Microscopic assessment of the effect of saw-textured Pinus radiata plywood surface on the distribution of a film-forming acrylic stain.

The micro-morphology of the wood-coating interface of saw-textured Pinus radiata plywood panels, which had been coated with a film-forming stain, was examined using a range of microscopic techniques. The surface appearance and tissue deformation characteristics of saw-textured plywood and the pattern of distribution of the applied coating were observed by combined light, confocal, and scanning electron microscopy. The saw-textured surface was highly irregular, with surface tissues greatly distorted, being bent, twisted, and occasionally detached, and showing extensive cell wall cracking. The applied coating conformed closely to the contours of the saw-textured surface, resulting in a highly variable film thickness. The film was thickest in the grooves and thinnest over the ridges, and in places discontinuous, particularly where wood cells protruded out of the wood surface. These and other salient features of the surface of saw-textured plywood are discussed in relation to the pattern of the applied acrylic stain topcoat.

Keywords: Pinus radiata, plywood, film-forming stain, microscopy, surface texture, film thickness, acrylics, semi-transparent stain, waterborne, architectural, coatings-substrate interface, wood

**********

Wood in exterior service has to be protected from the damaging effects of sunlight, rain, and microbial colonization, and the use of surface coatings can be an effective way of combating these elements. The selection of a particular coating is dependent on several factors, including physical and chemical properties of the coating, wood species and composition, surface texture, and environmental service conditions. Once a coating has been selected for its ability to perform in service, its performance is largely determined by film build (film thickness), coating adhesion, and weathering effects. The chemical and physical properties of wood are species specific, although within-species and even within-tree variations can occur. In a study of the exterior weathering of commercial coatings on pine sapwood in Germany and New Zealand, significant differences in performance relating to coating, species, site, and exposure were presented. (1) It was reported that coating type and site had the most significant effects, while species had the least. It was suggested that substrates from more genetically diverse species might produce performance differences.

It is a common practice to machine wooden surfaces prior to the application of coatings. Surfaces can be planed, sanded, or saw-textured. For exterior plywood siding, saw-textured plywood is widely used, primarily because it improves aesthetic appeal while overcoming plywood surface checking. Also, the cost of panel production is competitive, particularly where panels are supplied to the market as prefinished products. However, because of the highly irregular and weakened surfaces, due to mechanical damage of wood tissues, it may be more difficult to obtain optimal coverage and film-build on rough plywood surfaces. Therefore, the interaction of various coatings with rough textured wood surfaces needs to be examined using appropriate methods in order to facilitate the selection of coatings with suitable spreading properties and methods for their application.

Several studies have examined the effect of abrasive machining on morphological and structural characteristics of wood cells, (2-5) mainly with a focus to relate the features of abrasively machined wood to wood-adhesive bonding and adhesive performance. Studies by North American researchers have investigated the influence of rough surface texture on the weathering performance of coatings (finishes). (6-12) However, there is little information available at the microscopic scale that can provide an understanding of the interaction of coatings with wood surfaces featuring a rough texture. The wood-coating interface for varnished timber is an important region from the point of view of coating adhesion to wood, as this is the region where detachment of such coatings is initiated because of photodegradation of the outermost wood tissues. (13-23) For film-building stain and pigmented coatings, highly variable coating thickness can arise from surface attributes such as tissue damage. In the present work, we employed a combination of light microscopy (LM), confocal laser scanning microscopy (CLSM), and scanning electron microscopy (SEM) to understand how a rough textured plywood surface affects the distribution of a film-forming semitransparent acrylic stain. The primary focus was to check the film build of the topcoat and its surface characteristics and then consider possible improvements in manufacturing processes. The microscopic techniques used in the work described here revealed a complex nature of coating distribution associated with the rough plywood surface. This study does not contain any exterior weathering assessments. Such work has been initiated and will be reported subsequently, using a combination of microscopic techniques to assess weathering impacts.

