*2.8. Statistical Analysis*

The measured data of the adhesion strength were for individual groups of coated spruce samples evaluated on the basis of mean values and standard deviations. The e ffects of the HALS and BTZ UV-additives, added to coatings in concentrations (C) from 0 to 1 wt.%, were analyzed by the linear correlation "Adhesion = a + b × C". The e ffect of the outdoor weathering prolongation (τ) from 0 to 42 weeks was analyzed by the exponential correlation "Adhesion = a + b × exp (k × τ)". Within all correlations were determined the coe fficients of determination R2.

#### **3. Results and Discussion**

#### *3.1. Adhesion Evaluated at Artificial Weathering*

From the first experiment, connected with the artificial weathering of coated spruce samples, it is evident that the adhesion strength of the acrylic and alkyd coatings to the Norway spruce wood surfaces was not significantly influenced by the presence of the HALS and BTZ UV-additives in coatings in a range of concentration (C) from 0 to 1 wt.%. For example, for samples modified with plasma, the linear correlation "Adhesion = a + b × C" had in all cases a very small coe fficient of determination R<sup>2</sup> from 0.001 to 0.151 (Table 1). A similar result was obtained for the natural (plasma un-modified) samples.

**Table 1.** No e ffect of the hindered amine light stabilizer (HALS) and hydroxyphenyl-benzotriazoles (BTZ) used in the role of UV-additives on the adhesion strength between the coating (acrylic or alkyd) and the plasma modified Norway spruce wood—evaluation at none and 1 week artificial weathering—confirmed by the small coe fficients of determination R<sup>2</sup> from 0.001 to 0.151 for the linear correlations "Adhesion = a + b × C".


Notes: the mean value is from 6 measurements, i.e., from 3 replicates with a different mode of fungicidal pre-treatment (none, H3BO3, or benzalkonium chloride (BAC)), testing each replicate on 2 places. The standard deviation is in parentheses. The coe fficient of determination R<sup>2</sup> of the linear correlation (Adhesion = a + b × C) was determined from 24 values, i.e., 4 concentrations (C) of the UV-additive in the coating by 6 measurements.

The decrease in the adhesion strength between coatings and spruce samples due to the 1 week artificial weathering in Xenotest is documented by a drop in the parameter "a" in the linear correlations from 2.99–2.77 MPa to 2.38–2.23 MPa (Table 1). Analyzing the un-weathered samples, the acrylic and alkyd coatings—containing various amounts of the HALS or BTZ UV-additives—had a comparable

adhesion strength with the spruce samples, as is documented by comparable values of the parameter "a" in the linear correlations, i.e., 2.99 or 2.79 MPa for acrylic coatings and 2.84 or 2.77 MPa for alkyd coatings (Table 1).

The impact of the HALS and BTZ UV-additives on the adhesion strength between coatings and plasma modified spruce wood was also analyzed as by the complex linear correlations "Adhesion = a + b × C" (Figure 2). For both UV-additive types, a comparable parameter "a" (2.60 or 2.54 MPa) was determined, together with a very small and opposite parameter "b" (−0.123 or 0.122 MPa), and with very small values of the coefficient of determination "R<sup>2</sup> = 0.01" and *p* > 0.1. These results confirm that the adhesion strength of the acrylic and alkyd coatings to the Norway spruce wood samples was not influenced by the presence of the UV-additives in the coatings.

**Figure 2.** No effect of the UV-additives (**a**) HALS and (**b**) BTZ on the adhesion strength between the coatings and the plasma modified Norway spruce wood—confirmed by the small coefficient of determination R<sup>2</sup> equal to 0.01 and *p* > 0.1 for the linear correlations. Notes: the summary effect of UV-additives at the defined concentrations (C = 0, 0.25, 0.5, or 1 wt.%) is from 24 data measured for 6 replicates of 2 coating types "acrylic and alkyd" and 2 weathering modes "none and 1 week in Xenotest". The coefficient of determination R2 of the linear correlation (Adhesion = a + b × C) was determined from 96 values, i.e., from 24 values for one concentration of the UV-additive multiplied by 4 concentrations.

#### *3.2. Adhesion Evaluated at Outdoor Weathering*

The second experiment with the un-weathered and also with the 14 week, 28 week, and 42 week outdoor weathered coated spruce samples obtained other findings related to the adhesion strength.

