Printer Friendly

Wood surfaces protected with transparent multilayer UV-cured coatings reinforced with nanosilica and nanoclay. Part II: application of a standardized test method to study the effect of relative humidity on scratch resistance.

Abstract Coated wood surfaces of components constituting flooring and furniture for interior end uses that exhibit good tribological properties are highly desirable. Surfaces of yellow birch wood (Betula alleghaniensis Britton) were protected with six different types of multilayer coatings (MCs) developed in this study. Each MC consisted of three layers: primer, sealer, and topcoat. UV-curable primer and topcoat formulations were, respectively, reinforced with a hydrophobic fumed silica (NS: 0 and 0.5 wt% in the formulation) and nanoclay (NC: 0, 1, and 3 wt% in the formulation). The scratch resistance of MCs on wood surfaces conditioned at 40% and 80% relative humidity (RH) was quantitatively and qualitatively studied. Quantitative evaluation was performed according to a standardized scratch test, while scanning electron microscopy (SEM) analysis was used for qualitative evaluation. Statistical results have shown that NS, NC, and NS x NC do not have a significant effect on scratch resistance of coated wood surfaces, whereas the effect of RH is significant. Regardless of RH, SEM images reveal that: (i) there is no sign of lack of adhesion between coating layers and the MCs/wood surfaces interface and (ii) all the MCs seem to have a ductile/brittle response to scratching. Qualitative information was in accordance with quantitative results.

Keywords Wood, UV cure, Multilayer coatings, Nanoparticles, Scratch resistance, Relative humidity

Introduction

Secondary wood products, such as wood flooring and wood furniture for interior end uses, are very useful in our daily lives. Among various materials, such as metals and polymers, used as substrate in flooring and furniture manufacturing, solid wood remains the most appreciated by most customers. The choice of solid wood is related not only to its properties, (1,2) but also to its appearance (color, grain, and texture). Wood is a hydrophilic material due to the chemical composition (cellulose, hemicelluloses, and lignin) of its cell walls. (3,4) The hydrophilic nature of wood drives its relationship with water. (5) For this reason, surfaces of wood components must be protected to improve the lifetime of secondary wood products. Wood surface protection by using organic (6,7) and inorganic (8,9) coatings is the most favored technique in the flooring and furniture industries for two reasons: economy and simplicity. Other methods such as chemical modification, (10,11) physical, (12,13) and thermal (14,15) treatments are also used to reduce the hydrophilic nature of wood surfaces.

Thanks to advances in nanotechnology, (16-18) it is possible to add and disperse different types of nanoparticles into finishes, thereby improving the physico-mechanical properties of the resulting coatings. Nanoclays present a lower hardness (1.5-2.0 Mohs) (19) than nanosilica (Quartz) (7 Mohs). (20) Silica nanoparticles, although having a low modulus (88.7 GPa) (21) compared to that of the nanoclays (170 GPa), (22) if well dispersed in the primer formulation, could enhance the mechanical properties of primer which has partly diffused in wood cell cavities at the surface. In this research, nanosilica was added in the primer formulation to improve wood surface hardness, whereas the reasons for the addition of nanoclay in the topcoat formulation are the enhancement of barrier properties such as surface hardness and scratch resistance.

With respect to the reduction of volatile organic compounds (VOCs) from coating formulations, according to the literature, (23-26) the development of photocurable formulations was and continues to be the most favored technique used in several industries. Due to lower equipment costs (27) for UV (ultraviolet light)-curing technology than for EB (electron beam)-curing technology, the former is more widely used in the flooring and furniture industries. UV-curing technology has three main advantages (28): (i) performance (high scratch and chemical resistance), (ii) greenability (VOCs and waste reduction), (29) and (iii) productivity (fast polymerization at room temperature).

Wood-coating performance studies are important to perform since they allow scientists and engineers to optimize the lifetime of the coated wood components constituting secondary wood products. Van Meel et al. (30) have studied moisture transport in coated wood. The results obtained by these authors were different as a function of applied coatings and wood type. Many studies (31,32) based on the exterior durability of wood coatings have been published, based on optical properties. In the case of interior durability of wood coatings, very few studies (33) have been published on the effect of light and climatic variations on the wood-coating performance.

The appearance of coated wood surfaces of components constituting wood flooring and wood furniture is very important. However, surfaces of these components are susceptible to various types of damages that significantly degrade their surface appearance. Fracture of the coating, delamination between coating layers, detachment of the coating from wood substrate, etc. are some examples of the types of damages in industrial wood coatings. In this research, in order to replicate some of these damages, scratch tests were used for the following reason: by using indenters, it is possible to apply a progressive or constant stress in the coating, at the coating/wood substrate interface and in the primer which has partly diffused in wood cell cavities at the surface. The scratch resistance of coatings (34-37) is considered to be a tribological property of polymer coatings. (38,39) Coated surfaces that exhibit good tribological properties are highly desirable. From a mechanical point of view, the scratch resistance of a coating can be considered as its ability to resist damage caused during the contact between the scratcher and the coating surface. According to Wong et al., (40) crazing, shear yielding, microvoiding, cracking, and debonding are the main defects observed when a polymeric surface is scratched. These defects allow the determination of the type of polymeric material responses to scratching, which depend in turn on the polymer chemistry (thermoplastics, thermosets, and elastomers). As reported by Dasari et al.,39 there are four types of polymeric material responses to scratching: (i) ductile response, (ii) ductile and brittle response, (iii) brittle response, and (iv) elastomeric response.

Lorinczova and Decker (41) have investigated scratch resistance of UV-cured acrylic coatings applied on glass substrate by using a Taber scratch tester equipped of a conical diamond indenter. The method used by these authors is based on weight change. Sow et al. (42) have evaluated scratch resistance of UV-cured acrylate waterborne coating films on glass substrate. The method used by these authors was based on gloss retention (gloss at 60[degrees] before and after scratch wear with abrasive pad). Dasari et al. (39) have reported the main scratch testing techniques for polymeric materials. As mentioned by Sangermano and Messori, (43) ASTM Standards such as: (i) ASTM D6677 (Standard Test Method for Evaluating Adhesion by Knife) and (ii) ASTM D3363 (Standard Test Method for Film Hardness by Pencil Test), are the most industrially widespread test methods for the evaluation of scratch resistance of coatings on porous and nonporous substrates. However, these standard test methods have many disadvantages, which are generally related to two factors: (i) no quantitative results and (ii) lower reproducibility of results.

A recent standardized test method (ASTM D7027: Standard Test Method for Evaluation of Scratch Resistance of Polymeric Coatings and Plastics Using an Instrumented Scratch Machine) has been used by several authors to evaluate the scratch resistance of coatings on glass (36,44) and metal (34,35) substrates. A hard indenter is pressed onto the material under load and moves relative to the material; scratch resistance is given by its ability to withstand mechanically induced surface damage under these conditions. (39) Contrary to ASTM D6677 and ASTM D3363, ASTM D7027 allows quantitative and qualitative results to be obtained: good repeatability of quantitative data and scratch panorama obtained by means of an optical microscope in the case of new scratch machines. Quantitative results such as penetration depth, residual depth, critical load, tangential force, and acoustic wave emission can be obtained as a function of normal load (progressive load scratch tests) and scratch length (constant load scratch tests). No investigation of scratch resistance of coatings and nanocomposite coatings on wood surfaces by using ASTM D7027 has been published to our knowledge. Besides scratch tests of UV-cured coatings (with and without nanoparticles) on wood surfaces, to our knowledge, no results based on the effect of relative humidity (RH) on scratch resistance of coated wood surfaces have been published to date.

