In the cork oak, the phellogen forms a continuous layer surrounding stem and branches and may live as long as the tree, although the intensity of its activity decreases with age (Natividade, 1950).

The phellogen activity begins in the first year of the shoot; the first cork cells keep the tangential form of the phellogen initial and build up radially aligned rows (Graça and Pereira, 2004). Besides the periclinal divisions of the phellogen cells, some anticlinal divisions also occur that increase the number of phellogen initials, and therefore of the radial rows of cork cells.

The cork in the first periderm is called virgin cork (Figure 3A). It shows numerous and deep cracks that run mostly longitudinally due to the radial enlargement of the tree (Pereira, 2007).

If the initial phellogen is destroyed, as it happens by the stripping of the cork layer, a new (traumatic) phellogen is formed in the inner tissues of the phloem and begins its meristematic activity in the same way as it happened in the first periderm, thereby forming a new regular cylindrical layer of cork cells around the tree; this cork is called reproduction cork. This new periderm is covered externally by the tissues that remained to the outside of where the phellogen was formed; therefore they include the first phelloderm (formed by the initial phellogen) and a layer of the non-functional phloem. As these tissues become exposed to air, they dry out and develop thin fissures building up what is called the cork back (Figure 3B) (Taco et al., 2003).

The first reproduction cork produced by this second phellogen (also called second cork) is often longitudinally fissured due to the still high tangential growth stress of the young tree (Figure 3B). When the second cork is removed, the process is repeated as described above, and the new cork layers no longer fissure since the tangential growth stress is much smaller (Figure 3C).

In the commercial exploitation of cork oaks, this process is repeated successively and a new phellogen and periderm are formed as described. This ability to develop each time a new periderm is the basis for the sustainability of the cork production and cork oak exploitation.

The cork cells are mostly hexagonal prisms that are stacked by their bases in radially aligned rows disposed in parallel without intercellular voids (Figure 4). Therefore the cork cells appear as a honeycomb structure in the tangential section (Figure 4A) and as a brick-wall structure in transverse (Figure 4B) and radial sections (Figure 4C). On average the cell prism height is 30–40 µm and the cell wall thickness 1–1.5 µm (Table 2). It is possible to observe annual rings that are marked by the presence of a layer of latecork cells at the end of the growth season with a shorter prism height (10–15 µm) and a thicker cell wall (2–3 µm) in comparison with the earlycork cells.

The solid fraction in the cork is 8–9% in the earlycork and 15–22% in the latecork region (Pereira, 2007), which justifies the low density of cork.

An important structural characteristic is the corrugation of the radial aligned cell walls that arise from compression stresses during cork growth, i.e., the new cell layers compress the already existing cork cells by pushing them toward the exterior. Sometimes the cell wall corrugation may be strong, especially in virgin cork (Pereira et al., 1987).

Lenticels are present and develop as lenticular channels that cross radially the cork layers; they cause the so-called porosity of cork that appears as more or less circular in tangential sections of cork and as thin strips in the other sections. Cork porosity has been extensively characterized because it is a visual quality parameter that defines the commercial quality of cork stoppers (Pereira et al., 1996; Costa and Pereira, 2007; Oliveira et al., 2012, 2015).

Reproduction cork has on average 16% extractives, 43% suberin, 22% lignin, and 19% cellulose and hemicelluloses (Table 3). The extractives include non-polar compounds that are extracted by solvents such as dichloromethane (representing on average 5.8% of cork) and polar compounds solubilized by ethanol and water (5.9 and 4.5%, respectively) (Pereira, 2013).

Suberin is the most important component of Q. suber cork. Its composition regarding the long-chain lipid monomers is shown in Table 5. In reproduction cork, the α,ω-diacids are the most abundant monomers (53.0% of the monomers), followed by ω-hydroxyacids (30.6%). The most abundant single monomer is the 9-epoxyoctadecanedioic acid (26.7%), followed by 22-hydroxydocosanoic acid (9.2%), 9,10-dihydroxyoctadecanodioic acid (9.0%), and 9-epoxy-18-hydroxyoctadecanoic acid (8.5%). In terms of chain length, most fatty acids have 18 and 22 carbons, representing, respectively, 66.2 and 14.4% of the total monomers. Suberin also contains, as a major monomer, glycerol that accounts on average to 8.5 or 14.2% of the suberin (Graça and Pereira, 2000; Pereira, 2015).

