Ferroelectrics, 330:93–101, 2006 Copyright ? Taylor & Francis Group, LLC ISSN: 0015-0193 print / 1563-5112 online DOI: 10.1080/00150190600605676
Surface Electroclinic Effect in Low PS Smectic Materials Without SmA Phase
ARTUR ADAMSKI, KRISTIAAN NEYTS,? AND HERMAN PAUWELS
LCD Research Group, Electronics and Information Systems Department, Ghent University, Sint-Pietersnieuwsstraat 41, B-9000 Gent, Belgium
The molecular axis of a chiral smectic material deviates away from a desired direction induced by the rubbing process. This small angular deviation is caused by the Surface Electroclinic Effect and it is usually studied near the phase transition N-SmA. In the material without SmA phase thus after N-SmC transition the phenomenon looks quite similar except the fact that the deviation of a molecular axis occurs in two opposite directions. Keywords Liquid crystals; smectic C; surface electroclinic effect; LCD
The liquid crystal alignment is very important in order to have a good electrooptical performance of the display. It is widely known that LC molecules orient themselves in a desired direction because of the anisotropy of the surface forces induced by the surface orientation process. Several methods have been already employed. The most common techniques are: polyimide or polyamide rubbed substrates, surfactant coated surfaces, obliquely evaporated SiOX ?lms, photo-induced alignment etc. All these methods align the long optical axes of molecules parallel to the substrate in a certain direction however the forces which govern the induced alignment are still not well understood. In a nematic LC where only the orientational order is present the molecules tend to align easily along the desired direction, however in smectics where also the positional order is present the averaged orientation shows a small deviation of the molecular director away from the desired direction. Experimentally, ellipsometry and some other linear optical methods have been used to study the molecular arrangement inside the cell but only the recent optical second harmonic generation technique gives a good understanding for this phenomenon . This small angular deviation is caused by the so called Surface Electroclinic Effect  and has been studied elsewhere [3, 4] usually near the phase transition between the N and SmA. We have investigated this effect in a material without the SmA phase [5, 6] thus after the N-SmC transition. In this case the orientation in two domains tilt in two opposite directions thus deviation of the optical axis occurs at two different angles. The magnitude of this effect is different in different test cells. Therefore the surface interactions responsible for the SEC effect can be
Received November 20, 2003; in ?nal form August 1, 2005. ? Corresponding author. E-mail: firstname.lastname@example.org
A. Adamski et al. Table 1 FLC materials and their properties
LC material LZ1 LZ2 LZ3
Ps [nCcm?2 ] 3.9 4.5 4.0
Phase sequence temperature [? C] Cr–(-25)–SmC? –69.7–N? –102.5–I Cr–(-40)–SmC? –67.9–N? –89–I Cr–2–SmC? –66–N? –85–I
Tilt angle [? ] 22.5 22.5 22.5
varied in a controlled way in order to ?nd some relations between the cell parameters and the electrooptical performance.
2. Experiments and Results
The test cells are produced in the clean-room facilities of Ghent University. During the technological process special attention was paid to reproducibility. All the test cells have a gap of 1.6 um with pixel area of 2 cm2 . The cells are assembled in an antiparallel rubbing con?guration, buffed in the direction of the pixel edge and have an alignment layer thickness of 50 nm. A standard rubbing process is employed to induce the surface anisotropy. The polyimide used for an alignment layer is provided by JSR (AL-1254). Three experimental FLC samples (LZ type) are provided by CLARIANT GmbH, Germany. The materials feature small spontaneous polarization value and the lack of the SmA phase. Table 1 presents selected parameters of the investigated materials. 2.1 Multidomain Structure The lack of the SmA phase during the phase transition leads to very dif?cult alignment with the standard methods. This feature however gives also a bene?t namely the absence of zig-zag defects and thus good electrooptical performance. The standard alignment method gives rise to the creation of a so-called multi-domain structure in which the molecules are aligned along the rubbing direction on two opposite sides of the smectic cone (and PS points in opposite directions). In principle in the SmA phase the smectic layer normal (LN) coincides with the rubbing direction (R) and the molecules tilt within the ?xed layers in a latter SmC phase. However for the materials without SmA phase the creation of layers and the tilting has to be done at the same time during transition between N and SmC. Therefore the smectic layer normal will be inclined with respect to the rubbing direction over an angle which is in fact the tilt angle of the director. The layer inclination can occur in two symmetric directions because of the same probability of an alignment on two sides of the smectic cone. Thus the developed structure consists of two symmetric domains in which the average net polarization points in opposite directions (Fig. 1). During the application of an electric ?eld these two domains will react in a different way thus the response of the pixel will be different for positive and negative voltage (Fig. 2). This multidomain structure can be reduced to a monodomain with the dc-?eld alignment technique . The lack of the SmA phase and a speci?c con?guration of the cell (monodomain) quali?es the above FLC mixtures to operate with a low threshold and continuous switching mode (half-V-shape with no hysteresis) with very small saturation voltages. This operating mode, called Continuous Director Rotation (CDR-mode) is described elsewhere
Surface Electroclinic Effects
Figure 1. Smectic layer inclination during the phase transition.
