Dopamine-Melanin Nanofilms for Biomimetic Structural Coloration

Article pubs.acs.org/Biomac

Dopamine-Melanin Nanofilms for Biomimetic Structural Coloration Tong-Fei Wu and Jong-Dal Hong* Department of Chemistry, Research Institute of Natural Sciences, Incheon National University, 119 Academy-ro, Yeonsu-gu, Incheon 406-772, Republic of Korea S Supporting Information *

ABSTRACT: This article describes the formation of dopaminemelanin thin films (50−200 nm thick) at an air/dopamine solution interface under static conditions. Beneath these films, spherical melanin granules formed in bulk liquid phase. The thickness of dopamine-melanin films at the interface relied mainly on the concentration of dopamine solution and the reaction time. A plausible mechanism underlining dopamine-melanin thin film formation was proposed based on the hydrophobicity of dopamine-melanin aggregates and the mass transport of the aggregates to the air/ solution interface as a result of convective flow. The thickness of the interfacial films increased linearly with the dopamine concentration and the reaction time. The dopamine-melanin thin film and granules (formed in bulk liquid phase) with a double-layered structure were transferred onto a solid substrate to mimic the (keratin layer)/ (melanin granules) structure present in bird plumage, thereby preparing full dopamine-melanin thin-film reflectors. The reflected color of the thin-film reflectors depended on the film thickness, which could be adjusted according to the dopamine concentration. The reflectance of the resulted reflectors exhibited a maximal reflectance value of 8−11%, comparable to that of bird plumage (∼11%). This study provides a useful, simple, and low-cost approach to the fabrication of biomimetic thin-film reflectors using full dopamine-melanin materials.

■

reflectors are governed by film thickness and the complex refractive indices of the dielectric materials employed in the thin film and the substrate.7 The difference between the refractive index (n) and extinction coefficient (k) strongly affects the light reflectivity of a thin-film reflector.8 The reflectance (R) at the surface of a reflector is equal to r2, where r is the amplitude reflection coefficient.9 For unpolarized light at normal incidence, R at the thin film/substrate interface may be approximated as [(nfilm − nsubstrate)2 + (kfilm −ksubstrate)2]/ [(nfilm + nsubstrate)2 + (kfilm + ksubstrate)2],9,10 where nfilm and nsubstrate are the refractive indices of the two different materials, respectively, while kfilm and ksubstrate are their extinction coefficients. Thus, a larger difference between refractive indices (nfilm − nsubstrate) or a larger difference between extinction coefficients (kfilm − ksubstrate) promotes strong reflection. The refractive indices of keratin and melanin were empirically estimated to be 1.50 to 1.58 and 2.00, respectively, whereas a large extinction coefficient of k = 0.6 was estimated for melanin.5 The bright colors of feather barbs result from a large difference between the complex refractive indices of keratin and melanin, which provides strong reflection. Melanin, as a pigment, can also contribute to the bright colors of feather barbs independently of its reflection contributions. A melanin

INTRODUCTION The term melanin encompasses a class of dark biological materials found in the skin, feather, hair, eyes, and even brain living organisms. Melanin has attracted attention over past few decades due to its strong UV−visible absorption, low radiative quantum yield, and remarkable antioxidant and free radical scavenging properties.1 As an important color-producing biomaterial in animals, melanin functions as a dark, sunlightprotective pigment, but it also accounts for much of the structural colors in nature.2,3 For instance, cross-sectional transmission electron microscopy of feather barbs has revealed that the barbules generate strong reflective structural colors using a single keratin layer cortex covering a substrate composed of melanin granules.4 The air/(keratin layer)/ (melanin substrate) configuration, which produces thin-film interference, is especially important for producing blues, greens, and iridescence in bird plumage.3,5 Nature uses three main approaches for producing colors; pigments, structural colors, and bioluminescence.6 Melanin is involved in two of these approaches, namely, pigments and structural colors. This curious biomaterial attracted interest for its potential utility in optical biomimetics, especially once melanin was recognized as contributing to a broad range of structural coloration. Thin-film reflectors composed of dielectric materials with an ambient/thin film/substrate configuration can reflect specific wavelengths of light by creating conditions that promote selective interference. The optical properties of thin film © 2015 American Chemical Society

Received: December 6, 2014 Revised: January 9, 2015 Published: January 14, 2015 660

