Natural Organic UV-Absorbent Coatings Based on Cellulose and

Article pubs.acs.org/Biomac

Natural Organic UV-Absorbent Coatings Based on Cellulose and Lignin: Designed Effects on Spectroscopic Properties Arayik Hambardzumyan,†,‡,§ Laurence Foulon,†,‡ Brigitte Chabbert,†,‡ and Véronique Aguié-Béghin*,†,‡ †

INRA, UMR614 Fractionnement des AgroRessources et Environnement, F-51100 Reims, France Université de Reims Champagne-Ardenne, UMR614 Fractionnement des AgroRessources et Environnement, F-51100 Reims, France § Physics − Chemistry Laboratory, State Erevan University, Erevan, Armenia ‡

S Supporting Information *

ABSTRACT: Novel nanocomposite coatings composed of cellulose nanocrystals (CNCs) and lignin (either synthetic or fractionated from spruce and corn stalks) were prepared without chemical modification or functionalization (via covalent attachment) of one of the two biopolymers. The spectroscopic properties of these coatings were investigated by UV−visible spectrophotometry and spectroscopic ellipsometry. When using the appropriate weight ratio of CNC/lignin (R), these nanocomposite systems exhibited high-performance optical properties, high transmittance in the visible spectrum, and high blocking in the UV spectrum. Atomic force microscopy analysis demonstrated that these coatings were smooth and homogeneous, with visible dispersed lignin nodules in a cellulosic matrix. It was also demonstrated that the introduction of nanoparticles into the medium increases the weight ratio and the CNC-specific surface area, which allows better dispersion of the lignin molecules throughout the solid film. Consequently, the larger molecular expansion of these aromatic polymers on the surface of the cellulosic nanoparticles dislocates the π−π aromatic aggregates, which increases the extinction coefficient and decreases the transmittance in the UV region. These nanocomposite coatings were optically transparent at visible wavelengths.

■

nanocomposite films that have low environmental impact.3 These coatings could be used in optical,4,5 biological,6 and medical7 sensors or in electronic applications.8 These films are based on cellulose nanocrystals (CNCs), the width (3−20 nm) and length (100 nm to 2 μm) of which depend on their biological origin, and are composed of linear chains of highmolecular-weight cellobiose, which are stabilized with numerous hydrogen bonds to obtain a highly crystalline polymer.9 In the same way, lignins offer considerable potential as adhesives,10,11 additives in biodegradable composite materials,12,13 and stabilizing agents in ceramic and aqueous alumina suspensions for use in advanced materials,14 a source of aromatic chemicals15 for polyurethane synthesis.16 Lignins are complex and highly branched phenolic polymers that contribute to the many properties of lignocellulosic cell walls, such as mechanical resistance, recalcitrance to biodegradation, and hydrophobicity. The polymer is composed of three phydroxycinnamyl alcohols as primary precursors and displays a broad range of possible chemical structures in terms of the various linkages between the aromatic units and multiple

INTRODUCTION Since 2001, the global nanotechnology market for nanostructured materials has been estimated to be worth $40 billion (http://www.goodplanet.info). The exponential increase in nanostructured materials has been reinforced by the emergence of new nanotechnology-based firms, which are mainly involved in the nanomaterials, nanodevice, nanoelectronics, and nanobiotechnology sectors.1 The nanoparticles incorporated in these materials are generally synthetic (e.g., carbon nanotubes) or mineral-based (clay, mica, or limestone). UV blockers have been developed as coatings using inorganic particles (titanium dioxides or cerium oxide) doped with iron, silica, alumina, or organic liquids.2 Some of these UV blockers display a high degree of absorption in the visible region, which explains the less transparent or opaque nature of these coatings. However, UV blockers have many advantages in terms of their resistance to discoloration under UV light and antimicrobial properties. Cellulose and lignins are two new candidate biopolymers that could offer considerable potential as nanocomposites. Cellulose and lignins are prepared from the fractionation of lignocellulosic biomass in biorefineries, which generate large quantities of residues in connection with the pulp and paper industries. Cellulose coatings have already received widespread attention on the laboratory scale regarding their use in high-performance © 2012 American Chemical Society

