Dielectric barrier discharge (DBD) plasma pretreatment of lignocellulosic materials in air at atmospheric pressure for their improved wettability: a literature review

The principle of DBD plasma generation

DBD discharges can be generated in many configurations (Wolf 2013) and a typical atmospheric pressure DBD plasma setup for surface treatment of wood is presented in Figure 1. It consists usually of metallic, mostly chilled, conductive electrodes connected to a high voltage supply of alternating polarity (Král et al. 2015). Thin alumina sheets can also be used, on which silver-based electrodes are painted (Hardy et al. 2015). Another electrode or the treated lignocellulosic material itself may serve as a counter electrode (Manolache et al. 2008). An electric field of sufficient field strengths is generated by application of high voltage in the gap between the electrode and material, where the electrical breakdown takes place (Wolf 2013) and a stable plasma state is formed (Zanini et al. 2008a). The total charge is approximately zero, i.e. the number of electrons is equal to the sum of positive and negative ions multiplied with their respective charges (Magalhães and De Souza 2002). Free electrons are accelerated towards the positive electrode. Upon impact, they dissociate gas molecules yielding ionization and fragmentation. Ions and electrons continue colliding with further gas atoms and molecules, and promote an electron avalanche (Wagner et al. 2003). Plasmas consist at least partially of ionized gas molecules, electrons, free radicals, positive or negative ions, excited atoms and molecules. Further, they emit photons ranging from ultraviolet to infrared (Denes and Young 1999). A dielectric barrier covering the metal electrode, made of glass, quartz, ceramics, thin enamel or polymer coating on the metal electrodes (Kogelschatz 2002) prevents the formation of an arc discharge, thus leading to a better stability and homogeneity of the discharge (Rehn and Viöl 2003). Further, the dielectric barrier prevents an excessive heat transfer and the thermal damage of the material (Kogelschatz 2003).

Figure 1:

Schematic presentation of a DBD plasma applied on a wood surface (based on Rehn and Viöl 2003; Viöl et al. 2012)

Insertion of a dielectric material into the electric discharge modifies the dielectric’s surface. As the voltage at the electrodes increases, space charges accumulate in the discharge gap, generating many micro discharge channels (filaments) via the electron avalanche mechanism described above. In every filament, with typical diameters being 0.1 mm, currents of a few amperes flow for a few nanoseconds (1–100 ns) (Conrads and Schmidt 2000). When a discharge channel hits a dielectricʼs surface, the diameter of the discharge channel expands. Electric charge builds up on the dielectric, which reduces the electric field at the current spot. Electrons, with energy levels 4–8 eV at 100–300 Td reduce electric fields (Kogelschatz 2003), reach the substrate and are able to break chemical bonds between molecules, where ions, free radicals and other thermodynamically unstable species are created. Consequently, many of surface properties are changed. These activated species are capable of surface modification in a variety of ways (Evans et al. 2007; Acda et al. 2012a). Energy must be continuously supplied into the system to supply the lost energy on the surface (Lekobou et al. 2016).

History of DBD plasma sources

Plasma technology has been utilized since 1960s to improve the surface properties of fibers in many applications. Early investigations mainly focused on the improvement of wettability, shrinking resistance, twisting and desizing of fibers (Li et al. 1997). The first DBD reactor was built in 1857 by the Siemens Company for the production of ozone, and these are still the most popular cold atmospheric plasma sources (Kogelschatz 2002; Denes et al. 2005).

