286
Biomacromolecules 2008, 9, 286–288
Monitoring of Hydroxyl Groups in Wood during Heat Treatment Using NIR Spectroscopy Katsuya Mitsui,* Tetsuya Inagaki, and Satoru Tsuchikawa Gifu Prefectural Human Life Technology Research Institute, Yamada, Takayama 506-0058, Japan, and Graduate School of Bioagricultural Sciences, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8601, Japan Received July 23, 2007; Revised Manuscript Received October 23, 2007
This paper deals with the evaluation of thermally treated wood by near-infrared (NIR) spectroscopy. In the NIR second derivative spectrum, the absorption band at 6913 cm-1 appeared with the procession of heat treatment, which conclusively assigned to the phenolic hydroxyl groups due to the lignin in comparison with the spectrum of acetylated spruce wood. As a result of the changes in the ratio of the areal integral calculated from spectral separation in the region of hydroxyl groups (7200-6100 cm-1) by the Gauss-Newton method, it was clear that the degradation of hydroxyl group in the cellulose started predominantly from the amorphous region and followed to semicrystalline and crystalline region. There was an obvious correlation between the weight decrement of wood and the decrement of hydroxyl groups in the cellulose by heat treatment.
Introduction The thermal modification of wood has long been carried out to change its properties. Although there is a disadvantage that the thermal treatment causes the decrement of strength,1–4 there are some advantages such as an increment of dimensional stabilization of wood4–7 and coloration without paint including organic solvents such as toluene and xylene.8–11 The dimensional stabilization is caused by the control of volume swelling by water because the hydroxyl groups of cellulose and hemicellulose decrease by thermal treatment and the adsorption of water to wood decreases. Fengel and Wegener12 have shown the probable thermal degradation pathway for cellulose and hemicelluloses of wood. Near-infrared (NIR) spectroscopy is one of the useful methods for nondestructive measurement. Recently, there have been various reports on NIR spectroscopy assay for wood.13–23 In comparison with infrared (IR) spectroscopy, it is difficult to assign the functional groups because of obscureness or overlapping of peaks in the NIR spectrum; however, in the region of 7200-6100 cm-1, clear absorption peaks in terms of the hydroxyl group are observed by second derivatives of the spectrum. On the experiment of wood, Tsuchikawa and Siesler16 reported that the absorption peaks at 7003, 6722, 6460, and 6281 cm-1 were assigned to the amorphous region, the semicrystalline region, and two kinds of crystalline regions, respectively, by means of a deuterium exchange method and FT-NIR polarization spectroscopy. However, there are no reports on detailed discussion on hydroxyl groups of thermally treated wood using NIR measurement. In this study, the behavior of hydroxyl groups in wood by thermal treatment was monitored using NIR spectroscopy. In addition, the acetylated wood, which is a typical chemically modified wood for hydrophobicity, was also employed to clarify the spectroscopic characteristics of hydroxyl groups. * Corresponding Author. E-mail:
[email protected]. Telephone: +81-577-33-5252. Fax: +81-577-33-0747.
Figure 1. NIR spectra of untreated, heat-treated for 100 h, and acetylated wood.
Experimental Section Materials. The species used in this study was Sitka spruce (Picea sitchensis) having 50 mm × 10 mm × 1 mm in the longitudinal, tangential, and radial directions, respectively. The specimens were collected from sliced veneer, and the early wood was used for measurement. Thermal Treatment. The five unacetylated specimens were treated by steaming at 140 °C for 5, 10, 20, 50, and 100 h, respectively. Acetylation. The oven-dried specimens were treated with acetic anhydride in the liquid phase without catalyst for 6 h at 120 °C. At the end of the reaction, the specimens were washed with water until the pH of water used for washing remained neutral and then dried over P2O5 at room temperature. The weight percent gain was 20.8 ( 0.9%. NIR Spectroscopic Measurement. Before NIR measurement, the specimens were dried under reduced pressure at 60 °C. The NIR spectra were measured on a FT-NIR spectrometer by using optical fiber cable that had 1 m of length and 7.2 mm of diameter, of interactance mode (MATRIX-F, Bruker Optics Inc., Ettlingen, Germany). To improve the signal-to-noise ratio, 128 scans were co-added at a spectral resolution of 8 cm-1. Two spectra were collected and averaged from each sample.
