Nanoparticles and LCDs: It's a Surprising World

Doping nanoparticles into liquid-crystal-display (LCD) host media induces the modification of almost all the physical properties of liquid crystal, causing a reduction of both the operating voltage and the response time. These techniques may be an alternative approach for improving the properties of liquid crystal other than chemical synthesis.

by Shunsuke Kobayashi and Naoki Toshima

LIQUID-CRYSTAL DISPLAYS (LCDs) doped with nanoparticles exhibit novel, interesting, and useful characteristics such as the reduction of operating voltage and response times. Sometimes, there is a change of operating mode, even though the volume occupation factor of nanoparticles is very small, i.e., on the order of 10-3–10-4. Several types of nanoparticles whose size is several nano-meters, such as metals, inorganic materials, semiconductors, polymeric nanostructures, etc., have been utilized in our research. Herein, we present some examples of the enhancement of the performance of LCDs doped with nanoparticles, and we demonstrate a field-sequential full-color LCD using pseudo-spin-valve (PSV) LCDs that exhibit beautiful moving images with reduced color break. We also show a super-twisted-nematic (STN) LCD exhibiting fast response at a low temperature, i.e., –30°C. Furthermore, we present some basic experimental results obtained through our research and try to give a theoretical explanation for these phenomena.

What Are Nanoparticles?

Nanoparticles may be defined as particles whose size is between 2 and 100 nm. Metal nanoparticles have a history dating back more than several hundred years. A good example is stained glass exhibiting beautiful colors. These colors come from metal nanoparticles such as gold (Au) dispersed in the glass host media. Another example is Au colloid dispersion in water. Nanoparticles have been widely utilized as cosmetics and as fillers dispersed in several industrial materials such as paper and inks. We have extended these technologies and applied them to LCDs, using not only thermotropic types such as nematic, smectic, cholesteric, ferroelectric, and polymer, but also lyotropic types of LCDs.

As far as we know, several types of nano-particles have been utilized for LCD applications. They include metals,1,2 inorganic materials including oxides such as MgO,3,4 SiO2,5 BaTiO3,6,7 Sn2P2S6,6,7 ferromagnetic materials,8 polymeric materials, carbon 60,11 and others (in this article, we omit carbon nanotubes).

Generally, nanoparticles are difficult materials in terms of obtaining a good spatial dispersion in a host medium (water is the easiest host material for this, followed in order of increasing difficulty by organic solvents plus water, organic solvents, and liquid crystals as they are used for host media).

To realize good spatial dispersion, it is necessary to use the so-called ligand molecules to cover nanoparticles to protect them from their aggregation in a media; however, in some cases, there is no need for special ligand molecules. In our study, we chose nematic liquid-crystal molecules as ligand molecules for the nanoparticles of metals, but in other nanoparticles we adopted several different approaches by using newly synthesized molecules or not using special ligand molecules at all.

Figure 1 depicts the molecular configuration of the cross-sectional view of an LCD cell with nanoparticles. However, this is drawn by our conjecture. In this case, the ligand molecules take a radial configuration (an alternative is a circular configuration).12

Synthesis of Nanoparticles

Metal nanoparticles such as Pd, Ag, and Pt have been used as catalysts dispersed in a liquid phase, where they are normally covered with polymers as ligand molecules. However, this approach has failed for LCD application because it results in the destruction of LC ordering. An example of synthesizing a colloidal dispersion of Ag/Pd nanoparticles for our purpose follows. Silver perchlorate and palladium (II) acetate are dissolved in ethanol solution along with nematic liquid-crystal molecules; then, this solution is irradiated with a UV light source or subjected to an alcohol reduction.1,2 Semiconductor nano-particles are similarly synthesized.13 For polymers, monoacrylate or diacrylate photo-curable monomers that take on the NLC phase before photocuring are doped into liquid crystal and photocured under the application of an electric field.9,10


Shunsuke Kobayashi is Professor of Electronics and Information at the Graduate School of Science and Technology (GSST) and Director of the Liquid Crystal Institute at the Tokyo University of Science, Yamaguchi, 1-1-1, Daigaku-dori, Sanyo-Onoda, Yamaguchi 756-0884, Japan; telephone +81-8-3688-4540, e-mail: kobayasi@ed.yama.tus.ac.jp. Naoki Toshima is Professor of Materials and Environmental Science and the GSST Director of the Advanced Material Institute at the Tokyo University of Science, Yamaguchi, Japan.

