Optimization and kinetic parameter estimation of bioethanol production from freely available Sri Lankan fruits in batch fermentation

Bioethanol is used in many countries as an alternative to gasoline, mainly due to better emission characteristics. Since Sri Lanka imports fuel for its whole transportation requirement, this research aims optimization and kinetic parameter estimation of bioethanol production from freely available Sri Lankan fruit wastes. Bioethanol production was carried out using three different freely available overripe fruits in Sri Lanka: banana (Musa spp., embul kesel variety), papaya (Carica papaya) and jackfruit (Artocarpus heterophyllus). Kinetic evaluation of the fermentation process and optimization of bioethanol yields were carried out by changing the type of fruits, type of microorganism (Saccharomyces cerevisiae and Pseudomonas mendocino (PM)), concentration of the substrate (1:1, 1:1.5, 1:2 w/w ratio of fruit and water), pH (4.3, 5.0, 5.7), and temperature (27, 32, 35 oC). Estimation of the kinetic parameters was carried out using Monod and Modified Gompertz equations by fitting experimental data. The best parameters found for the fermentation of banana using PM microorganism were 1:1 (w/w) concentration at pH 5 and 35 oC with a maximum bioethanol yield of 13 vol%. Experimental data were well fitted with Monod model and Gompertz model, with a higher regression coefficient of R2 99.81 and R2 > 99.37 respectively. Comparative to the reported, 12.8 vol% bioethanol yield from banana by fermentation using PM microorganism at room temperature a higher bioethanol yield of 13 vol% could be obtained by fermentating banana fruit mass using PM microorganism with 1:1 (w/w) concentration at pH 5 and 35 oC temperature.


INTRODUCTION
Bioethanol has been identified as an attractive alternative biofuel source for fossil fuel due to several merits such as the possibility of using bioethanol as a transportation fuel, reduction of its greenhouse gas emission, the availability of various renewable feedstocks, etc. Presently, bioethanol is produced from sugar-containing feedstocks such as sugar cane, molasses, etc., starch-containing grains; corn, wheat, etc., lignocellulosic materials such as agricultural residues, wood, etc. and waste materials: municipal waste, vegetable waste, fruit wastes, etc. (Goyal et al., 2008;Manochioa et al., 2017;Mustapa et al., 2008). Bioethanol, blended with gasoline (less than 10 % of ethanol) is widely used in many countries as a transportation fuel with existing gasoline engines because of improved emission characteristics (Mustapa et al., 2008).
Nevertheless, owing to the high production cost of ethanol, many researchers have focused more on investigating into freely available efficient bioethanol substrates and microorganisms for higher bioethanol yield (Faradiella et al., 2017). Overripe fruit mass (obtained from waste fruits) has been identified as an ideal substrate for bioethanol production.
Many fruits are well grown in Sri Lanka. Out of the most available fruits, banana, papaya, and jack fruits have been identified as the most sugar-rich fruits which could be used efficiently for bioethanol production (Kularathne et al., 2016). Overripe fruits have been identified as proven feedstocks to produce bioethanol by fermentation using Saccharomyces cerevisiae since those fruits contain simple sugars such as sucrose, fructose, maltose as well as reducing sugars such as glucose, xylose etc. (Jahid et al., 2018). However, fruit wastes such as peels which contain cellulose are with a great constraint to access for the production of fermentable sugars. Such lignocellulosic materials should be subjected to pretreatments such as physical, chemical and biochemical pretreatments before the fermentation process (Behera et al., 1996).
Carbohydrate is available in three types as sugars, starches and fibres. Starches are considered as the complex sugars. Unripe fruits mainly contain starch. The amount of starch content of fruits varies depending on the type of fruit and the ripen stage. At the harvest, fruits contain a high percentage of starch and low percentage of sugar. Comparative to other fruits, unripe banana contains 20% starch and 1% sugar at the time of harvest. As the fruits ripen, formation of sucrose takes place first, however, nearly a constant amount while the amount of fructose and glucose content increases until the end of the ripening processes. A fully ripped banana, which is yellow with some brown spots, contains 14% fructose, 20% glucose and 66% sucrose (Macrae et al., 1992). .
Fruit ripening process takes place in association of transient respiration. It involves a series of events including changes in physiological and biochemical levels resulting in softening, sweetening, decreased bitterness, colour alterations, etc. The major changes can be categorized into three main stages such as starch degradation, tricarboxylic acid (TCA) cycle, glyoxylate cycle, sucrose synthesis and degradation (Kumar et al., 2019).
Starch is composed of amylose and amylopectin. During ripening, starch is broken down into simple sugars. Various hydrolytic enzymes such as α-amylase, β-amylase, debranching enzyme, α-1 -4 glucosidase, etc. carry out its degradation processes by breaking glycosidic bonds between glucose monomers of starch. Unripe fruits contain organic acids and these acids are stored in cell vacuoles. These acids are utilized in the tricarboxylic acid (TCA) cycle and glyoxylate cycle for the respiration processes. In most ripen fruits, there is a decrease in acidity due to TCA utilization in the respiratory process (Boggio et al., 2000). The cell wall degradation and softening of fruit during ripening take place with the action of various enzymes such as pectinases, cellulases, hemicelluloses and other texturesoftening enzymes such as e-glycosidases, α-galactosidase, α-glucosidase, β-glucosidase, α-mannosidase, β-mannosidase, α-xylosidase, and β-xylosidase.
Presently, Sri Lanka depends on imported fossil fuel to fulfil the transportation fuel requirement. The possibility of producing first-generation bioethanol in Sri Lanka is limited due to the unavailability of extra agricultural lands. However, it has been estimated that approximately 270000 tons of the harvested fruits in Sri Lanka are wasted annually due to the improper post-harvesting activities such as improper packing, handling, transportation and storage throughout the marketing chain starting from the farmer to collector, whole seller, retailer and finally to the consumer (Kodippili, 2016). About 30% of banana, 46% of papaya, 18% of pineapple, 40 of lime, 41% of avocado and 90% of jackfruit are wasted from the annual production in Sri Lanka (Sarananda, 2000;Perera et al., 2009). Thus, it is a timely requirement to study the possibility of producing anhydrous bioethanol in Sri Lanka using these fruit wastes, as a gasoline substitute. It would minimize the dependence on imported fossil fuel, environmental pollution, solid waste management in the country, etc. (Borah & Mishra, 2011).
Microorganisms are used for the fermentation of substrates in the bioethanol production. Microorganisms employ several fermentation characteristics such as rapid and relevant carbohydrate fermentation ability, appropriate flocculation and sedimentation characteristics, genetic stability, ethanol tolerance, elevated concentrations of ethanol production, high cell viability for repeated recycling, and temperature tolerance (Doran, 2012). Out of many available microorganisms, yeast and bacteria are the most proven microorganisms for the production of ethanol through fermentation. Yeast species, Saccharomyces cerevisiae (commonly known as Baker's yeast) in Saccharomyces spp variety, are the more commonly used microorganism (Casey & Ingledew, 1986). PM, a novel microorganism (Gen Bank number of KU056820.1) has been identified recently as an efficient microorganism to produce bioethanol with fermentation. (Hewawasam et al., 2019). Presently, many varieties of genetically modified strains are used worldwide to produce bioethanol through fermentation. During the process of bacterial fermentation, the growth of bacterial cells takes place. This involves various biochemical networks and chemical reactions. Bioethanol yield depends on the proper manipulation of fermentation factors such as agitation to improve nutrient transfer to the cell surface, aeration to improve cell growth, temperature, pH, etc. (Lee, 2008). As such, it is essential to have prior knowledge of the kinetic behaviour of the fermentation process along with kinetic models before the large-scale bioethanol production (Farah et al., 2011;Ogbebor et al., 2014;Nanthaporn, & Niron, 2013).
Accordingly, this research aims at optimizing the kinetic modelling of bioethanol production by batch fermentation of selected three types of overripe fruits.
Fermentation is carried out at three different substrate concentrations, temperatures, and pH values for the optimization of bioethanol production. Optimization and kinetic modelling are carried to predict kinetic parameters and thus to select the best fermentation factors. This has been done by considering the highest correlation between the theoretical and actual curves by using the Monod model for glucose utilization and modified Gompertz equation for bioethanol production (Tussanee et al., 2015).

