SSRG International Journal of Chemical Engineering Research (SSRG-IJCER) - Volume 7 Issue 2 May to Aug 2020 ISSN: 2394 - 5370 www.internationaljournalssrg.org Page 24 Modelling of Soxhlet Extraction of Lemongrass Oil Samson Onoriode Okpo 1* and Ipeghan Jonathan Otaraku 2 1 Department of Chemical Engineering Technology, Delta State Polytechnic, Ozoro, Delta State, Nigeria 2 Department of Chemical Engineering, University of Port Harcourt, Nigeria ABSTRACT Soxhlet extraction of lemongrass oil using ethanol as a solvent was carried out, and regression modeling was done with Microsoft Excel 2010 and SPSS version 23. The results obtained from the extraction process were fitted into different regression models to select an appropriate model for the extraction process using their coefficients of regression (R 2 ) and significance values (p-value) as the basis for selection. The effects of particle size, contact time, and solvent volume on oil yield were considered in the modeling. The proposed regression model for effect of particle size, contact time and solvent volume on yield are y = -0.246In(x) +1.4147, y = - 0.00002510x 2 + 0.01548x- 0.6898 and y = - 0.0000001714x 2 + 0.000113x +1.600 respectively. The optimum yield on applying the proposed models was 1.586 for 0.5cm particle size, 1.696 for 300 minutes contact time, and 1.6179 for 300ml solvent volume. To maximize the lemongrass's oil yield using Soxhlet extraction, a particle size of 0.5cm, contact time of 300 minutes, and solvent volume of 300ml is recommended. Finally, the proposed model equations can be used satisfactorily to predict any value of yield for Soxhlet extraction of lemongrass essential oil within the defined experimental range of values. Keywords: Essential oil, Lemongrass, Modelling, Operating parameters, Soxhlet extraction I. INTRODUCTION Lemongrass is a tropical perennial (all seasons) plant belonging to Graminae(Poaceae) family and genus Cymbopogon. The plant has long, thin leaves and is largely cultivated as a medicinal plant in parts of tropical and subtropical areas of Asia, Africa, Australia, Europe, and America ([11]; [21]; [25]; [5]). The leaves of lemongrass and its oil have a lemon-like flavor due to its citral content. Dry leaves of lemongrass contain approximately 1%-2% essential oil [3]. The oil has a light yellow color. The essential oil composition of lemongrass does vary with agronomic treatment, climatic conditions, and geographical locations. Many techniques of extracting essential oil of plant origin include steam distillation, solvent extraction, Soxhlet extraction, hydro-distillation, hydro-diffusion, enfleurage, maceration, expression, destructive distillation ([9]; [11]; [7]). Soxhlet extraction techniques involve solid/liquid contact for removing one or more chemical compounds from solid materials by dissolution in liquid reflux. In a conventional Soxhlet extractor, the solid material is put into the thimble of the extractor. It is gradually filled up with the extracting liquid phase by condensing the vapors from the distillation flask. When the solvent gets to a particular level, a siphon pulls the thimble contents into the distillation flask, thus carrying the extracts into the bulk liquid [20]. The process is continued for the chosen contact time, and each is replicated. Moreover, [4] has reported that many industrial applications require mathematical models for design and effective systems control. Models are simplified mathematical representations of systems at a particular point in time and intended to promote understanding of the real system [2]. Therefore, process modeling involves relating together the properties of a system that are influenced by the process. The outcome is a set of mathematical equations, which is the process model [27]. The process model comprises a set of mathematical formulations or equations that permit us to predict a chemical process's dynamics. Sometimes, to optimize or maximize process operating variables, engineers cannot choose the best operating variables that will minimize operating costs or maximize the profit of a chemical process plant. In a situation like this, the process model and appropriate economic information are used to analyze the prevailing situation and determine the most profitable process conditions [2]. It is worthy to note that mathematical models are useful in developing scale-up procedures from laboratory scale up to pilot plant scale and then industrial scale-up allowing alternative strategies to evaluate the selection of the process variable conditions [18]. In modeling with Microsoft Excel, different trendlines, including linear, polynomial (quadratic and cubic), exponential, logarithmic, and power regression models, can be obtained. But, Middleton [14] suggested that the exponential and power model transform data before the fit, resulting in inaccurate best fit and regression (R 2 ). In addition, both power and exponential curves are used to fit data that increase or decrease at a high rate, and