【作者】 周冠武；

【导师】 段新芳；

【作者基本信息】 中国林业科学研究院 ， 木材科学与技术， 2011， 博士

【摘要】 杨树(Populus spp.)生长迅速、繁殖容易、适应性强、抗逆性强、木材用途广,在我国分布地域广泛,杨树人工林木材的加工利用与木材性质及木质素含量联系紧密；傅里叶变换红外光谱分析技术是一种有很大潜力和前途的无损评价预测技术,木材物理化学性质的红外光谱快速评价研究一直倍受世界各国研究人员的广泛关注。红外光谱技术快速预测杨木木质素含量不仅是研发绿色酶法胶合人造板技术中关键的一环,而且还是培育生物能源及造纸用材林树种必不可少的筛选分析技术。人造板中以游离甲醛为主的挥发性有机物污染问题日益受到人们的重视,环境友好型人造板也已经成为世界性难题,因此急需开发既减轻环境负荷又利用可再生资源的人造板制备新技术。通过傅里叶变换红外-衰减全反射光谱技术结合多元数据分析方法,建立多元木质素预测模型,筛选适宜酶法胶合人造板制造的人工林杨树品种,为后期板材制备实验的原材料选定提供指导,进一步阐明漆酶活化木材木质素催化反应的反应机理,探索酶法纤维板的干法制备工艺及对所得板材性能指标进行检测对比,确立漆酶活化杨木纤维制备酶法纤维板的工艺参数,实现环境友好型木基复合材料的清洁生产,为推动杨树人工林木材纤维漆酶活化制备纤维板技术朝向产业化方向发展提供理论参考,对解决人造板中的游离甲醛污染、改善人们生活居住环境、保护人类赖以生存的自然环境都具有重要意义。本论文探讨傅里叶变换红外光谱技术结合多元数据分析算法快速预测杨木木质素含量建立标定模型及优化的各种影响因素,考察红外光谱技术定性评价木质素、纤维素等木材主要化学组分的可行性,探索木纤维漆酶催化氧化干法工艺制备酶法纤维板,杜绝使用合成树脂胶,从源头上排除甲醛污染,建立漆酶活化杨树木纤维中间产物直接检测及定量的电子自旋共振波谱分析方法,研究漆酶催化氧化反应的自由基影响及作用。本论文的主要研究结论如下：1.毛果杨×美洲黑杨杂交种(P. trichocarpa×deltoides)无抽提物木材试样的木质素含量范围为23.45%到32.07%(w/w),呈正态分布,均值为27.02%。采用改进的乙酰溴木质素测定方法,本批次样品分析的标准偏差为0.18%,合并标准离差仅为0.042%,适用于次级分析方法标定样品的木质素含量分析。2.毛果杨×美洲黑杨杂交杨树木材无抽提物试样化学组分的种内天然变异性适用于构建木质素含量预测模型。傅里叶变换红外-衰减全反射光谱技术与偏最小二乘回归建模技术相结合,经过外部验证的最佳标定模型决定系数高,R2(标定)为0.906,R2(交叉验证)为0.806,交叉验证均方根误差低,仅为0.77%,独立木材试样数据集验证最佳预测模型所得R2为0.88。杨树木材木质素含量与能量含量的种内变异性互不相关,通过主成分分析识别对这两个特性差异起关键作用的红外吸收峰波数,木质素的前4个因子载荷谱包含32个最大差异波数中的14个指认为芳香族化合物,而能量含量的前4个因子载荷谱中仅有7个最大差异波数指认为芳香族化合物,普遍为碳氢化合物的环振动。3.基因改良无异戊二烯释放灰杨(Populus×canescens)与哥廷根野生型杨树木材木质素含量范围为24.06%到26.59%,α纤维素含量为42.95%到47.73%。聚类分析所得树形图表明,基因改良无异戊二烯释放灰杨木材与野生型杨树木材的木质素、纤维素、可溶性抽提物等化学组分没有显著差异。传统湿化学方法测定可溶性抽提物(0.96%至1.83%)、综纤维素(70.17%至74.47%)、α纤维素、木质素含量、能量含量(17690 J/g至18280 J/g)的研究结果均验证了聚类分析所得的结果。对因子载荷谱中最高的七个峰进行尝试性特征化学振动官能团指认分析发现,第二、三、四个因子载荷谱中21个差异显著的吸收峰中的12个对应于碳氢化合物的环振动所占据的波数范围,但也含有少部分木质素的官能团所对应的波数范围,仅7个差异较大波数。4.参考干法工艺中密度纤维板热压曲线,通过热电偶测定板坯芯层温度,确定干法工艺漆酶纤维板热压工艺的具体参数,具体热压工艺参数为：热压温度190℃,最大压力保持段压力5 MPa,时间1.5 mmin；低压塑化段压力3.5 MPa,时间3 min。在此热压条件下压制干法工艺漆酶纤维板,酶法纤维板的内结合强度显著高于对照板,在同样工艺条件下,对照纤维板的内结合强度为0.19 MPa,酶用量为5.58 U/g绝干木纤维漆酶纤维板的内结合强度为0.53 MPa。纤维板密度达到一定水平才能得到内结合强度较高的酶法纤维板。铜离子可显著提高酶法纤维板的内结合强度。5.采用电子自旋共振波谱自旋捕集剂技术以N-叔丁基-α-苯基硝酮为自旋捕集剂,然后进行乙酸乙酯抽提,鉴别并定量漆酶催化氧化木纤维自由基反应中间产物,N-叔丁基-α-苯基硝酮所捕集自由基的电子自旋共振波谱图的g值为2.005,aN为15.0 G,为超氧化物和羟基自由基等活性氧物质的电子自旋共振波谱图,这表明活性氧物质是漆酶催化反应的主要自由基中间产物。芬顿反应所得羟基自由基标准曲线的决定系数为0.9799,依据电子自旋共振波谱信号强度与羟基自由基未配对自旋数的定量曲线,确定自由基反应中间产物的绝对自旋数为3.74±0.005×1018自旋数/克木纤维干物质。基于存在活性氧物质的研究发现和与漆酶氧化木纤维自由基反应有关的前人文献,我们提出漆酶催化氧化杨树木纤维的可能反应机制：漆酶介导反应不能直接触及木质素的大部分结构域,因此低分子量可溶性木质素可能重新附着到纤维表面,起着与胶黏剂类似的活性化合物的作用。更多还原

【Abstract】 Poplars(Populus spp.) with various traits including rapid growth, easy propagation, adaptability, stress resistance and wide use of wood are distributed in a wide range of geographical locations. The processing and utilization of poplar wood are closely associated with the plantation wood properties and lignin content. Fourier transform infrared spectroscopy has great potential and prospects as a non-destructive evaluation and forecasting technique. The evaluation of the physical and chemical properties of wood by means of infrared spectroscopy has lured attention from researchers around the world. Rapid prediction of poplar lignin content by Infrared Spectroscopy is not only a crucial chain in the development of technology for the green enzyme-bonded wood composites, but also essential screening analysis in the cultivation of timber tree species in bio-energy and paper-making. The pollution of volatile organic compounds including free formaldehyde has been paid more attention, environment-friendly wood-based panels has become a worldwide problem, therefore there is the urgent need to reduce environmental load and the development of renewable resources, preparation of new technology wood-based panels. The site conditions, genetic characteristics, growth climatic conditions, forest management methods and other factors of wood species have different effects on the physical and chemical properties of wood, in order to achieve high efficiency and optimal use of materials it is required to use Fourier transform infrared-attenuated total reflection spectroscopy combined with multivariate data analysis method to build the prediction model calibration for the overall chemical nature of wood by making fast and accurate forecasts, and thus fully understand the nature of wood. This work is of great significance to address free formaldehyde