1. Introduction
Anisotropic nanomaterials have been of great interest because their peculiar optical, electrical, catalytic, and mechanical properties. One-dimensional (1D) anisotropic nanomaterials such as nanowires, nanorods, and nanotubes exhibit good transport properties along with their long axis so that they have potential in applications such as flexible transparent electrodes[1-5], nano-composites with high strength[6,7], transistors[ 8], thermo-electrics with high figure of merit[9], and high thermal conductive materials[10]. Two-dimensional (2D) anisotropic materials including nanoribbons and nanoplates can easily cover the surface of certain object and consequently can be used in applications such as flexible/stretchable electrode[11-15], barrier films[16,17], and pigments [18-23].
In particular, highly anisotropic 2D inorganic nanoplates such as metals and metal oxides exhibit lustering effect due to the strong reflection of light from their surface. Generally the 2D metal oxide nanoplates without absorption in visible region show white or silver white in color due to the reflection or back scattering of whole visible light. Therefore, to make a specific lustering color from the 2D metal oxide nanoplates without absorption of visible light, a certain light with specific wavelength should be selectively reflected from the 2D nanoplates. One way to demonstrate selective reflection of specific wavelength in light is to use interference phenomenon by constructing periodic structure of metal oxides with different dielectric constant like one-dimensional (1D) photonic crystals.
Generally, the interference color is dependent on refractive index and thickness of metal oxide layer on the 2D nanoplate substrate because the optical path difference between the reflected light at air/metal oxide interface and the reflected one at metal oxide/anisotropic 2D nanoplate substrate interface is related to the refractive index and thickness of metal oxide layer. Therefore, the refractive index and the thickness of metal oxide layer should be controlled to control the lustering color of 2D nanoplate substrate in order to control the constructive interference at air/metal oxide shell/2D nanoplate substrate interface[ 23,24]. To make constructive interference, the optical path length difference between the reflected light at air/metal oxide shell and the reflected light at metal oxide shell/2D nanoplate should be m+ λ/2 (m = integer, λ = wavelength of reflected light). Hence, the thickness of metal oxide shell is 40~170 nm in order to control visible interference color (λ = 400~700 nm).
In contrast, the lustering effect is related to the intensity of reflected light (reflectance), which is function of refractive index (n) of metal oxide shell. When the light is reflected from air to metal oxide shell, the reflectance is (1 - nshell)2 / (1 + nshell)2 (where nshell is refractive index of metal oxide shell). Accordingly, the metal oxide shell with higher refractive index is more desirable to maximize the lustering effect. The crystalline anatase and rutile TiO2 has high refractive index of ~2.5 and ~2.7, respectively, so it is generally coated on the substrate to maximize reflectance.
Recently, Lee et al. reported that metallic gold color can be appeared by coating α-Al2O3 nanoplates with TiO2[23]. However, the roughness control of the TiO2 shell on α-Al2O3 nanoplates was not systematically studied and the pristine α-Al2O3 nanoplates coated by TiO2 did not have good mechanical stability due to easily detachment of TiO2 shell from the mother α-Al2O3 nanoplate by mechanical impact such as mechanical mixing and sonication. In this regard, here we systematically controlled the roughness of TiO2 shell on α-Al2O3 nanoplates via reaction controlled sol-gel method and improved the adhesiveness of TiO2 shell on α-Al2O3 nanoplate by heat-treatment. Furthermore, we compared the photocatalytic effect of the synthesized α-Al2O3 nanoplates with robust TiO2 shell with the commercialized effect pigment of α-Al2O3 nanoplates with TiO2 shell.
2. Experimental
2.1. Materials
The α-Al2O3 nanoplate substrates (TCERA), TiCl4 (Fluka), HCl solution (Samchun), and NaOH (Samchun), methyl orange (Aldrich), commercially available α-Al2O3 nanoplates with rutile TiO2 shell (MERCK, Xirallic T60-10SW) lustering effect pigment were used as it received without further purification.
