Nelson, Paul N., and Baldock, Jeffrey A. (2005) Estimating the molecular composition of a diverse range of natural organic materials from solid-state 13C NMR and elemental analyses. Biogeochemistry, 72 (1). pp. 1-34.

PDF (Published Version) - Published Version Restricted to Repository staff only

 

139

7

Abstract

Most techniques for determining the chemical nature of natural organic matter in soil, sediment and water require prior extraction or concentration steps that are not quantitative and that create artifacts. 13C nuclear magnetic resonance (NMR) analysis can avoid these problems, but it gives little information at the scale of molecules. Here we show that the molecular composition of a diverse range of natural organic materials could be inferred from 13C NMR analysis combined with C and N analysis. Forty-six different organic materials including undecomposed and decomposed plant materials, soil organic matter, phytoplankton, and the organic matter found in freshwater, estuarine and marine sediments were examined. A mixing model simultaneously solved a series of equations to estimate the content of four biomolecule components representing the organic materials produced in greatest abundance by plants and other organisms (carbohydrate, protein, lignin and aliphatic material) and two additional components (char and pure carbonyl). Based on defined molecular structures for each component, signal intensities for 13C NMR spectra were predicted and compared with measured values. The sum of the absolute differences in signal intensity between the measured and predicted spectral regions was <7% for the terrestrial materials. For aquatic materials the fit of the predicted to measured signal intensities was not as good. Predicted molecular compositions correlated well with independent analyses of cellulose, protein and lignin contents of plant samples and char contents of soil samples. Across all samples, carbohydrates accounted for 10-76% of the sample C (40-76% in plants and 10-42% in soils, sediments and phytoplankton), protein for 2-80% (21-80% in phytoplankton and marine water column samples and 2-36% in plants, soils and sediments), lignin for 0-36%, aliphatic materials for 2-44%, char for 0-38% and carbonyl for 0-22%. For the soils, sediments and decomposed plant materials, the close correspondence between actual signal intensities and those predicted using known biomolecular components, suggested that either ‘humic’ structures can be approximated by mixtures of common biologically derived molecules or that humic structures did not exist in significant amounts.

Actions (Repository Staff Only)

Item Control Page