Lignin-Laccase-Mediator-Systems (LLMS) for the Production of Binderless Medium Density Fiberboards (MDF)

Medium density fiberboards (MDF) are produced mainly by urea-formaldehyde resins (UF) as binding agent, which are synthesized from finite fossil resources. Those boards may emit critical amounts of formaldehyde, which can influence the health of humans and animals. In recent times the wood panel board industry is looking for alternative glues which contain less or no formaldehyde. In order to avoid potential formaldehyde emissions altogether it would be preferable not to use binders which are formaldehyde based at all. One possibility is to use natural binders or to activate the wood fibers’ own binding forces by applying Laccase-Mediator-Systems (LMS). As a support of the LMS interactions with wood fibers technical lignin can be added. In this study it was found out that the addition of technical lignin intensified the fiber to fiber bindings. Two Lignin-Laccase-Mediator-Systems (LLMS) were analysed by Gel-Permeation-Chromatography and Cyclic voltammetry. Later the LLMSs were tested in the pilot scale production of MDF. The determination of the physical technological properties revealed that the LLMS treated MDF have higher dimension stabilities than only LMS treated MDF and approximately the same thickness swelling after 24 h. The results indicate that the application of LLMSs have a high potential for natural bonded MDF.


Introduction
Approximately ten million square meters of medium density fiberboards (MDF) are produced yearly in Europe (EPF, 2013). About 20% of the production costs can be attributed to the use of mainly petrochemical binders (Kharazipour, 2004). Here we refer to mainly the use of urea-formaldehyde resins (UF) and melamine-urea-formaldehyde resins (MUF) which are synthesized from finite fossil resources. Due to fluctuating raw oil prices, as well as the dependence of the wood-based materials industry on these petrochemical binding agents, there has been a noticeable price increase in wood-based products in recent years (Kharazipour, 2004). In addition the formaldehyde emissions from wood-based building materials are regarded critically and can be attributed to the use of formaldehyde based binders. At certain concentrations formaldehyde can induce irritation of the eyes, mucus membranes and skin (Yang et. al., 2001;IARC, 2004). In 2004 formaldehyde was classified as a carcinogen by the International Agency for Research on Cancer (IARC, 2004). Accordingly, increased research attention has focused on appropriate measures of reducing formaldehyde emissions in products in which it is contained (Oberdorfer, 2008).
At the beginning of 2009 in California, the first stage in reducing the formaldehyde emissions in the field of wood based products came into force. The "CARB" standards (CARB = California Air Resources Board) as an example, required for medium density fiberboards (MDF) a reduction of the formaldehyde emissions from the local norm of 0.3 ppm to 0.21 ppm (CARB, 2008). In 2011 the second stage in the reduction of emissions to an upper limit of 0.11 ppm came into force. For economic reasons, the European manufacturers of wood-based products oriented themselves on CARB and EN13986 with the emissions class E1 and E2 and also followed the recommendations of IKEA IOS MAT 003 as well as EPF-S (European Panel Federation-standard) in which the actual permitted formaldehyde emissions are lower than the top limit recommended by the CARB standards (SUBSPORT, 2013). It remains to be seen what the future will bring in terms of regulations and measures to be taken with regard to formaldehyde emissions, considering that in 2014 formaldehyde was upgraded in terms of carcinogenicity from category 2 (suspected of carcinogenic properties) to category 1B, and in terms of mutagenicity from category 3 (possibly causing genetic mutations) to category 2 according to EU regulation 605/2014 (2014). Category 1B means that formaldehyde now is considered to be a probable cause of cancer in humans and that the cancer causing effect has been proven in animals (EG 1272(EG /2008(EG , 2008. Category 2 in terms of mutagenicity, classifies formaldehyde as a material potentially harmful for humans in that it could possibly cause genetic mutations in germ cells (EG 1272(EG /2008(EG , 2008. Many manufacturers of petrochemical products containing formaldehyde are exploring new ways for handling formaldehyde, while taking into consideration all current regulations (Formacare, 2015). On the one hand discussions are taking place regarding a reduction of formaldehyde in these products and on the other a reduction in the use of formaldehyde in the actual production process (Wiechmann & Kohla, 2014). In local industry, no harmful emission levels for employees in the manufacturing process of wood-based materials with binders containing formaldehyde have been measured. However various possibilities to reduce emissions are being evaluated (Wiechmann & Kohla, 2014). In order to avoid potential formaldehyde emissions altogether it would be preferable not to use binders which are formaldehyde based at all.