2.2. Preparation of α-Al2O3 nanoplate-TiO2 core-shell pigments
For the preparation of the α-Al2O3 nanoplate-TiO2 core-shell pigments, 0.05 g α-Al2O3 nanoplates and 1 mL HCl solution in 20 mL deionized water charged in a vial. The solution was then mixed by magnetic stirring bar for certain time (1, 2, 3, and 4 h) for pre-treatment process. To coat TiO2 shell on the α-Al2O3 nanoplates, 0.25 mL 3M TiCl4 aqueous solution was added in the reactant solution and 2.8 mL of 0.8 M NaOH aqueous solution was added in the solution drop by drop. The reaction was proceeded for 1 h at 98 ℃. Then 0.25 mL 3M TiCl4 aqueous solution and 2.8 mL of 0.8 M NaOH aqueous solution were added in the same manner and the reaction was proceeded for 1 h at 98 ℃. Finally, we added again 0.25 mL 3M TiCl4 aqueous solution and 2.8 mL of 0.8 M NaOH aqueous solution in the same manner. The reaction was also further proceeded for 1 h at 98 ℃ and then was cooled down to room temperature (S8-5 sample in Table 1). Various reaction conditions were summarized in Table 1. To check the effect of HCl pre-treatment time on the uniformity and morphology of TiO2 shell, we changed the HCl pre-treatment time from 1 h to 2, 3, and 4 h while keeping the other experimental conditions constant.
2.3. Characterization
The morphology of α-Al2O3 nanoplates and α-Al2O3 nanoplates with TiO2 shell was characterized by scanning electron microscopic (SEM: Supra 55, Carl Zeiss). For this, we dropped the α-Al2O3 nanoplate and the α-Al2O3 nanoplate with TiO2 shell dispersion solution on a Si wafer and dried them at convection oven at 60 ℃ for 30 min. Pt was then sputtered on the sample prior to examine SEM image. The crystal structure of α-Al2O3 nanoplate was characterized by X-ray diffraction (XRD: D8 advance, Bruker) machine. For this, we dropped the anisotropic α-Al2O3 nanoplate dispersion solution on a Si low background sample holder and subsequently dried it. We scanned the samples from 20° to 80° at a scan rate of 6 °/min under irradiation of Cu K-α (λ = 0.15406 nm).
3. Results and Discussion
Figure 1(a) is a scanning electron microscopic (SEM) image of α -Al2O3 nanoplate indicating that the size of nanoplates is 10~30 μm and the inset SEM image shows that the thickness of the nanoplate is ~500 nm. The aspect ratio (lateral size to thickness) of α-Al2O3 nanoplates is 20~60 and their surface is smooth enough not to scatter the incident light to the nanoplates. Therefore, the α-Al2O3 nanoplates can be used as substrate for lustering pigments. The X-ray diffraction (XRD) pattern of the nanoplates was exactly matched with α-Al2O3 crystalline phase as shown in Figure 1(b).
Generally, the α-Al2O3 nanoplates are known to be synthesized by flux method requiring high reaction temperature to form crystalline α -phase and some additives such as Zn, Sn, and etc. are added in the reaction bath to promote the anisotropic growth[25]. The surface of α -Al2O3 nanoplates often has some impurities. Therefore, we etched the impurities with HCl aqueous solution to guarantee the uniform surface of the α-Al2O3 nanoplates.
To check the effect of HCl pre-treatment time on the uniformity and morphology of TiO2 shell, we changed the HCl pre-treatment time from 1 h to 2, 3, and 4 h while keeping the other experimental conditions constant. Figure 2 shows the SEM images of produced 2 α -Al2O3 nanoplates coated by TiO2 shell with different HCl pre-treatment time. The 1 h HCl pre-treated sample [Figure 2(a)] shows that most α-Al2O3 nanoplates were coated by TiO2 [upper inset in Figure 2(a)] but some nanoplates were not [lower inset in Figure 2(a)]. The SEM image of 2 h HCl pre-treated sample [Figure 2(b)] shows that all α-Al2O3 nanoplates were coated by TiO2 shell with relatively smoother roughness [insets in Figure 2(b)]. Over 3 h HCl pre-treated samples in Figure 2(c,d) show that all α-Al2O3 nanoplates were covered by TiO2 shell but the surface of TiO2 shell became more rough possibly due to the surface etching of α-Al2O3 nanoplates by HCl. Therefore, we fixed the HCl pre-treatment time to 2 h hereafter.
