Eur. J. Wood Prod. (2011) 69:521–525DOI 10.1007/s00107-010-0480-4

O R I G I NA L S O R I G I NA L A R B E I T E N

Properties of furfurylated wood (Pinus pinaster)

Bruno Esteves · Lina Nunes · Helena Pereira

Received: 14 September 2009 / Published online: 9 September 2010© Springer-Verlag 2010

Abstract Sapwood samples of Pinus pinaster wood weretreated with a 70% furfuryl alcohol mixture. Weight percentgain (WPG), equilibrium moisture content, dimensional sta-bility, MOE, bending strength, hardness, density and dura-bility were determined.

WPG was on average 38%. Equilibrium moisture con-tent decreased more than 40%. Dimensional stability in-creased reaching an ASE of 45%. MOE was little affected bythe treatment but bending strength increased by about 6%.Hardness increased by about 50%, and density by 37%.Mass loss due to Postia placenta and Coniophora puteanadecreased by 96% and 86%, respectively. Furfurylation ofPinus pinaster wood shows an interesting potential to im-prove the wood quality for solid timber products.

Eigenschaften von furfuryliertem Kiefernholz(Pinus pinaster)

Zusammenfassung Kiefernsplintholzproben (Pinus pinas-ter) wurden mit einer 70% Furfurylalkohol-Mischung be-handelt und die prozentuale Massezunahme (WPG), die

B. Esteves (�)Centro de Estudos em Educação, Tecnologias e Saúde, ESTV,Instituto Politécnico de Viseu, Av. Coronel José Maria V.de Andrade, Campus Politécnico, 3504-510 Viseu, Portugale-mail: bruno@demad.estv.ipv.pt

L. NunesLaboratório Nacional de Engenharia Civil, Av. do Brasil, 101,1700-066 Lisboa, Portugal

H. PereiraCentro de Estudos Florestais, Instituto Superior de Agronomia,Universidade Técnica de Lisboa, Tapada da Ajuda,1349-017 Lisboa, Portugal

Gleichgewichtsfeuchte, die Dimensionsstabilität, der Elas-tizitätsmodul, die Biegefestigkeit, die Härte, die Dichte unddie Dauerhaftigkeit wurden bestimmt.

Die prozentuale Massezunahme lag im Durchschnitt bei38%. Die Gleichgewichtsfeuchte nahm mehr als 40% ab.Die Dimensionsstabilität nahm zu und erreichte ein Quellre-sistenzvermögen (ASE) von 45%. Die Behandlung hatte nurwenig Einfluss auf den Elastizitätsmodul, während die Bie-gefestigkeit um 6% zunahm. Die Härte erhöhte sich um ca.50% und die Dichte um 37%. Der Masseverlust durch Postiaplacenta und Coniophora puteana ging um 96% bzw. 86%zurück. Die Furfurylierung von Pinus pinaster Holz zeigtein hohes Potential zur Verbesserung der Holzqualität vonMassivholzprodukten.

1 Introduction

Several wood modification processes have emerged in thelast few years, the most important one being thermal, chem-ical, surface and impregnation modification. Furfurylationis often considered as an impregnation modification since itis believed that furfuryl alcohol is not chemically bound towood and instead, it is its polymerization inside the woodthat inhibits the water molecules to reach the wood polysac-charides, thus reducing equilibrium moisture content andincreasing dimensional stability and durability. Neverthe-less Lande et al. (2004a) suggested grafting between ligninand furfuryl alcohol whereas Venås et al. (2006) found noproof of such linkages on treated wood using ATR-IR spec-troscopy. Recent NMR studies by Nordstierna et al. (2008)showed that some model compounds similar to lignin formcovalent bonds with poly (furfuryl alcohol) which mightconfirm the results reported by Lande et al. (2004a).

522 Eur. J. Wood Prod. (2011) 69:521–525

The use of furfuryl alcohol to modify wood was first sug-gested by Stamm (1977). Wood was treated with furfurylalcohol solutions at 90% and dimensional stability, fungaldurability and some mechanical properties were increased.The process used zinc chloride as catalyst and was notsuitable for lumber-size material due to a chromatographicseparation when the catalyst solution penetrated the wood.More recently. Schneider (1995) and Westin (1995) devel-oped similar processes with new catalysts based on cycliccarboxylic anhydrides.

Epmeier et al. (2004) studied some mechanical proper-ties of differently modified wood and concluded that MOEwas not significantly changed by furfurylation and MORwas slightly increased. Furfurylated wood with 48% WPG(weight percent gain) showed stiffness stabilization effi-ciency (SSE) of 40–70% and an increase of over 100%of the Brinell hardness for furfurylated wood with 92%WPG.

