1. Introduction

Colloidal crystal monolayers have attracted a lot of interest because of their potential application as templates for the fabrication of metallic nanostructures. Different approaches have been proposed and many kinds of periodically nanostructured plasmonic materials were obtained based on the close-packed arrangement of dielectric colloids (usually polystyrene, PMMA, or silica): metal-coated sphere arrays [1–5], hole arrays [6–9], arrays of triangular particles [10–13], split-ring resonators [14–17], nano-cup arrays [18,19], and many others [20–22]. Due to their refractive index periodicity (e.g. air and polymer), colloidal crystals are also very interesting on their own as a photonic system. They prompted several studies focused on their optical response, in particular the study of their photonic stop-gaps [4,23–28]. One of the simplest approaches to produce plasmonic crystals is the deposition of a thin metallic film on top of the dielectric colloidal sphere arrays. These materials are known from the early work of Van Duyne as Metal Films over Nanospheres [29] or, more recently, when more focus was set on their optical response, as hybrid colloidal plasmonic-photonic crystals (HPPCs) [30]. The extended capabilities and functionalities of hybrid metal dielectric photonic crystals have been very well synthesized by Romanov [30]. A highlight of these structures is the coupling/hybridization of pure photonic modes from the dielectric sphere array and surface plasmon modes in the overlaying metal film. This structure was also studied in the context of the Extraordinary Optical Transmission (EOT) phenomenon, originally observed on periodic subwavelength hole arrays [31]. HPPCs exhibit a transmission band which was attributed to propagative surface plasmons, as in EOT. One distinctive advantage of HPPCs is that the use of spheres of different diameters allows to tune very precisely the enhanced transmission band from the UV-visible farther to the infrared [4]. HPPCs thus offer a unique platform allowing one to perform relevant and systematic studies on fundamental plasmonics. Besides EOT, other interesting phenomena, such as directional emission from fluorophores located within the colloidal spheres, were also observed [30]. However, based on the easy tunability of the optical/plasmonic response and their low production costs, HPPCs also prompted a range of applications covering notably Surface Plasmon Resonance sensing, Surface Enhanced Raman Spectroscopy, and Surface Enhanced Fluorescence [29,32–35].

More recently, researchers proposed modifying the morphology of the colloidal crystals (CCs) through reactive ion etching (RIE), thermal treatment or chemical etching [6,36–39]. Almost 20 years ago, RIE was already used to reduce the size of polymer colloids, but only as a mean of modifying the monolayer as mask for patterning the underlying substrate. However, few examples exist in which the non-closed-packed arrays (period remains the same while sphere diameters are decreased through etching) were used to fabricate HPPCs (for convenience, we will refer to the etched HPPCs, as Et^HPPCs). Such studies demonstrated a dependence of the optical response on the degree of etching of the colloidal crystals. Li et al [40] demonstrated that by controlling the etching time and thickness of the gold film deposited on top the CC, the reflectivity spectra of the Et^HPPCs can be tuned. Xia et al. [41] analyzed the SERS activity of such structures, and observed higher enhancements on Et^HPPCs than on normal HPPCs. Despite these interesting existing results, a thorough analysis of the plasmonic properties, both in the far-field and near-field zones, e.g. how the enhanced transmission of HPPCs is modified in Et^HPPCs, is still missing in the literature.

In this work we investigate the optical transmission of Et^HPPCs in the visible spectral range. We start by preparing a monolayer of 500nm diameter polystyrene spheres which we then etch in gradual and controlled steps, studying its optical response as a function of the etching degree. The next step is the deposition of a 50nm thick gold film on top of the Et^CCs, so that the different morphologies each provides its own optical response. In order to confirm our experimental results, a series of numerical simulations using Lumerical Inc. FDTD [42] software were conducted. Furthermore, the electromagnetic near-field distributions generated within these Et^HPPCs have been numerically obtained in order to get some insight about the main differences between the optical responses of the gradually etched Et^HPPC. Our results show how the optical/plasmonic response of HPPCs can be modified in a controlled way and constitute a novel contribution to the understanding of the physics underlying these nano-patterns.