Lignin is the second most important component in the cork cell wall. Cork lignin is a G-type lignin composed by 95% guaiacyl units, 3% syringyl units, and 2% 4-hydroxyphenyl units (Marques et al., 2006; Marques and Pereira, 2013) with 80% of the inter-unit linkages as β-O-4-alkyl-aryl ether bond (Marques et al., 2016).

The hemicelluloses are mainly composed of xylose and arabinose, with smaller amount of galactose and mannose (Table 4).

There is a substantial chemical variability of Q. suber cork regarding between-tree and between-site differences that has been addressed in several studies (Pereira, 1988, 2013; Conde et al., 1998; Bento et al., 2001; Sen et al., 2016a). The content of suberin is the most important chemical attribute of cork since it is its chemical fingerprint and is directly related to most of its typical properties (Pereira, 2015).

Cork oak forests extend to about 2.2 million ha and produce annually up to 200,000 tons of cork that feed an important industry (APCOR, 2015). Portugal is the main producer of raw cork and of cork products that include mainly cork stoppers for the wine industry and insulation and surfacing boards (Pereira, 2007). The production of cork is based on the periodical removal of the cork layer by cutting large planks that are pulled out from the stem (Figures 3D–F). The cork removal is made usually in June–July, when the tree and the phellogen are active and allow an easy separation of the cork. The period between cork removals is usually 9 years in the major producing regions, which allows a thickness of the cork plank suitable to produce the cork stoppers (>24 mm). The cork oak silvicultural system is directed for cork production and is regulated, e.g., the minimum period between cork removals or the extension of the debarking, aiming at maintaining the overall sustainability of the forests.

The cork planks are primarily directed for the production of natural cork stoppers that are bored out from the planks in the axial direction (Costa and Pereira, 2010). If the cork planks are too thin to allow this, they are directed to produce cork discs to be used for sparkling wine stoppers. The residues of these production lines, as well as other unsuitable cork pieces (for instance, virgin cork) are triturated, and the cork granules are agglomerated with suitable adhesives for production of technical wine stoppers, insulation boards, flooring and surfacing boards, joint sealants, and damping and shock absorption layers, among other usages (Pereira, 2007).

The cork of Q. suber has been comprehensively studied. Nevertheless, areas of interest remain where more knowledge and research are needed, e.g., fundamental studies on the topochemistry and 3D architecture of the cork cell wall, the extent and causes of natural variability in chemical composition and cellular features, and how these relate with properties, as well as innovative approaches to product development and new uses.

The cork cells from Q. variabilis are polygonal in a honeycomb-like arrangement in the tangential section, while the transverse and radial sections show a brick-wall structure with an alignment in parallel rows (Figure 2) (YaFang et al., 2009; Miranda et al., 2013; Ferreira et al., 2016a). Cork cells are prismatic, and the cell walls present corrugations, especially the earlycork cells that form against the previous latecork cells (Miranda et al., 2013; Ferreira et al., 2016a).

Like in the cork oak, the Chinese cork has rings with larger earlycork cells and with thinner walls than latecork cells. The average ring width of Q. variabilis is significantly smaller than in Q. suber cork (0.82 vs. 2.06 mm), and each ring has two to six latecork cells representing 14% of the ring width (Kim, 1993; Miranda et al., 2013; Ferreira et al., 2016a).

Prism height in the earlycork cells of the reproduction cork of Q. variabilis is smaller than in Q. suber (23 vs. 30–40 µm, Table 2). The solid fraction in earlycork is, therefore, higher in the Chinese cork than in Q. suber cork (13 vs. 8–9%) (Miranda et al., 2013; Ferreira et al., 2016a). Consequently, Q. variabilis cork has higher density than Q. suber, what constitutes a disadvantage as an insulating material (Kim, 1993; Miranda et al., 2013). Another difference between these two corks is the appearance of the cell wall: the lumen side of the cork surface of Q. variabilis is rough with deposits of several dimensions, while in Q. suber, it is smooth and with only occasional deposits (Miranda et al., 2013).