To form the crystalline TiO2 shell, TiCl4 precursor has been generally used because it forms stable TiOCl2 intermediate phase. The diluted TiOCl2 begins to form [Ti(OH)mCln]2- (m + n = 6) octahedral intermediate structure and the intermediate is polymerized by sol-gel reaction[ 26]. In more acidic condition under presence of Cl- ions, each octahedral intermediate is grown by sharing corner oxygen via condensation reaction due to the blocked corner by Cl- which will be further hydrolyzed and condensed so that the rutile crystalline phase is produced. Meanwhile, under less acidic condition the octahedral intermediate has more hydroxyl groups so that each octahedral intermediate tends to be grown by sharing edge, thereby producing anatase crystalline phase. The typical morphology of rutile and anatase nanoparticles are needle and spherical in shape. In terms of surface roughness, the formation of anatase phase TiO2 shell seems to be better than the rutile phase shell. However, the lattice parameter mismatch between α-Al2O3 nanoplates and rutile phase TiO2 is smaller than that of anatase TiO2 so that the rutile phase TiO2 crystal seems to be grown easily on the surface of α-Al2O3 nanoplates. Accordingly, if we grow the rutile phase TiO2 on the surface of α-Al2O3 nanoplates in early reaction stage and grow the anatase phase TiO2 in late stage, we can expect that the thickness and roughness of TiO2 shell on α-Al2O3 nanoplates can be controllable. Therefore, we checked the effect of TiCl4 to NaOH mole ratio on the morphology of TiO2 shell on α-Al2O3 nanoplates as summarized in Table 1. We fixed the HCl pre-treatment process and the moles of TiCl4 in the reactant solution, but changed the added moles of NaOH in each feeding step. The SEM images for the morphology of produced α-Al2O3 nanoplates coated by TiO2 with different mole ratio of TiCl4 and NaOH were shown in Figure 3. When the mole ratio of TiCl4 : NaOH is 1 : 1 (S8-1) and 1 : 1.5 (S8-2) as shown in Figure 3(a) and (b), some α-Al2O3 nanoplates were not fully coated by TiO2 shell and the coated ones had very rough TiO2 shells, thereby showing reduced lustering effect due to the random back scattering effect. Over 1 : 2 mole ratio of TiCl4 : NaOH, the surface of all α -Al2O3 nanoplates were fully coated by TiO2 shell. The 1 : 2 mole ratio of TiCl4 : NaOH sample (S8-3) still showed relatively rough TiO2 shell as shown in Figure 3(c). Over 1 : 2.5 mole ration of TiCl4 : NaOH samples (S8-4~7), apparently smoother TiO2 shell was formed as the mole ratio of TiCl4 : NaOH is more increased as shown in Figure 3(d~f) and Figure 2(b) due to the promoted formation of small spherical anatase TiO2 by the increased concentration of NaOH. The inset photographs in Figure 3(d~f) and Figure 2(b) indicate that the α -Al2O3 nanoplates with smoother TiO2 shell have stronger lustering effect (more shinny) due to the reduced random back scattering. Therefore, we could control the roughness of TiO2 shell on the α -Al2O3 nanoplates by controlling the mole ratio of TiCl4 : NaOH in reaction process and could find the condition for the formation of smooth TiO2 shell by adjusting the mole ratio of TiCl4 : NaOH, as well.
To apply the synthesized α-Al2O3 nanoplates coated by TiO2 shell as lustering effect pigments, the coated TiO2 shell should be tightly adhered on the α-Al2O3 nanoplates because the lustering pigments will be mechanically mixed with resins. To check the adhesiveness of TiO2 shell on the α-Al2O3 nanoplates, we sonicated the aqueous dispersion solution of α-Al2O3 nanoplates with TiO2 shell and compared the SEM images before and after sonication as shown in Figure 4. This clearly shows that some α-Al2O3 nanoplates with TiO2 shell have weak adhesiveness between TiO2 shell and α-Al2O3 nanoplates. Accordingly, additional process is required to improve the adhesiveness between TiO2 shell and α-Al2O3 nanoplates.
