Optimisation, experimental validation and thermodynamic study of the sequential oil extraction and biodiesel production processes from seeds of Sterculia foetida.

Non-edible seeds are not used in any commercial applications, which implies that they can be used for biofuel applications. The present study aimed to maximise the process conditions for oil extraction and sterculia biodiesel production from Sterculia foetida (poon oil). GC-MS identified the methyl esters of sterculia oil as sterculic acid (32%), palmitic acid (15.88%), oleic acid (10.00%), linoleic acid (9.95%) and malvalic acid (9%). Response surface methodology (RSM) based parametric optimisation of oil extraction was carried out by choosing process variables such as sample weight, volume of solvent to seed ratio and time. The optimum sample amount of 7.5 g and the volume of solvent to seed ratio of 40 mL/g resulted in a maximum oil yield of 45.27% at 3 h. The results were statistically significant (P < 0.05) with a regression coefficient (R2) of 0.9988. Furthermore, the artificial neural network (ANN) resulted in an R2 value greater than 0.9, which validates the RSM. Conventional optimisation of the temperature (55 °C), feedstock to methanol ratio (1:12), catalyst proportion (1.5%) and transesterification reaction time (60 min) yield 90.87% biodiesel production. The physicochemical characteristics of oil and biodiesel complied with the requirements of the ASTM standards. The rate constant and thermodynamic variables at the optimum temperature (333 K) were calculated from the experimental data. The activation energy (Ea), activation enthalpy in transition state theory (ΔH++), activation entropy in transition state theory (ΔS++) and Gibbs free energy in transition state theory (ΔG++) were 37.91 kJ mol-1, 35.14 kJ mol-1, - 239.58 J mol-1 K-1 and 79.81 kJ mol-1 respectively. Graphical abstract.