Like in the cork oak, lenticular channels cross radially the cork layer with a circular or elliptical shape in the tangential section and with filling tissue (Kim, 1993; Ferreira et al., 2016a).

Virgin and reproduction cork of Q. variabilis are chemically very similar: on average 0.9% ash, 9.2% extractives, 37.8% suberin, and 28.5% lignin. The composition is within the variation range of reproduction cork from Q. suber (Table 3) although with less extractives, with lower levels of non-polar compounds extracted by dichloromethane (YaFang et al., 2009; Miranda et al., 2013; Ferreira et al., 2016a). The ratio suberin/lignin is 1.4 (virgin cork) and 1.3 (reproduction cork), lower than the average of 2.0 in Q. suber (Ferreira et al., 2016a). Given that most of the properties of cork result from the joint presence of suberin and lignin in the cell wall and their relative proportion, the higher proportion of lignin will give Q. variabilis cork higher compressive strength than Q. suber (Pereira, 2013).

Polysaccharides (Table 4) show a large presence of xylan-based hemicelluloses and a comparable proportion of cellulose (glucose amounts to 52% of the neutral sugars).

Suberin composition (Table 5) shows a substantial difference in relation to the reproduction cork from cork oak: Q. variabilis cork has more ω-hydroxyacids and less α,ω-diacids. The main monomers in Q. variabilis suberin are 22-hydroxydocosanoic acid (35.1–21.4%), 18-hydroxy-9-octadecenoic acid (22.0–17.1%), and 9,10-epoxy-18-hydroxyoctadecanoic acid (10.0–10.2%) (Ferreira et al., 2016a). Epoxyacids have lower contents (13.3–14.4%, Ferreira et al., 2016a) than in Q. suber suberin (30%, Graça and Pereira, 2000). The relation between the saturated and the substituted acids in Q. variabilis suberin is 1.0–0.7 (0.4 in Q. suber) (Ferreira et al., 2016a). Therefore, it is to be expected that suberin spatial development may be more compact in Q. variabilis cork. Ferulic acid is also present in the suberin extracts (3.6–3.0%), enforcing the conclusions about its function in the cross-linking between suberin and lignin (Marques et al., 2016).

Cork of Q. variabilis is at present already harvested, industrially processed, and commercialized. Due to the characteristics of the raw material, it cannot be used for production of solid cork products, namely for wine stoppers; instead, it is granulated and used in agglomerates for various surfacing and insulation applications.

The cork from Q. variabilis presents lower quality than that from Q. suber, namely, higher density, compressive strength, and elasticity. However, this cork presents characteristics that are compatible with its use as an insulating, sealing, and energy absorption material and may be considered as a complementary raw material for cork from cork oak (Miranda et al., 2013; Ferreira et al., 2016a). The cork from Q. variabilis can also be a potential source of friedelin or of betulinic acid, due to the high levels of these triterpenes in the extractives (Ferreira et al., 2016a).

The exploitation of this species as a cork provider is far from being developed and optimized as it is for Q. suber and its cork, e.g., the development and improvement of Q. variabilis silviculture and forest management as well as of the respective cork industry may overcome the present constrains. More knowledge and research are certainly needed in the tree physiology related to the bark and cork formation that would complement applied silvicultural studies. The variation of cork characteristics between trees and geographical locations as well as with tree age is also important for improving products and industrial processes.

The cellular structure of K. coriacea cork is similar to that of Q. suber cork. The cork cells are hexagonal prisms, disposed in a honeycomb-like arrangement (Figure 2); cell wall thickness is between 1.5 and 2 µm, and cell length in the radial direction is significantly higher than in cork oak cork (Table 2).

The chemical analysis of K. coriacea virgin cork, shown in Table 3, includes not only the cork tissue but also its phloem inclusions. The content of suberin is from 16 to 30%, and this species presents higher contents of lignin than cork oak cork with 43.6–55.3% (Rios et al., 2014).

Table 4 shows that xylans are the most important hemicelluloses in the virgin cork from K. coriacea; the glucose proportion is higher than in cork of Q. suber and of all the other reported species.