EXPERIMENTAL

Coated Panels

Coated Pinus radiata (radiata pine) plywood panels were supplied to us by a New Zealand coated wood products company. This company obtained freshly manufactured plywood. It profiled and band-sawed them to produce a saw-textured surface, strapped them together in packets ready for preservative treatment, and dispatched them to a treatment plant where they were treated with a light organic solvent preservative (tributyltin naphthenate). Nondurable timbers such as radiata pine need to be preservative-treated to the decay hazard level H3.1 (New Zealand Standard NZS 3640:2003, Chemical Preservation of Round and Sawn Wood). Once treated and returned to the company, the plywood panels would typically have a residual white spirit solvent level of about 35 L/[m.sup.3] of plywood. The plywood panels were then filleted and the solvent flashed-off. Each panel was held vertically and banged on its edge and then brushed to remove any loose material on the panel surface.

The panels were then coated in the factory with a proprietary, semitransparent acrylic stain system in the following manner. One coat of acrylic primer was spray applied, as was the first acrylic topcoat. No back brushing was performed after either spray application. The second acrylic topcoat was brush applied. Details of the coating composition and coverage rates are given in Table 1.

Light Microscopy

Sliding microtome sections, 90 [micro]m thick, cut perpendicular to the surface, were taken from the coated panels in the region of the wood-coating interface. The staining of sections was performed in the following manner. All sections were stained with a 0.5% aqueous solution of Toluidine Blue to enhance the contrast of the wood tissue. To examine the depth of penetration and the pattern of distribution of the primer, sections were also stained with a 1% Sudan IV solution prepared in 95% ethanol. Because Sudan IV stained both the primer and topcoat, we included panels with primer applied only and those with both primer and topcoat applied, for comparative purposes. The resulting turquoise color of the wood cell walls contrasted well against the natural brown color of the coating in sections that had been stained only with Toluidine Blue, and the red color of the primer facilitated visualization of the primer. Stained sections were mounted in glycerol on a glass slide and examined with a Zeiss Photomicroscope II after placing a cover glass over the sections.

Confocal Laser Scanning Microscopy

Similarly prepared and stained sections as for light microscopy, and which had been stained with Toluidine Blue, were examined with confocal laser scanning microscopy. The sections were imaged with a Leica TCS/NT CLSM. Confocal images were acquired using an argon/krypton laser with excitation wavelengths of 568 and 647 nm and emission wavelengths of 600 and 660 nm. A 16x multi-immersion lens with a numerical aperture of 0.5 was used for all images. Projection images produced from 0.9 [micro]m sequential optical sections, through 12-20 [micro]m thicknesses, were captured and stored for analysis. Wood cell walls were colored strong pink and the coating appeared in shades of brown, enabling the coating material to be clearly differentiated from wood cell walls.

Scanning Electron Microscopy

Uncoated surfaces of small plywood blocks with planed and saw-textured faces were mounted on aluminum stubs with adhesive carbon discs, coated with gold in a sputter coater, and examined with a Cambridge Stereoscan 240 SEM, operating in the secondary electron detection mode.

Correlative Microscopy

For correlative microscopy of coated surfaces, the same sections were examined using LM, CLSM, and SEM. This allowed for a comparison of the three different types of microscopy for their suitability to resolve the highly distorted surface tissues in order to understand their interaction with the applied coating.

Sections were prepared and stained with Toluidine Blue, as described in the Light Microscopy section. The same sections were examined sequentially using LM, CLSM, and SEM, in that order, after being mounted on glass slides or stubs, as appropriate, and imaging the same wood-coating region. The procedures for LM and CLSM were the same as described earlier. For SEM, the only difference was that the samples were coated with carbon (instead of gold) in a vacuum coater and examined using a backscattered detector.

