Achieving Color and Function with Structure: Optical and Catalytic Support Properties of ZrO2 Inverse Opal Thin Films.

Taking inspiration from natural photonic crystal architectures, we report herein the successful fabrication of zirconia inverse opal (ZrO2 IO) thin-film photonic crystals possessing striking iridescence at visible wavelengths. Poly(methyl methacrylate) (PMMA) colloidal crystal thin films (synthetic opals) were first deposited on glass microscope slides, after which the interstitial voids in the films were filled with a Zr(IV) sol. Controlled calcination of the resulting composite films yielded iridescent ZrO2 IO thin films with pseudo photonic band gaps (PBGs) along the surface normal at visible wavelengths. The PBG position was dependent on the macropore diameter (D) in the inverse opals (and thus proportional to the diameter of the PMMA colloids in the sacrificial templates), the incident angle of light with respect to the surface normal (θ), and also the refractive index of the medium filling the macropores, all of which were accurately described by a modified Bragg's law expression. Au/ZrO2 IO catalysts prepared using the ZrO2 IO films demonstrated outstanding performance for the reduction of 4-nitrophenol to 4-aminophenol in the presence of NaBH4, which can be attributed to the interconnected macroporosity in the films, which afforded a high Au nanoparticle dispersion and also facile diffusion of 4-nitrophenol to the catalytically active Au sites.

Introduction

Intriguing, ingenious, and diverse are the mechanisms by which nature uses submicron-scale periodic architectures called photonic crystals to create structural color.[1−14] A characteristic feature of structurally colored materials is their iridescence or color change with viewing angle with respect to a fixed white light source (cf. pigment-based colored materials that generally look the same color at all illumination and viewing angles). The striking iridescence shown by photonic crystals arises from the selective diffraction of certain wavelengths of white light on one-dimensional, two-dimensional (2D), or three-dimensional (3D) periodic structures within the material, where typically the periodicity is comparable to visible wavelengths (400–700 nm). Figure S1 shows digital photographs for some common naturally occurring structurally colored objects, as well as scanning electron microscopy images of the periodic dielectric structures responsible for the iridescent colors of these objects. These examples serve to highlight key concepts exploited here in our research aimed at the fabrication of ZrO2 inverse opal thin-film photonic crystals. In Paua shell (Haliotis iris), 400 nm-thick tiles of calcium carbonate add mechanical strength to the molluscs shell, but also cause the shell to diffract strongly at both near-infrared (first-order diffraction) and visible (second-, third-, and fourth-order diffraction) wavelengths.[12] The eye in the display feathers of the peacock (Pavo cristatus) contains 2D arrays of melanin rods, with the diameter of the rods depending on the location in the eye (rods in the green part have a diameter of around 140 nm, whereas rods in the orange part have a slightly larger diameter of ∼150 nm).[13] Finally, precious opal obtains its color from a face-centered cubic (fcc) (3D) arrangement of monodisperse spherical silica (SiO2) colloids, with colloid diameters typically in the range 200–300 nm.[14] The diffracted wavelengths in opals increase with colloid diameter (i.e., at normal incidence, opals made of 200 nm colloids reflect in the violet, those made from 250 nm colloids in the green, and those made from 300 nm colloids in the red). Light refraction and diffraction leading to color in opal-based photonic crystals are depicted in Figure S2 and have been discussed in depth below.[15−24] In each of the examples in Figure S1, a dark background provided by protein or a sandstone matrix enhances the color play of the photonic crystal architecture, making the iridescence more vivid. These are but a brief snapshot of the diverse architectures used in nature to generate structural color. Physicists and engineers are now using “top-down” approaches[25−28] and chemists and biologists are using “bottom-up” approaches[29−69] to fabricate photonic crystals operating at near-infrared wavelengths (for optical communications, especially waveguides, fiber-optic cables, fast optic switches, etc.) and visible wavelengths (sensing, high-efficiency laser cavities). Nature serves as a valuable teacher for technology breakthroughs in these areas.

Among the methods used by chemists for photonic crystal fabrication, the colloidal crystal template approach is arguably the most popular due to the diverse range of photonic crystals that can be obtained.[42−69] The first step in this approach is the synthesis of monodisperse polymer or silica colloids via surfactant-free emulsion polymerization or the Stöber method,[30−33,42,48] respectively, followed by the crystallization of these colloids to form a colloidal crystal or synthetic opal (typically a face-centered cubic array of colloids with a solid fraction of 74 vol %). Due to the high interfacial surface free energies of submicron-sized spherical colloids, polymer and silica colloids tend to spontaneously self-assemble to form well-ordered colloidal crystals, which assists this fabrication route. Strategies used to form colloidal crystals from aqueous or ethanoic colloidal dispersions include gravitational sedimentation, centrifugation, evaporation, and various “falling meniscus” methods.[33−42,70−74] The latter methods are especially amenable for colloidal crystal thin-film deposition on planar substrates (this is preferable if detailed optical analyses are going to be performed on the colloidal crystals and their inverted replicas). In the subsequent stage of the colloidal crystal templating strategy, the colloidal crystal (i.e., opal) is used as a sacrificial template to fabricate inverse opal replicas, comprising a face-centered cubic array of air spheres (macropores) in a solid matrix.[42−69] To obtain the inverse opal, the interstitial voids in the colloidal crystal template (26 vol % of the structure) are filled with a solid material via sol–gel, electrochemical, or nanoparticle infiltration methods, after which the original colloidal crystal template is selectively removed by dissolution or calcination (heating in air). A multitude of inverse opal-based photonic crystal materials have been synthesized by this strategy, including metals, metal oxides, conducting polymers, and carbon materials.[29,42−69] The three-dimensionally ordered macroporous structure of inverse opals, together with their inherent photonic crystal properties, creates new opportunities in optical sensor development, separation, photocatalysis, and catalysis, among others. Particularly valuable in this context are inverse opal thin films suitable for “on chip” diagnostic or sensing applications, though reliable design strategies toward large-area inverse opal thin films are currently limited, prompting further research in this area.

