SciELO - Scientific Electronic Library Online

Home Pagelista alfabética de periódicos  

Serviços Personalizados




Links relacionados


Journal of the Chilean Chemical Society

versão On-line ISSN 0717-9707

J. Chil. Chem. Soc. v.53 n.4 Concepción dez. 2008 

J. Chil. Chem. Soc, 53, N° 4 (2008) págs: 1722-1725





aForestry Science Faculty and Biotechnology Center, Universidad de Concepción, Casilla 160-C, Concepción, Chile. Fax: 56 41 224 7517; Tel: 56 41 2203850; E-mail:
bMASISA S.A. División Tableros. Concepción. E-mail:
cAnalytical and Inorganic Chemistry Department, Chemical Sciences Faculty, Universidad de Concepción, Concepción, Chile. E-mail:
dChemical Sciences Faculty and Biotechnology Center, Universidad de Concepción, Concepción, Chile.


Resin-free fiberboards were manufactured using industrial fiber from Pinus radiata activated by an oxidative treatment using the Fenton reaction (H202/ Fe(II)). A multivariate analysis was used to study the effect of fiber moisture content (MC), press temperature (T), and the H202/Fe(II) ratio on the board internal bond strength (IB). Using response surface methodology, a set of máximum IB conditions was obtained. Validation of these conditions which included 25% MC, 170°C press temperature and a H202/Fe(II) relation of 25 produced an optimal board with an IB strength of 0.888 MPa. Without the addition of sizing agents or other additives, the dimensional stability properties were 16% of thickness swell and 40% of the water absorption of control boards.

Keywords: Fenton reaction, Fiberboard, multivariate analysis


Wood based panels are traditionally manufactured via the application of thermosetting resins to fibers, particles or strands under pressure and temperature1-3. Boards consolidated with urea formaldehyde (UF)-type resins are characterized by their high mechanical strength, dimensional stability, hardness, resistance to microorganisms and abrasión 45. However, because synthetic resins are produced from high demand non-renewable resources, such as petrochemicals, bonding mechanisms for wood and lignocellulose composites using bio-based chemicals could be an attractive alternative. Considering environmental concerns with the use of formaldehyde-based resin, studies have been conducted to determine if self-bonding of wood fibers, without synthetic resins, can be achieved6-9. Most of these self-bonding processes are based on the generation of free radicais on the fiber surface so that the fiber can be pressed into panels without additional adhesives. Some of the applications involve thermosetting binder applications using oxidoreductases, such as laceases and peroxidases, for polymerization or cross-linking of wood components to promote binding 10,11. Direct bonding of wood surfaces with inorganic chemical oxidants has been studied12-15. As a complement to an added oxidant, ultrasound or photo- or electro-Fenton reactions (in situ generation of hydroxyl radicáis by radiation or electric currents) can be used16. The chemical reactions involved in these self-bonding systems are not fully understood, although the oxidative coupling of phenolic and non-phenolic lignin units contained in wood is either the principal, or at least one of the predominan! reactions leading to self-adhesion of lignocellulosic materials11-13. Phenolic free radical formation and subsequent coupling probably occurs precisely when the surfaces to be bonded are in cióse contact13. Some polysaccharide-to-polysaccharide or lignin-to-polysaccharide bonding may also occur during oxidation6,13.

The Fenton Reaction [eqn.(1)] has been proposed as an oxidative treatment for the self-bonding of material containing lignocellulose particles for the non-conventional manufacturing of restituted wood boards 16. When wood fibers are treated with hydrogen peroxide and a ferrous salt (Fenton reaction), fiberboard panels with enhanced internal bond strength can be produced due to the increase in interfiber bonds formed by hydroxyl radicáis. In addition, other reactive groups, and bonding generated by the Fenton reaction, may develop within the fiber under elevated pressure temperature conditions 14,15.

Considering that board adhesión using fiber activation by the Fenton reaction does not involve a resin curing process, traditional manufacturing process conditions could potentially be altered to reduce energy and costs associated with bonding. However, at present, there is little information on conditions required for appropriate bonding to occur. Experimental process conditions such as press temperature and fiber moisture content at the press entry must be determined for Fenton reaction activated fibers. Stofko et al. 13 patented a plywood bonding process employing a mixture of oxidants in solution, using a wide range of the press temperatures, between 80° and 200°C. Recent studies15,17 used temperatures between 170° and 200°C to produce fiberboard activated by the Fenton reaction and by a synthetic chelator mediated free radical system (CMFR), respectively. A patented process by Westermark et al. 16 employed press temperatures between 120° and 200°C The manufacture of fiberboards via an oxidative enzymatic process also used press temperatures between 180° and 200°C, to obtain acceptable mechanical strength properties 11,12,18.

