The further use of wood resources is expected in an environmentally conscious society. Added-value, such as durability enhancement and preservation by painting, are needed to expand the applicability of wood. Assessment of wood properties such as surface and coat adhesion can be made by studying perviousness to liquid oils, with the aim of developing wood products that deter insects and are weather-resistant; hence, discriminant analysis of oil type is important. Near-infrared (NIR) spectroscopy is a powerful tool for nondestructive characterization of organic materials and has been widely used in many industries. Here, NIR detection of oil on wood surfaces is applied for the distinguishing of three different types of oil (hereafter, “Oil_1”, “Oil_2” and “Oil_3”) via soft independent modeling of class analogy (SIMCA). Oil_1 was antiseptic vehicle or cutting oil. Oil_2 was used as a motor oil for an oil pressure machine. Oil_3 was plant-derived oil. Two types of wood that are commonly used in Japanese construction (Cryptomeria japonica and Chamaecyparis obtuse) were analyzed after applying oil. The NIR spectra measured after the oil was applied were greater in the ranges 1700–1800 nm and 2300–2500 nm than spectra for the bare wood sample. As SIMCA analyses were performed by using spectral data that included the moving average, baseline correction and second derivatives, good results were obtained for Oil_3 for both wood samples. However, the correct classification percentages were low for Oil_1, and the percentage of samples classified within several categories was high. If the components are very different, such as those for Oil_3, NIRS can be a powerful non-destructive method for identifying oil in the context of wood products testing.
Read full abstract