Photonic properties of titania inverse opal ... - OSA Publishing

1 downloads 59 Views 6MB Size Report
Hooi Sing Lee,1,* Roman Kubrin,2 Robert Zierold,3 Alexander Yu. ... P. V. Braun, S. A. Rinne, and F. García-Santamaría, “Introducing defects in 3D photonic ...
Photonic properties of titania inverse opal heterostructures Hooi Sing Lee,1,* Roman Kubrin,2 Robert Zierold,3 Alexander Yu. Petrov,1 Kornelius Nielsch,3 Gerold A. Schneider,2 and Manfred Eich1 1

Hamburg University of Technology, Institute of Optical and Electronic Materials, Eissendorfer Str. 38, 21073 Hamburg, Germany 2 Hamburg University of Technology, Institute of Advanced Ceramics, Denickestr. 15, 21073 Hamburg, Germany 3 University of Hamburg, Institute of Applied Physics, Jungiusstr. 11, 20355 Hamburg, Germany *[email protected]

Abstract: Titania inverse opal heterostructures demonstrating two distinctive photonic stopgaps were fabricated by repetitive vertical selfassembly and atomic layer deposition (ALD). Angle resolved reflectance measurements of the inverse opal heterostructure are reported for the first time. The comparison with the spectra of constituents show that the ΓL stopgaps of the heterostructure obey the superposition principle and the angular dispersion of their stopgaps is well-fitted with the modified Bragg’s law at low incidence angles. Numerical simulations were used to predict the dominant features in the reflectance spectra. The total (specular and diffuse) transmission and reflectance measurements of the single inverse opals and the heterostructure reveal that the diffuse scattering could severely impair the photonic properties of the buried layers in the multi-stack photonic crystal (PhC) configurations. Ascending stacking is proposed as a means to improve the performance of the multi-layer coatings. ©2013 Optical Society of America OCIS codes: (120.5700) Reflection; (120.5820) Scattering measurements; (160.5293) Photonic bandgap materials; (160.5298) Photonic crystals; (310.6188) Spectral properties.

References and links 1.

P. V. Braun, S. A. Rinne, and F. García-Santamaría, “Introducing defects in 3D photonic crystals: state of the art,” Adv. Mater. 18(20), 2665–2678 (2006). 2. K.-S. Lee, D.-Y. Yang, S. H. Park, and R. H. Kim, “Recent developments in the use of two-photon polymerization in precise 2D and 3D microfabrications,” Polym. Adv. Technol. 17(2), 72–82 (2006). 3. J. Ballato and A. James, “A ceramic photonic crystal temperature sensor,” J. Am. Ceram. Soc. 82(8), 2273–2275 (1999). 4. A. Sutti, C. Baratto, G. Calestani, C. Dionigi, M. Ferroni, G. Faglia, and G. Sberveglieri, “Inverse opal gas sensors: Zn(II)-doped tin dioxide systems for low temperature detection of pollutant gases,” in Proceedings of the Eleventh International Meeting on Chemical Sensors IMCS-11 IMCS 2006 IMCS 11 130 (2008), pp. 567– 573. 5. A. Bielawny, J. Üpping, P. T. Miclea, R. B. Wehrspohn, C. Rockstuhl, F. Lederer, M. Peters, L. Steidl, R. Zentel, S.-M. Lee, M. Knez, A. Lambertz, and R. Carius, “3D photonic crystal intermediate reflector for micromorph thin-film tandem solar cell,” Phys. Status Solidi A 205(12), 2796–2810 (2008). 6. P. Bermel, C. Luo, L. Zeng, L. C. Kimerling, and J. D. Joannopoulos, “Improving thin-film crystalline silicon solar cell efficiencies with photonic crystals,” Opt. Express 15(25), 16986–17000 (2007). 7. B. Hatton, L. Mishchenko, S. Davis, K. H. Sandhage, and J. Aizenberg, “Assembly of large-area, highly ordered, crack-free inverse opal films,” Proc. Natl. Acad. Sci. U.S.A. 107(23), 10354–10359 (2010). 8. A.-J. Wang, S.-L. Chen, P. Dong, X.-G. Cai, Q. Zhou, G.-M. Yuan, C.-T. Hu, and D.-Z. Zhng, “Fabrication of colloidal photonic crystals with heterostructure by spin-coating method,” Chin. Phys. Lett. 26(2), 024210 (2009). 9. P. Jiang, G. N. Ostojic, R. Narat, D. M. Mittleman, and V. L. Colvin, “The fabrication and bandgap engineering of photonic multilayers,” Adv. Mater. 13(6), 389–393 (2001). 10. R. V. Nair and R. Vijaya, “Three-dimensionally ordered photonic crystal heterostructures with a double photonic stop band,” J. Appl. Phys. 102(5), 056102 (2007). 11. Q. Yan, L. K. Teh, Q. Shao, C. C. Wong, and Y.-M. Chiang, “Layer transfer approach to opaline hetero photonic crystals,” Langmuir 24(5), 1796–1800 (2008).

