**1. Introduction**

Oak (*Quercus petraea* L.) wood is often used for the exterior applications, mostly in construction of bridges, pergolas, balconies or garden furniture, where higher natural durability [1] is required. Oak contains a relatively high amount of phenol extractives, mainly vescalagin, castalagin, gallic and ellagic acids [2], creating problems in the field of surface treatment durability [3,4]. Tannins in oak wood also retard coating hardening [5]. The complex open vessel morphological structure of oak wood complicates the overall application of coatings. A photodegradation process of oak wood accompanied with significant discolouration and leaching of extractives from the surface takes place during the initial phases of outdoor exposure [6], more intensely in the heartwood zone [7], which leads to the need to protect oak wood by coatings to maintain its natural appearance.

Exterior wood coatings are used to improve the properties of substrate wood [8], reduce the e ffects of degradation factors [9–11] and prolong the service life of the material. The exterior coating generally protects against moisture uptake and related dimensional changes, protects against photochemical

degradation, and prevents microbiological degradation [12,13]. The problems with transparent coatings have been well discussed by Evans et al. [13]. Their advantage is the ability to protect wood and preserve the natural look and colour [13], but they have the disadvantage of not protecting the substrate wood against UV and visible light radiation as well as pigmented coatings [14]. The di fferent performance of coatings is also caused by their polymer base [15], the type of solvent [16] or even by the underlying wood species [17,18]. Coatings protect underlying wood, but they are exposed to weathering process causing their degradation [17,19,20]. The durability of coatings against atmospheric degradation is assessed via natural weathering (NW) or artificial weathering (AW) tests [8,21–25] with parameters given in international standards. The older and more common method is natural weathering [26], which provides reliable results of coating durability due to the synergistic action of outdoor factors. Accelerated artificial weathering [27] is carried out in laboratory conditions, to simulate the exterior environment [22,28]. Artificial weathering can significantly reduce the required testing time, however its reliability is still often questioned. Correlations between weathering methods were done in several studies [23,29–31] however the results were ambiguous. Valverde and Moya [32] developed a model to predict total colour di fference between natural weathering and accelerated weathering for di fferent kinds of finishes of three tropical species. Cogulet et al. [11] focused on how the impacts of di fferent weathering methods challenges the reliability of AW and states that it is necessary to test coating systems in an end-use environment for accurate assessment of their likely performance.

In these works, where wood weathering was studied, more qualitative parameters of wood coatings during both exposures were evaluated–change of colour parameters [3,11,30], coating thickness [23,30], coating adhesion [30,33] or gloss [3,34]. Especially the change of colour during weathering serves as a basic indicator of the rate of degradation [18,22,30]. In a study of Moya et al. [30], colour change was higher in all species after NW than after AW due to the constant variation of solar radiation, moisture, water, air contamination and biotic agents, which accelerates colour degradation processes [8,35]. These findings are consistent with previous studies indicating the di fficulty of reproducing the synergistic action of NW factors during AW [11,17,23,24,36].

Currently there is very limited information on characteristics of transparent and pigmented coatings on oak wood when they are exposed to both natural and artificial weathering. Therefore, the objective of this study was to compare the performance of eight di fferent transparent and pigmented coating systems applied on oak samples using natural and artificial weathering tests. The e fficiency of specific coating systems was determined by measurements of colour, gloss and surface wettability changes and by regular visual macroscopic and microscopic evaluation of the samples.

#### **2. Materials and Methods**

#### *2.1. Sample Preparation, Coatings and Weathering Process*

Samples of oak (*Quercus petraea* L.) wood harvested in the Czech Republic having an average oven dry density of ρ0 = 705 kg/m<sup>3</sup> [37] were used for the experiment. The samples were conditioned to 12 ± 2% moisture content. Test samples were prepared from the heartwood zone and they were visually sorted in order to minimalize the colour variability of the tested wood materials. The dimensions of the samples were 375 mm × 78 mm × 20 mm for natural weathering and 45 mm × 45 mm × 8 mm for the artificial weathering tests. Tangential surfaces were exposed to weathering in both cases. Two samples for NW and four samples for AW for each type of treatment were used.