pollution from wood-based panels and improve the living conditions of people’s lives, protect the natural environment of human life.In this thesis, Fourier Transform Infrared Spectroscopy with Multivariate Data Analysis algorithm makes fast prediction of poplar lignin content. Optimized calibration model is used to establish factors affecting the qualitative evaluation of wood properties through infrared spectroscopy. The establishment of poplar tree-laccase direct detection of intermediate fibers and quantitative analysis is fulfilled through electron spin resonance spectroscopy to study the catalytic oxidation of laccase and role of free radicals, to explore the manufacture of fiberboards made of laccase catalyzed oxidation of wood fibers, which eliminates the use of synthetic resin and excludes the source of formaldehyde contamination.The main findings of this paper are as follows:1. The lignin content of Extractive-free wood samples from hybrid poplar (P. trichocarpa×deltoides) ranges from 23.45% to 32.07%(w/w), showing a normal distribution with a mean value of 27.02%. Modifications to the traditional wet chemical analysis - acetyl bromide lignin assay consist of:The weighing range during the period of weighing wood powder sample strictly centered on 0.99 and 1.01 mg, and wood flour together with polypropylene weighing paper was transferred into micro-tube. Prior to the addition of acetyl bromide/acetic acid mixture, the micro-tubes filled with wood flour was placed into the icy water bath, and acetyl bromide/acetic acid mixture was stored at room temperature before use. Within 10 min immediately after the final dilution by acetic acid, the UV absorbance measurement was supposed to complete. By means of the modified acetyl bromide lignin analysis the standard deviation for the current batch of samples was 0.18%, and combined standard deviation was as low as 0.042%, which proves suitable for analysis of lignin content of samples used for secondary calibration. As for preparation flour of for Fourier transform infrared-ATR spectra, we recommend the following steps:finely ground flour as possible, preferably less than 20 microns particle size; to avoid overheating in the course of grinding, milling under liquid nitrogen environment if possible. Wood flour should be placed for at least 3 days in the laboratory for infrared spectrometer before recording IR spectrum for the purpose of equilibrating moisture content in the wood powder.2. The intra-specific natural variability of chemical composition in extractive-free wood samples of hybrid poplar (P. trichocarpa×deltoides) is sufficient for building lignin content prediction model. Fourier transform infrared spectroscopy combined with partial least squares regression modelling technique yielded optimized calibration model which went through external validation. The resultant optimized calibration model was of high quality, with high coefficients of determination (R2 (calibration)=0.906 and R2 (cross-validation)=0.806) and low root mean square error of cross validation (RMSECV=0.77%) respectively. Verification of the best model through independent data sets of wood samples resulted in R2=0.88. These results showed that FTIR-ATR spectroscopy is suitable as a high-throughput method for poplar lignin estimations in large-scale breeding or genetic engineering projects. Poplar lignin content and energy content within the species variability is unrelated. By means of principal component analysis to identify the key wavenumbers influencing the above-mentioned two traits,14 most divergent wavenumbers in the first four factor loading spectra of lignin containing 32 maximum difference in wavenumbers mainly attributes to aromatic compounds, whereas only 7 most divergent wavenumbers in the first four factor loading spectra of the energy content attributes to aromatic compounds, and is generally hydrocarbon ring vibration.3. The lignin content in genetically modified non-Isoprene releasing grey poplar (Populus×canescens) and wild-type poplar cultivated in Gottingen wood ranges from 24.06% to 26.59%, and a-cellulose content from 42.95% to 47.73%. The dendrogram obtained from Cluster analysis showed that the lignin content, cellulose content, soluble