RESULTS AND DISCUSSION

SEM micrographs provided a comparison of two different types of surface textures produced on plywood by planing (Figure 1A) and band-sawing (Figure 1B). The information obtained was a useful complement to the results obtained by light microscopy of sectional views, which provided valuable comparisons between early- and latewood, illustrating the extent to which grain-raising and distortion had occurred due to band-sawing. Ten randomly selected spots on each surface type from uncoated veneers were imaged. The illustrations included in Figure 1 are representative of the images acquired. The planed surface is relatively smooth, with minimal grain raising and distortion of surface tissues. Wood cells (tracheids) are largely oriented in their original direction. In sharp contrast, the saw-textured surface is irregular. Wood grains exhibit highly raised ridges in places, with tissue masses being often dislocated, resulting in a wavy appearance of the plywood surface. Some pockets of tissues are severely distorted, with a grossly changed orientation.

The following information on the characteristics of the wood-coating interface is based entirely on an examination of the two topcoats of the stain. The pattern of distribution of the primer is described only briefly, because the emphasis of our work was on coating distribution (film thickness) on the saw-textured plywood surface and the fact that the primer did not form a film.

Light microscopy of transverse sections taken through the coated wood proved valuable in observing relatively large areas at low magnification, providing important information on the distribution of primer and enabling useful comparisons to be made of ridges, grooves, and moderately flat regions of the saw-textured surface in relation to the distribution of topcoat. Ten randomly selected regions of coated saw-textured plywood were sectioned perpendicular to the surface and examined by LM. The illustrations provided are representative of the images acquired from all sections examined.

The primer applied to panels was not visible under the light microscope and staining with a suitable dye (Sudan IV) was necessary to get an estimate of the depth of penetration into wood tissues from the surface. Panels coated with only the primer as well as those coated with primer plus topcoat were sectioned and examined for this purpose. It is apparent from the illustrations presented in Figures 2A, B, and C that the primer penetrated into the surface and subsurface tissues to a depth of 2-10 cells, filling cell lumina in both early- and latewood tissues (Figures 2A, B), as well as the damaged parts of cells. However, cell walls do not appear to be penetrated except in the regions where gaps have developed due to cracking. Absence of a film on the panel surface with only the primer applied (Figure 2C) is a clear indication that the primer is a non-film forming formulation. The primer appears to be present only in the cell lumen and there are no indications of cell wall penetration.

Low magnification LM views of the wood-coating interface (Figures 3 and 4) show a wide range of tissue distortions in early- and latewood surface layers of the plywood, reflecting the severity of damage resulting from the "plowing" and "ripping" actions of band-sawing, with damage being more severe for earlywood than latewood.

A wide range of surface morphologies in the earlywood regions of the saw-textured coated plywood, as well as variations in the film thickness of the topcoat, are illustrated in Figures 3A-E Figure 3A shows a relatively flat region of the plywood, although curving of the outer tissues has changed their original radial alignment to some degree. The coating film in this region is relatively uniform in its thickness but contains prominent air bubbles, including one close to the outer margin of the coating. The plywood surface shown in Figure 3B has a wavy appearance, reflecting perhaps more severe damage as compared to the surface illustrated in Figure 3A. Although the thickness of coating film is variable, the veneer surface is well covered by the coating. The damage to the surface of the saw-textured plywood in the earlywood region was generally more severe than illustrated in Figures 3A and B, and a range of morphologies of damaged outer tissues is shown in Figures 3C-F.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

In Figure 3C, the "plowing" action of band-sawing resulted in localized detachment of a large tissue mass, and its displacement relative to the parent tissues (parent tissues refer to the underlying tissues to which distorted surface tissues are connected, loosely attached, or once belonged) and the tissue damage may be related to a large crack in this region that has developed at an acute angle relative to the surface. Elsewhere, small pockets of surface tissues are torn because of the "ripping" action of band-sawing. In Figure 3D, a large mass of outer tissues has detached from the parent tissues along a deep crack that developed at a shallow angle relative to the plywood surface. The coating film is irregular in thickness and contains a large open air bubble within a groove. Figure 3E shows an effect of the "ripping" action of band-sawing which has resulted in a surface that is highly irregular, containing variously shaped voids among loosely attached pockets of tissues. The coating film is irregular in its thickness, being thinnest over a detached mass of tissues. An extreme example of the "ripping" action of band-sawing is shown in Figure 3F where masses of surface tissues are torn all along the surface and are detached from or loosely connected to underlying parent tissues. This figure also shows extreme variability in the thickness of coating film.

