Colour in thermally modified wood of beech, Norway spruce and Scots pine. Part 1: Colour evolution and colour changes

Transcription

1 Holzforschung, Vol. 63, pp , 2009 Copyright by Walter de Gruyter Berlin New York. DOI /HF Colour in thermally modified wood of beech, Norway spruce and Scots pine. Part 1: Colour evolution and colour changes Marcos M. González-Peña* and Michael D.C. Hale School of the Environment and Natural Resources, Bangor University, Bangor Gwynedd, LL57 2UW, UK *Corresponding author. Present address: Centre for Advanced Wood Processing, The University of British Columbia, Main Mall, Vancouver, V6T 1Z4, Canada (present address); Abstract Colour evolution and colour changes were analysed from surface images of small specimens of three thermally-modified timber species using the CIEL*a*b* colour space. Upon heat exposure, the wood substance became orange and then approached grey irrespective of species; this was accompanied by a steady reduction in lightness. Colour changes were similar in the three woods at any given level of heat-induced weight loss (WL), whilst changes in the three coordinates of the CIEL*a*b* space in function of WL were different regardless of the wood species. For DL*, the profile was curvilinear and monotonous, while Da* and Db* bear a complex, non-linear profile. In turn, DE* was found to be highly influenced by the behaviour of DL*. It is proposed that DE* in thermally modified wood originates from chemical changes in the main wood polymers, more so in lignin than in polysaccharides, due to the darkening of the lignin itself. This was associated with the generation of chromophoric groups, mainly the increase in carbonyl groups appearing in the Fourier transform infrared spectra of lignin between 1710 and 1600 cm -1, particularly the emergence of quinone species. Keywords: aesthetics; beech (Fagus sylvatica L.); chemical composition; heat treatment; IR spectroscopy; Norway spruce (Picea abies L.); Scots pine (Pinus sylvestris L.). Introduction Wood colour is frequently one of the most valuable characteristics for its use as a material, as it is an important aesthetic component (Usta 2007). There is also an emotional dimension of colour which is central to species selection for wood trim, owing to the major contribution that colour makes to the perceived warmth and beauty of the material (Hon and Minemura 2001). One method of changing wood colour without adding dyes is by the thermal modification of wood, albeit this gives mainly darker tones. In early research, when Tiemann (1920) was investigating the effect of high-temperature drying of wood, he contended that the colouring due to heat treatment would be potentially valuable for wood decoration, and has ever since been considered an advantage and also as an alternative to conventional staining of light coloured species (Chang and Keith 1978; Johansson and Morén 2006; Brischke et al. 2007). Thermally modified wood (TMW) is quite homogeneous in colour throughout its thickness, which is helpful when further machining is required; the darker colour can also mask many blemishes and discolourations. Currently, the heat-induced colouring of wood is more relevant owing to the global trend for reducing the use of solvent-borne finishes mainly in furniture processing (Fonseca 2004). The study of colour conversion in TMW has aroused interest lately, because a relationship between these changes and weight loss (WL), equilibrium moisture content (EMC) and mechanical strength of TMW has been determined (Bekhta and Niemz 2003; Patzelt et al. 2003; Brischke et al. 2007). However, information on the colour development and colour changes in TMW is often limited to few colour variables (DE*, DL* and/or L*), and comparisons amongst the industrially relevant wood species in relation to the modification and resultant colour characters are even more scarce. In this study, we describe the effect of 20 schedules of thermal modification on 11 colour characters in three commercially important woods, determined by image analysis. This information is then used to predict the changes in several physical properties of TMW; this is presented in a second paper of this series (González-Peña and Hale 2009a). Methods Thermal modification Matched, small specimens (150 l mm=20 r mm=10 t mm) of beech (Fagus sylvatica L.), Norway spruce (Picea abies L.) and Scots pine (Pinus sylvestris L.) woods were modified in N 2 gas atmosphere at 1908C, 2108C, 2308C and 2458C, for 0.33, 1, 4, 8 and 16 h (treatment at 2458C for 8 h for beech wood was abandoned). Oven-dried specimens were loaded into the