Owing to the constant increase of prices of the process for petrochemical resin and the possibility of harmful formaldehyde emissions from industrial produced medium-density fiberboards (MDF), enzymatic binder systems are discussed as an environmentally friendly alternative for gluing lignocelluloses such as wood fibers. In this work laccase-mediator systems (LMSs) were used to activate the lignin on wood fiber surfaces. Two different mediators were tested, vanillic acid (VAN) and 4-hydroxybenzoic acid (HBA), of which HBA performed best. Carbon-13 nuclear magnetic resonances (13C-NMR) and electron spin resonances (ESR) of LMS-treated thermomechanical pulp (TMP) fibers were determined for qualitative and quantitative analysis of lignin activation. Analysis outputs were transferred to produce MDF using a dry process. 13C-NMR revealed more structural changes in the wood fibers using LMS with HBA than LMS with VAN. ESR spectroscopy indicated a higher amount of phenoxy radicals after treatment with LMS containing HBA as a mediator. The data correlated well with the quality of MDF. The best mechanical technological properties were achieved by using HBA within the LMS, so that the European Norms could be fulfilled. But VAN also performed well, which showed a high potential to produce ecofriendly MDF by using LMSs in the future.ABSTRACT
Production of medium-density fiberboard (MDF) is an important industrial process with about 11.5 million m3 panels produced in Europe per year (European Panel Federation [EPF] 2011). When producing MDF, 20 percent of the production costs are attributed to adhesive agents alone, which are usually urea-formaldehyde (UF) resins (Kharazipour 2004). Owing to the constant increase of crude oil prices and with it the prices of adhesives, the production costs and the sales prices of wood-based panels rose noticeably in the last years (Europäischer Wirtschaftsdienst GmbH [EUWID] 2007, 2010). The levels of emission of formaldehyde from the products are another problem for the industry, mainly linked to formaldehyde resins (Bulian et al. 2003, Roffael 2006, Weigl et al. 2009). In 2004 the International Agency for Research on Cancer (IARC) established new regulations for reduced formaldehyde emissions, CARB I and II (IARC 2004). The European wood-based panel industry has implemented the European Panel Federation-Standard (EPF-S). It defines new maximum permitted emission values, which shall be reduced to 50 percent (Oberdorfer 2008, Sauerwein 2009). A substitution of conventional glues by formaldehyde-free or low formaldehyde–containing resins was also recommended at that time (Oberdorfer 2008).
Several nonconventional processes have been developed to use the natural components of wood as adhesives in MDF without any addition of resins. The nonconventionally produced MDF offer several advantages because they do not produce any formaldehyde emissions. The first attempts in this direction were made using inorganic oxidants. Linzell (1945) described a process for the bonding of wood particles by compressing and heating the wood fibers with an oxidant, ferric salt. Another process involving oxidation of lignocellulose with nitric acid (HNO3) or oxygen (O2) has been patented by Brink (1975). More recent reviews in this area have been published by Zavarin (1984) and by Ellis and Paszner (1994).
The use of enzymes for oxidation of wood particles was first suggested by Nimz et al. (1972), who used peroxidase for the binding process. Haars et al. (1988) used phenoloxidase and a mixture of lignin and resin polymeric methylene diphenyl diisocyanate (pMDI) to produce particleboards. Kharazipour et al. (1997) showed that it is possible to use the bonding strength of laccase-activated lignin on the wood fiber surface for the production of wood composites without any addition of resins. It is generally known that the enzyme laccase is rather unspecific with respect to substrate and will catalyze a number of oxidative reactions with different polyphenols, aryl amines, and lignins as well. Laccase (phenoloxidase) has been found to cause depolymerization as well as polymerization of lignin and lignin derivatives. Owing to its low redox potential and the need for a free phenolic hydroxyl group for its action, in several studies certain organic redox systems, so-called mediators, were used to enhance and accelerate lignin oxidation. It was shown that laccase in the presence of mediators is also able to activate nonphenolic compounds (Bourbonnais and Paice 1990). Actually some mediators like 1-hydroxybenzotriazole, acetosyringone, syringaldehyde, or vanillic acid (VAN) were successfully applied with laccases as laccase-mediator systems (LMSs) in biochemical bleaching of paper pulp and in degradation of textile fibers (Chakar and Ragauskas 2004; Rochefort et al. 2004; Camarero et al. 2005, 2007; Ibarra et al. 2006). Chandra et al. (2005) described a positive effect of the mediator 4-hydroxy-benzoic-acid (HBA) in the modification of paper. With regard to the manufacture of MDF boards, it is suggested that the wood fibers are treated with LMS to generate binder- and formaldehyde-free products (Müller et al. 2009, Euring et al. 2011a). The main aim is to activate the exposed middle lamellae lignin, which becomes plasticized after the thermomechanical pulping (TMP) process. Euring et al. (2011a) have shown that MDF boards can be produced in pilot scale (dry process) using different LMSs within a short incubation time. However, little is known about the detailed reaction mechanism underlying LMS oxidation of wood fibers for making MDF.