Latewood tissues are not completely resistant to the damaging effects of band-sawing, although the extent of distortion to wood tissues appears to be considerably less in comparison with earlywood. The range in variability in the surface morphology of band-sawn plywood in the latewood region is illustrated in Figure 4. In Figure 4A, the plywood surface is relatively flat. Only small masses of tissues are torn, resulting in a slightly irregular surface. However, the coating is not uniform, the film being thicker over the grooves relative to the torn protruded tissue masses. In Figure 4B, a large mass of tissue is slightly raised above the surface and is associated with a diagonal crack that has penetrated into the subsurface layers. The thickness of the coating film is variable.

As can be seen in the illustrations presented in Figures 3A-F, the thickness of coating film is highly variable--from a relatively uniform thickness (Figure 3A) to displaying extreme thickness ranges over a relatively small region of the veneer surface. In relatively flat surface regions of the saw-textured plywood, where tissue distortion is much less, the coating film is relatively uniform in thickness over a large area (Figure 3A). The film is thinnest in the apical regions of ridges composed of loose tissue masses and thickest in the grooves. The planed and coated plywood is not illustrated because the appearance of surface tissues and the distribution of coating were similar to those of the relatively flat regions on saw-textured plywood, as indicated by the arrowheads in Figure 3A.

Because of the high variability of coating thickness, dry film coating thickness measurements were carried out at randomly selected areas representing grooves (film thickness up to 200 [micro]m), ridges (0 to 20 [micro]m), and the uniformly flat regions (50 to 70 [micro]m). Where air bubbles were present in the coating film, effective film thickness was reduced to between 10 and 40 [micro]m.

LM proved unsuitable to view sections at higher magnifications, as it was not possible to obtain the entire wood-coating complex of the surface region of saw-textured plywood in the same focal plane, particularly where surface tissues were greatly distorted. CLSM proved to be an ideal tool for this purpose, providing high definition images of both distorted surface tissues and the pattern of coating distribution. A combination of the information contained in the composite image prepared from 20 sequential optical sections, and contrasting colors of wood tissues (pink) and the coating (brown), is the key factor for achieving a high image resolution as shown in Figure 5. The cell walls protruding beyond the outer margin of the coating film are in the same focal plane as the tissues underneath the coating, and thus are sharply defined. The bulk of coating fills the groove at the base of the tissue ridge and a large crack alongside it, filling spaces among loosely connected cells. Only a thin layer of the coating material covers the ridge. The coating material is also present in the lumina of surface and subsurface cells, the cell walls of which are extensively cracked, containing both trans- and intrawall fractures. Air bubbles are present in the coating film (Figure 5).

[FIGURE 3 OMITTED]

Correlative LM, CLSM, and SEM imaging of the same transverse section taken through the coated saw-textured plywood afforded a direct comparison of the three different types of microscopy for their suitability in viewing the wood-coating complex. Images illustrated in Figure 6 were obtained using LM, CLSM, and SEM of the same transverse section through coated saw-textured plywood. Figure 6A provides a low magnification LM view of large masses of surface tissues, which are greatly distorted because of the "plowing" action of the band-saw, as the presence of large cracks at the bases of these tissue masses would indicate. Cell walls of the distorted tissue masses are not clearly resolved. In particular, the tissue mass indicated by the arrow in Figure 6A appears very fuzzy. In the CLSM view of this fuzzy tissue mass (Figure 6B), individual cells are discernible and the presence of coating in cell lumina and cell wall cracks is resolvable. Coating is also present in the large crack all along the region where the surface tissue mass is loosely connected to underlying parent tissues. The SEM micrograph (Figure 6C) illustrates the severely distorted tissue mass shown in Figures 6A and B, and also the features of other surface tissues. Cells in all surface regions are resolvable, with a clear distinction between cell walls and cell lumina. The coating film is clearly identifiable, although detection of coating in cell lumina is marginal. The differentiation between the coating and cell walls is not as pronounced as in the CLSM image shown in Figure 6B. Additionally, there are indications of electron beam damage to tissues in the section, as indicated by a large crack formed along one of the rays and the presence of marked cell wall delamination.