Results and Discussion Changes in NIR Spectra by Thermal Treatment. Figure 1 and Table 1 show the NIR spectra of untreated, heat-treated
10.1021/bm7008069 CCC: $40.75 2008 American Chemical Society Published on Web 12/08/2007
Monitoring of Hydroxyl Groups in Wood by NIRs
Biomacromolecules, Vol. 9, No. 1, 2008 287
Table 1. Assignment of absorption bands of NIR spectrum wavenumber [cm-1]
assignment
a
7003
b
6722
c
6460
d
6281
e
5981
f
5800
g h i j k
5219 4808 4401 4281 4202
OH stretching first overtone, amorphous regions16 OH stretching first overtone, semicrystalline regions16 OH stretching first overtone, crystalline regions in cellulose16 OH stretching first overtone, crystalline regions in cellulose16 CH stretching first overtone, aromatic skeletal due to lignin25,26 CH stretching first overtone, furanose, or pyranose due to hemicellulose27 OH stretching + OH def., H2O25,26 OH stretching + OH def., OH25,26 CH stretching + CH def., CH325,26 CH stretching + CH def.27 OH def. second overtone, OH25,26
for 100 h, and acetylated wood and the assignment of absorption bands, respectively. The spectra have been shifted along the absorbance axis due to a standard white plate as reference so that the spectra of the control sample and the acetylated sample showed negative absorbance values under 8000 cm-1. The absorbance from 10000 cm-1 to 7200 cm-1 in the spectrum of the heat-treated sample was higher than that of acetylated or control one. It was caused by the remarkable darkening of wood by thermal treatment. Schwanninger et al.24 showed that the absorbance in this region increased with treatment temperature on the experimental of thermal treatment of beech wood. In the case of wood, the absorbance at wavenumbers 7003, 6722, 6460 and 6281 cm-1 is assigned to hydroxyl groups in the cellulose,16 and that at 5981 and 5800 cm-1 is assigned to the CH stretching first overtone of aromatic skeletal due to lignin25,26 and the CH stretching first overtone of furanose or pyranose due to hemicellulose,27 respectively. After acetylation, the absorption of hydroxyl groups ranging from 7200 to 6100 cm-1 decreased, and the absorption of CH stretching at 5981 and 5800 cm-1 increased. These results indicate that the acetyl groups reacted to the cellulose, hemicellulose, and lignin. It was suggested that the hydroxyl group at 4808 cm-1 was assigned to that of the cellulose or hemicellulose because it decreased by thermal treatment or acetylation. Figure 2 shows the expanded second derivative spectra of spruce wood. The second derivative spectra were calculated by moving average of 25 points based on Savitzky-Golay method. We were able to observe two weak absorption peaks at 6913 cm-1 and 6598 cm-1. The absorption peak at 6913 cm-1 was clearly observed after thermal treatment. Comparison with the spectrum of control specimen explains that the absorption peak overlapped to that of the amorphous region in the cellulose might be clearly separated by the thermal treatment. Furthermore, it is suggested that the absorption peak at 6913 cm-1 is assigned to the stable substances against thermal treatment because the second derivative values (d2a/dV˜ 2) around 6913 cm-1 before and after treatment are almost the same. On the other hand, there was no absorption peak around 6913 cm-1 in the spectrum of acetylated wood. The functional groups that have high reactivity to acetylation should directly relate to this band. Therefore, it was concluded that the absorption peak of 6913 cm-1 was assigned to the phenolic hydroxyl groups originated from lignin. The absorption peak at 6598 cm-1 was tentatively assigned to the hydroxyl groups in either the semicrystalline region or the crystalline region because the absorbance peak appeared between those of the semicrystalline region at 6722 cm-1 and the crystalline region at 6460 cm-1. Their behavior
Figure 2. Expanded second derivative spectra of spruce wood.