There are no studies on Fenton reaction fiberboard production that define the moisture content at the time of press entry. In traditional resin curing processes, a moisture content of 10% to 13% is used for an optimal transfer of mass and energy 2. These conditions permit heat transfer to the center layers of the fiber mat while limiting strength loss as material compression and fiber deformation occurs 19. In earlier research 17 with wet process fiberboard production, high moisture contents (70% to 400%) were tested. The authors of this work proposed that and the CFMR system rapidly penetrated into fiber wall interior, producing free radicáis in fiber's cellulosic core, decreasing the internal bond strength 17. They suggested that lower MC fiber may help to elimínate this effect as the free radical reaction could be limited to the fiber exterior where high lignin fractions were located. Westermark et al.16 patented the manufacture of composite lignocellulosic materials activated with an oxidant treatment that employing a wood particle moisture content between 3% and 9% before hot pressing. In fibers activated with oxidative enzymes, no difference in internal bond strength was observed between boards that were pre-dried at 50°C and those flash dried before hot pressing, although they suggested that forced air drying may have decreased the dimensional stability properties of the final boards produced 11,12.

The objective of this study was to determine experimental process conditions for the production of fiberboard with sufficient internal bond strength for general use in dry end-use environments. The influence of MC, T and the H202/Fe(II) ratio were evaluated to determine the optimal interaction conditions appropriate for maximal IB strength in thin fiberboards as defined by EN 622-5.

Materials and Methods


In all the experiments, nanopure distilled water and industrial external mould reléase agent (Wurtz) were used. All reactants were grade p.a. 30%(w/ w) H202, ferrous sulfate heptahydrate FeS04x7H20 were purchased from Merck The Fe(II) solutions were adjusted to pH 3.5 with HC1. All the solutions were prepared immediately before each experiment.

Multivariate Analysis

Response surface modeling (RSM) was used for the multivariate analysis 20,21. The model was based on a two-level factorial design with three central points and their respective star points (central composite model). The data were analyzed using Modde 7.0 software. The variables were coded and normalized using unitary valúes where -1 was defined as the variable with the lowest valué and +1 as the one with the highest valué. With these external valúes for the variables, central points (coded as 0) were established and the analyses performed in triplícate. The star points were distributed at a distance of-2 and +2 from the central point.

The influence of the following three variables was studied: fiber moisture content (X1), press temperature (X2), and the H202/Fe(II) ratio (X3). The response evaluated was internal bond strength (Y) expressed in MPa. The variable range was as follows: X between 10% and 20%, X2 between 140°C and 180°C, and X3 between 10 and 20.

A second-order function describing the behavior of the system was determined using a múltiple linear regression (MLR). The optimized valúes of the variables analyzed were obtained from the SIMPLEX algorithm. Statistical validation was performed using ANO VA at a 95% confidence level.

The different valúes of MC, T, and the H202/Fe(II) ratio are provided in Table 1. The H202 charge was added to pro vi de a final concentration of 4.5% based on dry fiber. The Fe(II) amount (%) was changed according to the experimental design.

Deflbering and Fiber Treatment with the Fenton Reaction

Pinus radiata wood fibers produced by an Asplund Sunds process were obtained from an industrial board manufacturing plant (MASISA S.A. Chile). For pulping, the wood chips were preheated for 5 minutes, reaching a temperature of 160° C. The fibers were collected from a start up cyclone with 120% moisture content and dried in a forced air oven (100°C) until reaching 10% MC. The fibers were placed in a batch reactor and agitated to prevent dumping. The Fe(II) solution was injected followed by H202 following the experimental sequence used by Pino et al. 22 Following treatment, the fiber consistency was 55%. The fiber activated by the Fenton reaction was then agitated about 2 minutes and dried in an oven at 75°C.

Preparation and Evaluation of the Fiberboards

Activated fiber furnish was placed in a 30 cm x 30 cm frame and cold pressed at low pressure for a two-minute period. The fiber matt was then transferred to a stainless steel caul píate in a Dieffenbacher press. All boards were pressed using a closure rate of 20 s mm-1. Target thickness and density of the boards were 6111111 and 850 kg m-3, respectively. The boards were conditioned for three days at 60% relative humidity (RH) and 25°C before evaluating their IB strength according to EN- 319. Thickness swell and water absorption were determined according to the European board standard EN- 317.



Polynomial Response

From the results of the experimental design (Table 1), a polynomial function was obtained [eqn (2)] for the internal bond strength, which considers the relative importance of the different variables and their interaction. The coefficients were normalized according to codified variables value.