In this respect several wood processing factories, in particular those for wood-based products, are currently occupied with testing low-emission or emission free binders. A possible solution can be found in the development of natural binding agents (Müller et al., 2007, Euring et al., 2011, 2013. The Lignin-Laccase-Mediator-System is a naturally based binding agent and can offer the wood-based product industry a chance to establish environmentally acceptable production methods. Lignin occurs in quantities of approximately 50 million tons worldwide as a byproduct of cellulose manufacture (15 million tons of lignin-sulfonates and 35 million tons of kraft-lignin) (Puls, 2009). In fact lignin is a mass product for which there are hardly any uses. According to Glasser (2009) only 2% of the technical lignins worldwide are used in industrial processes, although they belong to the most inexpensive of raw materials (approx. 0.15 €/kg). The remaining lignins are burnt for energy generation, or not used at all. For this and other reasons, technical lignins which are available widely and in large quantities, are the most promising polyphenol for the development of binding agents.
It makes sense to use these technical lignins where the natural lignins are already present on the wood fibers due to the high content of phenol groups. This is amongst others, the case with fibers which have been through a thermo-mechanical pulp process (TMP). The middle-lamella lignins in the wood fibers are incrusted on the fiber's surface after the thermo-mechanical process, an effect which is recognized as a significant chemical transformation (Sixta, 2006). The laccase can activate these surface lignins through oxidation resulting in free lignin fractions, which when hot pressed to fiber board cause the fibers to adhere to one another creating a fiber to fiber bond (Hüttermann et al., 2001, Kharazipour et al., 1997, 1998. The utilization of technical lignins and mediators as catalyzers should accelerate and intensify this process, which should lead to a more potent adhesion of the wood fibers (Euring, 2008). This study will investigate formaldehyde free wood fiber products, specifically medium density fiber-boards (MDF). These wood fiber products should be produced without the addition of conventional, petrochemical binding agents, using a Lignin-Laccase-Mediator-System (LLMS). The use of Laccase-Mediator-Systems has already been comprehensively tested. The results showed that when using an LMS of laccase and 4-hydroxicbenzoic acid, not all mechanical and physical properties were satisfactory (Euring et al., 2011(Euring et al., , 2013. The use of technical lignins in LLMS activates additional bonding strength. The wood fiber surface lignins which have been activated with the laccase and mediator should be supported by the use of technical lignin components. For the first time this study will test the applicability of the mediators caffeic acid (CA) and vanillyl alcohol (VA) in the production of MDF.
LLMS should make a stronger bond between the individual wood fibers possible. The implementation of LLMS should be integrated seamlessly into the conventional fiberboard production process. In preliminary stage to the production and testing the physical-technological properties of the MDF, analyses of lignin molecular weight through gel permeation chromatography (GPC) and also of the redox potential of Laccase-Mediator-Systems using a potentiostat, will take place. Those analyses are helpful to understand LMS or LLMS actions. The results and experiences e. g. the optimal LMS or LLMS concentrations will directly be transferred to the MDF production

Laccase
Novozym 51003 (from the company Novozymes, Bagsvaerd, Denmark) fermented from Trametes vilosa in recombination with Aspergillus oryzae, at hand in liquid form and brown in color due to stabilizer. Its specific activity was measured by determining the oxidation of ABTS (Matsumura et al. 1986). The activity of the laccase stock was about 1,000 U/ml.