The originality of this research is based on two aspects. The first one is related to the investigation of the effect of addition of nanoparticles [(i) nanosilica (NS) in the primer formulation, (ii) nanoclay (NC) in the topcoat formulation, and (iii) NS x NC] and that of RH on scratch resistance of UV-cured multilayer nanocomposite coatings on wood surfaces. On the other hand, the second aspect is based on the application of a standardized test method (ASTM D7027) to study scratch resistance of these multilayer coatings (MCs) on wood surfaces at different humidity conditions. Approaches based on nanotechnology, UV-curing technology, and MCs have been combined in our previous research (45) to protect wood surfaces and evaluate adhesion strength of MCs on wood surfaces. In the present study, scratch resistance of these MCs on wood surfaces was studied as a function of RH. The investigation was performed quantitatively by micro-scratch tests and qualitatively by scanning electron microscopy (SEM). The surface roughness and the effect RH on optical properties of these coated wood samples will be investigated in the last part of this research.

Experimental

Materials

UV-curable formulations with and without nanoparticles

Primer, sealer, and topcoat UV-curable formulations were used in this study to develop different types of multilayer coatings (see Table 5) on wood surfaces. These formulations were prepared according to experimental methods described in our previous studies. (45,46) The UV-curable formulations with and without nanosilica were used as primer, while those with and without nanoclay were applied (see Fig. 3; Table 6) on wood surfaces as topcoat and sealer, respectively. The main functions of primer, sealer, and topcoat formulations are mentioned in Fig. 1. The chemical composition of UV-curable formulations with and without nanosilica in one case and with and without nanoclay in another is shown in Tables 1 and 2, respectively. A detailed description of the chemical compounds and nanoparticles used to prepare these formulations are presented in "Chemical compounds" and "Nanoparticles" sections, respectively.

CHEMICAL COMPOUNDS: All commercial chemical compounds used in this study were components of a radical photopolymerization system. Table 3 shows the main characteristics of oligomers, monomers, photoinitiators, and a thermal initiator.

The three selected oligomers (CN 961E75, CN 104A80Z, and CN 131B) were of an acrylate type. CN 961E75 (urethane acrylate) and CN 104A80Z (epoxy acrylate) are two diacrylate oligomers with good water resistance. CN 131B, an oligomer with a lower viscosity than the first two, is fast curing and makes cured films strong and flexible. Three monomers (CD 501, SR 9003B, and SR 350, as shown in Fig. 1) with low viscosity (see Table 1) were used in this study as reactive diluents to adjust the viscosity of UV-curable formulations. CD 501 and SR 9003B are characterized by acrylate functions, while SR 350 has a methacrylate double-bond in its chemical structure. All of them offer fast cure response in free radical photopolymerization. Irgacure 184, Irgacure 819, and Darocur 1173 used in this study are all Type I photoinitiators which form free radicals by [alpha] cleavage. The first two photoinitiators are in powder form, while the last one is liquid. In order to increase the degree of acrylate bond conversion in primer formulations during the UV-curing process, a thermal initiator (benzoyl peroxide) was added to the acrylate matrix.

NANOPARTICLES: Two different types of commercial nanoparticles (Table 4), namely, Aerosil R711 (NS) and Cloisite 10A (NC), were used in this study as reinforcing agents to prepare UV-curable nanocomposite formulations. Aerosil R711 (produced by Evonik Industries) is a hydrophobic fumed nanosilica obtained after surface modification of Aerosil 200 with a methacrylsilane. Cloisite 10A (produced by Southern Clay Products) is a natural montmorillonite (MMT) modified with a quaternary ammonium salt, as shown in Table 4. Since our objective in this research was to protect wood surfaces with transparent UV-cured multilayer nanocomposite coatings, we have chosen nanoparticles having a similar refractive index (1.46 (20) for nanosilica and 1.5 (47) for nanoclay) compared to that generally reported (1.5 (48)) for resins. In this way, the color of resulting formulations containing NS is not affected and the coating remains clear. The others reasons related to the choice of NS are based on their intrinsic properties as mentioned in "Introduction" section. With respect to NC, its choice was related to our previous study, (46) in which we prepared UV-cured topcoats reinforced with three different types of commercial nanoclays (Cloisite 10A, Cloisite 15A, and Cloisite 30B). Among the findings in using those nanoclays, good results (water vapor transmission rate, and optical clarity) were obtained with Cloisite 10A.

Solid wood

Yellow birch wood (Betula alleghaniensis Britton) was used in this research as a substrate to be protected with UV-cured nanocomposite multilayer coatings. The wood boards were selected in a sawmill (Scierie Bois Poulin, Quebec). Our selection criterion was based on the color of the external bark of the wood trunk, since, from an anatomical point of view, it is not possible to differentiate wood boards from yellow birch trees from those from white birch trees. These two hardwood species are often found in the same wood batch. We chose yellow birch wood for two main reasons: first, it is one of the most preferred by consumers and second, it has good anatomical characteristics (3,49) (diffuse porosity and bordered pits without torus). These characteristics ease both penetration and diffusion of coating formulations into wood cell cavities, depending on wood surface topography, which is affected by wood surface processing. (50,51) Each selected board was made up: (i) heartwood, (ii) sapwood, or (iii) both heartwood and sapwood.

Statistical design

Treatment design

Table 5 describes the composition of the six different types of multilayer coatings (MCs) developed in this study. Each of these MCs is a combination of the levels of primer and topcoat formulations containing nanosilica (NS) and nanoclay (NC), respectively. The first three combinations (MC1, MC2, and MC3) evaluate only the effect of NC in the topcoat, while the combined effect of NS and NC was evaluated in the case of the other three combinations (MC4, MC5, and MC6). The effect of two levels (40% and 80%) RH was also investigated. Based on the above-mentioned information, we have selected a treatment design as a factorial experiment [Primer (2) x Topcoat (3) x Relative humidity (2)]. A schematic representation of the cross-section of a typical MC on a wood surface described in Table 5 is shown in Fig. 1.

Experimental design

When using porous and heterogeneous materials like solid wood as the substrate on which coating properties are assessed, the way in which wood samples are prepared (see "Wood surface preparation" section), depending on the effects to be investigated, is a key factor. In this way, unwanted effects related to wood properties (anatomy, chemistry, etc.) can be limited. Moreover, it is very important to choose an adequate experimental design in order to carry out a factorial experiment, as the type of experimental design will be taken into account for the statistical analysis of ensuing data. The experimental design selected in this study is shown in Fig. 2. As it can be seen, three wood boards were used for the development of each type of multilayer coating (MC) described in Table 5. Wood boards represent whole-plots, where the effect of primer (with and without nanosilica) and topcoat (with and without nanoclay) formulations was evaluated. As shown in Fig. 2, each whole-plot was split into two split-plots, where the effect of RH was investigated. For each type of MC, six samples were used for microscratch tests (see "Microscratch tests" section) after sample conditioning at 40% RFI and six others for microscratch tests after sample conditioning at 80% RFI. Based on the above-mentioned information, our experimental design appears like a split-plot design.

Wood surface preparation

Before the wood surface sanding process, flat-sawn wood boards were dried, conditioned, crosscut, and planed. Dried flat-sawn wood boards were conditioned for 23 months at 20 [+ or -] 0.25[degrees]C and 40 [+ or -] 3.5% RH, to reach ~8% equilibrium moisture content (EMC). This EMC is typical (50,51) for solid wood used for interior end uses, such as in wood flooring and wood furniture. Each conditioned wood board was crosscut into three sections (whole-plots, see Fig. 2) having the following sizes: 500 mm (longitudinal) x 150 mm (tangential) x 30 mm (radial). Before the wood surface planing process (orthogonal planing), wood grain was oriented as much as possible for each section, to obtain boards of 60 mm width (tangential). Planed wood surfaces were sanded by means of paper sanding belts (aluminum oxide type, Sia Abrasives, Inc.) mounted on an automatic sander (Costa Levigatrici). The sanding program used in this study was the following: (i) feed speed (parallel to wood fibers) = 8 m/min and (ii) successive stages = 80-120-180 grit size sandpapers.