In response, our group has been targeting new and approved approaches toward inverse opal thin films, primarily metal oxides[42,62−66] and conducting polymers.[67−69] Zirconia (ZrO2) attracts particular interest owing to its high thermal stability and chemical inertness, leading to end applications in high-performance ceramics, catalysis, sensing, and high-temperature fuel cell applications.[48,49,75−80] Herein, we report the first successful fabrication and detailed optical characterization of ZrO2 inverse opal thin films. Evaluation of the performance of the zirconia inverse opal (ZrO2 IO) films for refractive index sensing and catalytic applications was a further aim of the research. At this point, it should be mentioned that most of the methodologies introduced here for the fabrication and characterization of the ZrO2 IO thin films were developed in undergraduate teaching labs in Advanced Physical Chemistry (CHEM 310) and Materials Chemistry (CHEM 380) at the University of Auckland. Indeed, 4–12 weeks undergraduate research projects based on photonic crystal fabrication by the colloidal crystal template method have proved extremely popular with final year BSc students and honors program students, offering a rich tapestry of open-ended learning experiences that encompass, but are not limited to, emulsion polymerization for polymer colloid synthesis, colloid crystallization techniques, colloidal crystal templating strategies, sol–gel chemistry, optical characterization techniques (UV–vis transmittance and reflectance), computational modeling of photonic band-gap structures, as well as hands-on exposure to advanced characterization techniques not normally accessed by undergraduates (scanning electron microscopy (SEM)/energy-dispersive spectrometry (EDS), transmission electron microscopy (TEM), X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), N2 physisorption). Parallel research around natural structurally colored materials such as iridescent beetles, bird feathers, and so forth has allowed students to greatly extend their knowledge of color and how it is generated (i.e., extending beyond the norms such as pigment-based color, transition-metal complexes, charge-transfer complexes, fluorescence and phosphorescence, chemiluminescence), culminating in enhanced student interest in higher postgraduate programs. Methodologies introduced here for inverse opal thin-film fabrication are designed to be easily adopted elsewhere to enhance chemistry-based teaching programs.

Experimental Section

Materials

Methyl methacrylate (99%, Aldrich), 2,2′-azobis(2-methylpropionamidine) dihydrochloride (97%), zirconium(IV) propoxide (70 wt % in 1-propanol), concentrated HCl (37 wt %), tetrachloroauric acid trihydrate (≥99.8%), urea (99%), sodium borohydride (%), and 4-nitrophenol (%) were all obtained from Sigma-Aldrich and used without further purification. Methanol (99%), ethanol (99.4%), acetone (99%), n-heptane (99.5%), benzene (99.7%), dichloromethane (99%), carbon tetrachloride (99.5%), toluene (98%), and bromobenzene (95%) were obtained from local suppliers. Milli-Q water (18.2 MΩ cm resistivity) was used in all experiments.

Synthesis of Monodisperse Poly(methyl methacrylate) (PMMA) Colloids

Three batches of monodisperse PMMA colloids with diameters of 317, 360, and 441 nm (denoted PMMA #1, PMMA #2, and PMMA #3, respectively, in the text below) were synthesized by the surfactant-free emulsion polymerization of methyl methacrylate at temperatures between 70 and 80 °C (Table S1, Figure S3), following procedures introduced by Schroden et al. with a few modifications.[48] Briefly, a volume of methyl methacrylate (0.3 or 0.4 L depending on the batch) and water (1.6 L) was heated under vigorous mechanical stirring and a N2 atmosphere to the desired reaction temperature (70, 75, or 80 °C). Subsequently, 2′-azobis(2-methylpropionamidine) dihydrochloride (1.5 g) was quickly added to initiate the polymerization reaction, which was then continued for 3 h at the specified reaction temperature. Finally, the PMMA colloid suspensions were quickly cooled to room temperature, filtered through glass wool, and stored in glass Schott bottles for later use.

Fabrication of PMMA Colloidal Crystal Thin Films

PMMA colloidal crystal (PMMA CC) thin films on glass microscope slides were fabricated using the flow-controlled vertical deposition (FCVD) method outlined in Figure S4a,b. Briefly, 20–25 mL of the PMMA colloid suspensions obtained in Section was diluted with water to a volume of 500 mL, which was then poured into a vertical-walled glass beaker. Glass microscope slides were then immersed vertically in the diluted colloid suspension, after which a peristaltic pump slowly removed the colloidal suspension (pump rate about 0.3 mL min–1). As the meniscus of the colloid suspension moved down the glass slide, colloidal crystals of thickness 5–7 μm (18–25 layers of colloids) were deposited. In general, the FCVD method proved superior to an evaporation method (Figure S4c) also trialed for colloidal crystal fabrication.