[5–7] and will not be discussed here. It suf?ces to say that the aligning ?eld  orients the macroscopic polarization along the ?eld lines whereas the molecules orient themselves parallel to the rubbing direction. It results in a stable monodomain structure, where the molecules are aligned on one side of the smectic cone. The second position on the cone (which would be the other stable position in SSFLC) is de-stabilized, due to the fact that a large deviation from the rubbing direction leads to an increase of the potential energy. Application of a ?eld opposite to the aligning ?eld rotates the molecules gradually, leading to gray levels in the transmission. An electric ?eld applied in the same direction as the aligning ?eld does not affect the position of the molecules, resulting in an asymmetric characteristic of the device—half V-shape mode. In conventional (A)FLC materials hysteresis can also vanish at certain driving frequencies leading to the V-shape modes [8–12]. However the CDR-mode is working in a very different con?guration compared to classical SS(A)FLC, which makes it less sensitive to the operation frequency. 2.2 SEC Effect Ideally, in conventional smectic materials, the layer normal in the smectic phase is parallel to rubbing direction R, resulting in a planar or bookshelf geometry. However, it has been found that already in the SmA phase (the ?rst smectic phase to occur) the optic axis is slightly rotated with respect to the rubbing direction . Further detailed studies revealed that the molecules remain anchored at the surface along the rubbing direction while only the optic axis of the bulk rotates layer by layer progressively as one moves into the cell’s interior. This rotation ?nally saturates and stays at ?xed position in the bulk . Such an effect occurs during the phase transition (N-SmC), when the molecules at the interface start to create a mono-layer, which is formed due to the polar interactions between the
Figure 2. The microscopic texture of the multidomain structure and its response for an external electrical ?eld.
A. Adamski et al.
Figure 3. Deviation of the optical axis away from the rubbing direction (R) caused by the Surface Electroclinic Effect.
liquid crystal molecule and a solid alignment ?lm. The arrangement of the molecules in a monolayer induces a certain direction of the smectic layer normal, which does not match the direction of the layer normal created in the bulk. In order to match them a slight twist of the smectic layers occurs between the bulk and the interface. This effect has been attributed to a polar interaction between the LC molecules and the interface and received the name Surface Electroclinic Effect (SEC effect) [2–4]. The multidomain cell (?lled with the material without SmA phase) viewed between two crossed polarizers does not show a perfect extinction when aligned along the rubbing direction (Fig. 3, middle picture). After some rotation of the sample over a small angle (α0 ) one can ?nd much better extinction for each domain respectively (Fig. 3, left and right picture). The rotation angle α0 is different for different domains and seems to be symmetric with respect to the rubbing direction. This deviation of the averaged optical axis of molecules away from the rubbing direction is caused by the surface electroclinic effect  described above. Analysis of Fig. 3 indicates that starting from the rubbing direction (Fig. 4a) towards the bulk the molecular director twists together with smectic layers over a small angle until it reaches a ?xed position (Fig. 4b). Figure 4b compares the orientation of smectic layers in the bulk (solid line) and at the interface (dashed line). The thickness of a twisted layer is much thinner than the wavelength of the visible light therefore a good extinction is observed for a rotated cell (Fig. 4c and Fig. 3, optimum dark). In  it has been shown that the spontaneous polarization at the interface points into the surface, thus in the material without SmA phase, two domains with opposite polarizations develop in one cell. In this case the rotation of a molecular director occurs in two opposite directions (Fig. 3) because of the opposite orientations of the mono-layers at both interfaces with respect to the bulk.
