1. Introduction

Like the conventional molecular liquid crystal (LC) display [1], lyotropic 2-dimentional (2D) colloid of graphene oxide (GO) can serve as an electrical switching material for display applications [2–4]. Interestingly, the Kerr coefficient of the aqueous GO dispersion is the highest among known Kerr materials, which ensures that a low driving voltage suffices for display applications [2]. However, the GO colloid has several limitations in these applications: chemical instability in water [5, 6], and optical absorption in the visible range. GO particles spontaneously cleave into smaller particles, and produce H + ions in a water solvent, resulting in a decrease of pH of the colloid [7]. As a result of the low birefringence of the GO colloid, the thickness of the GO layer must be larger than that in the usual molecular LC cell. Hence, the optical absorption tail in the blue region in GO makes its optical absorption considerably high [8, 9]. For practical applications these issues must be overcome, but these are intrinsic properties in GO materials and cannot be avoided by simply modifying the functional groups. Hence, a new type of 2D colloid is required that does not have such problems.

In this study, we introduce tetrabutylammonium (TBA)-tethered α-ZrP colloid in various solvents for use in electro-optical devices; it exhibits significantly improved electro-optical (EO) performance compared to GO dispersions. TBA-ZrP colloid is transparent in the visible range and chemically stable over long periods. In particular, TBA-ZrP can disperse in a wide range of solvents, and its selection (e.g., acetone) can significantly improve the EO performance.

2. Materials and methods

Inorganic crystalline layered alpha-zirconium phosphate (α-ZrP), Zr(HPO₄)₂H₂O, was synthesized using the hydrothermal method [10]. Under constant stirring, 16 g of zirconyl chloride octahydrate (ZrOCl₂.8H₂O, 98%, Alfa Aesar) was added homogeneously to 160 mL of 14 M phosphoric acid (H₃PO₄, 85%, Daejung, Korea). The mixture was sealed in a Teflon-lined autoclave and kept at 200 °C in an oven for 16 h. The product was α-ZrP crystals with an irregular hexagonal shape, as shown in the scanning electron microscopy (SEM) images in Fig. 1(a) (FESEM/EDS, JSM-7600F, JEOL).

 

Fig. 1 a) SEM images of α-zirconium phosphate (α-ZrP) crystals, b) opaque 1w% aqueous GO and transparent 1 wt% aqueous α-ZrP colloids before and after two months’ storage, and c) zeta potential of ZrP colloids in various solvents before and after two months’ storage.

Download Full Size | PDF

To exfoliate α-ZrP crystals, its aqueous dispersion was mixed with tetra-n-butylammonium hydroxide (TBA⁺ OH^–, 40% in water, Alfa Aesar) at a molar ratio of 1:1 (ZrP:TBA⁺ OH^–) using a vortex. The suspension was ultrasonicated (SD-250H, Mujigae Co., Korea) at 0 °C for 2 h and left for 3 days to ensure full exfoliation of the crystals [11, 12]. The water content in the synthesized aqueous α-ZrP colloid was replaced by other organic solvents such as acetone, N,N-dimethylformamide (DMF), or N-methyl-2-pyrrolidone (NMP) by using a supernatant replacement method [13, 14]; the concentration of α-ZrP in solvent was then adjusted as an experimental parameter.

A dynamic light scattering instrument (Zetasizer Nano, Malvern Instrument Ltd, UK) was used to measure the mean lateral size (~447 nm) and zeta potential of the exfoliated α-ZrP monolayers dispersed in various solvents at 0.01 wt%. The viscosity and conductivity of the samples were measured using a Micro-Ostwald viscometer with an inner capillary diameter of 0.4 mm (SI-Analytics GmbH, Germany) and a conductivity/pH meter (Cyberscan PC 300, Eutech Instruments, Singapore). To measure the EO response of α-ZrP dispersions including their switching times, experiments were carried out using a cell with 1 mm thickness and indium tin oxide (ITO) electrodes that are horizontally separated by 1 mm. The optical transmittance was measured using the method described in a literature; we only replaced the cell in the experimental set-up shown in Supplementary Figure S6 in [2]. 10kHz square electric signals were used for the measurement.

3. Experimental results and discussion

Exfoliated α-ZrP dispersions in acetone, DMF, NMP, and water are transparent in the visible spectral range [Fig. 1(b)] [15], which is an important advantage for electro-optical applications in comparison to the opaque GO dispersion [16,17]. The macroscopic appearances and zeta potentials of α-ZrP dispersions did not change for two months’ long storage [Figs. 1(b) and 1(c)], which confirms their excellent colloidal stability. This shows that the functional groups and surrounding ionic circumstance on α-ZrP particles are stable as well. An exfoliated aqueous α-ZrP dispersion is basic, with a pH in the range 8–11, depending on the concentration, indicating that the number of OH⁻ ions from TBAOH exceeds the H⁺ ions coming from the negatively ionized α-ZrP basal plane [18].

