1. Introduction

In our everyday life, coloration has a particular significance due to its psychological effects and diverse applications with various functionalities. For humans, colors serve as information carriers. They are linked to feelings, individual preferences and experiences. Therefore, colors determine strongly our behavior and emotions [1]. In nature, colors often increase the chance for survival of the species. For example, many plants and animals use camouflage through specific optical properties of the outer surface structure [2]. Furthermore, optical phenomena in organisms can also serve as a warning signal against enemies [3] or for communication in courtship [4]. The multitude of the natural color systems often originate from the principles of selective light scattering and absorption by organic pigments. The same physical principles are also used by modern industry for the production of e.g. textile apparel packaging materials, coatings/paints or cosmetics [5]. For this purpose, synthetic pigments, such as e.g. metal oxides, are often processed. However, these pigments are not biodegradable and represent an enormous problem in case of recycling [6]. In addition, various synthetic pigments are suspected of being carcinogenic [7]. For these reasons, the development of alternative engineering methods for the technical production of colors remains an important topic in research and development. In this context, an enormous potential is provided by the biomimicry of optical properties which are found in structural color systems in nature.

Natural structural colors are highly intense. They may have iridescent or non-iridescent characteristics and do not fade [8]. The coloration arises from the interaction between the visible light and nanoscaled surface structures. The color formation and the optical properties are based on individual fundamental physical processes or a superposition of these. This includes thin-film interference, multilayer interference, diffraction gratings, photonic crystals and scattering [9]. In nature, organisms possess sophisticated constructions, e.g. periodical systems from hierarchically organized micro-/nanostructures, which are responsible for those effects [10]. Therefore, the geometry of the structures, their spatial arrangement and the material properties can be particularly important for the optical color appearance. The optical properties of structural colors are often angle-dependent due to regular interference or diffraction systems. However, there are also species whose color formation is independent of the viewing angle, e.g. the famous Morpho butterfly [10–13] and the tarantula Poecilotheria metallica [14]. In the last years, those special optical properties have attracted enormous scientific attention not only for extending the knowledge of functional morphology of biological systems but also for the technical fabrication of biomimetic surfaces. In this context, the review articles by Yu et al. [15], Sun et al. [16], Burg and Parnell [17] and Niu et al. [18] give a detailed overview of already well-explored color phenomena in flora and fauna, their mimicry and potential technical application [19–21].

The Morpho butterfly wing structure was previously often used as an inspiration for the technical fabrication of structural coloration. For this purpose, previous authors have successfully used different approaches such as electron beam lithography [22–24], multilayer deposition [25,26], focused-ion-beam chemical-vapor-deposition (FIB-CVD) [27], nanostructuring by synthesizing metal oxides [28] and nanoimprint technology [29]. Nevertheless, a uniform methodology could not been established up to now for the industrial production of eco-friendly structural colors with angle-independent properties. In order to achieve this goal, further investigations are required, for example, to study the influence of definite structure geometries on the optical properties of the color formation. In this case, the aforementioned approaches are only hardly suitable. A primary reason here is the level of complexity for each fabrication process. Coating or deposition fabrication techniques are based on multiple sophisticated and expensive fabrication steps. Conventional lithographic methods, e.g. electron beam lithography, are using individual masks and a vacuum for nanostructuring.

A potential alternative in the field of additive manufacturing offers the Two-Photon Polymerization (2PP) [30]. Arbitrary 3d microstructure geometries with sub-100 nm features can be generated directly from a computer-aided design file in a photosensitive material. The smallest structures size could be achieved in a range of approximately 26 nm [31]. The high resolution and the simple manufacturing process has made 2PP suitable for a variety of applications until now [32,33] including the production of artificial structural coloration. In the this research field, for example, an experimental study of different colors was performed by 2PP-generated transmission gratings [34]. However, a replica of natural color systems is limited by the resolution of 2PP. Nevertheless, artificial structural coloration with similar optical properties to biological systems have been mimicked based on the generation of equivalent photonic systems. Thus, 2PP has been used to study the artificial manufacturing of the iridescent color of Maratus robinsoni [35] and the tarantula Poecilotheria metallica color with angle-independent optical properties [36]. In our previous publication, structural coloration with similar optical properties compared to the Morpho butterfly was fabricated with 2PP [37]. In this context, the photonic structure geometry of the Morpho butterfly was mimicked by a porous structure composition of air layers and polymer layers.