Figure 4. The orientation of an optic axis (OA) of the molecule with respect to the rubbing direction (R) and crossed polarizers (A,P) for one domain type (a) without the SEC effect or just at the interface (b) rotation of smectic layers caused by the SEC effect (corresponding to Fig. 3 middle) (c) rotation of the sample over a deviation angle α0 (corresponding to Fig. 3 left).
Surface Electroclinic Effects
Figure 5. Transmission as a function of the orientation of the rubbing direction for two single domains.
Figure 5 shows the transmission of two single domains versus position of the sample with respect to the rubbing direction. One observes the minimum not along the rubbing direction but shifted over the deviation angle α0 . The minimum shifts symmetrically in two opposite directions for two different domains. The maximum transmission is reached for 2· ± α0 which is in this case 45? ± 3? .
2.3 Different Alignment Conditions Unidirectional rubbing of the cell surface allows for aligning the long axes of molecules along the rubbing direction. The molecular arrangement is controlled by the anchoring forces between the LC layer and the alignment surface. The non-polar forces can be varied technologically in order to change the molecular alignment of LC layer at various surfaces. We noticed that the deviation angle is different in different test cells and strongly depends on the surface conditions. A number of test cells with different parameters has been prepared in order to ?nd some relations between the surface interactions and the angular deviation α0 . The quality of the obtained alignment can be observed under the microscope within the dark state with the high illumination (Fig. 6). Due to the different surface interactions one observes different molecular order in particular cells. Cell 1–1 shows very inhomogeneous alignment with a large deviation of the molecular directors in different domains. The alignment of the cell 1–4 looks quite homogeneous, however additional texture in a larger scale is observed caused by a strong surface treatment which worsens the uniformity. The middle cell 1–3 shows the best molecular orientation. All these cells have a different displacement of the optimal dark state for particular domains. Note that without the SEC effect the shift does not occur and the darkest state appears along the rubbing direction. This displacement has been measured carefully and
Figure 6. Textures of cells with different surface interactions (no voltage, high illumination).
A. Adamski et al.
Figure 7. Transmission of a single domain for two cells with very different surface interactions, as a function of the orientation of the rubbing direction.
seems to be the largest in the cell 1–1 and decreases progressively for the cells 1–2, 1–3 and 1–4 respectively. Figure 7 shows transmission measurements for single domains in two extreme cells (1–1 and 1–4) as a function of the angle between the rubbing direction and the ?xed polarizers. The minimum for the cell 1–1 appears at about 4.0? whereas for the cell 1–4 at 2.9? . The cells 1–2 and 1–3 possess minima at 3.6? and 3.3? respectively. Figure 8 presents the summarized angular deviations, this is the optimal dark state angles for the cells with different surface interactions and different LC materials. One observes the tendency of decreasing the deviation with increasing the surface interactions ?nally going to saturation for very strong surface treatment. The angular deviation seems to be larger for the materials with higher spontaneous polarization value in identical cells, which is logic because a higher PS induces a stronger ?eld at the surfaces which afterwards causes higher tilt. It explains the angular deviations as large as 18? (recent measurements in ) for high PS materials which can not be neglected during the design of a display. The angular deviations have been measured also in the specially prepared test cells with asymmetric surface interactions. The parameters of these cells are described in . These cells develop more domains of one type because of the asymmetry of the nucleation centers, thus the SEC effect exhibits asymmetric behavior as well. The displacement of an optimal dark state for a domain which dominates in the cell (domain attracted to the stronger surface) is higher than for the domain attracted to the weaker surface. This tendency is observed in all asymmetric cells. The effect is more enhanced for stronger differences between the surfaces.
Figure 8. Optimal dark state angles for the cells with different surface interactions and different PS .
Surface Electroclinic Effects
Figure 9. Transmission-voltage characteristic for two different single domains.
