High-throughput Quality Characterization of Warm Season Lignocellulosic Feedstocks

The potential of using plants as a renewable resource for energy and high-value products has fueled research in developing high-biomass dedicated feedstock crops for conversion to liquid fuels. Feedstock quality (cell wall composition and energy density) is a major component of liquid fuel yield from plant biomass. Variation in quality traits can significantly impact conversion efficiency and costs. Rapid and accurate characterization of feedstock quality in terms of the relative proportions of nonstructural carbohydrates, cellulose, hemicellulose, and lignin is important in determining the suitability of different feedstocks for various biofuel conversion technologies. In this study, calibration equations for rapid, nondestructive estimation of feedstock quality using near-infrared spectroscopy (NIRs) were developed using cell wall composition data obtained from wet chemistry analysis. Ninety-six genotypes of energy grasses including Saccharum spp., Miscanthus, Panicum virgatum, and Pennisetum purpureum were used to develop calibration models which were subsequently validated with over one thousand samples from different genera. In all samples, cellulose was the dominant cell wall component with dry mass concentrations ranging from 32% to 46%. Hemicellulose and lignin concentrations ranged from 15–24 and 19–48, respectively. Energy grasses and hybrids from the Saccharum family had the highest lignin and cellulose concentrations. Least squares calibration equations relating NIRs spectra and wet chemistry data had high coefficients of determination (R² > 95%) for cellulose and lignin. Factors affecting R² values included number of samples used to develop calibration models, species and genotype, sample age, and sample origin with respect to soil type and fertility management. Whereas high-lignin varieties would be suited for thermochemical conversions, those high in structural sugars would be more appropriate for biochemical conversion processes. The NIRs technique and calibration equations developed in this study are also useful in breeding programs for high throughput screening and genetic improvement of potential feedstocks.