Biodiesel production from used cooking oil is sustainable alternative, for bio-energy production. The process generates residual crude glycerol (RCG) as the major energy-rich waste which can be used to produce various bio-based chemicals like 1,3-propanediol (1,3-PDO) through biotechnological interventions. This RCG contains several impurities like methanol, soap, organic materials, salts non-transesterified fatty acids and metals in varied concentrations. These impurities significantly affect yield and productivity of the bio-process due to their marked microbial toxicity. In this work, previously isolated Clostridium butyricum L4 was immobilized on various abundantly available cheap bio-wastes (like rice straw, activated carbon and corn cob) to explore advantages offered and improve tolerance to various feed impurities. Amongst these, shredded rice straw was found most suitable candidate for immobilization and results in maximum improvement in 1,3-PDO production (18.4%) with highest porosity (89.28%), lowest bulk density (194.48Kg/m3), and highest cellular biofilm density (CFU/g-8.4 ×1010) amongst the three matrices. For practical purposes, recyclability was evaluated and it was concluded that even after reusing for five successive cycles the production retained to ∼82.4%. Subsequently, polynomial model was developed using 30 runs central composite factorial design experiments having coefficient of regression (R²) as 0.9520, in order to predict yields under different immobilization conditions for 1,3-PDO production. Plackett-Burman was employed (Accuracy= 99.17%) to screen significant toxic impurities. Based on statistical analysis six impurities were found to be significantly influential on PDO production in adverse manner. With negative coefficient of estimate (COE) varying in decreasing order: Linoleic acid >Oleic acid >Stearic acid >NaCl>K2SO4 >KCl. The study illustrates practical application for repurposing waste glycerol generated from biodiesel plants, thus developing improved agnostic process along with yield production models.
Read full abstract