2.4 Orientation of Optic Axis The application of an external electrical ?eld to the multidomain structure gives a response for both voltage polarities (Fig. 2). We measured the transmission-voltage characteristic of two separate domains and found a quite symmetric response in the symmetric cells (Fig. 9). Domain A responds only for positive voltages whereas domain B only for negative ones. This characteristic has been measured when the sample (cell 1–3) was aligned exactly along the rubbing direction. Figure 9 shows there is no perfect extinction for 0V; this is caused by the SEC effect which is symmetric for both domains. This low transmission remains unchanged with the external ?eld directed along the averaged spontaneous polarization. On the other hand, when the ?eld is directed opposite to the averaged PS , the domain reacts for the applied voltage. That is why the response is asymmetric for a single domain. Note that the change of the transmission is continuous (V-shape like) however for small voltages the transmission decreases ?rst going to some minimum (for V = ±1 V in this case) and increases afterwards with higher voltages. This behavior is again related with the SEC effect and depends on the angular deviations. The minimum shifts towards smaller voltages for the cell 1–4 and towards higher voltage for the cell 1–1. It means that the cells with stronger surface treatment and smaller deviations of molecular orientation, need a smaller voltage to obtain the optimum dark state along the rubbing direction. This tendency is also visible for the asymmetric cells when the optimum dark state is found for asymmetric voltages as well and the minimum shifts according to the same rules as previously. This measurement allows us ?nally to ?nd out the averaged molecular orientation in the bulk for a particular domain. Fig. 10 presents schematically the position of the molecule
Figure 10. The positions of the molecule director with respect to the rubbing direction (R) in two separate domains.
A. Adamski et al.
and its polarization (P) with respect to the rubbing direction (R) between two crossed polarizers (A,P). The optical axis (OA) of molecules in domain A tilts right away from the rubbing direction over an α0 angle whereas the optical axis of the molecules in the domain B tilts left away from R over an opposite α0 angle. Considering only domain B we ?nd some residual birefringence and thus some light leakage when the cell is aligned along the rubbing direction (Fig. 3 middle). The right rotation of the cell over an angle α0 aligns the molecular director along the polarizer and good extinction is observed (Fig. 4c, Fig. 3 left). Positive voltages do not change the orientation of domain B because the ?eld is directed downwards along the averaged polarization. Small negative voltages slightly rotate the director on the cone, resulting ?rst in a better extinction and at higher voltages in a switching towards the bright state (Fig. 9). The behavior of domain A is found by a similar explanation. The T-V measurements for this structure show a ?at line for small voltages and a continuous response for higher ?elds, which could be interpreted as a small threshold value. We do not ?nd here any threshold, except for the cells with high surface treatments .
3. General Conclusions
The Surface Electroclinic Effect causes the deviation of the molecular orientation away from a desired direction. This deviation strongly depends on the cell parameters as well as on the liquid crystal material, and it is lower for better aligned cells and smaller spontaneous polarization values. The deviation saturates for a high surface treatment and can not be lowered any more. The rotation of the optical axis of molecules caused by the SEC effect decreases the contrast ratio of the display and should not be neglected. First it introduces light leakage for the relaxed state so the extinction is not perfect and secondly the brightest state is never achieved because the molecule on the cone is rotated away from the 2· position. For multidomain textures the opposite deviation of the smectic layers due to SEC effect makes it dif?cult to improve the electrooptical performance of the display. Enhancement of the dark state for one domain worsens the performance of the other one and vice versa. In case of conventional materials with a SmA phase, as well as for monodomain structures, one just has to take into account the angular deviations during the cell fabrication process and apply skewed rubbing with an adapted angle to compensate the deviation.