In the present paper, a simple and novel option for the fabrication of biomimetic photonic systems with blue coloration is demonstrated using interference assisted Two-Photon Polymerization. The cross-sectional morphology of a single polymerized volume resembles the nanostructures of the Morpho butterfly. Furthermore, the role of lateral structure geometry for the color appearance and its optical properties are analyzed systematically. In this way, structures with a specific degree of symmetry inside a single color area can be used to adjust the angle-independent properties and to create different optical effects which strongly dependent on the sample orientation.

2. Concept of biomimicry using interference assisted 2PP as manufacturing process

2.1 Fabrication procedure and design of biomimetic system with photonic characteristics

The visual pleasing of the Morpho butterflies have received the most attention due to its brilliant optical properties over a wide observation angle. Highly regular structures, named ridges, are responsible for the intense blue reflection of the incident light [9,10]. Those ridges are located inside individual butterfly cover scales. Furthermore, they possess a lamellar structure shape which is formed by alternating stacks of air and chitin. Consequently, the geometry of the lamellar ridges meets the requirements for a multilayer system. The number of lamellae can vary between 6 and 8 based on the height of a single ridge which can be 0.7 - 1.4 µm [18]. However, there are offsets and diverse inclines between the lamellar ridges inside a single wing scale. Those differences lead also to diffraction and scattering which are responsible for the angle-independent properties [9,10,15,17].

A biomimetic fabrication of these optical appearance requires the mimicry of the sophisticated construction of the Morpho ridges. However, the technical detailed replica of those ridges cannot fabricated with the conventional 2PP-manufacture procedure due to the low axial resolution of a single polymerized volume. Nevertheless, hierarchical microstructure systems can be generated by 2PP using a thin-film of the base material for the polymerization process. In this case, a wavy substructure feature arises inside the cross section of a single polymerized line by the interference between the focused laser and the back reflection at the boundary layer [38–41]. The size of the wavy substructures depends on the adjusted laser intensity. Thus, similar structure features compared to those in the Morpho ridges can be fabricated along the polymer structures’ cross section.

Arbitrary lateral geometries can be produced by diverse deflections of the laser beam with the use of a 2D galvo scanner. In order to study the influence of the lateral structure geometry on the color formation, a grid and an arc structure were used (see Fig. 1). In the first experiments, an edge length $l_{g} = 100$ µm was applied to the grid structure. For the generation of larger color arrays, the edge length was reduced to $l_{g} = 20$ µm to avoid an inhomogeneous polymerization due to spherical aberration. The length of a single arc structure is $l_{a} = 20$ µm and the width is $w_{a} = 8.5$ µm. While the grid structures have full photonic properties, only the horizontal bars of the arc structure generate coloration (see Fig. 1). The color hue of all samples is determined by the wavy substructures in the cross-section which form a multilayer system through polymer and air (see Fig. 1). The dimensions for the polymerized volume and the air gaps were adjusted using the maximum constructive interference condition of a multilayer stack with two different materials $\lambda _{max}=2(n_{1}d_{1}+n_{2}d_{2})$ [42] in the case of perpendicular incident light.

 

Fig. 1. Visualization of the concept for the generation of biomimetic photonic structures with 2PP. A computer-aided design of the grid structure and the arc structure are illustrated on the left and in the center. Furthermore, different views of the cross sectional morphology are shown in A-A and B-B. These section views demonstrate the theoretical structure design in axial direction and the color formation process. The axial structure design is the same for the grid and the arc structure.