#185113 - $15.00 USD (C) 2013 OSA

Received 11 Feb 2013; revised 24 May 2013; accepted 24 May 2013; published 2 Jul 2013

1 August 2013 | Vol. 3, No. 8 | DOI:10.1364/OME.3.001007 | OPTICAL MATERIALS EXPRESS 1007

12. W. Khunsin, S. G. Romanov, C. M. Sotomayor Torres, J. Ye, and R. Zentell, “Optical transmission in triple-film hetero-opals,” J. Appl. Phys. 104(1), 013527 (2008). 13. Z.-Q. Liu, T.-H. Feng, Q.-F. Dai, L.-J. Wu, and L. Sheng, “Fabrication of high-quality three-dimensional photonic crystal heterostructures,” Chinese Physics B 18(6), 2383–2388 (2009). 14. S. G. Romanov, M. Egen, R. Zentel, and C. M. Sotomayor Torres, “Propagation and scattering of light in opal heterojunctions,” in Proceedings of the 12th International Conference on Modulated Semiconductor Structures Proceedings of the 12th International Conference on Modulated Semiconductor Structures (2006), Vol. 32, pp. 476–479. 15. C. M. Soukoulis, Photonic Crystals and Light Localization in the 21st Century (Kluwer Academic, 2001). 16. A. Mihi, M. E. Calvo, J. A. Anta, and H. Miguez, “Spectral response of opal-based dye-sensitized solar cells,” J. Phys. Chem. C 112(1), 13–17 (2008). 17. D.-K. Hwang, H. Noh, H. Cao, and R. P. H. Chang, “Photonic bandgap engineering with inverse opal multistacks of different refractive index contrasts,” Appl. Phys. Lett. 95(9), 091101 (2009). 18. A. Wang, S.-L. Chen, P. Dong, and Z. Zhou, “Preparation of photonic crystal heterostructures composed of two TiO2 inverse opal films with different filling factors,” Synth. Met. 161(5-6), 504–507 (2011). 19. Z. Cai, Y. J. Liu, J. Teng, and X. Lu, “Fabrication of large domain crack-free colloidal crystal heterostructures with superposition bandgaps using hydrophobic polystyrene spheres,” ACS Appl. Mater. Interfaces 4(10), 5562– 5569 (2012). 20. H. S. Lee, R. Kubrin, R. Zierold, A. Y. Petrov, K. Nielsch, G. A. Schneider, and M. Eich, “Thermal radiation transmission and reflection properties of ceramic 3D photonic crystals,” J. Opt. Soc. Am. B 29(3), 450–457 (2012). 21. R. Kubrin, H. S. Lee, R. Zierold, A. Yu. Petrov, R. Janssen, K. Nielsch, M. Eich, and G. A. Schneider, “Stacking of ceramic inverse opals with different lattice constants,” J. Am. Ceram. Soc. 95(7), 2226–2235 (2012). 22. S. Johnson and J. Joannopoulos, “Block-iterative frequency-domain methods for Maxwell’s equations in a planewave basis,” Opt. Express 8(3), 173–190 (2001). 23. Available at www.cst.com. 24. A. F. Koenderink and W. L. Vos, “Optical properties of real photonic crystals: anomalous diffuse transmission,” J. Opt. Soc. Am. B 22(5), 1075–1084 (2005). 25. H. M. van Driel and W. L. Vos, “Multiple Bragg wave coupling in photonic band-gap crystals,” Phys. Rev. B 62(15), 9872–9875 (2000). 26. S. G. Romanov, T. Maka, C. M. Sotomayor Torres, M. Müller, R. Zentel, D. Cassagne, J. Manzanares-Martinez, and C. Jouanin, “Diffraction of light from thin-film polymethylmethacrylate opaline photonic crystals,” Phys. Rev. E Stat. Nonlin. Soft Matter Phys. 63(5), 056603 (2001). 27. J. F. Galisteo-López, E. Palacios-Lidón, E. Castillo-Martínez, and C. López, “Optical study of the pseudogap in thickness and orientation controlled artificial opals,” Phys. Rev. B 68(11), 115109 (2003). 28. M. Ishii, M. Harada, A. Tsukigase, and H. Nakamura, “Three-dimensional structure analysis of opaline photonic crystals by angle-resolved reflection spectroscopy,” J. Opt. A 9(9), S372–S376 (2007). 29. R. C. Schroden, M. Al-Daous, C. F. Blanford, and A. Stein, “Optical properties of inverse opal photonic crystals,” Chem. Mater. 14(8), 3305–3315 (2002). 30. Y. Cao, Y. Wang, Y. Zhu, H. Chen, Z. Li, J. Ding, and Y. Chi, “Fabrication of anatase titania inverse opal films using polystyrene templates,” Superlattices Microstruct. 40(3), 155–160 (2006). 31. M. Lanata, M. Cherchi, A. Zappettini, S. M. Pietralunga, and M. Martinelli, “Titania inverse opals for infrared optical applications,” Opt. Mater. 17(1-2), 11–14 (2001). 32. J. E. G. J. Wijnhoven, L. Bechger, and W. L. Vos, “Fabrication and characterization of large macroporous photonic crystals in titania,” Chem. Mater. 13(12), 4486–4499 (2001). 33. J. S. King, E. Graugnard, and C. J. Summers, “TiO2 inverse opals fabricated using low-temperature atomic layer deposition,” Adv. Mater. 17(8), 1010–1013 (2005). 34. Y. A. Vlasov, M. Deutsch, and D. J. Norris, “Single-domain spectroscopy of self-assembled photonic crystals,” Appl. Phys. Lett. 76(12), 1627–1629 (2000). 35. J. F. Galisteo Lòpez and W. L. Vos, “Angle-resolved reflectivity of single-domain photonic crystals: effects of disorder,” Phys. Rev. E Stat. Nonlin. Soft Matter Phys. 66(3 Pt 2B), 036616 (2002). 36. A. F. Koenderink, A. Lagendijk, and W. L. Vos, “Optical extinction due to intrinsic structural variations of photonic crystals,” Phys. Rev. B 72(15), 153102 (2005).