Fabrication of Zirconia Inverse Opal Thin Films

ZrO2 IO thin films were prepared using the colloidal crystal templates fabricated above. Briefly, a Zr(IV) sol was first prepared by adding Zr(IV) propoxide in 1-propanol (5 mL) to 5 mL of methanol under vigorous stirring, followed by the addition of concentrated HCl (2 mL) and water (2 mL). The resulting mixture was stirred for 30 min to give a clear and homogeneous sol, after which it was diluted 5-fold with methanol. For the colloidal crystal infiltration step, the coated glass slides were placed on a slight incline (∼5°), after which 3–4 drops of the diluted Zr(IV) sol were placed on the uppermost edge of the colloidal crystal film. Gravity and capillary action then allowed the sol to infiltrate all the interstitial spaces in the film. After hydrolyzing and drying for 24 h at room temperature, the resulting ZrO2/PMMA CC composites were then calcined using the following protocol: heating from room temperature to 300 °C at 2 °C min–1, holding at 300 °C for 2 h, heating from 300 to 400 °C at 2 °C min–1, then holding at 400 °C for 2 h. Finally, the films were cooled to room temperature and then subjected to detailed characterization.

Fabrication of Au/ZrO2 IO Catalysts

Gold nanoparticles were deposited on ZrO2 IO films at nominal weight loadings of 5 and 18 wt % using a modified version of the procedure described by Zanella.[81] Briefly, the ZrO2 IO films were immersed in a 100 mL aqueous solution containing a certain amount of HAuCl4·3H2O and urea (1 g) and then heated to 80 °C and kept at this temperature for 4 h. By increasing the amount of HAuCl4·3H2O in the solution, the amount of gold deposited on the films could be systematically increased. After reaction for 4 h, the films were rinsed with water, dried in air at 90 °C, and then heated to 300 °C for 1 h to reduce adsorbed Au(OH)3 or Au2O3 to metallic Au0 nanoparticles.

Characterization

SEM micrographs were taken using a Philips XL-30S Field Emission Gun scanning electron microscope (FEGSEM) operated at an electron accelerating voltage of 5 kV in high vacuum. Prior to analysis, specimens were mounted on black carbon tape and sputter coated with platinum for 60 s with a Quorum Q150RS.

TEM images were collected on a Tecnai T12 transmission electron microscope operating at an accelerating voltage of 120 kV, equipped with a 4 megapixel camera. Specimens were dispersed in ethanol, and then a drop of the resulting dispersion was placed on a holey carbon-coated copper TEM grid. The ethanol was then allowed to evaporate at room temperature for several hours before sample imaging.

X-ray powder diffraction patterns were obtained on a Siemens D5000 X-ray diffractometer equipped using a Cu Kα1 source (λ = 1.5418 Å, 40 mA, 40 kV). Data were collected over the 2θ range of 2–100°, with a step size of 0.02° and a scan rate of 2° min–1. ZrO2 crystallite sizes (L) were estimated from the X-ray powder diffraction data using peak line widths and the Scherrer equation.[82]

XPS data was collected on the soft X-ray Beamline at the Australian Synchrotron. The endstation was equipped with a hemispherical electron energy analyzer and an analysis chamber of base pressure ∼1 × 10–10 Torr. Spectra were excited using monochromatic Al Kα X-rays (1486.69 eV). Core-level scans were collected over the Zr 3d and Au 4f regions with an analyzer pass energy of 20 eV. The binding energy scale was calibrated against the C 1s signal at 285.0 eV from adventitious hydrocarbons.

N2 physisorption isotherms for the ZrO2 inverse opals were collected at 77 K on a Micromeritics TriStar 3000. Specific surface areas and pore sizes were calculated from N2 physisorption isotherms according to the Brunauer–Emmett–Teller (BET)[83] and Barret–Joyner–Halenda[84] methods. Prior to analysis, samples were heated at 180 °C overnight in a vacuum oven.

UV–vis transmittance spectra for the PMMA colloidal crystal films and ZrO2 inverse opal films were recorded over the wavelength range 300–1100 nm on a Shimadzu UV-1700 spectrometer, equipped with a custom-built sample holder for UV–vis transmittance measurements at different film angles to the sample beam path.

The photonic band-gap structure diagram for the ZrO2 IO films was created using the MIT photonic-band package. Fully vectoral eigenmodes of Maxwell’s equations with periodic boundary conditions were computed by preconditioned conjugate-gradient minimization of the block Rayleigh quotient in a planewave basis, along the high-symmetry directions U-L-Γ-X-W-K for an fcc first Brilluoin zone.[85] The additional parameters for ZrO2 calculations are as follows: rsphere = 1.06 and neff = 2.1.

Catalytic Activity Tests

The reduction of 4-nitrophenol at room temperature with sodium borohydride (NaBH4) as the reductant was used to test the activity of the ZrO2 IO and Au/ZrO2 IO catalysts. In a typical experiment, catalyst (2 mg) was added to aqueous 4-nitrophenol (0.15 mM, 25 mL) with stirring. Then, 2.75 mL of the resulting dispersion was added to a standard quartz cuvette (path length 1 cm). To start the reaction, 0.25 mL of freshly prepared NaBH4 (0.1 mol L–1) was then added. The reaction progress was monitored by collecting UV–vis absorption spectra every 30 s from 250 to 600 nm. The absorbance at 400 nm was monitored as a function of time, and pseudo-first-order rate constants (k′) for the reduction of 4-NP were calculated from the slopes of plots of ln(A/A0) versus t.

