We report single-wall carbon nanotube (CNT) doped liquid crystal materials which show significant improvement in the response time for optical controlled birefringence (OCB) cells. Four different types of liquid crystals (LCs) were chosen to mix with CNTs and they demonstrated similar results in improving the response time. Experimental results show that the anchoring energy at alignment layers has been changed by CNT doping. CNTs were attracted to the alignment layer and modified the property of the surface. The anchoring enhancement is due to the π-π electron stacking between the CNTs, LC molecules and alignment layers.
©2008 Optical Society of America
In the demand for high quality motion pictures, among the available liquid crystal display technologies, the optical compensated birefringence (OCB) cell has been a promising switching mode. It has a wide viewing angle feature because the bent director orientation self-compensates the phase retardation when light enters from oblique angles. The advantage of the wide viewing angle arises from the bend structure where the director orientation self-compensates the phase retardation when light enters from oblique angles. Compared to other switching modes such as twisted nematic (TN), vertically aligned (VA) nematic and in-plane-switching (IPS) nematic modes, the OCB mode also excels in switching speed because of the reduced backflow effect . The major issue of OCB resides in the transition between splay and bend states in the OCB cell with a low pretilt angle. A holding voltage is required to prevent switching to the splay configuration; thus, the device can maintain high switching speed between the bend and homeotropic states. When high voltage is applied to the cell it can be switched to a homeotropic, in which most of liquid crystal molecules are aligned in the direction perpendicular to the cell. The response time for the “voltage-on” state is much faster than those of the most nematic devices. Yet for most nematic devices including the OCB cell, the response time for the “voltage-off” relaxation is the limiting factor which is related to the long image refreshing time, because no torque is applied to the local directors near the center of the cell. In the case of a nematic liquid crystal that is aligned parallel to the substrates, the switching mechanism resides in the fact that liquid crystal is the elastically deformed by the surface such that the liquid crystal is initially maintained at a bend state by a warm up voltage and transformed to a homeotropic state at high voltage. When the electric field is turned off, the fall time of the OCB cell, compared with the rise time is relatively long.
Since the discovery of carbon nanotubes (CNTs) by Iijima in 1991 , CNT has drawn a great deal of studies in fundamental and applied research because of its electronic and dielectric properties. Recently, several groups reported studies of carbon nanotubes doped nematic liquid crystals. Lee and Shih [3,4] demonstrated that in TN cells, carbon nanotubes doped nematic liquid crystals improve the rise time due to the lowering of the threshold voltage; however, the fall time was increased because of the increase in viscosity . Lee et al., shows improvement both in rise time and fall time in TN cells when the carbon nanotubes were doped with a concentration of 10-4% . They credit the improvement in both the rise and fall time with the reduction in viscosity. The application of CNTs presents challenge in obtaining reproducible results and improving the electro-optical properties of the host. From the reported results, uniform dispersion of CNT in LC is the key to explore new device applications.
Conventionally, the response time is considered with the assumption that the anchoring energy is infinite. It has been shown that the response times are sensitive to anchoring energy in those cells with intermediate to weak anchoring. In this paper, we use functionalized CNTs to enhance the switching time of the OCB cells by changing surface anchoring with CNTs. The dispersed CNT is functionalized or surfactant-treated CNTs as described in literature by Liu et al. .
2. Materials and cell preparation
The principle operation of the OCB cell with a low pretilt angle is illustrated in Fig. 1. At the non-activated state (a), liquid crystal molecules are elastically deformed at a splay configuration because of surface constraints such that the long axis of the nematic material is oriented in a direction parallel to the rubbing direction of both alignment layers. At a low applied voltage V1, the cell is switched to a bend state (b) in which the middle layer of liquid crystal molecules are perpendicular to the cell substrates while those molecules increase the angle of tilt away from the surface responding to the field. Some of the dispersed CNTs were attracted to the alignment layer because of the π-π electrons between CNTs and polyimide while others remain in the bulk and follow the direction of liquid crystal molecules throughout the cell.
