Facile solvothermal synthesis of high refractive index ZrO₂ spheres: estimation of the enhanced light extraction efficiency

Abstract

Sub-micron sized ZrO₂ spheres were synthesized by a facile solvothermal reaction method. The influence of Zr(OPr)₄ concentration on the size and size distribution of ZrO₂ spheres were examined by taking the scanning electron microscope and transmission electron microscope images. When the concentration of Zr(OPr)₄ was increased from 0.031 to 0.045 mol, the average size of the spheres also increased from 305 to 920 nm. The diffused reflectance spectra of the calcined ZrO₂ spheres exhibited the higher reflectivity than that of as-prepared ZrO₂ spheres. As compared to the bare LEDs, light extraction efficiency was improved to 23% when the ZrO₂ spheres were coated on the ITO surface, however, 70.12% was improved when the ZrO₂ spheres were coated on the GaN surface due to the spherical morphology induced photon escape cone enhancement.

---

Introduction

Self-assembly is a promising method to fabricate monodispersed spheres based photonic crystals in the visible and near-infrared wavelength ranges.^1–3 Particularly, sub-micron sized monodispersed spheres are necessary for their applications in visible range photonics. Silicon dioxide (SiO₂), polystyrene (PS) and poly(methyl methacrylate) (PMMA) spheres are widely used to assemble the opal-like photonic crystals because of their ease of synthesis with monodispersion properties and a size distribution of less than 5%.^4–7 Monodispersed silica colloidal spheres with tetraethyl orthosilicate (TEOS) hydrolyzing methods were first prepared by Stöber et al. in 1968.⁸ Polymer colloidal spheres such as PS and PMMA have been produced by an emulsion polymerization method.⁹ However, the properties of the prepared photonic crystals by using these colloidal spheres are not shown to be impressive due to the low refractive indices of SiO₂ (refractive index (n) = 1.45), PS (n = 1.58) and PMMA (n = 1.49).^10–12

To overcome the low refractive index problem, one of the ways is to synthesize the spheres (without visible light absorption) with high refractive index because the high refractive index spheres scatters more light at the large scattering angles than that of low refractive index spheres.¹³ By the literature survey, many research methods were developed to synthesize the submicron spheres with high refractive index. Jiang et al.¹⁴ reported the synthesis of sub-micron size TiO₂ spheres (200–500 nm) via co-precipitation method and used the calcination process over 500 °C. However, the structure of as-prepared TiO₂ sphere (n = 2.488 (anatase), and n = 2.6 (rutile)) is amorphous. After annealed at 500 °C, TiO₂ spheres were crystallized into anatase phase but the refractive index reduced to 2.06. The lower refractive index is due to the mesoporous nature of TiO₂ spheres, which yields the less scattering of light. Wang et al.¹⁵ reported the synthesis of sub-micron size ZnO spheres (200–500 nm) by using Nd:YAG laser, however, ZnO spheres could not be synthesized in large scale.

Among the various sizes of metallic spheres, high refractive index material such as monodispersed submicron ZrO₂ spheres (n = 2.0) are of particular interest and comparative particle size to an optical wavelength, which make them ideal candidates for photon-related application such as solar cells and photonic crystals. Unfortunately, only few researchers have been able to prepared the monodispersed ZrO₂ spheres with submicron size. Widoniak et al. reported the synthesis of ZrO₂ spheres with 0.2–2 μm by a sol–gel process.^16,17 In this method, ZrO₂ particles were synthesized by adding various salts or polymers. The structure of as-prepared ZrO₂ particles is amorphous. After annealed at 450 °C, they are crystallized into tetragonal phase. In this context, we have focused on the development of high refractive index ZrO₂ spheres by a facile route.

Herein, we reported a more convenient and cost effective method to fabricate the monodispersed ZrO₂ spheres with sub-micro meter sized diameter by solvothermal method. This process facilitate the direct formation of cubic phase ZrO₂ submicron spheres. Subsequently, the sizes of spheres were modified with a functional factor of the Zr(OPr)₄ precursor concentration. Furthermore, light extraction efficiency and far-field emission distributions of LED coated with ZrO₂ spheres were carried out by LightTools simulation process.