2. Experimental / Methods

2.1. Sample fabrication

Colloidal crystals made of polystyrene (PS) colloids were obtained using the convective assembly technique, in a configuration similar to the one proposed by Velev [43]. Our home-built system allows controlling the translation speed, temperature of the substrate, angle and position of the deposition blade. The method consists in injecting a water droplet containing polystyrene spheres of 500nm diameter into the wedge between two glass slides positioned in close proximity to each other and having a tilted angle between them. The bottom slide was kept at a constant temperature of 25°C using a cooling home built system that flows water through the aluminum block that holds the slide. As the bottom slide slowly translates, a 2D array of microspheres self-assembles on the substrate due to water evaporation and the continuous flow of spheres from the solution droplet.

Reactive Ion Etching was used in order to etch the polystyrene nanospheres. By controlling the flow of oxygen inside the etching chamber, the power of the generator as well as the etching time, we were able to alter the diameter of the polystyrene microspheres, while maintaining the 500 nm lattice period. The samples discussed in this paper were prepared at a constant pressure of 0.75 mbar in the chamber for an etching time of 120 s. The variation parameter was the power of the generator, adjusted from 10% to 50% (10% steps) of the total 300 W power of the 13.56 MHz RF generator.

On top of the polystyrene sphere arrays a 50 nm gold film was thermally evaporated inside a vacuum chamber having pressures below 10⁻⁶ mbar. The rate of deposition was beforehand stabilized using a quartz balance in order to calculate the time needed for the deposition. The substrates were rotating around an axis perpendicular to them during the deposition process, in order to maintain a constant deposition rate on the entire area of the substrate.

2.2. Characterization

The optical transmission and reflectivity spectra were measured using the unpolarized light source of either a Zeiss Axio Observer Z1 microscope, or a WITec alpha 300 Raman microscope, focused by means of a 10x objective and driven by an optical fiber (0.6mm entrance) to a portable spectrometer (Ocean Optics USB400 UV-Vis). Scanning electron microscopy pictures were recorded by using a field emission scanning electron microscope JEOL JSM 6700F operated at 5 kV. AFM measurements were performed on a WITec alpha system. To obtain reliable height determinations a small sample area was sacrificed, by scratching the surface of the sample with a needle. AFM measurements were made at the border of that scratch.

2.3. FDTD simulations

The optical and electromagnetic properties of the substrates were analyzed using a three-dimensional finite-difference time-domain software (FDTD Solutions, from Lumerical Solutions, Inc.) The geometry of these substrates consists in a hexagonal close-packed two dimensional array of 500 nm polystyrene spheres having their centers in the XY plane. The gold metal deposition was simulated using gold spheres with a diameter larger than the polystyrene spheres by 10 nm and having their centers shifted with 40 nm in the + Z direction in order to obtain a 50 nm thick gold nano-cap. The triangular nanostructures formed directly on the glass substrate were also modeled, by a thin metallic film from which the geometrical projections of the polystyrene spheres were excluded. The simulation space was constructed using periodical boundary conditions along the X- and Y-axes. Excitation of the structure was done using a plane wave source placed above the nanostructures propagating along the Z-axis and polarized parallel to the X- or Y- axis. For the optical properties of the materials we have used a wavelength independent index of 1.585 for the polystyrene spheres and of 1.578 for the glass substrate together with the permeability and permittivity of gold from the software database as described by Johnson and Christy. Transmission spectra were obtained by placing a 2D monitor parallel to the XY plane, below the nanostructures inside the substrate, and from which only the zero-order was extracted in order to better match experimental conditions.