Received: August 31, 2012 Revised: October 22, 2012 Published: October 22, 2012 4081

dx.doi.org/10.1021/bm301373b | Biomacromolecules 2012, 13, 4081−4088

Biomacromolecules

Article

interaction patterns between the lignins and polysaccharides in the plant cell-wall network.17−20 Lignins can interact with polysaccharides through covalent (LCC) and noncovalent bonds.21−23 The two main types of noncovalent interactions are electrostatic interactions (such as hydrogen bonds formed by phenolic and alcoholic hydroxyl, carbonyl, and methoxyl groups) and weaker van der Waals interactions between cellwall polymers. Lignins absorb UV light, although their spectral properties may vary as a function of lignin composition (plant origin) and the chemical nature of its environment. Novel coatings or nanocomposite films based on industrial lignin (e.g., lignosulfates, lignin kraft, and acetylated lignin) containing cellulose fibers or commercial derivatives or noncellulosic polysaccharides have recently been developed24−27 to enhance the hydrophobicity, mechanical resistance, and oxygen-barrier properties of these materials. Recently, transparent and colored coatings have been developed based on chemically modified lignins.28 The present study aimed to design cellulose and lignin-based transparent coatings with controlled UV-absorbent properties. Cellulose/lignin nanocomposite films were first produced using lignin analogs (dehydrogenation polymers (DHPs)) and later with lignin oligomers isolated from spruce wood and corn stalks. The structural and optical properties of these coatings were investigated by spectroscopy and ellipsometry, combined with atomic force microscopy (AFM) measurements to determine the homogeneity of their topography, roughness, and thickness. One of the challenges when producing UVabsorbent coatings is to determine suitable manufacturing conditions, including the cellulose/lignin ratio that will ensure optimum optical properties (such as UV absorbance, transparency, and colorlessness) without chemical modification of the cellulose and lignin fractions.

■

Figure 1. AFM image of cellulose nanocrystals after acid hydrolysis showing the average length (A) and the chemical structure of the monomer units found in lignins (B). organosolv lignins was determined by high-performance anion exchange chromatography (HPAEC) after acid hydrolysis.35 The chemical characteristics of each sample are summarized in Table S1 of the Supporting Information and Figure 1. The lignin samples were dissolved in DMSO (99%, purchased from SDS, France) or a mixture of dioxane and water (9/1 v/v) to prepare stock solutions (0.2 to 2.0 g L−1) before they were used in the coating process. All measurements were performed at ambient temperature. Coating Process and Characterization. The cellulose-lignin films were prepared in two steps: The first step involved mixing CNC and lignin solutions to obtain cellulose/lignin mixtures with weight ratios ranging from 0.1 to 10 (Table S2 of the Supporting Information). The concentration of each polymer was adjusted to obtain CNC/lignin mixtures differing in weight ratios but retaining a constant proportion of organic solvent: dioxane/water (0.45/0.55 v/v) and DMSO/water (0.5/0.5 v/v). In the second step, 200 μL of each cellulose-lignin mixture was applied to clean quartz slides (3.14 cm2) and dried either under ambient temperature and humidity (∼50% RH) conditions for >24 h for the mixtures in dioxane/water or under a vacuum for >48 h for the same mixtures in DMSO/water. Each dried coating was stored in a desiccator until optically and spectroscopically characterized. Before deposition, the quartz was washed with piranha solution (a 7:3 mixture of concentrated sulfuric acid and hydrogen peroxide) for 0.5 h, rinsed with ultrapure water (18 MΩ) and dried under a stream of N2. UV/vis Absorption Spectrophotometry. UV/vis spectra and the transmittance between 240 and 800 nm of each dried and coated film were recorded using a Shimadzu Scientific Instrument (UV2401PC) in double-beam mode using an uncovered and cleaned quartz slide. The extinction coefficient value (ε, 10−4 dm3 g−1 μm−1) was calculated using the Beer−Lambert eq 1