1. Introduction Three dimensional photonic crystals (3D-PhCs) have been extensively studied in the past few decades due to their unique photonic properties that allow confining and manipulating of light. This class of material has been envisaged for various applications such as optical devices [1,2], sensorics [3,4], and photovoltaics [5,6]. Among the 3D-PhC structures, socalled colloidal crystals obtained by self-assembly of monodisperse particles are particularly attractive because they provide a much simpler and cost-effective approach for large scale fabrication than the top-down lithography techniques. Furthermore, recent refinements of the

#185113 - $15.00 USD (C) 2013 OSA

Received 11 Feb 2013; revised 24 May 2013; accepted 24 May 2013; published 2 Jul 2013

1 August 2013 | Vol. 3, No. 8 | DOI:10.1364/OME.3.001007 | OPTICAL MATERIALS EXPRESS 1008

self-assembly techniques allow for producing 3D-PhC structures with low defect densities [7]. This further suggest self-assembly as an intuitive way to fabricate PhCs. Stacking of several PhC film on top of each other can be considered as a mean to expand the functionality of the PhCs, e.g. to create structures either with multiple bandgaps or a broadened bandgap. So far, the reported studies of the PhC heterostructures have been mainly focused on the hetero-opals, consisting of two or more opaline films with different particle size [8–14]. It is known that ‘inverse opals’ obtained by inversion of the colloidal crystals can provide more pronounced photonic bandgap properties since a stronger photonic interaction is expected in these structures [15]. Up to now only a few studies have been conducted to deal with the inverse opal heterostructures [16–19] and they are mostly concentrated on the fabrication processes, without thorough investigation their optical properties. Similar to the direct hetero-opal, the most straightforward consequence for stacking two inverse opal films is the superposition of the bandgap features from the constituents [20]. Whether or not the product of spectra from the individual films is sufficient to describe the resultant properties of an inverted heterostructure has not been proven experimentally. In this paper, the ΓL spectral characteristic of an inverse opal heterostructure is compared to that of the single inverse opal and a sandwiched inverse opal. The angle dependent characteristics of reflected light by the inverse opal heterostructure were investigated. Furthermore, the transmission and the reflection of the diffusely scattered light through the inverse opals and heterostructures were studied. We show that the scattering effects within the inverse opal are important in the heterostructure. A simple FEM model was set-up to compare with the experimental observations. 2. Experimental technique The typical fabrication of inverse opal heterostructures involves several sequential steps. A detailed description of the synthesis can be found elsewhere [21]. Monodisperse Polystyrene (PS) particles with the sizes of 608 nm and 756 nm (coefficient of variation < 3%) purchased from the Microparticles GmbH (Berlin, Germany) were used in the experiment. The suspension of PS particles were diluted with deionized water in a beaker to a concentration – of about 1 mg/ml. Then a microscope slide (76x26mm) was immersed nearly vertically into the beaker. The beaker was placed in a thermostatic oven at atmospheric pressure, high relative humidity and at temperature of 60° for several days. Evaporation of water induces self-assembly of colloidal particles in the meniscus on the surface of the microscope slide thus creating an opal layer as the meniscus slowly slides down along the substrate with the growth rates of approximately ~4 mm/day. The resulting PS templates were then infiltrated with titanium dioxide (TiO2) by using low temperature atomic layer deposition (ALD) in a custommade ALD reactor. The precursor titanium tetraisopropoxide (TTIP, purity 97%) was obtained from Strem Chemical Inc. (Kehl, Germany). The chamber temperature was set to 95°C in exposure mode. Alternate pulses of TTIP and water vapor followed in each case by a long exposure time of 60 s as well as 120 s N2 purging result in conformal deposition of TiO2. An average growth rate of 0.045 nm/cycle was achieved at the optimized conditions. Thus more than 1000 cycles were necessary for the complete infiltration. After the first PS template was coated with titania, the steps of self-assembly and infiltration are repeated to obtain the heterostructure. As previously reported [21], there is a practical limit of the infiltration depth with ALD technique due to the complex diffusion paths within such porous structures. Thus the infiltration step was repeated for every stack in order to ensure the best quality of the inverse shell opal heterostructure. Before the second ALD-infiltration of the sample used for the present study, the TTIP-precusor was completely exchanged for a fresh one, as opposed to the previously reported heterostructures [21]. Finally, the PS templates were removed by calcination in air at 500°C, leaving behind a mechanically stable inverse opal heterostructure. The quality of the fabricated structures was assessed by scanning electron microscopy (SEM: Leo 1530 Gemini) and the optical properties were measured with a Fourier transform