 

Electro-Optical Effects of LCDs Doped with Nanoparticles

It has been reported that the physical properties of LC media in LCDs are modified by the existence of nanoparticles and, hence, greatly affect the electro-optical (EO) characteristics of nanoparticle-doped LCDs, depending on the nature of the nanoparticles, their size, and concentration. This approach is an alternative to the conventional chemical synthesis technique to improve the properties of LCs. Below, we present these topics in some detail.

Ferromagnetic Particles

In this case, each nanoparticle has a large permanent-magnetic moment that induces an alignment of LC molecules, and thus the physical properties of the LC medium will be modified and altered depending on the concentration of nanoparticles.8

Ferroelectric Nanoparticles

When each nanoparticle takes on a single-crystal phase, then it has a large permanent electric dipole moment that induces an alignment of surrounding LC molecules. This results in the increase of clearing point and order parameter.6,7 And thus, there occurs the reduction in the threshold voltage and the operating voltage.6,7 However, the occurrence of these properties are very much dependent upon the combination of nanoparticles and liquid-crystal host material.6 (The E-O effects of BaTiO3-doped LCDs are very much dependent on the synthesizing process of BaTiO3 nanoparticles.15)

 

p27a_tif

Fig. 1: Molecular structure of a nanoparticle-embedded NLCD.

 

p27b_tif

Fig. 2: Frequency modulation response of a TN-LCD with 5CB-Ag particles. The upper trace is the optical transmission.

 

Metallic Nanoparticles

So, what happens in an LCD doped with metal nanoparticles? Surprisingly, an LCD doped with metal nanoparticles works as an electro-optical switching device or even as a dot-matrix LCD, where both show fast response speed of about 500 μsec in the fastest component where the thickness of the cell is 5 μm. Figure 2 demonstrates an example of the frequency-modulation response with short response times in a TN cell, particularly with 5CB-Ag nanoparticles. This device is called an FM/AMLCD.16 Before presenting a detailed discussion on these results, keep in mind the following characteristics of metal nanoparticles: (1) A metal nanoparticle is never ionized by ejecting electrons from a nanoparticle due to an extremely strong coulomb force (Kubo theory)17 and (2) a nanoparticle floating in a capacitor filled with an insulator such as LC and under the application of an electric field virtually becomes an insulator due to a strong depolarization electric field.16

Due to the existence of electrons in a nanoparticle of metal, there exists a difference of the electrical relaxation times between the nanoparticle of metal and the surrounding LC molecules such that Δτ = ε011–ε22), where εand σ1 are the dielectric constant and electrical conductivity of the LC, respectively, and ε2 and σ2 are the dielectric constant and the electrical conductivity of the metal nanoparticle. This Δτ results in the Maxwell– Wagner effect (M–W),18,19 which produces extra oscillating electrical charges around the boundary surfaces of each nanoparticle. This in turn generates a jump in the dielectric constant, [ε(0) – ε(∞)], called the dielectric strength in a low-frequency region centered about the dielectric relaxation frequency fR(= 1/2πτ) that determines the frequency range of the FM/AMLCD (τ represents the dielectric relaxation time). The dielectric function of our system obeys the Debye dispersion formula

p28_eq1  (1)

 

The actual values of the dielectric relaxation frequency range from several tens of Hertz to several thousands of Hertz, depending on the materials of metal nanoparticles and their concentrations. Figures 3(a) and 3(b) show the data of dielectric spectroscopy on the sample TN-LCD cells with 5CB-Ag/Pd.