3. Results and discussion

3.1. Morphology

Figure 1 schematically depicts the morphology of the periodic metal-dielectric nanostructures investigated in this paper, shown both as a top view (top row) and a cross-section view (bottom row). ‘Standard’ HPPCs (Figs. 1(e) and 1(f)) are obtained by preparation of a monolayer of polystyrene microspheres (Figs. 1(a) and 1(b)), which are then coated by a metal film (gold in our case). Alternatively, the PS sphere monolayers are etched by means of plasma RIE (Figs. 1(c) and 1(d)), and then coated by a metal film, in order to construct a morphology such as the one depicted in Figs. 1(g) and 1(h). The etching process can be controlled by three main parameters: oxygen pressure inside the chamber, power of the generator and duration of the etching. We observed that, in order to increase the etching effect, we need to either decrease the pressure, increase the power or increase the duration. These observations are also in agreement with some previously published literature [6,36–39].

 

Fig. 1 Scheme depicting a top view (top) and a cross-section view (bottom) of a CC (A and B), Et^CC (C and D), HPPC (E and F) and Et^HPPC (G and H).

Download Full Size | PDF

For the purpose of this study we selected a series of samples which were prepared by gradually increasing the etching power. Morphological changes induced by the etching process were investigated by scanning electron microscopy and atomic force microscopy. A first observation is that by increasing the power of the plasma, the colloidal spheres are etched to an increasing degree, i.e. their diameter decreases, as can be seen in the SEM images in Fig. 2.

 

Fig. 2 SEM images of: (A,E) HPPC, and (B-D and F-H) Et^HPPC. (I) The dependence of the etched spheres diameter on plasma power for the Et^HPPC samples. Scale bars are 2 µm (A-E) and 0.5 µm (E-H).

Download Full Size | PDF

The diameters of the etched spheres reach average values of 496 nm, 468 nm, and 446 nm in samples S1 (Fig. 2(b) and 2(f)), S3 (Fig. 2(c) and 2(g)), and S5 (Fig. 2(d) and 2(h)) respectively. The dependence of the etched spheres average diameter on plasma power for the Et^HPPC samples is represented in Fig. 2(i), demonstrating that the diameter of the colloidal spheres can be adjusted at will within an interval at least 50 nm wide. As can also be observed in the SEM images, the colloids are not decoupled: as the degree of etching increases, a bridge is formed, that interconnects each sphere to its neighbors. This bridge is most probably built by polymerization during the plasma processing. Another observation is the slight increase of roughness from the non-etched to the most etched samples.

The Et^HPPCs have also been characterized by AFM, which allowed evaluating the height of the etched colloids, and to compare these with their diameters measured by SEM. Somewhat surprising, we determined only small differences between diameter and height, with a relative error in the range of 2.0%-6.5% (relative to the nominal initial sphere diameter). As an example, the spheres in the most etched samples have an average diameter of 446 nm and an average height of 451 nm. This is a strong indication that the colloids remain very close to their spherical shape after etching.

3.2. Far Field optical properties

To ease further analyses and discussion we start with the description of a typical transmission spectrum of HPPC (Fig. 3, red line). The figure also presents, as references, the transmission spectra of a monolayer of bare polystyrene spheres (same sample without metal) - black line, and that of a bare gold film of the same thickness as that deposited on the spheres in the HPPC - blue line.

 

Fig. 3 Transmission spectra of the bare colloidal crystal (CC) (black line), hybrid plasmonic-photonic crystal (HPPC) (red line), and flat Au film (blue line).

Download Full Size | PDF

The features of the bare CC transmission spectrum include a transmission dip, located around 609 nm, and attributed to light coupling into eigenmodes of the array [44,45]. This light couples into the CC and is guided to propagate in the monolayer plane; one can also understand this as photons travelling in the (10) direction of the 2D lattice, i.e. along rows of spheres. It is also known that the spectral features of these CCs can be tuned by changing the diameter of the spheres [4], which is one of their most attractive features.