However, the zeta potentials of α-ZrP dispersed in acetone, DMF, NMP, and water are all negative [Fig. 1(c)], which means that the negative surface-ionic groups (−O⁻) on α-ZrP are not fully bonded by TBA⁺. Hence, TBA⁺ ions act as bulk ions that screen the negative surface ions and form an electrical double layer. The absolute values of the zeta potentials are large for all the colloids, indicating a stable electrostatic repulsion between nanoparticles. However, α-ZrP colloids in acetone and DMF exhibit the largest absolute zeta potential values, and in NMP, the lowest, which indicates a different ionization rate in different solvents for OH groups on the ZrP surface.

Although the phase transition from an isotropic to a biphasic phase exhibits a first-order transition accompanying phase separation in long storage in the lyotropic nanoparticle liquid crystals (LCs), the experimental determination of the exact phase transition concentration (Φ_C) is not easy in a short time observation. We roughly determined Φ_C by finding the minimum concentration at which a clear optical birefringence pattern is observed under crossed polarizers in a stationary bottle containing colloids, as shown in Fig. 2. The phase behaviors of α-ZrP colloids were significantly dependent on the type of solvent. The concentration Φ_C was the lowest in acetone (~0.4 wt%), a middle value in DMF (~0.6 wt%), and the largest in water and NMP (~0.7 wt%). This order roughly coincides with the zeta potential, which means that the large electrostatic repulsion enhances the nematic interaction in the colloid. We could not clearly distinguish the biphasic from nematic phases, but a partially dark biphasic-like pattern was observed in the middle range of concentrations. Interestingly, the birefringence patterns in acetone and water are gradually enlarged with increasing concentration, but the change is relatively abrupt in DMF and NMP, as shown in Fig. 2.

 

Fig. 2 Liquid crystalline phase behavior of ZrP colloids in various solvents. Birefringence patterns under crossed polarizers are shown, distinguishing isotropic (solid blue line) from biphasic (dotted red line) and nematic (red line) phases. Lowest phase transition concentration (Φ_C) is indicated by vertical dotted white line.

Download Full Size | PDF

The EO performance of the isotropic, biphasic, and low-concentration nematic phase colloids was investigated by introducing them into cells with parallel electrodes, as shown in Fig. 3(a). The application of an electric field generates a Maxwell-Wagner (MW) polarization due to the flux of TBA⁺ ions near surface, and the field-induced ordering of MW polarization produces optical birefringence Δn in α-ZrP colloids. In fact, the Kerr effect can be defined only in isotropic phase. However, the cell appears perfectly dark in biphasic state and partially dark in low concentration nematic state, as shown in Fig. 3(a). As the applied voltage increases, the apparent birefringence increases even in the low concentration nematic phases. Δn as a function of applied voltage was measured for the colloids in various solvents and at different concentrations. The field-induced Δn was relatively large in water and acetone, but it was low in DMF and NMP, as clearly indicated in Fig. 3(b). The maximum Δn was about 2.5 × 10⁻⁴ in acetone and water, which is larger than that reported in aqueous GO dispersions [2, 19]. While the field-induced Δn increases monotonically in isotropic phase, those in nematic phases decreases at low voltages and then increases again. By deliberately ignoring the decreasing Δn region at very low voltages, the effective Kerr coefficient (K_EFF) can be calculated even in the low concentration nematic phase [see the inset graph in Fig. 3(b)], although the effect may be attributed to partially the director reorientation as well as the field induced order parameter.

 

Fig. 3 a) Cell configuration (top) and microscopic image for cells containing 0.4, 0.6 or 0.9 wt% ZrP in acetone at various voltages. b) Birefringence Δn as a function of applied voltage for ZrP colloid in various solvents at specified concentrations (wt%). The inset graph in the right top panel shows the birefringence vs V² at low voltages for 0.9 wt% nematic and 0.6 wt% biphasic colloids in DMF.

Download Full Size | PDF

To investigate the EO sensitivity, K_EFF for each colloid was calculated using the data at low applied voltages, and the results are shown in Fig. 4(a). The K_EFF was largest near 0.8–0.9 wt% in all solvents, but the maximum values varied significantly. The K_EFF for the acetone-ZrP colloid is the largest, while that for water is about half of that value, and for DMF and NMP it is an order of magnitude smaller than for acetone. Although the maximum birefringence values in the ZrP colloids in acetone and water are similar, their maximum K_EFF values are different owing to the different slopes of birefringence curves at low voltages. The K_EFF of the acetone-ZrP colloid is similar to the maximum value in GO dispersions reported in the literature. Interestingly, the K_EFF in ZrP colloids keeps increasing beyond Φ_C, in contrast to the results reported in aqueous GO dispersions, where it rapidly decreases beyond Φ_C [2]. Specifically, the Φ_C of the acetone-ZrP colloid is about 0.4 wt%, but the K_EFF keeps increasing with concentration and reaches its maximum at 0.8 wt%.