DBD plasma in wood technology

An electrical gas discharge is one of the most significant high-tech methods for modification (Chu et al. 2002) of wood surfaces (Papadopoulos and Mantanis 2011; Petrič 2013) and other natural fibers, especially with respect to their low wettability (Wascher et al. 2014) after chemical or thermal treatments. The activation begins with the radicalization of the molecules on the surface (Zanini et al. 2008a). A DBD is also applicable to initiate grafting of polymers onto lignocellulosic materials, in the course of which fibers are coated by grafted material via free-radical reactions (Vander Wielen and Ragauskas 2006). PTs have been successfully applied to improve adhesion between lignocellulosics and thermoplastics (Mahlberg et al. 1998). Plasma pre-treatment of wood and wood-based material have many beneficial effects for higher productivity and lower cost of wood industry (Avramidis et al. 2009, 2010). This is especially true when water-based adhesives are used, also for paintings and impregnation agents, due to the shorter processing times. These materials are also suited to a continuous in-line processing (Tóth et al. 2007; Odrášková et al. 2008). DBD has a great potential for modifying a series of chemical and physical properties of lignocellulosic fibers (Vander Wielen et al. 2005). Physical-chemical transformations occurring during the exposure of the substrate to plasma determine the strength of adhesion of coatings on the substrate, the chemical resistance, water absorption and frost resistance (Volokitin et al. 2016). The surface etching of the fiber leads to a physical modification, while reactive ions produce an enlargement of contact area, which increases the friction between the fiber and polymer matrix. A chemical modification is the implementation of active polar groups on the fiber surface, increasing the surface energy and promoting chemical bonding between the fiber and the polymer matrix (Li et al. 1997).

Different mechanical processes (sanding or planning) and chemical pre-treatments are usual to improve wettability, reactivity and smoothness of wood surfaces. Via PTs, smoother surfaces can be achieved without decreasing the thickness of the material, and in contrast to conventional wet-chemical modification, it does not require vacuum conditions or special solvents, nor does it include energy-intensive drying processes, and chemical waste is not produced (Vander Wielen et al. 2005; Demirkir et al. 2014). According to Viöl et al. (2012), electric gas discharges can be divided into low pressure (1–100 Pa) and atmospheric pressure plasmas. The determination of the plasma parameters is based on relationships, which differ for plasmas in local or non-local thermodynamic equilibrium plasmas (Tendero et al. 2006). Gas temperatures in the thermodynamic equilibrium plasmas are well above 1000°C, which are therefore not suited for wood treatment (Jamali and Evans 2011). NTP have low-caloric capacity of electrons (Zanini et al. 2008a), and neutrals are preferably below 120°C or at room temperature. For NTP, high voltages or high frequency electromagnetic fields are used to heat the ambient electrons above the ionizing energy of the surrounding gas (Král et al. 2015) in such a way, that the influence on ions is negligible.

DBDs were the NTP sources for activating wood surfaces (Král et al. 2015) and have been identified as the most relevant modification approaches (Prégent et al. 2015). DBDs are more practical than RF and microwave plasmas for surface treatment, where usually a vacuum is required (Vander Wielen and Ragauskas 2006). The three most extensively studied types of DBD plasmas are volume DBD (Avramidis et al. 2009), plasma jet (Gascón-Garrido et al. 2016, 2017) and surface DBD (Král et al. 2015). Peters et al. (2017) compared three plasma configurations for modification of maple (Acer sp.) solid wood, wood plastic-composite (WPC) and high density fiberboard (HDF) and found only minor differences in terms of the temperature influence on surface energy and reduction of the electrical field strength. A low temperature DBD plasma (Wolkenhauer et al. 2007) works between 50 and 60°C (Avramidis et al. 2011, 2012a; Tang et al. 2015a) and does not entail a thermal effect to wood (Busnel et al. 2010).

Depending on the processing conditions and plasma device design, primary applications of NTPs in surface processing and coating include: (i) surface pre-treatments like cleaning (decontamination, grease removal) and activation (adherence or anti-adherence properties) or passivation (ablation and degradation); (ii) deposition of films and (iii) post-treatments of coated surfaces in order to change the chemical composition or crystallinity of the coating (cross-linking, polymerization) (Podgorski et al. 2000; Bárdos and Baránkova 2010). The typical effects of air PTs on wood surfaces are summarized in Figure 2.

Figure 2:

Schematic presentation of different effects of plasma treatment in air on macroscopic, microscopic and molecular level of wood surface.