The impact bending strength is decreased by the treat-ment. Lande et al. (2004b) reported 57% decrease with fur-furylated southern yellow pine wood with 70% WPG andEpmeier et al. (2004) reported a 75% decrease for furfury-lated wood with 48% WPG.

Lande et al. (2004b) reported that wood dimensional sta-bility of furfurylated wood increased with the increase inweight percent gain. Epmeier et al. (2007a) reported a reduc-tion of about 30% in EMC (equilibrium moisture content)and in the swelling strain with an increase in the swellingcoefficient, while no significant change was observed for thedynamic MOE. Epmeier et al. (2007b) also concluded withmodified Scots pine that furfurylation reduces creep deflec-tion and relative creep.

Furfurylated wood is slightly more resistant to weather-ing than untreated wood (Temiz et al. 2007). In relation todurability, furfurylated wood is resistant to brown and whiterot (Gobakken and Westin 2008), termites (Hadi et al. 2005)and marine borers (Westin et al. 2006).

Wood furfurylation might have a promising future, sincethe price of furfuryl alcohol might go down substantiallyin the near future because it can be obtained from the sec-ondary products in the production of bioethanol.

With this first work on furfurylated Pinus pinaster woodit is intended to determine if the species is suitable for thiskind of treatment and what are the main properties of thetreated wood. P. pinaster (maritime pine) is an importantsoftwood species in southern Europe but its timber valueis low due to poor performance mainly regarding dimen-sional variation and sapwood biodegradation. The furfury-lation treatment therefore aims at increasing wood stabilityand durability without considerable loss of mechanical prop-erties required for structural applications.

2 Material and methods

2.1 Treatment

Two pine boards (Pinus pinaster Aiton) with 2 m length,20 mm thickness and approximately 300 mm width werecut from the same tree and air dried until reaching approx-imately 10% equilibrium moisture content. Each board wascut into four boards with dimensions 1000 × 150 × 20 mm3

and numbered from 1 to 8. All the boards included sapwoodand heartwood. Boards 1, 5 and 7 were used for splittingprior to the treatment and the resulting samples were markedas heartwood (h) or sapwood (s). The transition zone on theboundary of heartwood and sapwood was taken out to obtainpure sapwood and heartwood pieces.

Boards 3 and 4 were kept for control and the remainingboards were treated with a furfuryl alcohol mixture (FA 70mix) at Kebony (Norway). The treatment was carried outin an autoclave by a vacuum and pressure stage and subse-quently cured and dried in a vacuum drying kiln. After thetreatment the samples were kept in a conditioned room forthree weeks and the equilibrium moisture content was deter-mined. The weight percent gain was calculated in relation todry wood.

2.2 Equilibrium moisture content and dimensional stability

For the determination of equilibrium moisture content anddimensional stability ten cubic samples with 20 × 20 ×20 mm3 were cut from the treated and untreated boards. Thesamples were subsequently conditioned in a controlled envi-ronment at 20°C and 35, 65 and 85% relative humidity for atleast 4 weeks in each relative humidity and until mass varia-tion was less than 5% in two consecutive days. Mass was de-termined and the equilibrium moisture content (EMC) wascalculated.

The samples dimensions were measured in radial and tan-gential directions and the dimensional stability of the treatedsamples was calculated as Anti Swelling Efficiency (ASE)for the three values of relative humidities (ASE35, ASE65,ASE85) as reported earlier for heat treated wood (Esteves etal. 2007).

2.3 Density

Four cubic samples with 20×20×20 mm3 from treated anduntreated wood were conditioned at 65% relative humidityfor 4 weeks. Density was determined by measuring the sam-ple dimensions and weight.

2.4 Hardness

Hardness was measured according to ISO 3350 standard(1975) with small adaptations. The force measured was that

Eur. J. Wood Prod. (2011) 69:521–525 523

required to embed a 11.28 mm steel ball into wood to a quar-ter of its diameter (2.82 mm), rather than to half of its diam-eter as mentioned in ISO 3350 because untreated pine woodwas very soft.

2.5 MOE and bending strength

MOE and bending strength were determined with samplesof 340 × 20 × 20 mm3 (axial × radial × tangential) by athree point bending device. MOE measurements were madeusing a constant velocity of 0.3 mm/min and for bend-ing strength the velocity was estimated to cause rupture inabout 3 min. Both properties were determined according toNP-619 (1973) as:

MOE (N/mm2) = �F ∗ L3

�x ∗ 4 ∗ b ∗ h3

Bending strength (MPa) = 3 ∗ F ∗ L

2 ∗ b ∗ h106

where F is the load on rupture measured in N, �F�x

is theslope of the elastic zone in N/mm, L is the arm length, h theheight and b the width, all expressed in mm.