Results

Structural and Optical Characterization of the PMMA Colloidal Crystal Films

The colloidal crystal template technique relies heavily on the synthesis of monodisperse spherical colloids and their subsequent self-assembly to form well-ordered colloidal crystals. Accordingly, preliminary SEM and UV–vis characterization studies were performed on PMMA CC #1, PMMA CC #2, and PMMA CC #3 thin films to assess their suitability as templates for ZrO2 inverse opal fabrication. Figure shows the SEM images along the [111] direction for films fabricated from each PMMA colloid batch. In each case, the films crystallized with their fcc (111) planes parallel to the underlying glass substrate. Further, each batch was found to be highly monodisperse, with all colloids in each batch having the same size resulting in very sharp spots in the Fourier transforms performed on the SEM images (see insets in Figure a,c,e). In the higher magnification images, small necks can be seen between neighboring colloids, which likely formed during the drying of the colloidal crystals. This necking adds a degree of mechanical strength to the PMMA colloidal crystals, which is beneficial for colloidal crystal templating (see below). The diameters of the colloids, calculated as the center-to-center distance between adjacent colloids on the (111) planes, were 317, 360, and 441 nm for PMMA CC #1, PMMA CC #2, and PMMA CC #3, respectively (Table ). The colloidal crystal films developed some cracks during fabrication and SEM imaging (Figure S5a) but generally had very acceptable structural uniformity. Colloidal crystal film thicknesses were typically in the range of 5–7 μm (Figures S4d and S5b).

Figure 1

SEM images for the PMMA colloidal crystal films viewed along the [111] direction. (a, b) PMMA CC #1; (c, d) PMMA CC #2; and (e, f) PMMA CC #3. Images on the left were taken at 20 000× magnification, those on the right at 100 000× magnification. The insets in (a), (c), and (e) show Fourier transforms of the corresponding image and confirm a high degree of structural order.

Table 1

Summarized Structural and Optical Data for the PMMA Colloidal Crystal Films and Their ZrO2 Inverse Opal Replicasa

		optical data and derived constantsb
sample	BET surface area (m2 g–1)	DSEM (nm)	DUV–visb (nm)	navg	ϕ	λmax (nm) θ = 0°	Δλ/λmax	a/λmax
PMMA CC #1		317	319	1.35	0.72	707	0.073	0.63
PMMA CC #2		360	362	1.36	0.72	802	0.074	0.63
PMMA CC #3		441	439	1.37	0.74	983	0.074	0.64
ZrO2 IO #1	47.7	260	242	1.16	0.16	462	0.125	0.79
ZrO2 IO #2	44.9	287	281	1.15	0.15	526	0.120	0.77
ZrO2 IO #3	47.6	342	345	1.12	0.12	635	0.128	0.77

D = center-to-center distance between spheres on fcc (111) planes.

DUV–vis, navg, and ϕ determined from plots of λmax2 versus sin2 θ.

The PMMA colloidal crystals all showed structural color, especially when viewed at grazing angles that shifted their photonic band gaps (PBGs) along the [111] and [200] directions into the visible region. UV–vis transmittance spectra for the PMMA colloidal crystal films collected along the [111] direction, that is, normal to the (111) planes, are shown in Figure . PMMA CC #1, PMMA CC #2, and PMMA CC #3 gave intense PBGs along the [111] direction (θ = 0°) at 707, 802, and 983 nm, respectively (Table ). Note that here we have collected transmittance spectra rather than reflectance spectra, and as such the PBGs will appear as sharp dips in transmittance (by convention we still use the term λmax here to describe the PBG position). The shift in the PBG position (λmax) to longer wavelengths with increasing colloid diameter (D), and hence increasing interlayer spacing along the [111] direction, is in perfect accord with predictions of the modified Bragg’s law expressions below,[48,63] which consider both refraction and diffraction of light in opal-based photonic crystals (Figure S2). For first-order diffraction on any fcc plane of an opal-based photonic crystal (including inverse opals), the position of the PBG can be described by the following equationwhere λmax is the PBG position (in nm), d is the interlayer spacing for a particular (hkl) plane, θ is the incident angle of light with respect to the normal of the plane, and navg is the average refractive index of the photonic crystal, typically calculated as navg = [ϕnsolid + (1 – ϕ)nvoid], where ϕ is the solid volume fraction and nsolid and nvoid are the refractive indices of the solid and void components, respectively. Note for a PMMA colloidal crystal in air with an ideal solid volume fraction of 0.74, navg = (0.74 × 1.492 + 0.26 × 1.00) = 1.364. For first-order diffraction on fcc (111) planes, eq transforms to eq since d111 = 0.8165D.When θ = 0°, eq reduces further toThus, the PBG position for first-order diffraction along the [111] direction should be directly proportional to D (i.e., λmax = 2.23 × D), in excellent accord with the shift in the PBG position to longer wavelengths seen on going from PMMA CC #1 to PMMA CC #3 (Figure and Table ). The normalized PBG bandwidths (Δλ/λmax) and PBG angular frequencies (a/λmax) were near identical for all three different colloidal crystals, consistent with their identical construction (differing only in the colloid diameter). As predicted by eq , the PBGs along the [111] direction blue-shifted as the incident angle of light with respect to the [111] direction increased (Figure ). Plots of λmax2 versus sin2 θ yielded straight lines (r2 > 0.995, Figure d). Values of D, navg, and ϕPMMA determined from the slopes (−1.6332D2) and y-axis intercept (1.6332D2navg2) for the colloidal crystal films were in excellent accord with expectations (Table , columns 4–6).

Figure 2

UV–vis transmittance spectra collected in air at incident angles of 0, 5, 10, 15, 20, and 25° with respect to the [111] direction for (a) PMMA CC #1; (b) PMMA CC #2; and (c) PMMA CC #3. As the incident angle of light increased from 0 to 25°, the [111] PBGs blue-shifted in accordance with eq . (d) Plot of λmax2 versus sin2 θ for the PMMA CC films.