Formulations application and UV-curing processes

According to the literature, (52-54) the adhesion of coatings on wood surfaces is mainly attributed to the penetration of the primer formulation into wood cell cavities, with subsequent cure of the coating, while part of this adhesion could take part at the wood-coating interface. Contrary to UV-curable waterborne primer formulations, (42) high solid UV-curable formulations (55) are generally more viscous. For this reason, before applying primer formulations, sanded and cleaned (by using a manual brush) tangential wood surfaces were heated by means of an electric IR lamp with a wavelength emission in the middle IR (MIR). Experimental method (Fig. 3) used for this purpose was the same as that described in our previous paper. (45) As shown in Fig. 3, each type of MC on wood surfaces was carried out in three steps, as described in Table 6. In step # 1, after wood surface heating, each type of primer formulation was directly applied and UV-cured in three passes. Then, UV-cured impregnated wood surfaces were manually sanded with aluminum oxide paper (400 grit size supplied by Carborundum Abrasives Products). After that, surfaces were manually cleaned by means of a brush. The manual sanding process was carried out for two reasons: first, to remove raised wood grain at the surface, caused by the penetration of primer into wood cell cavities, and second, to increase adhesion between the wood surface and [sealer + topcoat]. This process simulated the industrial process as much as possible. Finally, as shown in Fig. 3, steps # 2 and 3 were carried out to apply and cure sealer and topcoat formulations.

Preparation and conditioning of coated wood samples

As shown in our experimental design (see Fig. 2), three coated wood boards for each type of multilayer coating were used to prepare samples for microscratch tests. Sample preparation was done by crosscutting coated wood boards by means of an automatic circular saw (SI 16 SCM). The effect of RH on scratch resistance of the different types of multilayer coatings (see Table 5) was investigated in this study. For this purpose, one half of the coated wood samples for microscratch tests were conditioned at [20 [+ or -] 0.25[degrees]C-40 [+ or -] 3.5% RH] and the other at [20 [+ or -] 0.1[degrees]C-80 [+ or -] 2% RH]. In both cases, coated wood samples were conditioned until they reached equilibrium, which was verified by weighing coated wood samples each week. The six samples constituting multilayer coating 6 [MC6] for microscratch tests (see "Microscratch tests" section) were used to control reached equilibrium. Coated wood samples were conditioned for 35 and 63 days at [20[degrees]C-40%] and [20[degrees]C-80%], respectively. These conditioning times were sufficient in order to reach weight equilibrium of wood samples.

Characterization

Microscratch tests

The microscratch tests were performed according to ASTM D7027 "Standard Test Method for Evaluation of Scratch Resistance of Polymeric Coatings and Plastics using an Instrumented Scratch Machine." The scratch machine used for this purpose was a Micro Combi Tester (CSM Instruments, Peseaux, Switzerland) equipped with an optical microscope (i) allowing indenter positioning before scratch tests pass and (ii) to visualize scratches traced. For each scratch test, the sample is fastened on the sample stage, which has a horizontal (A"-axis), lateral (T-axis), and vertical (Z-axis) displacement, while the indenter is immobile.

Scratch tests in progressive and constant mode were performed in wood fiber direction with a spherical Rockwell C-type diamond indenter having a tip radius of 800 [micro]m. An indenter having a large radius was chosen in order to deform (applied normal load) several wood cells (radii, fibers, and vessels) during scratching. As mentioned by Piao et al., (56) the radius of the fibers and vessels in hardwood is comprised between 10-30 and 20-350 [micro]m, respectively. Both indenter tip and the surface of the sample were manually cleaned before performing scratch tests. The indenter tip was cleaned with a wipe (moistened with the isopropanol), while the surface of the samples was cleaned with a dry wipe. As shown in Fig. 2, five scratches were performed in progressive and constant mode for each sample. The distance between two consecutives scratches was 5 mm. The indenter tip was manually cleaned (as above described) after each scratch test pass. In this way, possible chips (turn out coating materials, wood fibers, dust, etc.) at the surface of the indenter tip produced during the previous scratch test pass were removed before performing the next scratch test pass.

The following experimental parameters were used in progressive load scratch tests: (i) applied normal load = from 0.1 to 10 N, (ii) scratch speed = 2 mm/ min, which correspond to 3.96 N/min as loading rate and (iii) scratch length = 5 mm. With respect to constant load scratch tests, the following experimental parameters were used: (i) applied normal load = 10 N, (ii) scratch speed = 2 mm/min, and (iii) scratch length = 5 mm. As shown in Fig. 2, 60 measurements (progressive and constant mode) were done for each type of coated wood sample conditioned both at 40 and 80% RH. Note that ambient conditions in the laboratory during microscratch tests were: ~20[degrees]C and ~30% RH.

Progressive and constant load scratch tests described above were done in three consecutives steps, namely, pre-scan, loaded-scan, and post-scan, (i) Pre-scan measures the initial surface profile of the sample, which is subtracted from the loaded-scan profile to determine the depth of surface penetration, (ii) loaded-scan is a scratch test pass that measures the actual scratch penetration depth while scratching, and (iii) the post-scan measures the final residual scratch depth along the scratch direction after completing the scratch test. The following parameters: (i) scanning load = 30 mN and (ii) scanning speed/scanning direction = same as in the case of scratch tests pass were used in this study for both pre-scan and post-scan.

Scanning electron microscopy (SEM) analysis of the scratched coated wood samples

When evaluating scratch resistance of coatings applied on various types of substrates (wood, metals, polymers, etc.), it is important to pay attention to qualitative evaluation of scratches traced. After performing microscratch tests on coated wood samples, the scratches were qualitatively analyzed by means of a high-resolution scanning electron microscope (SEM) (Quanta 3D FEG, FEI). For this purpose, one of the six samples (see Fig. 2) from each type of multilayer coating (MC, as shown in Table 5) tested at 40 and 80% RH was used as an SEM sample and qualitative evaluation of the scratches was carried out from the surface.

SEM samples were prepared in two steps. In step # 1, coated wood samples were dried in order to obtain anhydrous samples: the samples conditioned at 40% RH were vacuum-dried at 40[degrees]C for 7 days, while those conditioned at 80% RH were placed in a conditioning chamber having 40 [+ or -] 3.5%RH-20 [+ or -] 0.25[degrees]C for 14 days, then vacuum-dried at 40[degrees]C for 7 days. In step # 2, SEM samples were gold-palladium coated, then mounted on the sample holder for SEM analysis. SEM images were taken in two different areas of each scratch (scratch # 3: from SEM samples # 2 and 1, respectively, for coated wood samples conditioned at 40 and 80% RH, as shown in Fig. 2), at an accelerating voltage of 5 kV. These images were taken at low (24x) and high (175x) magnification. Low magnification allows the evaluation of the entire scratch length, while high magnification shows more information about scratch morphology.

Statistical analysis

Scratch resistance data obtained in this study was analyzed by means of Statistical Analysis Software (SAS 9.3). Statistical analysis consisted of computing an analysis of variance (ANOVA) of scratch parameters (slope, penetration depth, and residual depth). Probability (P value) was computed at the 0.05 confidence level. Last, multiple comparison tests were performed according to the Stepdown Bonferroni method to determine significant differences between groups.