#185113 - $15.00 USD (C) 2013 OSA

Received 11 Feb 2013; revised 24 May 2013; accepted 24 May 2013; published 2 Jul 2013

1 August 2013 | Vol. 3, No. 8 | DOI:10.1364/OME.3.001007 | OPTICAL MATERIALS EXPRESS 1009

infrared spectrometer (FTIR: Bruker Tensor37) in the near infrared spectral region. In this study, four measurement configurations were used: specular transmittance, specular reflectance, hemispherical transmittance and hemispherical reflectance. In the specular transmittance measurement, only the transmitted intensity within a solid angle of 1.5° is being captured, whereas in the total transmittance configuration the transmitted intensity from the entire hemisphere of the sample is being recorded. In the specular reflectance mode, the recorded power spectrum of a mirror-like reflected beam from the sample is normalized with the spectrum taken from an aluminum mirror. In this measurement configuration, the incident and the reflected beam, as well as the normal to the surface of the sample, are all contained in the diffraction plane. The diffusely reflected beam which results from the defects, surface roughness etc. was collected with a gold coated integrating sphere and the spectra of the samples were normalized with the spectrum of a gold diffuser. The probe beam sizes used in the specular mode and in hemispherical mode are ~3 mm and ~10 mm, respectively. Thus several regions of the sample with different thicknesses and different domains were simultaneously illuminated. The refractive index of the titania from a control sample was determined with the variable angle spectroscopic ellipsometry. The photonic band structures of photonic crystals were calculated by using plane wave expansion method (MPB software) [22]. It allows the calculations of the spectral position and the width of the bandgaps of infinite structures. The optical properties of finite PhC structures were calculated with the finite element method (FEM) in frequency domain [23]. 3. Results and discussion Figure 1 shows the SEM images of the fabricated titania inverse opal heterostructure. The top layer film contains air pores with a nominal diameter of 608 nm while the bottom layer contains pore size of 756 nm (assumed equal to the corresponding PS particle size specified by the supplier). The actual sizes of the pores could be slightly smaller due to the volume shrinkage of titania during the calcination step. The close-packed arrangement of air spheres in both layers is evident, preserving the original face-centered-cubic arrangement of the PS particles in the opal templates. High magnification SEMs on the {111} surfaces (Fig. 1(a) insets) reveal that the top- and the bottom- PhC layers have different surface roughness. The abrupt boundary between the films (Fig. 1(d)) clearly demonstrates that the growth of opal on top of an infiltrated PhC can still retain good ordering and the PS spheres are assembled on top of the first PhC film similar to crystalline on a flat surface. Inspite of many broken shells resulting from the cleaving process, a high degree of order in both PhC layers is visible.

#185113 - $15.00 USD (C) 2013 OSA

Received 11 Feb 2013; revised 24 May 2013; accepted 24 May 2013; published 2 Jul 2013

1 August 2013 | Vol. 3, No. 8 | DOI:10.1364/OME.3.001007 | OPTICAL MATERIALS EXPRESS 1010

Fig. 1. The top-view (a-b) and the cross-sectional view (c-d) SEM images of the 608 nm/756 nm inverse opal heterostructure. Each stack of the film contains PhC with the average thickness of ~10 µm (top) and ~5 µm (bottom), respectively. Insets in (a) and (d) show the high magnification SEMs of {111} surfaces and the interface between the two PhC layers, respectively.