Activation of wood surfaces

Topala and Dumitrascu (2007) summarized the various ways how wood surfaces are inactivated via humidity, temperature, sunlight, volatile organic compounds, heat, dust and grime and microorganisms, beetles and termites. Activation via atmospheric NTPs (Lütkemeier et al. 2016) is limited only to the functionalization of material close to wood surfaces without influencing the bulk material’s mechanical properties. Strength, hardness or density remain untouched, and no material is lost during the treatment (Potočňáková et al. 2013).

Drying of wood

Drying of wood inactivates its surface. This process causes a migration of the extractives, molecular reorientation of functional groups or irreversible closure of large micropores in cell walls and consequently, a significant reduction in bonding ability occurs (Šernek et al. 2004). Altgen et al. (2015) found that PT is a useful technique to improve bonding of less wettable wood surfaces including those of recycled, modified or dried wood. Huang et al. (2011) proved, that cold PT significantly enhances wettability of hybrid poplar (Populus hybrids) veneers’ surface. Three-ply panels made from extremely oven-dried veneer bound with urea-formaldehyde (UF) resin, met glue bond shear strength standard requirements.

Effect of natural aging

The status achieved by plasma modification is not permanent and the surfaces tend to return to their untreated condition. The elapsed time is an important parameter in this context. The polar component of the surface free energy is relatively stable in the first 2 days, but afterwards, values rapidly decrease to the original state. Aging mechanisms of the surface are species dependent, and the PT and other parameters influence the results. The complex reactions of free radicals including recombination and oxidation reactions lead to non-reversible etching lower bonding strengths of aged surfaces (Zhang et al. 2015; Chen et al. 2016). Higher power and longer time of glow discharges may keep a degree of wettability, even after the exposure to atmospheric air (De Cademartori 2016a). After 2-weeks’ exposure to air at 20°C and 50% relative humidity, the samples of sugar maple (Acer saccharum March.) and black spruce (Picea mariana Mill.) lost their polished wettability received by an initial PT (Riedl et al. 2014). Similar aging effects were reported by Odrášková et al. (2008) on oak wood surfaces, who explained the effects by reactions of ambient air with the active chemical groups produced on the plasma treated surfaces. Energetically favorable migration of compounds with low surface energy from the bulk wood to the surface may also be an explanation. Applying coating systems right after PTs is recommendable to take advantage of the potentially best hydrophilization effect, and in this way, the best bonding between the activated surface and the applied coating may be achieved (Jablonský et al. 2014). The CAs of UF adhesive significantly increase, indicating a reduced wettability on aged spruce wood [Picea abies (L.) Karst.] surfaces compared to the freshly planned surface. PT performed on aged surfaces leads to a clear recovery of wettability, which is manifest in the elevated tensile shear strength of UF bonds (Konnerth et al. 2014).

Surface structure and roughness

The DBD treatment shapes the fibers’ topochemistry. At high power densities or long duration of the PT, cellulosic fibers become smoother (Xiao et al. 2015), and the coefficient of friction and the surface roughness decrease (Vander Wielen et al. 2005). The structure of the PT surface can be exactly visualized and recorded by atomic force microscopy (AFM) (Bente et al. 2004). Clearly, PT enlarges the surface contact area between the wood and the coating or adhesive leading to a stronger mechanical entanglement (Tang et al. 2015b; Lütkemeier et al. 2016).

PT techniques modify only the outermost layer of the substrate surface (in a depth of around 330 nm) without changing the chemical structure of the bulk wood (Král et al. 2015). DBDs are able to amend chemical and physical properties by removing weak boundary layers (WBLs). These layers arise, when low molecular impurities migrate from the bulk to the surface (Hse and Kuo 1988). Mechanical processing such as sawing, planning, sanding and polishing leads to mechanical WBLs (Mertens et al. 2006), which may also contribute to paint adhesion failure (Král et al. 2015). They act as a low cohesion barrier between the substrate and a surrounding phase (Baltazar-y-Jimenez et al. 2008). PT partly removes the WBLs and has a similar effect to that of sanding (Hardy et al. 2015).

