Light-Induced Colour Changes in Wood Surfaces in Indoor Conditions Determined by an Artificial Accelerated Test: Influence of Wood Species and Coating Materials

Abstract

Stabilising the natural colour of wood species exposed to light in indoor conditions is a challenge that could be better addressed based on a deeper understanding of the occurring phenomena and influencing factors. This paper investigates comparatively the light-induced colour changes for three hardwood species, namely, European maple (Acer pseudoplatanus L.), European ash (Fraxinus excelsior L.) and European walnut (Juglans regia L.), as well as the influence of finishing with three types of clear, colourless waterborne lacquers: acrylic-polyurethane (F1), acrylic (F2) and polyurethane (F3) on their colour stability. Colour measurements in the CIELab system in conjunction with an artificial accelerated ageing test under the action of UV-VIS radiation, simulating natural light passing through window glass, and two types of test samples were employed to highlight the influence and contributions of the wood substrate and of the coating films to the global colour modifications. Coating films applied on 1 mm clear glass slides were employed as a sort of “detachable” finish for this purpose. Direct exposure to UV-VIS light caused visible colour changes for both uncoated and coated wood surfaces, the values of the calculated colour differences (ΔE) after a 72 h exposure being dependant on both the wood species and the coating material. Excepting two situations for walnut, statistically significantly higher colour differences were obtained for the uncoated samples: maple (9.36 units), ash (8.39 units), walnut (6.20 units), compared to the coated ones: maple (4.92–5.71 units), ash (2.25–3.94 units), walnut (4.74–7.70 units). The wood substrates underlying the coating films were found to bring the maximum contribution to the overall colour changes in the clear coated surfaces, while the coating films employed in this research demonstrated a fairly good colour stability to UV-VIS light exposure, with maximum colour changes (∆E) up to only 1.30 units. Overall, the wood species and the type of coating were found as influencing factors in interaction with the light-induced colour changes in wood surfaces in indoor conditions.

1. Introduction

Colour, alongside texture and grain, represent particular and valuable aesthetic features of the various wood species [1]. The coating of wood surfaces with transparent, colourless finishing materials is the usual treatment applied to highlight and preserve their beauty and natural aspect over time in various practical applications. Nevertheless, colour changes in both uncoated and coated wood surfaces occur over time when exposed to light/UV radiation, both indoors and outdoors. This practical issue has been only partly solved by the outcomes of extensive research worldwide, remaining an actual research topic [2,3,4,5,6,7].

Colour changes in wood surfaces exposed to light are primarily the result of a complex and dynamic photo-degradation process of the wood surface, having as main actors the UV radiation from the light source and the lignin in wood. The absorption of UV radiation by lignin results in chemical bond cleavage and the formation of reactive free radicals (e.g., phenoxyl), which are then transformed into carbonyl- and carboxyl-containing chromophores and quinoid structures, responsible for colour changes, especially yellowing of the exposed wood surfaces [8,9,10,11,12,13]. Not only UV light but also visible violet light and blue light up to 515 nm may cause surface and subsurface colour changes without affecting lignin [14,15]. Furthermore, a possible contribution of wavelengths in the visible-near infrared region was highlighted by hyperspectral imaging [16].

In addition to lignin, certain classes of wood extractives may play an important role in determining the natural colour of wood species and their behaviour when exposed to light. Poly-phenolic extractives, such as tannins, can absorb UV light, scavenge free radicals and undergo photo-oxidation, retarding the photo-degradation of lignin and wood [17,18,19,20]. On the other hand, extractives from dark coloured wood species are responsible for their sensitivity to visible light, resulting in a bleaching effect after longer exposure times [21].

The result is, therefore, that the light-induced colour changes in wood surfaces are significantly influenced by the wood species due to the differences in their chemical composition, as proven by comparative research on more than two different species [1,22,23,24]. Colour changes also depend on the type of light (natural/artificial), spectral range (wavelengths), type of light source, exposure conditions and duration [25,26,27]. Accordingly, data from different research publications are often difficult to compare.

The coating of wood can impart some protective effect against light-induced colour changes by reducing the amount of light/UV radiation reaching the underlying wood substrate [28], depending on the chemical structure of the film-forming resin and the specific additives in a certain formulation. The coating materials can also undergo photo-degradation processes (photolysis, photo-oxidation) under the action of UV light, resulting in property changes, including glass transition temperature, UV light transmittance, colour and their protective behaviour [29,30]. Different types of film-forming materials, such as: acrylic-polyurethanes [31], acrylics [28,29,32,33,34] and polyurethanes [35], as well as natural products [36], have been studied in terms of behaviour and protective effect against light/UV-induced colour changes in wood surfaces.

Due to their transparency for both visible and UV light, clear finishes offer generally less protection against the action of light/UV radiation, resulting in more pronounced colour changes [28,29,31,33,35], compared to semi-transparent and opaque finishes that contain colouring pigments with a UV protective effect.

The interaction of UV light with the underlying wood at the interface wood/coating material results in the photo-oxidation of lignin and associated colour changes, representing an important contribution to the overall colour changes in clear-coated wood surfaces [10,28,35]. Different principles for stabilising the colour of clear-coated wood have been in research focus: (i) pre-treatments of the wood substrate for an improved UV resistance, (ii) modification of the coating materials with additives for improved UV resistance and (iii) combined methods. Nanoparticles of TiO₂ [5], ZnO [37], CeO₂ [3], CH₂O [38] or α/γ Fe₂O₃ [6], UV absorbers (benzotriazole, [39], benzophenone [40]), HALS (hindered amine light stabiliser), (Tinuvin 292, Tinuvin 293 [41], Tinuvin 5151 [39]), or chemical modification treatments with succinic anhydride [42], maleic anhydride [43] or soybean oil modified epoxy resins [40] were employed as pre-treatments. Organic UV absorbers and HALS stabilisers [34,36] and inorganic UV absorbers and nanoparticles [6,44] were employed as additives for clear coatings. Approaches combining the two principles were also tested [4,13,33,39,41].

Most of the studied references reporting light-induced colour changes in the coated wood surfaces refer strictly to the changes observed and measured on the directly exposed coated wood surface [1,5,17,22,45], without any information on the possible photo-degradation of the wood substrate under the coating film. However, colour changes in clear-coated wood may be caused by colour changes in both the clear-coating film and underlying wood, or either of them [35], and their contribution should be acknowledged and evaluated to deeper understand the occurring phenomena. References addressing this important issue are rather rare [35,36,46,47]. Uncoated wood specimens and free coating films were employed to simulate clear-coated wood. A light reflection model was used to elucidate separately the colour changes caused by the clear-coating film and the underlying wood, showing a major contribution of the underlying wood compared to the coating film for two types of polyurethanes (approx. 60%/40%; 95%/5%) [35]. Different research concluded that in the case of spruce wood clear coated with a water-borne acrylic, the colour changes after 200 h UV irradiation were due almost entirely to the wood substrate, as colour changes in the coating film were negligible [47].