The flat Au film presents spectral features which are typical of thin gold films, i.e. a transmission peak in the 510-520 nm range, attributed to interband transitions. At longer wavelengths, outside of that spectral region, a 50 nm thin film is rather opaque. The most important feature in the HPPC spectrum is the strong transmittance band that can be observed in the 600-800nm range, contrasting the flat films' opaqueness in this spectral region. In this case, the maximum is located at 653nm and has a FWHM of ~120 nm. This transmission band is resembling very well (shape, width, intensity) that observed previously on periodic array of subwavelength holes in metallic films, in the EOT context [1,2,31,46]. In the following sections, we present our results concerning the dependence of the Et^HPPC transmission spectra on the degree of etching applied to the PS spheres.

Figure 4(a) shows the transmission spectra of the etched colloidal crystals (Et^CC) to which increasing degrees of etching were applied, together with the reference, a standard non-etched CC crystal (black line, corresponding to sample S0). Let us first note that, the more the polystyrene spheres are etched the more the transmission minima shift towards shorter wavelengths. Another important observation is the deepening of the transmission gap. The transmittance reaches values below 50% for Et^CC, compared to about 65% of the standard CC at the dip. The inset in Fig. 4(a) shows the dependence of the transmission dip position on the etching power (expressed as percent of the total power of the RIE setup), on a larger set of samples. While the minimum position can be found at 609 nm for the bare CC, this minimum can shift down to 560 nm for the most etched Et^CC in this series. In the spectral range to the red side of this dip, the transmittance is invariably high for all fabricated samples, regardless of the etching degree.

 

Fig. 4 (A) Transmission experimental spectra of the unetched (S0) and gradually etched (S3-S5)PS CC; Inset depicts the dip wavelength of each experimental sample. (B) Comparison of the experimental CC (thin, black line) and their respective best fitted simulated spectra (thick, red line). The diameters of the simulated PS spheres are 500, 490, 460, 450, 410 and respectively 390nm. (C) Transmission experimental spectra of the HPPC (S0) and Et^HPPCs (S3-S5). (D) Comparison of the experimental HPPC and Et^HPPCs (thin, black line) and the simulated spectra (thick, red line) using the PS radius that best fitted their respective CC spectra.

Download Full Size | PDF

As described above, the transmission dip is attributed to resonant light coupling into eigenmodes of the sphere array. However, one can also view this resonantly coupled light as being diffracted by the lattice at a 90 degree angle. Thus, in a simplified approach, based on Bragg diffraction and an effective refractive index of the scattering medium, the behavior of this guided-mode could be predicted, by${\lambda }_{res}=2\pi n_{ef}/G_{ij}$, where$n_{ef}=\left(V_{PS}{n}_{PS}+V_{air}{n}_{air}\right)/V_{total}$is an effective refractive index of the sphere array (containing air and PS) and$G_{ij}=\left(4\pi /\sqrt{3}P\right)\sqrt{i^2+j^2+ij}$is the reciprocal vector of the array. P represents the period (distance between the centers of the spheres). Accordingly, in CCs made of large spheres only, P is large, which implies a large λ_res, as is known from previous results [4]. Coming back to the present results, by considering that, through the etching process the centers of the PS spheres remained in the same positions, the G_ij vector also remains the same in all samples. As such, in the Et^CC samples, the decrease of n_ef (because parts of the PS beads are removed), is responsible for the decrease of λ_res, in agreement with our experimental results shown in Fig. 4(a).