Where Y is the IB strength, X is the moisture content of the fiber mat, X2 is the press temperature, and X3 is the H2O2 /Fe(II) ratio. The error with a confidence level of 95% is also indicated.

Table 1 shows that the treatment with 15% MC, T of 160°C, and a H202/ Fe(II) ratio of 25, presents the highest internal bond strength valué.

For a better understanding of the importance of each variable, their coefficients in the polynomial were plotted in decreasing order (see Fig. 1). Considering the first-order and the interaction coefficient of each variable in Eqn. (2) and Fig. 1, it can be seen that, the main effects are due to the MC-H202/Fe(II) (X1X3) interaction, which is more important than the effect of moisture content (Xt) and the H202/Fe(II) ratio (X3) by themselves. The positive coefficients of these factors mean that they improve the IB strength as the valué of these variables increase. The negative coefficient of second-order of MC (X21) y T (X22) indicates a máximum region described by a parábola. The X1X2 and X2X3 interaction terms were omitted in the polynomial (reduced model), because its coefficients were not significative with P= and P= respectively. This imply that the interaction MC-T and T-H202, are not important for IB in the studied range. In this way T is only important as independent variables and do not modified the behavior of the other studied variable.


Fig. 2 shows a proportional relation of the polinomial predicted values and the exprimental values. The linear correlation coeficient (R2) was 0.992 indicating a good prediction by the model. The ANO VA test results over the regresión (table 2) shows the validity of the regression.


Fig. 3 illustrates the contour prediction graphs showing two variables, while the other are kept constant at a máximum value. The contour graphs are predicted by the polynomial function [eqn. (2)], where the área of the experimental domain is within the circle formed by the dotted line. In this area the validity of the model is guaranteed at confidence level 95%, but outside this it must be considered with caution. The máximum IB strength regions are shown by a 0.9 MPa, and the valúes chosen for each variable are shown by black points in their respective área (see Fig.3).


In Plot A of Fig. 3, which presents IB strength as a function of MC and T, a máximum región with MC ranging from 22% to 38% and T ranging from 145° to 190° C can be observed.

For moisture contents above 20%, the IB strength is above the European board standard EN 622-5 and coincides with the moisture content valúes used by other researchers11. However, increases in MC would improve the IB strength, which would cause a time increase in the press cycle for degasification23. However, in all the experimental treatments, the same rate of 20 mm s-1 was used, although press cycle times needed to be redistributed for the different moisture contents, avoiding possible delaminating and "blown" boards. Yelle et al. " used a rate of 76 mm s-1 for a dry thin fiberboard production in order to obtain similar valúes for internal bond strength.

In Plot B of Fig. 3, the internal board strength is illustrated as a function of T (X2) and the H202/Fe(II) (X3) ratio. The model's predicted the press temperatures between 160° and 180°C increased the IB strength for a H202/ Fe(II) ratio of 25.

Model Validation

The suggested of conditions from the SIMPLEX optimization to maximize IB strength correspond to the two treatments shown in table 3. In treatment A, a higher MC (25%) is required, although this valué is lower than the MC of 40% used by Yelle et al. " for fiberboard elaboration in the "dry process" with a CMFR system. Additionally, the H202/Fe(II) ratio is equal to the one used by Widsten et al. 15 with an acceptable IB strength.

In condition B, a higher T (180°C) is required, valué below the temperature traditionally used for resin curing in board industry24.

After analyzing the experimental data, the best suggested conditions for the IB strength are shown in table 3, also in this table the experimental valué of IB is shown.

The improved IB strength is explained by the superficial oxidation of the fibers with the Fenton Reaction preventing their diffusion to fiber interior due to the low MC. These results are in agreement with the conclusions obtained by Yelle et al. 17 and Qian et al.25 in similar systems.

Finally, the optimized boards were evaluated for thickness swell and water absorption according to European board standard EN 317, reaching valúes within the range indicated by standard EN 622-5 with respect to thin fiberboards for general use in a dry environment. These valúes were reached without sizing agent or other additives like that were used by Widsten ".


The multivariate analysis presented a statistically validated response surface that identifies conditions to improve the IB strength in fiberboards activated by oxidative treatment with the Fenton reaction with a 95% confidence level. These experimental conditions resulted in a board at lower press temperatures (170°C and 180°C) than used industrially (200°C to 220°C) and furthermore using fiberwood obtained at defibration temperature allowed to obtain below than those used routinely. The polynomial response indicates an important synergic interaction between MC and the H202/Fe(II) ratio with respect to the effects of the remaining variables. Fiberboards without resin, having IB strength according to standard EN 622-5, were obtained through fiberwood activation using the Fenton reaction.