Experimental

In a typical procedure, all reagents were taken without any further purification. 0.031 mol L⁻¹ of zirconium propoxide Zr(OPr)₄ dissolved in 550 mL of 2-propanol to form solution 1, and solution 2 was prepared by dissolving the 0.018 mol L⁻¹ of HNO₃ in 320 mL ethanol. The solutions were shaken with ultra-sonication for 5 min to dissolve the solutes and then the solutions were independently stirred with a magnetic stirrer until the homogeneous solution (1 day) was formed. The solution 2 was then slowly added drop-wise into solution 1 with magnetic stirring and then the magnetic stirring of the mixed solution was continued until the homogeneous solution (approximately 1 day) was formed. The homogeneous solution was then transferred into a Teflon liner (1730 mL capacity and 50% filling) and was placed into stainless steel autoclave. The temperature of the autoclave was increased to 200 °C at a rate of 2 °C min⁻¹ and held at this temperature for 30 min with continued magnetic stirring. The temperature of the autoclave was then raised to 290 °C at a rate of 1 °C min⁻¹, and held at this temperature for 15 h. After gradually cooling down to room temperature, the precipitate was separated and then dried at 50 °C for a day in the ambient atmosphere. The experiment was repeated without changing the parameters except Zr(OPr)₄ concentration. Additionally, ZrO₂ spheres were annealed at 500 °C. The temperature was increased at a rate of 4 °C min⁻¹.

The phase purity of the synthesized ZrO₂ spheres were analyzed with a powder X-ray diffraction (XRD) using a X'PERT PRO diffractometer with a CuKα = 1.5406 Å and beam voltage of 40 kV and current 30 mA. The morphology and sizes of the spheres were examined by using a field emission scanning electron microscopy (FE-SEM), model JEOL JSM-6700 FESEM. Osmium coating was sprayed on the sample surfaces by using Hitachi fine coat ion sputter E-1010 unit to avoid possible charging of specimens before FESEM observation was made on each time. Transmission electron microscopy (TEM) studies were carried out using a JEOL JEM-2010 operating at an accelerating voltage of 200 kV. The sample was prepared by dispersing the ZrO₂ powder in ethanol and the dispersion solution was drop on a carbon coated cupper grid and then finally dried into ambient atmosphere. The reflectance spectra were measured by using JASCO V-670 UV-visible spectrophotometer.

Results and discussion

The morphology and structure of ZrO₂ spheres

The morphology and phase formation of the as-synthesized and calcined ZrO₂ spheres were characterized using XRD, SEM, TEM, and selected area electron diffraction (SAED) pattern, as shown in Fig. 1. The XRD profiles of the as-prepared (prepared in 870 mL with a ratio of 5 : 3 (2-propanol : ethanol), 0.031 mol L⁻¹ of Zr(OPr)₄, and 0.018 mol L⁻¹ of HNO₃ at 290 °C and calcined at 500 °C) samples were shown in Fig. 1(a). All the diffraction peaks are in good agreement with the cubic phase of ZrO₂ (JCPDS file no. 03–0640). In particular, the broad peaks of (111), (200), (220), (311) and (222) reflect the nanocrystalline nature. Based on the Scherrer's equation, the average crystallite sizes of ZrO₂ spheres before and after calcined at 500 °C are about 3.67 and 5.08 nm, respectively. Fig. 1(b) shows the SEM image of the sample, which exhibited that the product is composed of a large amount of sub-microspheres. Most of the sub-microspheres exhibits monodispersion characteristics. The average diameter of the sub-microspheres is about 305 nm with a diameter distribution from 200 to 400 nm and an approximately 15% standard deviation. Fig. 1(c) and (d) shows the SEM images and Fig. 1(e) and (f) shows the TEM image of ZrO₂ spheres before and after calcining at 500 °C, respectively. The diameter of annealed ZrO₂ spheres was reduced by ∼2%. TEM images displayed well defined monodispersed sub-micron spheres with an average diameter of about 300 nm, which is in good agreement with SEM observation. The corresponding SAED patterns are shown in the Fig. 1(g) and (h), which exhibited the polycrystalline characteristics by displaying the ring patterns of ZrO₂, and all the diffraction rings form the cubic phase of ZrO₂ with (111), (200), (220) and (311) planes. The SAED patterns show relative change indicated by the slight increase in spots for the spheres annealed at 500 °C.