Figure 2(a) shows the 13° incidence specular reflectance of the 608/756nm heterostructure. For comparison, the reflectance spectra of single inverse opal films with the pore sizes 608 nm and 756 nm are plotted in black dashed lines. The transmission dips at wavelength 1150 nm and 1470 nm correspond to the first order ΓL stopgaps of the individual PhC films. The spectra show one-to-one correspondence suggesting that the optical property of a heterostructure can be relatively well approximated by multiplying the individual optical transmission spectra of the constituent films. The normal incidence hemispherical transmittance spectrum of the heterostructure is shown in Fig. 2(b). By comparing both the transmission curves, one can easily deduce that the contribution of diffuse light to the overall photon transportation within the inverse opal heterostructure is fairly large. Scattering of light by the surface roughness and the structural defects within the photonic crystals is responsible for the low specular transmittance within the entire measured spectral range. Typical defects observed in our opal structures include point/line defects, stacking faults, position randomness of spheres and cracks. These defects are developed during the self-assembly process and are transferred to the final structure. The surface roughness of the coating arises from the ALD process (discussed below). Due to their presence in the structure, the propagation of light is no longer solely ballistic as in the case of perfect photonic crystals but consist of both ballistic and complex diffuse components. Thus after repetitive of diffuse scattering, light exits the structure with a broad k-vector distribution. The specular transmittance, which constitutes only part of this diffusely distributed light shows significantly lower intensity. The question ‘which defects have the dominant effects on the optical properties’ will require further investigation. The fact that the transmission for the wavelengths outside of the stopgaps is not 100% is caused by the surface reflectance and the diffuse back-scattering, which are readily captured with an integrating sphere. The 13° specular – and hemispherical- reflectance spectra of the heterostructure are plotted in Fig. 2(c). Both Bragg peaks from the constituent PhC films are #185113 - $15.00 USD (C) 2013 OSA

Received 11 Feb 2013; revised 24 May 2013; accepted 24 May 2013; published 2 Jul 2013

1 August 2013 | Vol. 3, No. 8 | DOI:10.1364/OME.3.001007 | OPTICAL MATERIALS EXPRESS 1011

clearly displayed in the spectrum. Each peak corresponds to a gap width-to-midgap ratio (Δω/ω) of 12.4% and 13.6%, which is relatively close to the value 14.1% obtained from MPB software by using the refractive index of titania = 2.41 (measured on a control thin film sample). The discrepancy of the relative gap width between the top and bottom PhC films could be due to the unequal degree of infiltration, since the bottom layer was exposed to additional ALD cycles during the infiltration of top layer [21]. The top PhC film has a lower filling fraction of titania and thus exhibits a smaller Δω/ω. Again, a remarkable difference between the specular- and hemispherical- reflectance is observed. Apart from the Bragg peaks corresponding to the stopgaps of the photonic heterostructure, there is an intensive background signal in the hemispherical reflectance increases monotonically for the shorter wavelengths. Such a strong background reflectance is attributed to light scattering and usually has λ−2 dependence [24]. It should be noted that there is no new feature is registered in the transmittance- and reflectance spectra, so it can be assumed that the interface between PhC films does not play an important role at first glance. In our previous work, the diffuse scattering practically precludes the detection of the second stopgap in photonic heterostructures obtained by the standard ALD infiltration [21]. Obviously, the exchange of precursor resulted in an improvement of the photonic performance of the top PhC layer. As shown in the inset of Fig. 1(a), the surface of the top layer appears to be very smooth. Therefore, it can be concluded that the roughness of TiO2 coatings deposited at 95°C must be related to degradation of the TTIP-precursor. The exact causes of the change in the properties of TTIP are being investigated and will be discussed in our upcoming publications. We believe that by improving the existing opals fabrication techniques [7] and the infiltration process, inverse opal multi-stacks with excellent photonic properties should be feasible.

#185113 - $15.00 USD (C) 2013 OSA

Received 11 Feb 2013; revised 24 May 2013; accepted 24 May 2013; published 2 Jul 2013

1 August 2013 | Vol. 3, No. 8 | DOI:10.1364/OME.3.001007 | OPTICAL MATERIALS EXPRESS 1012

Fig. 2. (a) Normalized 13° specular reflectance spectrums (red line) of a titania inverse opal heterostructure. The reflectance spectra of the constituent PhC films (black dotted lines) are plotted for comparison. The normal incidence total transmittance and 13° incidence total reflectance of the heterostructure are plotted in (b) and (c), respectively. Black lines represent measurements in the specular mode while the red dash-lines represent the transmitted or reflected intensity collected with an integrating sphere. Insets show the integrating sphere measurement configurations. S represents the sample and M represents the mirror.

