STATISTICA® software was utilized to perform regression modeling of bio-oil yield. The standard least square model was used to run the full model as described in Eq. (5). The model results were analyzed statistically and graphically. The deviation of actual experimental data was observed from the predicted values. The overall fitness of the predicted model was checked by the coefficient of determination R2 value, which provides the percentage of valid model. Moreover, the significance of higher R2 value was reviewed by F-test method using ANOVA table. The mean square is based on degree of freedom and sum of square of regression model. The degree of freedom is the number of estimated model based parameters and the number of observations. The Fvalue calculated from table was compared with F0.05, which account for degree of freedom and desired confidence level. For this study, 95% confidence level or p value ⩽0.05 was selected for all experimental runs.
2.1. GC–MS analysis of bio-oil
The bio-oil chemical composition were analyzed using Agilent Technologies 6890 GC–MS with HP-5MS capillary column 30 m length and 0.00025 m diameter. The GC oven temperature was raised from initial temperature of 80–200 °C at the rate of 10 °C/min, and then to 300 °C by maintaining 5 °C/min. The final oven temperature of 300 °C was held constant for about 10 min. Helium gas flow rate was maintained at 2 ml/min. The GC was connected to Agilent Technologies 5975 series Mass Spectroscopy (MS) equipped with inert Mass Selective Detector (MSD) operated at scanning acquisition mode. The MS conditions were: mode Electron Ionization (EI), ion source temperature 230 °C, emission current 34.6 μA, ionization energy 70 eV, full scan range of 50–550 and quantification by Selected Ion Monitoring (SIM) mode. The Agilent Chemstation software was used to indentify chemical compounds and peaks with the help of NIST library.