Sodium Acetate Buffer
To carry out the potentiostatic trial, a sodium acetate buffer was used in order hold a pH level of 6 constant throughout the tests. The buffer was made by producing 0.1 M sodium acetate (NaAc) solution. Subsequently the pH value was brought to 5 using acetic acid.

Determination of the molecular weight of the lignin using Gel-Permeation-Chromatography (GPC).
Gel-permeation chromatography was used to determine the molecular weight distribution of the lignin prior to, and following, incubation with laccase or respectively laccase and mediator. The measurements were made using a high pressure liquid chromatography (HPLC) unit from Waters. The HPLC method used here is referred to as Gel-permeations chromatography as reference to the use of gel columns in the HPLC instrumentation. In this case it is a special method of column chromatography which is suitable in particular for separating molecules and therefore for establishing molecular weights.
The HPLC instrumentation is comprised of the following: Sample preparation was as follows: 0.5 g Ca-Ligninsulfonate were added to 27 ml McIlvain buffer (pH 6.0). Depending on the type of sample either 10 U/ml laccase or 10mM mediator were added to the buffer solution. The samples were incubated for 15, 30 and 60 minutes in order to compare the molecular weights. Following the incubation time the samples containing laccase were stopped through being briefly brought to the boil.
Subsequently 20 µl from each sample was taken, put into vials, and placed into the GPC, where measurement of the samples took place. The previously recorded values for the polystyrene standards were used as reference values for the evaluation software in calculation of the lignin molecular weights.
Measurements of both controls and samples took place in a UV detector at a wavelength of 254 nm. The eluent used was a water/acetonitrile mixture (80 : 20, v : v, pH 7.8). All results were recorded and calculated using software from Waters, and can be portrayed as spectra using the weight average molecular weight (Mw), the number average Molecular weight (Mn) and the poly-dispersity (Mw/Mn).

Electrochemical measurements of LMS by means of Potentiostat
Cyclic voltammetry is an electrochemical method of investigation used to analyze redox reactions. Not only simple parameters, such as redox potential, but also reaction kinetic parameters such as electron transfer stages and www.ccsenet.org/jmsr Journal of Materials Science Research Vol. 5, No. 2;2016 chemical reactions which are potentially taking place, can be recorded and analyzed by means of cyclic voltammetry.
For the purpose of this study the electrochemical measurements were used solely to analyze the redox potential of the mediators caffeic acid and vanillyl alcohol with laccase, in order to assess their reaction-kinetic suitability.
The potentiostat used was a PCSTAT101 from Metrohm Autolab B.V. from Utrecht in the Netherlands. It exhibits a three-electrode system with a carbon working electrode and a reference electrode which is immersed in a 3-molaric potassium chloride solution. The software used was also from Metrohm Autolab B.V. in the version 1.10.4.
For the analysis of the mediators caffeic acid and vanillyl alcohol with a potentiostat, the mediators were first dissolved in a sodium acetate buffer and diluted to a concentration of 0.2 mM. The laccase used in the trials diluted 100 times. For the samples with lignin 2 mM Ca-Ligninsulfonate were added to evaluate the effect of the lignin within the LLMS expecting changing redox-potentials. The measurements were made in a range from 100 mV to 1000 mV with a feed velocity of 5 mV/s. The output potential was held for 10s, before the voltage change occurred. 5 mL of mediator solution was used in each measurement. At first three measurements as a control were taken using pure mediator solution. In order to record the interaction between mediator and laccase, the laccase solution was dripped onto the working electrode. After the solution had completely dried out, a measurement was taken using a 5 ml mediator solution and the same parameters as with the pure mediator solution. Here it is important to note that after each trial the electrode was cleaned completely. Each measurement was repeated three times. The results were reported in a voltage-current strength diagram. Press Time: 12 s/mm The LLMS treated wood fibers were directly transferred from the Blender to the tube dryer to guarantee an uninterrupted production of the fibers to a fiber mat formation and then pressing to MDF (equivalent to the industrial process of MDF).