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

It is apparent, from the information obtained from the three different types of microscopy used, that the combination of LM and CLSM proved most valuable. While LM alone was satisfactory for examining the depth of penetration of the primer into plywood tissues, when scrutinizing surface features of plywood and the distribution of the topcoat, the combination of LM and CLSM provided information that was complementary and which could not be obtained using one form of microscopy without the other. The ability to scan sections rapidly with LM at a low magnification, where depth of focus was of little consequence, proved vital for examining and comparing surface irregularities in early- and latewood regions of plywood and associated variability in the thickness of coating over relatively large areas. However, despite a satisfactory differentiation in contrast achieved between the natural brown color of the coating and the turquoise color of wood cell walls after Toluidine Blue staining, LM proved unsatisfactory for examining the wood-coating interface regions at high magnifications because of its inability to bring the distorted surface wood tissues, underlying parent tissues, and the coating film into the same focal plane. The use of CLSM readily overcame this difficulty, as it became possible to bring into focus significantly large parts of the entire wood-coating polymer composite. The ability to form a composite image of a series of successive optical sections, free of any distortions, is a unique capability of CLSM, which enables one to obtain information from relatively thick biological tissues.

In our work, we examined the composite images made from optical sections taken through depths of 12 to 20 [micro]m from the surface of sections, which made it possible to resolve the details of even the most highly distorted surface tissues. Also, reasonably sharp color differentiation between the coating and the wood enabled the distribution of coating and features of wood tissues associated with the coating to be clearly examined. Although the distorted surface tissues were clearly resolved with the SEM because of its superior depth of focus, it did not provide any additional information beyond that achieved by CLSM. Besides, SEM sections tended to split along the rays and differentiation between the coating and wood tissues was inferior to both LM and CLSM, despite the use of the backscattered mode, which has been widely used in composite image studies to differentiate components based on their atomic number differences. However, it is possible that field emission SEM fitted with a high-resolution backscatter detector and with the capability of providing high definition images at low kVs will prove more useful.

Keeping in mind that coating formulation and the method of coating application are also important contributing factors to nonuniform coating thickness that warrant attention, it is clear from the LM and CLSM images illustrated that coating distribution is greatly affected by the surface irregularities of saw-textured plywood. On one hand, numerous cell wall cracks, formed from band-sawing, appear to have facilitated greater coating penetration into surface and subsurface tissue layers, which may be considered a positive factor with regard to coating adhesion to wood, as coating penetration-assisted physical anchoring is one of the key factors that influence coating adhesion to wood. (24-28) On the other hand, saw-texture-induced disproportionate deposition of the coating, with consequent irregularities in the thickness of the coating film, will no doubt have an undesirable effect on coating performance. Potential issues with regard to the durability of the applied coating are that the film was very thin over ridges, bubbles were common in the coating film, and the coating did not cover cell walls when fibers protruded sharply from the surface.