treating chamber already heated to the treatment temperature. A standard specimen heating up time of 50 min was used for all treatments; the reported treatment time was measured from this point on. Following treatment, all sets were conditioned at 65% relative humidity at 208C for at least 6 months in the absence of light. Image acquisition Digital images (3550=450 pixels) of TMW surfaces (radial plane) were recorded on an Epson Perfection 4990 flatbed scanner on conditioned specimens without further preparation. Ten specimens per treatment (a combination time-temperature) were scanned for each species; only the after-treatment colour was measured. The before-treatment colour used for each modified

2 386 M.M. González-Peña and M.D.C. Hale specimen was the average of those specimens of the untreated control sets (ns10). Images were transferred to Photoshop 7.0 software (Adobe Systems Inc., USA), and analysed in the CIEL*a*b* system. A rectangle of 3500=380 pixels was sampled from each wood specimen. Each pixel was assigned a value in the internal scale of the software for each coordinate (between 0 and 255), and the software gave the histogram for each of these; only the mean value for each coordinate (L*, a* and b*) was recorded. A linear transformation was used to convert these values to the data range of the CIEL*a*b* system (100 for L*, and 120 for each of the a* and b* coordinates, see below). Conditioned powdered Klason lignin specimens from TMW and unmodified wood controls for each species were also scanned and the images stored and analysed similarly as for the solid wood specimens. Only one lignin specimen per treatment was imaged for each species. The CIEL*a*b* colour space In the three dimensional CIEL*a*b* colour space, each colour can be expressed as a point in the Euclidean space defined by three coordinates correlating with the subjective colour perception (Mononen et al. 2002) (Figure 1). The vertical coordinate for lightness L* represents the position on the black-white axis (L*s0 for total blackness, L*s100 for pure white), whilst the chromatic coordinates a* and b* describe the position on the horizontal plane. The chrome value a* defines the position on the green-red axis (-60 for green, q60 for red), while b* defines the position on the blue-yellow axis (-60 for blue, q60 for yellow) (Thompson et al. 2005). A difference in colour called DE* is expressed as a distance between two points in the colour coordinate system (Figure 1). The equation for the total colour difference DE* is given according to Tappi (1994): DE sž DL qda qdb. U U2 U2 U 2 $ where: DL* is the difference in lightnesssl*-l* control Da* is the difference in a* coordinatesa*-a* control Db* is the difference in b* coordinatesb*-b* control Figure 1 The CIEL*a*b* colour space. The deduced colour components C* and h* are also included. Insert: the physical interpretation of DE*. where L* control, a* control and b* control are the mean values of these colour components for untreated control sets (ns10), and L*, a* and b* are the colour variables of each treated specimen (ns10 per treatment). Alternatively, a cylindrical coordinate system may be used to describe colour, with the a* and b* values being substituted by the saturation C* and the hue angle h* on the colour circle around the lightness axis L* (Figure 1). On the hue circle, hs08 denotes redness and hs908 denotes yellowness. C* is the distance between the colour and the centre of the chromaticity plane; it is a measure of colour intensity. From the a* and b* values, the following variables are calculated (Bourgois et al. 1991, Tappi 1994): h*shue anglesarctang (b*/a*) C*ssaturations(a* 2 qb* 2 ) 1/2 DC* ab ssaturation differencesc*-c* control DC*schromaticity changes(da* 2 qdb* 2 ) 1/2. Total hydrolysis, analysis of the hydrolysate and the unhydrolysables Oven-dry WL was measured relative to the initial oven-dry wood weight. Wood was submitted to total hydrolysis in sulphuric acid (Klason procedure according to Effland 1977). Total lignin content is reported here as the sum of the unhydrolysable residue (Klason lignin) and acid soluble lignin (Tappi 1991). Quantification of monosaccharides in the hydrolysate was performed by high performance anion exchange liquid chromatography (Worrall and Anderson 1993). From monosaccharide composition, the two main hemicellulose types (glucomannan, GluMan, and glucuronoxylan, GluXylan), and cellulose content were calculated based on several assumptions on the structure of the major wood hemicelluloses, with the algorithm of Janson (1970). Fourier transform infrared (FTIR) spectra