Results and discussion

Quantitative evaluation of scratch resistance of multilayer coatings on wood surfaces

In general, scratch tests performed on coated substrates according to ASTM D7027 provide much useful information (penetration depth, residual depth, acoustic wave emission, critical load, tangential force, and coefficient of friction) (34-36) about the response of the materials to scratching. From this information, penetration depth (Fig. 5; Tables 7, 8), residual depth (Fig. 6; Tables 7, 8), and slope (Tables 7, 8) computed from the penetration depth curves were used in this study to discuss quantitative scratch resistance ("Penetration and residual depth: experimental curves analysis" and "Slope, penetration, and residual depth: statistical comparisons" sections) of the different types of multilayer coatings (MCs, as shown in Table 5) on wood surfaces. The wood coating industry is interested in the deformation of the coating and also cares about the critical load. However, coatings tested in this study do not fracture (see qualitative results in "Qualitative evaluation of scratch resistance of multilayer coatings on wood surfaces" section) probably due to the large tip size used. Consequently, critical load was not observed. The main reason why a large tip size was used to perform scratch tests was to deform several wood cells during scratching.

Penetration and residual depth: experimental curves analysis

Experimental curves showing scratch behavior of penetration depth (Pd) and residual depth (Rd) are presented in Figs. 5 and 6, respectively. Figures 5a-5c and 5a'-5c' show, respectively, the Pd curves plotted as a function of normal load and those plotted as a function of scratch length for all the tested MCs. On the other hand, Figs. 6a-6c and 6a'-6c' present, respectively, the Rd curves plotted as a function of normal load and those plotted as a function of scratch length for all the coated wood samples. As shown in Fig. 2, for coated wood samples conditioned at 40% RH, each experimental curve was obtained from sample # 2 (scratch # 3), while in the case of coated wood samples conditioned at 80% RH, each curve was obtained from sample # 1 (scratch # 3). In this way, the effect of wood's anatomical characteristics (porosity, fiber direction, etc.) is minimized and the effect of RH is maximized. Results (not presented in this paper) obtained for the other scratches (scratches # 1,2,4, and 5, as shown in Fig. 2) from a same sample regardless of RH are very similar to those presented in Figs. 5 and 6. This means that there is a good repeatability of the findings.

For all the MCs assessed in progressive load scratch tests (Figs. 5a-5c and 6a-6c), the increase in normal load (Fn: from 0.1 to 10 N) linearly increases the Pd (Figs. 5a-5c) and Rd (Figs. 6a-6c). Unfortunately, from these curves (Figs. 5 and 6), it is not possible to observe the transition between (i) the coating layers (primer, sealer, and topcoat, as shown in Fig. 1) and (ii) the MC/wood surface interface. According to the fact that coating layers and wood surface do not generally have the same physico-mechanical properties (hardness, Young's modulus, etc.), the above-mentioned transitions should be noted from Pd and Rd curves as expected. It is possible that the absence of these transitions means that performed scratch tests, as described in the "Microscratch tests" section might not be discriminant enough. The behavior of experimental curves in Figs. 5a-5c and 6a-6c are common for scratch tests in progressive mode regardless of types of materials on which tests are performed and have been observed by several authors. (34-36) At low Fn (interval: 0.1-2 N), regardless of MC type and RH, all the Pd curves in one case and Rd curves in another are very close to each other. As Fn increases up to reach its maximum value (10 N), regardless of MC type, a difference between the Pd curves for coated wood samples conditioned at 40% RH and those conditioned at 80% RH is observed. A similar trend is also noted for Rd curves. As also observed by Barletta et al., (34) the Rd curves (Figs. 6a-6c) are characterized by a decrease at the last position of the contact between the indenter and the coating surface. This decrease is related to the pile-up formation in front of the indenter and is due to the compressive residual stress, as shown in Fig. 4. (57) Regardless of MC type, the extent of this decrease is higher for coated wood samples conditioned at 80% RH than for those conditioned at 40% RH. There would be less front pile-up material after progressive load scratch tests on samples conditioned at 80% RH and less in the case of samples conditioned at 40% RH, which could be due to water plastification. Scratch resistance of coated wood samples is superior at low RH and weak at high RH, as both the Pd and Rd values at the maximum of Fn are higher for samples conditioned at 80% RH than that of those conditioned at 40% RH.

With respect to Figs. 5a'-5c' and 6a'-6c', for all the MCs evaluated in constant load scratch tests, it is possible to observe that at the early phase of scratching (interval: 0-1 mm), Pd and Rd values at constant Fn (10 N) are not the same for those obtained at the maximum value of the Fn, as shown in Figs. 5a-5c and 6a-6c. Two regions might be considered from Figs. 5 a'-5c' and 6a'-6c'. The first region (nonlinear behavior) ranging from 0 to ~1 mm shows a progressive increase in Pd and Rd and has been observed for all the tested samples. The nonlinear behavior could represent the ability of the coated wood surfaces to withstand the indenter penetration into the sample at the beginning of scratching: regardless of RH, this behavior could be related to the surface hardness of the MCs, which is slightly affected in turn by the wood surface hardness. The second region from 1 mm to the maximum value of the scratch length (5 mm) shows as observed in the case of the Pd (Figs. 5a-5c) and Rd (Figs. 6a-6c) curves in progressive mode, that scratch resistance of coated wood samples is negatively affected by RH, as Pd and Rd increase with the increase in RH. Moreover, the Pd and Rd values at constant Fn along the scratch length are irregular. This irregularity (Fig. 6c' for example) could be attributed to the porous nature of wood surfaces: wood porosity explains the low hardness of this material compared to that of nonporous materials such as polymers, metals, and ceramics.

According to Amerio et al., (58) the beginning of the scratch test can be taken as truly representative of resistance of the investigated materials toward penetration of the indenter before scratching. In this study, the slope of the penetration depth curves in progressive mode (Figs. 5a-5c) was computed in the linear region (from 100 to 150 mN). We care about this scratch parameter in order to identify at a low load the scratch resistance of the topcoat. The coefficient of determination ([R.sup.2]) from linear regression analysis of each penetration depth curve was computed using a Data Analysis and Graphing Software (OriginPro 8.6) and the obtained values were near 0.99. The average of the slope was computed from the penetration depth curves obtained in progressive mode, whereas in the case of both Pd and Rd, the average of these scratch parameters was computed from their respective experimental curves obtained in constant mode. The average of each scratch parameter was calculated on five scratches per sample for six samples per MC type. Statistical comparisons of the scratch parameters are presented and discussed in the "Slope, penetration, and residual depth: statistical comparisons" section.

Slope, penetration, and residual depth: statistical comparisons

The effect of RH on the slope, penetration depth (Pd), and residual depth (Rd) of coated wood surfaces is shown in Figs. 7a-7c, respectively. Statistical results of these scratch parameters are presented in Tables 7 (analysis of variance) and 8 (multiple comparison tests). As shown in Table 7, significant effect at a 0.05 confidence level has been observed only for humidity, as a variation source. Thus, all multiple comparison tests of the slope, Pd, and Rd were performed only for humidity.

According to multiple comparison tests, results shown in Table 8, RH has a significant effect on each scratch parameter: the increase in RH increases the slope (~30%), Pd (~38%), and Rd (~34%). These findings are explained by the plasticization effect caused by water vapor absorbed by multilayer coatings (MCs) and wood fibers near the MC/wood surface interface during conditioning time, as described in the "Preparation and conditioning of coated wood samples" section. From a theoretical point of view, as mentioned by Bolon et al.,59 during conditioning time of coated wood samples, (i) water molecules reduce intersegmental attractions of polymer chains caused by preferential bonding to water and (ii) since wood is a hydrophilic material, the wood cell walls hardness near MC/wood surface interface decreases. Consequently, scratch resistance of coated wood samples conditioned at 80% RH is lower than that of those conditioned at 40% RH. A decrease of the mechanical properties (Young modulus and tensile strength) of UV-cured topcoats (with and without nanoclay) due to the increase in RH has been observed in our previous study. (60) Similar results on UV-cured coatings have been obtained by Chawla and Poklacki (61) and Bolon et al. (59) Unfortunately, the amount of water in the coating at 40 and 80% RH was not measured in this study. However, it is important to note that it has a measurable effect on other mechanical properties. An increase of the adhesion strength of coatings on wood surfaces with the increase in RH (from 40 to 80% RH) has been observed in our previous study. (45) To our knowledge, no quantitative results based on the effect of RH on scratch resistance of UV-cured coatings on wood surfaces have been published to date.