Download Full Size | PDF

2.2 Experimental manufacturing setup

In the experiments a conventional photosensitive material, Femtobond 4B (Laser Zentrum Hannover e.V., Hannover, Germany), with transparent characteristics was used due to a similar refractive index (1.51 $\pm$ 0.02) in comparison to the structures of the Morpho butterfly which consists of chitin and proteins [43]. For this purpose, Femtobond 4B was diluted with 2-propanol at the ratio of 1:3. Each sample was spin-coated for 62 s with a rotation speed of 1000 rpm and an acceleration of 500 rpm s⁻¹. After that, a pre-bake was performed for 5 min at $50^\circ$C. The film thickness was 700 nm measured by a whitelight interferometer (TMS-1200, Polytec, Waldbronn, Germany). Therefore, thin polymer films were cured with UV-light of a Hg-lamp (LH-M100CB-1, Nikon, Tokio, Japan) for 24h. For the 2PP process, a mode-locked femtosecond, Ti:sapphire laser (Tsunami, Spectra Physics, Santa Clara, California, United States) with a wavelength of 780 nm was used. The laser system has a repetition rate of 82 MHz and a pulse duration of 90 fs. The total thickness of the polymer film was exposed alternating layer by layer starting from the bottom of the coverglas to ensure as uniform as possible 3d polymerization for each structure. The distance between each layer was defined with 200 nm. All samples were produced with a laser power of 12 mW and a scanning speed of 1 mm s⁻¹ of a galvo scanner (hurryScan II, Scanlab, Puchheim, Germany).

2.3 Spectral analysis and scanning electron microscopy

Spectral analysis was performed with an ellipsometer. The samples were illuminated at $45^\circ$ with a white light source (DH-2000-BAL, Ocean Optics Inc, Dunedin, Florida, USA) through an optical fibre (diameter = 200 µm) with an one lens condenser mounted on it. The reflected light intensity was detected by a monochromator (Ocean Optics Inc, Dunedin, Florida, USA) through another condenser mounted on an optical fibre (200 µm diameter). The detector position could be varied. A detail overview of the 2PP-setup and the ellipsometer setup can be found in [37]. The polymerized thin films were developed for 10 min in 2-propanol. For color monitoring a microscope (Eclipse LV 100, Nikon, Düsseldorf, Germany) was used. Here, the color formation was investigated at different sample tilt angles using an adjustable ramp to tilt the sample. The structure geometry was analyzed with a scanning electron microscope, Zeiss EVO MA 15 (Zeiss, Jena, Germany). The samples were sputtered using a AuPd-target. The sputtered layer thickness was 10 nm.

3. Results and discussion

First, the color formation was studied for the grid structure geometry. For this purpose, different grids were fabricated with varied cross-line distances ($x_{g}$, $y_{g}$) (compare Fig. 1). The applied cross-line values were the same ($x_{g}$=$y_{g}$) for both lateral dimensions. The initial cross-line distance was $x_{g}$=$y_{g}=0.6$ µm and the value was constantly increased with a step of 200 nm. The results are demonstrated in Fig. 2(A). Visual differences can be clearly identified between each grid. The grid with minimal cross-line distance ($x_{g}$=$y_{g}=0.6$ µm) is polymerized as a solid block. Therefore, the structure looks transparent. A blue coloration appears only with an increased cross-line distance value. The most intense blue reflection can be identified for a cross-line distance of $x_{g}$=$y_{g}=1.6$ µm. However, the wavelength of the reflection maximum does not change. Thus, the color formation is not affected by a diffraction grating effect.

 

Fig. 2. Biomimetic blue coloration fabricated with 2PP. The microscope image in (A) shows grid structures with different cross-line distance values for $x_{g}$ and $y_{g}$ (see Fig. 1). The cross-line distances were varied equally for both lateral sizes ($x_{g}$=$y_{g}$). An increment of the cross-line distance occurred stepwise with 0.2 µm. The edge length $l_{g}$ of each grid structure is 100 µm (A). A microscope image of the resulting color formation using the arc structure is illustrated in (E). The overall area size is 250 µm². A detail plan view of the two structure types is given in the SEM images (B) and (F). The section view of the grid structure (C) and the arc structure (G) demonstrate the lamellar construction which resembles the cross-sectional morphology of the Morpho butterfly ridges [8,10,12]. The SEM images in (D) and (H) illustrate the structural composition of the Morpho didius ridges in the plan view.