Fig. 1 (a) XRD patterns, (b–d) typical SEM images, (e and f) TEM images, (g and h) SAED patterns taken from ZrO₂ spheres. [(b, c, e, and g) images corresponding to as-prepared ZrO₂ spheres and (d, f, and h) images correspond to ZrO₂ spheres annealed at 500 °C].

In order to better understanding the effect of Zr(OPr)₄ concentration on the size and distribution of ZrO₂ spheres, the precursor (starting material)-dependent experiments were carried out with a fixed HNO₃ (0.018 mol L⁻¹), 2-propanol (550 mL), and ethanol (320 mL) under the same condition and resultant products were analyzed by the SEM and standard deviation analysis. SEM images and size distributions of the ZrO₂ spheres of various Zr(OPr)₄ concentration from 0.031 to 0.045 mol L⁻¹ are shown in Fig. 2. All the particles are dispersed relatively well and have an average sizes from 305 to 920 nm. It can be easily tuned the diameter of the ZrO₂ spheres from 305 to 920 nm by varying the precursor concentration within the range from 0.031 to 0.045 mol L⁻¹, which is calculated as the final concentration of the precursor in the mixed solvent of 2-propanol and ethanol. The details of sphere size and the concentration of precursors were represented in Fig. 2. When the concentrations of Zr(OPr)₄ are 0.031 and 0.045 mol L⁻¹, the sphere sizes are about 305 nm and 920 nm, respectively. The size of the ZrO₂ spheres were increased from 305 to 920 nm with increasing the concentration of precursor from 0.031 to 0.045 mol L⁻¹. Without changing the current procedure, use of solutions by varying the precursor concentration outside this region often led to the formation of polydispersed products. As the initial concentration of Zr(OPr)₄ becomes lower than 0.031 mol, nanoparticles (no spheres was observed) were obtained. As the initial concentration of Zr(OPr)₄ becomes greater than 0.045 mol, irregular sized spheres with sub-micron and micron sizes were obtained. The details of spheres sizes and concentration of precursors were represented in Fig. 3. The sizes of the ZrO₂ spheres were increased with increasing the concentration of precursor. Their standard distribution is about 15%.

Fig. 2 SEM images and size distribution analysis of the ZrO₂ spheres prepared at various concentrations of precursor: (a) 0.031 (size = 305 nm), (b) 0.033 (size = 370 nm), (c) 0.035 (size = 450 nm), (d) 0.037 (size = 580 nm), (e) 0.039 (size = 690 nm), (f) 0.041 (size = 740 nm), (g) 0.043 (size = 890 nm), (h) 0.045 (size = 920 nm) mol L⁻¹ (All the scale bars are 1 μm).

Fig. 3 Zr(OPr)₄ precursor concentration dependent sizes of ZrO₂ spheres.

In order to gather the optical properties, diffused reflectance measurements were carried out. Fig. 4 presents the UV-vis reflectance spectra of ZrO₂ spheres obtained before and after calcination process. As prepared ZrO₂ spheres with 305 nm size had higher diffused reflection capabilities in the near UV, visible and near infrared regions (from 350 nm to 1 μm). This phenomenon indicates that the incident light was significantly scattered within the powder of ZrO₂ spheres since the size of the sphere comparable to the wavelength of visible light. As compared with as-prepared ZrO₂ spheres, ZrO₂ calcined at 500 °C has higher diffused reflection capability between 350 nm and 1 μm region. While a red shift in the edge of reflection, indicating that the refractive index is higher than that of as prepared ZrO₂ spheres.¹⁴

Fig. 4 Diffused reflectance spectra of (a) as prepared 305 nm ZrO₂ spheres, (b) after calcining at 500 °C.

