Introduction I t is of great importance to investigate the degradation mechanism of antique wooden materials used for historical temples, buildings and crafts from the viewpoint of wood science and archaeology. Near infrared (NIR) spectroscopy is a powerful tool for measuring the physical, chemical and mechanical properties of wood whilst retaining the anatomical structure of the archaeological samples. Cell walls of wood are composed of cellulose, hemicelluloses, lignin and extractives. The concentration of extractives is relatively low. The cellulose molecules form microfibrillar aggregates which are embedded in a matrix of hemicelluloses as well as lignin. As physical, chemical and mechanical properties of wood are governed by the complex arrangement of these polysaccharides, their variation with degradation should strongly affect the wooden materials. Knowledge on the aging of archaeological wood has mainly been obtained from those preserved in saturated soils or bogs in water-logged conditions. On the other hand, many important wooden materials have been preserved and repaired in non-closed environments in Japan, i.e. the Horyuji temple (constructed in the 7th century). We examined the temporal change in properties of archaeological wood and the possibility of non-destructively evaluating archaeological wood properties (i.e. age, strength, chemical components) by NIR spectroscopy with the aim of establishing NIR archaeometry. One reasonable approach for investigation of the degradation mechanism of wood is to employ thermally-degraded wood samples. As thermal treatment is one degradation process, this technique is sometimes applied to archaeology when artificially degraded wood is utilised for the repair of historical wooden artefacts. In this article, the changes in mechanical properties and crystalline structure with aging or thermal treatment, as investigated by NIR spectroscopy, are explained. Prediction of compressive Young’s modulus of hydrothermally treated wood by NIR spectroscopy The changes in compressive Young’s modulus of hydrothermally treated softwood and hardwood as a function of thermal treatment time were measured. Hinoki cypresses (Chamaecyparis obtusa) and beech (Fagus creneta) were hydrothermally treated at 140°C under steamed atmospheric conditions in order to artificially degrade wood to create an analogue of archaeological objects. NIR spectra and compressive Young’s modulus of artificially degraded wood were systematically measured. Figure 1 shows the change in compressive Young’s modulus of (a) hinoki, (b) beech and the relationship between measured compressive Young’s modulus and that predicted from NIR spectra of (c) hinoki and (d) beech by partial least squares (PLS) regression analysis. Compressive Young’s modulus of hinoki decreased as a function of treatment time. Conversely, that of beech rapidly increased with treatment time and subsequently gradually decreased. These calibrations demonstrated that NIR spectroscopy could predict the compressive Young’s modulus of hydrothermally-treated hinoki and beech wood samples with sufficient accuracy, although this parameter varies greatly within the same thermal treatment time. Figure 2 shows the plot of the first and second PLS loadings. The first loading plot, ranging from 10,000 cm to 8800 cm, where absorbance of NIR spectra increased with increments in treatment time, showed a negative value. On the other hand, the loading ranging from 8800 cm to 5000 cm due to polysaccharide absorption bands, where absorbance of NIR spectra decreased with increments in treatment time, showed a positive value. Therefore, the first loading might be regarded as a parameter reflecting the degree of hydrothermal degradation (i.e. depolymerisation of polysaccharide). On the other hand, the second loading has characteristic positive peaks observed around 7000 cm due to the amorphous region in cellulose and 5200 cm due doi: 10.1255/nirn.1394
Read full abstract