of Klason lignin specimens of spruce wood were obtained in the range cm -1,at a resolution of 4 cm -1, on a Bruker spectrophotometer (Tensor 27, Bruker Optik GmbH, Germany) using KBr pellets. The spectra were base-line corrected at 3702, 1851, 709, and 410 cm -1, and then normalised at 858 cm -1 as a maximum, and at 906 cm -1 as a minimum. Results and discussion Effect of time and temperature in the evolution of colour components Heating of beech, pine and spruce woods induced noticeable modifications in all colour components, mainly a decrease in L* (darkening), irrespective of the temperature of treatment (Figure 2). Whatever the species, wood darkening deepened as the time of exposure increased, and evolved more rapidly with increasing temperatures as reported by Unsal et al. (2003). The rate of reduction of L* declined over time of exposure at any given temperature. This profile is similar to that of WL in TMW, and presumably manifests the chemical stabilisation of TMW upon increased lengths of heat treatment (Bourgois and Guyonnet 1988). In beech, L* decreased slowly only for the treatment at 1908C; colour conversion was fast for the treatment at 2108C, and almost instantaneous at 2308C and at 2458C, to reach a minimum of approximately L*s35. Evolution of L* in softwoods was gradual for treatments up to 2108C, and fast for treatments at 2308C and 2458C. The minimum at 16 h in softwoods varied with the temperature of exposure, from L*s70 and 75 at 1908C, to L*s28 and 34 at 2458C, for pine and spruce, respectively (Figure 2). In all the three species, a* and b* increased at the beginning of the treatment at any processing tempera-

3 Colour changes in TMW 387 Figure 3 Evolution of a* and b* for beech, Scots pine and Norway spruce woods in function of time and temperature of treatment. Each symbol is the mean of 10 replicates. Figure 2 Evolution of L* for beech, Scots pine and Norway spruce woods as a function of time and temperature of treatment. Each symbol is the mean of 10 replicates. ture, except for beech treated at 2458C (Figure 3). After reaching a maximum, both a* and b* declined with longer exposure; this also agrees with previous studies (Bekhta and Niemz 2003; Brischke et al. 2007). An increase in the temperature resulted in an acceleration of this increasedecrease cycle. Maxima in a* and b* varied between species. Maximum in a* was approximately 5 to 6 for beech, from 8 to 9.5 in pine and between 6 and 7 in spruce. In b*, maxima reached 18 22, and for beech, pine and spruce, respectively. The L*a*, L*b* plot shows that both b* and a* varied in a similar parabolic pattern with the decrease in lightness (Figure 4a). Regardless of the species, the coordinates reach a maxima at L*s60 for a* and at L*s72 for b*. The evolution of the a* and b* coordinates in beech is less conspicuous than in the softwoods at equivalent values of L*. At equivalent values of L*, the modification in b* is larger than in a* in all woods; b* increases more as L* decreases, but the reduction in b* is also larger after reaching its maxima, attaining values of nearly zero at L* values -30. Both a* and b* coordinates remain positive for most of the treatments, although a* becomes negative for L* values The direction of colour modification in the a*, b* quadrant was not linear (Figure 4b). Wood becomes redder and more yellow (i.e., more orange) as the treatment increases, and then the chromatic values decrease to reach almost zero and the wood specimens become essentially hueless. The description of the colour evolution is therefore better explained as an orange darkening towards dark grey in the three wood species. Colour saturation, C*, evolved differently according to the treatment temperature (Figure 5), in a similar way as described for the b* coordinate above, i.e., C* values increased at the beginning of the treatment, and then decreased after longer exposure. This was probably due to the large contribution of b* towards C* and was irrespective of species or temperature of treatment. The effect of increasing the temperature was the acceleration of this increasing-decreasing process. C* changes were smaller in beech than in softwoods. The behaviour of hue angle, h*, was somewhat similar but opposite to the one of C*. Angles decreased fairly abruptly at the beginning of the treatment in all three species, and then increased for longer treatments; the increase in the temperature of exposure led to speeding up these modifications. This profile and the fact that h* was always greater than 658 indicate a larger contribution of b* than a* towards colour of TMW.