Thus one part of this article examines wood fibers by carbon-13 nuclear magnetic resonance (13C-NMR) and electron spin resonance (ESR) techniques after treatment with two different LMSs. Another part deals with the making of MDF by using the same LMSs to transfer the results for technical application. The LMSs consist of a commercial laccase in combination with one of the two mediators. One mediator is VAN, which is also known as 4-hydroxy-3-methoxybenzoic acid, an oxidized form of vanillin. It can be found as a natural component in plants and was already successfully tested in laccase-catalyzed oxidative reactions (Gonzalez et al. 2009). The second mediator, HBA, was tested before for MDF making because of its low price and good availability (Euring et al. 2011a). For the comparison between natural bonded MDF (laccase or LMS) and conventional MDF, UF resin was used.
Materials and Methods
Materials
Commercially produced wood fibers
Spruce wood (containing 100% Picea abies wood) was defibrated into fibers by the TMP process by Pavatex SA in Cham, Switzerland.
Laccase
In this study the commercial laccase Novozym 51003 Trametes villosa, recombinantly produced in Aspergillus oryzae by Novozymes (Bagsveard, Denmark), was used. Its specific activity was routinely measured by monitoring the oxidation of diammonium salt of 2,2′-azinobis-3-ethylbenzothiazoline-6-sulfonic acid (Matsumura et al. 1986). The activity laccase stock was about 1,000 U/g.
Mediators
VAN “for synthesis” (Merck Chemicals, Darmstadt, Germany) had a purity of 98 percent and HBA (Alfa Aesar, Ward Hill, Massachusetts) a purity of 99 percent.
Buffer
McIlvain buffer, the reaction buffer, pH 6.0 (0.2 M dipotassium phosphate [K2HPO4], and 0.1 M citric acid [C6H8O7], both from AppliChem, Darmstadt, Germany) was used in all experiments.
UF resin
UF resin K 465 with a solid content of 66 percent and a pH value of 7.5 was used from BASF (Ludwigshafen, Germany).
Methods
13C NMR spectroscopy.—For 13C NMR measurements, 5 g of wood fibers was suspended in 150 ml of McIlvain buffer, pH 6.0, containing 100 U laccase or 100 U heat-inactivated laccase per g of wood fibers (cooked for 10 min), or without laccase and with either 10 mM VAN or HBA per g of wood fibers, or with no mediator. The mixture was stirred at room temperature for 30 minutes. The fiber samples were then separated from the supernatant by centrifugation (12,000 × g), washed two times with waterbidest, and centrifuged again at 12,000 × g. After water extraction, the samples were separated for different treatments. One part of the fiber samples was directly dried in a vacuum oven at 70°C for 12 hours (bi-distilled water [waterbidest]–extracted fibers). The second was suspended in 50 ml of dioxane, stirred for 12 hours, extracted with dioxane + 1 percent waterbidest three times by centrifugation, and dried in a vacuum oven for 12 hours (dioxane-extracted fibers). About half of the dioxane-extracted fibers were additionally extracted two times with 0.1 M sodium hydroxide (NaOH), centrifuged, and resuspended in 0.01 M hydrogen chloride (HCl) to adjust the pH to 7. The fibers were centrifuged and freeze-dried (dioxane- and alkali-extracted fibers). All treatments were repeated three times. The samples were measured using a Bruker Avance 400 MHz WB (Bruker, Bremen, Germany) at the Max-Planck Institute for Biophysical Chemistry, Division for NMR Based Structural Biology in Goettingen. 13C-NMR spectra were recorded in a region between 220 and 0 ppm.