[FIGURE 6 OMITTED]

The coverage rates of the stain topcoat on plywood in this study are slightly higher than those of Williams and Feist, (10) presumably due to the use of a primer in our study and application of more than one stain coat. When two semitransparent stain coats were applied on saw-textured yellow poplar, (11) the coverage was significantly less than with our primer coat, but the coverage of their second coat was comparable to ours. The coverage rates of two semitransparent stain coats on sanded yellow poplar surfaces were significantly higher than on saw-textured surfaces. (11) While these two exposure studies both reported better durability for saw-textured surfaces than smooth surfaces, due to higher stain loading on the surfaces of the saw-textured panels, they did not provide a microscopic description of the surfaces. Such a description would have provided a very useful bridge to the present work where discontinuous morphological features were observed in the surface tissues of plywood in parallel with the variation in coating film thickness. Since the semitransparent stain coverage rates for the studies are comparable, this suggests that gross coverage values are typical and satisfactory. The application method for the other two studies in question was by brush, while, for the present study, the primer and the first stain topcoat were sprayed on without back-brushing (unless shadowing occurred on airless spraying, resulting in regions of clearly different coverage). The mechanical action of back-brushing along the grain may result in a laying down, to some degree, of fibers protruding from the surface. The final coats in all studies were brushed on and this affords a final opportunity to lay down upstanding fibers. Whether or not the absence of back-brushing was a causal factor in the extent of upstanding fibers cannot be determined from the data presented.

Another contributor to the poor finish in the present study may relate to the impact of the saw-texture on the spread of the topcoat, and thus its thickness. The gross textural effects, demonstrated microscopically in this article, highlight the fact that abrupt discontinuities as well as more gradual local surface changes are present and are captured in the variable film thickness. Back-brushing after spraying may assist in the reduction of the variation in film thickness with the less abrupt textural features.

Stain formulation is the third possible contributor to poor finish. Data from three studies (8-10) are included in Table 2 (note: not stated whether these numbers were on a volume or solid material basis). The semitransparent stains are either acrylic or linseed oil based. The nonvolatiles range between 20-25% for acrylic stains and 20-76% for the linseed oil stains. The range of percent pigment was 11 to 21 and percent resin 4 to 12. Pigment volume concentration (PVC) was between 14 and 22. The PVC of the acrylic stain of the current work was <1%, indicating far less pigmentation and more transparency. It was also used as a topcoat after an acrylic primer application, unlike the other stains in Table 2. For comparable topcoat coverage rates, the primer application may result in better film build. The viscosity/rheology was not reported for any of the stains; viscosity/rheology may be an important factor in the resultant variation in observed film thickness.

In summary, surface morphology, stain coverage and formulation, and application methods play a role in determining the quality of finish over highly discontinuous features such as those present in the plywood substrate used in our work.

Based on the observations presented and information available on coating performance on textured wood surfaces, the following actions may reduce or eliminate these undesirable effects:

* Denibbing of panel surfaces, if practicable, to remove tissues that are likely to protrude beyond the outer extremities of coating film.

* More important than denibbing may be attention to the application methods. Back-brushing is important following spray application, particularly on saw-textured surfaces.6

* Adjusting formulation of stain, through the use of defoaming agents, in order to minimize foaming and thereby reduce air bubbles, or of additives to change the viscosity and rheology of the coating material to result in better uniformity in coating thickness over the saw-textured surface.

* Increasing the average film thickness in order to get a minimum film thickness at all points on the surface.

CONCLUSIONS

The work presented here fills an important gap in the understanding of wood-coating interaction for a highly textured substrate surface, as well as demonstrates the value of using a combination of microscopic techniques for assessing the quality of a potentially attractive product for use as external siding. An important utilization of plywood is that of exterior siding for buildings. Saw-texturing is effective as it significantly negates the otherwise rampant surface checking while creating increased fiber area for enhanced paint retention and thus better exterior performance, so long as considerations are given to using an appropriate coating formulation and application to ensure a proper film build. Importantly, the demonstrated variability of coating thickness, and its cause, has to be overcome. Because of the highly complex nature of wood-coating interaction in textured wooden products, as demonstrated in our work, we anticipate microscopy playing an important role in performance assessments of coated textured wooden products.