4 388 M.M. González-Peña and M.D.C. Hale Figure 4 (a) Relationship between the a* and b* coordinates and L* in thermally modified beech, Scots pine and Norway spruce woods, plotted irrespective of exposure conditions; (b) The a*, b* quadrant in the CIEL*a*b* colour space for thermally modified beech, Scots pine and Norway spruce woods. ns200 beech, ns220 pine and spruce. Figure 5 Evolution of C* and h* (in8) in the L*C*h* colour space for beech, Scots pine and Norway spruce woods as a function of time and temperature of treatment. Each symbol is the mean of 10 replicates. Changes in colour variables as a function of WL WL is an ideal surrogate response of the severity of the treatment, resulting from the combination of the processing time and temperature (González-Peña and Hale 2009b). Thus, changes in colour variables are presented below in function of WL for inter-species comparisons. A negative curvilinear relationship exists between DL* and WL in all the three woods (Figure 6). Moreover, at equivalent levels of WL, DL* is very similar in the three species, despite the initial difference in colour and chemical make-up between them. With regard to rate, however, DL* in pine wood increased at a slightly higher rate than in the other two species. On the other hand, a strong, curvilinear positive relationship exists between the WL and DE* in the three species (Figure 6). DE* is again very similar in the three woods at equivalent levels of treatment, although pine shows a slightly larger changing rate than the other two woods. Beech specimens also had somewhat smaller DE* than softwoods at WL )20%. DE* was highly influenced by DL, with Db* having more influence on DE* than Da* did; this is associated to the small initial contribution of both a* and b* coordinates towards wood colour. Figure 6 DE* and DL* in the CIEL*a*b* colour space for heattreated beech, Scots pine and Norway spruce woods as a function of the WL. ns200 beech, ns220 pine and spruce. The smallest colour difference in treated wood was visually rated at DE*s2.5. As the DE* value increased, colour differences were more conspicuous, being moderate at DE*s2.5 17, large at and very large at DE*)40. Interestingly, DE* appeared not to be linked to the schedules leading to comparable degree of WL. As noted earlier (Patzelt et al. 2003), higher treatment temperature and shorter reaction times lead to similar DE*as

5 Colour changes in TMW 389 lower temperatures and longer processing times at equivalent levels of WL; this may be advantageous for monitoring the reaction within the range of processing temperatures used here ( C). Differences in the chromic characters, Da* and Db*, were smaller at equivalent levels of treatment than DL*, and displayed a more complex behaviour (Figure 7) but the profile was similar in the three species. Both Da* and Db* follow a quasi-sigmoid, non-linear trend as the WL increases; Db* being greater than Da*. At low levels of WL, the differences are positive and increase sharply. After reaching their maxima, each coordinate declines as the treatment proceeds, reaching control values at WL)10% for Da* in the three species, and around 5% WL for beech and pine, and at 10% WL for spruce in Db*. Da* and Db* subsequently became negative (specimens are less red and less yellow than control specimens), but as it is shown in Figure 4b both coordinates remain largely in the positive quadrant. Saturation differences, DC* ab, follows a similar pattern to Db*, increasing sharply at the beginning of the treatment, and then decreasing before becoming negative at around 5% WL in beech and pine, and at 10% WL in spruce: specimens are more coloured than untreated controls at early levels of treatment, and then tend to become hueless (plots not shown). Chromaticity changes, DC*, were composite: in beech and pine these are almost linear at WL)5% and in spruce these follow a more complex, sigmoid behaviour, becoming linear at WL)10% (plots not shown). In any species, DC* is increasingly positive in line with the WL. Relationship between chemical changes and colour modifications Figure 7 Da* (above) and Db* (below) in the CIEL*a*b* colour space for heat-treated beech, Scots pine and Norway spruce woods as a function of the WL. ns200 beech, ns220 pine and spruce. A strong correlation (Pearson) was generally found between all chemical components and DL*, Db* and DE* (Table 1). Da* usually had the lowest correlation with any chemical constituent (Table 1). In pine and beech, the strongest correlation for Da* was also with the lignin content. DE* was positively related to lignin, and negatively related to the other chemical components, whilst the opposite was found for DL*, Da* and Db*. Whatever the species, the smallest correlation between colour differences and chemical changes was with cellulose. The relationship between DL*, Db* and DE* and chemical content was mixed. DL* has the strongest correlation with hemicelluloses in beech, with lignin in pine and with GluXylan in spruce. On the other hand, Db* was more strongly correlated with GluXylan, lignin and GluMan in Table 1 Pearson correlation between chemical constituents and DL*, Da*, Db* and DE* in termally modified beech, Scots pine and Norway spruce woods. Constituent a Spp. Lignin Glucuronoxylan Glucomannan Hemicelluloses Cellulose Beech DL* *** *** *** *** ** Da* *** *** ** *** *** Db* *** *** *** *** *** DE* *** *** *** *** ** Scots DL* *** *** *** *** *** pine Da* ** * ** ** ** Db* *** *** *** *** *** DE* *** *** *** *** *** Norway DL* *** *** *** *** *** spruce Da* * NS NS NS NS Db* *** *** *** *** * DE* *** *** *** *** *** Determination of the constituents: ligninsunhydrolysables by sulphuric acid (Klason residue by weight) plus acid soluble lignin by UV. Polysaccharidessby sugar analysis in the hydrolysate and recalculation of the data according to Janson (1970). a Content of the before treatment oven-dry weight, in %; ***Significant at p level; **Significant at p-0.01 level; *Significant at p-0.05 level; NS, not significant.