More studies have focused on the exposure of wood in outdoor conditions where UV radiation acts in conjunction with rainwater [2,32,45], while the topic of light-induced colour changes in indoor conditions by the natural light passing through the window glass [1,22,23,36] or light from artificial sources [26,27] has been less addressed, especially for clear-coated wood with natural aspect [46,48].

All these point to the idea that stabilising the natural colour of wood species exposed to light in indoors conditions remains a challenge that could be better addressed based on a deeper understanding of the occurring phenomena and influencing factors.

In this context, the objectives of the present study were: to evaluate comparatively the light-induced colour changes for uncoated and clear-coated wood surfaces with natural aspect in indoor conditions by accelerated testing; to determine the influence of the species and type of coating on the colour changes; and to highlight the specific contributions of the wooden support and of the coating films to the total colour changes in the finished surfaces. A dedicated laboratory method was employed for this purpose [46].

2. Materials and Methods

2.1. Wooden Material

Three hardwood species: European maple (Acer pseudoplatanus L.), European ash (Fraxinus excelsior L.) and European walnut (Juglans regia L.), were employed. The wood material was sourced from the company “Craiul Munților SRL”, Stupărei village, Valcea county, Romania. These species were selected based on their aesthetic appearance and importance for furniture and other interior design element manufacturing, including wooden floors and panelling. These species differ in natural colour, anatomical structure, texture and the presence or not of colour-differentiated sapwood and heartwood areas, so that they can provide beautiful contrasts that can be valorised by design, while also representing a group with sufficient variability in structure and properties to provide useful research data. Moreover, the expected research results will add new information to a data base, including their ageing properties as bare substrates or coated surfaces with natural traditional finishing materials [24], allowing useful comparisons.

Test samples of a rectangular form of 120 × 80 × 8 (mm × mm × mm) with radial or semi-radial faces were prepared from air-dried wood material. All samples were sanded manually with abrasive paper (120 and 180 grit sizes).

2.2. Coating Materials

Three types of clear, waterborne, colourless lacquers, falling into the classes of acrylic-polyurethane F1 (YO 20-M702), acrylic F2 (YO 15-M864) and polyurethane F3 (YO 20-M838) coatings (

Table 1. Waterborne transparent coating materials and additives (Renner–Italy) employed in the experimental research.

Lb. Code	Commercial Code	Film-Forming Resin	Solids Content, Csu, %	Density, g/cm3	Formulation	Gardner Gloss Index
F1	YO 20-M702	Acrylic-polyurethane (non-yellowing)	25 ± 1	1.02	1k	20
F2	YO 15-M864	Acrylic (non-yellowing)	34 ± 1	1.03	1k	15
F3	YO 20-M838	Polyurethane	30 ± 1	1.03	1k	20

), were employed for wood finishing with natural aspect. These are all products of Renner Italy (Minerbio, Italy) [49], commercialised in Romania by Kroncolor Brașov (Brașov, Romania) [50]. The selection of these materials took into account ecological aspects (preference for waterborne products instead of solvent-based), their market share as finishing materials, the quality of the cured coating films ensured by their chemistry, alongside the recommendation of the specialists from Kroncolor for transparent finishes with natural appearance.

2.3. Test Sample Preparation

Wood specimens (uncoated and coated) and clear microscopic slides (1 mm thickglass lamellae) as such and coated with the experimental finishing materials (F1, F2, F3 as 1k formulations) were employed in this research. The finishing of the wood samples was carried out in laboratory conditions, applying two successive layers of approximately 90–100 g/m² (corresponding to an approximately 100 μm thickness as a liquid), at an interval of 2–4 h. The same finishing technology was applied for the glass lamellae, employing a special device for coating film application (2 layers of 100 μm), with the intention to ensure reproducible testing conditions such as film thicknesses for the test samples V1 (uncoated wood + coating film of glass lamella) and F2 (coated wood). It should be noted, however, that some differences resulted from the practical procedure and the variation in the solids content and viscosity of the three coating materials. Moreover, the differences in the anatomical features of the three wood species: maple—diffuse porous with small vessels not visible with the naked eye; walnut—diffuse porous with larger vessels visible with the naked eye; and ash—ring porous with distinctly larger vessels clearly visible in early-wood, resulted in different coating penetration and coverage of the wood grain. Accordingly, semi-filled/nearly filled grain finishes were obtained for maple, and open grain finishes for walnut and ash. The preliminary examination of the transversal cross-cut edges of the coated samples confirmed visual perception in terms of pores filling and highlighted the interaction between the coating and the wood substrate at the interface, resulting in different penetrations of the coatings into the wood substrates. It could be observed that strongly interconnected “composite coating layers”, including coating films on top of wood surfaces and impregnated sub-surfaces, were obtained.

The actual spread rates applied and the (average values), determined by weighing each sample immediately after coating application, as well as the theoretical thickness of the resulting cured coating films, calculated considering the average spread rates and the solids content of each coating material but no wood penetration, are summarised in

Table 2. Experimental values referring to the preparation of the coated wood samples and glass lamellae with the three types of coating materials: application rates, theoretical film thickness on wood samples, thickness film on glass lamellae, alongside the solids content and viscosity of the applied coating materials F1, F2, F3.

Wood Species Substrate	Experimental/Calculated Value	Type of Coating
		F1	F2	F3
Maple	Application rate [g/m2]	191.1 ± 16.2	196.3 ± 19.7	203.5 ± 20.2
	Film thickness [μm]	47.8	57.9	55.6
Ash	Application rate [g/m2]	180.5 ± 25.8	203.3 ± 17.4	168.3 ± 4.0
	Film thickness [μm]	45.1	59.9	46.0
Walnut	Application rate [g/m2]	161.6 ± 16.2	184.9 ± 28.6	189.4 ± 17.4
	Film thickness [μm]	40.4	54.5	51.8
Glass lamella	Film thickness [mm]	0.034 ± 0.004	0.043 ± 0.008	0.046 ± 0.007
Coating material	Solids content [%]	25.0	29.5	27.3
	Flow time, ϕ 6 mm, 20 °C [s]	12	49	6

, The actual thicknesses of the cured coating films applied on glass lamellas, determined with a precision of 0.001 mm, are also included.

Two types of test samples for UV-VIS light exposure, coded V1 (uncoated wood support-Figure 1A) and V2 (coated wood support Figure 1B), resulted from assembling wood samples with glass lamellae, following an original concept [46]. This was developed to allow highlighting and measurement of the light-induced colour changes for the uncoated (V1/Z3) and coated wood surfaces (V2/Z1), the coating films (on glass lamellae) and the wood substrate under the coating film (simulated by V1/Z1, uncoated wood under coated glass lamella), in the various specific zones differentiated as exposure situations. Both types of test samples included a colour control area (V1/Z4, V2/Z2), tightly covered with black cardboard and aluminium foil to prevent the penetration and effect of light. The two types of test samples (V1, V2) were prepared in triplicate for each tested variant (wood species, type of coating), resulting in a number of 18 samples/species, respectively, 54 samples in total.

