Ultra-low viscosity liquid crystal materials

: We report five ultra-low viscosity nematic liquid crystal mixtures with birefringence around 0.1, dielectric anisotropy in the range of 3 to 6, and clearing temperature about 80°C. A big advantage of these low viscosity mixtures is low activation energy, which significantly suppresses the rising rate of viscosity at low temperatures. Using our mixture M3 as an example, the response time of a 3- μ m cell at − 20°C is only 30 ms. Widespread application of these materials for display devices demanding a fast response time, especially at low temperatures, is foreseeable.


Introduction
Fast response time is one of the most critical requirements for most liquid crystal display (LCD) devices [1] because it helps reduce motion picture image blur and crosstalk, enhance optical efficiency, and suppress color mixing for field-sequential displays [2,3]. Mobile displays, wearable displays, and car navigation systems are often used in outdoor and they have to endure harsh weather conditions, like low temperatures (−20°C). In such a cold ambient temperature, LC response time is usually as sluggish as several hundreds of milliseconds. As a result, the displayed image quality is severely degraded [4].
To shorten response time, a straightforward approach is to decrease the LC cell gap (d). However, for an LCD a certain dΔn value is required in order to obtain high transmittance; here Δn is the LC birefringence. For example, the commonly used fringe field switching (FFS) LCD requires dΔn≈320-340 nm in order to achieve high transmittance [5,6]. Although thin cell gap helps greatly to achieve fast response time [7], this approach imposes two problems: it reduces manufacturing yield and it demands a higher Δn LC, which has stronger wavelength dispersion [8]. To obtain white color, the transmittance of red, green, and blue sub-pixels should be balanced. From experimental studies, the preferred Δn for FFS is around 0.10 ± 0.01. Under such circumstance, the cell gap is about 3 μm, which is still manageable for high-yield manufacturing. With abovementioned constraints, the simplest way to reduce response time is to employ a low viscosity LC.
In this paper, we formulated five ultra-low viscosity LC mixtures with Δn≈0.1, dielectric anisotropy Δε≈3 to 6, and clearing temperature about 80°C. A big advantage of these low viscosity LC mixtures is their small activation energy, which significantly suppresses the rising rate of viscosity at low temperatures. Using our mixture M3 as an example, the response time of a 3-μm FFS cell at −20°C is about 30ms. These materials will find widespread applications for display devices that demand a fast response time.

Mixture formulation
Our low viscosity LC mixtures contain three major ingredients: 1) high Δn and large Δε compounds, 2) ultra-low viscosity diluters, and 3) wide nematic range compounds. Table 1 lists the chemical structures and compositions of our five mixtures. Compounds 1 and 2 have high Δn and large Δε (>25) [9,10], but their viscosity is also high. To lower the viscosity, we added more than 40% non-polar diluters (#3) [11]. To widen nematic range and achieve high clearing point, we added some terphenyl compounds (#4). To obtain different Δε values, we formulated five LC mixtures by varying the compound concentrations as Table 1 shows.

Material characterization
In experiment, we measured the dielectric anisotropy, birefringence, visco-elastic constant, and activation energy of these five mixtures. To avoid crowdedness of data presentation, here we only show the measured results of M3, M4, and M5 in the following Sections. Table 2 summarizes the key results of these five mixtures.

Dielectric anisotropy
Dielectric anisotropy affects the operation voltage, peak transmittance [12], and response time (through viscosity) of the FFS LCD. To reduce the power consumption of a mobile display, it is desirable to keep the on-state voltage below 5V. This requirement demands a fairly large Δε. On the other hand, to obtain low viscosity we should keep Δε as small as possible. Thus, there exist contradicting requirements for Δε between low operation voltage and fast response time. A compromised Δε value is in the range of 3 to 6. Table 2. Measured properties of the five LC mixtures at T = 23°C, λ = 633nm, and f = 1 kHz. In experiment, we used the capacitance method to measure the dielectric constants (ε // and ε ⊥ ) of our five LC mixtures at room temperature (23°C). Detailed procedures have been reported in Ref [13], and the measured results are listed in Table 2. From Table 2, the ε // and ε ⊥ of M3 is 6.26 and 2.76, respectively, i.e., Δε = 3.50, which is much lower than that used in conventional p-FFS LCD (Δε = 8~10) [14]. With such a low Δε, the operation voltage, which is inversely proportional to the square root of Δε, would undoubtedly increase [12]. Fortunately, the transmittance of p-FFS increases as Δε gradually decreases. As a result, we can still get high transmittance at a relatively low voltage (5V) using a low Δε LC material [15]. For M4 and M5, the Δε value is 4.60 and 6.18, respectively. Among these three mixtures studied, M3 contains the largest amount of diluters, thus its viscosity is the lowest but its dielectric anisotropy is also the smallest.