Light extraction simulation

To investigate the enhancement of light extraction efficiency of LEDs, LightTools experiments were carried out. Different sizes of monolayered ZrO₂ spheres coated on the top surface of ITO. The ZrO₂ spheres arrays were considered as a layer for enhancement of light extraction efficiency, as shown in Fig. 5. Where MQW is multi quantum wall and D is diameter of ZrO₂ spheres. The LED (16 μm × 16 μm × 0.7 μm: length × width × thickness) is set with 0.2 μm thick n-GaN layer, 0.1 μm MQW, 0.2 μm thick p-GaN layer and 0.2 μm thick ITO layer. In factor of the simulation, input flux was fixed at 100 lm, and 2.50 (GaN), 1.90 (ITO) and 2.00 (ZrO₂) were used as the refractive indices. The total flux of GaN → air is 12.394 lm and GaN → ITO → air is 13.247 lm. The relevant total flux of LEDs with variation size of ZrO₂ spheres is shown in Fig. 6. When the size of ZrO₂ spheres increases from 300 nm to 1 μm, the total flux decreases. Because the contact area between ITO top surface and ZrO₂ spheres on LEDs decreases with increasing the size of ZrO₂ spheres. Even though the total flux is reduced with increasing a size of ZrO₂ spheres, however, the observed total flux is better than that of GaN/air interface. The significant increase in the total flux of the LEDs with ZrO₂ sphere arrays can be attributed presumably due to the increase in its effective photon escape cone. In addition to the increase in its effective photon escape cone, the spherical morphology also can be attributed.

Fig. 5 Schematic diagram for LightTools simulation structure of ZrO₂ spheres addressed on GaN based LED.

Fig. 6 Total flux of ZrO₂ sphere arrayed LED prepared at various concentrations of Zr(OPr)₄ precursor.

Fig. 7 shows angular dependent emission intensity and luminance of the LEDs with ZrO₂ spheres, which was examined by LightTools simulations. The emission patterns were shown lambertian distribution for GaN LED and GaN LED with ITO. When the 300 nm sized ZrO₂ spheres coated on the ITO top surface, the emission pattern was narrow. With increasing the size of ZrO₂ spheres, the emission patterns lead to the lambertian distribution. When 300 nm sized ZrO₂ spheres put on the ITO top surface, the luminance exhibited the maximum brightness. When the size of ZrO₂ spheres increases from 300 nm to 1 μm, the luminance decreases, which might be due to the lambertian distribution of emission pattern. The above results suggested that the performance of LED has been enhanced when the ZrO₂ spheres were coated on the ITO top surface, and the optimum size of ZrO₂ spheres was found to be 300 nm. Fig. 8 shows the total flux, angular dependent emission intensity and luminance of the LEDs with/without ZrO₂ spheres on the top surface of GaN and ITO. The total flux of GaN → air, GaN → ITO → air, GaN → ITO → ZrO₂ → air and GaN → ZrO₂ → air is 12.394, 13.247, 15.094 and 22.537 lm, respectively. When the ZrO₂ spheres coated on top surface of GaN, total flux was increased to 70.12% and far field emission pattern was shown spot distribution. Although the total flux of the LED based on the ZrO₂ spheres coated on the top surface of ITO was less than that of the LED based on the ZrO₂ spheres coated on the top surface of GaN. The total flux of the LED was increased to 23% for the ZrO₂ spheres coated on the top surface of ITO and far-field emission pattern was shown middle of lambertian and spot distributions. The escape cone for the GaN/ZrO₂ interface is larger than that of GaN/ITO interface due to the smaller difference of refractive index of GaN(2.5)/ZrO₂(2.0) as compared with GaN/ITO(1.9). Although the critical angle of ZrO₂/air interface was smaller than that of ITO/air interface, spherical morphology of ZrO₂ spheres is helpful to escape of light. The above results suggested that the performance of LEDs has been enhanced when 300 nm sized ZrO₂ spheres coated on the GaN or ITO top surface.