DOI: 10.1021/bm501773c Biomacromolecules 2015, 16, 660−666

Article

Biomacromolecules

temperature (RT). The dopamine-melanin films were denoted to Tris0.01M, Tris0.02 M, and Tris0.04 M according to the dopamine concentration and alkali used, respectively. Fabrication of Dopamine-Melanin Thin-Film Reflectors. Two milliliters of dopamine HCl solutions containing an equivalent amount of Tris were placed in lidded dishes and allowed to react for 24 h under static conditions at RT. After carefully cutting the connections between the film edge and dish wall, the solutions under the films were slowly sucked off using a small syringe to settle down the dopaminemelanin film onto the dish bottom without agitating the original form. The full dopamine-melanin thin-film reflectors were obtained after drying in a fume cupboard at RT for 12 h to remove the residual water. Characterization. FE-SEM was executed on a JSM-7001F SEM microscope at an acceleration voltage of 10 kV. The samples were sputter-coated with platinum (Pt) using an SPI sputter coater for enhanced conductivity before SEM was conducted. UV−visible absorption and reflection spectra were recorded on a spectrometer (PerkinElmer, Lambda 40). For reflection spectra, Labsphere RSA-PE20 reflectance accessory was used with an incidence angle of 8° from normal. The white standard was SRS-99-010 (Labsphere). Raman spectroscopy was performed on the dopamine-melanin film and granules using a Raman-LTPL spectrometer equipped with an excitation laser (λex = 533 nm). The morphologies of dopaminemelanin films were characterized by microscopy (Nikon OPTIPHOT 2-POL). The top-surface topography of the polydopamine film on mica substrate was analyzed by a Veeco AFM. The AFM image with a scan area of 250 nm × 250 nm was obtained in tapping mode with a scanning rate of 0.5 Hz. Chemical surface analysis of the dopaminemelanin film and granules was performed by XPS. XPS spectra were recorded using a PHI 5000 Versa Probe II spectrometer with standard Al Kα X-ray source (1486.6 eV). Data analysis and curve fitting were performed using CasaXPS software with a Gaussian−Lorentzian product function (Gaussian/Lorentzian ratio = 70/30) and a linear background subtraction.

substrate absorbs most of the light that reaches it,5 which weakens the reflected background colors and further enhances the color saturation of bird plumage. Although the reaction pathways that contribute to the generation of natural melanin have been intensively investigated, the mechanistic details remain elusive. 5,6-Dihydroxyindole (DHI) and 5,6-dihydroxyindole-2-carboxylic acid have been widely accepted to be monomer precursors of melanin1 and enzymatic oxidation products of tyrosine.11 Oxidations of the monomer precursors then give rise to melanin. Several other biomolecules including 3,4-dihydroxyphenethylamine (dopamine) and its catecholic derivatives (e.g., 3,4-dihydroxyphenylalanine, DOPA) also produce monomer precursors through oxidative reactions via an even shorter and simpler pathway.12,13 As a result, dopamine and DOPA have been widely used as raw materials for preparation of synthetic melanin, which can mimic the structures and properties of natural melanins.14,15 Recently, melanin made from dopamine, known as dopamine-melanin or polydopamine, has been the subject of numerous studies in a variety of fields.16−23 The sizes and morphologies of dopamine-melanin aggregates that formed in bulk solution and produced a colloid were characterized,24−26 and a dopamine-melanin film prepared on a solid substrate was examined as a potential versatile surface modification platform.27 To the best of our knowledge, the formation of dopamine-melanin film at the air/solution interface has not previously reported. This article describes the properties of dopamine-melanin nanofilms formed at an air/solution interface during the development of dopamine-melanin aggregates in a dopamine solution under static conditions. The dopamine-melanin thin film was investigated using UV−visible spectroscopy (UV−vis), atomic force microscopy (AFM), and X-ray photoelectron spectroscopy (XPS). The film thickness was investigated using field-emission scanning electron microscopy (FE-SEM) as a function of the dopamine concentration and the reaction time. A plausible mechanism underlying the dopamine-melanin thin film formation process was proposed based on the hydrophobicity of the dopamine-melanin aggregates and the mass transport of aggregates to the air/solution interface via convective flow. The dopamine-melanin thin film and granules in the bulk liquid phase formed a double-layered structure at the solution surface, and this film was transferred to a solid substrate to mimic the (keratin layer)/(melanin granules) structure found in bird plumage. These structures displayed the properties of full dopamine-melanin thin-film reflectors. The structures of the full dopamine-melanin reflectors were confirmed using FE-SEM. The color-producing properties of the resulting reflectors were investigated by measuring the nearnormal reflectivity.