Chemical modification

Chemical modification may increase or decrease the surface wettability (Temiz et al. 2006; Blanchard et al. 2009; De Cademartori 2016a). As mentioned above, the same effects can be achieved by atmospheric PT (Tino and Smatko 2014; Avram et al. 2015). The separate consideration of the chemical changes on cellulose, hemicelluloses, lignin and extractives is a challenge (Prégent et al. 2015). Different kinds of atmospheres may give rise to different kinds of radicals, cations (e.g. N⁺, O⁺, OH⁺, H₂O⁺, N₂⁺, O₂⁺, Ar⁺, N₂O⁺, CO₂⁺) and anions (e.g. OH⁻ or O₂⁻) during PT (Baldus et al. 2015) and these act as chemical interlinks between the wood surface and the applied polymer (Wallenhorst et al. 2015). Hydroxyl (OH) and carboxyl (COOH) groups are the most abundant functional groups in wood and other lignocellulosic fibers. Of course, treatments generating these functional groups elevate the polar character (and hydrophilicity) of the surface (Setoyama 1996; Šernek et al. 2004; Avramidis et al. 2010). The surface chemistry of wood is characterized by optical microscopy, transmission electron microscopy (TEM), scanning electron microscopy (SEM), X-ray photoelectron spectroscopy (XPS) and Fourier-transform infrared (FTIR) spectroscopy (Toriz et al. 2008; Klarhöfer et al. 2010). PT in atmospheric air leads to a decrease in the C-C and C-H components, and to an increase in the polar components (C-O and O-C=O), indicating the oxidation of lignocellulosic materials (Avramidis et al. 2009; Hardy et al. 2015). By oxidative reactions, also other polar, oxygen containing functional groups, such as aldehyde, carbonyl, ether or peroxides are formed (Acda et al. 2012b; Jablonský et al. 2014).

External and internal parameters

Plasmas are characterized by external and internal parameters, but in general, the most important parameters are electron temperature, gas temperature (i.e. rotation temperature of the molecules), power density and energy density (Peters et al. 2017).

External parameters, with respect to the substrate and the desired application (Magalhães and De Souza 2002; Custódio et al. 2008; Blanchard et al. 2009; De Cademartori et al. 2015), include, e.g. 1. the material properties and its geometrical shape, dimensions of plasma reactor and electrodes, 2. plasma-gas composition, gas flow rates, total and partial pressure, as well as 3. electrical power and its mode (specific amplitude, driving-field frequency, duty cycle, etc.), 4. applied voltage or current, 5. reduced electric field strengths during breakdown, 6. frequency of alternating current and 7. the time of exposure.

Internal plasma parameters include densities of charged and neutral particles, such as electrons and ions of both polarities, free radical species, atoms or molecules. The different average energies and their distribution functions also belong to the internal parameters, such as the basic plasma properties, plasma frequency, Debye length and degree of ionization (Denes et al. 2005).

Effect of the gas for plasma generation

The organic or inorganic gases applied have a major importance (Podgorski et al. 2000; Tendero et al. 2006). Many plasmas generated from organic gases are capable of depositing thin films on surfaces. Other plasmas modify materials by ion or atom implantation, oxidation and etching of low molecular weight materials from surfaces (Evans et al. 2007). Most gases for PT are non-toxic and do not have a special effect on the material in their non-plasma state (Potočňáková et al. 2013). Bente et al. (2004) described two types of gases. The first one is a working gas, which activates the surface, making subsequent processing steps more effective. The second one is a processing gas creating protective layers on the surfaces. Usually, it is blown through the discharge gap at a fixed velocity (Rehn et al. 2003). By applying two gases, the developed layers are thicker and better formed (Bente et al. 2004). A major by-product of any DBD plasma in air is ozone (Tendero et al. 2006). As is known, ozone breaks down double bounds (Zanini et al. 2008b). Short DBD PT in synthetic air at atmospheric pressure leads to oxygen loss at the surface, but does not split off all OH groups (Klarhöfer et al. 2010). OH groups in cellulose and hemicelluloses are susceptible to oxidation, and form carboxylic acids lowering the surface’s pH (Aydin and Demirkir 2010; Wascher et al. 2014). Extractives may act as free-radical scavengers against the newly formed highly reactive radicals (Hardy et al. 2015).