2.6 Durability

Resistance to decay was evaluated following the method de-scribed in CEN/TS 15083-1 (2005). Two fungi were used,a wet rot fungus Coniophora puteana (Schum.:Fr.) Karst.(BAM Ebw.15), and a brown rot, Postia placenta (Fr.) Larset Lomb. (FPRL 280). Sixteen replicates of furfurylated pinewood samples and twelve replicates of untreated pine (ascontrols) with 15 mm × 25 mm × 50 mm, in radial, tangen-tial and longitudinal directions, respectively, were used foreach fungi.

To measure the mass loss of modified and non-modifiedwood, which was not caused by the fungi, three pairs ofblocks were placed in Kolle flasks with culturing media butwithout any fungus under the same sterile conditions. Alltest devices were placed in a conditioned room at 22 ± 1°Cand 70 ± 5% relative humidity for sixteen weeks. Afterthis period, the wood blocks were removed from the flasks,cleaned with a cloth to remove the hyphae from the surfaceof the specimen and weighed before and after drying in anoven at 103°C for 24 hours. Equilibrium moisture contentand mass loss were determined and the mass loss corre-sponding to the fungal biodegradation was determined bycorrection with the mass loss in the blank test, according to:

Equilibrium moisture content (%)

= final wet mass − final dry mass

final dry mass× 100

Mass loss (%) = initial dry mass − final dry mass

initial dry mass× 100

Corrected mass loss (%)

= mass loss − mass loss correction factor

The mass loss correction factor corresponds to the aver-age weight loss percentage of test pieces of wood in the testwithout the fungus. In the case of increasing mass in testwithout fungus this factor has a negative value.

3 Results and discussion

The penetration of furfuryl alcohol in the cell wall is neces-sary to impart new properties on the treated wood. In mar-itime pine wood, furfuryl alcohol penetrated in the wood;with a weight percent gain (WPG) ranging between 37.5–39.8% being on average 38%. In some of the heartwoodsamples the WPG was significant but there was an unevendistribution of the chemical. Even when the WPG was highit was clear that there was a gradient and that almost none ofthe furfuryl alcohol reached the middle of the board. Moreresearch is needed to determine the extent to which heart-wood of this species can be treated by furfurylation.

These results suggest that sapwood from this species maybe furfurylated and that it absorbs more furfuryl alcohol thanreported earlier for other pine species (Lande et al. 2004b).

Equilibrium moisture content (EMC) of furfurylated pine(39.8% WPG) decreased in the sapwood in relation to un-treated wood from 8.9% to 5.1% (at 35% relative humid-ity), from 12.9% to 7.3% (at 65%) and from 17.3% to 9.0%(at 85%), corresponding to 43%, 43% and 48% change inrelation to untreated wood (Table 1). These results show aclear decrease on EMC, however lower than the reportedEMC decrease from 8% to 2–4% at 30% RH for furfury-lated Scots pine with 48% WPG (Epmeier et al. 2004). Thedifferences might be due to the smaller WPG (39.8% against48%) and not because of the species. In fact, later reports byEpmeier et al. (2007a) showed much smaller reduction inEMC of Scots pine with larger samples. The larger size ofthe treated samples may also be a reason for the differences.

Nevertheless, the EMC reduction was higher than theone reported for other modification treatments. For exam-ple an equal reduction on EMC for thermally treated Pinus

Table 1 Equilibrium moisture content of untreated and treated pinewoodTab. 1 Gleichgewichtsfeuchte von behandeltem und unbehandeltemKiefernsplintholz

Sample EMC(35%)

EMC(65%)

EMC(85%)

Untreated Average 8.9 12.9 17.3

St. Dev. 0.1 0.2 0.3

Treated Average 5.1 7.3 9.0

St. Dev. 0.1 0.1 0.1

524 Eur. J. Wood Prod. (2011) 69:521–525

pinaster wood could only be attained at high temperatures(210°C) and treating times (12 h) leading to a considerablestrength loss (Esteves et al. 2007).

3.1 Dimensional stability

Dimensional stability was measured as ASE at relative hu-midities of 35%, 65% and 85%, representing the normalconditions wood samples face outdoors. The dimensionalchanges and the respective ASE values are presented in Ta-ble 2. For all relative humidities the dimensional changeswere lower for the furfurylated wood with an increased di-mensional stability. ASE ranged between 35.6% and 41.8%at 35% RH, 29.0% and 43.4% at 65% RH and 31.4% and45.1% at 85% RH for radial and tangential directions, re-spectively.