References

(1) Dawson, B.S.W., Gottgens, A., and Hora, G., "Natural Weathering Performance of Exterior Wood Coatings on Pinus Sylvestris and Pinus Radiata in Germany and New Zealand," J. COAT. TECHNOL. RES., 2, No. 7, 539 (2005).

(2) Murmanis, L., River, B.H., and Stewart, H., "Surface and Subsurface Characteristics Related to Abrasive-Planing Conditions," Wood Fiber Sci., 18, No. 1, 107-117 (1986).

(3) Singh, A.P., Anderson, C.R., Warnes, J.M., and Matsumura, J., "The Effect of Planing on the Microscopic Structure of Pinus Radiata Wood Cells in Relation to Penetration of PVA Glue," Holz Roh Werkst., 60, 333-341 (2002).

(4) Carrano, A.L., Taylor, J.B., and Lemaster, R.L., "Machining-Induced Surface Damage of Wood," Forest Prod. J., 54 (1), 85-91 (2004).

(5) Sinn, G., Gindl, M., Reiterer, A., and Stanzl-Tschegg, S., "Changes in the Surface Properties of Wood Due to Sanding," Holzforschung, 58, 246-251 (2004).

(6) Williams, R.S., Knaebe, M.T., and Feist, W.C., "Finishes for Exterior Wood: Selection, Application and Maintenance," Forest Products Society, Madison, WI, c1996.

(7) Richter, K., Feist, W.C., and Knaebe, M.T., "The Effect of Surface Roughness on the Performance of Finishes. Part 1. Roughness Characterization and Stain Performance," Forest Prod. J., 45 (7/8), 91-97 (1995).

(8) Feist, W.C., Mraz, E.A., and Black, J.M., "Durability of Exterior Wood Stains," Forest Prod. J., 27 (1), 13-16 (1977).

(9) Feist, W.C., "Weathering Performance of Finished Southern Pine Plywood Siding," Forest Prod. J., 38 (3), 22-28 (1988).

(10) Williams, R.S. and Feist, W.C., "Effect of Preweathering, Surface Roughness and Wood Species on the Performance of Paints and Stains," J. COAT. TECHNOL., 66, No. 828, 109 (1994).

(11) Williams, R.S. and Feist, W.C., "Durability of Yellow-Poplar and Sweetgum and Service Life of Finishes After Long-Term Exposure," Forest Prod. J., 54 (7/8), 96-101 (2004).

(12) Mall, D.M. and Williams, R.S., "Finishes Checklist: A Guide to Achieving Optimum Coating Performance on Exterior Wood Surfaces," JCT COATINGSTECH, 1, No. 3, 36 (2004).

(13) Singh, A.P., Dunningham, E.A., and Plackett, D.V., "Assessing the Performance of a Commercial Wood Stain by Transmission Electron Microscopy," Holzforschung, 49, 255-258 (1995).

(14) Singh, A.P. and Dawson, B.S.W., "The Mechanism of Failure of Clear Coated Wooden Boards as Revealed by Microscopy," IAWA J., 24 (1) 1-11 (2003).

(15) Cassens, D.L. and Feist, W.C., "Exterior Wood in the South: Selection, Application and Finishes," Gen. Tech. Rep. FPL-GTR-69. Madison, WI: USDA, Forest Service, Forest Products Laboratory, 60 p. (1991).

(16) Chang, S.-T., Hon, D.N.-S., and Feist, W.C., "Photodegradation and Photoprotection of Wood Surfaces," Wood and Fiber, 14 (2), 104-117 (1982).

(17) Dunningham, E.A., Plackett, D.V., and Singh, A.P., "Weathering of Chemically Modified Wood. Natural Weathering of Acetylated Radiata Pine: Preliminary Results," Holz Roh Werkst., 50, 429-432 (1992).