Fig. 7 Angular dependent emission intensity and luminance of ZrO₂ sphere arrayed LED.

Fig. 8 Total flux, angular dependent emission intensity and luminance of ZrO₂ sphere arrayed LED.

Conclusion

The fabrication of monodispersed submicron ZrO₂ spheres by a facile solvothermal reaction method was demonstrated. By this method, ZrO₂ spheres with uniform diameter within the range of 305–920 nm were synthesized conveniently. The sizes of the ZrO₂ spheres were controlled by varying the concentration of precursor. After calcined at 500 °C, the density as well as refractive index of spheres were increased. The diffused reflectance spectra of ZrO₂ spheres annealed at 500 °C showed high reflectance within the range from 300 nm to 1 μm. The enhancement of light extraction efficiency of GaN LED by using ZrO₂ spheres was explored. From the simulation results, ZrO₂ spheres led to improve the light extraction efficiency of about 70.12% when coated directly on the GaN surface and 23% improved when coated on the ITO surface due to the increased effective photon escape cone and spherical morphology. This suggested that the established synthetic method could not only be extended to form the other sub-micron spheres that contain alkoxide compound, but these sub-micron spheres are also expected to allow a practical approach for enhancing the light extraction efficiency of GaN based LEDs.

Acknowledgements

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (2013R1A1A1202268).

References

1. A.-P. Hynnien, J. H. J. Thijssen, E. C. M. Vermolen, M. Dijkska and A. Blaaderen, Nat. Mater., 2007, 6, 202.
2. A. C. Arsenault, D. P. Puzzo, L. Manners and G. A. Ozin, Nat. Photonics, 2007, 1, 468.
3. N. Vogel, L. Viguerie, U. Jonas, C. K. Weiss and K. Landfester, Adv. Funct. Mater., 2011, 32, 3064.
4. J. F. Galisteo-López, M. Ibisate, R. Sapienza, L. S. Froufe-Pérez, Á. Blanco and C. López, Adv. Mater., 2011, 23, 30.
5. F. Marlow, Muldarisnur, P. Sharifi, R. Brinkmann and C. Mendive, Angew. Chem., Int. Ed., 2009, 48, 6212.
6. F. Li, D. P. Josephson and A. Stein, Angew. Chem., Int. Ed., 2011, 50, 360.
7. H. Cong and W. Cao, Langmuir, 2003, 19, 8177.
8. W. Stӧber and A. Fink, J. Colloid Interface Sci., 1968, 26, 62.
9. J. Zhang, Z. Chen, Z. Wang, W. Zhang and N. Ming, Mater. Lett., 2003, 57, 4466.
10. I. H. Malitson, J. Opt. Soc. Am., 1965, 55, 1205.
11. Y. K. Ee, P. Kumnorkaew, R. A. Arif, J. F. Gilchrist and N. Tansu, Appl. Phys. Lett., 2007, 91, 221107.
12. J. M. Cariou, J. Dugas, L. Martin and P. Michel, Appl. Opt., 1986, 25, 334.
13. J. P. Syvitski, Principles, methods and application of particle size analysis, Cambridge University Press, 2007.
14. X. Jiang, T. Herricks and Y. Xia, Adv. Mater., 2003, 15, 1205.
15. H. Wang, N. Koshizaki, L. J. Li, K. Kawaguchi, X. Li, A. Pyatenko, Z. Swiatkowska-Warkocka, Y. Bando and D. Golberg, Adv. Mater., 2011, 23, 1865.
16. J. Widoniak, S. Eiden-Assmann and G. Maret, Eur. J. Inorg. Chem., 2005, 15, 3149.
17. J. Widoniak, S. Eiden-Assmann and G. Maret, Colloids Surf., A, 2005, 207, 329.

---

Footnote

† equally contributed to this work.

---