Structural and Optical Characterization of ZrO2 Inverse Opal Films

After the colloidal crystal templating step, brightly colored films were obtained, indicating the success of the inversion process. Whereas the PMMA colloidal crystals (opals) were constructed from solid spheres with air in between, the inverse opal replicas comprised a face-centered cubic array of air spheres (macropores) in a ZrO2 matrix (Figure ). When viewed along the [111] direction, the ZrO2 IO #1, ZrO2 IO #2, and ZrO2 IO #3 films appeared purple, green, and orange, respectively. UV–vis transmittance spectra collected along the [111] direction (θ = 0°) revealed band gaps at 462, 526, and 635 nm (Figure ), consistent with the reflected color of each film. In agreement with eq , the PBG along the [111] direction shifted to shorter wavelengths as the incident angle of light was increased with respect to the [111] direction. Values of D, navg, and ϕZrO obtained from plots of λmax2 versus sin2 θ (Figure d) are listed in Table . Figure e,f shows a ZrO2 IO #2 film illuminated and viewed along the [111] direction and at an angle of 30° with respect to the angle of illumination, respectively. Note the striking color change with angle—graphic evidence for the successful realization of structural color in the films.

Figure 3

SEM images for ZrO2 inverse opal films viewed along the [111] direction; (a, b) ZrO2 IO #1; (c, d) ZrO2 IO #2; and (e, f) ZrO2 IO #3. Images on the left were taken at 50 000× magnification and those on the right at 100 000× magnification. The insets in (a), (c), and (e) show digital photographs of the ZrO2 inverse opal thin films in air, illuminated and viewed along the [111] direction.

Figure 4

(a–c) UV–vis transmittance spectra collected in air for the ZrO2 inverse opal films at incident angles of 0–25° with respect to the [111] direction. The spectra were collected in 5° increments. (d) Plot of λmax2 versus sin2 θ for the ZrO2 inverse opal films. (e) Digital photograph of ZrO2 IO #2 illuminated and viewed along the surface normal, and (f) the same film illuminated along the surface normal but viewed at an angle of around 30° with respect to the surface normal.

As with the PMMA colloidal crystal templates, the PBG position along the [111] direction for the ZrO2 IOs shifted to longer wavelengths as D increased (Table and Figure a). The normalized PBG bandwidths (Δλ/λmax) and PBG angular frequencies (a/λmax) were also very similar for the three different samples, consistent with expectations for ZrO2 IOs of comparable solid volume fraction (0.12–0.16). The a/λmax value of 0.77 found for ZrO2 IO #2 and ZrO2 IO #3 agreed perfectly with the value predicted in the photonic band-gap structure diagram for a ZrO2 IO with a solid volume fraction of around 0.12 (Figure b). In the diagram, a pseudo photonic band gap opens between the second and third bands along the L → Γ or [111] direction at a/λmax = 0.77, as indicated by the blue shaded area.

Figure 5

(a) Plot of position of the photonic band gap along the [111] direction and the normalized band-gap width (Δλ/λ) versus the center-to-center distance between macropores (D) on fcc (111) planes in the ZrO2 inverse opal films. (b) Photonic band-gap structure diagram for a ZrO2 inverse opal, showing a pseudo photonic band gap at a/λ = 0.77 (indicated by the blue shaded area) between the second and third bands along the fcc L → Γ direction.

It should be noted here that on transforming the PMMA colloidal crystals to their inverse opal replicas, the center-to-center distances between spheres on the fcc (111) planes decreased by ∼18–22%. The theoretical maximum solid volume fraction for the inverse opals is 0.26, but the actual values were approximately one half of this (ϕZrO = 0.12–0.16). The data suggests that the during the template removal process, the films cracked and shrunk to form small islands on the glass microscope slides, though clearly the islands retained the desirable (111) orientation of the PMMA colloidal crystal templates (Figure ). Further evidence in support of this theory is found in the SEM images of Figure S5c–f, where an island of ZrO2 IO is seen (Figure S5e). In addition, some cracks and other defects can be seen in film cross sections (Figure S5c,d), although these defects seem to have little detrimental effect on the optical properties of the ZrO2 IO #2 film.

ZrO2 can exist in three different polymorphic forms, monoclinic, tetragonal, and cubic, with the cubic form being the most thermodynamically stable.[86,87] Powder X-ray diffraction patterns (Figure S6) for the ZrO2 IOs confirmed that all were composed of tetragonal ZrO2 (space group P42/nmc), consistent with the literature related to ZrO2 crystals obtained by mild calcination of amorphous zirconia precursors.[86,87] The broadness of the XRD peaks is evidence for the presence of nanocrystalline ZrO2 in the walls of the inverse opals. Using the Scherrer equation,[82] ZrO2 crystallite sizes were estimated to be 3–5 nm. The ring structures seen in the selected area diffraction pattern for ZrO2 IO #2 (Figure a) validated this size estimate.

Figure 8

(a, b) TEM images for ZrO2 IO #2 and (c, d) 5 wt % Au/ZrO2 IO #2. The images in (c) and (d) reveal well-dispersed Au nanoparticles of size around 5–7 nm; (e, f) shows digital photographs of the 5 wt % Au/ZrO2 IO #2 film in air and after wetting with ethanol. The inset in (a) shows a selected area diffraction pattern (scale bar = 0.2 Å–1). The well-developed ring structure confirms the presence of small nanocrystals of tetragonal ZrO2 (P42/nmc).