2.4. Colour Measurements

Colour measurements in the CIELab reference system were performed using an Ava-spec-USB2 spectrometer from Avantes (Apeldoorn, The Netherlands) [51] equipped with an integration Ava Sphere with a diameter of 80 mm, with a circular measuring aperture of 8 mm in diameter. Measurements were carried out under a standard D65 illuminant and the measurement angle set at 10 degrees, according to the relevant standard methods [52]. The light connection to the sphere was via glass fibres and a COL-UV-VIS collimating lens connected to the light source. The colour data were processed employing Ava soft 7.0.

The measured colour parameters were: the lightness value (L*), varying from 0 for black to 100 for white; the degree of red (redness) on the green–red chromatic axis (−a*, +a*); and the degree of yellow (yellowness) on the blue–yellow colour axis (−b*, +b*). The colour measurements were performed on 12 fixed points (circular areas of 8 mm diameter) on each wood sample (uncoated or coated) and on 3 fixed points for the coated glass lamellae, prior to assembling the wood specimens with the glass lamellae. After each period of light exposure, the samples were disassembled and colour measurements repeated in the same areas on the wood substrate and the coated lamellae. The colour measuring points on the wood samples corresponded to 3 fixed locations on each zone of the test samples, excepting zones V2/Z1 with 6 measuring points. As the test samples were in triplicate, a minimum number of 9 and maximum 18 measurements resulted for each zone/exposure situation for each tested variant. All the colour data were collected and further processed in Excel to calculate the light-induced variation in colour coordinates (ΔL*, Δa* and Δb*) and the total (global) colour difference (ΔE) according to the relationships (1)–(4):

$${∆\mathrm{L}}^{\mathrm{*}}={\mathrm{L}}_{\mathrm{e}}^{\mathrm{*}}-{\mathrm{L}}_{\mathrm{i}}^{\mathrm{*}}$$

(1)

$$\mathsf{\Delta }{\mathrm{a}}^{\mathrm{*}}={\mathrm{a}}_{\mathrm{e}}^{\mathrm{*}}-{\mathrm{a}}_{\mathrm{i}}^{\mathrm{*}}$$

(2)

$$\mathsf{\Delta }{\mathrm{b}}^{\mathrm{*}}={\mathrm{b}}_{\mathrm{e}}^{\mathrm{*}}-{\mathrm{b}}_{\mathrm{i}}^{\mathrm{*}}$$

(3)

$$\mathsf{\Delta }\mathrm{E}=\sqrt{\left({\left(\mathsf{\Delta }{\mathrm{L}}^{\mathrm{*}}\right)}^2+{\left(\mathsf{\Delta }{\mathrm{a}}^{\mathrm{*}}\right)}^2+{\left(\mathsf{\Delta }{\mathrm{b}}^{\mathrm{*}}\right)}^2\right)}$$

(4)

where: e—exposed; i—initial.

A series of ΔL*, Δa*, Δb* and ΔE values calculated for each measuring point of a certain area/exposure situation (9 or 18 values) and their means were employed for statistical analysis of the colour changes after 72 h of light/UV exposure.

2.5. Accelerated UV–VIS Light-Induced Ageing Procedure

A Feutron 400 FKS environmental climatic chamber (Feutron Klima Simulation–GmbH-Greiz, Germany), equipped with a UVA Spot 400 T lamp (Hoenle UV Technology–Gilching, Germany), fitted with a glass H2 filter, was employed to expose the wood samples to UV-VIS light in the range of 295 to 600 nm, simulating natural light filtered by window glass, similarly to the previously reported research [23,53]. Considering the maximum emission of the lamp in the UV range (mainly UV-A, as indicated also by its name), this type of radiation with higher energy was considered to be the main ageing factor causing colour and chemical changes in our experiments. However, the expression UV lamp and UV exposure were still employed in this paper, including the graphs, for commodity reasons. It should be acknowledged, however, that visible light with wavelengths up to 515 nm may contribute to the surface and sub-surface wood colour changes, while visible violet radiation up to 430 nm may have a contribution to lignin degradation [14]. Accordingly, the visible light emitted by the source employed might have had a contribution to the colour changes measured in this research.

The samples were placed vertically on a rack at a distance of 60 cm from the UVA light source. The actual irradiation procedure, similar to the previously reported research [53], included an initial conditioning of 0.5 h (20 °C, 55% RH, no light), followed by 3 cycles of 24 h UV exposure separated by conditioning phases of 24 h (20 °C, 55% RH, no light), inserted to allow the removal of samples for measurements. Each cycle of 24 h UV exposure consisted of four steps of 6 h UV irradiation at 40 °C, alternated with dark periods of 0.5 h. Cycles of 24 h UV exposure are in accordance with a standardised procedure of exposure to daylight behind window glass (method B) [54]. This procedure was repeated three times so that the data refer to samples exposed for 24, 48 and 72 h to UV radiation. All the samples were in triplicate and exposure was carried out in two series. The position of the samples on the rack was changed after each period of 24 h exposure in order to ensure a similar degree of irradiation.

2.6. Data Processing and Statistical Analysis

Microsoft Excel version 2019 and the operating system Windows11 was employed for data collection and their basic processing by calculation of the average values and standard deviations for each group of 9 or 18 measurements. One-way Anova test in conjunction with Tukey’s honestly significant difference (HSD) as post hoc tests were employed to check the significant differences between the means of different groups at a significance level α = 0.05. Two-way Anova with replication was employed to evaluate the statistical significance of the wood species, type of coating material or exposure situation (specific areas of the test samples) as influencing factors on the global colour changes ∆E after 72 h light/UV exposure. The data analysis tool-pack in Excel 2019 and Statistica 8 software were used for statistics.

3. Results and Discussion

3.1. Light-Induced Colour Changes in the Uncoated Wood Surfaces

Light exposure of the uncoated wood surfaces (samples V1) resulted in progressive colour changes, visually perceived mostly as darkening, with some shade modifications, depending on the exposure situation and wood species. The evolution of these colour changes, expressed by the variation in the ∆E, ∆L*, ∆a*and ∆b* values as a function of the exposure time, is illustrated by the graphs in Figure 2, comparatively for the areas V1/Z3 directly exposed and the area V1/Z2 covered with a clear glass lamella.