To improve the adhesiveness between TiO2 shell and α-Al2O3 nanoplates, we tried to examine the effect of heat-treatment because metal oxides tend to be sintered at high temperature. To check the morphological variation of TiO2 shell by heat-treatment, we intentionally used the α-Al2O3 nanoplates with rough TiO2 shell as a model system in order to catch the morphological changes distinctively. So we synthesized the model sample by stirring the solution mixture of 0.05 g α -Al2O3 nanoplates, 1 mL HCl solution, and 20 mL deionized water for 2 h and then reacting it at 98 ℃ for 2 h after adding 0.5 mL 3M TiCl4 aqueous solution into the solution mixture. The SEM image of model sample with rough TiO2 shell was shown in Figure 5(a). The magnified SEM surface image of the TiO2 shell indicates that the very rough needle shaped TiO2 is formed by the reaction. Apparently the heat-treated α-Al2O3 nanoplates with rough TiO2 shell at 500 ℃ did not show significant morphological difference, whereas the heat-treated sample at 700 ℃ showed significant morphological change due to crystal grain growth. Therefore, the heat-treatment at 500 ℃ seems more desirable to maintain the topology and roughness of α-Al2O3 nanoplates with TiO2 shell and to improve the adhesiveness between TiO2 shell and α -Al2O3 nanoplates. From the results of model experiment, we checked the SEM morphological variation of S8-7 sample before and after heat-treatment at 500 ℃ as shown in Figure 6(a) and (b). As expected from the model experiment, apparently there did not significant morphological difference with and without heat-treatment at 500 ℃. The magnified SEM images of their surface confirmed that the surface topology and roughness are not significantly changed by heat-treatment at 500 ℃. To check if the adhesiveness between TiO2 shell and α-Al2O3 nanoplates is improved or not, we sonicated the heat-treated S8-7 sample at 500 ℃ after dispersing the powder in deionized water. The SEM image of the sonicated sample in Figure 6(c) clearly indicates that TiO2 shells are tightly adhered on the surface of α-Al2O3 nanoplates, whereas the without heat-treated S8-7 sample exhibited weak adhesiveness between TiO2 shell and α-Al2O3 nanoplates. Accordingly, we could synthesize lustering pigment appearing metallic gold in color with robust TiO2 shell by heat-treatment.
The lustering effect pigments are generally formulated with organic resin to coat certain object so we compared the synthesized α-Al2O3 nanoplates with TiO2 shell with the commercialized effect pigment of α-Al2O3 nanoplates with TiO2 shell (MERCK, Xirallic T60-10SW). For comparison, we measured the UV-visible absorption spectra of the methyl orange/synthesized α-Al2O3 nanoplates with TiO2 shell (heat treated S8-7)/H2O solution and the methyl orange/commercialized effect pigment of α-Al2O3 nanoplates with TiO2 shell/H2O solution with UV exposure time (500 Watt) as shown in Figure 7(a) and (b). Here, we fixed the concentration of methyl orange and effect pigments and continuously stirred the solution to prevent the sedimentation of effect pigments in the solution during UV exposure. The normalized UV-visible absorption by methyl orange at its maximum absorption peak with UV exposure time in Figure 7(c) clearly indicate that the synthesized α-Al2O3 nanoplates with robust TiO2 shell (heat treated S8-7) have similar photo-stability against photo-degradation by photo-catalytic effect due to UV exposure compared with the commercialized effect pigment of α-Al2O3 nanoplates with TiO2 shell. This implies that the synthesized α-Al2O3 nanoplates with robust TiO2 shell can be applicable to effect pigments by mixing conventional organic resins.
4. Conclusions
We could successfully coat robust smooth TiO2 shell on the α- Al2O3 nanoplates by controlling the mole ratio of TiCl4 and NaOH and heat-treatment. Through the combination of the HCl pre-treatment for 2 h and the control of the mole ratio of TiCl4 and NaOH from 1 : 1 to 1 : 1.5, 1 : 2, 1 : 2.5, 1 : 3, 1 : 3.5, and 1 : 4, we found that more increase of the mole ratio of TiCl4 and NaOH leads the formation of smoother TiO2 shell due to the promoted formation of anatase crystalline TiO2 than the rutile one. Consequently, we could obtain the α-Al2O3 nanoplates with TiO2 shell having strong lustering effect by controlling the mole ratio of TiCl4 and NaOH due to the reduced random scattering by the formation of smooth TiO2 shell. Moreover, we could improve the adhesiveness between TiO2 shell and α-Al2O3 nanoplates by heat-treatment at 500 ℃ because the formed TiO2 nanocrystals in the shell can be sintered with each other while maintaining their topology and roughness, whereas the TiO2 shell sintered at 700 ℃ showed crystal grain growth causing significant topology and roughness change. Finally, we found that the synthesized α-Al2O3 nanoplates with robust TiO2 shell (heat treated S8-7) has similar photo- stability against UV exposure compared with the commercialized effect pigment of α-Al2O3 nanoplates with TiO2 shell.