With respect to addition of nanoparticles in the MCs (nanosilica [NS] and nanoclay [NC] in the primer and topcoat formulations, respectively, as shown in Table 5), according to multiple comparison tests results (Table 8) and regardless of RH, no enhancement of scratch resistance of the resulting nanocomposite coatings on wood surfaces has been observed. According to the literature (62-65) on nanotechnology applied to polymeric materials, the improvement of mechanical properties of pure polymer matrices by the addition of nanoparticles is related to the interfacial adhesion between the polymer chains and the dispersed phase. Thus, the absence of these interfaces or their low percentage could explain the lack of improvement of scratch resistance of MCs on wood surfaces developed in this study. The absence of an enhancement in scratch resistance upon the addition of nanoparticles might also be due to the parameters related to: (i) dispersion method, (ii) amount of loading nanoparticles, and (iii) characterization method. The optimization of these parameters could allow an enhancement in scratch resistance of UV coatings cured on wood surfaces. Amerio et al. (58) have observed a high scratch resistance of UV-cured coating containing polyhedral oligomeric silsesquioxane (POSS) compared to a low scratch resistance for pure UV-cured coating. Sangermano et al. (66) have further observed a similar trend on methacrylated UV-cured coatings reinforced with silica, titanium, and alumina nanoparticles.

Qualitative evaluation of scratch resistance of multilayer coatings on wood surfaces

Effect of relative humidity on scratch morphology: surface analysis

Qualitative information about the scratches traced on coated materials is very important since it reveals scratch morphology, which allows scientists and engineers optimizing scratch resistance of coatings on substrates. According to Mohamadpour et al., (37) among the scratch parameters mentioned in the "Quantitative evaluation of scratch resistance of multilayer coatings on wood surfaces" section, residual depth (Rd) provides information on how the coatings resist mechanical scratch. SEM images showing scratch morphology of the different types of multilayer coatings (MCs) on wood surfaces (see Table 5; Fig. 1) are shown in Figs. 8, 9, 10, and 11. Figures 8 and 9 present the Rd scratch patterns after progressive load scratch tests, whereas those after constant load scratch tests are shown in Figs. 10 and 11. Three types of MCs (MC1-MC2-MC3 and MC4-MC5-MC6) on wood samples conditioned at 40 and 80% RH are presented in each figure, for comparison. As shown in Fig. 2, for each type of MC, only the first three scratches (# 1, 2, and 3) were analyzed. Each SEM image in Figs. 8, 9, 10, and 11 was obtained from scratch # 3. SEM images from the other two scratches (# 1 and 2) are very similar to those in Figs. 8, 9, 10, and 11. This means that there is a good repeatability of the obtained results.

From Figs. 8, 9, 10, and 11, it is possible to conclude that: (i) the point where the scratch becomes visible, (ii) the scratch width, and (iii) the scratch depth for all the coated wood samples conditioned at a same RH are similar. This means that the presence of nanoparticles (nanosilica and nanoclay in the primer and topcoat, respectively) does not improve scratch resistance of MCs on wood surfaces. The scratch width and the scratch depth are higher for samples conditioned at 80% RH than for those conditioned at 40% RH. The increase in RH decreases scratch resistance of MCs on wood surfaces. Regardless of MC type and RH, there is a lack of the following defects: (i) coating layers-MC delamination, (ii) cracks inside the scratches and on the surface of coating, and (iii) cracks-crazes at the subsurface. This means that all the MCs seem to withstand scratching in the same way. As proved in Fig. 6 (decrease that represents the front pile-up), there would be a very low front pile-up formation (see Fig. 4) for all the MCs although these front pile-up are not easily identified at a low magnification (see Figs. 8 and 9). All the MCs on wood surfaces conditioned at 80% RH compress more than those conditioned at 40% RH.

Our previous results (45) based on the morphological study (SEM and TEM analysis) of the cross-sections of the coated wood samples tested in the present research as term of scratch resistance has revealed that the total thickness of the MCs (primer + sealer + topcoat) on wood surfaces ranged from 30 to 40 [micro]m. Based on these findings and according to multiple comparison tests results of Rd (see Table 8), for coated wood samples conditioned at 40% RH, scratches traced are probably in the MC thickness, as the average Rd value (27.32 [+ or -] 1.30 [micro]m) is lower than that of the MCs thickness. With respect to coated wood samples conditioned at 80% RH, scratches traced seem to reach the wood cells near MC/wood surface interface, as the average Rd value (41.70 [+ or -] 1.32 [micro]m) is slightly higher than that of the MCs thickness.

According to theory of polymeric material responses to scratching as reported by Dasari et al., (39) all the MCs on wood surfaces developed in this study seem to have a ductile/brittle response to scratching. Several authors (34,67) have used scratch damage and its features to discuss scratch resistance of coatings on metal substrates. To our knowledge, there are no published qualitative results (SEM images) based on scratch resistance of UV-cured multilayer coatings on wood surfaces. According to Dasari et al., (39) (i) Young's modulus, (ii) yield and tensile strengths, and (iii) scratch hardness are the main factors that affect scratch resistance of polymers.

Conclusion

Yellow birch wood surfaces were protected in this study with six different types of multilayer coatings (MCs), namely, MC1, MC2, MC3, MC4, MC5, and MC6. Each MC consisted of three layers: primer, sealer, and topcoat. The primer and topcoat UV-curable formulations were, respectively, reinforced with a commercial hydrophobic fumed silica (NS: 0 and 0.5 wt% in the formulation) and nanoclay (NC: 0, 1, and 3 wt% in the formulation). With respect to sealer UV-curable formulation, no nanoparticle was added. Coated wood samples were conditioned at 40 and 80% RH. The scratch resistance of these MCs on wood surfaces was quantitatively (standardized scratch tests in progressive and constant mode) and qualitatively (scanning electron microscopy analysis) studied.

The average of the slope, penetration depth (Pd), and residual depth (Rd) were used in this study to statistically compare the different types of MCs. Statistical results have shown that: (i) NS, NC, and NS x NC do not have a significant effect on scratch resistance of MCs on wood surfaces, while the effect of RH is significant and (ii) regardless of MCs, the increase in RH increases the average value of each scratch parameter (slope, Pd, and Rd). These findings mean that nanoparticles used in this study do not seem to confer additional scratch resistance of MCs on wood surfaces and that scratch resistance of samples conditioned at 40% RH is higher than for those conditioned at 80% RH. Qualitative evaluation (SEM images) of the different types of MCs has revealed important information in accordance with quantitative results. Regarding the residual scratch pattern, it is worth taking SEM images at low (evaluation of the entire scratch length) and high (evaluation of each scratch end) magnification in order to accurately evaluate scratch morphology. Regardless of RH, SEM images reveal that: (i) there is no sign of lack of adhesion between coating layers and the MCs/wood surfaces interface and (ii) all the MCs seem to have a ductile/brittle response to scratching.

Further studies should be conducted to: (i) improve (0, 0.5, 1, 2, and 3 wt% NS in the primer formulation--0 wt% in the topcoat formulation) scratch resistance of MCs on yellow birch wood surfaces and (ii) investigate the role of the wood substrate (ring [hackberry], semiring [beech], and diffuse [sugar and red maple] porous hardwood species) on scratch resistance of MCs.

DOI 10.1007/s11998-014-9609-4

Acknowledgments The authors would like to acknowledge the Conseil de recherches en sciences naturelles et en genie (CRSNG), ArboraNano and NanoQuebec for their financial support, FPInnovations (Secondary wood products manufacturing) for its collaboration with Universite Laval, Departement des sciences du bois et de la foret as well as the technicians who greatly contributed to laboratory experiments.