In order to get a better understanding of how the etching process affects the optical properties, we conducted a series of FDTD numerical simulations in which the diameter of the etched colloids was varied. From all of our simulation results, the ones in which the transmission dip position best fitted the position of the corresponding experimental dip are presented in Fig. 4(b). Specifically, the simulated spectra were obtained with spheres having diameters 500 nm - S0, 490 nm - S1, 460 nm - S2, 450 nm - S3, 410nm - S4 and 390 nm -S5. The obtained numerical dimensions are in good agreement with the AFM and SEM results presented above. Furthermore, from the overall good overlap between the simulated and experimental transmission spectra, we can conclude that the structure's geometry has been correctly determined, even without taking into account the small bridges between the etched colloids. This has also been verified by additional simulations which compared structures with and without bridges, and showed barely visible differences. The good agreement at this step is a necessary starting point for an accurate comparison of experimental and numerical spectra in the next steps, where the geometry of the simulation setup becomes more complex, with the addition of the metal components (gold hemispherical nanocaps on top of the dielectric spheres and triangular nanoparticles on the substrate).

Moving on to the Et^HPPCs experimental spectra (Fig. 4(c)), one can observe that the position of the EOT-like transmission band shifts to the blue with the increase of the etching degree. The transmission maximum shifts from 650 nm on the reference HPPC (S0) down to 590 nm on the Et^HPPC with most etched spheres (S5). This behavior is similar to the one discussed above for the bare Et^CCs. Being known that the photonic mode in CCs (i.e. guided mode, transmission minimum) and the hybrid photonic-plasmonic mode in HPPCs (transmission band) are strongly linked, the observed behavior is to some extent expected. Note here that the main transmission dip of Et^CC is still present in the transmission spectra of the Et^HPPC, on the short-wavelength side of the transmission peak. This suggests that the same hybrid photonic-plasmonic modes are involved in the description of the transmission in both HPPC and Et^HPPC. At the same time the intensity of the transmission band of Et^HPPCs decreases with increasing the etching degree. There are several factors which could contribute to this effect. Firstly, by moving from HPPC to the more etched Et^HPPC, the contacts between the adjacent Au nanocaps are physically smaller and of different shape; this altered surface pattern can be less favorable for an efficient propagation of surface plasmons. Secondly, we have noted above that the transmission dip in the Et^CC spectra has deepened; the coupling to the photonic resonance of the dielectric sphere lattice becomes stronger, and this can inhibit the out-coupling process needed in EOT plasmon-mediated transmission. An alternative interpretation of this can be that, in the coupled photonic-plasmonic mode of the Et^HPPC, the photonic part is now stronger. The third factor which can visibly affect the EOT peak in Et^HPPCs is the fact that the surface plasmon response of the Au particles on the substrate changes: these particles become larger with more etching, and can eventually merge into a continuous patterned film. Therefore, while on the upper metallic structures (caps on top of spheres) a transition from propagative to localized plasmons can occur, on the lower metallic structures (particles on substrate) the opposite situation can take place, i.e. a transition from localized to propagative plasmon modes.

Figure 4(d) exhibits the transmission spectra obtained by numerical simulations on realistically modeled structures, compared to the experimental spectra. Despite some narrower features in the simulated spectra, a very good agreement is observed, especially with respect to the main transmission band. Both its spectral position and shape are correctly reproduced by our simulations. From the simulations, one can also note that the transmission band is actually made of two overlapping peaks, in agreement with the asymmetric shape of the experimental spectrum. The main differences between experiment and simulation occur in the long-wavelength region (above 800 nm), where the transmission is higher in the simulated spectra of the most etched Et^HPPCs (Au@S3 – Au@S5 in Fig. 4(d)). By performing an additional set of simulations, that take into account also the bridges connecting the etched spheres, we observed that the long-wavelength high transmission is absent in these structures. This result is in agreement with the SEM observations of the bridges, and confirms that indeed these are connecting the metal nanocaps on the neighboring spheres. Another difference between simulated and experimental spectra is the appearance in the simulated spectra of the highly etched Et^HPPC samples (Au@S4 and Au@S5) of a sharp dip at about 680 nm and a local maximum to the long-wavelength side. Although not as well defined, the dip at 680 nm and the broad maximum centered around 700-720 nm are also present in the experimental transmission spectrum of the most etched samples (Au@S5 in Fig. 4(d)). To tentatively explain these features, we performed some near-field numerical simulations of this structure, which we present farther below.