Kielmeyera coriacea cork suberin (Table 5) is chemically composed mainly by ω-hydroxyacids and α,ω-diacids have a lower proportion (Rios et al., 2014). The main ω-hydroxyacids are the C18 9,10 mid-chain substituted ones, namely, with an unsaturated group (18-hydroxyoctadec-9-enoic-acid) and a vicinal diol group (9,10,18-trihydroxyoctadecanoic acid). The α,ω-diacids include the octadec-9-ene-1,18-dioic acid and 9,10-dihydroxyoctadecane-1,18-dioic acid. This suberin does not contain acid monomers with saturated chains, either ω-hydroxyacids, α,ω-diacids, or n-alkanoic acids (Rios et al., 2014).

There is no industrial utilization of K. coriacea bark as a raw material for cork products. However, its exploitation as a cork provider has been already suggested (Natividade, 1950), namely, for the production of cork agglomerates (Rios et al., 2014) since the homogeneity of the cork layer does not seem suitable as a raw material for solid cork products. The cellular and chemical characteristics of K. coriacea cork are compatible with the common uses of cork, namely, as an insulation and surfacing material after trituration and agglomeration. Its use as a source for bio-based chemicals was also suggested, e.g., xantones and long-chain fatty acids from the extractives, or the bifunctional ω-hydroxyacids and α,ω-diacids monomers from suberin (Rios et al., 2014). However, any further considerations on the potential use of K. coriacea cork will require studies on its characterization as well as on tree-related aspects of cork development.

The bark of this species presents an internal thick brown layer of phloem that is covered by a periderm containing a conspicuous light brown layer of cork with deep fissures that result from the tangential growth stress. The periderm develops continuously around the stem with a lifespan probably equal to the tree age and with an annual regular activity of the phellogen (Mota et al., 2016).

Plathymenia reticulata cork cellular structure has no substantial difference from the one of Q. suber cork. The cork cells are polygonal, in a honeycomb-like arrangement (as in Figure 2), cell wall thickness is within the variation range observed for Q. suber but prism base edge is much higher than in cork oak (Table 2).

The cork from P. reticulata shows growth rings, with fewer and smaller latecork cells (two to four cells in a radial row) and more and larger earlycork cells (six to nine cells), although the differences between earlycork and latecork are not so pronounced as in the cork oak cork, leading to a more homogeneous material (Mota et al., 2016). Large variations of size between cells were observed, either in the tangential or in the non-tangential sections. Like in Q. suber, the cork cells of this species have undulations in the non-tangential sections as well lenticular channels, sometimes with filling tissue.

The summative chemical composition is included in Table 3 (Mota et al., 2016). The important feature is that lignin is the major structural component and not suberin (34.5 and 24.7%, respectively). This indicates a low value of the suberin/lignin ratio of 0.7; this was also found by Rios et al. (2014) for the other cerrado species K. coriacea. In comparison with the cork of Q. suber although there is a large variability in the chemical composition values, the ratio suberin:lignin is quite above this value (Pereira, 2013). This chemical difference certainly will bring differences in the chemical behavior, as discussed in Pereira (2015).

In P. reticulata cork, xylan-based hemicelluloses prevail with a considerable proportion of arabinose (Table 4), including acetyl substitutions and uronic acid monomers (Mota et al., 2016).

The suberin composition (Table 5) shows more ω-hydroxyfatty acids and less α,ω-diacids than in Q. suber reproduction cork suberin. The main individual suberin monomers are the 18-hydroxyoctadecanoic acid (saturated and substituted), the octadecanodioic acid, the tetracosanoic acid, and the hexacosanoic acid while only minor values of the 9-epoxyoctadecandioic acid (<1%) were found (Mota et al., 2016), contrarily to what happens in the suberin from cork oak where it is a main component.

The cellular and chemical characteristics of P. reticulata cork are compatible with the common uses of cork, namely, as an insulation and surfacing material after trituration and agglomeration (Mota et al., 2016). Other potential utilizations were identified, namely, as a biosorbent and as a raw material for extraction of valuable compounds like lupeol, fatty acids, and terpenoids.