6 390 M.M. González-Peña and M.D.C. Hale beech, pine and spruce, respectively. In beech, total colour difference had the largest correlation with hemicelluloses, followed by GluXylan; in pine the largest correlation was with lignin followed by GluMan, and in spruce with GluXylan followed by hemicelluloses. In previous studies on colour change of TMW, the decrease in L* and the increase in DE* resulting from modification of Scots pine wood at 1008C and at C were found to result from the degradation of pentosan and relative increase of lignin (Bourgois et al. 1991). A strong relationship between lightness reduction and glucose (R 2 s0.96), hemicelluloses (R 2 s0.92), lignin (R 2 s0.86) and extractives (R 2 s0.62) was later determined for Pinus pinaster Aiton (Esteves et al. 2007). Results from previous investigations as well as the correlation analysis reported here probably reveal the contribution that all components make to colour change in TMW. However, gravimetric determinations do not completely reflect the qualitative modifications undergone in each polymer of the wood substance upon heating (e.g., colour change) and therefore analyses of qualitative chemical changes are also needed. Colour changes in individual wood constituents and its relationship with colour changes in TMW Wood discolours when it is subject to high temperatures; if left for long periods, it usually becomes brown. This has been ascribed to the formation of coloured substances from a phenolic compound oxidized with air and the formation of dark materials from hydrolysis of hemicelluloses (Hon and Minemura 2001). If wood is heated for short periods, wood colour changes to yellow, brown, red, grey, etc. In other practices in the woodworking industry involving heating (e.g., wood drying), these colour changes are associated with superficial staining of the material, as a result of enzyme-mediated (Maillard) reactions between sugars, phenolic compounds and amino acids, or to the oxidative reactions between extractives and the atmosphere (Keey 2005; Thompson et al. 2005). Kiln stains are also thought to develop due to the migration of darkened coloured compounds to the surface of boards as free water is removed upon drying. However, extractives rapidly disappear from heat-treated Scots pine wood at early stages of thermal modification (Nuopponen et al. 2003). In the present study, only ovendry sapwood was used and colour changes were not only large (DE*)17) but also fairly even and throughout the thickness. Therefore, extractive-induced or enzymemediated discolorations are unlikely to be major factors in the colour changes noted here. An investigation of the relationship between DE* in Klason lignin and in the wood Figure 8 Relationship between DE* in heated softwoods (above) and beech wood (below) and DE* in Klason lignin. substance showed that DE* in wood is closely though non-linearly associated to DE* in lignin in treated softwood (Figure 8, Table 2). In beech this relationship was curvilinear, but the data show a larger scatter than in softwoods, probably indicative of a less strong relationship between both events. Regardless of the species, colour differences in lignin accounted for 40% or more of DE* in wood at WL)7 8%. At lower WL, polysaccharides (and extractives) possibly have a larger involvement in the colour changes of TMW. Colour changes in isolated cellulose and hemicelluloses were only quantified for a few treatments in spruce wood. These polymers also changed in colour, but to a much smaller degree than lignin. For instance, cellulose from spruce wood heated at 2108C for 16 h (to WLs8.7%) has a DE* values10.2 units, while in hemicelluloses this was 3.2 units. For comparison, DE* in Klason lignin was 22.8 units. Apparently the rule of mixtures does not apply; the difference in wood colour does not reflect the simple accumulative difference in colour of individual components (36.2 units), because for this treatment DE* in wood was greater (39.4 units). Probably the absorption in the visible range of chromophoric groups in lignin is amplified by