The angle-resolved specular reflectance of the single inverse opals and heterostructure are presented in Fig. 3. The reflectance spectra of single inverse opals in Figs. 3(a) and 3(b) have the appropriate pore sizes (608 nm and 756 nm) and similar degree of infiltration. All spectra are normalized to the peak intensities for the sake of comparison. The Bragg peaks of the inverse opal films and the heterostructure have a maximum reflectivity of 86% and 66%, respectively. The well-known blue shift of {111} stopgaps with increasing angle of incidence are observed for the single inverse opals. The same shift can also be seen in the spectra of the heterostructure (Fig. 3(c)). At large incidence angles, the intensities of the Bragg peaks gradually decrease and a new diffraction peak appears in the lower wavelength (~900nm) region. This new reflectance peak demonstrates a red shift in the spectra and becomes intense and prevails as the angle of incidence increases. This can be attributed to the specular reflectance satisfying simultaneously {111} planes and {200} family of planes. The behavior of this high energy peak has been well documented by Driel et al. [25], Romanov et al. [26] and Galisteo et al. [27]. The superposition of the angle-resolved reflectance spectra in Figs. 3(a) and 3(b) is shown in Fig. 3(d). As already shown in the near normal incidence specular reflectance spectrum (Fig. 2(a)), the superposed spectrum resemblances to the measured spectrum of the

#185113 - $15.00 USD (C) 2013 OSA

Received 11 Feb 2013; revised 24 May 2013; accepted 24 May 2013; published 2 Jul 2013

1 August 2013 | Vol. 3, No. 8 | DOI:10.1364/OME.3.001007 | OPTICAL MATERIALS EXPRESS 1013

heterostructure, especially in the low incidence angles range. This further supports the conclusion that the photonic properties of a stacked inverse opal with different spectral positions of the stopgaps can be predicted by multiplying the angle resolved spectra of individual photonic crystals [20]. At oblique incidence angles, however, there is a considerable deviation between the linear superposition of the spectra of the constituents and the spectrum of the heterostructure. The high angle diffraction of the bottom PhC is no longer registered in the spectra. We attribute this vanishing intensity of the peaks from the buried PhC layers to the fact that photons with corresponding wavelengths have a longer propagation length at oblique angles. In this way, the contribution of the ballistic photons to diffuse scattering should increase and the stopgaps of the bottom layers are no longer registered in the specular mode. The center wavelengths of each reflection peak in Fig. 3 were then extracted and plotted against the incidence angles. The angular dispersion of the both stopgaps for the heterostructure can be well described with the modified Bragg’s law for the angle below 45°:

λ = 2 d111 neff2 − sin 2 θ

(1)

where λ is the center wavelength of Bragg peak, d111 is the interplanar spacing of FCC lattice, neff is the effective refractive index and θ is the incidence angle. By fitting the Bragg relation to the plotted data, the average pore sizes of the film were determined. The results estimated that the pore sizes of the structure have shrunk ~18% in the direction perpendicular to the sample surface. The comparison was carried out with respect to the nominal polystyrene particle sizes as specified by the manufacturer. The reason of this reduced pore sizes is due to the combination of the shrinkage of polystyrene sphere during drying (in the range of few % [28]) and densification of titania during the phase transformation from amorphous to anatase phase. Such shrinkage is smaller than the reported values (25-30%) observed from structures obtained from the sol-gel method [29–32]. The obtained values of the effective refractive index neff = 1.438 and 1.48 for the first and the second peak, are remarkably close to the value 1.44 calculated from the weighted average. The calculation assumes that titania is conformally deposited on the template and the maximum infiltration is achieved after ~86% of void space has been filled [33]. Further infiltration with ALD is not possible because of the complete closure of the pores (Fig. 1(b)) in the {111} surfaces of the opal structure so that the transport of the precursor vapor by diffusion is disrupted.

#185113 - $15.00 USD (C) 2013 OSA

Received 11 Feb 2013; revised 24 May 2013; accepted 24 May 2013; published 2 Jul 2013

1 August 2013 | Vol. 3, No. 8 | DOI:10.1364/OME.3.001007 | OPTICAL MATERIALS EXPRESS 1014

Fig. 3. Normalized angle resolved specular reflectance of: single inverse opal with pore sizes (a) 608nm (b) 756nm and (c) inverse opal heterostructure. The measurements were carried out in 2° resolution. Inset in (a) shows the angle resolved measurement configuration. The superposed spectrum (by multiplication of spectra from (a) and (b)) is plotted in (d) for comparison. The dotted lines in all graphs represent the curves fitting using Eq. (1).