Eight di fferent transparent or pigmented coatings were applied to the samples based on the producer's recommendation. Their specification and application details are given in Table 1. One group of samples was left untreated as control samples to compare coatings performance on treated samples. The cross ends of samples were sealed with silicon to minimize additional water uptake.


**Table 1.** Specifications of the tested coatings according to the producers.

Note: IPBC = 3-iodo-2-propynyl butylcarbamate.

The natural weathering (NW) test was performed at Suchdol, Prague (50◦0749.68" N, 14◦2213.87" E) for 12 months. The climatic conditions during exposure are given in Table 2. The samples were exposed at a 45◦ inclination, facing south, and placed approximately 1 m above the ground according to the procedure previously described [26].


**Table 2.** Climatic conditions during NW. use a period (.) for decimals.

Note: based on the data from http://meteostanice.agrobiologie.cz [38].

The artificial weathering (AW) test was performed in UV-chamber QUV (Q-Lab, Cleveland, OH, USA) according to a modified EN 927-6 method [27]. The total time consisted of six cycles (1008 h) of weathering in the UV chamber and 36 h of temperature cycling. During the each weekly cycle of irradiation and spraying, the samples were transferred to a Discovery My DM340 conditioning chamber (ACS, Massa Martana, Italy) and exposed to three cycles each lasting 2 h of temperature changes from −25 to +80 ◦C (with 25% RH). The alternation of UV radiation, spray, and low temperature cycles, leading to a better imitation of the exterior conditions in Central and Northern Europe, was previously used by Van den Bulcke et al. and Pánek et al. [22,39].

#### *2.2. Colour Change (*Δ*E\*) Test*

Colour variations of the specimens were evaluated through natural and artificial weathering exposure of oak samples with 8 different coatings. The colour parameters L\*a\*b\* [40] of the test specimens were measured after 0, 6 and 12 months of NW and after 0, 1, 3 and 6 weeks of AW using CM-600d Spectrophotometer (Konica Minolta, Osaka, Japan). For the observation of reflection, the specular component was included (SCI mode) at a 10◦ angle and d/8 geometry with an illumination standard of D65 (corresponding to daylight in 6500 K). Six measurements of each tested sample were carried out for each weathering time. Colour changes evaluations were done in CIE L\*a\*b\* colour space where L\* is lightness from 0 (black) to 100 (white); a\* is chromaticity coordinate + (red) or − (green); b\* is chromaticity coordinate + (yellow) or − (blue).

The total colour difference ΔE\* [40] was subsequently calculated from relative changes of colour (ΔL\*, Δa\*, and Δb\*) using Equation (1):

$$
\Delta \mathbf{E}^\* = \sqrt{\left(\Delta \mathbf{L}^\*\right)^2 + \left(\Delta \mathbf{a}^\*\right)^2 + \left(\Delta \mathbf{b}^\*\right)^2} \tag{1}
$$

#### *2.3. Gloss Change (*Δ*G\*) Test*

The gloss of the different coatings before and during weathering tests was measured using MG268-F2 glossmeter (KSJ, Quanzhou, China) on the basis of [41]. Six measurements at a 60◦ angle per sample after 0, 6 and 12 months of NW and after 0, 1, 3, and 6 weeks of AW were carried out to evaluate gloss changes of the samples.

#### *2.4. Surface Wettability Change (*Δ*W\*) Test*

The sessile drop method with static contact angle measurement was performed using a Krüss DSA 30E goniometer (Krüss, Hamburg, Germany) with the methodology used in previous studies [42,43]. Twenty measurements were taken for each sample, with distilled water drops with a dosing volume of 5 μL. The contact angle values were determined after 5 s of drop deposition on surface of the sample before weathering and after 0, 6 and 12 months of NW and after 0, 1, 3 and 6 weeks of AW.

#### *2.5. Macroscopic and Microscopic Evaluation*

Tested surfaces of the samples were regularly macroscopically evaluated using a Canon 2520 MFP scanner with 300 DPI resolution (Canon, Tokyo, Japan). Creations of cracks, defoliation of coating systems were visually analysed. Microscopic structural changes of coatings and surface of the samples, creation of ruptures, 3D-images of surface profiles were also studied employing confocal laser scanning microscope Lext Ols 4100 (Olympus, Tokyo, Japan) with 108-fold magnification.
