`supported in part by NSF Science and Technology
`Center AICOM. DMR 89-20147.
`
`References
`[l] M Nishikawa et al, Proc. Japan Display '92, p.819 (1992).
`[2] A. Dyadyusha et. al, Ukrainian Fiz. Z1 , 36, 1059 (1991).
`[3] M Schadtwt. :1, Jpn. J. Appl. Phys , 31, p.21 55 (1992).
`[4] M. Haseywa et al., Proc. HJRC '94, 1:213 (1994).
`[5] J. L. West etal, Proc. sm '95, p.703 (1995).
`[6] X Wang ct a.l , Proc. SH) '96, p.654 (1996).
`[7] x Wang et al, Proc. mac '97, p5 (1997).
`[8] Y. Iimura ct a1,.I. Photapalym. Sci. Tech , 8, 257 (1995).
`[9] K. Y. Han et al, Proc. SID '97, p.707 (1997).
`[10] M Schadt et :1, Nature, 331, 11.212 (1996).
`[11] M Schadt et :1, Proc. $11) ’97, p.397 (1991).
`[12] J. A. Movie et :1, Chem. Man, l,p.l63 (1989).
`[13] M. Nishikawa er al Mol. Cryst. Liq. cryst, 275, p. 15
`(1995).
`[14] M Nishkawa 6 al, Mol. Cryst. Liq. Cryst, 259, p.93
`(1995).
`[15] M Nishamva 3 al, Jpn. J. Appl. Phys., 33, 15.1.1113
`(1994).
`
`4. Heat Stability ofLC alignmatt
`fior heat
`To check the stability of IE alignment
`treatment, IE cells were fabricated using PI-30F annealed
`at 50 — 250 °C afier polarized UV exposure. Resuls of
`dithroh ratbs and pretilt angles ofIJC cells are shown in
`Fi§. 13 and 14, respectively. Results of di:hro'r ratio
`and pretilt angk measurement suggest that unidirectional
`IE alignment is stable up to 200 °C. However, IE cells
`annealed at 250 °C showed decrease of dichroi: ratio and
`
`increase of pretilt angle probably due to the relaxation of
`PI chain by annealing.
`
`5. Wm ofLC cells
`The values of voltage holding ratio (VHR) were
`measured using a measurement unit (Ekicon Inc., model
`VI-IR-100) where the pulse voltage of 5 V height and 60
`[sec duration was
`applied to LC cells with
`fluorinated III (ZLI-4792).
`In this experiment, we
`prepared three types of LC cell using PI coated substrates
`with ditferem surface trefiments as bellows.
`
`a) Rubbed PI substrates wihout washiig
`b) UV exposed PI substrates in air
`c) UV exposed PI substrates in nitrogen
`
`Figure 15 shows the dependence ofVI-IRS using treated PI
`substrates. VI-IRS ofI£ celk using rubbed PI shows the
`
`ratio(%)
`Holdingratioofdichmic
`
`
`0
`
`:so
`in
`{so
`18.
`in
`Hattreatrmnttnnpaature(°Cx1hr)
`
`Fig 13 Dichroicratiodiangeoflflcdlsahr
`heattreatmmt
`
`angleofLC(°) I‘{H
`Pletilt
`
`I
`
`250
`:00
`150
`100
`so
`I-Ieattreatrmnt ternperaturc (°Cx 1 hr)
`
`Fig 14 Preiilranglcdnngeofutocllsafin
`hcartmmcm
`
`Page 4 of 4
`
`|SSN0098-0966XI98/2901—$1.00 + .00 (c) 1998 SID