Temperature dependent birefringence
Birefringence of an LC is mainly governed by the conjugation length and order parameter [16]. To measure Δn, we filled the LC mixture into a homogeneous cell made of indium tin oxide (ITO) glass substrates. The inner surface of the ITO-glass was over-coated with a thin polyimide alignment layer. The pretilt angle was about 2°. The cell was sandwiched between two crossed linear polarizers. By measuring the voltage dependent transmittance through LabView system, we can obtain Δn easily. Detailed method has been described in [17]. From Table 2, the measured birefringence at room temperature is Δn = 0.100 for M3, 0.097 for M4, and 0.099 for M5. These values are very close to our ideal one, which is 0.1.
Next, we measured the temperature dependent birefringence. We placed the LC cell on a Linkam heating stage controlled by the temperature program (Linkam TMS94). Results are shown in Fig. 1, where dots stand for measured data and solid lines for the fittings using Haller's semi-empirical equation [18]: where Δn 0 is the extrapolated birefringence at T = 0, S is the order parameter, T is the temperature, T c is the clearing point, and β is a material parameter. Through fittings, we found Δn 0 = 0.138 and β = 0.174 for M3, Δn 0 = 0.133 and β = 0.177 for M4, and Δn 0 = 0.135 and β = 0.165 for M5, respectively. Using these fitting parameters, we can calculate the order parameter (S), which will be used later.

Elastic constant and viscosity
In an LCD, the response time is proportional to the visco-elastic coefficient (γ 1 /K ii ), where K ii is the corresponding elastic constant depending on the LC alignment. For examples, for vertical alignment, K ii = K 33 is the bend elastic constant, and for in-plane switching (IPS) cell [19] K ii = K 22 is the twist elastic constant. However for FFS, the electric field has transversal and longitudinal components so that both K 22 and K 11 are involved, although twist dominates [20]. Several approaches have been proposed to measure γ 1 and K 11 . Here, we used the time dependent transmittance method described in Ref [13].
For a homogeneous cell, the threshold voltage is related to K 11 and Δε as [21]: where K 11 is the splay elastic constant and ε 0 is the permittivity of vacuum. From the measured threshold voltage and dielectric anisotropy, we can extract K 11 from Eq. (2). As listed in Table 2, all the five mixture we prepared have a very similar K 11 value (~12pN) because they basically consist of same compounds except at different compositions. Next, we used the same setup as described in Sec. 3.2 to measure γ 1 /K 11 . Detailed method has been described in [13]. Since K 11 has already been obtained from Eq. (2), we can extract γ 1 from the measured γ 1 /K 11 . The measured γ 1 is 45.1 mPas, 50.4 mPas, and 52.3 mPas for M3, M4, and M5, respectively. These γ 1 values seem to correlate with Δε linearly, as will be examined in more detail later.

Activation energy
As the temperature decreases, rotational viscosity increases exponentially as [22,23]: where E is the activation energy and k B is the Boltzmann constant. From Eq. (3), activation energy determines the rising rate of rotational viscosity in the low temperature region. Key parameters affecting E include molecular structure and conformation, and intermolecular interactions [13]. As Table 1 shows, the low Δε LC mixture contains more non-polar diluters. As a result, its activation energy is relatively small, which in turn only causes a mild increase as the temperature decreases. To extract E, we measured the temperature dependent viscoelastic coefficient of these mixtures using the same method discussed above. In theory, temperature dependent γ 1 /K 11 (homogenous cell) can be described as follows [22]:   [15,24]. Our low viscosity LC mixtures exhibit much lower activation energy. In experiment, we tested a 3.5-μm FFS cell with electrode width l = 3μm, electrode gap g = 4μm using M3. Peak transmittance (90.4%) was achieved at 7.1 V rms under λ = 514nm. The measured response time [rise, decay] is [10.3ms, 10.7ms] at room temperature. As the temperature decreases to −20°C, the decay time increases to 42ms. If we use a thinner cell gap (e.g. d = 3μm), the expected decay time, which is proportional to d 2 , is ~30ms. This result is >10X faster than that of the MVA cell reported in Ref [4] at the same temperature. More details about the electro-optic properties using ultra-low viscosity and low dielectric anisotropy materials have been reported in Ref [15]. Table 2 summarizes the measured physical properties of the five mixtures we prepared. Their Δn is around 0.1 and clearing point ≈80°C, which is desirable for FFS LCD applications. As Δε decreases from 6.2 to 3.1, γ 1 decreases from 53 mPas to 41 mPas. The correlation seems to be linear between these two parameters. To further investigate this empirical relation, more mixtures using the compounds listed in Table 1 are prepared for comparison. Figure 3 depicts the results, from which a linear relation between Δε and γ 1 is indeed observed. The extrapolated γ 1 is about 30 mPas for the employed non-polar diluters whose Δε≈0. For some LCDs, such as desktop computers and TVs, they can afford to have a higher operation voltage, say 7.5V. Thus, we can use a lower Δε LC mixture and achieve a faster response time.

Conclusion
We have formulated five ultra-low viscosity LC mixtures with positive Δε and characterized their physical properties. In addition to low viscosity, their Δn is around 0.1 and T c~8 0°C, which is ideal for FFS LCDs. Another big advantage is their small activation energy, which significantly suppresses the rising rate of viscosity at low temperatures. Widespread applications of these ultra-low viscosity LC mixtures are expected.