Production of LLMS bonded
The MDF were produced using the following steps.

1) Blending of fibers with LLMS
The combination of fibers and each LLMS in buffer solution (McIlvain) was achieved using the Blender process in the MDF pilot plant of the Biotechnikum (Büsgen-Institute, Goettingen).
2) Drying of the treated fibers As known from previous studies a fiber moisture content of 10-15% was aimed for in the drying process, in order to support the natural binding forces, but also to avoid a loss of MDF quality associated with higher moisture levels (>15%) in the following hot-pressing process (Euring, 2008, Felby et al., 2011, Euring et al. 2011). The drying temperature was adjusted so that the incubation process with LLMS would not be stopped or interrupted.

3) Fiber Mat formation
Monolayer fiber mats were generated.

4) Pre-pressing and hot-pressing
Hot-pressing of the fiber mat followed in the hot press at 190 °C for 12 s/mm board thickness after pre-pressing. Both temperature and time were determined as optimal parameters for pressing of LLMS bonded MDF in pilot-scale. Each type of MDF was replicated six times.

5) Determining the physical-technological properties of the pilot scale MDF
Determination of the following properties was carried out after a conditioning phase of each MDF in a standard climate: -Determination of bending strength (EN 310, 1993), sample size: 5 x 20 x 0,1 cm³, 6 samples/MDF -Determination of internal bond strength (EN 319, 1993), sample size: 5 x 5 x 0,1 cm³, 8 samples/MDF -Determination of the hygroscopic behavior of the material (thickness swelling and absorption) (EN 317, 1993), sample size: 5 x 5 x 0,1 cm³, 8 samples/MDF Significant difference was calculated by using Tukey's HSD (Honestly Significant Difference) test with a significance level of P < 0.05. 6) Production of the reference MDF boards.
The boards used as references were produced in dry-process using the same process parameters and target sizes, as used in the production with LLMS treated wood fibers. The reference boards were produced with inactivated LLMS (cooking for 10 minutes), only with laccase, with technical Lignin and laccase, with simple LMS and with 10% UF resin like in a typical MDF in Europe. Lastly the physical-technological properties were analyzed and compared. For the first time in the course of this study lignin molecular weights were determined using a water-based eluent in a GPC instrumentation unit. Up till now GPC runs of lignin were only carried out using dangerous solvents such as tetrahydrofuran (THF) (Euring, 2008, Majcherzyk & Hüttermann, 1996. During the trials of Ca-Ligninsulfonate treated with laccase and laccase mediator with incubation times of 15, 30 and 60 minutes, which were successful right from the start, it was observed that both treatment methods resulted in an increase in lignin molecular weight (see Figure 3). Untreated Ca-Ligninsulfonate kept a molecular weight of 55,000 Dalton throughout the entire analysis time, while in the case of incubation with laccase the weights increased to over 70,000 Dalton with the highest value being reached after 30 minutes. It can be assumed that the increase in molecular weight is due to oxidation of the phenol group of the lignin to phenoxy radicals (Ortner et al., 2015), which would then be available to facilitate the subsequent bonding of the fibers to MDF. After treatment with LLMS (CA) and LLMS (VA) even higher lignin molecular weights were achieved, which can be associated to a higher effectiveness of LLMS than laccase alone. With results almost as high as 100,000 Daltons after 30 minutes incubation with LLMS containing caffeic acid, this is the most effective LLMS. The GPC tests with LMS (CA) and LMS (VA) resulted in only low molecular weights because here there is no lignin present. However, the detected molecular weights are a result of the partly phenolic structures of the mediator. The fact that all laccase, LMS or LLMS samples showed molecular weights are highest at 30 minutes incubation and sink after 60 minutes incubation can be explained by the fact that phenoxy radicals possibly react with each other and are deactivated. Thus it can be seen that the optimal incubation time is that of 30 minutes, especially when considering the process of MDF board production which usually requires 30-40 minutes from the gluing to fiber mat formation and pressing (Euring et al., 2011).