ESR spectroscopy
For ESR spectroscopy 5 g of wood fibers was suspended for 30 minutes at room temperature in 150 ml of McIlvain buffer, pH 6.0, containing 100 U laccase or 100 U heat-inactivated laccase per g of wood fibers (cooked for 10 min), or without laccase and with either 10 mM VAN or HBA per g of wood fibers, or with no mediator. After incubation the samples were centrifuged at 12,000 × g. After centrifugation, 1 ml of the supernatant was pipetted into 4-mm CFQ-quartz tubes (RototecSpintec, Biebesheim, Germany). The tubes were transferred into liquid nitrogen (−196°C). Three repeats per sample type were analyzed. The spectra were taken in a magnetic field from 280 to 3,480 Gauss at −196°C by a Bruker ELEXSYS 500 (Bruker) at the Institute of Inorganic Chemistry, University of Goettingen.
Production of MDF in pilot scale
The production of MDF boards took place at the institute's own MDF pilot plant (Binos, Springe, Germany). The MDF pilot plant consists of a horizontal blender system with injectors for spraying binders onto fibers and a tube dryer into which the blended material will be conveyed. The tube dryer is directly followed by a cyclone from which the material is spread to a conveyer carrying it forward to the fiber bunker. Leaving the bunker, the fibers are strewed into a thick fiber mat by spreader rolls onto the form conveyer. There it is pressed in a prepress (without heat). The prepressed mat is finally moved to a hot press (Siempelkamp, Krefeld, Germany).
For each treatment, 20 kg (based on a moisture content of 0%) of wood fibers was incubated at room temperature by spraying on 1.5 liters/min of a total of 10 liters of buffer solution containing 100 U laccase per g of wood fibers with or without 10 mM mediator (VAN or HBA) per g of wood fibers onto 1.5 kg/min wood fibers. For all treatments, wood fibers were parallel blended with heat-inactivated laccase or heat-inactivated LMS. As general control samples, wood fibers were treated only with buffer. After blending, the fibers were directly moved into the tube dryer using 100°C warm air to reduce the fibers' moisture content to about 8 percent. Conveyed through the fiber bunker, the dried fibers were directly strewed onto the form conveyer to be prepressed into a one-layer mat. The handling procedure of blending, drying, and mat generating lasted about 30 minutes. Finally, the hot pressing to 10-mm-thick MDF boards with a molded density of 750 kg/m3 took place for 12 s/mm at 190°C. For the UF resin–bonded MDF, fibers were blended with 12 percent UF resin, relating to a fiber's moisture content of 0 percent. The making of MDF with UF resin was equivalent to the procedure described above except the drying was done by the tube dryer. For all treatments, no hydrophobing agent was used. The production of each type of MDF board was replicated six times. After conditioning (climate chamber at 20°C ± 2°C and 65% ± 5% humidity for 24 h), the MDF boards were sanded and trimmed (final size, 800 by 400 mm). The physical–technological properties, internal bond (IB) strength, thickness swelling (TS), and modulus of rupture (MOR, known as bending strength) were tested according to European Norms EN 310, EN 319, and EN 317 (European Committee for Standardization [CEN] 1993a, 1993b, 2003) using an universal test machine (Zwick/Roell ZO10, 10 kN test load, Zwick/Roell, Ulm, Germany) with software TestXpert (Version 10.1.1 from Zwick/Roell). Significant difference was calculated by using Tukey's honestly significant difference (HSD) test with a significance level of P < 0.05.