In order to explain the frequency range of the FM response, the fast response, and the dielectric function of metal-nanoparticle-doped TN-LCD cells, we formulated a dielectric function based on an equivalent circuit as shown in Fig. 4.16 In the Appendix, we enumerate the components of Eq.(1) derived by the authors in comparison with those by Maxwell–Garnett in 1906.20 The reason for our formulae is that the equations for ε′(ω), ε″(ω), and τ derived by Maxwell–Garnett in 1906 shown in Appendix Table 1 do not contain the volume occupation factor φ2.20 Contrary to this, actually obtained data shown in Figs. 3(a) and 3(b) depend on the concentration of nanoparticles and, thus, on φ2. By using these formulae, we succeeded in explaining the experimental data of dielectric spectroscopy as follows. First, we define the gain in the ε(ω) as G = [ε(0)–ε(∞)]/ε(∞). The experimental value of G is about 10; this value agrees with the calculated value using our formulae.

Next, we will discuss the dielectric relaxation time. An approximate formula for τ by assuming σ> σ1, which is relevant to the FM frequency, will be

p28_eq2(2)

The square dots shown in Fig. 5 indicate the data of τ obtained from the curves in Figs. 3(a) and 3(b), and the solid line is a theoretically calculated one and fitted to the data using Eq.(2). We achieved good agreement between theory and experiment by choosing τ= 0.6 sec, ε2= –13.6, and σ2= 3 x 104; we have the same agreement on other samples such as 5CB-Pd/5CB and NLCs–Ag/Pd.

 

p28a_tif

Fig. 3: Data of dielectric spectroscopy on TN-LCD cells with 5CB-Ag/Pd nanoparticles.

 

 

p28b_tif

Fig. 4: A crystal model and equivalent circuit model of nanoparticle-embedded LC.

 

Looking at these results, two issues arise: one is on the value of τ that is on the order of 10-4 sec. If we let σ2 ≈ 107 S/m for the metal, then we have τ ≈ 10-19 sec. This gives a huge discrepancy from 10-4 sec. We consider that the order of σ2 will be σ2 ≈ 10-8S/m. The second concern is, Why this order? Regarding this problem, Kittel provides an exercise in his book about the dielectric constant of a vacuum medium that contains floating metal particles under the application of a dc field.21 An answer is that the dielectric constant of this system is almost nearly equal to that of vacuum.21 This is attributed to the situation that electrons in a metal particle are immobilized due to the existence of a strong depolarization electric polarization. Our situation is slightly different from this example because we apply an ac field and each nanoparticle is covered with LC molecules that have a finite electrical polarization and contain some ionic impurities. Thus, in our situation, the metal nanoparticles embedded in an insulating medium such as LC behave like a semi-insulating material. This is an answer for the value of τ.

The M–W effect appears in LCDs because they comprise a multi-layered structure. The M–W effect is essentially attributed to the appearance of oscillating extra electrical charges, Δq, on the surface regions of a nanoparticle that is symbolically given as point A in our equivalent circuit (Fig. 4). The formula for these extra charges Δq is given as 

p29_eq3 (3)

For an applied ac voltage, sin(ωt). The phase shift φ in the Δq reads tanφ = ωτ, where

τ = (C1 + C2)/(G1 + G2), (4)

The numerator of Eq. (4) must be small because this is the origin of the gain G in the ε(ω). This means that the M–W effect originates from the extra oscillating charge Δq, but the phase shift φ must be small, i.e., φ << π/2. The fast response in the rising process in an FM-AMLCD may be attributed to the amplification of the dielectric torque such that ΔεE2 → GΔεE2.

LCDs exhibiting the FM response are called FM/AMLCDs because they also show the ordinary amplitude modulation. FM/AMLCDs feature a peculiar EO performance that is free from the temperature dependence at a specific frequency, say, 2 kHz in a 5CB-Ag/Pd system.

As a final topic on the metal-nanoparticle-doped LCD, we show the result of a direct multiplexed matrix STN-LCD with NLC-Ag/Pd nanoparticles and 480 x 160 pixels. If we use these nanoparticles with a low concentration below 0.1 wt.%, then the dielectric relaxation frequency becomes less than 50 Hz. Thus, it is possible to operate a matrix STN-LCD by direct multiplexing driven at the frequency of 100 Hz. Surprisingly, this type of STN-LCD shows a three-times-faster response time at a low temperature, i.e., at –30°C.22 Figure 6 demonstrates this example.