However, in order to fully develop the utilization of P. reticulata cork as a raw material for cork products under a sustainable cork production management, further research is needed to understand how periderm regenerates after cork removal and the characteristics of the subsequent cork layers. Should the cork be obtained as a by-product of tree exploitation for timber, i.e., the bark is produced as a residue, then its valorization for cork agglomerates is possible straightaway. In any case, studies on P. properties, e.g., density, compression, and insulation properties are needed.

Quercus cerris has a conspicuous cork presence in the rhytidome as layers with a radial width from about 1 to 10 mm, while the phelloderm is composed of a layer of only two or three thin-walled cells (Sen et al., 2011b).

The phellem layers present 2–5 growth rings, and each ring is composed of about 6–12 layers of phellem cells more or less radially aligned without intercellular voids. The phellem cells are suberized, with thin walls with a uniform thickness in the tangential and radial walls and are sometimes radially flattened. Like in Q. suber, at the beginning of the growth ring, cells are larger with thin walls, while at the end of the growth ring, they are smaller and have thicker walls. Despite this similarity, both species have different intensities of phellem growth. In fact, in turkey oak, each phellogen cell produces only 6–12 phellem cells/ring in each radial row (Sen et al., 2011b), and in Q. suber, this figure is about 10–20 in young plants (Graça and Pereira, 2004) and up to 100 in mature trees (Pereira et al., 1992). It is also to remark that in this species, but not in Q. suber, one or two layers of the cork cells, in the limit of each growth ring, can thicken up and become very lignified (Sen et al., 2011b).

Lenticular channels crossing radially the periderm are rare and without filling material. This feature can be explained because in turkey oak cork layers are not continuous around the stem, as it happens in cork oak, and therefore, gas exchange between the innermost living tissues and the exterior does not require the development of these channels (Sen et al., 2011b). The cork tissue also includes lignified phloem cells (fibers and sclereids).

In what concerns phellogen longevity, it is probable that its lifespan is around 25 years (Sen et al., 2011b) but, the fact that each periderm presents from two to five rings suggests that phellogen activity is not annual and that the rings cannot be considered as successive annual rings (Sen et al., 2011a).

The cork cellular structure of Q. cerris is very similar to the one presented in Section “Cork Structure and Chemical Composition” for Q. suber cork (Figure 2). In fact, the cork cells form a bidimensional network of edges and vertices without intercellular voids. The cells are mostly hexagonal prisms, stacked base-to-base disposed in parallel rows aligned in the radial direction, in a compact space-filling arrangement. Frequently, in adjacent rows prism bases coincide with each other or lay in staggered positions (Sen et al., 2011a).

The cell walls also present corrugations. The intensity of corrugation is greater in Q. cerris than in Q. suber, possibly due to the thicker cell walls and to the more irregularity in the stress radial-growth distribution along the cork tissue as a consequence of the presence of phloem between the cork layers (Sen et al., 2011a).

Cork ring width is on average 201 µm and contains seven to nine earlycork cells and two to four latecork cells in each radial row (Sen et al., 2011a). Dimensions of earlycork and latecork cells are presented in Table 2. Prism base edge and cell wall thickness are higher in Q. cerris cork than in Q. suber cork, and therefore, Q. cerris cork has a higher density (Sen et al., 2011a).

The chemical composition of Q. cerris cork is given in Table 3. Suberin is the major cell wall structural component of the cork from Q. cerris but almost in the same proportion as lignin; the suberin:lignin ratio is 1.0. In the polysaccharides, that represent almost 16.7%, glucose is the major monosaccharide (Table 4) but important contents of xylose, arabinose, and galactose are also present. This composition suggests that the main hemicelluloses are xylans (Sen et al., 2010).