Figure 3. Spectral separation in the region of hydroxyl groups by Gauss-Newton method.
Figure 4. Changes in the ration of areal integral after thermal treatment to that before treatment. b: amorphous region (7003 cm-1), O: semicrystalline region (6722 cm-1), 9: crystalline region (6460 cm-1), 0: crystalline region (6281 cm-1), 2: phenolic hydroxyl group (6913 cm-1).
in the case of thermally treated wood and acetylated wood was almost the same as that of the control one. Next, the spectrum was separated to six peaks by the Gauss-Newton method, and curve fitting was applied to separate the NIR difference spectrum by taking into account these absorption bands, assuming that each component had Gaussian distribution after baseline correction between 7200 and 6100 cm-1. The mean value of error of the fit (the deviation between measured and calculated curve) was around 0.007. Figure 3 shows an example of the spectrum of the control sample and its resolution into six component bands. Figure 4 shows the changes in ratio of the areal integral after thermal treatment to that before treatment. The ratio of areal integral of all the hydroxyl groups decreased with thermal treatment time. The order of degree of decrement was that of the amorphous region > the semicrystalline region > the crystalline region (6460 and 6281 cm-1) in the cellulose > the phenolic hydroxyl group. This explains that the degradation of hydroxyl group in
288 Biomacromolecules, Vol. 9, No. 1, 2008
Mitsui et al.
In this study, the discussion on other functional groups (for example, CH groups in wood or the hydroxyl group in hemicellulose) has not been carried out yet. We will discuss this point by measuring some kinds of lignocellulosic samples.
Conclusions
Figure 5. Changes in weight of spruce wood by thermal treatment.
We discussed the variation of molecular structure of wood with thermal treatment using NIR spectroscopy. The absorption peak at 6913 cm-1 was conclusively assigned to the phenolic hydroxyl groups due to the lignin in comparison with the spectrum of acetylated spruce. It was suggested that the hydroxyl groups in the cellulose degraded in following order: amorphous, semicrystalline, and crystalline region. There was a clear correlation between the changes in weight of wood and the decrement of hydroxyl groups in the cellulose by thermal treatment.
References and Notes (1) (2) (3) (4) (5) Figure 6. Relationship between the ratio of weight and the ratio of areal integral after heat treatment to that before treatment. b: amorphous region (7003 cm-1), 0: crystalline region (6281 cm-1).
the cellulose starts predominantly from the amorphous region and follows to the semicrystalline and the crystalline regions. On the other hand, the phenolic hydroxyl group decreased more than the hydroxyl groups in the cellulose by acetylation. The phenolic hydroxyl group was stable to thermal treatment; however, its reactivity was the highest against acetylation in the hydroxyl groups in wood. Nuopponen et al.28 and Windeisen et al.29 reported that the percentage of phenolic hydroxyl group increased with rising temperature for thermal treatment in comparison with untreated wood. However, the phenomenon was observed in this study because the treatment temperature was relatively lower than their experimental results. Both the amorphous and semicrystalline regions in the cellulose had a comparable reactivity to acetylation, and it was higher than those of the crystalline region. The reactivity of crystalline region at 6460 cm-1 was higher than that of the crystalline region at 6281 cm-1; however, the difference of cellulose structure between the crystalline regions at 6460 and 6281 cm-1 has not been discussed yet. Both heat treatment and acetylation of wood make the dimensional stability improve, however, the mechanism might be different because of the difference of reactivity in hydroxyl groups against thermal treatment and acetylation. Changes in Weight of Wood by Thermal Treatment. Figure 5 shows the changes in weight of the spruce wood by thermal treatment. The residual weight decreased with treatment time, and this result concurred with previous reports.1,30,31 Figure 6 shows the relationship between the ratio of weight and the ratios of areal integral in amorphous region (7003 cm-1) and crystalline region (6281 cm-1) after heat treatment to that before treatment. The correlation coefficients between the ratio of weight and the ratios of the areal integral in amorphous and crystalline regions after heat treatment to that of untreated wood were 0.97 and 0.88, respectively. This indicates that the decrement in weight of wood caused by thermal treatment is directly related to the decrement in the hydroxyl groups in the cellulose.