Four liquid crystals, BL006, ZLI4792, MLC6080, and ZLI4792 were chosen to study the effect of doping single-wall CNTs in nematic liquid crystals. The characteristics of these liquid crystals are listed in Table 1. The doping concentration of CNTs in the host liquid crystal was 10-3%, and the mixture was put in ultra-sonic bath for 3hrs until the CNTs were well-dispersed in the host liquid crystal. The mixtures were filled in electro-optical cells with 5-micron cell gap by a capillary action along the rubbing direction at 100°C.
Two types of cells used to characterize the physical properties except anchoring energy of liquid crystal materials were laboratory made. The planar cells were made with two indium-oxide coated glass substrates, spin-coated with PI2555 (DuPont), separated by 10 micron spacers, and assembled with the rubbing directions in 180 degree. The vertical aligned cells were spin-coated with SE1211 (Nissan Chemical Industries, LTD) without rubbing and the cell gap was 10 microns. Planar cells with 20 µm cell gap made from E.H.C. Co are only used in anchoring energy measurement. The liquid crystals were capillary filled in all cells. All the material characterizations were done in room temperature (~20 °C).
3. Material characterization
The dielectric spectroscopy of pure and CNT doped liquid crystals is depicted in Fig. 2. The dielectric constant of a nematic liquid crystal was measured both in the direction parallel (ε‖) and perpendicular (ε⊥) to the long axis of the molecule with vertical aligned and anti-parallel aligned cells, respectively. The applied voltage for the dielectric permittivity measurement is 0.2V and with a frequency ranging from 10Hz to 1MHz, which is smaller than the Freedericksz transition threshold. The values of dielectric permeabilities for all four liquid crystals are listed in Table 1. The measured values are about 3~13% smaller than the values from data sheets, the difference may be resulted from the different alignment layers we used. The critical voltage Vc is designated as the threshold voltage when the splay to bend transition of liquid crystal occurs. The CNT doped liquid crystal materials showed no change in both the ε⊥ and ε‖ in the measured frequency range. Therefore the dielectric anisotropy, Δε, at operating frequency, which is 1 kHz in this experiment, remained the same. It indicates that the electro-optical properties of CNT-doped materials should remain similar to the pure nematic liquid crystal host. The imaginary part of the dielectric permittivity was increased for about one order in magnitude. The increase is more significant in the low frequency region which indicates that the CNTs increase the conductivity of the material.
The threshold voltage, Vth, of a planar cell can be described by equation (1)
where K1 is the splay elastic constant, and ε0 is the electric permittivity in vacuum. The 10 µm planar cells meet the condition of strong anchoring energy (~10-3) and low pretilt angle (~2°). Thus, Eq. (1) is valid. The dielectric anisotropy Δε of a liquid crystal is measured in the previous experiment. Accordingly, the elastic constant K1 can be determined by measuring the threshold voltage with T-V curve from the planar cell.
For a planar cell between crossed polarizers, the relation between the phase change δ and time t after the removal of applied voltage can be written as
where K1 is the splay constant in a planar cell, d is the cell thickness, and γ1 is the rotational viscosity . By plotting the data points of ln(δ0/δ) and t from oscilloscope trace (Fig. 3(a)), a slope representing can be obtained as shown in Fig. 3(b). With the effective elastic constants of a planar cell from the measurement of threshold voltage, the rotational viscosity can be evaluated. The rotational viscosity we obtained from this method is generally smaller than the data from Merck. The cell gap values we used in the calculation were measured with empty cells. So the cell gaps might be smaller after the injection of liquid crystals while we use the cell gap value measured at empty cell and thus cause the smaller calculated values in γ 1 comparing to datasheets.
The measured values of rotational viscosity for the liquid crystal materials are listed in Table 1. According to the table, the dielectric permeabilities and rotational viscosities show negligible changes after doping CNTs and do not endorse significant improvements in response time.