extractive content between genetically modified non-Isoprene releasing grey poplar wood and wild-type poplar wood displayed no significant difference. The semi-quantitative analysis of the fingerprint region of IR spectra collected from poplar wood four showed that absorption peak height ratio of the characteristic absorption peak of cellulose and hemicelluloses to lignin ranges from 2.18 to 2.23, whereas the corresponding ratio for wild-type poplar wood was 2.16; therefore the major chemical components including lignin and cellulose in the two categories of wood showed no significant difference. The determination of the traditional wet chemical methods with soluble extractives (0.96 to 1.83%), holocellulose (70.17% to 74.47%), a-cellulose, lignin content, and energy content (17690 J/g to 18280 J/g) verified the results of cluster analysis. The projection of three-dimensional data of genetically modified grey poplar wood, wild-type poplar wood and hybrid poplar (P. trichocarpa×deltoides) wood as a result of principal component analysis showed that there was slightly better degree of differentiation between genetically modified grey poplar wood and wild-type poplar wood, which corroborated the results of wet chemistry analysis:The soluble extractives content with an average of 1.11% of one progeny in genetically modified grey poplar wood was significantly lower than that of wild-type poplar grown in Gottingen (mean 1.50%). Through tentative assignment of the characteristic functional groups in the highest seven peaks of the factor loading spectra, it was found that 12 out of 21 most divergent absorption bands corresponds to ring vibration with hydrocarbon origin, but also contains a small number of lignin functional groups with only 7 significantly different wavenumbers.4. In reference to MDF dry process hot-press curve, slab core temperature was measured by thermocouples to determine the dry process technology of laccase-bonded fiberboard. The specific parameters of hot pressing are:pressing temperature 190℃, the maximum pressure to maintain the pressure at 5 MPa, time 1.5 min; low pressure plastics section 3.5 MPa, time of 3 min. The enzyme binding strength within the fiberboard was significantly higher than the control board in the same process conditions, control the internal bond strength of fiberboard 0.19 MPa, enzyme dosage of 5.58 U/g oven-dry wood fiber laccase the internal bond strength of fiberboard 0.53 MPa. Fiber density is less than 0.9 g/cm3, the enzyme is very low internal bond strength cf fiberboard, is only 0.17 MPa, fiber density reached a certain level to get a higher bond strength within the enzymatic fiberboard. Copper ions can significantly improve the internal bond strength fiber enzymatic, within the control enzyme binding strength of fiberboard 0.38 MPa, adding copper ions within the enzyme binding strength of fiberboard increased to 0.59 MPa.5. To investigate the radical reaction intermediates during the laccase-catalyzed oxidation of poplar wood fibers, electron spin resonance spin trapping technique using N-tert-butyl-a-phenylnitrone as a spin trap followed by ethyl acetate extraction were empioyed to identify and quantify the reaction intermediates. The electron spin resonance spectrum of free radicals trapped by N-tert-butyl-a-phenylnitrone had a triplet at g value=2.005 and an= 15.0 G, showing reactive oxygen species is the major products of laccase oxidation of wood fibers. The Fenton system produced a standard curve for hydroxyl radical with a determination coefficient of 0.9799. Using reactive oxygen species standard curve, the reactive oxygen species detected via N-tert-butyl-a-phenylnitrone spin trap method was quantified as 3.74±0.05×1018 spins/g wood fiber dry substances. Based on the findings of the presence of reactive oxygen species during the activation of wood fiber with laccase and previous studies on related free radical reactions, we propose the possible reaction mechanism for laccase oxidation of wood fibers:the laccase-mediated reaction cannot directly reach the majority of the lignin domain, the low molecular weight soluble lignin may function as reactive compounds like adhesives, clinging back to the fiber surface.更多还原