Results and Discussion
13C NMR spectroscopy of wood fibers
In this article we have analyzed the composition of the insoluble residue of different treated spruce fibers. To remove the unreacted components (laccase, HBA, and VAN) and to show structural changes more effectively, the samples were extracted with waterbidest and afterward partly with dioxane and successively with dioxane and alkali. The fibers were then characterized using 13C-NMR spectroscopy (Fig. 1), recording spectra between 220 and 0 ppm. The 13C-NMR spectra clearly imply that the incubation of wood fibers with laccase and LMSs resulted in chemical modification of insoluble lignocellulosic material (Table 1). Because no quantitative differences were detected in the region 60 to 45 ppm corresponding to O14CH3 (methoxyl) groups, the area of this signal was set to 100 percent in relative calculation of changes in other regions of the spectra. This may be caused in the low activation potential within methoxyl groups by laccase or LMS treatment (Euring et al. 2011a, 2011b). It should be noted that the aromatic groups of laccase-treated and even more LMS-treated wood fibers were decreased noticeably (region 160 to 195 ppm; Table 1). Aromatic groups are important components in the lignin structure of wood fibers (Lüdemann and Nimz 1974). A decrease of aromatic groups indicates a change in lignin and especially in the surface lignin, which might be caused by oxidation processes or by repolymerization during enzymatic exposure (Widsten 2002). For this work it is an important fact because the lignin is not supposed to be degraded within an incubation time of 30 minutes but rather essentially modified, e.g., to phenoxy radicals (Felby et al. 1997).
Citation: Forest Products Journal 63, 1-2; 10.13073/FPJ-D-12-00075
The content of carbohydrates/alkyl-OR (region 95 to 60 ppm) was decreased comparing untreated and laccase- or LMS-treated samples. In the aliphatic region 50 to 0 ppm, significant differences were observed in the water and dioxane-insoluble residues of laccase- and LMS-oxidized wood fibers. There were only low signals in the region of 220 to 160 ppm corresponding to carboxyl groups (COOH/C=O); however, a decrease of COOH/C=O between untreated and treated samples could be detected, which may be affected by the processes during lignin oxidation and depolymerization through laccase and LMSs. This is in good agreement with earlier data of Kharazipour et al. (1997), which showed a higher content of COOH/C=O in water-soluble products after laccase oxidation of wood fibers.
The enhanced solubility of laccase- and LMS-treated wood fibers likely effects change in the structure of lignin components. The data in Table 1 confirm the suggestion of Yamaguchi et al. (1992) that the lignin structure is loosened by laccase oxidation. In summary, all data reveal that the most structural changes have been found in the LMS-treated samples and especially with the mediator HBA, which shows the high potential of enhancing and intensifying laccase reactions on the wood fiber surface.
ESR spectroscopy
ESR spectra between 400 and 4,400 Gauss for wood fibers after different treatments were recorded (Fig. 2). The most noticeable differences between the samples appeared in the region between 3,000 and 3,500 Gauss. Focusing the control samples within this area, the peak agglomeration is effected by some simple organic radicals (org.), which have been generated during the thermomechanical pulping process of the wood fibers (Euring 2008). In all spectra of laccase- or LMS-incubated wood fibers, Cu2+ signals have been detected in the region between 3,000 and 3,250 Gauss resulting in the presence of laccase. Especially in the region between 3,250 and 3,500 Gauss, large signals of org. appeared, which are caused by laccase or LMS treatment. The run of the spectra in this region is characteristic for a high amount of org. on the wood fiber surfaces, which have been treated with laccase or LMS (Felby et al. 1997, Euring 2008). More precisely the org. in this case are mostly phenoxy radicals, which are generated by single-electron reactions between laccase and the wood fiber surface lignin (Felby et al. 1997). The intensities of the signals were even stronger once the LMS (VAN or HBA) was used to incubate the wood fibers and follow the order VAN < HBA. The results coincide well with the 13C NMR analysis, in which the results confirm structural changes in relevant, and especially lignin relating, components.