The polar anchoring is measured with planar cells manufactured by E.H.C. Co with a cell gap of 20µm. By measuring phase retardation R of liquid crystal cells in response to an applied voltage V, the anchoring strength can be obtained by the following equation: 
R0 is initial phase retardation, J̃ 0 is the intersection in x-axis in the plot of R(V-V′)/R 0 vs. (V-V′),V′=α(1-ε ⊥/ε ‖)Vth, W is polar anchoring, κ=(K 3–K 1)/K 1, yp=sin2 θp, and θp is pretilt angle. An example of anchoring measurement is as shown in Fig. 4. With a linear fit at region larger than 6Vth, it insures that the liquid crystal molecules in middle layer are parallel to the electric. The polar anchoring can be evaluated with slope and the K1 and d from previous measurement.
(Note: The discrepancy between the measured dielectric permeabilities are about 3~13% lower than those of the data sheets from Merck.)
4. Response time measurement
The OCB cells used in our experiment are manufactured by E.H.C. Co. The transmittance versus applied voltage measurements for the OCB cells were determined with a set up consisting of a He-Ne laser (λ=633 nm), a pair of crossed polarizers, the cell and a photodiode detector, where the rubbing direction of the cell is placed at a 45° angle between the crossed polarizers. The applied voltage was square wave at 1 KHz. As shown in Fig. 5(a), the T-V curves of OCB cells have similar responses with respect to the voltage ramping. It was also found that the critical voltage of the splay to bend transition remained the same after the liquid crystal was doped by the CNTs as shown in Table 1. The slight change in transmission at zero voltage in Fig. 5(a) is due to the different phase retardation caused by the cell thickness variation from different cells. According to the T-V curve of the liquid crystal, we selected 8 gray levels to evaluate the improvement in response time. The bright state was selected as gray level 0 and dark state as gray level 7. In between, 6 gray levels were chosen. An example of the transmission of gray levels from CNT-doped BL006 is shown in Fig. 5(b).
Because a splay to bend transition exists in an OCB cell, the response times are measured by switching between gray levels and a bias voltage as indicated in Fig. 6. The rise time is defined as the time when the transmission goes from 90% to 10%, and the fall time is from 10% to 90%. The voltage difference between bias voltage and gray-level voltages is expressed as δV. It is shown in Fig. 7 that after doping CNTs, the fall time of all four liquid crystals in OCB cells are improved. The percentage of improvement is listed in Table 2, it shows significant change in fall time. Although the rise times slightly increase in the cases of BL006 and MLC6080, the decrease in fall time is larger than the increase in rise time. Thus, the total response time is improved. As was demonstrated earlier, the rotational viscosity γ1, dielectric anisotropy Δε of the material does not show significant change after doping CNTs. The OCB cells are with a weak polar anchoring at the range of 10-4 to 10-5 J/m2 as shown in Table 1, which indicate the change in anchoring energy will lead to the change in response time. Our results show that the CNTs have increased the surface anchoring by one order. A general LC response time considering polar anchoring can be written as ,
where d is the cell gap, γ1 is the rotational viscosity of liquid crystal, and K is effective elastic constant of liquid crystal. The equation is derived under the condition of homogeneous alignment. For OCB cell ignoring the flow effect, the response time should still be proportional to same parameter. Therefore, a modified expression for the fall time of OCB mode, considering polar anchoring, has an additional term because of the warm up voltage Vb is written as
where ε0 is the electric permittivity in vacuum, Δε is the dielectric anisotropy of liquid crystal, and K is the effective elastic constant of liquid crystal at the bend state. From the observation of TV curve in Fig. 5, the saturation voltages of dark state are quite similar in the cases of without and with CNTs. It shows that the effective constant K in OCB cell doesn’t show significant change with CNTs. Thus, as the anchoring strength from the alignment layer is increased, the response time will be shortened. The increase in anchoring energy is believed to be from the π-π electron stacking between CNTs, surface alignment layers and LC molecules. For rise time, the bias voltage Vb in response time Eq. (5) is replaced by driving voltage V. The effect of change in response time from anchoring W is small; instead, the driving voltage V is a dominating factor. As shown in Fig. 7, the rise time decreases with increasing δV. Overall, a faster response is achieved by doping nematic liquid crystals with CNTs.