■

■

RESULTS AND DISCUSSION Formation of Dopamine-Melanin Film at Air−solution Interface. Once dopamine HCl had been deprotonated by Tris, the dopamine-melanin began to form. The initial colorless solution turned to a cloudy pale yellow and then gradually changed to a dark brown. The color of dopamine solution surface gradually turned golden under static conditions over 24 h (Figure 1a), giving rise to the formation of a well-defined thin film at the air/solution interface. The film edge adhered to the dish wall, consistent with the strong adhesive forces between dopamine-melanin structures and solid surfaces reported previously.26 The resultant film was carefully detached and released from the Petri dish by submerging the Petri dish in a water reservoir. The dopamine-melanin film was hydrophobic, and the detached film floated to the surface of the water reservoir, as shown in Figure 1b. The bulk liquid phase (dark gray) in the dish was collected in a container (Figure 1c). The released dopamine-melanin film could be readily transferred onto a variety of substrates, including quartz, silicon, PET, or glass. FE-SEM images of the Tris0.01 M film (Figure 1d) transferred onto a silicon wafer suggest that the dopaminemelanin film that formed at the air−solution interface was continuous and smooth with a uniform thickness (105 nm), as shown in the inset of Figure 1d. The presence of shrinkage cracks suggested that a sol−gel process might have might have occurred during the formation of dopamine-melanin structure. Ruptures caused by granules under the film (Figure S1 in Supporting Information SI) implied that the dopamine-melanin was brittle. The size of dopamine-melanin film generated using our method was limited by the area of solution containers used in experiments. In principle, the film size could exceed tens of

EXPERIMENTAL SECTION

Materials. 3,4-Dihydroxyphenethylamine hydrochloride (dopamine HCl) and tris-(hydroxymethyl)aminomethane ((HOCH2)3CNH2, denoted by Tris) were purchased from SigmaAldrich and used as received. Preparation of Dopamine-Melanin Films at Air−Solution Interface. Two milliliters of aqueous dopamine HCl solutions with various concentrations (0.01, 0.02, and 0.04 M) containing equivalent amount of Tris were poured into plastic Petri dishes (35 mm × 10 mm), which were afterward lidded. Tris is an alkalescent compound employed to deprotonate dopamine HCl.27 Dopamine-melanin films were obtained at the air/solution interface after being allowed to react under static conditions for a desired period of time at room 661

DOI: 10.1021/bm501773c Biomacromolecules 2015, 16, 660−666

Article

Biomacromolecules

ture, the oligomers may be further linked through covalent bonding. Monomeric species may also be involved as assumed in the other hypothesis: a physical, self-assembled complex of (dopamine)2/DHI entrapped in melanin aggregates.31,32 New aggregates formed constantly and the aggregate size grew gradually through integration with monomeric species, free oligomers, and other aggregates over prolonged polymerization time. This process led to the formation of different melanin aggregate structures. The random integration of aggregates of different sizes generated melanin colloidal granules in bulk liquid phase and formed 2-D dopamine-melanin nanofilms at the solution/solid interface and the air/solution interface. The reaction pathways that led to the dopamine-melanin films and drove the aggregation process in the dopamine solution under static conditions are schematically illustrated in Figure 2 based

Figure 1. (a) Golden color appeared on the surface of dopamine solution (Tris 0.01M) in a Petri dish. (b) Dopamine-melanin film afloat on open water surface. (c) Dopamine solution remained after the removal of the dopamine-melanin film. FE-SEM images of (d) the top surface of the dopamine-melanin thin film (inset: the cross section) and (e) dopamine-melanin granules formed in bulk liquid phase. (f) AFM tapping-mode image of the top surface of the dopamine-melanin thin film together with a height profile along the line (Y = 100 nm).

Figure 2. Schematic illustrations of (a) the reaction pathway for dopamine-melanin and (b) the aggregation evolution of dopamine solution under static condition.

centimeters. Capped containers yielded more homogeneous surface colors. The dopamine-melanin separated from bulk liquid phase revealed the presence of granule-like particles (Figure 1e) with granule sizes that varied from 0.5 to 1.5 μm. These sizes were much larger than those values (100−300 nm) developed using the precious protocol.24 The size and morphological differences resulted from variations in several protocol variables including the dopamine concentration, dopamine/base molar ratio, and reaction temperature; The granule size generally changes in proportion to the dopamine concentration.24 AFM image of the dopamine-melanin film revealed that the top surface was smooth (Figure 1f), with a surface profile root-mean-square roughness of 0.59 nm for the whole area (250 nm × 250 nm). This value was comparable to the value obtained from the dopamine-melanin thin film formed on solid substrates (0.31 nm).20 Proposed Formation Mechanism of Dopamine-Melanin Films at Air−Solution Interface. The mechanism by which the monomer precursor DHI polymerizes (or oligomerizes) remains the subject of debate; however, DHI has been suggested to transition from a monomeric species to large melanin aggregates through oligomers, as suggested by AFM studies of eumelanin obtained from sepia ink.28 The oligomers (which were expected to contain two to eight DHI units29) assemble in an orderly manner through π−π stacking to form small fundamental aggregates (approximately 2−20 nm) in a supramolecular architecture.30 Within this architec-

