A robust, non-invasive and fast routine for the quantification of the nutritional composition of microalgae biomass slurries based on near-infrared spectroscopy

The use of microalgae as a food ingredient is currently limited by both the quantity and quality of biomass. In particular, there is a lack of reliable monitoring systems for the real-time analysis of the nutritional value of microalgal biomass. The analytical tools for measuring the macromolecular composition of microalgae biomass are based on time-consuming laboratory routines that can take up to several days, including resource-intense drying processes and wet-chemical analyses. Consequently, there is a significant delay between the sampling of biomass and the acquisition of detailed information regarding its nutritional quality. This limits the ability to monitor the product in real-time or to manage production processes with a quality-oriented focus. In this study, near-infrared spectroscopy (NIR) was used to generate NIR datasets for microalgal eukaryotes ( Chromochloris zofingiensis ) and cyanobacteria ( Limnospira maxima ) biomasses with high (slurry with approximately 85 %) and low (powder with approximately 3 %) water content. The generated datasets were employed in chemometric modeling, which enabled the prediction of the major macromolecular biomass compounds (protein, lipid, water, ash/mineral content). The chemometric analysis yielded highly predictive regression models (R 2 > 0.88) for protein, fat, ash, and water content in microalgal slurries of C. zofingiensis . Similar outcomes were observed for L. maxima slurry datasets, which demonstrated ‘very good’ predictive capability, and were only restricted when the analytical concentration fell below a critical level of 1 % such as the low lipid content of L. maxima slurries (< 0.5 %). This study indicated that the combination of NIR datasets among different microalgae can extend the measurement range while maintaining the predictive power of the chemometric models. The utilization of NIR spectroscopy on freshly harvested microalgal slurries with a dry weight concentration of appromimately 150 g/L has demonstrated that a non-invasive, multi-component analysis of the nutritional quality of microalgal biomass is feasible with high accuracy within minutes. • NIR analysis of microalgae samples with high and low water content • Chemometric model establishment based on reference analytics • Prediction of the nutritional composition of microalgae with high accuracy • Dataset fusion of different microalgae allow an extension of measurement range