The total colour changes (ΔE) in the unfinished wood support directly exposed to light/UV radiation (V1/Z3 areas) increased gradually during exposure, though these changes were more rapid in the first 24 h. The highest values after 72 h exposure were registered for maple (9.36 units), followed by ash (8.39 units) and then walnut (6.20 units). A similar order of light sensitivity for these three species was determined by [1], ΔE values of approximately 14.1 units, 12.5 units and 9.8 units resulting after longer exposure (600 h). Colour changes of 12.5 units for maple and 11.7 units for walnut were reported after 7 years exposure to natural light passing through the window glass [48].

The total colour changes resulted mostly from the decrease in lightness, highlighted by the negative ΔL* values evolving following a mirrored pattern when compared to ∆E and reaching the final values of (–5.07) units for maple, (−4.13) units for ash and (–3.25) units for walnut (Figure 2(A2)).

From a chromatic point of view, the positive Δa* values (up to 2.23 units) indicate only a slight increase in redness, while the positive Δb* values (up to 7.58 units) indicate a more important increase in the yellowness, specific to each wood species. The previous research and literature data indicated that the redness of wood, as well as UV/light-induced changes in redness, might be associated to the extractives [55], while the increase in yellowness can be correlated to UV-induced lignin degradation and formation of new quinones and stilbenes chromophores [8,9,11,12]. As lignin photo-degradation is the first and main photo-chemical process induced by UV light, higher changes in yellowness compared to redness are expected. In the present research, the highest increase in redness (Δa*) was determined in the case of walnut (2.23 units), followed by maple (1.48 units) and ash with quite similar values (1.43 units). Contrarily, the lowest change in yellowness was determined for walnut (Δb* = 4.42 units), compared to ash (7.05 units) and maple (7.58 units), with the highest value. This is in accordance with the previous research results and may be explained by the role of extractives [23,53]. Wood extractives may act as UV absorbers and free radical scavengers while themselves undergoing the photo-oxidation processes. All these effects result in retarding the photo-degradation of lignin in wood [18,19,20], which means less yellowing. A higher content of extractives in the walnut wood compared to maple and ash wood may explain both the darker natural colour of walnut and the reduced colour changes compared to maple following UV exposure, at least for the duration of the test carried out in this research (72 h).

The light-induced colour changes for the uncoated wood surfaces that were covered with a 1 mm thick clear glass slide (V1/Z2) are presented in Figure 2B. It can be observed that the total colour changes (ΔE) are smaller than those registered for surfaces directly exposed (V1/Z3) to UV light for all the three species, though the decrease in ∆E values varied among species, being higher for maple and ash (approximately −4 units) than for walnut (−1.24 units). This shows an effect of the glass by diminishing the intensity/energy and very likely the spectral distribution of light reaching the wood surface underneath. The literature data [56,57] and our own experience have shown reduced transmittance indexes of glass for UVB and UVA wavelength compared to visible radiation. The effect of glass is visible especially in the light-induced changes in lightness (∆L) and yellowness (∆b*), while it is almost absent in the case of redness (∆a*). This may support the assumption of selective partial blocking of UV radiation from the spectral distribution of the light emitted by the source, so that wood under the glass is exposed not only to light of lower intensity/energy but also to light with a different spectral distribution, with less UV radiation and more VIS radiation. This may also explain why the colour changes were less affected by the cover glass slide in the case of walnut, with a darker, reddish natural colour due to the extractives, which may absorb more visible radiation. The previously published research has demonstrated that dark natural wood species undergo intensive photo-bleaching when exposed to daylight, the most detrimental effect being caused by wavelengths of the visible spectral range, which affects mostly the low molecular extractives [21].

The glass effect is clearly a limitation of the method employed in this research in the attempt of highlighting and quantifying the colour changes in the wood substrate occurring under the coating film by simulating this situation with coated glass lamella on top of uncoated wood (V1/Z1).

3.2. Light-Induced Colour Changes in the Coated Wood Surfaces

Exposure of the coated wood surfaces to light (V2 samples) resulted generally in smaller colour changes, dependant on the wood species and the type of finishing material.

The graphs in (Figure 3A) represent the colour changes in the wood samples coated with the acrylic-polyurethane lacquer, coded F1, in the areas directly exposed to light/UV radiation (F1/V2/Z1) as a function of the exposure duration. The evolution of the global colour changes, expressed by the ΔE values, show a quite similar pattern for maple and ash, with maximum changes occurring during the first 24 h of exposure (3.96 units for maple and 2.16 units for ash), followed then by a slower increase up to the 5.71 units for maple and only 3.37 units for ash. The evolution of colour changes for the walnut samples was totally different, almost linear in time, reaching the highest value of 7.70 units after 72 h of light exposure.

The changes in lightness varied depending on the wood species: an increase in lightness, expressed by positive ΔL* values, reaching (+3.40) units after 72 h exposure was registered for walnut, whilst for the maple and ash coated specimens the lightness decreased, as shown by the negative ΔL* values, reaching (−3.83) units for maple and only (−0.52) units for ash following a reversed evolution after 24 h exposure.

From a chromatic point of view, only slight variations in redness were registered as positive, with Δa* values up to (1.09) units for maple and values up to (1.14) units walnut, while those measured for ash were negative but very close to zero (−0.06 units). Positive Δb* values indicated for all the coated wood species a continuous increase in yellowness, up to 2.65 units for ash, 4.03 units for maple and 6.69 units for walnut. To conclude, the colour stability of the wood samples coated with the acrylic-polyurethane water-based lacquer decreased in the following order: F1-ash > F1-maple > F1-walnut, which is different from that observed for the uncoated wood samples: walnut > ash > maple.

The light-induced colour changes for the same coated wood surfaces covered with 1 mm thick clear glass (V2/Z3) are shown in (Figure 3B). It can be seen that the total colour changes ΔE are smaller than those recorded on the surfaces directly exposed (V2/Z1) to light for all the three species, but to a different extent. The effect of covering glass in reducing the global colour changes (∆E) for the coated wood (Figure 3(B1)) was similar to uncoated wood, minimal for walnut and clearly more accentuated in the case of maple (maximum) and ash (medium). However, according to the graphs plotted in (Figure 3(B2–B4)), the effect of glass seemed to be more uniformly distributed on all the colour parameters L*, a*, b*, with some specificity related to the wood species. In the case of maple, for instance, it is clearly visible that the yellowness increase due to light exposure was considerably reduced (from approximately 4 units to approximately 1 unit) for the coated samples covered by glass. It is important to highlight that the order of colour stability among the F1 coated wood samples was not changed by the effect of glass cover.

A similar analysis of the light-induced colour changes for wood surfaces coated with the acrylic water-based lacquer, coded F2, is presented in (Figure 4), comparatively for the directly exposed areas (F2/V2/Z1) (Figure 4A) and areas exposed under a glass covering slide (F2/V2/Z3) (Figure 4B).