The determination of surface energies by means of sessile drop CA measurement is an established method (Wolkenhauer et al. 2009a) based on the shape change of the liquid droplets and their spreading on the surface as a function of time (wetting) (Topala and Dumitrascu 2007). Surfaces with a water CA<90° are called hydrophilic and those with CA>90° hydrophobic. Many applications require high wettability with water CA<0° (Rehn and Viöl 2003; Fang et al. 2016).

Modifications of working gases

The working gas (single gas or gas mixture), introduced via a gas shower, has a predetermined flow rate in the discharge gap (Topala and Dumitrascu 2007). Gas-dependent discharge modes can be very different leading to special surface layers (Dahle et al. 2012). Atmospheric air, nitrogen or oxygen increase the hydrophilicity (Wolkenhauer et al. 2008a). PT in nitrogen and later exposure to air manly leads to the formation of C=O type bonds (Tóth et al. 2007). Argon PTs, in contrast to air PTs, degrade OH groups (Klarhöfer et al. 2010). Gas mixtures containing oxygen augment the polar components of the surface energy, and thus increase wettability (Wascher et al. 2014). Ambient air as the working gas of PTs distinctly improves the wetting characteristics of wood surfaces (Avramidis et al. 2012b; Sinic and Weigl 2012) and is the best industrially-applicable atmosphere, because it provides similar results to both CO₂ and N₂ atmospheres, but is considerably less expensive (Tino and Smatko 2014). Rehn and Viöl (2003) reported that the shortest water droplet uptake times on pine wood surface were recorded after cold PT in atmospheric air, whereas treatment times were a few seconds longer for N₂ or He, and the longest times were found for Ar. Still, water droplet uptake times on all PT surfaces are shorter than on the untreated surfaces. Riedl et al. (2014) treated sugar maple and black spruce with DBD plasma of Ar, N₂, CO₂ and ambient air. Pull-off tests on cured films of water-based coatings revealed that the PT significantly increased the adhesion of the coating, independently of the working gas and exposure time.

Deposition from process gases

The working gases described above are usually used in the first step of PTs (Bente et al. 2004). PTs in the dry state allow the deposition and polymerization of a wide range of monomers at low temperatures. Precursors of polydimethylsiloxane (O’Neill et al. 2005), such as hexamethyldisiloxane (HMDSO) (Avramidis et al. 2009; Konnerth et al. 2014), are most commonly employed for the plasma deposition of coatings. Other approaches include the deposition of fluorocarbon thin films on wood and wood-based substrates from carbon tetrafluoride (CF₄) (Vander Wielen and Ragauskas 2006) or from octafluoropropane in Ar (De Cademartori et al. 2017). The final characteristics of the created layers are controlled by the parameters of plasma polymerization. Due to small penetration energies, a deeper penetration into the wood is not expected (Konnerth et al. 2014; Moghaddam et al. 2016). Additionally, the cross-linking of macromolecules, generating protective layers on the surface, leads to an increased resistance against scratching, a better abrasion resistance and improved barrier properties (Novák et al. 2013; Jablonský et al. 2014).

Effect of gap distance

To ensure a stable plasma operation, the gas-filled gap separating the electrodes can range from less than 0.1 mm to about 100 mm (Kogelschatz 2002). With longer gap distances, the plasma is liable to a higher energy consumption and more reactions may occur (Fang et al. 2016). A typical feature of atmospheric discharges at high voltages and plasma power densities is the formation of filaments, also called streamers, that are statistically distributed over the dielectric surface (Tendero et al. 2006; Černák et al. 2008). Not all DBDs at atmospheric pressure are of the filamentary type (Sankaranarayanan et al. 1999; Ono and Oda 2003), but most of them certainly are. The treatment quality is a function of the gap distance, and therefore complex pieces with 3D shape are difficult to treat (Custódio et al. 2008). The distance between the sample and the electrode influences the relative water uptake time, indicating that electron impact ionization and excitation processes are most significant for the wood surface treatment (Odrášková et al. 2008). Tino and Smatko (2014) observed increasing hydrophobicity with increasing distance from the electrode in case of the wood of Norway spruce, European larch (Larix decidua Mill.), sessile oak [Quercus petraea (Matt.) Liebl.] and European ash (Fraxinus excelsior L.).