7 Colour changes in TMW 391 integrating these into a polysaccharide matrix (Pew and Connors 1971). The connection between the profile found for DL* and changes in wood polymers is probably analogous to the description as for DE*, i.e., due to a proportionally higher acid-insoluble lignin content and development of colouring compounds in all the three main wood components. In turn, the profile of Da* ordb* is more complex, and hence it is more difficult to assign to chemical changes. Positive, rapid changes in these coordinates were found at low WL, whilst further modification lead to negative changes. This may be linked to the production and subsequent destruction of chromophoric species, probably originated from extractives, and/or due to condensation and re-polymerisation of the lignin substance (Wikberg and Maunu 2004). FTIR spectroscopy of lignin and its relationship to colour changes Although development of coloured compounds is possible in all the three main wood polymers, lignin is richer than carbohydrates in latent chromophoric groups. Thus, a closer examination of chemical changes by the analysis of the FTIR spectra was restricted herein to the Klason lignin, in the belief that it is still possible to detect the actual changes due to the thermal modification even against the background of the modifications impinged on lignin by the harsh isolation method. The difference spectra of Klason lignin for Norway spruce, for treatments resulting in WL between 1 and 15%, reveal an absorbance increment at several wavenumbers in the carbonyl region (Figure 9). Several positive peaks appear in this region, markedly at 1708, 1637, 1630 and 1600 cm -1. The band at 1708 cm -1 is for the CO stretching of various functional groups (unconjugated ketones, carbonyl and ester groups; Faix 1992). Although the spectra are for Klason lignin, the large peak in this region may also contain the vibration from CO groups derived from the carboxylation of polysaccharides (Chow 1971). These species possibly percolated and remain trapped within the lignin network even after the acidic removal of carbohydrates (Nikitin 1966; Garrote et al. 1999; Inari et al. 2007a,b). The sharp decreasing band for pure aromatic skeletal vibrations seen at 1506 cm -1 has previously been reported in heated wood (Grandmaison et al. 1987; González-Peña and Hale 2009c). However, the reason for this decline in aromaticity in materials with increased lignin content is not completely understood. Degradation of aromatic structures involving the opening of the aromatic ring leading to the formation of muconic acids rich in carbonyl structures, carboxylic acids and/or their lactones are common results of wood bleaching proceedings and in light-irradiated wood; these give rise to the CO vibration above 1700 cm -1 (Tolvaj and Faix 1995; Mononen et al. 2005). Moreover, the intensities of carbonyl bands at cm -1 and the aromatic skeletal vibrations at cm -1 show characteristic patterns when plotted as a function of the time in irradiated wood. The two curves act as mirror images: the CO band increases to a comparable extent as the aromatic band decreases. In lignin from heat-treated wood, this is also noted if the result for the difference spectra at 1708 and 1506 cm -1 is plotted against the WL (Figure 10), but the reduction of the latter is around 30 50% smaller than the increase at 1708 cm -1. However, other vibrations arising from the aromatic skeletal in lignin (e.g., at 1140 cm -1 ) also decline as the treatment proceeds, and probably contribute towards the increase of the CO band at 1708 cm -1. Based on these results, we suggest that some of the main reaction products formed during heating of wood are likely to be of the muconic acid type, probably with the concourse of trace oxygen in the heating chamber, resulting in the generation of chromophoric carbonyl groups. Evidence of the opening of the aromatic ring in lignin of hydrothermally treated wood has also been detected by UV spectroscopy (Boonstra and Tjeerdsma 2006). On the other hand, the absorbance increase at 1600 cm -1 is associated with quinone formation, since these have a peak assignable at this frequency due to ring stretching vibration (Kimura et al. 1994). Likewise, the peak at 1637 cm -1 most likely arises from quinone carbonyl groups, since these are expected at cm -1 (Kemp 1991). Quinone structures absorb in the visible region and are probably the main reason for colour changes in heated wood. Although bands indicating its presence were not very pronounced, quinones can cause an intense colour at very low concentrations (Tolvaj and Faix 1995). Condensation reac- Figure 9 Difference spectra of Klason lignin for seven levels of thermal modification in Norway spruce wood. Treated specimen minus control spectrum shown. Figure 10 Absorbance vs. WL for the bands at 1708 cm -1 and 1506 cm -1 in Klason lignin of Norway spruce wood, from the difference spectra of treated specimen minus control spectrum.