EXPERIMENTAL SECTION

Preparation of Cellulose Nanocrystals. CNCs were prepared from ramie fibers (Boehmeria nivea). Small pieces of fiber were cut and treated with 2% NaOH at 20 °C for 48 h to remove any residual proteins, hemicelluloses and traces of pectin. The homogeneous suspensions obtained from washed ramie fibers were hydrolyzed overnight (∼16 h) with 65% (w/w) H2SO4 at 35 °C with stirring. The resulting suspensions were washed with water until neutral in pH and dialyzed using a 6000 molecular weight cutoff regenerated cellulose membrane. The resulting colloidal CNC suspensions were stored at 4 °C and sonicated for several minutes using a Sonics vibra-cell (750W, Fisher-Bioblock) at an appropriate concentration before use. The average crystal length was 166 ± 5 nm, and the mean section was 5.5 nm (Figure 1). Considering the cross-sectional morphology of the nanocrystal to be approximately square with a density of 1.59 g/cm3,29 the mean total specific surface area may be calculated30 to be ∼465 m2/g CNC. Lignin Preparation and Characterization. Organosolv lignin samples were isolated from spruce wood (Picea abies) and corn stalks (Zea mays L.) using a mixture of dioxane and 0.2 M hydrochloric acid (9/1 v/v) according to the procedure described by Monties et al.31 Synthetic lignins (DHPs and model lignin) were synthesized by the peroxidase/hydrogen peroxide-mediated polymerization of coniferyl alcohol (guaiacyl (G) unit), as previously described.32 DHP-G was collected as the insoluble fraction by centrifugation and then washed with water before freeze-drying. DHP and lignin samples were characterized by thioacidolysis33 to examine their relative chemical structures. Each sample was analyzed by steric exclusion chromatography after acetylation.34 The weight-average molecular weight (Mw) and the number-average molecular weight (Mn) values were determined using the relative calibration method based on the elution of polystyrene standards. Possible sugar contamination of the

ελ = Aλ /dC

(1)

where Aλ is the absorbance value at the maximum wavelength, λmax, d (μm) is the thickness of the film determined from ellipsometry measurements, and C (g dm−3) is the lignin concentration in the film. Fourier Transform Infrared spectroscopy. Fourier transform infrared (FT-IR) spectra were recorded on coated films using a Nicolet 6700 spectrophotometer using 16 scans at a resolution of 4 cm−1 from 800 to 4000 cm−1. The background (measured in air) was subtracted from the FT-IR spectra, which were normalized at 1510 cm−1. Spectroscopic Ellipsometry. Ellipsometric measurements were performed using a spectroscopic phase-modulated ellipsometer (UVISEL, Horiba-Jobin Yvon, Palaiseau, France), as previously described.36 Each coated film was analyzed by spectroscopic measurements recorded between 240 and 820 nm at the air/quartz 4082

dx.doi.org/10.1021/bm301373b | Biomacromolecules 2012, 13, 4081−4088

Biomacromolecules

Article

Figure 2. Coating films prepared from cellulose nanoparticles and lignins: (a) quartz slide, (b) cellulose nanocrystals (CNC), (c) microcrystalline cellulose, (d) synthetic lignin, (e) CNC-synthetic lignin at a weight-ratio of R = 10, (f) organosolv lignin, and (g) CNC-organosolv lignin at a weight ratio of R = 10. interface and at an incidence angle of 56°. All ellipsometric experiments were performed in an air-conditioned room at 20 ± 0.5 °C at ambient humidity (∼50% RH). The thickness and complex refractive index (real, n, and imaginary, k, parts) were calculated by fitting the ellipsometric spectra data using the Tauc−Lorentz dispersion law.37 Atomic Force Microscopy. AFM measurements were performed under ambient conditions using a Nanoscope V controller from Bruker Instruments. The AFM instrument was placed on an active vibration isolation table to prevent external vibrations from affecting the imaging process. A scanner with a maximum scan area of 120 μm2 was used and calibrated following standard procedures provided by Digital Instruments. A silicon tip with a nominal resonance frequency of ∼320 kHz, a nominal spring constant of ∼40 N m−1, and a nominal tip radius of 5 to 10 nm was used to record simultaneously both height and phase images in tapping mode. Depending on the image size, scanning rates of either 1 or 0.5 Hz were used; the resolution was 512 × 512 data points. The RMS roughness of the films was calculated from 10 μm × 10 μm AFM scans.

were visible on the quartz slide when compared with those obtained with microcrystalline or synthetic lignin coatings (Figure 2c,d), which suggests that the composite films were more homogeneous than the pure lignin coating. Selected topographic AFM images of the two series of nanocomposite films (prepared using either dioxane/water or DMSO/water as solvent) are shown in Figure 3. The RMS roughness of the