In Q. cerris cork suberin, ω-hydroxyacids represent almost 90% of all long-chain monomers (Table 5). The single most abundant compound is the vic-diol mid-chain substituted ω-hydroxyacid in C18: the 9,10,18-trihydroxyoctadecanoic acid represents about 54% of all long-chain monomers. The saturated ω-hydroxyacid in C22 represents 20% and the unsaturated ω-hydroxyacid 18-ω-hydroxyoctadec-9-enoic acid 11%. The α,ω-diacids are present in less than 8% of the total suberin and are constituted mostly by vic-diol substituted 9,10-dihydroxyoctadecanedioic acid with 5% of the long-chain monomers (Sen et al., 2010). Alkanoic acids in C16 and C18 as well as alkanols in C20, C22, and C24 represent a small part of the suberin (Sen et al., 2010). The C18 monomers are the most abundant, followed by C22 monomers, and their relative amounts are not very different from the ones found in cork oak (Sen et al., 2010). The major quantitative difference between Q. cerris suberin and Q. suber reproduction cork suberin lies on the relative proportion of α,ω-diacids and ω-hydroxyacids which is much lower in the first one (Table 5).

Due to the characteristics of Q. cerris bark, the industrial utilization of its cork requires previous trituration and fractioning by size and density in order to obtain pure cork or cork-enriched fractions. Therefore, the use of cork is limited to granulates and the production of cork agglomerates. A pilot-scale fractioning of Q. cerris bark was already made (Sen et al., 2016b): a yield of 8.4% cork pure granulates suitable for agglomerated stoppers was obtained as well as 18.5% of cork-rich fractions usable in the production of surfacing and other cork agglomerates.

Considering the comparative cost advantage of the Q. cerris cork, it seems economically interesting to integrate it into the present cork industrial production lines, thereby allowing to enlarge the cork raw-material supply base. However, further studies are required for process optimization and for the development of Q. cerris-specific cork-based products.

Betula pendula bark has a rhytidome with numerous tightly packed periderm layers that consist of thick-walled cork cells inwards and thin-walled cells outwards (Schönherr and Ziegler, 1980), that resemble annual increments produced in the early and late season of the growth period (Ferreira et al., 2016b). The birch periderm contains alternate layers of strongly suberized cells (five layers of cells with tangential walls up to 2 µm thickness) with a variable number of layers with little or no suberization (Schönherr and Ziegler, 1980). The cells are mostly flattened in the radial direction and stretched in the tangential direction and show many cell wall deposits on the lumen side (Ferreira et al., 2016b). Figure 9 shows SEM photographs of the cross section of birch outerbak.

Birch bark is described as having a rhytidome composed by periderms with successive cork layers and phloem layers with several sclerenchymatic tissues (fibers, sclereids, cortical, and parenchyma cells) and showing conspicuous lenticels that contain suberized and non-suberized cells disposed in stratified layers (Bhat, 1982; Trockenbrodt, 1991; Ferreira et al., 2016b). Suberized layers are continuous across the lenticels and maintain in between the loose tissue together, breaking up successively as a new phellogen is formed. Intercellular spaces with diameters up to 5 µm are seen in the tangential section across the lenticels that represent about 3% of the total periderm area (Schönherr and Ziegler, 1980).

There are a few studies on B. pendula bark chemical composition made in the 70s and 80s of last century [e.g., Holloway (1972, 1983), Holloway and Deas (1973), Ekman (1983), and Ekman and Eckerman (1985)] and more recently on suberin composition [e.g., Gandini et al. (2006) and Ferreira et al. (2013, 2016b)]. It should be noticed that several of the studies were made with the whole birch outer bark instead of only the cork fraction, thereby including the non-suberized cells.

Betula pendula cork contains a large amount of extractives of about 32% (Table 3) mostly corresponding to lipophilic extractives, i.e., 98% of the total extractives are dichloromethane solubles. Their composition shows the striking dominance of triterpenoids (90–97% of all compounds) where betulin corresponds on average to 71% of the extract; the other compounds are lupeol (on average 14%) and betulinic acid (4%) (Ferreira et al., 2016b).

Suberin represents 36% of cork (Ferreira et al., 2016b) corresponding to 53% on extractive-free cork that are in agreement with previous reported values of 59% (Holloway, 1983) and 51% (Holloway and Deas, 1973). Lignin amounts are 14% of the cork, and therefore, the suberin:lignin ratio is 2.5.

Cellulose and hemicelluloses represent only 10% of B. pendula cork with most hemicelluloses formed by arabinose and xylose (Table 4).