8 392 M.M. González-Peña and M.D.C. Hale tions of lignin lead to the production of these coloured compounds: under heating the most reactive sites of lignin side chains (a positions) are attacked rapidly by adjacent phenyl nuclei regardless of the ph conditions to form C a -aryl linkages (diphenylmethane type structures). These units greatly influence the colour in the lignin substance, and are readily oxidised to yield coloured mesomeric quinone-methides, which form strong hydrogen bonds with hydroquinones or phenols (Sudo et al. 1985; Funaoka et al. 1990; Nakamura et al. 2007). Conclusion Results from the present study confirm earlier reports on the trend of colour coordinates in the CIEL*a*b* system, whereby it is recognised that L* decreases at any given treating schedule, whilst the a* and b* coordinates have a paraboloid profile in line with the severity of the treatment. We have extended previous observations on the relationship between colour changes and chemical changes in TMW, and have determined that colour changes are presumably more linked to changes in the acid-insoluble lignin substance than in the carbohydrate fraction of TMW. Acknowledgements M.M.G.P. would like to thank the National Council for Science and Technology (CONACYT), Mexico (grant no ) for funding to carry out this research. References Bekhta, P., Niemz, P. (2003) Effect of high temperature on the change in color, dimensional stability and mechanical properties of spruce wood. Holzforschung 57: Boonstra, M., Tjeerdsma, B. (2006) Chemical analysis of heat treated softwoods. Holz Roh Werkst. 64: Bourgois, J., Guyonnet, R (1988) Characterization and analysis of torrefied wood. Wood Sci. Technol. 22: Bourgois, J., Janin, G., Guyonnet, R. (1991) La mesure de couleur: une méthode d étude et d optimisation des transformations chimiques du bois thermolysé. Holzforschung 45: Brischke, C., Welzbacher, C.R., Brandt, K., Rapp, A.O. (2007) Quality control of thermally modified timber: interrelationship between heat treatment intensities and CIE L*a*b* color data on homogenized wood samples. Holzforschung 61: Chang, C.I., Keith, C.T. (1978) Properties of heat-darkened Wood. II. Mechanical properties and gluability. Eastern Forest Products Laboratory, Ottawa. Report OPX214E. 19 pp. Chow, S.Z. (1971) Infrared spectral characteristics and surface inactivation of wood at high temperatures. Wood Sci. Technol. 5: Effland, M.J. (1977) Modified procedure to determine acid-insoluble lignin in wood and pulp. Tappi 60: Esteves, B., Velez-Marques, A., Domingos, I., Pereira, H. (2007) Influence of steam heating on the properties of pine (Pinus pinaster) and eucalypt (Eucalyptus globulus) wood. Wood Sci. Technol. 41: Faix, O. (1992) Characterisation in solid state: Fourier transform infrared spectroscopy. In: Methods in Lignin Chemistry. Eds. Lin, S.Y., Dence, C.W. Springer-Verlag, Berlin. pp Fonseca, A. (2004) Environmental management in wood processing industries and the European legislation on VOC emission control. In: 1st International Conference on Environmentally-compatible Forest Products, September Fernando Pasoa University, Porto. pp Funaoka, M., Kako, T., Abe, I. (1990) Condensation of lignin during heating of wood. Wood Sci. Technol. 24: Garrote, G., Domínguez, H., Parajó, J.C. (1999) Hydrothermal processing of lignocellulosic materials. Holz Roh Werkst. 57: González-Peña, M.M., Hale, M.D.C. (2009a) Colour in thermallymodified wood of beech, Norway spruce and Scots pine. Part 2: Property prediction from colour changes. Holzforschung 63: González-Peña, M.M., Hale, M.D.C. (2009b). On the prediction of the mechanical performance of thermally modified wood. II. From treatment parameters and from other mechanical properties. Holz Roh Werkst. (submitted). González-Peña, M.M., Hale, M.D.C. (2009c). Rapid assessment of physical properties and chemical composition of thermally-modified wood by mid-infrared spectroscopy. Wood Sci. Technol. (in press). Grandmaison, J.L., Thibault, J., Kaliaguine, S. (1987) Fourier transform infrared spectrometry and thermogravimetry of partially converted lignocellulosic materials. Anal. Chem. 59: Hon, D.N.