Citation: Forest Products Journal 63, 1-2; 10.13073/FPJ-D-12-00075
Production of MDF in pilot scale
The results of 13C NMR and ESR spectroscopy (see “13C NMR spectroscopy of wood fibers” and “ESR spectroscopy”) reveal that application of mediators in LMS induces an enhanced reactivity on wood fibers as it was shown in stronger structural changes and higher amounts of phenoxy radicals compared with laccase-treated fibers or control. For this reason, MDF boards were produced considering the most important facts relating to sample preparation for the spectroscopic analysis. Thus the amounts of enzyme and mediators, the buffer system, or the incubation time were all the same (see “Methods”). Laccase-bonded MDF were produced in a pilot plant by spraying laccase alone or one of the LMSs (VAN or HBA) onto wood fibers. Together with fiber drying and mat generation the process lasted 30 minutes, during which the laccase or an LMS had time to react on the wood fibers for activation of the surface lignin (incubation time). The board properties are presented in Figures 3 through 5. Additionally, MDF boards were also made from wood fibers blended solely with buffer and UF resin (samples “control” and “UF-resin,” Figs. 3 through 5). The production of UF resin–bonded MDF followed the same production process but without drying by the tube dryer (see “Methods”). MDF boards that were prepared from wood fibers treated with heat-inactivated laccase or treated with any one of the heat-inactivated LMSs all resulted in performances very similar to the control boards (data not shown).
Citation: Forest Products Journal 63, 1-2; 10.13073/FPJ-D-12-00075
Citation: Forest Products Journal 63, 1-2; 10.13073/FPJ-D-12-00075
The IB (Fig. 3) of control boards was very low, around 0.15 N/mm2, followed by laccase-bonded MDF, with about 0.4 N/mm2. Both LMS-bonded MDF types performed better. The boards made with LMS (VAN) just met the required norm of 0.6 N/mm2 defined in DIN 622‐5 (Deutsches Institut für Normung 2010) and according to EN 319 (CEN 1993b) for MDF with 750 kg/m3 raw density and a thickness of 10 mm. In contrast, using the mediator HBA in LMS led to a value of 0.65 N/mm2, which was equal to the UF resin–bonded MDF with an IB of 0.66 N/mm2.
When testing the TS after 24 hours, according to EN 317 (CEN 2003), MDF board thickness must not change more than 15 percent. Closely related to the IB values (Fig. 3), the control, laccase-, and LMS (VAN)–bonded MDF failed to meet the required value (Fig. 4). Nonetheless, the TS of LMS (VAN) lies closer to the TS of LMS (HBA), which had a value of 13.5 percent, being below the required standard, although no hydrophobing agent was used. The UF resin–bonded MDF fulfilled the requirement with a TS value 12 percent.
Citation: Forest Products Journal 63, 1-2; 10.13073/FPJ-D-12-00075
By evaluating the MOR values it was found that the control and laccase-bonded boards did not meet the required value of 22 N/mm2 (EN 310, CEN 1993a; Fig. 5). With 18 N/mm2, MOR values of laccase-bonded MDF came close to this value. The MDF bonded with LMS (VAN), LMS (HBA), and UF resin fulfilled the norm with values of 31, 34, and 33.5 N/mm2, respectively.
To summarize, the effect of the mediators VAN and HBA in LMSs after MDF production was clearly proven in higher physical–technological properties compared with the laccase-bonded and control MDF boards. The LMS (HBA) always performed better than LMS (VAN). This correlates well with the results of former studies, in which a combination of laccase with the mediator HBA was very efficient (Chandra et al. 2005; Euring et al. 2011a, 2011b). In comparison with the conventional binder system (UF resin) the properties were always close together, which shows the high potential of this LMS.
Conclusions
This research has shown that activation of wood fibers by LMSs has great potential as a suitable alternative to conventional binder systems for MDF production. MDF boards obtained after LMS (VAN) or LMS (HBA) treatment gained mechanical–technological properties that are comparable to those of UF resin–bonded MDF. Also, the LMS (HBA) always performed better than LMS (VAN). This result was also supported by evaluating 13C NMR and ESR spectra, which revealed the most structural changes and most intensified radicalization of organic compounds by using LMS (HBA). The turnover of aromatic groups and side chains on the wood fibers' surface as well as the generation of phenoxy radicals implied a positive effect for MDF production. Another important factor for MDF making is the timing. As an improvement of former studies in which wood fibers had to be incubated with laccase for 1 or 2 hours under constant conditions like temperature, pH value, moisture content (Kharazipour 1996, Felby et al. 2002, Widsten 2002), the production process with LMSs correlates well with conventional blending, drying, formatting, and hot pressing in MDF production. Further studies need to be done to improve performances, especially regarding the established process of MDF production with conventional binders. This could be variations in the concentrations of LMSs or in the technical process during blending with LMS. Also the cost efficiency of LMS should be surveyed in future works.