Refractive Index Sensing Using ZrO2 Inverse Opal Films

One of the most fascinating and intriguing features of the ZrO2 IO thin films was their dramatic color change on being wetted by water or other liquids. As indicated in eqs –3, the photonic band-gap position (λmax) is highly dependent on the average refractive index of the photonic crystal. On filling the macropores in the ZrO2 IO films with liquids of increasing refractive index, navg will increase, and thus the PBG along the [111] direction is predicted to shift progressively to longer wavelengths (i.e., red shift). Figure a,b shows UV–vis transmittance spectra collected for ZrO2 IO #1 and ZrO2 IO #2 films, respectively, in air and then liquids of increasing refractive index. Liquids used were methanol (n = 1.329), ethanol (n = 1.361), n-heptane (n = 1.388), dichloromethane (n = 1.4241), carbon tetrachloride (n = 1.4601), benzene (n = 1.501), and bromobenzene (1.558). As the refractive index of the liquid increased, the PBG along the [111] direction shifted progressively to longer wavelengths. The large change in the color of the ZrO2 IO #1 and ZrO2 IO #2 films on going from air to ethanol is apparent in the digital photos provided in Figure . ZrO2 IO #1 changed from purple to yellow, and ZrO2 IO #2 changed from green to red. Once the ethanol evaporated from the films, the original purple and green colors of the films returned.

Figure 6

(a, b) UV–vis transmittance spectra collected along the [111] direction for different ZrO2 inverse opal thin films in air (n = 1.000) and solvents of increasing refractive index (ranging from methanol, n = 1.329, to bromobenzene, n = 1.558). For both samples, the PBGs progressively red-shifted as the refractive index of the solvent filling the macropores in the films increased. Insets in (a) and (b) shows contact angle data for a water droplet on each ZrO2 IO film. The digital photographs below the spectra show each ZrO2 IO film in air and ethanol, separately.

Surprisingly, the ZrO2 IO films were quite hydrophobic, giving water droplet contact angles ranging from 120 to 124° (Figure a,b, insets). These values were independent of the macropore diameter in the ZrO2 IOs. By comparison, SiO2 IO and TiO2 IO samples prepared by the same colloidal crystal templating method (discussed briefly below) were hydrophilic and easily wetted by water. This result suggests that the combination of inverse opal structure (i.e., very spherical and regularly sized macropores) and the unique physical properties of the very small ZrO2 crystallites in the walls of the inverse opals was responsible for the unexpected hydrophobicity displayed by the ZrO2 IO films.[88−90]

Returning to the data in Figure a,b, we noted the PBG positions for each ZrO2 IO in air and each liquid, which were then used to create plots of PBG position versus solvent refractive index (Figure ). The plots for all three ZrO2 IO samples were remarkably linear (r2 > 0.999). This linearity can be understood by substituting [ϕnsolid + (1 – ϕ)nvoid] for navg in eq , which leads to the following equationHere, nvoid represents the refractive index of the medium (i.e., solvent) filling the void space in the inverse opals. In agreement with experimental observations, a plot of λmax versus nsolvent is predicted to be linear, with slope = 1.633D(1 – ϕ) and y-axis intercept = 1.633Dϕ nZrO. Using the slopes and intercepts of the straight lines in Figure , we calculated values of D and ϕZrO for the ZrO2 IO #1, ZrO2 IO #2, and ZrO2 IO #3 films using the following expressions[48,63]Using nZrO = 2.1, the D values obtained were 251, 284, and 331 nm, respectively, and the ϕZrO values obtained were 0.158, 0.122, and 0.115, respectively, all in excellent accord with the values determined for the same parameters using other methods (Table ). The very strong linearity of the plots in Figure suggested the ZrO2 IO films could be used in refractive index sensing. To determine the refractive index of an unknown liquid, all that would be required is to determine the PBG position along the [111] direction using one of the ZrO2 IO films. Once the PBG position was determined, the refractive index of the unknown liquid could be extrapolated from the plot in Figure for that same ZrO2 IO or alternatively calculated directly using the linear regression equation.

Figure 7

Plot of the PBG position (λmax) along the [111] direction versus the refractive index (nsolvent) of the solvent filling the macropores in the ZrO2 inverse opal thin films. For all samples, a strong linear relationship was observed (r2 > 0.999), in agreement with eq .

To test the sensitivity of the ZrO2 IO films for refractive index sensing, we collected UV–vis spectra for the different films in acetone (n = 1.3587) and ethanol (n = 1.3614), which differ by only 0.0027 refractive index units. For the ZrO2 IO #1 film, a PBG shift to longer wavelengths of 1.2 nm was detected on going from acetone to ethanol, with a slightly large shift of 1.4 nm found for the ZrO2 IO #2 film. Based on this result, we thus estimate that the ZrO2 IO films should be able to discern two liquids with a refractive index difference as small as 0.0005.