 

p29a_tif

Fig. 5: Dielectric relaxation time vs. volume-occupation factor in a TN-LCD cell with 5CB-Ag/Pd nanoparticles. Square dots indicate experimental data and the solid line is a theoretically fitted curve.

 

p29b_tif

Fig. 6: Photographs of an STN-LCD doped with Ag/Pd nanoparticles exhibiting faster response at low temperature.

 

Inorganic Nanoparticles

To obtain good spatial dispersion of inorganic nanoparticles in a liquid-crystal host media, the necessity for the ligand molecules depends on what combination of host LC and nano-particles are adopted.

MgO Nanoparticles

Doping nanoparticles of MgO into TN or ECB cells causes the reduction of the threshold voltage and the operating voltage and also the reduction of response times, particularly at low temperatures (at –30°C as shown in Figs. 7 and 8. These phenomena may be attributed to the reduction of the order parameter owing to the existence of nanoparticles in an LCD medium. This understanding is supported by a computer simulation; by direct measurements of the order parameter through optical retardation; and measurements of the elastic constants, dielectric anisotropy, and rotational viscosity.4 This may be true for all the types of nanoparticles. Actually, the observed values of the threshold are shown to obey

p30_eq

because the elastic constant K ∝ S2 and the dielectric anisotropy Δε ∝ S, where S is the order parameter. It is interesting that due to the doping of nanoparticles, the elastic constant K11 decreases, K22 remains almost the same, and K33 increases depending on the LC materials.4 The full understanding of these phenomena is now under way and the results will be published elsewhere.

Regarding the M–W effect in an NLC cell doped with MgO nanoparticles, no M-W effect is observed in its dielectric function because Δτ = ε01– ε22) is almost null even though their mechanical properties are completely different.

Semiconductor Nanoparticles

In comparison with metals, the number of mobile electrons is smaller than in semiconduc-tors such as CdS, ZnS, ZnO, etc., so that the M–W effect is weak and, hence, the dielectric relaxation frequencies are low, i.e., below 100 Hz. Thus, the FM response rarely occurs.

PSV-FLCDs

Photocurable nanoacrylate, diacrylate, triacrylate, and their mixtures are useful for fabricating polymer-stabilized NLCs and other liquid crystals. These photocurable monomers form a polymeric nanostructure and polymeric networks that play a role in the polymer stabilization of LC molecules. Herein, we represent some results of polymer-stabilized (PS) V-FLCDs exhibiting a threshold-less continuous gray-scale operation. Photocurable monomers, whose molecular weight is small, may form a polymeric nanostructure called side-chain polymer networks. Our system is believed not to be the conventional ferroelectric LC state but a special novel LC state because our system has no spontaneous polarization at the quiescent condition (V = 0). For this reason, we call our device a PSV-FLCD instead of an ordinary FLCD. Figure 9 demonstrates VT curves of our PSV-LCD.

 

p30a_tif

Fig. 7: Shift in the V-T curve in an ECB cell toward low voltage by doping MgO nanoparticles.

 

p30b_tif

Fig. 8: Increase in the optical throughput in a TN-LCD cell with Mgo nanoparticles at low temperature at 100 Hz, 3 V, and at –30°C. In a pure cell, the NLC is almost frozen; a large-scale delay time occurs and bouncing is caused by the mismatch from the Gooch-Tarry condition. In a cell with Mgo nanoparticles, a super cooling may occur that generates a finite electro-optical response accompanied by a long delay time.

 

By adjusting the synthesis and the blend of photocurable monomers, we succeeded in improving the reduction of operation voltages and their temperature stability, and the optical throughput. Response times are always between 200 and 400 μsec.9,10

By using our PSV-FLCDs, we fabricated a prototype field-sequential full-color (FS-FC) 4-in.-diagonal SVGA (800 x 600) LCD; an example of a displayed image on our FS-FC-LCD is shown in Fig. 10. The color gamut is wide compared to those of conventional LCDs, and a color beak is scarcely recognized.10

Conclusion

Doping nanoparticles into LCD host media induces the modification of almost all the physical properties of LC by several percent to several tens of percent. This results in the reduction of both the operating voltage and the response time. In particular, PSV-LCDs have been fabricated by adopting polymeric nanostructure and polymer networks. These techniques may be an alternative approach for improving the properties of LC other than chemical synthesis. By using our PSV-LCD, we succeeded in fabricating a field-sequential full-color LCD with SVGA specifications.

