Raman Spectroscopy as a Tool for Real‐Time Nutrient Monitoring in Bioreactor Cultivation of Microalgae

In this work, aqueous nutrient solutions replicating bioreactor culture media for microalgae were analyzed using spontaneous Raman spectroscopy. Focusing on nitrate, sulfate, glucose, and phosphate, the study evaluated their potential for real-time monitoring in cell cultivations such as Chlorella vulgaris. Univariate analysis, based on Raman intensities of specific nutrient peaks, was conducted and compared to multivariate analysis results. Four multivariate calibration models were developed using partial least squares regression (PLSR), achieving high calibration and validation performance, with R2 values above 0.99 and low RMSECV, indicating strong calibration accuracy. The study also examined the limit of detection (LOD) for each nutrient, finding that LODs for nitrate, sulfate, and glucose reached levels relevant for algae bioreactors even without the application of enhanced Raman techniques. To further validate the PLS models, independent real bioreactor samples were analyzed, showing strong predictive accuracy (RP2: 0.9661–0.9892) and low RMSEP values. Additional testing with five samples collected over a Chlorella vulgaris cultivation run (day 0 to day 9) confirmed the models' robust performance under real bioprocess conditions. Limitations in practical applications, such as phosphate's relatively high LOD, were also identified. The results suggest that Raman spectroscopy, combined with multivariate analysis, could deliver precise and reliable detection of critical nutrients and their concentrations in bioreactor culture media. This potential of the Raman technique, along with insights into nutrient LODs, PLS model accuracy, and practical application challenges, provides a solid foundation for future research and development in industrial bioprocess monitoring.