Prediction of Nuclear Magnetic Resonance Carbon Fractions in Decomposing Forest Litter Using Diffuse Reflectance Infrared Fourier Transform Spectroscopy and Partial Least Squares Regression

Abstract

A diffuse reflectance infrared Fourier transform (DRIFT) spectroscopy method was developed to enable DRIFT to be used as a substitute for 13C-nuclear magnetic resonance (13C-NMR) spectroscopy in predicting specific functional groups containing carbon. As part of the Canadian Intersite Decomposition Study, samples of 10 foliar litter types (trembling aspen, American beech, bracken fern, black spruce, Douglas-fir, plains rough fescue, jack pine, tamarack, white birch, western redcedar) and one wood type (western hemlock) at one site and a subset of three foliar litters (trembling aspen, black spruce, plains rough fescue) at three other colder sites undergoing field exposure for 12 years were annually collected. The DRIFT spectra were collected for all samples, with a subset of litter samples also analyzed by 13C-NMR spectroscopy with cross-polarization and magic-angle spinning. Partial least squares calibrations were calculated from the DRIFT spectra for the seven NMR regions representing specific carbon-containing functional groups. These calibrations were then used to predict the proportion of each NMR region in each sample. A single nondestructive sampling using as little as 0.5 g of sample gave measurements for all of the NMR regions. The DRIFT was demonstrated as a fast and simple analysis method for analyzing large numbers of samples to give fair estimates of the NMR regions for each litter type at all four sites.