Cyclic Voltammograms of LLMS
The tests with caffeic acid show a significant change in the curve after addition of laccase (Figure 4). The electric current is elevated during oxidation and a shift into the negative range occurs during reduction. The cyclic voltammogram of the tests with caffeic acid show a significant interaction between laccase and mediator (CA and LMS (CA). It can be seen that there is a reduction of current during the oxidation process and on the other hand a shift of the reduction peak into the negative. Together with Ca-Ligninsulfonate the currents become more intensive caused by the presence of lignin (Lignin + CA and LLMS (CA)). The oxidative effect of laccase is indicated by these results. Furthermore, the oxidation and therefore the radical formation are partly reversible. There is virtually a reversible electron transfer present, and the so formed radicals are partially stable. The potentiostatic test of vanillyl alcohol as mediator shows a decrease of measured current during oxidation and stronger negative current during the reduction process independently if in presence of lignin ( Figure 5). The cyclic voltammographic results of the mediator vanillic acid, display similar results to those with caffeic acid. The oxidative effect of the laccase on the mediator can be recognized by the decreased current during the oxidation process. The change in current during the reduction is however slightly weaker than with caffeic acid, which indicates less stable radicals (Diaz Gonzalés et al., 2009). Based on the higher stability of the radicals, caffeic acid would appear to possess a higher potential to function in LLMS (Siebert, 2015). The resulting bending strengths of the MDF differed depending on the treatment of the fibers ( Figure 6). In both LLMS, (CA) and (VA), the required bending strength norm of at least 23 N/mm² was exceeded by far, and almost equaled that of the control MDF with 10% UF. When applying only LMS, the bending strengths fluctuate around the required norm. The control boards, the laccase and also the laccase-Ligninsulfonate bonded MDF did not reach the required norm value. The values for internal bond strength of MDF bonded with 10% UF exceed those of the required norm of 0.6 N/mm² (Figure 7). The required norm is also reached distinctly with LLMS with the mediator caffeic acid or vanillyl alcohol. In the case of the MDF with the two different LMS, it didn't quite reach the norm values for IB and therefore do not comply with the required norm. The control samples and the laccase bonded MDF fall short of the norm. Thickness swelling values after 24 h water immersion are still too high in the LMS or LLMS bonded MDF (Figure 8).

Physical-Technological Properties of MDF Made with LLMS in Pilot Scale
However the values are almost as good as those with 10% UF bonded MDF. All other MDF in the study showed thickness swelling values which are too high. As has been discovered in earlier studies, the hydrophilic properties of technical lignins have a negative effect on water absorption (Müller et al. 2007). The use of hydrophobic agents would be of assistance in this case. Another possibility could be to choose higher pressing temperatures than 190 °C to improve the physical-technological properties of the boards, but in industrial process temperatures of 200 °C and above are considered uneconomic, as well as too long pressing times (Dunky & Niemz, 2002).

Conclusion
The study demonstrates that the Lignin-Laccase-Mediator-System as a natural bonding agent is very suitable for the production of MDF. In both the analytical tests and in the production of LLMS bonded MDF, significant advantages over the previous use of LMS can be seen (Euring et al., 2011(Euring et al., , 2013.
The GPC analysis results of LLMS in comparison with the previously investigated LMS already show a significant increase in the lignin molecular weight after 15 minutes incubation, with the highest results after 30 minutes. 30 minutes incubation coincides with the duration of MDF production from blending to the pressing process of the completed product. In production terms therefore, the use of LLMS in the production of MDF is appropriate.
As shown in the positive technological mechanical properties of the so produced MDF, laccase and mediator activated wood fiber surface lignin are supported by the additional phenoxy radicals contributed by the technical lignins, and therefore facilitate stronger bonding of the individual wood fibers. Caffeic acid proved to be more effective than vanillyl alcohol in LLMS. Further tests need to follow in order to optimize the LLMS particularly with regard to thickness swelling properties.