Effect of plasma power

The density of charged particles and their energies are generally dependent on the feeding power and power type (e.g. AC, DC, pulsed, frequency, etc.) (Kogelschatz 2002; Tendero et al. 2006). Further, these properties strongly alter the influence of PT material (Vander Wielen and Ragauskas 2006). The heat transport in wood during PT depends on the density, structure and thickness (Lux et al. 2013), and the high combustibility of wood has to be taken into consideration. Higher electrical power densities lead to increased free radical concentration, hotter discharge channels and local heating or burning of the surface (Rehn et al. 2003; Chen et al. 2016). XPS spectra show that increasing the PT power of DBD leads to coupling of free radicals and thus the carbonyl moiety of the peaks is increased (De Cademartori et al. 2016a). Wascher et al. (2014) were the first who reported on the plasma-induced effects occurring within deeper regions of PT beech wood bulk material as a function of discharge power. Water CA data showed that the penetration time was reduced at increased power density.

DBD treatments at high energy density (up to 9.3 kW min m⁻²) cause smoothing of the cellulosic fibers in cell walls of fully bleached southern pine kraft pulp (KP) and unbleached Norway spruce thermomechanical pulp (TMP), and diminish the dispersive part of the surface energy. Increased treatment power causes damage on the over-treated fibers in bleached KP, resulting in the presence of bumps and nodules, formed due to localized melting or the degradation and distribution of surface components. Increased DBD treatment power entailed decreased viscosity of bleached KP fibers (Vander Wielen et al. 2005). However, the less etched surfaces at high power levels can be interpreted as the effect of quenching of free radicals leading to a recombination of free radicals (Chen et al. 2016).

Natural wood

Absorbed gas and water molecules in wood effect the plasma, its stability and uniformity (Setoyama 1996). Rehn et al. (2003) found that the moisture content (MC) of the wood before PT (3, 7, or 10%) has no influence on the final water uptake. This is important for industrial scale application in terms of a constant product quality. Wood with higher MC (10–15%) can be submitted to PT with thicknesses from a few mm (veneers) up to 30 cm (boards) (Custódio et al. 2008).

During the PT of wood species with wide vessels (e.g. sugar maple), an electrical breakdown occurs inside the natural cavities of the vessel elements. The possible breakdown gap is given by the vessel diameter (Wascher et al. 2014). Such modified walls of wood capillaries accelerate and increase the capillary uptake of liquids due to their increased surface energies (Prégent et al. 2015; Wascher et al. 2015).

PT conditions leading to the maximal adhesion improvement differed from one wood species to another (Busnel et al. 2010). Etching of wood with plasma depends on the treatment time, but is also influenced by the structure and the chemical composition of the wood (Jamali and Evans 2011). Abundant extractives, phenolic compounds with hydrophobic character, and their migration from the bulk to surface, may reduce the effect of PT. They contribute to less susceptibility of modified surfaces to establish a close contact between molecules of wood and coatings, thereby the hardening or curing is retarded. Minimizing the effect of extractives is wood species dependent (Šernek et al. 2004; Busnel et al. 2010; Konnerth et al. 2014). The DBD PT of tropical species can significantly improve their ability to form bonds with coatings (Acda et al. 2012b). During PT of some wood species, cell wall material or extractives can be broken down and may later re-polymerize on the surface. During air driven DBD at atmospheric pressure on Scots pine (Pinus sylvestris L.), the degradation and removal of the extractives were clearly demonstrated by XPS analysis. An increased O/C ratio indicates oxidative activation of the treated surface, but only a minor impact on their surface energetic characteristics was seen (Avramidis et al. 2012a).

Substantial improvement of wettability by water of pedunculate oak (Quercus robur L.) heartwood was achieved in the experiments of Odrášková et al. (2008). The wetting behavior of radial and tangential cutting planes of spruce, beech and ash wood in atmospheric air became nearly omnidirectional after PT (Lux et al. 2013).

