Optimization of oil extraction from locust beans (Parkia biglobosa) using central composite rotatable design of RSM was carried out. Independent variables were roasting temperature (90, 100, 110, 120, 130°C) and roasting duration (5, 10 15, 20, 25 min), while the responses were oil yield (OY), free fatty acid (FFA), color (CO), stability, anisidine value (AV), and specific gravity (SG). Data obtained from this study were analyzed using ANOVA and regression analysis. OY, FFA, CO, oil stability, AV, and SG of the oil ranged between (19.0–22.5%), (0.65–3.1%), (37.9–199.9 mg/L), (3.6–6.4%), (11.4–59.9 mg/L), (0.7–0.9), respectively. Roasting conditions significantly influenced all dependent variables at p<0.05. Coefficients of determination (R2) of the generated models ranged from 0.49 to 0.98. Optimum roasting condition was 108°C for 10 min. This gave 22.3% OY, 1.7% FFA, 54.9 mg/L CO, 4.1% stability, 16.5 mg/L AV, and 0.9 SG. Desirability of 0.8 was obtained for optimum conditions.
Practical applications: Generally, African locust bean is fermented to give condiment. It has not been used for commercial oil production despite its richness in oil, but soybean with a similar oil content of about 18–20% has been used both for condiment and commercial oil production. Research into the potential of locust bean as a source of edible oil is desirable. Roasting is among pre-treatment methods employed in vegetable oil industry. Information on degree of influence of these parameters on quantity and quality of oil recovered oil from African locust benefits both entrepreneur and researchers. Also, the developed mathematical expression for OY, FFA, CO, oil stability, AV, and SG, as influenced by roasting temperature and duration is an effective tool in predicting these parameters. The research output has both immediate and potential application.