The modified Bragg equation above provides a convenient formula for estimating the center wavelength of the first order diffraction peaks. However, it does not provide any information regarding the stopgaps beside the spectral position (e.g. bandwidth and suppression of the stopbands). Especially for the case of ALD-infiltrated inverse opals, the optical properties of structure no longer depend solely on the index contrast and pore size, but also on the shell thickness of the infiltrated materials and the unfilled interstitial volumes of the structure. Therefore, a three-dimensional FEM simulation model of an inverse shell opal was developed to more accurately describe the fabricated structure. In this model, periodic boundaries are applied in X- and Y-direction while the Z-direction is finite. The structure contains air spheres in FCC lattice and the titania is conformally distributed on top of the spheres with a defined thickness. Two waveguide ports were applied at both end of the structure and were used to excite the opal model with a defined angle of incidence. The simulated transmission spectra of inverse opals as a function of shell thickness (increases from 0.086r to 0.155r, where r is the radius of the pores) are plotted in Fig. 4(a). The transmissions dip shifts toward long wavelength and increases in suppression as its shell thickness increases, which is consistent with the increase of average dielectric constant in the structure. The calculated angular response of the stopgap of the inverse shell opal as a function of incidence angle is plotted in Fig. 4(b). The spectra are normalized with the lattice parameter of the structure. Despite of the simplicity of the model, the simulated spectra show a relatively good agreement with the measured results in Figs. 3(a) and 3(b). The reflectance peak corresponding to the stopgap is much more pronounced as compared to the experimental results because the simulation model consists solely of perfectly ordered air spheres. Furthermore, in experiment the probe beam cross section increases from a circle with size of ~3 mm to an ellipse with size of ~8 mm at the long axis for oblique angles. The beam size is

#185113 - $15.00 USD (C) 2013 OSA

Received 11 Feb 2013; revised 24 May 2013; accepted 24 May 2013; published 2 Jul 2013

1 August 2013 | Vol. 3, No. 8 | DOI:10.1364/OME.3.001007 | OPTICAL MATERIALS EXPRESS 1015

significantly larger than the domain size of the sample and, therefore, will cause the measurements averaging over many domains including defects and cracks. Especially at high incidence angles, reflectance of the real PhC is strongly affected by diffuse scattering, which could not be accounted for in our simulation model. The maximum reflectance up to 96% could be demonstrated experimentally by performing the measurement on single domains [34,35].

Fig. 4. (a) Calculated transmission spectra of a 10 layers titania inverse shell opal for various shell thicknesses. The grey arrow indicates the direction in which the shell thickness increases. b) Contour plot of reflectance spectra of the titania inverse opal for different incidence angles. The oscillations outside stopgap correspond to the Fabry-Pérot fringes due to the finite sample geometry.

Potential applications of 3D photonic crystal coatings are not limited to the single layers and heterostructures. For example, the realization of broadband reflectors of thermal radiation would require stacking of up to ten layers with different periodicity constants [20]. It is implied that high reflectance would be achieved by superposition of photonic stopgaps and not due to diffuse back-scattering on the defects in such structures. However, one could anticipate from the spectra of the heterostructure that the diffuse component would inevitably increase, if further PhC layers were added. There exist estimates that in the inverse opals produced by conventional methods, photons can be controlled over the distances corresponding to approximately 50 lattice constants (or 15 µm) [36]. On one hand, we expect that this limitation can be overcome by further developing the manufacturing techniques. On the other hand, it is well known that the scattering losses outside the stopgap of the PhC decrease for the longer wavelengths. Therefore, the multi-stacks structured in an ascending manner (the top PhC layer immediately exposed to the incident light has the smallest pore size and the lattice constant of each deeper layer is larger than that of the previous one) potentially could provide a sufficient penetration depth for ballistic photons in a wide range of wavelengths even with already existing fabrication methods. Our ALD-infiltration technique, however, still requires optimization in order to deposit high-quality photonic multi-stacks. At the current stage of experiments, only single inverse opals and heterostructures could be produced. We also could not simulate the effect of diffuse scattering, as discussed above. Nonetheless, it was possible to address the issues of stacking several photonic crystal films in a “sandwich” experiment. We deliberately place two single inverse opals with different pore sizes (608 nm and 756 nm) on top of each other and compared its optical properties to that of the heterostructure. As illustrated in the insets of Fig. 5, the opal films are sandwiched between two glass substrates. Light is transmitted through the glass substrate and first incident on the 608 nm inverse opal film. The illuminated areas of the single inverse opals had a thickness of ~9 µm and 17 µm, respectively, which was estimated from the Fabry-Pérot fringes in the individual specular reflectance spectra. The obtained transmittance and reflectance spectra of the sandwich

#185113 - $15.00 USD (C) 2013 OSA

Received 11 Feb 2013; revised 24 May 2013; accepted 24 May 2013; published 2 Jul 2013