To confirm if the response time improvement is due to the dielectric heating from the increase in imaginary part, a one-step pulse was applied in order to prevent the dielectric heating. A critical response time was performed with TL204 with pulse duration of 215ms, divided into warm up, switching on and switching off stages. As was indicated earlier, the addition of CNT causes the increase in the imaginary part of dielectric permittivity. An example of the applied waveform is shown in Fig. 8. The liquid crystal in the OCB cell goes from splay to bend transition at stage A with a warm up voltage. The dip shown in region A shows the cell is going through a splay to bend transition and phase retardation before it reached the bias voltage. Then the cell was switched to the dark state in which the liquid crystal is aligned homeotropically at stage B. At stage C, the cell was switched back to the bright state with a warm up voltage. The results of response times using the one-pulse measurement is summarized in Table 3. The results show that the improvement in response times using the continuous waveform method is validated. The discrepancy in percentage of improvement between Tables 2 and 3 is due to the gray scale switching and the other is on-off switching.
The results show that CNTs have changed the property of alignment layer, causing the increase in anchoring energy for one order. The response time is shortened because of the increased anchoring energy. The improvement in response time with the CNT doped liquid crystal OCB mode display is confirmed with both the continuous waveform and one-pulse switching methods. The CNT-doped liquid crystal OCB cells are compatible with field sequential color LCDs. Furthermore, it is possible to extend the enabling fast-switching CNT doped materials to other types of liquid crystal displays, such as Blue phase LCDs, Cholesteric LCDs, STN LCDs, or PDLCs.
The authors gratefully acknowledge Samsung Electronics Company, Ltd. for providing a Graduate Research Scholarship under the Chemical Physics Interdisciplinary Program at Kent State University.
References and links
1. P.J. Bos and K. Koehler, “The pi-cell: A fast liquid crystal optical switching device,” Mol. Cryst. Liq Cryst. 133, 329 (1984). [CrossRef]
2. S. Iijima, “Helical microtubules of graphitic carbon,” Nature 354, 56–58 (1991). [CrossRef]
3. W. Lee, C.-Y. Wang, and Yu-Cheng Shih, “Effects of carbon nanosolids on the electro-optical properties of a twisted nematic liquid crystal host,” Appl. Phys. Lett. 85, 513–515 (2004). [CrossRef]
4. Wei Lee and Yu-Cheng Shih, “Effects of carbon-nanotubes doping on the performance of a TN-LCD”, Journal of SID , 13, 9 (2005).
5. H.-Y. Chen and W. Lee, “Faster electro-optical response characteristics of a carbon-nanotube-nematic suspension,” Appl. Phys. Lett. 90, 033510-1/033510-3 (2007). [CrossRef]
6. Seung Hee Lee, Hee-Kyu Lee, Seung-Eun Lee, and Young Hee Lee, “Effects of carbon nanotubes on physical properties of nematic liquid crystal and liquid crystal device,” Proc. SPIE 6487, OU-1, (2007). [CrossRef]
7. X. Nie, R. Lu, H. Xianyu, T. Wu, and S-T Wu, “Anchoring energy and cell gap effect on liquid crystal response time,” J. Appl. Phys. 95, 5502 (2004).
8. J. Liu, A. Rinzler, H. Dai, J Hafner, A.R. Bradley, P. Boul, A. Lu, T. Iversion, A.K. Shelimov, C. Huffman, F. Roderiguez-Macias, Y. Shon, R. Lee, D. Colber, and R.E. Smalley, “Fullerene pipes,” Science 280, 1253– 1256 (1998). [CrossRef] [PubMed]
9. Yu. A. Nastishin, R. D. Polak, S. V. Shiyanovskii, V. H. Bodnar, and O. D. Lavrentovich, “Nematic polar anchoring strength measured by electric field techniques,” J. Appl. Phys. 86, 4199 (1997). [CrossRef]
10. L. M. Blinov and V.G. Chigrinov, Electrooptic Effects in Liquid Crystal Materials, (Springer, 1993), Chapter 2.