The time prolongation (τ) of the outdoor weathering from 0 to 42 weeks caused in all cases (i.e., for three modes of fungicidal pre-treatment of wood; for two modes of plasma modification of wood; for two coating types) a significant decrease in the adhesion strength between the coatings and the wood surfaces, in summary from 3.01–2.12 MPa to 1.81–1.20 MPa (Tables 2–4, Figure 3). This knowledge was confirmed by the exponential correlations "Adhesion = a + b × exp (k × τ)" with high values of the coefficient of determination R2: (a) taking into account the different fungicidal pre-treatments of wood, the R<sup>2</sup> ranged from 0.903 to 0.997 (Tables 2–4), and (b) without taking into account the different fungicidal pre-treatments of wood, i.e., at the summary evaluation of the experiments when only the effects of coating type "acrylic or alkyd" and the plasma mode of wood modification "without or with" were taken into account, the R<sup>2</sup> ranged from 0.827 to 0.991 (Figure 3).



Notes: the mean value is from 16 measurements (2 UV-additives "HALS or BTZ"; 4 concentrations of each UV-additive "0, 0.25, 0.5 and 1 wt.%"; 2 testing places on replicate). This mode of evaluation could be used on the basis of the knowledge that the adhesion strength was not significantly influenced by the UV-additive, i.e., by its type and concentration in the coating (see Section 3.1; Table 1 and Figure 2). The standard deviation is in the parentheses. A significant drop in the adhesion at the prolonged time of weathering (τ) was evaluated by the exponential equation "Adhesion = a + b × exp (k × τ)" from 64 values, with R<sup>2</sup> always above 0.95.


**Table 3.** The significant negative e ffect of outdoor weathering on the adhesion between the coating (acrylic or alkyd) and the Norway spruce wood pre-treated with fungicide H3BO3 —subjected to or not subjected to plasma modification.

Notes: the mean value is from 16 measurements (see the first note in Table 2). The standard deviation is in the parentheses. A significant drop in the adhesion at the prolonged time of weathering (τ) was evaluated by the exponential equation "Adhesion = a + b × exp (k × τ)" from 64 values, with R<sup>2</sup> always above 0.95.

Another basic finding was obtained for the plasma modified spruce samples in the un-weathered state. Their adhesion strength with the acrylic and alkyd coatings was higher in comparison to the natural (plasma un-modified) spruce samples: i.e., the adhesion strength of the plasma modified wood was on average higher by 16.5% in testing the native chemically un-pre-treated wood (2.785 to 2.39 MPa), by 26.6% in testing the H3BO3 pre-treated wood (2.835 to 2.24 MPa) and by 13.6% in testing the BAC pre-treated wood (2.845 to 2.505 MPa)—(Tables 2–4). The plasma modification of the wood surfaces had a more positive e ffect on the adhesion strength of the acrylic coatings, increased by 20.5% (from 6.5 to 7.83 MPa), than the alkyd coatings, increased by 5.5% (from 7.77 to 8.2 MPa). This result can be explained by the better wettability of the hydrophilic plasma modified spruce surfaces with more polar water-borne acrylic coatings (see Section 3.3). However, as a result of the outdoor weathering, the positive e ffect of plasma on the adhesion strength disappeared in most cases (Tables 2–4), which was more apparent for samples painted with acrylic coatings, which manifested less durable to sunlight, water and other weathering agents (Figure 3).



Notes: the mean value is from 16 measurements (see the first note in Table 2). The standard deviation is in the parentheses. A significant drop in the adhesion at the prolonged time of weathering (τ) was evaluated by the exponential equation "Adhesion = a + b × exp (k × τ)" from 64 values, with R<sup>2</sup> above 0.95 or 0.90.

**Figure 3.** The summary evaluation—negative e ffects of the outdoor weathering prolongation (from 0 to 42 weeks) on the adhesion strength between the acrylic (**<sup>a</sup>**,**b**) or alkyd (**<sup>c</sup>**,**d**) coatings and the plasma modified spruce wood (**<sup>a</sup>**,**<sup>c</sup>**) or the natural spruce wood (**b**,**d**). Notes: each point in the graphs represents the mean value of the adhesion from 16 measurements—performed on 2 places of 8 di fferent replicate types (2 UV-additives used in 4 concentrations). The coe fficients of determination R<sup>2</sup> for the exponential equation "Adhesion = a + b × exp (k × τ)" (evaluating 12 mean values, i.e., 3 fungicidal pre-treatments: ∇ native wood + H3BO3 pre-treated wood, • BAC pre-treated wood, multiplied by the 4 outdoor weathering times) were high, from 0.827 to 0.991—confirming they have significance.

The presence of fungicides H3BO3 and BAC in spruce samples did not a ffect the adhesion strength of acrylic and alkyd coatings to their surfaces. In the complex evaluation—taking into account two coating types, two modes of wood modification with plasma, and the four outdoor weathering times—the average adhesion strength was comparable for the native wood (1.853 MPa), the wood pre-treated with H3BO3 (1.837 MPa), and the wood pre-treated with BAC (1.947 MPa) (averages calculated from Tables 2–4).