Download Full Size | PDF

While Fig. 2(B) illustrates a grid structure with a cross-line distance of $x_{g}$=$y_{g} = 1$ µm in the top view, Fig. 2(C) shows its structure in the cross section. The wavy substructures are generated by the aforementioned fabrication method [40,41] based on interference (see Sec. 2.1). The coherence length $l_{c}$ in the 2PP-setup can be calculated by the equation $l_{c}=\lambda ^2/\Delta \lambda$ [44]. Thus, the value of $l_{c}$ is approximately 30 µm for $\lambda$ = 780 nm and $\Delta \lambda$=20 nm. The reflection of the boundary layer is approximately 4% calculated by $R = |(n_{1}-n_{2})/(n_{1}+n_{2})|$ [44] with $n_{1}$=1.51 and $n_{2}$=1, respectively. These wavy structures form a multilayer system with four periods consisting of polymer and air (see Figs. 1, 2(C) and 2(G)). Assuming a total structure height of 700 nm, the axial size of a single polymer layer is approximately 125 nm, respectively, an air gap is approximately 50 nm thick. Therefore, the blue coloration can be attributed to the maximum constructive interference condition [42] with a refractive index of $n_{1}$ = 1.51 for the polymer and a refractive of $n_{2}$ = 1 for air. Those results are comparable to the color formation of the Morpho butterfly [9,10]. Differences in the color contrast (see Fig. 2(A)) can be explained by varying numbers of photonic structures inside the grids. A blue coloration could be generated as well using the arc structure shape (see Fig. 2(E)). The line distance $x_{a}$ was set to 1.6 µm (see Fig. 2(F)) since here, the most intense blue reflection could be detected for the grid structure. The color formation is based on the same multilayer system (see Fig. 2(G)) as in the case of the grid structure. As a consequence, the perceived blue reflection is independent of the structural geometry. Rather, the color arises due to the layering of artificial lamellae.

In the following course of this work, the optical properties of the biomimetic generated coloration was studied with angle-resolved spectroscopy. For this purpose, larger color arrays with an overall area of 500 µm² were fabricated using the grid and the arc structures (see Figs. 3(A) and 3(D)). The spectral measurements were performed for two illumination directions (I1 and I2) as shown in Figs. 3B and 3(E). The normalized spectral results of variable observation angles ($30^\circ$ - $70^\circ$) are illustrated in Figs. 3(C) and 3(F) for illumination direction I1. Spectra processing includes the dark noise subtraction and normalization of the spectra on the illumination source spectrum.

 

Fig. 3. Biomimetic blue coloration fabricated with 2PP and the ellipsometric measurements. Microscope images of color arrays are shown in (A) using the grid structures (A) and in (D) using the the arc structures. The overall color areas are 500 µm². The illumination angle was constantly $45^\circ$. The observation angle could be varied. Related reflection spectra are shown in (B) and (E) for an observation angle of $50^\circ$. In this process, two illumination direction (I1 and I2) were adjusted to study the reflection properties regarding the structure orientation. Normalized angle-resolved spectra are illustrated in (C) and (F) for the illumination direction I1.

Download Full Size | PDF

Studying the optical properties of the color array of the grid structures for an observation angle of $50^\circ$ in the ellipsometer setup, the reflection peaks are equally in the blue wavelength range for both illumination directions (see Fig. 3(B)). The reason for this fact is the high symmetry of the grid structure. Therefore, an identical number of photonic structures is illuminated for I1 and I2. Differences in the reflection intensity occurred because the sample could not orientated perfectly to the illumination source. On the contrary, the color arrays of the arc structure show different optical results. Here, only one orientation (I1) where light comes from left to right on the structure orientation (see Fig. 3(E)) leads to a visible color formation in the blue wavelength range. This is because only in this orientation, a sufficient number of horizontal aligned photonic structures contribute to the color formation process. Nevertheless, the optical properties for the grid and the arc structure types are similar considering the direction in which the coloration appears. Furthermore, both color formations are nearly independent of the viewing angle in a total range of $40^\circ$ as can be seen in Figs. 3(C) and 3(F). Therefore, the coloration has the similar characteristics as those which can be found in the Morpho butterflies [8,10,12], caused by the composition and arrangement of their ridge structures (see Figs. 2(D) and 2H).

On the basis of previous work [37], the color’s origin in accordance with its optical properties can be explained by a combination of Rayleigh scattering and multilayer interference. According to this, the color hue is determined due to the interaction between the incident light and the multilayer system inside the polymerized volume and can be adjusted by taking into account the maximum constructive interference condition [42]. The unique characteristic of angle-independence is engendered by a random scattering centers [45] which also could explain the non-iridescence of the coloration. We also suggest that a certain disorder inside each polymerized structure supports those specific optical properties [46]. For example, disorder can be induced by the limitation of the axial stage resolution in the 2PP-setup. This limitation could also affect the uniform axial polymerization for a certain number of structures and therefore, this issue could affect minimal shifts between the reflection peaks (compare Figs. 3(B) and 3(E)). Nevertheless, all reflection intensity peaks are clearly in the blue wavelength range. Furthermore, disorder can be induced by the minimal shrinkage of the photosensitive material which depends on the polymerization degree [47]. In this case, the shrinkage causes a local variation of the height for defined structure parts which are more strongly polymerized (compare Figs. 2(C) and 2(G)).