The SEC effect is a quite well-known phenomenon usually investigated in high PS smectic materials near the phase transition between the N and SmA. We have reported this effect in low PS materials without SmA phase, thus after the transition N-SmC. The phenomenon looks similar except for the fact that the tilt of the smectic layers occurs in two opposite directions for different domains. We found the tilt to be symmetrical for symmetric cells and asymmetrical (one domain tilts larger than the other) for asymmetric cells. Thanks to accurate T-V measurements of a single domain we were able to ?nd the exact orientation of the molecular axis and the tilt with respect to the rubbing direction. Although the multidomain patterns presented in Fig. 2 can optically be observed for both conventional FLC materials and materials without the SmA phase, the arrangement of smectic layers, and therefore the performance of the display is very different. In case
Surface Electroclinic Effects
of ordinary FLC materials, the smectic layers are oriented perpendicular to the rubbing direction with the molecules tilted over an angle with respect to the layer normal and the rubbing direction. In case of materials without the SmA phase, the smectic layers are tilted over an angle 900 - away from the rubbing direction with the molecules aligned along the rubbing direction. The above results of the opposite tilt during the SEC effect are therefore obtainable only for materials without SmA phase. The in?uence of the SEC effect in low PS materials is relatively small, however it is present and observable. We would like to thank Dr. Hans-Rolf Dubal and his team for providing the LC materials. This work is supported by Belgian project “Photon” IUAP V/18; the EU project “SAMPA.”
1. W. Chen, M. B. Feller, and Y. R. Shen, “Investigation of anisotropic molecular orientation distributions of LC monolayers by optical SHG,” Ph. Rev. Lett 63, 24 (1989). 2. J. Xue and N. A. Clark, “Surface Electroclinic Effect in Chiral Smectic-A Liquid Crystals,” Rev. Lett. 64, 3 (1990). 3. M. S. Spector, S. K. Prasad, B. T. Weslowski, R. D. Kamien, J. V. Selinger, B. R. Ratna, and R. Shashidhar, “Chiral twisting of a smectic-A liquid crystal,” Ph. Rev. E 61, 4 (2000). 4. W. Chen, Y. Ouchi, T. Moses, and Y. R. Shen, “Surface electroclinic effect on the layer structure of a FLC,” Ph. Rev. Lett. 68, 10 (1992). 5. H. R. Dubal, T. Nonaka, J. Li, A. Ogawa, B. Hornung, W. Schmidt, and R. Wingen, “Material characteristics of an active matrix LCD based upon chiral smectics,” Liquid Crystals 26, 11, 1599 (1999). 6. Y. Asao, T. Togano, M. Terada, T. Moriyama, S. Nakamura, and J. Iba, “Novel FLC mode for AMLCD using cholesteric-chiral smectic C phase transition material,” Jpn. J. Appl. Phys. 38, 5977 (1999). 7. J. Patel and J. W. Goodby, Jpn. J. Appl. Phys. 59, 2355 (1986). 8. L. M. Blinov, E. P. Pozhidaev, F. V. Podgornov, S. A. Pikin, S. P. Palto, A. Sinha, A. Yasuda, S. Hashimoto, and W. Haase, Ph. Rev. E 66, 021701 (2002). 9. V. Panov, Yu. P. Panarin, O. Kalinovskaya, and J. K. Vij, Ferroelectrics 246, 35 (2000). 10. S. Inui, N. Iimura, T. Suzuki, H. Iwane, K. Miyachi, Y. Takanishi, and A. Fukuda, J. Mater. Chem. 6, 671 (1996). 11. S. S. Seomun, V. P. Panov, J. K. Vij, A. Fukuda, and J. M. Oton, Ph. Rev. E 64, 040701(R) (2001). 12. S. S. Seomun, T. Fukuda, A. Fukuda, J.-G. Yoo, Yu. P. Panarin, and J. K. Vij, J. Mater. Chem. 10, 2791 (2000). 13. K. Nakagawa, T. Shinomiya, M. Koden, K. Tsubota, T. Kuratate, Y. Ishii, F. Funada, M. Matsuura, and K. Awane, Ferroelectrics 85, 39 (1998). 14. W. Chen, Y. Ouchi, T. Moses, Y. R. Shen, and K. H. Yang, Phys. Rev. Lett. 68, 1547 (1992). 15. A. Adamski, K. Neyts, and H. Pauwels, “Controlled domain growth in low PS smectic material without SmA phase,” SID proceedings, IDRC’03, p.240 (2003). 16. J. S. Patel, Sin-Doo Lee, and J. W. Goodby, “Nature of Smectic Ordering at a Solid-Liquid Crystal Interface and its In?uence on Layer Growth,” Ph. Rev. Lett. 66, 14 (1991).