3.3. Near field properties

In order to get a better understanding of why and how the peculiarities of the transmission spectra vary from one corrosion step to the other we also gathered information concerning the near-field of the simulated structures. The middle row in Fig. 5(e, f, g, h) presents the electric field intensity maps on the simulated Et^HPPCs with different PS sphere diameters (350nm, 400nm, 450nm and 500nm, depicted on the top row of Fig. 5), at the wavelength corresponding to the EOT-like transmission maximum in the transmission spectra. One can observe in these images that for the unetched spheres (Fig. 5(h)) the most intense fields are localized on the upper surface of the gold nanocaps, on their sides, while they are much less intense in the other regions of the structure. A brighter area can also be observed inside the PS sphere, extending through the glass substrate. This near-field distribution is in very good agreement with what has been observed previously [4,47] on standard HPPCs, and explained as having the origin in the guided mode of the CC part of the structure.

 

Fig. 5 On the top row the cross section of each structure is schematized. The middle row depicts the electric near-field intensity maps for the simulated Et^HPPC structures. The bottom row represents a vectorial plot of the electric fields, where the magnitude of each vector is described by both its size and color (from blue to red). The diameter of the PS spheres are: 350nm (A, E and I), 400nm (B, F and J), 450nm (C, G and K)and 500nm (D, H and L).

Download Full Size | PDF

By observing the field distribution on the gradually etched geometries, it is evident that the region of high intensity fields shifts progressively towards the bottom of the simulated area. In all cases the upper regions (metal nanocaps) bear intense fields, but the intensity in the lower regions increases only with the etching. While for the reference non-etched HPPC the enhanced field is located mostly on the metal nanocaps, outside and near the rims (Fig. 5(h)), for the moderately etched Et^HPPCs samples the enhanced fields are located both on the caps' outside region and the sharp rim/edge (Fig. 5(g)), and for the most etched samples the enhanced fields are located on the nanostructures on the substrate and at the nanocap rim towards the inside of the PS sphere (Fig. 5(e)). The possibility of adjusting the region of enhanced fields in Et^HPPCs by the degree of etching is an interesting result that can prove very useful in sensing applications: one desires to control the near-field distribution in such a way that the regions of high field intensity are at the same time easily reachable by the analyte to be detected.

Further useful information regarding the near-field can be obtained by observing the vectorial representations of the electric field maps in the bottom row in Fig. 5. These representations help visualizing how the electric currents flow through the structures and where the concentrations of electrical charges are higher. One can see that the higher the etching degree, the more structured the electric field is. For example, for the unetched, standard HPPC the electric field beneath the nanocaps tends to flow in a single direction following the polarization of the incident light, while in the Et^HPPCs some well-defined current loops are established, as can well be seen in Figs. 5(i) and 5(j).

The morphology of these Et^HPPC structures is quite complex, with many changes taking place as the etching degree changes: the PS spheres and the metal caps on top of the spheres decrease in size, the spacing between neighboring caps increases, the metal structures on the substrate increase and change shape. For these reasons, in the last part of this paper, we present results concerning our analysis of the individual contributions of the metal nanostructures on top of the spheres (array of caps) and those below the spheres, on the substrate. This was possible by performing a dedicated set of simulations, in which some of the metallic parts were removed from the simulated geometry. More precisely, the whole Et^HPPC structure (scheme in Fig. 6(b)) was separated in its components: one simulation was performed on an array of metal coated spheres on the substrate (without the metal on the substrate) schematized in Fig. 6(c), and another with only the PS spheres and the metal nanostructures on the substrate (without the metal nanocaps on top of the spheres), as schematized in Fig. 6(d). The simulated transmission spectra are presented in Fig. 6(a).