However, the results emphasise a non-complex generation method of biomimetic structural coloration with angle-independent properties using highly symmetric structure geometries. Furthermore, the direction of the angle-independence property can be adjusted simply through the manipulation of the reflection intensity using a specific orientation of photonic structures which are not perfectly symmetric (compare Fig. 3(E)). Additionally, this fact leads also to differences in the degree of angle-independence for the visible blue color appearance which is confirmed with the upper illustration in Fig. 4. For a specific tilt direction (tilt B), the microscope images show a blue color formation over an observation angle of $40^\circ$. On the contrary, the blue coloration disappears after an observation angle of $20^\circ$ for the contrary tilt direction (tilt A) due to a less number of photonic structures which reflect the incident light.

 

Fig. 4. A demonstrator for biomimetic application. In the upper illustration, upright microscope images visualize the reflection characteristics depending on the tilting direction (tilt A, tilt B) of color array consisting of photonic structures with an arched geometry. An adjustable ramp was used to tilt the sample in a range between $0^\circ$ and $40^\circ$. The lower images show a color array with an internal butterfly shape for three tilt positions. The arc structures are aligned with different orientations of the structure symmetry axis with respect to the incidence plane for the image contour and butterfly shape. The left microscope image shows the color formation without sample tilting. The other microscope images were taken at a tilt angle of $40^\circ$ for different tilting axis of the sample. In the middle image the arc structures in the contour of the color array are tilted around the internal structure axis “tilt B”, whereas the arc structures which form the butterfly shape are tilted around “tilt A”. The reverse case is shown in the left picture. As a result, the contour or the butterfly shape is blue colored with a tilt around the vertical or horizontal axis of the sample.

Download Full Size | PDF

Finally, the acquired knowledge was used to demonstrate the high potential of 2PP for future applications which are based on biomimetic structural coloration, for example, in the area of sensing [48,49] or anti-counterfeit protection [50,51]. Therefore, a blue color array with an overall size of 700 µm² was fabricated by 2PP with the aforementioned manufacturing method in Sec. 2.1 using the arc structure type. The arc structures were positioned specifically inside the color area. One structure orientation forms a butterfly geometry (see Fig. 4). The other structures are aligned perpendicular around the specific shape. Depending on the tilt direction of the sample, the contour or the artificial butterfly appears in blue coloration (see Fig. 4).

4. Conclusion

In this work, a simple manufacturing method was demonstrated experimentally to print reproducibly biomimetic color systems with nearly angle-independent properties using interference assisted 2PP. For this purpose, the fact was used that wavy substructure features inside the cross section of a single polymerized line arise during a polymerization process in a thin polymer film [40,41]. The reason for this can be explained by interference between the focused laser and the back reflection at the boundary layer of the photosensitive material [38,39]. Thus, the biomimetic systems consist of hierarchically micro- and nanostructure which form a multlayer system (see Figs. 2C and 2G). With a specific adjustment of the laser power, the scanning speed of the galvo scanner and a defined layer by layer process, the lamellar ridges of the Morpho butterfly could be mimicked by polymer layers and air gaps regardless of the lateral structure geometry (see Fig. 2).

The Morpho-like optical properties of the artificial color arrays were supported by color monitoring using an incident light microscope and angle-resolved measurements with an ellipsometer. The origin of the coloration was explained with scientific findings from a previous work [37]. Additionally, the role of disorder inside the artificial color arrays was discussed. Furthermore, a specific orientation of a photonic structures with less degree of symmetry leads to different color reflection intensities depending on the tilting angle of the sample (see Figs. 3 and 4). Therefore, the property of angle-independence of the coloration is tunable in a definite degree and for a specific observation direction. However, the overall angle-independence of the blue coloration is not affected by further diffraction grating effects or thin-film interference. In summary, 2PP is an exceptional research tool in the field biomimetic and offers advantages not available in coating, conventional lithographic or nanoimprint manufacturing processes due to its maskless fabrication of arbitrary structure geometries.