The reduction in the dimensional changes was higherin the tangential direction and therefore, the treatment de-creased wood anisotropy since these changes are generallyhigher in the tangential direction. The ratio of tangential toradial dimensional changes was on average 1.7 for untreatedwood and 1.4 for treated sapwood.

The results reported by Epmeier et al. (2004) showed aclear change in dimensional stability for furfurylated Scotspine (48% WPG) with ASE reaching 50%. However, it isdifficult to compare results since these authors determinedthe volumetric ASE between 30% RH and 90% RH. Ifthe volumetric ASE is determined between 35% RH and85% RH an ASE of about 46% is obtained for furfurylatedwood with 39.8% WPG which is not much different fromthe Scot pine one presented by Epmeier et al. (2004).

Higher ASE values are possible with higher WPG, as re-ported by Baysal et al. (2004) for furfurylated Scots pine andJapanese cedar. Nevertheless there is a limit above which afurther increase in WPG does not lead to a significant higherdimensional stability. According to Lande et al. (2004b),ASE was about 50% and 70% for WPG of 32% and 47%,

Table 2 Dimensional changes in radial and tangential directions andAnti Shrinking Efficiency (ASE) of untreated and treated wood at rel-ative humidities of 35%, 65% and 85%Tab. 2 Formänderung in radialer und tangentialer Richtung sowieQuellresistenzvermögen (ASE) von unbehandeltem und behandeltemHolz bei relativen Luftfeuchten von 35%, 65% und 85%

Sample Sapwood

Radial [%] Tangential [%]

35% 65% 85% 35% 65% 85%

Untreated Average 1.8 2.7 3.5 2.9 4.6 6.2

St. Dev. 0.4 0.4 0.5 0.4 0.5 0.4

Treated Average 1.2 1.9 2.4 1.7 2.6 3.4

St. Dev. 0.4 0.4 0.5 0.4 0.5 0.5

ASE (%) 35.6 29.0 31.4 41.8 43.4 45.1

respectively but the increase in WPG above this level didnot significantly improve dimensional stability (75% ASEfor 125% WPG).

3.2 Mechanical properties

The treatment did not change significantly the static bend-ing properties of Pinus pinaster sapwood (Table 3). MOEwas similar for untreated and treated wood and bendingstrength increased by 6%. Analogous results were reportedby Epmeier et al. (2004) for furfurylated Pinus sylvestriswood.

There was a clear increase in sapwood hardness on ra-dial and tangential surfaces of about 50% (Table 4), slightlyhigher for the radial surface. The increase is higher than theone reported earlier for Scots pine for which Brinell hard-ness increased by about 20% for 32% WPG and by about30% for 47% WPG (Lande et al. 2004b). It is likely thathigher WPG will significantly increase the surface hardnessas reported by Epmeier et al. (2004) who obtained a Brinellhardness increase of 100% for furfurylated wood with 92%WPG.

Sapwood density (at 65% relative humidity) increased byabout 36% similar to the one reported by Epmeier et al.(2007a) for furfurylated Scots pine. The density increase issmaller than the WPG which shows that there is also someincrease in cell wall volume leading to wood swelling.

Table 3 Mechanical properties of untreated and treated pine sapwoodTab. 3 Mechanische Eigenschaften von behandeltem und unbehan-deltem Kiefernsplintholz

Sample MOE[MPa]

Bending strength[MPa]

Untreated Average 10,924 166

St. Dev. 1021 14

Treated Average 10,833 176

St. Dev. 907 12

Table 4 Hardness and density of untreated and treated woodTab. 4 Härte und Dichte von unbehandeltem und behandeltem Holz

Sapwood

Sample Hardness [N] Density[kg/m3]Rad Tang

Untreated Average 4505 4363 0.62

St. Dev. 322 668 0.02

Treated Average 7013 6534 0.85

St. Dev. 416 454 0.01

Eur. J. Wood Prod. (2011) 69:521–525 525

Table 5 Durability of untreated and treated wood (mass loss in per-cent)Tab. 5 Dauerhaftigkeit (prozentualer Masseverlust) von unbehandel-tem und behandeltem Holz

Fungus

Sample Coniophoraputeana

Postiaplacenta

Untreated Average 5.69 28.23

St. Dev. 2.13 3.99

Treated Average 0.78 1.11

St. Dev. 0.36 0.38

3.3 Durability

Durability is probably the most important property in treatedwood. Some wood modifications have shown an increasedsusceptibility to brown rot attack, mainly by Postia placenta(Rapp et al. 2008). With the furfurylation process there wasa clear decrease in mass loss due to fungal attack for bothtested fungi. Mass loss due to Postia placenta and Conio-phora puteana decreased from 28.23% to 1.11% and from5.69% to 0.78%, respectively, corresponding to a 96% and86% decrease in relation to untreated wood (Table 5). Landeet al. (2004b) also reported a decrease in mass loss of Pinussylvestris wood due to Postia placenta. Nevertheless, sincethe tests were made with furfurylated wood with about 25%and 125% WPG different from this study and with woodburied in different kinds of soil for a period of 16 weeks, theresults cannot be directly compared.

