Dunaliella salina is able to produce simultaneously several valuable compounds (such as lipids, carotenes and functional proteins) within the biorefinery concept. However due to the lack of rigid cell wall, this microalgae can easily disrupt during harvesting, losing valuable compounds to the saline water, affecting the downstream processing. Therefore, the development of non-invasive tools able to monitor cell concentration and integrity in real-time, can assist the development of harvesting methodologies. In the present work, a monitoring approach was developed based on two-dimensional (2D) fluorescence spectroscopy. Mathematical analysis of the monitoring data involved the use of Principal Component Analysis (PCA) and Projection to Latent Structures (PLS) modelling. For green D. salina, the models developed for prediction of cell number and percentage of viability captured 90.6% and 86.3% of variance, respectively. Both models have R2 of 0.8 and 0.9, respectively for validation and training. Similar values were found for the prediction of cell number when using data from growth kinetics and harvesting combined. Orange D. salina rupture was also successfully modelled with 95% of variance captured and R2 of 0.9 for both training and validation. The combined approach using 2D fluorescence spectroscopy and the mathematical analysis proved to have the potential to monitor D. salina during cell growth and harvesting within a biorefinery concept.
ASJC Scopus subject areas
- Agronomy and Crop Science