Rapid determination of the lignin content in sitka spruce (Picea sitchensis (Bong.) Carr.) wood by Fourier transform infrared spectrometry

The rapid FTIR (Fourier transform infrared) spectroscopy technique was used to indirectly measure the lignin content of Sitka spruce wood samples. A model was constructed to relate FTIR data to lignin content measured in w ood by chemical analysis (i.e. the modified acetyl bromide method), through the application of the principal component regression (PCR) approach to a set of calibration observations. The acetyl bromide method provided lignin determinations ranging from 24% to 34% with a measurement error of 0.6%. A residual standard deviation of 0.8% and an average prediction error of 0.9% were calculated for lignin content when employing the selected PCR model ( based on a normalized infrared data set, obtained using the band at 1374 cm (-1) as a reference) to indirectly measure this wood property. Furthermore, a large portion of the variability in lignin content was explained by the two principal components retained in the selected calibration model, as ind icated by the magnitude (i.e. 0.93) of the coefficient of multiple determin ation. The model was subsequently employed to predict the lignin content of wood s amples collected in a clonal experiment. Two multivariate diagnostic measur es were applied to assess the quality of the individual predictions, and th e results indicated that the spectral information contained in the new samp le vectors was suitable for use with the selected calibration model.