4 Conclusion

Furfurylation of Pinus pinaster wood shows an interestingpotential to improve the wood quality for solid timber prod-ucts. Furfurylation improved the wood behavior in relationto moisture by decreasing the wood equilibrium moisturecontent and increasing its dimensional stability and reduc-ing anisotropy, without significant effect on bending proper-ties. Hardness was significantly increased by the treatment.Durability increased significantly, since there was a clear de-crease in mass loss due to fungal attack, both for wet rot andbrown rot.

References

Baysal E, Ozaki S, Yalinkilic M (2004) Dimensional stabilization ofwood treated with furfuryl alcohol catalysed by borates. WoodSci Technol 38(6):405–415

CEN/TS 15083-1 (2005) Durability of wood and wood-based prod-ucts. Determination of the natural durability of solid wood againstwood-destroying fungi, test methods. Part 1: Basidiomycetes.CEN, Brussels

Esteves B, Velez Marques A, Domingos I, Pereira H (2007) Influenceof steam heating on the properties of pine (Pinus pinaster) andeucalypt (Eucalyptus globulus) wood. Wood Sci Technol 41:193–207. doi:10.1007/s00226-006-0099-0

Epmeier H, Johansson M, Kliger R, Westin M (2007a) Bending creepperformance of modified timber. Holz Roh-Werkst 65:343–351

Epmeier H, Johansson M, Kliger R, Westin M (2007b) Material prop-erties and their interrelation in chemically modified clear wood ofScots pine. Holz Roh-Werkst 61:34–42

Epmeier H, Westin M, Rapp A (2004) Differently modified wood:comparison of some selected properties. Scand J For Res19(5):31–37

Gobakken L, Westin M (2008) Surface mould growth on five modifiedwood substrates coated with three different coating systems whenexposed outdoors. Int Biodeterior Biodegrad 62:397–402

Hadi Y, Westin M, Rasyid E (2005) Resistance of furfurylated wood totermite attack. For Prod J 55(11):85–88

ISO 3350 (1975) Wood—determination of static hardnessLande S, Westin M, Schneider M (2004a) Chemistry and ecotoxicol-

ogy of furfurylated wood. Scand J For Res 19(Suppl 5):14–21Lande S, Westin M, Schneider M (2004b) Properties of furfurylated

wood. Scand J For Res 19(Suppl 5):22–30Nordstierna L, Lande S, Westin M, Karlsson O, Furó I (2008) Towards

novel wood-based materials: chemical bonds between lignin-likemodel molecules and poly(furfuryl alcohol) studied by NMR.Holzforschung 62(6):709–713

NP 619 (1973) Wood determination of static bending (In Portuguese).Instituto Português da Qualidade, Lisboa

Rapp A, Brischke C, Welzbacher C, Jazayeri L (2008) Increased resis-tance of thermally modified Norway spruce timber (TMT) againstbrown rot decay by Oligoporus placenta— study on the mode ofprotective action. Wood Res 53(2):13–25

Schneider M (1995) New cell wall and cell lumen wood polymer com-posites. Wood Sci Technol 29:121–127

Stamm A (1977) Dimensional stabilization of wood with furfuryl al-cohol. In: Goldstein I (ed) Wood technology; Chemical aspects.ACS symposium series, vol 43. American Chemical Society,Washington, pp 141–149

Temiz A, Terziev N, Eikenes M, Hafren J (2007) Effect of acceleratedweathering on surface chemistry of modified wood. Appl Surf Sci253:5355–5362

Venås T, Thygesen L, Barsberg S (2006) Chemical reactions in-volved in furfurylation of solid wood—an investigation by ATR-IR spectroscopy. Int Res Group Wood Pre. Document No.IRG/WP/40347. Stockholm

Westin M (1995) Development and evaluation of new alternativewood preservation treatments. Mid-term status report to theSwedish Council for Forestry and Agricultural Research (SJFR).(In Swedish with English summary)

Westin M, Rapp A, Nilsson T (2006) Field test of resistance of modi-fied wood to marine borers. Wood Mater Sci Eng 1:34–38