References

(1.) Panshin, AJ, deZeeuw, C, "Introduction." In: McGraw-Hill (ed.) Textbook of Wood Technology: Structure, Identification, Properties and Uses of the Commercial Woods of United States and Canada, pp. 1-7. McGraw-Hill, New-York, 1980

(2.) Winandy, JE, "Wood Properties." In: Arntzen, CJ (ed.) Encyclopedia of Agricultural Science, pp. 549-561. Academic Press, Orlando, 1994

(3.) Panshin, AJ, deZeeuw, C, "The Woody Cell Wall." In: McGraw-Hill (ed.) Textbook of Wood Technology: Structure, Identification, Properties and Uses of the Commercial Woods of United States and Canada, pp. 85-124. McGraw-Hill, New-York, 1980

(4.) Stevanovic, JT, Perrin, D, "Formation et structure du bois." In: Romandes, PPeU (ed.) Chimie du bois, pp. 47-67. Press Polytechniques et Universitaires Romandes, Lausanne, 2009

(5.) Rowell, RM, "Moisture Properties." In: Rowell, RM (ed.) Handbook of Wood Chemistry and Wood Composites, pp. 77-98. CRC Press, Boca Raton, 2005

(6.) Bongiovanni, R, Montefusco, F, Priola, A, Macchioni, N, Lazzeri, S, Sozzi, L, Ameduri, B, "High Performance UV-Cured Coatings for Wood Protection." Prog. Org. Coat., 45 359-363 (2002)

(7.) Rahman, MBA, Ghani, NA, Salleh, NGN, Basri, M, Rahman, R, Salleh, A, "Development of Coating Materials from Liquid Wax Esters for Wood Top-Based Coating." J. Coat. Technol. Res., 8 229-236 (2011)

(8.) Sun, Q, Yu, H, Liu, Y, Li, J, Lu, Y, Hunt, JF, "Improvement of Water Resistance and Dimensional Stability of Wood Through Titanium Dioxide Coating." Holzforschung, 64 757-761 (2010)

(9.) Ximenes, FA, Evans, PD, "Protection of Wood Using Oxy-Aluminum Compounds." For. Prod. J., 56 116-122 (2006)

(10.) Wallstrom, L, Lindberg, KAH, Johansson, J, "Wood Surface Stabilization." Holt Als Roh-und Werkst., 53 87-92 (1995)

(11.) Namyslo, JC, Kaufmann, DE, "Chemical Improvement of Surfaces. Part 1: Novel Functional Modification of Wood with Covalently Bound Organoboron Compounds." Holzforschung, 63 627-632 (2009)

(12.) Podgorski, L, Chevet, B, Onic, L, Merlin, A, "Modification of Wood Wettability by Plasma and Corona Treatments." Int. J. Adhes. Adhes., 20 103-111 (2000)

(13.) Poaty, B, Riedl, B, Blanchet, P, Blanchard, V, Stafford, L, "Improved Water Repellency of Black Spruce Wood Surfaces After Treatment in Carbon Tetrafluoride Plasmas." Wood Sci. Technol., 47 411-422 (2013)

(14.) Korkut, S, Budakci, M, "The Effects of High-Temperature Heat-Treatment on Physical Properties and Surface Roughness of Rowan (Sorbus Aucuparia L.) Wood." Wood Res., 55 67-78 (2010)

(15.) Kocaefe, D, Saha, S, "Comparison of the Protection Effectiveness of Acrylic Polyurethane Coatings Containing Bark Extracts on Three Heat-Treated North American Wood Species: Surface Degradation." Appl. Surf. Sci., 258 5283-5290 (2012)

(16.) Mathiazhagan, A, Rani, J, "Nanotechnology--A New Prospective in Organic Coating--Review." Int. J. Chem. Eng. Appl., 2 225-237 (2011)

(17.) Zhou, SX, Wu, LM, "Development of Nanotechnology-Based Organic Coatings." Compos. Interfaces, 16 281-292 (2009)

(18.) Fernando, RH, "Nanocomposite and Nanostructured Coatings: Recent Advancements." In: Fernando, RH, Sung, L-P (eds.) Nanotechnology Applications in Coatings, pp. 2-21. American Chemical Society, Washington, DC, 2009

(19.) Utracki, LA, "Basic Elements of Polymeric Nanocomposite Technology." In: Smithers Rapra Technology (ed.) Clay-Containing Polymeric Nanocomposites, p. 82. Smithers Rapra Technology, Shrewsbury, 2004

(20.) Haynes, WM, "Hardness of Minerals and Ceramics." In: Weast, R, Astle, M (eds.) Handbook of Chemistry and Physics, p. 225. CRC Press, Boca Raton, 2013

(21.) Chung, H, Kong, S, Kim, D, "Study on the Compressive Modulus of Nylon-11/Silica Nanocomposites." J. Nanomater., 11 1-7 (2012)

(22.) Yasmin, A, Luo, JJ, Abot, JL, Daniel, IM, "Mechanical and Thermal Behavior of Clay/Epoxy Nanocomposites." Compos. Sci. Technol., 66 2415-2422 (2006)

(23.) Dowbenko, R, Friedlander, C, Gruber, G, Prucnal, P, Wismer, M, "Radiation Curing of Organic Coatings." Prog. Org. Coat., 11 71-103 (1983)

(24.) Senich, GA, Florin, RE, "Radiation Curing of Coatings." J. Macromol. Sci. A, 24 239-324 (1984)

(25.) Hoyle, CE, "Photocurable Coatings." In: Hoyle, CE, Kinstle, JF (eds.) Radiation Curing of Polymeric Materials, pp. 1-16. American Chemical Society, Washington, DC, 1990

(26.) Schrof, W, Menzel, K, "Formulations." In: Schwalm, R (ed.) UV Coatings: Basics, Recent Developments and New Applications, pp. 140-159. Elsevier, Amsterdam, 2007

(27.) Clough, RL, "High-Energy Radiation and Polymers: A Review of Commercial Processes and Emerging Applications." Nucl. Instrum. Methods Phys. Res. B, 185 8-33 (2001)

(28.) Schwalm, R, "Introduction to Coatings Technology." In: Elsevier (ed.) UV Coatings: Basic, Recent Developments and New Applications, pp. 1-18. Elsevier, Amsterdam, 2007

(29.) Gustafsson, L, Borjesson, P, "Life Cycle Assessment in Green Chemistry: A Comparison of Various Industrial Wood Surface Coatings." Int. J. Life Cycle Assess., 12 151-159 (2007)

(30.) van Meel, PA, Erich, SJF, Huinink, HP, Kopinga, K, de Jong, J, Adan, OCG, "Moisture Transport in Coated Wood." Prog. Org. Coat., 72 686-694 (2011)

(31.) Ozgenc, O, Hiziroglu, S, Yildiz, UC, "Weathering Properties of Wood Species Treated with Different Coating Applications." BioResources, 7 4875-4888 (2012)

(32.) Cristea, VM, Riedl, B, Blanchet, P, "Enhancing the Performance of Exterior Waterborne Coatings for Wood by Inorganic Nanosized UV Absorbers." Prog. Org. Coat., 69 432-441 (2010)

(33.) Scrinzi, E, Rossi, S, Deflorian, F, Zanella, C, "Evaluation of Aesthetic Durability of Waterborne Polyurethane Coatings Applied on Wood for Interior Applications," Prog. Org. Coat., 72 81-87 (2011)

(34.) Barletta, M, Gisario, A, "The Role of the Substrate in Micro-scale Scratching of Epoxy-Polyester Films." Appl. Surf. Sci., 257 4449-4463 (2011)

(35.) Sung, L-P, Comer, J, Forster, A, Hu, H, Floryancic, B, Brickweg, L, Fernando, R, "Scratch Behavior of Nanoalumina/Polyurethane Coatings." J. Coat. Technol. Res., 5 419-430 (2008)