on previously reported results.1,30 Dopamine-melanin films formed at both the air/solution interface and in bulk liquid phase. The higher oxygen concentration compared with that in solution suggested that the dopamine-melanin films might form faster at the air/solution interface than in bulk liquid phase. The favorable alignment of the aggregates at the air/solution interface was attributed to the hydrophobic nature of the aggregates and the transport.30,33,34 Dopamine-melanin aggregates formed at the air/solution interface, or they were transferred to the interface by the convective flow. At the interface, the aggregates became trapped at the air/solution interface due to their hydrophobicity and formed the dopamine-melanin nanofilms.34 Strong convective flows caused by rapid water evaporation must be avoided during the preparation of dopamine-melanin films. Strong agitation at the solution interface can hinder the formation of films with a uniform thickness. In our experiment, rapid water evaporation from the dopamine solution was prevented using capped containers. Characterization of Dopamine-Melanin Films. Figure 3a shows the UV−visible absorption spectra of dopamine solution (0.1 mM), a dopamine-melanin film, and a sample of the granules. The absorbance at 280 nm in both spectra of the dopamine-melanin film and granules was attributed to phenolic groups on the dopamine.35 The dopamine-melanin film exhibited a monotonic broad absorption band, consistent 662

DOI: 10.1021/bm501773c Biomacromolecules 2015, 16, 660−666

Article

Biomacromolecules

Figure 4. XPS survey spectra of the dopamine-melanin samples.

ment most likely arose from the heterogeneity in molecular spatial distribution and orientation in dopamine-melanins, which formed from aggregates of different size and structure. In our case, we achieved values of N/C = 0.0902 and N/O = 0.327 in dopamine-melanin granules and N/C = 0.0837and N/ O = 0.250 in the dopamine-melanin film, respectively, indicating the presence of more O and C atoms in the outermost layer of the dopamine-melanin film than granules. The XPS peak binding energy assignments are summarized in Table 1. The best fits to the O 1s spectra (Figure S2 in SI) were

Figure 3. UV−visible spectra (a) and Raman spectra (b) of the dopamine-melanin samples.

with the spectra of natural and synthetic melanins, as reported in the literature.24,36 By contrast, the dopamine-melanin granules showed a broad strong absorption over the wavelength range of 350−1100 nm. This band has not previously been reported in either the dopamine-melanin film or in other types of melanins. The origin of the new absorption band was explored by characterizing the π-conjugated structure structures of the dopamine-melanin film and granules using Raman spectroscopy, as shown in Figure 3b. The Raman spectra of the dopamine-melanin film and granules exhibited two broad bands at 1387 and 1582 cm−1, which were ascribed to the characteristic bands of dopamine-melanin.20,37 The band at 1387 cm−1 was attributed to the stretching of hexagonal carbon rings in the molecule structure, whereas the band at 1582 cm−1 resulted from the stretching of three out of the six C−C bonds within the rings.38,39 The area ratios of two bands are calculated to 0.41 and 0.44 for the dopamine-melanin film and granules, respectively, suggesting strong similarities between π-conjugated structures of the dopamine-melanin film at air/solution interface or the dopamine-melanin granules formed in bulk liquid phase. The difference observed between UV−visible absorption spectra of the dopamine-melanin film and granules may have originated from the size effects of the colloidal nanoparticle assemblies, in which interparticle coupling redshifted the absorption spectra.40 The red shift was expected to increase with the particle assembly growth and size.41 The dopamine-melanins formed structures assembled from small dopamine-melanin aggregates, in which π−π coupling occurred.1 The dopamine-melanin granules displayed a strongly red-shifted and broad absorption band due to the large difference between particle size and the interfacial film thickness (0.5 to 1.5 μm vs 105 nm, respectively). Note that the particle size in the interfacial film was estimated based on the film thickness, as shown in the inset of Figure 1d. The elemental compositions of the dopamine-melanin film and granules were analyzed by collecting the XPS spectra (Figure 4). The theoretical N/C and N/O atomic ratios of pure dopamine C8H11NO2 are 0.125 and 0.500, respectively. The value of N/C observed in the previous XPS studies on dopamine-melanins ranged from 0.08 to 0.17 in different cases, much smaller than the theoretical value.25,27,42 This disagree-