 

Fig. 6 (A) Simulated transmission spectra of different compositions of the Et^HPPC: i - all the components, corresponding to the structure schematized in (B), ii - metal coated spheres, corresponding to structure in (C), iii - metal nanostructures on the substrate and uncoated spheres, corresponding to structure in (D); (E), (F) and (G) electric near-field intensity maps at 595 nm, on the structures schematized above each image.

Download Full Size | PDF

By comparing the transmission spectrum of the whole Et^HPPC structure (graph (i)) with that of the metal cap on sphere array (graph (ii), it becomes even more evident that the EOT-like transmission peak is determined by the periodic metallic structure on top of the colloidal crystal. Through these simulations, we were also able to find out that the dip at 680 nm, mentioned above in the discussion regarding Fig. 4(d), is primarily caused by the plasmonic response of the nanostructured film formed on the substrate. Another observation that arises from the simulated spectra is that the spectra of the whole Et^HPPC structure is not a linear combination of the independent contributions, but rather selective couplings take place in some spectral regions. It can also be observed in Fig. 6(a), that the minimum corresponding to the eigenmode of the PS sphere array can be shifted to different degrees, depending on the simulated structure, which can indicate that the coupling of this mode to the metal nanostructures can have different strengths in the three analyzed cases.

By visualizing the electric field maps at a wavelength of the Et^HPPC transmission maximum, 595 nm (Figs. 6(e)-6(g)), one can further observe: i) the response of the metal nanocaps is not sensitive to the metal nanostructures on the substrate, as the electric field distribution on the nanocaps is similar in Figs. 6(e) and 6(f); ii) the electric field intensity bound to the metal nanostructures on the substrate are slightly modified by the presence of the nanocaps, especially in its lower area (Fig. 6(e) vs Fig. 6(g)); this might not come as a surprise, by having in mind that the structure is illuminated from above; iii) the near-field distribution inside the PS sphere is strongly altered by the presence of the metal nanocaps on top. These observations suggest that only a very weak or no interaction occurs between the metal nanostructures on the substrates (below the spheres) and the gold caps on top of the spheres. Since the distance between the metal nanostructures is still large for an effective plasmonic coupling, the only coupling occurring between the lower and upper metallic nanostructures is mediated by the photonic modes in the dielectric spheres array. Finally, although the enhanced optical transmission in the Et^HPPC is inhibited by increasing the etching degree, the different spatial distribution of the near-fields (different localization, higher field extension, mode volume) can prove very useful for plasmonic applications, such as sensing, SERS, or SEF.

4. Conclusions

The optical properties of etched colloidal crystals (Et^CCs) and hybrid plasmonic-photonic crystals based on etched colloids (Et^HPPCs) were systematically investigated as a function of the etching degree. We evidenced a clear dependence of the spectral position and shape of the transmission spectra on the degree by which the colloidal spheres were etched. Specifically, by increasing the etching, the wavelength of spectral features (maxima/minima) shift towards shorter wavelengths. The experimental optical transmission properties were confirmed by means of FDTD simulations performed on realistic structures mimicking the actual geometrical parameters provided by means of SEM and AFM analyses. By analyzing the near-field properties of Et^HPPCs with different degrees of etching, we were able to demonstrate that i) the distribution of electric field enhancement is controlled by morphology, ii) charges accumulate in given positions on the surface of the metal nanostructures and iii) observe how their couplings evolve. On a more practical level, these results not only enable a fine tuning of the optical properties of hybrid plasmonic-photonic crystals in the spectral region best suited for specific applications, but also allows us to adjust the regions of large near field enhancement, e.g. a shaping of the near fields. In the case of the Et^CCs, the observed deepening of the transmission band gap can prove useful for photonic applications such as fabricating stop-band filters with submicron thickness. The possibility to tune not only the wavelength of the enhanced fields, but also their locations within the Et^HPPC nanostructure on a microscopic level, via etching of the colloids, expands the currently known application capabilities of hybrid plasmonic-photonic crystals, and furthers the understanding of their fundamental properties.