Optical Properties of Gold Nanoparticle-Decorated ZrO2 Inverse Opal Films

Fascinated by the optical properties of the bare ZrO2 inverse opal films, we decided to explore this aspect further by decorating the films with gold (Au) nanoparticles. Unlike bulk gold, which is yellow, gold nanoparticles display a range of size-dependent colors due to localized surface plasmon resonances (LSPRs).[63,91,92] The LSPRs are a collective oscillation of valence electrons on the surface of the gold nanoparticles induced by the electric field of light at the resonance frequency. For spherical gold nanoparticles of size <20 nm in water, the LSPR absorption maximum is typically observed around 520 nm (i.e., the nanoparticles absorb light strongly in the green part of the visible spectrum), and we see the other reflected wavelengths (especially the red part of the visible spectrum). For this reason, a vial of small spherical Au NPs in water will look red to the human eye. Similarly, stained glass windows containing Au nanoparticles appear red to the observer. We were particularly interested in decorating the ZrO2 IO #2 film with gold nanoparticles since that film has a PBG along the [111] direction at 526 nm where the gold nanoparticles were expected to absorb. This offered the opportunity to study two competing optical phenomena (selective diffraction and reflection by the ZrO2 IO #2 film and strong absorption at comparable wavelengths by the Au nanoparticles).

Figure shows transmission electron microscopy (TEM) images for the ZrO2 IO #2 film (Figure a,b) and the 5 wt % Au/ZrO2 IO #2 film (Figure c,d). In both cases, the films were scraped off the underlying glass substrate for imaging, which did not adversely affect the ordered structure of the films. The selected area diffraction pattern for the ZrO2 IO #2 film showed a characteristic ring structure, which supported the earlier XRD findings that the films were composed of very small nanocrystalline tetragonal ZrO2 crystallites (<5 nm). After decoration with Au nanoparticles at a loading of 5 wt %, small Au nanoparticles of size 5–8 nm could be discerned in the TEM images. The small size and excellent dispersion of the Au nanoparticles over the ZrO2 IO support are consistent with the known strong metal–support interaction between Au and ZrO2. The XRD pattern for the 5 wt % Au/ZrO2 IO #2 film (Figure S8) confirmed the successful deposition of Au nanoparticles, with new broad peaks characteristic for small Au crystals appearing (especially apparent in the XRD pattern of the 18 wt % Au/ZrO2 IO #2 film, Figure S8). XPS data in Figure S9 also confirmed the successful deposition of metallic Au nanoparticles, with the Au 4f signal intensities increasing in proportion to the nominal Au loadings. As expected, the Zr 3d signals were attenuated slightly as the Au loading increased. When illuminated and viewed along the [111] direction, the Au/ZrO2 IO #2 film appeared green in air (Figure e). Thus, the photonic crystal was rejecting wavelengths of light typically used by the Au nanoparticles for absorption (i.e., the PBG properties of the ZrO2 IO film appeared to dominate the perceived color along the [111] direction; note: a noninverse opal Au/ZrO2 specimen would certainly appear purple under the same illumination conditions). On adding a drop of ethanol to the film, the PBG for the 5 wt % Au/ZrO2 IO #2 film was red-shifted considerably and decoupled from the Au LSPR, resulting in a brightly red-colored film (Figure f) whose color is the sum of both PBG diffraction and Au LSPR absorption.

The LSPRs of Au nanoparticles are known to be sensitive to the refractive index of the surrounding medium[91,92] and have been used previously to develop refractive index sensors for liquids. The data in Figures and 7 revealed that the ZrO2 IO films were extremely effective for refractive index sensing. Accordingly, it was of novelty to see if the 5 wt % Au/ZrO2 IO #2 film could be used to get a double determination of liquid refractive index from a single UV–vis spectrum. Figure a shows UV–vis transmittance spectra for the 5 wt % Au/ZrO2 IO #2 film in air (n = 1.000), methanol (n = 1.329), ethanol (n = 1.362), n-heptane (n = 1.388), dichloromethane (n = 1.424), carbon tetrachloride (n = 1.460), and benzene (n = 1.501). In air, the ZrO2 IO #2 PBG along the [111] direction and the Au LSPR absorption were almost perfectly overlapped, giving a broad feature centered at 532 nm (i.e., responsible for the green reflected color seen in Figure e). On immersion in the various organic solvents, the Au LSPR and ZrO2 IO PBG decoupled, giving two distinct features. The Au LSPR feature was centered between 545 and 556 nm, and the ZrO2 IO #2 PBG centered between 663 and 754 nm. In both cases, these features shifted linearly to longer wavelengths on increasing the refractive index of the liquid the film was immersed in (Figure b). Accordingly, it can be concluded that the 5 wt % Au/ZrO2 IO #2 film allows two parallel determinations of the refractive index of a liquid from a single UV–visible transmittance spectrum measurement. Using the slope and intercept of the PBG regression line in Figure b, values of D = 288.4 nm and ϕZrO = 0.11 were calculated using eqs and 6, respectively (nZrO = 2.1 was assumed). These values are in good agreement with data for bare ZrO2 IO #2 (D = 287 nm, ϕZrO = 0.15, Table ), indicating that gold Au nanoparticle deposition did not really alter the geometric properties of the ZrO2 IO film significantly.

Figure 9

(a) UV–vis transmittance spectra collected along the [111] direction for a 5 wt % Au/ZrO2 IO #2 thin film in air (n = 1.000) and solvents of increasing refractive index (ranging from methanol, n = 1.329, to benzene, n = 1.501). Both the PBG and the Au LSPR progressively red-shifted linearly as the refractive index of the solvent in the film increased, as shown in (b). The inset in (b) shows how the hydrophobicity of the ZrO2 IO #2 film decreased after Au nanoparticle deposition (top = before Au deposition; bottom = after Au deposition).