For the directly exposed areas (F2/V2/Z1) (Figure 4A), the evolution of the global colour changes, expressed by the ΔE values, showed a quite similar pattern for maple and ash, with changes occurring rapidly during the first 24 h of exposure (3.34 units for maple and about 2 units for ash), followed then by a slower increase up to the 5.23 units for maple and 3.49 units for ash. The colour changes for the walnut samples evolved somehow differently, almost linearly in time, reaching the highest value of 5.53 units after 72 h of light exposure. A trend of darkening expressed by negative ΔL* values was registered for all wood species after the first 24 h of light exposure, but this trend was reversed by prolonged exposure. After 72 h light exposure, a slight increase in lightness was registered for walnut (ΔL* = +0.96 units), while the ΔL* values for ash (−0.12 units) and maple (−0.48 units) remained slightly negative but very close to zero.

From a chromatic point of view, the variation in redness ∆a* and yellowness ∆b* depended on the wood species. An increase in redness Δa* was determined only for walnut (2.16 units), the values for ash and maple being close to zero. Positive Δb* values indicated for all the coated wood species a continuous yellowing, following species dependant patterns, up to 3.45 units for ash, 4.96 units for walnut and 5.11 units for maple. To conclude, the colour stability of the wood samples coated with the acrylic water-based lacquer decreased in the following order: F2-ash > F2-maple ≅ F2-walnut. This is different from that observed for the uncoated wood samples (walnut > ash > maple) and those coated with the acrylic-polyurethane finish F1 (F1-ash > F1-maple > F1-walnut), pointing out that both the wood substrate and the coating film influence the colour light stability of the coated wood.

The light-induced colour changes for the same coated wood surfaces, covered with 1 mm thick transparent glass (F2/V2/Z3 areas), plotted in (Figure 4B), were again smaller than those determined for the directly exposed areas (Figure 4A), the influence of the covering glass being also dependant on the wood species. The highest decrease in the global colour changes (ΔE after 72 h of exposure) was observed again for maple (about −3 units), while the lowest was observed for walnut (about −0.5 units) (Figure 4(B1)). According to the graphs plotted in (Figure 4(B2–B4)), light exposure through the covering glass influenced mostly the evolution of the yellowness changes ∆b*, which were reduced for all species, with a more pronounced influence (about −3 units) for the maple and ash samples. It is important to highlight that the order of colour stability among the F2 coated wood samples was slightly changed by the effect of the glass cover (S), so that a clear difference could be observed between walnut and maple: S-F2-ash > S-F2-maple > S-F2-walnut.

Finally, the light-induced colour changes for wood surfaces coated with the polyurethane water-based lacquer (1 K), coded F3, are presented in Figure 5 comparatively for the directly exposed areas (F3/V2/Z1) (Figure 5A) and areas exposed under a glass covering slide (F3/V2/Z3) (Figure 5B).

For the directly exposed areas (F3/V2/Z1) (Figure 5A), the evolution of the global colour changes, expressed by the ΔE values, show more specific wood species patterns. For ash (coated F3), fast and important colour changes occurred during the first 24 h, a maximum ∆E value of approximately 4 units being registered after this period. The process was reversed with about the same speed by continuing the light exposure, the ∆E values determined after 48 h and 72 h being considerably lower. The individual patterns for maple and walnut remained very similar to those presented for the samples coated with the other two finishes in Figure 3(A1)(F1) and Figure 4(A1)(F2), being characterised by: (i) changes occurring more rapidly during the first 24 h of exposure, followed then by a slower increase in the case of maple; (ii) changes increasing almost linearly in time in the case of walnut. After 72 h light exposure the measured ∆E values for the F3 coated samples, listed in increasing order, were: −2.25 for ash, 4.74 for walnut and 4.92 for maple.

A trend of darkening expressed by negative ΔL* values was registered after the first 24 h of light exposure, especially for ash and maple. This trend was reversed by continuing the light exposure in the case of ash but continued at a lower slope in the case of maple. In the case of walnut, only a very slight darkening was registered after 24 h light exposure and the further reversing of the process resulted in a lightening effect after 72 h, contrarily to the other two species. The comparative ∆L* values (glass/direct exposure) after 72 h were as follows: −3.17/−3.56 for maple, −0.17/−0.51 units for ash and +0.63/2.14 units for walnut.

From a chromatic point of view, the variation in redness ∆a* and yellowness ∆b* depended also on the wood species. A progressive increase in redness during light exposure was registered for maple (up to +2.60 units) and walnut (up to +1.61 units), while the values for ash had a slight wavy variation, resulting in only a very low increase (+0.49 units after 72 h).

The variation in the chromatic component b* was totally different. A trend of a very slight but continuous decrease in yellowness was registered for maple (up to about −0.5 units). A more advanced decrease in yellowness (up to approximately −3 units) was registered for ash samples but only during the first 24 h of exposure; afterwards, this trend was reversed so that after 72 h light exposure, almost no variation in yellowness could be observed any longer (∆b* = +0.24 units). An increase in yellowness after 72 h light exposure could be measured only for walnut (∆b* = +3.24 units).

Based on the global colour differences ∆E after 72 h light exposure, the colour stability of the wood samples coated with the waterborne polyurethane (1 K) lacquer F3 decreased in the following order: F3-ash > F3-walnut ≅ F3-maple. This is different from those observed for the uncoated wood samples (walnut > ash > maple) and those coated with the acrylic-polyurethane finish F1 (F1-ash > F1-maple > F1-walnut) but closer to those resulting for the samples coated with the acrylic finish F2 (F2-ash > F2-maple ≅ F2-walnut).

The light-induced colour changes for the same coated wood surfaces, exposed behind 1 mm thick clear glass (F3/V2/Z3 areas), plotted in Figure 5B, were again different but not always smaller than those determined for the directly exposed areas. Looking comparatively at the corresponding evolution patterns of the colour differences in the Figure 5(A1–A4,B1–B4), it can be observed that the most obvious pattern change due to the effect of the covering glass appeared for the yellowness variation (∆b*), especially for maple and ash. When exposed through the 1 mm clear cover glass, the F3 coated maple samples yellowed progressively, whilst when uncovered, there was a very slight trend of decrease. The pattern for ash was also different, with a less pronounced decrease in yellowness in the first 24 h, but finally after 72 h, there were no significant changes in yellowness, similarly to the uncovered samples. For walnut, the evolution pattern did not change shape, but the final increase in yellowness was slightly increased. Exposure through the covering glass reduced the variation in lightness to different extents depending on the wood species but did not change visibly the evolution patterns. It can be noted a smaller lightening for walnut, while the darkening effect for maple and ash seems less affected. Accordingly, there were some small differences in the evolution patterns of ∆E values for the three species and the final values recorded after 72 h light exposure. Based on these values, the colour stability of the wood samples coated with the waterborne polyurethane (1 K) lacquer F3 and covered by a 1 mm thick clear glass slide (S) decreased in the following order: S-F3-ash ≥ S-F3-walnut > S-F3-maple, which is slightly different from those observed earlier for direct exposure (F3-ash > F3-walnut ≅ F3-maple), especially as walnut is concerned. It has to be pointed out that for this type of finish (F3), the glass cover showed sort of a shielding effect only in the case of walnut, reducing the final ∆E value by 1.41 units, whilst the corresponding values for maple and ash substrates actually increased with 0.16, respectively, 0.35 units.