The suberin from B. pendula (Table 5) contains aliphatic ω-hydroxyacids of chain lengths between C16 and C26, with a large proportion of components in C18 (73% of all long-chain monomers) and C22 and with smaller quantities of fatty acids and neutral constituents (Holloway, 1972). In comparison with Q. suber reproduction cork, birch cork suberin contains much more ω-hydroxyfatty acids and much less α,ω-diacids. In fact, the main individual suberin monomers are the 9,10-epoxy-18-hydroxyoctadecanoic acid (30% of total monomers), the 22-hydroxydocosanoic acid (13%), the 9,10-dihydroxyoctadecandioic acid (12%), and the 18-hydroxyoctadecanoic acid (12%) (Ferreira et al., 2016b).

Birch cork does not have the structural and cellular characteristics that impart the properties that have made cork so interesting as a material. The cell biometry, e.g., the radially flattened cells and the cork thickness of the periderms, the weak tangential cohesion, and the heavy inclusion of lumen deposits do neither allow a prospective utilization for stoppers nor for the most common applications of cork as a cellular material (Ferreira et al., 2016b). Consequently, this cork main utilization and valorization will be as a chemical source.

Due to its high content in suberin, birch outer bark has been proposed as a source of monomers, e.g., of ω-hydroxyfatty acids, α,ω-dicarboxylic acids, and homologous mid-chain dihydroxy or epoxy derivatives for production of novel macromolecular materials (Gandini et al., 2006), polyesters (Olsson et al., 2007; Sousa et al., 2007), polyols (Evtiouguina et al., 2000, 2002), polyurethanes (Cordeiro et al., 1997, 1999; Evtiouguina et al., 2001) and has also been considered as a source of extractives, namely, of triterpenoids and in particular of betulinol (Pinto et al., 2009). Some patents have already addressed the possibility of using birch bark as a source of chemicals (Krasutsky et al., 2003, 2005, 2009).

The development of the first periderm in P. menziesii is similar to the process reported in Section “Bark Structure and Formation” and when it dies, a new phellogen forms inside the phloem (Grillos, 1956). The new phellogen does not form a whole circumference around the tree, but only in short portions around and along the length of the stem; therefore, the cork layer is interspersed with small parts of inactive crushed phloem (Grillos, 1956; Krahmer and Wellons, 1973). Douglas-fir rhytidome is composed by approximately 49% (in volume) of dead phloem and 51% of periderms (Patel, 1975).

Although the first phellogen can be functional for 25 to 35 years, only the deeper phellogens produce enough cork that show distinct increments (Grillos, 1956).

Most of the periderm is formed by cork cells disposed in well-oriented radial rows and the phelloderm layer is between one and three cells thick (Chang, 1954; Grillos, 1956; Krahmer and Wellons, 1973). The cork thickness is very variable between trees, in average 1.8 cm at breast height ranging from 2 cm to less than 1 mm at root collars and from 5 to 40 cell layers (Kurth, 1953; Grillos, 1956; Ross and Krahmer, 1971). At any particular tree height, the number of cork growth increments is inferior to the age of the tree (Ross and Krahmer, 1971).

One striking feature of Douglas-fir cork is the major presence of bands of compressed cork cells that build up a compact layer of crushed cells due to a severe folding of the radial cell walls (Ross and Krahmer, 1971; Cardoso et al., 2016). These bands of high solid content appear as dark layers, with a successive tangential alignment, separated by light colored layers. These layers have been called as growth increments (Chang, 1954; Ross and Krahmer, 1971; Krahmer and Wellons, 1973).

Pseudotsuga menziesii cork cells are in general thin-walled and uniform in thickness. Some thick-walled cells with small pores or pits are occasionally observed either disposed in bands or scattered (Chang, 1954; Grillos, 1956); when their proportion is high the cork is referred as “woody cork” (Grillos, 1956; Krahmer and Wellons, 1973).

Cork cells are more or less isodiametric, symmetrical in shape, and disposed without intercellular voids. On the transverse (Figure 10) and radial directions, they are similar, square to rectangular (if the cells have not been crushed); on the tangential direction, the cells are mostly pentagonal or hexagonal (Krahmer and Wellons, 1973; Cardoso et al., 2016).