(18) Evans, P.D., Thay, P.D., and Schmalzl, K.J., "Degradation of Wood Surfaces During Natural Weathering. Effects on Lignin and Cellulose and on the Adhesion of Acrylic Latex Primers," Wood Sci. Technol., 30, 411-422 (1996).

(19) Hon, D.N.-S., "Photochemical Degradation of Lignocellulosic Materials," in Developments in Polymer Degradation, Grassie, N. (Ed.), Applied Science Publishers, London, 229-281, 1981.

(20) Hon, D.N.-S., "Weathering Reactions and Protection of Wood Surfaces," J. Appl. Polym. Sci., 37, 845-864 (1983).

(21) Kuo, M. and Hu, H.N., "Ultrastructural Changes of Photodegradation of Wood Surfaces Exposed to UV," Holzforschung, 45, 347-353 (1991).

(22) Plackett, D.V., Dunningham, E.A., and Singh, A.P., "Weathering of Chemically Modified Wood," in Chemical Modification of Lignocellulosic Materials, Hon, D.N.-S. (Ed.), Marcel Dekker, Inc., New York, 277-294, 1996.

(23) Turkulin, H., Arnold, M., Derbyshire, H., and Sell, J., "Structural and Fractographic SEM Analysis of Exterior Coated Wood," Surf. Coat. Int. Part B. Coatings Transactions, 84 (B1), 67-75 (2001).

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

(25) Rijckaert, V., Stevens, M., Van Acker, J., De Meijer, M., and Militz, H., "Quantitative Assessment of the Penetration of Water-Borne and Solvent-Borne Wood Coatings in Scots Pine Sapwood," Holz Roh Werkst., 59, 278-287 (2001).

(26) Van den Bulcke, J., Rijckaert, V., Van Acker, J., and Stevens, M., "Quantitative Measurement of the Penetration of Water-Borne Coatings in Wood with Confocal Laser Microscopy and Image Analysis," Holz Roh Werkst., 61, 304-310 (2003).

(27) Singh, A.P. and Dawson, B.S.W., "Confocal Microscope--A Valuable Tool for Examining Wood-Coating Interface," J. COAT. TECHNOL. RES., 1, No. 3, 235 (2004).

(28) Dawson, B.S.W., Gallagher, S., and Singh, A.P., "Microscopic View of Wood and Coating Interaction and Coating Performance on Wood," Forest Research Bulletin, No. 228, New Zealand Forest Research Institute (2003).

Adya P. Singh and Bernard S.W. Dawson ([dagger])--Ensis, a joint venture of CSIRO and Forest Research*

* Private Bag 3020, Rotorua, New Zealand. Email: [e-mail].

([dagger]) Author to whom correspondence should be addressed.

Table 1 -- Coating Composition and Coverage Rates

                                               Volume
                                               Solids-Wet  Nonvolatile
                Pigment in Dry  Resin in Dry   Coating     Content
                Film (%, v/v)   Film (%, v/v)  (%, v/v)    (%, m/m)

Primer          0               99.3           30          30
First topcoat   0.5-1.0         95             35          40
Second topcoat  0.5-1.0         95             35          40

                Pigment
                Volume                              Coverage
                Concentration (a) (%)  Application  ([m.sup.2]/L)

Primer          0                      Spray        4
First topcoat   0.5-1.0                Spray        5
Second topcoat  0.5-1.0                Spray        7

(a) PVC determined by dividing the volume of pigments by the combined
volume of pigments and resin x 100.

Table 2 -- Properties of Stains from Other Studies

Stain        Reference  Pigment (%)  Resin (%)  Nonvolatiles (%)  PVC

Linseed oil   8         10.5          8.9       20                14.3
Linseed oil   8         16.1          3.9       20                30.2
Acrylic       8         11.7         12.3       20                14.3
Acrylic       8         20.8          5.2       20                30.2
Acrylic       9         --           --         25                --
Linseed oil   9         --           --         46                --
Linseed oil   9         --           --         76                --
Linseed oil  10         --           --         76                --