More research on an increased number of replicates and complementary investigation methods, including spectral analysis of the light emitted by the source and passing through the 1 mm thick clear glass cover, as well as the interaction of the respective light with the different wood substrates and the coating films in terms of absorption, reflection or transmittance are necessary to better understand and explain the experimental results obtained so far.

3.3. Light-Induced Colour Changes in the Wood Substrate under the Coating Film, Simulated by V1/Z1 Zones—Contribution of the Underlying Wood Substrate to Overall Colour Changes

The experimental method employed in this research envisaged a possibility to highlight, measure and compare the light-induced colour changes in the wood substrate under the coating film and in the coating film itself. For this purpose, sort of a “detachable” coating film was necessary, and this goal was achieved by applying the coating films on 1 mm thick clear glass slides. These were tightly attached to the uncoated wood substrate during light exposure and removed after each exposure period to allow evaluation of the colour changes in the coating film (on glass slide) and in the substrate under the protective coating film (areas V1/Z1). As was already mentioned, the influence of the 1 mm thick clear glass slide was a limitation of this method and was considered by comparative evaluations with areas V1/Z2 (exposure through clear glass). The graphs plotted in (Figure 6) illustrate the in-time evolution of the global colour changes, expressed as ∆E values, for wood substrates exposed to light under the three different types of coating films: acrylic-polyurethane (F1), acrylic (F2), polyurethane (F3) applied on glass slides (SF1, SF2, SF3), alongside the control curves highlighting the effect of the clear glass (S). It can be clearly observed that light penetrated through the transparent coating film and the clear glass slide, reaching the wood substrate underneath and causing colour changes that evolved with the duration of exposure. The evolution patterns in the colour changes were very similar for exposure under clear glass (S) and exposure under coated glass (SF1–SF3), with only some small differences in the actual ∆E values. Slightly lower values under the coated glass may suggest some protective effect of the respective coating films, whilst slightly higher values may indicate an opposite effect. However, when compared to the colour differences for wood exposed under clear glass the differences brought about by the coatings applied on glass were very low, as can be observed in

Table 3. Comparative data of colour changes after 72 h light/UV exposure for wood substrate (maple, ash and walnut) under the coating films (coating materials F1, F2, F3), simulated in V1/Z1 areas by uncoated wood exposed under coated glass slides; the data on colour changes in the same substrates under clear glass slides (areas V1/Z2) are presented as reference values to evaluate the effect of the coatings.

Wood Species		Colour Changes after 72 h Light /UV Exposure
		V1/Z2 Uncoated Wood under Clear Glass (Reference Values)			V1/Z1 Uncoated Wood under Coated Glass
		(F1)	(F2)	(F3)	F1	F2	F3
Maple	∆E	4.90 (1.39)	5.80 (1.13)	5.20 (1.86)	4.44 (1.36)	6.02 (1.55)	5.17 (1.73)
	∆L*	−3.56 (1.42)	−4.29 (0.62)	−3.43 (1.46)	−3.51 (1.11)	−4.37 (1.16)	−3.36 (1.85)
	∆a*	0.99 (0.52)	1.74 (0.51)	0.70 (0.68)	1.00 (0.32)	1.26 (0.71)	0.45 (0.71)
	∆b*	2.95 (1.23)	3.40 (1.16)	3.69 (1.47)	2.33 (1.24)	3.82 (1.28)	3.64 (1.15)
Ash	∆E	4.84 (0.76)	4.23 (0.82)	4.23 (1.34)	4.32 (1.15)	4.20 (0.77)	4.03 (0.96)
	∆L*	−3.25 (0.55)	−2.94 (0.76)	−3.23 (1.30)	−3.07 (1.11)	−3.16 (0.59)	−2.98 (1.39)
	∆a*	0.80 (0.32)	0.67 (0.58)	0.92 (0.65)	0.79 (0.54)	0.76 (0.27)	0.50 (0.82)
	∆b*	3.45 (0.73)	2.70 (1.18)	2.22 (1.25)	2.79 (0.88)	2.62 (0.64)	2.18 (0.94)
Walnut	∆E	4.77 (1.38)	4.51 (1.13)	5.52 (1.52)	4.51 (1.54)	4.83 (1.35)	5.93 (1.20)
	∆L*	−3.37 (1.88)	−2.07 (0.74)	−2.13 (1.43)	−3.18 (1.60)	−2.32 (0.79)	−2.22 (1.38)
	∆a*	2.26 (0.66)	2.55 (1.10)	2.51 (0.74)	2.19 (0.61)	2.08 (1.12)	2.41 (0.72)
	∆b*	2.99 (1.03)	2.99 (1.34)	4.15 (1.55)	1.83 (1.33)	2.94 (2.35)	4.56 (1.77)

Note: The data represent average values of colour changes in the 9 measuring points; standard deviations are presented in brackets in italics.

, where the average data after 72 h light exposure are presented.

Analysis of paired data by using the one-way Anova test at significance level α = 0.05 found no statistically significant differences between the mean ∆E values in the areas V1/Z1 and V1/Z2 (p values higher than 0.05), irrespective of the wood species and the type of coating material (F1, F2, F3). This means that no protective effect of the coating films on the colour changes in the wood substrate under the coating could be observed by the applied method. This statistics result was in good accordance with visual perception when analysing comparatively the light-induced colour changes in the areas V1/Z1 and V1/Z2 on the tested samples.

On the other hand, the fact that notable colour changes in the wood substrate occurring under the coated glass (V1/Z1) could be observed and quantified represents an important original contribution of this research. Moreover, the wood species and the type of coating were found as statistically significant influencing factors on the colour change values ∆E, determined for the V1Z1 areas after 72 h light/UV exposure, when the respective colour data were analysed by the Anova two-factor test with replication (summary in

Table 4. Summary of two-factor Anova test results with replication of the colour change values ∆E, after 72 h light exposure, in the areas V1/Z1, considering the wood species (1) and the type of coating (2) as influencing factors. (α-0.05).

Anova: Two-Factor with Replication (α = 0.05)
Source of Variation	SS	df	MS	F	p-Value	F Crit
Wood species (1)	44.95563	2	22.47781	12.13096	2.88142 × 10−5	3.123907
Type of coating (2)	13.78771	2	6.893853	3.720516	0.028997068	3.123907
Interaction (1)x(2)	18.01943	4	4.504858	2.431209	0.055229542	2.498919
Within	133.4109	72	1.852929
Total	210.1737	80

). The higher F value indicates wood species as a more important influencing factor than the coating film on the colour changes in the underlying substrate. The interaction of the two factors was not found statistically significant.