Douglas-fir cork cell dimensions are presented in Table 2. The cell prism height and base edge are high and cell wall thickness is low; the radial walls are wrinkled due to growth stresses after the cell walls are formed (Grillos, 1956; Krahmer and Wellons, 1973; Cardoso et al., 2016); 60% of the cells observed by Cardoso et al. (2016) were crushed.

Pseudotsuga menziesii cork has a high content of extractives (Table 3) that are mostly polar compounds soluble in ethanol and water (23.5% of the cork) (Ferreira et al., 2015a). The lipophilic extracts (5.4% of the cork) contain as major component catechin (49% of the total), as well as pinoresinol and β-sitosterol (Ferreira et al., 2015a).

The cell walls are composed mainly of suberin (Krahmer and Wellons, 1973; Graça and Pereira, 2000). Suberin represents 36.2% of the Douglas fir cork and lignin 16.8% (Ferreira et al., 2015a). The suberin:lignin ratio is 2.2, a value similar to the one in Q. suber cork (Pereira, 2013).

The monosaccharide composition (Table 4) shows that hemicelluloses contain arabinoxylans and also galactoglucomannans, e.g., galactose and mannose represented 20.4% of the sugars, and xylose and arabinose 24.2%. This differs from the compositions of the other corks and is associated to the fact that Douglas-fir is a softwood (Ferreira et al., 2015a).

Suberin composition (Table 5) shows that the most important monomers are the ω-hydroxyacids, representing 36.2% of the total, followed by the α,ω-diacids (18.6%); chain lengths of C16 and C22 are more relevant (49% of the identified long-chain monomers) (Ferreira et al., 2015a).

When comparing with Q. suber cork suberin, the Douglas-fir suberin contains more saturated α,ω-diacids (hexadecanedioic and octadecanedioic acids), less unsaturated α,ω-diacids, and no epoxyacids (Graça and Pereira, 2000; Ferreira et al., 2015a).

The potential of Douglas-fir bark as a raw material, namely, of its cork component, was recognized since the 50s of last century and several studies were performed (Kurth and Kiefer, 1950; Hergert and Kurth, 1952; Kurth, 1953; Grillos, 1956).

Although there is a considerable proportion of cork in Douglas-fir bark, the small dimensions of the cork layers and their discontinuous distribution with interspersed phloem restrict its direct use and require previous bark milling and particle size separation (Ferreira et al., 2015a,b). An integrated bark valorization approach should contemplate the separation of cork-rich fractions to be further purified and used as a raw material for cork-based products (e.g., composites), while phloem-rich fractions can be used for the extraction of polar soluble compounds (e.g., phenols and polyphenols), while the polysaccharide–lignin matrix may be considered for biorefineries (Ferreira et al., 2015b).

The utilization of the Douglas-fir cork granulates in products that rely on the properties given by the typical cork cellular structure is hindered by the high proportion of crushed cells in this cork (Cardoso et al., 2016). Studies on structural improvement, e.g., cell expansion might therefore increase its utilization potential.

The first phellogen initiates in the outermost layer of the cortex below the epidermis. It is long lived and remains active, producing a relatively large amount of cork before a new one is formed, deeper inside the phloem, initiating the formation of a rhytidome. The rhytidome starts at 43–88 years of age, depending on sun exposure (Mogensen, 1968). Before rhytidome formation, the external bark has a soft and spongy nature given by the phellem but becomes rough and hard to the touch after the formation of the rhytidome. The cork layers do not form a complete circumference around the stem and crack due to diameter growth, and the outer layers may wear away (Mogensen, 1968).

Within the cork, there are periodic rows of cells with thick and sclerified outer tangential walls that mark the separation between the growth increments which, in young stems, can be correlated with the age of the stem. The other cork cells are thin-walled and usually slightly elongated radially.

The use of this species as a raw material for cork products seems possible since the initial phellogen remains functional for a long period and produces a large quantity of cork. However, little is known on the structure, chemical composition, and properties of this cork, and research is needed to evaluate its potential utilization.

The potential commercial value of the cork fraction of white fir bark was recognized early (Hergert, 1958). However, the knowledge is very scarce, and further research should be carried out to characterize this cork and its potential application value.