■

RESULTS AND DISCUSSION Cellulose/Synthetic Lignin - Coated Films. Structural and Morphological Properties. CNCs were used as a colloidal aqueous suspension. Using the coating assembly process to coat ramie CNC on quartz slides, a homogeneous and transparent film was prepared (Figures 2a,b). However, using microcrystalline cellulose, the same assembly process produced a heterogeneous white coating (Figure 2c). For preparing composites, the use of CNCs combining a higher aspect ratio (L/d ≈ 30) with a negative charge was an interesting alternative to the use of microcrystalline fibers. Nanosized cellulose fibers, such as CNC, which have negatively charged ester sulfate groups on the surface rod (0.4 charge/nm2), have been shown to form a strong hydrogen-bond network that does not require the use of an additional strengthening matrix.38 Interactions involving cellulose and synthetic lignin were studied first. These oligomer lignins were not contaminated by sugar monomers (Table S1 of the Supporting Information), which enabled the selective investigation of interactions between phenolic compounds and cellulose in the coating films. The synthetic lignin was solubilized in organic solvent before it was mixed with the nanocrystal and film-coating assembly. Two organic solvents were used: dioxane/water (9/1 v/v) or pure DMSO. The CNC aqueous suspension and synthetic lignin solutions were mixed to obtain the final cellulose/lignin weight ratios detailed in Table S2 of the Supporting Information. These ratios corresponded to surface concentrations of lignin ranging from 15 to 96 μg cm−2 in the coating films. Each cellulose/lignin mixture was stirred vigorously until a homogeneous dispersion was obtained, which was then deposited (200 μL) on quartz slides. The slides were completely dried under a vacuum or ambient atmosphere. The obtained films were as transparent and colorless as the pure CNC film (Figure 2e). Neither white nor yellow traces

Figure 3. Topographical images (5 μm × 5 μm) of CNC/synthetic lignin coatings using mixtures prepared with 45% dioxane (v/v) (A) or 50% DMSO (v/v) (B). The lignin contents were 32 and 64 μg cm−2, respectively. Weight ratio: R = 4.

composite films averaged 25−30 and 15−20 nm on 10 μm ×10 μm images when dioxane and DMSO were used, respectively. With both types of coating film, lignin was distributed in the cellulose matrix as nodules, as it was previously observed on cellulose model substrates.36,39 The size varied as a function of the solvent used (300 ± 65 and 89 ± 26 nm when using dioxane and DMSO, respectively). These lignin nodules were separated and well-dispersed, particularly when using DMSO as the solvent. These observations demonstrate that (1) CNC can interact with lignin in a dioxane/water or DMSO/water solvent through noncovalent interactions such as hydrogen bonding or hydrophobic interactions and (2) the presence of lignin 4083

dx.doi.org/10.1021/bm301373b | Biomacromolecules 2012, 13, 4081−4088

Biomacromolecules

Article

mixture and from 1.560 to 1.628 when the film was prepared using the DMSO/water mixture. Similarly, the imaginary part, k, at the maximum absorbance wavelength (277 nm) ranged from 0.15 to 0.30 a.u. and from 0.05 to 0.15 a.u. when the films were prepared using the dioxane/water and DMSO/water mixtures, respectively. These values suggest that films based on DMSO/water mixtures might be denser than films prepared using dioxane/water. UV/vis spectrophotometry and FTIR measurements were performed on each film to obtain data regarding their light absorption/transmission properties. Absorption/Transmission Properties of Films Within the UV/Visible Regions. Lignins exhibit characteristic UV/vis spectra depending on their origin and chemical structure and on the surrounding medium. The maximum absorbance wavelength (λmax) of synthetic lignin (DHP G) solubilized in dioxane/water or DMSO was 277 ±0.5 nm, and the extinction coefficients in these solvents were 26.7 and 24.8 L g−1 cm−1, respectively (Table S1 of the Supporting Information). When the synthetic lignin solution was mixed with the CNC suspension at a CNC/lignin weight ratio (R) from 1 to 10, the colloidal mixtures remained stable over time without any precipitation. After each mixture was dried on quartz slides, the transparent films were analyzed using UV/visible spectrophotometry. A comparison of the extinction coefficient values (ε) calculated from eq 1 for each coating showed that ε increased with the CNC/lignin weight ratio (Figure 5A,B), although the amount of lignin in the films remained constant. For example, at R = 10, the value of ε was similar to that obtained for a solution of pure lignin, that is, (26.7 and 24.8) × 10−4 dm3 g−1 μm−1 when the films were prepared using a dioxane/water mixture and DMSO/water mixture, respectively. The conditions necessary to achieve the optimal UV spectroscopic properties in the film depended on the initial lignin content and the CNC/lignin weight ratio within the film (Figure 6). The total accessible surface area of CNC increased from 0 to 1 m2 when the CNC/lignin weight ratio was increased from 0 to 10. The extinction of the film (εfilm) became similar to that in solution (εsol) when the surface area of cellulose per mg of lignin (previously solubilized in 45% of dioxane) was >2 m2/ mg, which corresponded to a theoretical surface coverage of 0.5 mg lignin/m2 of cellulose surface area (Figure 6). The same behavior was observed for lignin solubilized in 50% DMSO