(36.) Amerio, E, Fabbri, P, Malucelli, G, Messori, M, Sangermano, M, Taurino, R, "Scratch Resistance of Nano-silica Reinforced Acrylic Coatings." Prog. Org. Coat., 62 129-133 (2008)

(37.) Mohamadpour, S, Pourabbas, B, Fabbri, P, "Anti-scratch and Adhesion Properties of Photo-curable Polymer/Clay Nanocomposite Coatings Based on Methacrylate Monomers." Sci. Iran., 18 765-771 (2011)

(38.) Kenneth, H, Allan, M, "Tribological Properties of Coatings." In: Kenneth, H, Allan, M (eds.) Coatings Tribology: Properties, Techniques and Applications in Surface Engineering, pp. 125-256. Elsevier, Amsterdam, 1994

(39.) Dasari, A, Yu, Z-Z, Mai, Y-W, "Fundamental Aspects and Recent Progress on Wear/Scratch Damage in Polymer Nanocomposites." Mater. Sci. Eng. R, 63 31-80 (2009)

(40.) Wong, M, Moyse, A, Lee, F, Sue, HJ, "Study of Surface Damage of Polypropylene Under Progressive Loading." J. Mater. Sci., 39 3293-3308 (2004)

(41.) Lorinczova, I, Decker, C, "Scratch Resistance of UV-Cured Acrylic Clearcoats." Surf. Coat. Int. B, 89 133-143 (2006)

(42.) Sow, C, Riedl, B, Blanchet, P, "UV-Waterborne Polyurethane-Acrylate Nanocomposite Coatings Containing Alumina and Silica Nanoparticles for Wood: Mechanical, Optical, and Thermal Properties Assessment." J. Coat. Technol. Res., 8 211-221 (2011)

(43.) Sangermano, M, Messori, M, "Scratch Resistance Enhancement of Polymer Coatings." Macromol. Mater. Eng., 295 603-612 (2010)

(44.) Sangermano, M, Messori, M, Martin-Gallego, M, Rizza, G, Voit, B, "Scratch Resistant Tough Nanocomposite Epoxy Coatings Based on Hyperbranched Polyesters." Polymer, 50 5647-5652 (2009)

(45.) Nkeuwa, WN, Riedl, B, Landry, V, "Wood Surfaces Protected with Transparent Multilayer UV-Cured Coatings Reinforced with Nanosilica and Nanoclay. Part I: Morphological Study and Effect of Relative Humidity on Adhesion Strength." J. Coat. Technol. Res., 11 283-301 (2014)

(46.) Nkeuwa, WN, Riedl, B, Landry, V, "UV-Cured Clay/Based Nanocomposite Topcoats for Wood Furniture. Part I: Morphological Study, Water Vapor Transmission Rate and Optical Clarity." Prog. Org. Coat., 77 1-11 (2014)

(47.) Wang, SH, Zhang, Y. Ren, WT, Zhang, YX, Lin, HF, "Morphology, Mechanical and Optical Properties of Transparent BR/Clay Nanocomposites." Polym. Test., 24 766-774 (2005)

(48.) Wicks, ZW, Jones, FN, Pappas, SP. Wicks, DA, "Color and Appearance." In: Wiley Inc. (ed.) Organic Coatings: Science and Technology, pp. 382-416. Wiley Inc., Hoboken, 2007

(49.) Wiedenhoeft, A, Miller, R, "Structure and Function of Wood." In: Rowell, RM (ed.) Handbook of Wood Chemistry and Wood Composites, pp. 9-33. CRC Press, Boca Raton, 2005

(50.) de Moura, LF, Hernandez, RE, "Evaluation of Varnish Coating Performance for Two Surfacing Methods on Sugar Maple Wood." Wood Fiber Sci., 37 355-366 (2005)

(51.) Hernandez, RE, Cool, J, "Evaluation of Three Surfacing Methods on Paper Birch Wood in Relation to Water- and solvent-Borne Coating Performance." Wood Fiber Sci., 40 459-469 (2008)

(52.) de Meijer, M, Militz, H, "Wet Adhesion of Low-VOC Coatings on wood: A Quantitative Analysis." Prog. Org. Coat., 38 223-240 (2000)

(53.) Meijer, M, Thurich, K, Militz, H, "Comparative Study on Penetration Characteristics of Modern Wood Coatings." Wood Sci. Technol., 32 347-365 (1998)

(54.) Rijckaert, V, Stevens, M, van Acker, J, "Effect of Some Formulation Parameters on the Penetration and Adhesion of Water-Borne Primers into Wood." Holz Als Roh-und Werkst., 59 344-350 (2001)

(55.) Landry, V, Riedl, B, Blanchet, P, "Nanoclay Dispersion Effects on UV Coatings Curing." Prog. Org. Coat., 62 400-408 (2008)

(56.) Piao, C, Winandy, JE, Shupe, TF, "From Hydrophilicity to Hydrophobicity: A Critical Review: Part I. Wettability and Surface Behavior." Wood Fiber Sci., 42 490-510 (2010)

(57.) Barletta, M, Gisario, A, Trovalusci, F, Vesco, S, "Visual Appearance and Scratch Resistance of High Performance Thermoset and Thermoplastic Powder Coatings." Prog. Org. Coat., 76 244-256 (2013)

(58.) Amerio, E, Sangermano, M, Colucci, G, Malucelli, G, Messori, M, Taurino, R, Fabbri, P, "UV Curing of Organic-Inorganic Hybrid Coatings Containing Polyhedral Oligomeric Silsesquioxane Blocks." Macromol. Mater. Eng., 293 700-707 (2008)

(59.) Bolon, DA, Lucas, GM, Olson, DR, Webb, KK, "Effect of Humidity and Elevated Temperatures on Physical Properties of UV-Cured Polymers." J. Appl. Polym. Sci., 25 543-553 (1980)

(60.) Nkeuwa, WN, Riedl, B, Landry, V, "UV-Cured Clay/Based Nanocomposite Topcoats for Wood Furniture. Part II: Dynamic Viscoelastic Behavior and Effect of Relative Humidity on the Mechanical Properties." Prog. Org. Coat., 77 12-23 (2014)

(61.) Chawla, CP, Poklacki, ES, "Effect of Relative Humidity on the Mechanical Properties of UV-Curable Optical Fiber Coatings." In: Greenwell, RA, Paul, DK (eds.) Optical Materials Reliability and Testing, pp. 61-66. SPIE, Boston, 1993

(62.) Paul, DR, Robeson, LM, "Polymer Nanotechnology: Nanocomposites." Polymer, 49 3187-3204 (2008)

(63.) Bharadwaj, RK, Mehrabi, AR, Hamilton, C, Trujillo, C, Murga, M, Fan, R, Chavira, A, Thompson, AK, "Structure-Property Relationships in Cross-Linked Polyester-Clay Nanocomposites." Polymer, 43 3699-3705 (2002)

(64.) Goertzen, WK, Kessler, MR, "Dynamic Mechanical Analysis of Fumed Silica/Cyanate Ester Nanocomposites." Composites A, 39 761-768 (2008)

(65.) Corcione, CE, Frigione, M, "UV-Cured Polymer-Boehmite Nanocomposite as Protective Coating for Wood Elements." Prog. Org. Coat., 74 781-787 (2012)

(66.) Sangermano, M, Gaspari, E, Vescovo, L, Messori, M, "Enhancement of Scratch-Resistance Properties of Methacrylated UV-Cured Coatings." Prog. Org. Coat., 72 287-291 (2011)

(67.) Barletta, M, Gisario, A, Rubino, G, "Scratch Response of High-Performance Thermoset and Thermoplastic Powders Deposited by the Electrostatic Spray and 'Hot Dipping' Fluidised Bed Coating Methods: The Role of the Contact Condition." Surf. Coat. Technol., 205 5186-5198 (2011)