The colour changes in the wood substrate due to light exposure under coated glass slides (V1/Z1) were determined as a simulation of the behaviour (contribution) of wood substrate under the coating films (V2/Z1), situations where there is no effect of the glass. Accordingly, the real colour changes in the substrates underneath the coating films and the associated chemical changes should be higher than those determined by the method applied in this research. This major contribution of the wood substrate to the total colour changes in the clear-coated wood surfaces is an important key point in understanding the occurring phenomena with the view of their correction, as also pointed out by the previous research [35,47]. For equivalence as exposure situations, the colour and related chemical changes in the substrate in the areas V1/Z1 should be comparable with the colour changes in a wood substrate under the coating film after exposure to light passing through a covering clear glass slide, respectively, areas (V2/Z3) in this research.

3.4. Light-Induced Colour Changes in the Coating Films—Contribution of Finishing Film to the Overall Colour Changes

The colour changes (ΔE, ΔL*, Δa*, Δb*) in the finishing films applied on 1 mm clear glass slides as a result of light exposure for different periods of time are presented in Figure 7. It can be easily observed that for all the three types of coatings, there were registered similar wavy patterns in colour changes. These were maximal after the first 24 h of exposure and consisted mainly in slight darkening (∆L* negative) and yellowing (∆b* positive). By further light exposure, the initial trend reversed twice. Redness evolved similarly but the changes were very low, close to zero. Consequently, the total colour changes (ΔE) after 72 h light exposure for the three types of finishing films were very low: 1.30 units for F1(acrylic-polyurethane), 1.25 units for F2 (acrylic) and 1.07 units for F3 (polyurethane). However, these differences in the means for the three types of coatings were not statistically significant when analysed by one-way Anova in conjunction with Tukey’s HSD post hoc test (p values higher than 0.05 at a significance level α = 0.05).

These low values demonstrate a fairly good colour stability to light exposure for all the tested coating materials. It would result, therefore, in a minimal contribution by the coating films to the overall colour changes in the wood coated surfaces, which strengthens the previous finding about the major contribution of the wood substrate to the total colour changes in the coated wood surfaces. However, the spectral properties of the coating films, meaning their specific capacity of absorbing, transmitting or reflecting/dispersing the UV-VIS radiation remain a key point in understanding and potentially controlling the colour changes in the substrate under the coating films and, finally, of the coated wood surfaces. These issues will be approached in future research.

3.5. Influence of Wood Species and Coating Materials on the Light-Induced Colour Changes in Wood Surfaces

The experimental results presented and discussed so far highlighted, in most cases, some protective effect of the three types of waterborne coatings (F1-acrylic-polyurethane, F2-acrylic, F3-polyurethane) against the UV-VIS light-induced colour changes in the wood substrates from the three wood species (maple, ash, walnut) under study. This protective effect was dependant on both the coating film and the wood species, which demonstrated different sensitivity to light-induced colour changes, in accordance with the previous research [22,24,53].

In order to have a cumulative and comparative image of the effect of clear coating on the colour stability of wood surfaces when exposed to light in indoor conditions, the data on the colour changes (∆E, ∆L*, ∆a*, ∆b*) after 72 h light exposure in the conditions of this research are cumulated in

Table 5. Comparative data of colour changes after 72 h light/UV exposure for uncoated and coated wood surfaces with the three finishing materials (F1, F2, F3).

Wood Species		Colour Changes after 72 h Light /UV Exposure
		Uncoated V1/Z3	F1 V2/Z1	F2 V2/Z1	F3 V2/Z1
Maple	∆E	9.36 A (1.71)	5.71 B (1.20)	5.35 B (2.11)	4.92 B (1.64)
	∆L*	−5.07 C (1.61)	−3.83 B (0.84)	−0.11 A (1.18)	−3.56 B (1.76)
	∆a*	1.48 B (0.79)	1.09 B (0.58)	−0.14 C (0.66)	2.60 A (1.45)
	∆b*	7.58 A (1.38)	4.03 B (1.49)	5.11 B (2.30)	−0.52 C (1.48)
Ash	∆E	8.39 A (1.28)	3.37 B (1.37)	3.94 B (1.24)	2.25 C (1.02)
	∆L*	−4.13 B (1.13)	−0.52 A (1.74)	−0.48 A (1.50)	−0.51 A (1.57)
	∆a*	1.43 A (0.46)	−0.06 B (0.93)	0.06 B (0.86)	0.49 B (0.97)
	∆b*	7.05 A (1.38)	2.65 B (1.49)	3.45 B (1.45)	0.24 C (1.55)
Walnut	∆E	6.20 B (1.34)	7.70 A (1.91)	6.31 AB (2.04)	4.74 C (1.40)
	∆L*	−3.25 C (1.62)	3.40 A (1.36)	0.96 B (2.60)	2.14 AB (1.15)
	∆a*	2.23 A (0.85)	1.14 B (0.63)	2.16 A (0.71)	1.61 AB (1.02)
	∆b*	4.42 B (1.38)	6.69 A (1.76)	4.96 AB (2.67)	3.24 B (2.16)

Note: Values on the same line sharing a common (^A–C) code as superscript are not statistically different (one-way Anova, α = 0.05, p > 0.05).

. The data for uncoated and coated samples (F1, F2, F3) of the three species were also included to allow a final discussion on the influence of both wood species and type of coating. Lower values in colour changes (∆E, ∆L*, ∆a*, ∆b*) mean better colour stability so that the effects of the above-mentioned influencing factors can be easily observed. The statistical significance of the differences between the mean values of the uncoated and coated specimens was tested by one-way Anova and Tukey’s HSD as post hoc test, and this information is included as superscript codes.

Based on the comparative ∆E values, it can be observed that all the three coating materials F1, F2, F3 protected to some extent maple and ash wood substrates against the light-induced colour changes. This protective effect was higher for ash than maple. In the case of walnut, some protective effect was registered only for the coating F3. No effect was registered for the coating F2, while the samples finished with the coating F1 suffered colour changes slightly more intense (statistically significant) than uncoated walnut wood.

Differences in the protective effects of various waterborne clear coatings have been reported before. For instance, an increased protective effect of a 2 k polyurethane topcoat (based on hydroxyl-bearing polyacrylate and aliphatic poly-isocyanate) compared to an acrylic topcoat was reported for Radiata pine boards primed with a moisture curing isocyanate [58]. Colour change values ∆E varied in the range of approximately 2–10 units after 700 h light exposure (accelerated aging test simulating natural light passing through window glass) for spruce samples coated with four different three-layer coating systems (primer, middle layer, top layer) with waterborne coatings based on acryl/alkyd or polyurethane resins and their mixture [59].