macromolecules in the film does not disturb the extended CNC network generated. The thickness of these composite films was measured using spectroscopic ellipsometry and increased as the CNC/lignin weight ratio increased (Figure 4). The film thickness increased

Figure 4. Thicknesses of CNC/lignin nanocomposite films prepared from mixtures in dioxane/water (0.9/1.1 v/v) (black line) and DMSO/water (1/1 v/v) (gray line): 16 μg cm−2 (○); 32 μg cm−2 (+, ●); 48 μg cm−2 (□, ▼); 64 μg cm−2 (◆) and 96 μg cm−2 (■) of synthetic lignin.

linearly from 0 to 4 μm when the CNC content increased from 0 to 90.9% (w/w), which corresponds to an available cellulose surface area of between 0 and 1 m2 and a weight ratio of between 0 and 10. This result suggests that the assembly process is well-behaved and could be controlled using a solution containing a mixture of CNC and lignin depending on the type of organic solvent employed (dioxane or DMSO). Beyond a weight ratio of 10, the fitting of the ellipsometric spectra became more difficult (or ambiguous) because the growth of the film thickness became more uneven. The thickness values were linked to the real (n) and imaginary (k) parts of the refractive index of the composite films. The real part, n, increased from 1.396 to 1.573 at 589 nm (the sodium ray) when the film was prepared using the dioxane/water

Figure 5. Spectroscopic properties of nanocomposite films prepared from CNC/synthetic lignin mixtures in 45% dioxane (A) and 50% DMSO (B) in the UV/visible region. The lignin contents were fixed at 32 and 96 μg cm−2, respectively. Weight ratio (R). 4084

dx.doi.org/10.1021/bm301373b | Biomacromolecules 2012, 13, 4081−4088

Biomacromolecules

Article

DHP in dioxane (data not shown) when the CNC/lignin weight ratio was equal to 10. Similarly, more than 90% transmittance in the full visible spectrum (390−820 nm) was achieved when CNC was added to the coated film with a high CNC/lignin weight ratio and when the lignin was solubilized in DMSO (Figure 5B). Increasing the amount of lignin (and thus the CNC) did not markedly affect the light-transmission properties of the films. In this case, the composite film thicknesses were 1.5 to 2.0 times lower than those in the CNCdioxane lignin films despite the use of a similar amount of lignin (Figure 4). The density of these composite films was therefore higher than the density of films prepared using lignin solubilized in dioxane, as previously shown by comparing the film refractive indexes obtained using ellipsometry (from 1.560 to 1.628 in DMSO and from 1.396 to 1.573 in dioxane). Thus, the use of DMSO as a solvent to solubilize lignin in the CNC/ lignin mixture made it possible to obtain nanocomposite films with good anti-UV properties and high transmission values in the visible-light spectrum. Previous studies have indicated that thicker films (∼100 μm) made of bacterial or pulp cellulose nanofibers could display similar spectroscopic properties. However, these films needed to be impregnated with epoxy resin and sometimes polished to obtain light transmittances higher than 70% in the visible region and high absorbance in the UV region.40,41 Similarly, the reverse process, which used partially delignified cellulose nanofibers obtained from mature bamboo culms, produced UV-absorbent films with low light transmittance values in the visible spectrum (