Further studies should be performed to verify if the same conditions for máximum IB strength in P. radiata wood fiberboards activated with the Fenton reaction are valid in pilot plant.


This work was funded by Fondecyt Grant N° 1040619. The Support from MASISA S.A. División Tableros(Chile) is also acknowledged.


1.- O. Suchsland, and G. Woodson Fiberboard Manufacturing Practices in United States, USDAForest Service. Agricultural Handbook.Washington DC, pp.263; (1986)         [ Links ]

2.- T. Maloney, ForestProd. J. 46, 19, (1996).         [ Links ]

3.- J. Youngquist, A. Krzysik, P. Chow, and R. Meimban in Properties of Composite Panels. Paper and Composites from Agro-Based Resources, ed. R. Rowell, R. Young, J. Rowell, Lewis Publishers, New York; pp. 301, (1996).         [ Links ]

4.- Ch. B. Vick, Adhesive Bonding of Wood Materials. Wood handbook-Wood as an engineering material. General Technical Report FPL-GTR-113. USDA. Madison, (1999).         [ Links ]

5.- A. H. Conner in Urea-Formaldehyde Adhesive Resins, Polymeric Materials Encyclopedia, ed. J. C. Salamone, Florida, vol. 11; pp. 8496,(1996).         [ Links ]

6.- U.S. Pat, 2,388,487.         [ Links ]

7.- U.S. Pat, 20,030,186,036         [ Links ]

8.- U.S. Pat., 4,022,965.         [ Links ]

9.- U.S. Pat, 4,432,921.         [ Links ]

10.- A. Huttermann, C. Mai, and A. Kharazipour, Appl. Microbiol. Biotechnol,55, 387, (2001).         [ Links ]

11.- C. Felby, J. Hassingboe, and M. Lund, Enzyme Microb. Tech. 31, 736, (2002).         [ Links ]

12.- A. Kharazipour, A. Huttermann, and H. Ludemann, J. Adhes. Sci. Technol. 11,419,(1997)         [ Links ]

13.- U.S. Pat, 4,007,312        [ Links ]

14.- P. Widsten, Ph.D. Thesis, Helsinki University of Technology, 2002.         [ Links ]

15.- P. Widsten, P Qvintus-Leino, Tuominem and J. Laine,Holzforschung. 57,447, (2003).        [ Links ]

16.- U.S. Pat, 2005/0011621A1        [ Links ]

17.- D. Yelle, B. Goodell, D. Gardner, A. Amirbahman, P. Winistofer and S.Shaler, Forest. Prod. J. 54, 73, (2004).         [ Links ]

18.- A. Haars, A. Kharazipour, H. Zanker, and A. Huttermann in Adhesives from Renewable Resources, ed. R.W. Hemingway, A. Conner and S. Branham, Am. Chem. Soc, Washington D.C; pp. 126, (1989).         [ Links ]

19.- F.F. Kollmann, E. W. Kuenzi, and A.J. Stamm in Principies of Wood Science and Technology, ed. Springer-Verlag, vol. 2, pp703, (1975).         [ Links ]

20.- B. Barros, I. S. Scarminio, and R. E. Bruns in Como Fazer Experimentos.        [ Links ]

Pesquisa e Desenvolvimento na Ciencia e na Industria, ed. UNICAMP,Campinas,2nd edn; pp. 251-280, (2001).

21.- D. Contreras, J. Freer, and J. Rodríguez, Int. Biodeter. Biodegr. 57, 63, (2006).         [ Links ]

22.- F. Pino, R Mendonca, D. Contreras, J. P. Elissetche, J. Freer, J. Baeza and J. Rodríguez, Proceedings of Pre-Symposium: Chemistry and performance of composites containing wood and natural plant libres, Rotorua; pp. 167, (2005).         [ Links ]

23.- M. Kelly, Critical Literature Review of Relationships between Processing Parameters and Physical Properties of Particleboard. General Technical Report FPL-10, USDA. Madison WI, (1977).         [ Links ]

24.- USDA Forest Service, Fundamentáis of Composite Processing, Proceedings of a Workshop, General Technical Report FPL-GTR-149, Madison, (2003).         [ Links ]

25.- Y. Qian, B. Goodell, J. Jellison, and C. Félix, J. Polym. Environ. 12, 147,(2004).        [ Links ]


(Received: February 26, 2008 - Accepted: July 14, 2008)

Creative Commons License Todo o conteúdo deste periódico, exceto onde está identificado, está licenciado sob uma Licença Creative Commons