 

Fig. 4 a) Effective Kerr coefficient (K_EFF) as a function of concentration for various colloids. b) Conductance and (c) viscosity of ZrP colloids in various solvents. d) Switched optical response of 0.4 wt% ZrP in acetone. e-f) Rise and fall time of ZrP colloid in various solvents (note: logarithmic vertical axes).

Download Full Size | PDF

To clarify the dependence of the K_EFF on solvent, the conductance and the viscosity of ZrP colloids were investigated. The conductance is related to the bulk ion density in the colloid. The magnitude of the conductance was in the (decreasing) sequence: water, DMF, acetone, and NMP [see Fig. 4(b)], which does not correlate with the K_EFF. The viscosity of the ZrP colloids was in the (decreasing) sequence: NMP, DMF, water, and acetone, which is in the opposite order to the magnitude of the K_EFF. The viscosity reflects the interparticle friction, which hinders the rotational motion of nanoparticles under an applied electric field. Surprisingly, the viscosity of the ZrP colloid does not exhibit any sharp increase near the isotropic to nematic (or biphasic) phase transition concentration, unlike the GO dispersions [20]. Particularly, the viscosity of ZrP colloid in acetone is almost constant and the lowest for all concentrations up to 7 wt%; this may provide a clue to its high EO performance. The low viscosity implies that the interparticle friction is low, resulting in an enhanced electrical sensitivity. However, the viscosities of NMP- and DMF-ZrP colloids exhibit a rapid increase with increasing ZrP concentration, indicating the onset of the gel transition. The origin of the low viscosity in acetone-ZrP colloid is not clear at the current stage; it may be related to the intrinsically low viscosity of acetone.

Figure 4(d) shows a typical switching response for the 0.4 wt% ZrP colloid in acetone, and the corresponding rising and falling response times for all colloids are plotted in Figs. 4(e) and 4(f) with logarithmic vertical axes. The response time is also closely related to the viscosity. The colloids in DMF and NMP have slow rise and fall times compared to those in acetone and water, which can be explained by their higher viscosities. The ZrP in acetone has faster fall times than in water [Fig. 4(f)], which accords with the order of viscosity. However, unlike the almost constant viscosity with ZrP concentration, the response times are sensitively dependent on the ZrP concentration. Although the translational viscosity is low in acetone, the rotational particle motion is hindered when the concentration increases. For the rise time, the anisotropic polarizability is the origin of driving force, and a material having a larger Kerr coefficient will have a stronger rotational torque for particle rotational motion. As shown in Fig. 4(e), ZrP colloids in water exhibit slightly faster response than those in acetone, albeit small difference. ZrP colloids in water has slightly higher Kerr coefficient than that in acetone when the concentration is less than 0.7 wt%, and hence, the driving force will be stronger.

Thus, TBA-ZrP in acetone exhibits overall the best EO performance; this seems to be related to its low and concentration-insensitive viscosity. This viscosity phenomenon is unexpected because the interparticle interaction increases rapidly when the concentration exceeds Φ_C in lyotropic nano-colloids; a sharp increase in viscosity is expected as seen in the behavior of GO dispersions [20]. The biphasic region in the middle range of concentrations is wider in acetone and water than in DMF and NMP, as clearly shown in Fig. 2. This implies that the electrostatic repulsive force between ZrP particles is less sensitive to the mean distance in acetone, and the translational or rotational particle motion is relatively free compared to the colloids in DMF and NMP. This unexpected result may be related to a balance between the ionic screening and the surface charge density on the ZrP surface, which could depend on the solvent; further investigation will be required to clarify the underlying mechanism.

4. Conclusion

We have investigated the electrical switching of birefringence in the TBA-ZrP colloid in various solvents (water, acetone, DMF, NMP). The switching behavior is highly dependent on the type of solvent. TBA-ZrP in acetone exhibits the highest effective Kerr coefficient and the fastest switching time among those solvents used. The excellent switching behavior is related to the low viscosity of TBA-ZrP colloid in acetone. The viscosity of TBA-ZrP in acetone is relatively low and less sensitive to concentration than with the other solvents. This indicates that the individual motion of nanoparticles is less restricted in acetone. Thus, selection of the solvent is an important factor in improving the electrical switching of nanoparticles in a 2D lyotropic liquid crystal colloid.