Catalytic Activity of the Gold Nanoparticle-Decorated ZrO2 Inverse Opal Films for 4-Nitrophenol Reduction

As a further demonstration of the functionality of the Au/ZrO2 IO #2 films fabricated in this study, we decided to evaluate the catalytic activity of the as-prepared films, as well as films decorated with Au nanoparticles at two different loadings (5 and 18 wt %), for the reduction of 4-nitrophenol using NaBH4 at the reducing agent.[93,94] In the presence of a large excess of NaBH4, the kinetics of this reaction, which is catalyzed by metal nanoparticles, obeys pseudo-first-order kinetics, as described by the following integrated rate lawwhere A is the absorbance of the 4-nitrophenol solution in water at 400 nm at time t, A0 is the initial absorbance of the 4-nitrophenol solution in water at 400 nm (i.e., at t = 0 min), k′ is the pseudo-first-order rate constant (units min–1), and t is the reaction time (in min). The reaction commences as soon as the reducing agent (NaBH4) is added to the 4-nitrophenol solution containing the catalyst. Figure a shows the UV–vis absorption spectra for an aqueous solution containing 4-nitrophenol, NaBH4, and ZrO2 IO. A gradual decrease in the absorbance of the 4-nitrophenol peak at 400 nm is seen over time (24 min), which can largely be attributed to 4-nitrophenol adsorption on the ZrO2 IO surface (rather than catalytic reduction). The rate constant of 0.026 min–1 was understandably low. After decoration with Au nanoparticles, known active catalysts for 4-nitrophenol reduction in the presence of NaBH4, the rate constant jumped dramatically to k′ = 0.740 min–1 for the 5 wt % Au/ZrO2 IO #2 sample and then to k′ = 1.927 min–1 for the 18 wt % Au/ZrO2 IO #2 sample. The evolution of the main reaction product, 4-aminophenol, during the reaction is evident by the development of the peak at 295 nm in the spectra. In the case of the 18 wt % Au/ZrO2 IO #2 sample, the reaction proceeded so rapidly that it was almost impossible to follow by UV–vis spectroscopy (reaching completion in 2 min). The rates reported here accord well with the best literature data for 4-nitrophenol reduction over the supported gold catalysts.[94−97] This high activity can be understood in terms of the open macroporous structure of the ZrO2 IO support that allows facile transport of reagents to the Au active sites. The high gold nanoparticle dispersion realized in the Au/ZrO2 IO system (Figure c,d) is also advantageous in this context. Supported Au nanoparticle catalysts offer many advantages over unsupported Au nanoparticles in catalytic applications, especially the stabilization afforded against nanoparticle aggregation. After storage for 1 year at room temperature in closed vials, the 5 wt % Au/ZrO2 IO #2 and 18 wt % Au/ZrO2 IO #2 catalysts demonstrated identical performance as the freshly prepared catalysts for 4-nitrophenol reduction.

Figure 10

UV–vis absorbance spectra for aqueous solutions of 4-nitrophenol (0.15 mM, 2.75 mL) and NaBH4 (0.1 M, 0.25 mL) in the presence of (a) ZrO2 IO #2; (b) 5 wt % Au/ZrO2 IO #2; and (c) 18 wt % Au/ZrO2 IO #3. (d) Plots of ln(A/A0) versus time.

Author Commentary on the Versatility of Colloidal Crystal Template Strategy

The data above demonstrates the wonderful structural color palette that is available by fabricating a single type of inverse opal film. By varying the diameter of the PMMA colloids in the colloidal crystal templates, it is possible to controllably modify the position of PBGs along the [111] direction in the ZrO2 IO films. By rotating the samples with respect to the source of illumination, the PBG along the [111] direction can be shifted to shorter wavelengths, and by filling the macropores in the inverse opals with liquids of different refractive indices, it is possible to redshift the PBG over a large range of wavelengths. Through systematic manipulation of all of these parameters in the current study, we have been able to recreate, via structural engineering, all of the vivid colors exhibited by the natural photonic crystal specimens in Figure S1. Importantly, all of the optical phenomena we have reported relating to the ZrO2 IO films are accurately captured by the modified Bragg’s law expressions described in eqs –3. Further, by simply substituting the Zr(IV) propoxide precursor with tetraethylorthosilicate or titanium (IV) propoxide and changing the solvent from methanol to ethanol, keeping all other fabrication procedures the same, amorphous SiO2 IO or nanocrystalline TiO2 IO films, respectively, can readily be accessed (Figure S10). All of the aformentioned aspects make colloidal crystal templating a fascinating topic for advanced-level research as well as undergraduate teaching laboratories. We encourage others to enter this research space.

Conclusions

ZrO2 inverse opal thin films with striking angle-dependent structural color at visible wavelengths were successfully fabricated by the colloidal crystal template technique. The optical properties of the films were dominated by first-order diffraction on fcc (111) planes, with excellent accord established between the optical properties of the films and predictions based on a modified Bragg’s law expression. Refractive index sensing with a sensitivity estimated around ∼0.0005 units was demonstrated using the ZrO2 inverse opal thin films, based on the progressive linear shift in the PBG position along the [111] direction with increasing solvent refractive index. Further, by functionalizing the ZrO2 IO films with gold nanoparticles, a novel refractive index sensing platform was created that allowed parallel determinations of liquid refractive index (one strategy used the small shift in the Au LSPR position, and the other utilized the large shift in the PBG position as the refractive index of the solvent in the Au/ZrO2 IO films increased). Due to their inherent macroporosity, ZrO2 IOs represent near-ideal supports for noble metal catalysts, which we confirmed through the successful application of Au/ZrO2 IO films to the reduction of aqueous 4-nitrophenol in the presence of NaBH4. Results of this study are expected to encourage others to utilize the colloidal crystal template approach to fabricate novel photonic crystal thin films for sensing and catalytic applications.