Comprehensively, comparing the adhesion strength of acrylic and alkyd coatings to spruce samples—i.e., for the four outdoor weathering times (0, 14, 28, 42 weeks), taking into account three modes of fungicidal pre-treatment of wood and also two modes of plasma modification of wood surfaces—the adhesion strength of acrylic coatings was partly lower (1.74 MPa) than alkyd coatings (2.016 MPa) (averages calculated from Tables 2–4). The adhesion strength of tested coating types comparably worsened with the time prolongation of the outdoor weathering. For the coated natural (plasma un-modified) spruce samples, the adhesion decreased less using acrylic coatings (by 38.92%, from 2.166 to 1.323 MPa) than using alkyd coatings (by 44.13%, from 2.59 to 1.447 MPa). On the contrary, for the coated plasma modified wood, the adhesion strength when using the acrylic coatings decreased more (by 55.67%, from 2.91 to 1.29 MPa) than when using the alkyd coatings (by 38.22%, from 2.73 to 1.687 MPa)—(Figure 3, and Tables 2–4).

#### *3.3. Opinions of Some Researchers on the Adhesion of Coatings to Wood Surfaces under the Influence of Modification and Weathering Impacts*

Several studies have reported that the adhesion strength between an individual coating type and an individual wood species depends first of all on the chemical and physical characteristics of the contacting materials and the aging conditions of coated wood products.

Moya et al. [14] found that the addition of TiO2 nanoparticles, used as an UV-additive, to a water-based coating in the amount of 1% or 1.5% did not significantly a ffect the adhesion strength of the modified coating to seven (from nine tested) tropical wood species. This result is in line with our results concerning the un-significant impact of HALS and BTZ UV-additives in coatings on the adhesion strength (see Section 3.1).

Rehn et al. [39], Acda et al. [41] and Wolkenhauer et al. [55] found that the initial activation of wood surfaces with the atmospheric pressure plasma improved the paint adhesion and bondability by up to 30%. Their conclusions are in accordance with our results related to the water-borne acrylic coatings (see Section 3.2).

Riedl et al. [51] demonstrated that with prolonging plasma modification of maple sugar wood, its wettability (dispersive adhesion) improved more than its adhesion strength with the UV-cured polyurethane/polyacrylate coating. This is because, in addition to the free surface energy (dispersive adhesion), a number of other phenomena drive the adhesion dynamics, including surface roughness (mechanical adhesion), the formation of chemical bonds (chemical adhesion), presence of electrical discharges (electrostatic adhesion), and solubility of the coating (di ffusive adhesion). Jablonský et al. [45] recommend applying water-borne coating systems on wood surfaces right after the plasma treatment, when the activated wood surface with a higher portion of polar functional groups adopts the highest degree of hydrophilicity. This is in accordance with experiments performed by Král et al. [47], where the X-ray photoelectron spectroscopy study confirmed that in beech wood surfaces were created polar functional groups at exposure to atmospheric pressure plasma, which have a positive e ffect on the wettability.

Avramidis et al. [56] found a positive e ffect of plasma on the adhesion between polyvinyl acetate (PVAc) glue and beech wood pre-treated with synthetic waxes or montan-ester waxes, because a force in the sensitive peel test increased by about 50–60%. This result is not entirely in agreemen<sup>t</sup> with the results of our experiment, where the plasma modification of spruce wood pre-treated with fungicide not unequivocally increased the adhesion of acrylic and alkyd coatings (see Tables 2–4). In the un-weathered state of the coated samples, the combined e ffect of fungicide and plasma on the

adhesion strength was positive with the application of water-borne acrylic coatings (increased by 2.8% for boric acid, or by 6.7% for BAC); however, with the application of alkyd coatings, it was either negligibly positive (increased by 0.7% for boric acid) or even negative (drop by 2.55% for BAC). The different results achieved by us and by [56] can be explained by the non-polar characteristics of waxes and vice versa by the high polarity of fungicides used in our work; however, some other factors could play a role as well.

Moghaddam et al. [49] also searched the combined "chemical substance and plasma" effect on the wettability characteristics of wood surfaces. They created a thin transparent superhydrophobic layer on pine veneer surfaces from TiO2 with the liquid flame spray technique and following the deposition of plasma polymerized perfluoro-hexane or hexamethyldisiloxane. The hydrophobic effect of this approach was stable even after repeated wetting cycles and was higher compared to that of plasma modified surfaces.