-S., Minemura, N. (2001) Color and discoloration. In: Wood and Cellulose Chemistry, 2nd edn. Eds. Hon, D.N.-S., Shiraishi, N. Marcel Dekker Inc., New York. pp Inari, G.N., Mounguengui, S., Dumarcay, S., Pétrissans, M., Gérardin, P. (2007a) Evidence of char formation during wood heat treatment by mild pyrolysis. Polym. Degrad. Stabil. 92: Inari, G.N., Pétrissans, M., Gérardin, P. (2007b) Chemical reactivity of heat-treated wood. Wood Sci. Technol. 41: Janson, J. (1970) Calculations of the polysaccharide composition of wood and pulp. Pap. Puu 52: Johansson, D., Morén, T. (2006) The potential of colour measurement for strength prediction of thermally treated wood. Holz Roh Werkst. 64: Keey, R.B. (2005) Colour development on drying. Maderas: Cienc. Tecnol. 7:3 16. Kemp, W. Organic Spectroscopy. Macmillan, London, pp Kimura, F., Kimura, T., Gray, D.G. (1994) FT-IR study of the effect of irradiation wavelength on the color reversion of thermomechanical pulps. Holzforschung 48: Mononen, K., Alvila, L., Pakkanen, T.T. (2002) CIEL*a*b* measurements to determine the role of felling season, log storage and kiln drying on coloration of Silver birch wood. Scand. J. Forest Res. 17: Mononen, K., Jääskeläinen, A.-S., Alvila, L., Pakkanen, T.T., Vuorinen, T. (2005) Chemical changes in silver birch (Betula pendula Roth) wood caused by hydrogen peroxide bleaching and monitored by color measurement (CIELab) and UV-Vis, FTIR and UVRR spectroscopy. Holzforschung 59: Nakamura, T., Kawamoto, H., Saka, S. (2007) Condensation reactions of some lignin-related compounds at relatively low pyrolysis temperature. J. Wood Chem. Technol. 27: Nikitin, N.I. The chemistry of cellulose and wood. Academy of Sciences of the USSR, Institute of High Molecular Compounds, Moscow, pp. Nuopponen, M., Vuorinen, T., Jämsä, S., Viitaniemi, P. (2003) The effects of a heat treatment on the behaviour of extractives in softwood studied by FTIR spectroscopic methods. Wood Sci. Technol. 37: Patzelt, M., Emsenhuber, G., Stingl, R. (2003) Colour measurement as means of quality control of thermally treated wood. In: First European Conference on Wood Modification. Ghent University Press, Ghent, Belgium. April 3 4, pp Pew, J.C., Connors, W.J. (1971) Colour of coniferous lignin. Tappi 54:

9 Colour changes in TMW 393 Sudo, K., Shimizu, K., Sakurai, K. (1985) Characterization of steamed wood lignin from Beech wood. Holzforschung 39: Tappi standard (1991) UM 250. Acid-soluble lignin in wood and pulp. In: Tappi Useful Methods. Tappi Press, Atlanta, GA. pp Tappi standard (1994) T 527. Om-94. Color of paper and paper board (d/08 geometry). In: Tappi Test Methods. Tappi Press, Atlanta, GA. 5 pp. Tiemann, H.D. (1920) In: The Kiln Drying of Lumber. J.B. Lippincott Company, Philadelphia, PA. pp Thompson, D.W., Kozak, R.A., Evans, P. (2005) Thermal modification of color in red alder veneer. 1. Effects of temperature, heating time, and wood type. Wood Fiber Sci. 37: Tolvaj, L., Faix, O. (1995) Artificial ageing of wood monitored by DRIFT spectroscopy and CIE L*a*b* color measurements. 1. Effect of UV light. Holzforschung 49: Unsal, O., Korkut, S., Atik, C. (2003). The effect of heat treatment on some properties and colour in Eucalyptus camadulensis Dehn. wood. Maderas: Cienc. Tecnol. 5: Usta, I. (2007) Theory of aesthetics charm in furniture and an aesthetic evaluation of bleaching of the natural color of wood through the application of experimental bleaching process. In: 38th Annual Meeting of the International Group on Wood Protection, Wyoming, USA. Document IRG/WP pp. Wikberg, H., Maunu, S.L. (2004) Characterisation of thermally modified hard- and softwoods by 13 C CPMAS NMR. Carbohyd. Polym. 58: Worrall, J.J., Anderson, K.M. (1993) Sample preparation for analysis of wood sugars by anion chromatography. J. Wood Chem. Technol. 13: Received October 10, Accepted February 16, Previously published online May 7, 2009.