The coated maple and ash samples exposed to light darkened less compared to the uncoated ones, as reflected by the comparative negative ∆L* values for all the three finishing materials. In the case of maple, this effect appeared to be highest for the F2 coating material, while in the case of ash, all the three coatings seemed to have a similar effect. Contrarily, the coated walnut samples became lighter following 72 h light exposure, this effect being more pronounced for the coatings F1 and F3 and minimal for F2. This particular behaviour of walnut is due to its darker colour compared to the other two species, respectively, to the role of extractives and visible light in the occurring colour changes, as reported in the literature [18,20,21].

The statistical significance of wood species (1) and type of coating (2) as influencing factors on the light-induced colour change values ∆E was verified by an Anova two-factor with replication test, at a significance level α = 0.05. According to the summary results presented in

Table 6. Summary of two-way Anova results with replication of the colour change (∆E, after 72 h light exposure) in the coated wood surfaces (areas V2/Z1), considering the wood species (1) and the type of coating (2) as influencing factors.

Anova: Two-Factor with Replication (α = 0.05)
Source of Variation	SS	df	MS	F	p-Value	F Crit
Wood species (1)	266.8111	2	133.4055	52.71715	3.83802 × 10−18	3.055162
Type of coating (2)	77.40668	2	38.70334	15.29419	8.80515 × 10−7	3.055162
Interaction (1)x(2)	33.53909	4	8.384773	3.313366	0.012333886	2.430772
Within	387.1804	153	2.530591
Total	764.9372	161

, respectively, the p values lower than 0.05, it can be stated that both the wood species (factor 1) and the type of coating (factor 2) are statistically significant influencing factors. Comparing the F values of the two factors, it results that wood species (F = 52.72) influence more the colour changes than the type of coating (F = 15.29). The test shows also that there is a significant two-way interaction between the two factors (p < 0.05; F = 3.31). This would mean a common and interconnected influence of the two factors for each wood species and type of coating. A significant impact of wood species and type of lacquer/coating system, alongside pre-treatments of wood surfaces and modification of coatings, on the light-induced colour changes in wood surfaces was previously reported [59].

The interaction of the coating material with the wood substrate at the interface level, consisting in wetting and penetration, depending on both the structural features and surface properties of the substrate and the coating material, might be part of this significant two-way influence on the UV-induced colour changes determined in this research. According to the literature data, a combination of the wood anatomy and the ability of the coatings to flow into the wood capillaries, influenced by the viscosity, solids content and type of solvent, govern the degree of coating penetration [60]. Generally, waterborne coatings penetrate less than the solvent-borne ones, though large variations might be among different formulations. In this context, binder type, pigmentation, solids content and drying time were found to be important. Acrylic waterborne dispersions have been found to have a low penetration compared to alkyd emulsions [61,62].

Differences in the colour changes in clear-coated pine wood after an artificial weathering test were associated with different penetrations of an acrylic coating system (primer and topcoat), increased colour changes being explained as possibly due to a reduced coating penetration. Increased penetrations were obtained in the same research for beech wood, but no final conclusion was drawn on the influence of coating penetration on the colour changes caused by weathering [63].

An interesting study on the degradative behaviour of coated-wood surfaces exposed to artificial weathering using Raman microspectroscopy was recently published. Chemical imaging pointed out different artificial weathering degradation patterns between wood specimens (Japanese cedar) that were clear-coated with impregnating materials and film-forming materials. For the impregnating systems, the surface-distributed coating gradually decreased during irradiation and lignin in the top surface cell walls degraded similarly to uncoated wood [64]. The result is, therefore, that both good penetration and the formation of a top film of the coating are beneficial for improved UV protection.

The preliminary microscopic investigation of our samples highlighted such “composite coating layers” with total thicknesses varying in the range of approximately 25–75 μm, while the top layer film included varied in the range 5–20 μm, with maximum thickness in ash early-wood (vessels). From the waterborne coating materials employed in this research, the polyurethane lacquer (F3) was the most fluid and ensured the best penetration into the wood substrate, approximated to around 50 μm for maple, 30–60 μm for ash (with the high values in early-wood vessels) and 25–60 μm for walnut (large values in vessels). This might explain, at least partially, the best protective effect against light-induced colour changes demonstrated in this research by the F3 finish for all the three wood species tested. However, more research and complementary methods are necessary for a conclusion with this respect.

4. Conclusions

The paper investigated comparatively the light-induced colour changes for three hardwood species, namely, European maple (Acer pseudoplatanus L.), European ash (Fraxinus excelsior L.) and European walnut (Juglans regia L.), as well as the influence of finishing with three types of clear, colourless waterborne lacquers (F1: acrylic-polyurethane, F2: acrylic, F3: polyurethane) on their colour stability.

Colour measurements in the CIELab system in conjunction with an artificial accelerated ageing test under the action of UV-VIS radiation, simulating natural light passing through window glass, and two types of test samples were employed to highlight the influence and contributions of the wood substrate and of the coating films to the global colour modifications.

Coating films applied on 1 mm clear glass slides were employed as a sort of “detachable” finish for this purpose. A different transmittance of the UV and VIS radiation from the artificial source through this glass support and a decrease in light intensity reaching the substrate is a limitation of the experimental method employed, which has to be considered in future research.

Direct exposure to UV-VIS light caused visible colour changes for both uncoated and coated wood surfaces, the values of the calculated colour differences (ΔE) after 72 h exposure being dependant on both the wood species and the coating material, which were found to be statistically significant influencing factors in interaction.

Excepting two situations for walnut, statistically significant higher colour differences were obtained for the uncoated samples: maple (9.36 units), ash (8.39 units), walnut (6.20 units), compared to the coated ones: maple (4.92–5.71 units), ash (2.25–3.94 units), walnut (4.74–7.70 units), depending on the type of coating.

From the three coating materials tested, the waterborne polyurethane lacquer represents a good option for all the wood species. The effect of each coating material tested depended on the wood species, as a more powerful influencing factor. The interaction between the wood substrate and the coating material at the interface level needs further investigation.

The wood substrate underlying the coating film was found to bring the maximum contribution to the overall colour changes in the clear-coated surfaces, the actual ∆E values being influenced by the wood species and the coating film as significant factors. The colour changes ∆E in the wood substrates under the simulated “detachable” coating films varied in the ranges 4.44–6.02 units for maple, 2.97–4.43 units for ash and 4.51–5.93 units for walnut.

The coating films employed in this research demonstrated a fairly good colour stability to UV-VIS light exposure, with colour changes (∆E) up to only 1.30 units, with no statistically significant differences.

These results will be employed as a basis for further research looking at efficient methods to stabilise the natural colour of the clear coated wood surfaces against the effect of light in indoor applications, considering the specific contributions of the wood substrate and of the coating films, as well as the particularities of each wood species in relation to their chemical composition.