W. N. Nkeuwa, B. Riedl ([mail]), V. Landry

Departement des sciences du bois et de la foret, Centre de recherche sur les materiaux renouvelables, Universite Laval, Quebec City, QC G1V 0A6, Canada

e-mail: bernard.riedl@sbf.ulaval.ca

V. Landry

Secondary wood Products Manufacturing, FPInnovations, Quebec City, QC GIP 4R4, Canada

Table 1: Chemical composition of UV-curable primer
formulations with and without nanosilica

Chemical compound   Commercial name   Amount (wt%)

Oligomer              CN 961E75        5
Monomers              CD 501           47.5
                      SR 9003B         47.5
Defoaming agent       BYK 1798         0.4
Thermal initiator     BPO              1
Photoinitiator        IRG 184-819      4
Nanosilica            R711             0 and 0.5

Table 2: Chemical composition of UV-curable formulations
with and without nanoclay. Sealer (matrix + 0% of
nanoclay) and topcoat (matrix + 0.1 and 3% of nanoclay)

Chemical compound   Commercial name   Amount (wt%)

Oligomers           CN 104A80Z           32.5
                    CN 131B              32.5
Monomers            SR 350               17.5
                    SR 9003B             17.5
Defoaming agent     BYK 1798             0.4
Photoinitiator      Darocur 1173         4
Nanoclay            Cloisite 10A      0, 1, and 3

Table 3: Main characteristics of oligomers, monomers,
photoinitiators, and thermal initiator used in this study to prepare
pure UV-curable acrylate matrices

Chemical           Commercial                Chemical name
compound              name

Oligomers         CN 961E75       Urethane acrylate blended with SR
                                  454
                  CN 104A80Z      Epoxy acrylate blended with SR 306
                  CN 131B         Monoacrylate oligomer
Monomers          CD 501          Propoxylated (6) trimethylolpropane
                                    triacrylate
                  SR 350          Trimethylolpropane trimethacrylate
                  SR 9003B        Propoxylated neopentyl glycol
                                    diacrylate
Photoinitiators   Irgarcure 184   1-Hydroxy-cyclohexyl-phenyl-ketone
                  Irgacure 819    Bis(2,4,6-trimethylbenzoyl)-
                                    phenylphosphineoxide
                  Darocur 1173    2-Hydroxy-2-methyl-1-phenhyl-1-
                                    propanone
Thermal           BPO             Benzoyl peroxide
  initiator

Chemical           Commercial        Viscosity (cPs)       Refractive
compound              name                                   index

Oligomers         CN 961E75       4000 (at 60[degrees]C)      1.48
                  CN 104A80Z      500 (at 60[degrees]C)       1.53
                  CN 131B         250 (at 25[degrees]C)       1.52
Monomers          CD 501          125 (at 25[degrees]C)       1.45
                  SR 350          44 (at 25[degrees]C)        1.47
                  SR 9003B        15 (at 25[degrees]C)        1.44
Photoinitiators   Irgarcure 184
                  Irgacure 819
                  Darocur 1173
Thermal           BPO
  initiator

SR 454 is an ethoxylated trimethylol propane triacrylate added at 25%
into CN 961E75. SR 306 is a tripropylene glycol diacrylate added at
20% into CN 104A80Z

Table 5: Different types of multilayer coatings by
combination of UV-curable formulations with and without
nanoparticles (nanosilica and nanoclay)

Primer          Sealer      Topcoat     Multilayer coating
                                               (MC)

Nanosilica     Without      Nanoclay    Composition    Code
(NS)         nanoparticle     (NC)

0                              0       0% NS-0% NC     MC1
                               1       0% NS-1% NC     MC2
                               3       0% NS-3% NC     MC3
0.5                            0       0.5% NS-0% NC   MC4
                               1       0.5% NS-1% NC   MC5
                               3       0.5% NS-3% NC   MC6

Table 6: Experimental parameters used for formulations application
and UV-curing processes, as shown in Fig. 3

Step     Formulation            Application process

                          Belt       Dosing     Applicator
                          speed    roll speed   roll speed
                         (m/min)    (m/min)      (m/in.)

1      Primer with and     12          27           12
         without
         nanosilica
2      Sealer 1 (S1)       12          15           12
       Sealer 2 (S2)       12          15           12
3a     Topcoat 1 (T1)      12          15           12
       Topcoat 2 (T2)      12          15           12

Step     Formulation                   UV-curing process

                         Conveyor   UV-lamp   Energy density    Number
                          speed      power        of UV-A         of
                         (m/min)              (mJ/[cm.sup.2])   passes

1      Primer with and      4         100           529           3
         without
         nanosilica
2      Sealer 1 (S1)        4         50            331           1
       Sealer 2 (S2)        4         100           789           1
3a     Topcoat 1 (T1)       4         100           789           2
       Topcoat 2 (T2)

UV-curing process was done in air atmosphere at room temperature with
a medium vapor pressure Hg UV-lamp equipped with a conveyor

(a) Application and UV-curing process for topcoat (T1 and T2, as
shown in Figs. 1 and 3) with and without nanoclay was the same

Table 7: Analysis of variance (ANOVA) of the scratch parameters:
slope, penetration depth, and residual depth

Variation      Degree of                Slope
source          freedom
                           F value   P value   Significant
                                                 effect?

Primer (P)         1         0.92     0.3552       No
Topcoat (T)        2         0.45     0.6467
P x T              2         2.02     0.1718
Humidity (H)       1       100.32    <0.0001       Yes
P x H              1         0.57     0.4653       No
T x H              2         0.20     0.8199
P x T x H          2         1.21     0.3309

Variation             Penetration depth
source
               F value   P value   Significant
                                     effect?

Primer (P)       0.00     0.9501       No
Topcoat (T)      0.89     0.4352
P x T            1.02     0.3888
Humidity (H)   498.43    <0.0001       Yes
P x H            0.57     0.4630       No
T x H            0.39     0.6830
P x T x H        2.85     0.0969

Variation              Residual depth
source
               F value   P value   Significant
                                     effect?

Primer (P)       0.19     0.6698       No
Topcoat (T)      1.06     0.3765
P x T            1.61     0.2402
Humidity (H)   114.01    <0.0001       Yes
P x H            0.76     0.3990       No
T x H            0.07     0.9337
P x T x H        0.83     0.4595

Probability at 0.05 confidence level. Variation sources with P value
lower than that of the confidence level has a significant effect on
the slope, penetration depth, and residual depth

Table 8: Multiple comparison tests results of slope,
penetration depth, and residual depth for humidity:
effect relative humidity (RH)

Scratch              Relative   Average   Standard   Letter
parameter            humidity              error     group
                     (RH, %)

Slope                   40        8.27      0.25       B
                        80       11.88      0.30       A
Penetration             40       77.66      3.32       B
  depth ([micro]m)      80      125.84      3.37       A
Residual                40       27.32      1.30       B
  depth ([micro]m)      80       41.70      1.32       A

Average of each scratch parameter was computed for all the
multilayer coatings conditioned at 40% RH and those conditioned
at 80% RH since the effect of others factors was not
significant (see Table 7). For each scratch parameter,
average values with the same letter in each group are not
significantly different at 0.05 confidence level


----------

Please note: Some tables or figures were omitted from this article.
COPYRIGHT 2014 American Coatings Association, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2014 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Nkeuwa, William Nguegang; Riedl, Bernard; Landry, Veronic
Publication:Journal of Coatings Technology and Research
Geographic Code:1USA
Date:Nov 1, 2014
Words:9868
Previous Article:Application of CuCoMn[O.sub.x] coat by sol gel technique on aluminum and copper substrates for solar absorber application.
Next Article:The effect of Ti[O.sub.2], pigmentation on the hydrolysis of amino resin crosslinked epoxy can coatings.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters