Non-corroding α-alumina@TiO₂ core–shell nanoplates appearing metallic gold in colour

Abstract

Non-corroding α-alumina@TiO₂ core–shell nanoplates exhibiting a lustrous metallic gold colour were synthesized using sol–gel solution chemistry with controlled reaction conditions for the formation of anatase and rutile TiO₂ nanocrystals. Due to the small mismatch of lattice parameters between rutile TiO₂ and two-dimensional α-alumina nanoplates, the rutile TiO₂ could be conformably grown on the surface of the α-alumina nanoplates, however, the lustring effect was deteriorated by back scattering due to the formation of rough needle-shaped rutile TiO₂. The surface roughness of the TiO₂ nano-shell was controlled by controlling the pH of the reactant solution because rutile phase TiO₂ nanocrystals tend to be formed in more acidic conditions, whereas anatase phase TiO₂ is likely to be promoted in less acidic conditions, of which anatase small nanocrystals smooth the surface of the TiO₂ nano-shell during the crystal growth. The successful formation of a smooth TiO₂ nano-shell via controlled crystal growth was associated with the fact that the intermediate complex of [Ti(OH)_mCl_n]²⁻ (m + n = 6) tends to cause corner sharing in more acidic conditions and edge sharing in less acidic conditions. Apparently the α-alumina@TiO₂ core–shell nanoplates prepared in more and less acidic conditions showed a metallic gold colour, however, the core–shell nanoplates synthesized in less acidic conditions exhibited a better lustring effect than those from the more acidic conditions due to the formation of a smooth TiO₂ shell.

---

Introduction

Two-dimensional (2D) nanoplates such as graphene, metals, semi-conductors, and metal oxides have been of great interest because of their potential for application in flexible transparent electrodes,¹ barrier films,² thin film transistors,³ stretchable electrodes,⁴ and colour pigments.⁵ Among them, metal oxide 2D nanoplates including glass,⁶ mica,⁷ and alumina nanoflakes⁸ have been extensively utilized in the pigment industry because they can show a lustring colour and the colour can be controlled by the coating thickness and refractive index of shell material. Hence, the lustring colour is originated from the constructive and deconstructive interference of light at the air/shell/2D metal oxide nanoplate interface, thereby exhibiting angle dependent structure colouring like one-dimensional photonic crystals. For instance, if we assume that the incident light with θ angle from the normal direction is reflected at the shell (s)/2D metal oxide nanoplate (mo) interface, the constructive interference occurs when it satisfies following condition:⁹

2n_sd cos θ = mλ (1)

where, n_s is the refractive index of the shell, d is the thickness of the shell, m is an integer, and λ is the wavelength, respectively. The reflectance (R) can be written as the following equation:⁹where, n_a is the refractive index of the air, n_mo is the refractive index of the metal oxide nanoplate, and δ is 2πn_sd cos θ/λ, respectively. Therefore, the colour of the reflected light is dependent on the thickness (d), refractive index (n_s), and incident angle (θ) of the shell. Moreover, the intensity of the reflected light is closely related to the refractive index of each layer and the refractive index difference between each layer. In particular, here the n_a and n_m are fixed for air (n_a = 1.0) and α-alumina (n_mo = 1.7) so that the n_s is important for maximizing the lustring effect. Namely, higher refractive index materials such as crystalline TiO₂ are more desirable for shell materials to show a better lustring effect.

Initially, pearlescent 2D pigments were obtained from the fish scales of hairtails or herrings because the synthetic technology for artificial 2D pigments was not developed. Although crystalline 2D guanine or hypoxanthine natural pigments can be collected from fish scales, the yield was very low and insufficient to cope with the demands for pearlescent pigments. Therefore, artificial 2D pearlescent pigments such as basic lead carbonate (Pb(OH)₂·2PbCO₃), bismuth oxychloride (BiOCl), glass, and mica have been synthesized.⁵ Among them, the 2D mica pigments have been widely used until now because natural mica could be easily collected from mines. However, natural mica is often revealed to be of non-uniform colour or yellowish in colour because the composition of the impurities in natural mica differs depending on the location of the mine. Mica is composed of complex hydrous potassium-aluminium silicate and consequently has various names such as biotite, lepidolite, muscovite, phlogopite, and vermiculite according to its chemical composition.^7a Therefore, mica exhibits different colours according to the variation in its chemical composition. In addition, mica requires additives to coat TiO₂ on its surface and the control of the surface roughness of the TiO₂ shell is difficult, as well.^7c,d Accordingly, 2D α-alumina nanoplates will be better as a mother substrate than mica because α-alumina has a pure phase and small lattice mismatch with rutile TiO₂. Here, we try to coat a TiO₂ nano-shell on the surface of α-alumina nanoplates in order to synthesize non-corroding pigments with a metallic gold colour because TiO₂ is a very stable metal oxide which is not corroded by most acidic or basic chemicals.

Experimental

Materials

2D α-alumina nanoplates were purchased from TCERA. TiCl₄ (titanium tetrachloride), HCl, and NaOH were purchased from Fluka, Samchun, and Samchun, respectively. We used the chemicals as received without further purification.

Characterization

For the analysis using scanning electron microscopy (SEM: Supra 55, Carl Zeiss) images of the samples, we dropped a 2D α-alumina nanoplate, 2D α-alumina@TiO₂ core–shell nanoplate or TiO₂ nanoparticle dispersion solution on a Si wafer and dried it. The dried samples were then coated with Pt by sputtering. For the analysis using transmission electron microscopy (TEM: Jem-2100F, Jeol) images, we dropped the TiO₂ nano-sol solution on a carbon coated Cu grid. For the analysis using X-ray diffraction (XRD: D8 advance, Bruker) patterns, we collected the dried powder of the 2D α-alumina nanoplates or 2D α-alumina@TiO₂ core–shell nanoplates or TiO₂ nanoparticles and placed it on a Si low background sample holder. The samples were scanned from 20° to 80° at a scan rate of 6° min⁻¹ under the irradiation of Cu K-α (λ = 0.15406 nm).

Results and discussion

The scanning electron microscopy (SEM) image of the 2D α-alumina nanoplates in Fig. 1(a) indicates that the lateral size and thickness of the nanoplates is 10–30 μm and ∼500 nm, respectively. Fig. 1(b) is a magnified SEM image of the white rectangle area in Fig. 1(a), which indicates that the surface of the nanoplate is uniform. This indicates that the 2D nanoplates are adaptable for use as substrates for interference pearlescent pigments. Fig. 1(c) is a photograph of the 2D α-alumina nanoplates which were prepared by dropping a 2D α-alumina nanoplate/water dispersion solution on a glass plate and subsequent drying. The pristine 2D α-alumina nanoplates exhibited a metallic silver colour. Fig. 1(d) shows X-ray diffraction (XRD) patterns of the 2D nanoplates, indicating that the nanoplates have a crystalline α-alumina phase.

Fig. 1 (a) A scanning electron microscopy (SEM) image of the 2D α-alumina nanoplates (TCERA), (b) magnified SEM image of the white rectangle part in (a), (c) photograph of the 2D α-alumina nanoplates, and (d) X-ray diffraction (XRD) patterns: red peak = sample, blue line = JCPDS# 82-1468.

Table 1 summarizes the material properties of crystalline α-Al₂O₃, rutile TiO₂, and anatase TiO₂ indicating that the lattice parameter of rutile TiO₂ is slightly mismatched with the lattice parameter of α-Al₂O₃, whereas the anatase TiO₂ has great lattice mismatch with the α-Al₂O₃. In addition, the refractive index of rutile TiO₂ is higher than for anatase TiO₂ so the lustring effect is more enhanced by rutile TiO₂. Therefore, here, we try to coat a rutile TiO₂ nano-shell on the surface of 2D α-alumina nanoplates in order to create a pearlescent metallic gold colour.

Table 1 Summary of material properties of crystalline α-Al₂O₃, rutile TiO₂, and anatase TiO₂

Properties	α-Al2O3	Rutile	Anatase
Crystal system	Hexagonal	Tetragonal	Tetragonal
a	4.75 Å	4.59 Å	3.78 Å
b	—	4.59 Å	3.78 Å
c	12.99 Å	2.96 Å	9.51 Å
Density	3.95–4.03	4.3	3.8
Refractive index	1.7	2.7	2.5

---

To find the formation conditions of rutile TiO₂ in acidic conditions, we conducted experiments adding 0.5 mL of 3 M TiCl₄ aqueous solution and a HCl solution (0, 0.1, 0.2, 0.5, and 1 mL) into 20 mL of deionized water in a vial and subsequently reacting them at 60 °C for 1 h under magnetic stirring. The XRD patterns of the produced TiO₂ particles in Fig. 2(a) show that the pure rutile TiO₂ phase is formed in more acidic conditions (HCl = 0.5 and 1 mL under the current experimental conditions) and the anatase TiO₂ phase is dominantly formed in less acidic conditions (HCl = 0.2, 0.1 and 0 mL under the current experimental conditions). This tendency is well known because 3 M TiCl₄ aqueous solution forms stable TiOCl₂ and the diluted TiOCl₂ begins to form [Ti(OH)_mCl_n]²⁻ (m + n = 6), which tends to create corner sharing by condensation reaction in more acidic conditions due to more existence of Cl⁻, whereas it creates edge sharing in less acidic conditions.¹⁰ Fig. 2(b) is the TEM (transmission electron microscopy) image of the produced TiO₂ particles from less acidic conditions (HCl = 0 mL) indicating that small anatase TiO₂ nanocrystals with ∼5 nm in size were synthesized. Fig. 2(c) is the SEM image of the produced TiO₂ particles from more acidic conditions (HCl = 1 mL) indicating the formation of relatively larger needle-shaped rutile TiO₂ primary nanoparticles with ∼10 nm in size. These results imply that the rutile TiO₂ nano-shell can be directly coated on the 2D α-alumina nanoplates due to a small mismatch of the lattice parameter in more acidic conditions, but a rough surface will be formed. Meanwhile a smooth surface can be formed in less acidic conditions due to the formation of small anatase TiO₂ nanocrystals. Therefore, the coating process needs to be started in more acidic conditions at early reaction stages and finished in less acidic conditions at late stages in order to coat a smooth TiO₂ nano-shell with small surface roughness on the 2D α-alumina nanoplates. To confirm the crystal size of the synthesized TiO₂ nanoparticles, we calculated the representative crystal size of the anatase TiO₂ (HCl – 0 mL) and rutile TiO₂ (HCl – 1 mL) samples from the Scherrer equation:¹¹

where t is the nanocrystal size, λ is the wavelength of the X-ray irradiation (0.154 nm), and B is the line width at half maximum (in radians). For the calculation of the crystal size of anatase TiO₂ and rutile TiO₂, we used the (101) peak at 2θ = 25.3°, and the (110) peak at 2θ = 27.5°, respectively. The calculated crystal size of the anatase TiO₂ and rutile TiO₂ was ∼4 nm and ∼8 nm, respectively, which is well matched with the TEM and SEM results.

Fig. 2 (a) XRD patterns of the synthesized crystalline TiO₂ particles via sol–gel method with different acidic conditions. Reaction conditions: 60 °C for 1 h with a solution mixture of 20 mL deionized water, 0.5 mL 3 M TiCl₄ aqueous solution, and HCl aqueous solution – x mL (x = 0, 0.1, 0.2, 0.5, and 1), (b) TEM (transmission electron microscopy) image of the produced TiO₂ (HCl – 0 mL) from less acidic conditions, and (c) SEM image of the produced TiO₂ (HCl – 1 mL) from more acidic conditions.

To check if the rutile TiO₂ can be directly coated on the 2D α-alumina nanoplates, we coated TiO₂ nano-shells in more acidic conditions by adding 0.05 g of the 2D α-alumina nanoplates, 1 mL of HCl solution, and 0.5 mL of 3 M TiCl₄ aqueous solution into 20 mL of deionized water in a vial and subsequently heating at 98 °C for 2 h under magnetic stirring. To check the effects of pre-treatment of the 2D α-alumina nanoplates in acidic solution, we conducted again the above experiment by stirring a solution mixture of 0.05 g of the 2D α-alumina nanoplates, 1 mL of HCl solution, and 20 mL of deionized water for 2 h and then reacting it at 98 °C for 2 h after adding 0.5 mL of 3 M TiCl₄ aqueous solution into the solution mixture. The SEM images of the 2D α-alumina nanoplates coated by TiO₂ with and without HCl pre-treatment are shown in Fig. 3. They clearly indicate that the HCl pre-treatment is effective for forming a full covering of TiO₂ nano-shell on the surface of the 2D α-alumina nanoplates because all the 2D α-alumina nanoplates were coated by TiO₂ nano-shells in the HCl pre-treated sample (Fig. 3(a)), whereas some 2D α-alumina nanoplates were not fully covered by TiO₂ nano-shells in the sample without HCl pre-treatment (Fig. 3(b) and see sample area marked by arrows). The inset in Fig. 3(b) is a magnified SEM image of the uncovered part indicating the existence of incomplete TiO₂ coating.

Fig. 3 SEM images of the 2D α-alumina nanoplates coated by TiO₂ (a) with and (b) without HCl pre-treatment: the arrows indicate the parts of the 2D α-alumina nanoplates not covered by TiO₂ and the inset is a magnified image of the uncovered part.

To provide a metallic gold structure colouring effect to the 2D α-alumina nanoplates, we formed a TiO₂ rutile and anatase mixture coating layer on the surface of the 2D α-alumina to see if a smoother TiO₂ shell can be formed or not. For this, we pre-treated the 2D α-alumina nanoplates by stirring a 0.05 g α-alumina nanoplates/20 mL deionized water/1 mL HCl solution mixture for 2 h and adding 0.5 mL of 3 M TiCl₄ solution into the solution at 98 °C. 0.1 mL of 0.8 M NaOH aqueous solution was added into the reactant solution 25 times consecutively and the mixture was reacted for 2 h. Then 0.1 mL of 0.8 M NaOH aqueous solution was again added into the reactant solution 25 times consecutively and the mixture was reacted for 2 h. The SEM image of the synthesized 2D α-alumina@TiO₂ core–shell nanoplates is shown in Fig. 4(a) indicating that a smoother TiO₂ shell was formed using the modified reaction than the TiO₂ shell formed in Fig. 3(a) because the rough needle-shaped rutile TiO₂ shell is likely to be formed in more acidic conditions, whereas the smooth particulate anatase TiO₂ shell tends to be co-formed with rutile TiO₂ during pH adjustment by the addition of NaOH. Therefore, rutile TiO₂ might be preferentially grown on the surface of the 2D α-alumina nanoplates and the mixture of rutile and anatase TiO₂ might be co-grown on the TiO₂ layer, thereby reducing the surface roughness of the TiO₂ shell. The magnified SEM image in Fig. 4(b) shows that the surface morphology is composed of needle-shaped nanoparticles and irregular smaller nanoparticles. The photograph of the dried 2D α-alumina@TiO₂ core–shell nanoplates is shown in Fig. 4(c) confirming the successful formation of a TiO₂ nano-shell on the 2D α-alumina nanoplates due to the appearance of gold structure colouring, whereas the pristine 2D α-alumina nanoplates exhibit a silver colour. Fig. 4(d) is the corresponding XRD pattern of the 2D α-alumina@TiO₂ core–shell nanoplates indicating that the TiO₂ shell is composed of a mixed phase of rutile and anatase TiO₂ nanocrystals as we intended. The calculated crystal size of the anatase and rutile TiO₂ in the shell from the Scherrer equation was ∼8 nm and ∼16 nm, respectively. The crystalline phase composition of the TiO₂ shell can be calculated from the integrated intensity of the (101) peak for anatase and the (110) peak for rutile using the following equation:¹²

where W_R is the weight (mass) fraction of the rutile phase, A_A is an integrated peak intensity of the (101) anatase phase, and A_R is for the (110) rutile phase. The calculated W_R was ∼0.30 implying that the dominant crystalline phase is anatase which might help formation of a smooth surface, but a greater fraction of anatase phase might be required to obtain a much smoother surface. The relatively rough TiO₂ shell seems to back scatter the incident light so that the intense structure colouring effect might be weakened, thereby exhibiting a relatively weak lustring effect of gold colour.

Fig. 4 (a and b) SEM images of the 2D α-alumina@TiO₂ core–shell nanoplates prepared in more acidic conditions, (b) magnified SEM image of the top of the TiO₂ shell, (c) photograph of dried 2D α-alumina@TiO₂ core–shell nanoplates on a glass substrate, and (d) their XRD pattern (reaction conditions/procedure: 0.05 g α-alumina nanoplates/20 mL deionized water/1 mL HCl solution/0.5 mL of 3 M TiCl₄ solution/2.5 mL of 0.8 M NaOH aqueous solution/2.5 mL of 0.8 M NaOH aqueous solution).

Although the surface roughness of the TiO₂ shell can be reduced using the reaction design for co-formation of a rutile and anatase TiO₂ shell, the surface roughness needs to be more smooth in order to enhance the lustring effect. Hence, we increased the amount of added NaOH in the reaction and additionally added the Ti precursor under less acidic conditions to promote the formation of anatase TiO₂ nanocrystals. For this, we stirred a 0.05 g α-alumina nanoplates/20 mL deionized water/1 mL HCl solution mixture for 2 h for the pre-treatment process and 0.25 mL of 3 M TiCl₄ aqueous solution was added into the reactant solution at 98 °C. 0.1 mL of 0.8 M NaOH aqueous solution was added into the reactant solution 37 times consecutively and the mixture was reacted for 1 h. Then 0.25 mL of 3 M TiCl₄ aqueous solution was added and subsequently 0.1 mL of 0.8 M NaOH aqueous solution was again added into the reactant solution 37 times consecutively and the mixture was reacted for 1 h. Finally, we added again 0.25 mL of 3 M TiCl₄ aqueous solution into the reaction solution and added 37 times by consecutive addition 0.1 mL of 0.8 M NaOH aqueous solution. The reaction was finished after additional reaction for 2 h. The SEM image of the produced 2D α-alumina@TiO₂ core–shell nanoplates is shown in Fig. 5(a) and its magnified image is shown in Fig. 5(b) indicating that a smoother TiO₂ shell was formed using the modified reaction process. The photograph of the dried 2D α-alumina@TiO₂ core–shell nanoplates in Fig. 5(c) exhibited a more intense gold colour due to the reduced back scattering. Fig. 5(d) confirms that the composition of the TiO₂ shell is composed of anatase and rutile phases. As we expected, the calculated W_R is ∼0.22 indicating that crystalline anatase TiO₂ is formed more than in the above experimental result (see Fig. 4). Accordingly, the smoother TiO₂ nano-shell was formed by the greater formation of anatase TiO₂ nanocrystals. As expected, the synthesized 2D α-alumina@TiO₂ core–shell nanoplates were not corroded by acidic (except HF solution) or basic chemicals because the TiO₂ shell is a very stable metal oxide.

Fig. 5 (a and b) SEM images of the 2D α-alumina@TiO₂ core–shell nanoplates prepared in less acidic conditions, (b) magnified SEM image of the top of the TiO₂ shell, (c) photograph of dried 2D α-alumina@TiO₂ core–shell nanoplates on a glass substrate, and (d) their corresponding XRD pattern (reaction conditions/procedure: 0.05 g α-alumina nanoplates/20 mL deionized water/1 mL HCl solution/0.25 mL of 3 M TiCl₄ solution/3.7 mL of 0.8 M NaOH aqueous solution/0.25 mL of 3 M TiCl₄ aqueous solution/3.7 mL of 0.8 M NaOH aqueous solution/0.25 mL of 3 M TiCl₄ aqueous solution/3.7 mL of 0.8 M NaOH aqueous solution).

Conclusions

We could successfully synthesize non-corroding 2D α-alumina@TiO₂ core–shell nanoplates showing a metallic gold colour via sol–gel solution chemistry with controlled reaction conditions for the formation of anatase and rutile TiO₂ nanocrystals. When we used a 0.5 mL of 3 M TiCl₄ aqueous solution/0, 0.1, 0.2, 0.5, or 1 mL of HCl solution/20 mL deionized water solution mixture, the rutile phase of TiO₂ was formed in more acidic conditions (HCl = 0.5 and 1 mL) and the anatase phase of TiO₂ was formed in less acidic conditions (HCl = 0.2, 0.1 and 0 mL) because the [Ti(OH)_mCl_n]²⁻ (m + n = 6) intermediate complex tends to form corner sharing in more acidic conditions due to the greater existence of Cl⁻ and creates edge sharing in less acidic conditions. Due to the small lattice mismatch between rutile TiO₂ and the α-alumina nanoplates, rutile TiO₂ could be conformably grown on the surface of the α-alumina nanoplates. However, the rutile TiO₂ shell exhibited a rough surface due to the formation of rough needle-shaped crystals and consequently reduced the lustring effect due to a back scattering effect. The surface roughness of the TiO₂ nano-shell could be further reduced using a controlled reaction for co-formation of rutile and anatase TiO₂ nanocrystals during the TiO₂ coating. The successful coating of a smooth TiO₂ nano-shell on the 2D α-alumina nanoplates was associated with the fact that the rutile TiO₂ nanocrystals are preferentially coated on the 2D α-alumina nanoplates and the smaller irregular sphere-shaped anatase TiO₂ nanoparticles are co-formed on the surface of the α-alumina while smoothing the surface roughness of the TiO₂ nano-shell. We believe that this controlled reaction method to form a smooth inorganic TiO₂ nano-shell is helpful for the synthesis of opalescent colour pigments, angle dependent structure colour pigments, safety pigments, and cosmetic pearls.