Acknowledgments

This research was supported by the JSPS RFTF 98R14201, NEDO MF-H-15-028003, and the MEXT City Area collaboration project. We are deeply indebted to all the members of these research projects for their enthusiastic collaboration.

References

1Y. Shiraishi, N. Toshima, K. Maeda, H. Yoshikawa, J. Xu, and S. Kobayashi, Appl. Phys. Lett. 81, 2845-2847 (2002). 
2T. Miyama, J. Thisayukta, H. Shiraki, Y. Sakai, Y. Shiraishi, N.Toshima, and S. Kobayashi, Jpn. J. Appl. Phys. 43, 2580-2584 (2004).
3S. Sano, T. Miyama, J. Xu, K. Takatoh, and S. Kobayashi, Proc. IDW '05, 77-78 (2005).
4S. Sano, K. Takatoh, T. Miyama, and S. Kobayashi, SID Symposium Digest Tech. Papers 37, 694-696 (2006).
5K. Takatoh and S. Kobayashi (to be published).
6J. West, F. Li, K. Khang, and H. M. Atkuri, Proc. Asia Display '07, 113-118 (2007).
7J.West, F. Li, K. Zhang, H. Atkuri, and A.Glushenko, SID Symposium Digest Tech. Papers 38, 1090-1092 (2007).
8F. Brochard and P. G. de Gennes, J. Phys. 31, 691-708 (1970).
9S. Kawamoto, M. Oh-kochi, S. Kundu, H. Hasebe, H.Takatsu, and S. Kobayashi, Displays 25, 45-47 (2004).
10T. Fujisawa, M.Hayashi, H. Hasebe, K. Takeuchi, H. Takatsu, and S. Kobayashi, SID Symposium Digest Tech. Papers 38, 633-636 (2007).
11M.Suzuki, H. Furue, and S. Kobayashi, Mol. Cryst. Liq. Cryst. 368, 191-196 (2001). 12H. Yoshikawa, K. Maeda, Y. Shiraishi, J. Xu, H. Shiraki, N. Toshima, and S. Kobayashi, Jpn. J. Appl. Phys. 41, L1315-L1317 (2002).

 

p31a_tif

Fig. 9: V-T curves of PSV-LCDs.

 

 

p31_tif

Fig. 10: Photographs of a field-sequential full-color LCD using PSV-LCD; the specifications are 4 in. on the diagonal and an SVGA resolution (800 ´ 600). To perform an appropriate exposure, photos are taken the instant the backlight emits red, green, and blue light sequentially.

 

13N. Toshima et al. (to be published).
14Ube Material Ind., Data Sheet (2006).
15T. Kineri et al. (to be published).
16S. Kobayashi, T. Miyama, N. Nishida, Y. Sakai, H. Shiraki, Y. Shiraishi, and N. Toshima, J. Display Technology 2,121-129 (2006).
17R. Kubo, J. Phys. Soc. Japan 17, 975-980 (1962).
18J. C. Maxwell, Treatise on Electromagnetism (Dover, New York, 1945), Vol. 1, pp. 451-461.
19K. W. Wagner, Erkl. Arch. Electrotech. 2, 371-387 (1914).
20J. C. Maxwell-Garnett, Phil. Trans. 202, 385-419 (1904) and 205, 237-387 (1906).
21C. Kittel, Introduction to Solid State Physics, 7th ed. (Wiley & Sons, New York, 1996), pp. 412-413.
22Y. Toko, S. Yokoyama, S. Takigawa, S. Nishino, N. Toshima, and S. Kobayashi, SID Symposium Digest Tech. Papers 38, 158-160 (2007).
23A. Masutani, T. Robert, B. Schiiler, and A. Yasuda, SID Symposium Digest Tech. Papers 38, 1355-1357 (2007). •

p32_tbl_tif