Modified and treated wood

Heat treatment (HT) of wood causes a loss of OH groups, and therefore entails a reduced hygroscopicity on its surface. This adversely affects coating or gluing. However, HT wood might be very susceptible to PT because of its higher proportion of condensed and more hydrophobic lignin content (Wolkenhauer et al. 2009a,b). Therefore, DBD PTs are useful to modify the hydrophobic surface characteristics of HT beech, Norway spruce, as well as sap and heartwood of Scots pine. PT-HT wood showed lower water CAs with only minor differences concerning the PT times (Altgen et al. 2016a). Mertens et al. (2006) and Wolkenhauer et al. (2008b) treated HT beech wood in DBD plasma in atmospheric air; the resulting CA data of various liquids were similar to those of natural beech wood. Pre-treatments of HT and melamine treated beech and radiata pine wood with DBD plasma at atmospheric pressure decreased the delamination of polyurethane (PUR) and melamine-urea-formaldehyde (MUF) bonded specimens. An improved MUF bonding performance could be attributed to the enhanced wetting and liquid water up-take. Further, PT slightly reduced the PUR penetration (Lütkemeier et al. 2016).

Wax treatment of wood clearly affects its surface energy characteristics and therefore the adhesion of adhesives and paints. The peel force tested via cotton tissue glued with polyvinyl acetate (PVAc) adhesive onto wax treated beech wood surfaces increased after PT by 50–60% (Avramidis et al. 2011). Scholz et al. (2010) bonded the same type of wood with glues of PVAc, phenol-resorcinol and emulsion polymer isocyanate (EPI). A clear influence of the chosen wax, adhesive and PT duration was seen on the shear strength of the adhesive bond. Smoother fibers of fully bleached southern pine KP and unbleached Norway spruce TMP were found after PT. The presence of lignin in TMP protected the fiber from the DBD induced degradation, but a breakdown of the cell wall layers was not visible (Vander Wielen et al. 2005).

Wood-based composites

Natural fibers are widely used as fillers and reinforcing materials in wood plastic composites (WPCs) (Yuan et al. 2004). To overcome the incompatibility problem between non-polar thermoplastics and strong polar lignocellulosic fibers, cold PT can be applied as an alternative to conventional chemical treatment (Olaru et al. 2004; Kalia et al. 2009; Zivkovic et al. 2016). PT increases the low-surface free energy and reduces the non-polar character of the WPCs (Wolkenhauer et al. 2009b; Hünnekens et al. 2016). This was also shown with DBD PT of softwoods [Oregon pine (Pseudotsuga menziesii) and spruce (Picea abies)], hardwoods [beech (Fagus sylvatica), HT beech and oak (Quercus sp.)], and WPCs, as well as fiberboards (FBs) and particleboards (PBs) (Avramidis et al. 2009, 2012c). Samples of polyethylene-based WPCs, PBs and FBs were treated with DBD in ambient air and coated with linseed oil paint, solvent borne alkyd paint and water borne acrylic paint. Only the adhesion properties of polar coating systems could be increased considerably by the PT. After conducting cross-cut tests on coated WPCs, a decrease in percentage of delamination on PT samples was obvious (Wolkenhauer et al. 2007). The polar component of PE and PP in WPCs increased significantly after DBD PTs. Pull-off tests, determining tensile bond strengths of waterborne acrylic, solvent borne alkyd and oil-based paint cured films on PT-WPC, yield better coatings adhesion. Shear strengths of PUR and PVAc bonds on the same substrate are also increased considerably (Wolkenhauer et al. 2009b).

Hünnekens et al. (2016) compared three different types of air-driven DBD plasmas (coplanar surface barrier discharge, remote plasma and direct DBD) operated at atmospheric pressure, and their effect on the surface properties of treated WPC was studied. They found that the water CAs decreased independently of the WPC formulations and discharge type. The PT of the substrate had a positive effect on the acrylic coating uptake and its better adhesion, compared to the reference samples, which delaminated easily after cross-cut and pull-off tests.