Acknowledgements

This study was supported by the Basic Science Research Program (no. 2014R1A5A1009799) through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning.

References

1. (a) K. S. Kim, Y. Zhao, H. Jang, S. Y. Lee, J. M. Kim, K. S. Kim, J.-H. Ahn, P. Kim, J.-Y. Choi and B. H. Hong, Nature, 2009, 457, 706; (b) S. Bae, H. Kim, Y. Lee, X. Xu, J.-S. Park, Y. Zheng, J. Balakrishnan, T. Lei, H. R. kim, Y. I. Song, Y.-J. Kim, K. S. Kim, B. Özyilmaz, J.-H. Ahn, B. H. Hong and S. Ijima, Nat. Nanotechnol., 2010, 5, 574; (c) J. Wu, H. A. Becerril, Z. Bao, Z. Liu, Y. Chen and P. Peumans, Appl. Phys. Lett., 2008, 92, 263302.
2. (a) H. Kim, Y. Miura and C. W. Macosko, Chem. Mater., 2010, 22, 3441; (b) Y. Su, V. G. Kravets, S. L. Wong, J. Waters, A. K. Geim and R. R. Nair, Nat. Commun., 2014, 5, 4843.
3. (a) B. Radisavljevic, A. Radenovic, J. Brivio, V. Giacometti and A. Kis, Nat. Nanotechnol., 2011, 6, 147; (b) S.-K. Lee, H. Y. Jang, S. Jang, E. Choi, B. H. Hong, J. Lee, S. Park and J.-H. Ahn, Nano Lett., 2012, 12, 3472; (c) C. Yan, J. H. Cho and J.-H. Ahn, Nanoscale, 2012, 4, 4870.
4. (a) G. D. Moon, G.-H. Lim, J. H. Song, M. Shin, T. Yu, B. Lim and U. Jeong, Adv. Mater., 2013, 25, 2707; (b) Y. G. Seol, T. Q. Trung, O.-J. Yoon, I.-Y. Sohn and N.-E. Lee, J. Mater. Chem., 2012, 22, 23759.
5. M. Debeljak, A. Hladnik, L. Cerne and D. Gregor-Svetec, Color Res. Appl., 2012, 38, 168.
6. G. Pfaff, Inorg. Mater., 2003, 39, 123.
7. (a) V. Štengl, J. Šubrt, S. Bakardjieva, A. Kalendova and P. Kalenda, Dyes Pigm., 2003, 58, 239; (b) P. M. T. Cavalcante, M. Dondi, G. Guarini, F. M. Barros and A. B. Luz, Dyes Pigm., 2007, 74, 1; (c) M. Ren, H. Yin, A. Wang, T. Jiang and Y. Wada, Mater. Chem. Phys., 2007, 103, 230; (d) Q. Gao, X. Wu, Y. Fan and X. Zhou, Dyes Pigm., 2012, 95, 534.
8. (a) S. Teaney, G. Pfaff and K. Nitta, Eur. Coat. J., 1999, 4, 90; (b) S. Sharrock and N. Schuel, Eur. Coat. J., 2000, 1/2, 20.
9. E. Hecht, Optics, Pearson Education, Inc., Addison Wesley, 4th edn, 2001, ch. 4 and 9.
10. H. Cheng, J. Ma, Z. Zhao and L. Qi, Chem. Mater., 1995, 7, 663.
11. B. D. Cullity and S. R. Stock, Elements of X-ray diffraction, Pearson, 2001.
12. (a) J. Yu, Y. Su and B. Cheng, Adv. Funct. Mater., 2007, 17, 1984; (b) T. H. Han, H.-S. Moon, J. O. Hwang, S. I. Seok, S. H. Im and S. O. Kim, Nanotechnology, 2010, 21, 185601.

---

Footnote

† M. W. Suh and S. J. Lee contributed equally to this work.

---