1 August 2013 | Vol. 3, No. 8 | DOI:10.1364/OME.3.001007 | OPTICAL MATERIALS EXPRESS 1016

structure and its constituents are plotted in Figs. 5(a) and 5(b), respectively. As expected, the center wavelengths of the stopgaps in the sandwiched structure remain unchanged with respect to that in the individual films. The stopgap of the bottom PhC film shows a slightly reduced transmission dip/ reflectance peak. As compared to the sandwiched PhCs, the heterostructure demonstrates substantially higher diffuse components. This difference in intensity of back-scattering should be mainly attributed to a large concentration of defects near the interface between the top and the bottom PhC layers of the heterostructure. Even in the case of a perfect stacking of polystyrene spheres in the opal template deposited in the second self-assembly step, the top surface of the bottom layer is terminated by a coating of TiO2 in the first infiltration run and receives additional amounts of titania during the second ALD. Increased degree of infiltration of the bottom layer is evidenced by the broadening of the reflectance peak corresponding to the bottom layer in the spectra of the heterostructure and by the red-shift on the position of this peak. In this way, a plane-like 2D TiO2-rich defect (although it is not flat on the scale of the pore size) of locally increased refractive index is introduced between the inverse opal layers in the heterostructure. From our previous work, it is known that thick ALD deposits have a tendency to exaggerated roughness, similar to the surface in Fig. 1(d) which was also double-exposed to ALD [21]. Thus, roughness of the interface is assumed to cause additional diffuse scattering. Furthermore, it is expected that certain disorder due to the less favorable growth condition is introduced in the top PS opal template during the second self-assembly run because of unevenness of the surface of the bottom layer and possible surface contamination by dust particles and chemical species absorbed from the atmosphere before the sample is put in the beaker with the PS suspension. This simple experiment demonstrates that it is essential to control the properties of the interfaces for successful stacking of photonic crystals. Figure 5(c) shows the total reflectance of a “triple sandwich” structure, where we stack three single inverse opals with pore sizes of 476 nm (green), 608 nm (blue) and 756 nm (red) on top of each other, respectively. It shows that it is possible to stack more than two inverse opal films as long as the diffuse scattering on the defects inside the inverse opal films and on the interfaces between them is sufficiently low.

#185113 - $15.00 USD (C) 2013 OSA

Received 11 Feb 2013; revised 24 May 2013; accepted 24 May 2013; published 2 Jul 2013

1 August 2013 | Vol. 3, No. 8 | DOI:10.1364/OME.3.001007 | OPTICAL MATERIALS EXPRESS 1017

Fig. 5. Comparison of total transmittance (a) and total reflectance (b) of single inverse opals (dotted lines), the sandwiched structure (black solid line) and the heterostructure (olive solid line). Total reflectance of a “triple sandwich” inverse opal structure is plotted in (c). Insets on the right show the measurement configurations. The single inverse opals with a pore size of 476 nm, 608 nm and 756 nm are denoted by green-, blue- and red- colors, respectively.

5. Conclusion We have demonstrated that titania inverse opal heterostructures produced by ALD-infiltration of polystyrene templates with different particle sizes can possess two distinctive photonic stopgaps over a broad angle range. The comparison of the angle-resolved specular reflectance spectra of the heterostructure with the spectra of corresponding single inverse opal films demonstrates that the photonic properties of a stack can be approximated by linear superposition of the individual films and the angular dispersion of the stopbands can be well described by the modified Bragg’s law below the incidence angle of 45°. The dissimilarity between the superposed spectra and the spectra obtained from the heterostructure at high incident angles was explained by the increase of diffuse scattered intensity due to the elongation of the optical path of photons. Diffuse scattering is the main obstacle for applications of the multi-stack inverse opals that require multiple photonic stopbands or broadband reflectance. Although the realization of practical devices would require radical improvement of the structure quality, we show that successful stacking of at least three photonic crystals is feasible with the already existing fabrication technique of titania inverse opals.

#185113 - $15.00 USD (C) 2013 OSA

Received 11 Feb 2013; revised 24 May 2013; accepted 24 May 2013; published 2 Jul 2013

1 August 2013 | Vol. 3, No. 8 | DOI:10.1364/OME.3.001007 | OPTICAL MATERIALS EXPRESS 1018

Acknowledgments This research is supported by Hamburg Integrated Materials Systems Excellency Cluster and the German Research Foundation (DFG) via SFB 986 “M3”, projects C2, C3 and C4. The authors acknowledge the support from CST, Darmstadt, Germany with their Microwave Studio software.

#185113 - $15.00 USD (C) 2013 OSA

Received 11 Feb 2013; revised 24 May 2013; accepted 24 May 2013; published 2 Jul 2013

1 August 2013 | Vol. 3, No. 8 | DOI:10.1364/OME.3.001007 | OPTICAL MATERIALS EXPRESS 1019