Carbon-13 nuclear magnetic resonance spectra in the region 220 to 0 ppm (representing different components) of wood fibers after different treatments: wood fibers (control) = only buffer or wood fibers + inactivated laccase or wood fibers with mediator (VAN or HBA) or wood fibers + inactivated laccase LMS with VAN or HBA (no differences in the spectra have been detected); wood fibers + laccase = incubation with laccase in buffer; wood fibers + LMS (VAN) or (HBA) = incubation with laccase with the respective mediator VAN or HBA in buffer. VAN = vanillic acid; HBA = 4-hydroxybenzoic acid; LMS = laccase-mediator system; rel. = relative peak height.
Electron spin resonance spectra in the scanning region 400 to 4,400 Gauss of wood fibers after different treatments: wood fibers (control) = only buffer or wood fibers + inactivated laccase or wood fibers with mediator (VAN or HBA) or wood fibers + inactivated laccase LMS with VAN or HBA (no differences in the spectra have been detected); wood fibers + laccase = incubation with laccase in buffer; wood fibers + LMS (VAN) or (HBA) = incubation with laccase with the respective mediator VAN or HBA in buffer. org. = organic radicals; VAN = vanillic acid; HBA = 4-hydroxybenzoic acid; LMS = laccase-mediator system; rel. = relative peak height.
Internal bond (IB) strength of 10-mm-thick medium-density fiberboard with a raw density of 750 kg/m3 and a size of 800 by 400 mm. Treatments of wood fibers: control = only buffer; laccase = laccase in buffer; LMS (VAN) or (HBA) = laccase with respective mediator (VAN or HBA) in buffer; UF resin = solely UF resin (12%). Data are presented as mean ± SD. Dashed lines mark the values required by the actual technical norm EN 319. Means marked with different letters (a, b, c, d, e) are significantly different at P < 0.05. LMS = laccase-mediator system; VAN = vanillic acid; HBA = 4-hydroxybenzoic acid; UF = urea-formaldehyde.
Modulus of rupture (MOR) of 10-mm-thick medium-density fiberboard with a raw density of 750 kg/m3 and a size of 800 by 400 mm. Treatments of wood fibers: control = only buffer; laccase = laccase in buffer; LMS (VAN) or (HBA) = laccase with respective mediator (VAN or HBA) in buffer; UF resin = solely UF resin (12%). Data are presented as mean ± SD. Dashed lines mark the values required by the actual technical norm EN 310. Means marked with different letters (a, b, c, d, e) are significantly different at P < 0.05. LMS = laccase-mediator system; VAN = vanillic acid; HBA = 4-hydroxybenzoic acid; UF = urea-formaldehyde.
Thickness swelling (TS) after 24 h in water of 10-mm-thick medium-density fiberboard with a raw density of 750 kg/m3 and a size of 800 by 400 mm. Treatments of wood fibers: control = only buffer; laccase = laccase in buffer; LMS (VAN) or (HBA) = laccase with respective mediator (VAN or HBA) in buffer; UF resin = solely UF resin (12%). Data are presented as mean ± SD. Dashed lines mark the values required by the actual technical norm EN 317. Means marked with different letters (a, b, c, d, e) are significantly different at P < 0.05. LMS = laccase-mediator system; VAN = vanillic acid; HBA = 4-hydroxybenzoic acid; UF = urea-formaldehyde.
Contributor Notes
The authors are, respectively, Research Associate, Honorary Professor, and Professor, Univ. of Göttingen, Faculty of Forest Sci. and Forest Ecology, Büsgen-Inst., Dept. of Molecular Wood Biotechnol. and Technical Mycology, Göttingen, Germany (meuring@gwdg.de [corresponding author], jtrojan@gwdg.de, akharaz@gwdg.de). This paper was received for publication in July 2012. Article no. 12‐00075.