Taking into account all these benefits, the PT of medium density fiberboards (MDF) is a cost effective alternative to convert the hydrophobic MDF surface to a hydrophilic one (De Cademartori et al. 2016a). PT fibers and PBs showed an increased peel force of cotton tissue glued with PVAc on PT areas (Wolkenhauer et al. 2008a).

Effect of liquids

Wood is covered with dye, varnish and other protective layers to prevent natural and biological degradation, i.e. to improve durability. Most wood adhesives and paintings are water-based, thus the wetting is important for their good adhesion, bond formation and rate of hardening (Wolkenhauer et al. 2009b; Novák et al. 2013). The interaction between water and wood involves chemical and physical factors, such as polarity of the chemical groups in wood or capillary forces (Toriz et al. 2008), and their absence impairs wetting (Avramidis et al. 2011). However, a better wettability alone does not necessarily improve wood-coating adhesion, because surface roughness, chemical bonds, electrical discharges or solubility of the coating are acting synchronously during adhesion (Riedl et al. 2014). The wet adhesion of water-borne acrylic dispersions can be improved in the case of PT hardwood veneers, FBs or PBs by water-based modification agents (Lukowsky and Hora 2002; Wascher et al. 2015).

Water from PVAc glue penetrated into PBs surfaces faster during bonding after PT, which was explained by accelerated hardening of the adhesive. With this approach, pressing times could be reduced and the throughput could be accelerated (Wolkenhauer et al. 2009b). Better adhesion between melamine resin and PT pine wood was observed by Rehn and Viöl (2003). PT also accelerates adhesive curing on veneers of maple (Acer pseudoplatanus L.), oak (Quercus sp.), beech (Fagus sylvatica L.) and teak (Tectona grandis Linn.), and results in significantly higher shear bond strength of the overlap joints (Avramidis et al. 2010). Zhang et al. (2015) bonded DBD PT Yunnan pine (Pinus yunnanensis French) wood with UF, MF and MUF resins, finding significantly improved bond shear strength due to PT. An atmospheric NTP contributes to a more homogenous distribution of the UF adhesive on the spruce particles and enhances the internal bond strength of the produced PBs (Altgen et al. 2015). In a study by Altgen et al. (2016b), enhanced spreading and coverage by UF adhesive on HT beech particles and micro veneers’ surfaces was observed after DBD treatment in atmospheric air. Due to a better coverage of the particles with adhesive, the bending strength and modulus of elasticity (MOE) of the PBs were also improved.

Effect of treatment time

The treatment time is essential for the optimization of the process (De Cademartori et al. 2016b). A few seconds of DBD PT treatment time of pine (Král et al. 2015), spruce and beech wood (Rehn et al. 2003), and groundwood paper (Tóth et al. 2007) are sufficient to improve the hydrophiliciy without any visible color modification (Rehn and Viöl 2003). The PT duration has a higher influence on the water CA decrement on wood particles and FB than on WPCs (Wolkenhauer et al. 2007).

Prolonged periods of PT led to modifications of the surface microstructure and their chemistry (Toriz et al. 2008). Etching begins in the thinner regions of the cell walls (around pit apertures) and is more pronounced in cellulose-rich than in lignin-rich areas (Jamali and Evans 2011). The increment of water CA on Yunnan pine surfaces after extended DBD treatment times was interpreted by Zhang et al. (2015) as the consequence of breakage and the formation of new chemical bonds of small molecules. The uptake of the water-based modification agent [1,3-dimethylol-4,5-dihidroxyethylene urea (DMDHEU)] into PT beech veneer increased asymptotically to nearly 100% after 120 s of treatment time (Wascher et al. 2015). PT of wax impregnated wood elevates the surface energy as a function of time. After 120 s PT in atmospheric air, isocyanate and PVAc glue bond are resistant to delamination even after exposure to cyclic climates and they reach shear strength satisfying standard criteria (Scholz et al. 2010). In some cases of repeated short exposures with DBD of wood specimens, hydrophilization could only be achieved after a certain threshold number of passes. Changes are not induced from the thermal effect, as the surface temperature does not increase with repeated treatments because the surface cools down between the treatment periods (Prégent et al. 2015).