Table 1. XPS Peak Binding Energy Assignments for Granules and Films of Dopamine-Melanin C

O N

granules

film

C−C (283.6 eV) 32.5% C−N/C−O (284.8 eV) 47.4% CO (287.0 eV) 20.1% OC (530.3 eV) 22.4% O−C (532.4 eV) 77.6% R1-NH-R2 (399.2 eV)

C−C (283.9 eV) 45.9% C−N/C-O (284.9 eV) 35.3% CO (286.7 eV) 18.8% OC (530.9 eV) 26.6% O−C (532.2 eV) 73.4% R1-NH-R2 (399.4 eV)

obtained by assuming two components (OC and O−C) or three components (C−C, C−N/C−O, and CO) for the C 1s spectra (Figure S2 in SI). Only one component, that is, R1NH-R2, was found in the N 1s spectra (Figure S2 in SI), in good agreement with literature reports.42 The primary and imine nitrogen structures reported in other studies25,43 were not observed there. The spectral peak positions obtained from the dopamine-melanin film and granules remained unchanged in a given component. The thickness of dopamine-melanin film formed at air/ solution interface was correlated with the concentration of the dopamine solution and the reaction time, as shown in Figure 5. The film thickness of Tris0.01 M samples increased linearly with the reaction time within 24 h (Figure 5a). The growth rate of the film thickness was determined to be 3.2 nm/h from the linear fitting of the experimental data. The film thicknesses also increased linearly with the dopamine concentration (Figure 5b). The average thickness of dopamine-melanin films formed in Tris 0.01M, Tris 0.02 M, or Tris 0.04 M solutions within 24 h were determined to be 105, 132, and 179 nm, respectively. Note that the thicknesses of dopamine-melanin films were determined from the FE-SEM images (Figures S3 and S4 in SI). Structural Coloration of Dopamine-Melanin Thin Film Reflectors. Photographs of dopamine-melanin thin films transferred onto quartz substrate are shown in Figure 6a. Thin films of different thicknesses (105, 132, or 179 nm thick 663

DOI: 10.1021/bm501773c Biomacromolecules 2015, 16, 660−666

Article

Biomacromolecules

keratin/melanin structure in bird plumage. A dopaminemelanin thin-film reflector was fabricated on a plastic substrate in an air/thin film/substrate configuration, such that the middle clear solution was carefully removed, as schematically illustrated in Figure 7a. The dopamine-melanin granules in bulk liquid

Figure 7. (a) Schematic illustration of preparation of full dopaminemelanin thin-film reflectors. (b) FE-SEM cross section image of Tris0.01 M thin-film reflector on a plastic substrate. (c) Microscope images of reflector substrate made from dopamine-melanin granules (the control sample) and dopamine-melanin thin-film reflectors. Insets: Photographs of samples in which “+” shows the location for taking microscope images.

Figure 5. Plot for film thicknesses of dopamine-melanin film formed at air−solution interface as a function of (a) the reaction time and (b) the concentration of dopamine solutions (t = 24 h).

phase tend to sink to the dish bottom, leaving a clear solution layer between the top thin film and the bottom dopaminemelanin granules. FE-SEM cross-sectional images of the thin film reflector Tris0.01 M (Figure 7b) revealed the presence of double-layer structure comprising dopamine-melanin thin film (∼100 nm) underlying the granule (size ∼300 nm) layer, which very likely mimics the keratin/melanin structure in bird plumage. For example, a thick absorbing melanin layer underlying a keratin cortex produces violet-blue coloration, as found in the velvet satin bowerbird4 and some blackbirds and grackles.45 The dopamine-melanin thin film/granules in the double-layer structure, which were prepared from Tris 0.01M, Tris 0.02M, or Tris 0.04 M solution, were denoted to Tris0.01M, Tris0.02M, and Tris0.04 M reflectors, respectively. Microscopy images of Tris0.01M, Tris0.02M, and Tris0.04 M reflectors (Figure 7c) revealed characteristic bright colors that contrasted with the black color of the control sample comprising dopamine-melanin granules. The reflectance characteristics of different dopamine-melanin reflectors and the control sample comprising only dopaminemelanin granules were investigated using near-normal reflectance spectroscopy (incident angle = 8°), as shown in Figure 8. The bottom layer of dopamine-melanin particles in the thinfilm reflectors contributed to