(6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26)
(27)
(28) (29) (30) (31)
Stamm, A. J. Ind. Eng. Chem. 1956, 48, 413–417. Millett, M. A.; Gerhards, C. C. Wood Sci. 1972, 4, 193–201. Mitchell, P. H. Wood Fiber Sci. 1988, 20, 320–355. Bekhta, P.; Niemz, P. Holzforschung 2003, 57, 539–546. Stamm, A. J.; Burr, H. K.; Kline, A. A. Ind. Eng. Chem. 1946, 38, 630–634. Burmester, A. Holz Roh- Werkst. 1975, 33, 333–335. Inoue, M.; Norimoto, M.; Tanahashi, M.; Rowell, R. M. Wood Fiber Sci. 1993, 25, 224–235. Morita, S.; Yamazumi, T. Mokuzai Kogyo 1987, 42, 266–272. Tolvaj, L.; Horváth-Szováti, E.; Sáfár, C. DreV. Vysk. 2000, 42 (2), 25–32. Mitsui, K.; Takada, H.; Sugiyama, M.; Hasegawa, R. Holzforschung 2001, 55, 601–605. Mitsui, K.; Murata, A.; Tsuchikawa, S.; Kohara, M. Col. Res. Appl. 2004, 29, 312–316. Fengel, D.; Wegener, G. Wood. Chemistry, Ultrastructure, Reactions; Walter de Gruyter: Berlin, 1989. Tsuchikawa, S. Applied Spectrosc. ReV. 2007, 42, 43–71. Tsuchikawa, S.; Murata, A.; Kohara, M.; Mitsui, K. J. Near Infrared Spectrosc. 2003, 11, 401–405. Tsuchikawa, S.; Inoue, K.; Mitsui, K. Forest Prod. J. 2004, 54 (11), 71–76. Tsuchikawa, S.; Siesler, H. W. Appl. Spectrosc. 2003, 57, 667–674. Tsuchikawa, S.; Siesler, H. W. Appl. Spectrosc. 2003, 57, 675–681. Yonenobu, H.; Tsuchikawa, S. Appl. Spectrosc. 2003, 57, 1451–1453. Mitsui, K.; Tsuchikawa, S. Holz Roh- Werkst. 2003, 61, 159–160. Kelley, S. S.; Rials, T. G.; Groom, L. R.; So, C. L. Holzforschung 2004, 58, 252–260. Schimleck, L. R.; Evans, R. Holzforschung 2004, 58, 66–73. Schimleck, L. R.; Stürzenbecher, R.; Jones, P. D.; Evans, R. J. Near Infrared Spectrosc. 2004, 12, 55–61. Schwanninger, M.; Hinterstoisser, B.; Gradinger, C.; Messner, K.; Fackler, K. J. Near Infrared Spectrosc. 2004, 12, 397–410. Schwanninger, M.; Hinterstoisser, B.; Gierlinger, N.; Wimmer, R.; Hanger, J. Holz Roh- Werkst. 2004, 62, 483–485. Osborne, B. G.; Fearn, T. Near Infrared Spectroscopy in Food Analysis; Longman Scientific & Technical: Essex, 1988. Siesler, H. W.; Ozaki, Y.; Kawata, S.; Heise, M. Near Infrared Spectroscopy: Principle, Instrumentation, and Applications; John Wiley-VCH: Weinheim, 2002. Williams, P.; Norris, K. Near Infrared Technology in the Agricultural and Food Industries; American Association of Cereal Chemists, Inc.: St. Paul, MN, 1990. Nuopponen, M.; Vourinen, T.; Jämsä, S.; Viitaniemi, P. J. Wood Chem. Technol. 2004, 24, 13–26. Windeisen, E.; Strobel, C.; Wegener, G. Wood Sci. Technol. 2007, 41, 523–536. Rusche, H. Holz Roh- Werkst. 1973, 31, 307–312. Fung, D. P. C.; Stevenson, J. A.; Shields, J. K. Wood Sci. 1974, 7, 13–20.
BM7008069