Selection of fermentation parameters
Banana, papaya and jackfruits were used as substrates. Baker's yeast (Saccharomyces cerevisiae) and PM were used as inocula. Three different fermentation temperatures: 27, 32, and 35 oC, three different pH values: 4.3, 5.0, and 5.7 and three different concentrations of the fruit mass as 1:1, 1:1.5 and 1:2 (w:w) fruit to distil water ratio was selected. The settings of each variable were selected reference to the literature. (Hammond et al., 1996;Akin-Osaniye et al., 2008;Nirmal et al., 2012).
Separate fermentation experiments were carried out with all the possible combinations of fermentation parameters shown in Table 1.

Media for growing microorganisms
Yeast Extract-Peptone-Dextrose (YPD) media and glucose media were used to grow Saccharomyces cerevisiae and PM microorganism, respectively (Hewawasam et al., 2019).

Experimental procedure
Separate fruit juice samples with different solid contents were prepared by blending fruits and distilled water according to the required proportion in a domestic blender (Panasonic mixer grinder, model number -MX-AC300). 100.00 mL of each juice was then added in to sterile 250 mL conical flasks. The pH of each sample was set according to the sample number using a pH meter (HACH, Model No. 2786). The openings of the conical flasks were covered with cotton wool and aluminium foil for autoclaving. All the samples were then autoclaved for 30 min at 121 o C and allowed to cool up to room temperature. Experiments were carried out in duplicate.

Fermentation
The autoclaved fruit samples were used for fermentation.

Measurement of sugar and ethanol concentrations
The concentration of sugar in each filtrate was measured at 24-h time intervals using a calibrated refractometer (ATAGO, Model No.-1T) with a scale ranging from 0 to 30 % Brix unit. Ethanol concentrations of the filtrates were measured using Gas Chromatography (GC) (SHIMADZU, Model No. C 114850). The measured ethanol concentrations using the GC were verified randomly by the standard dichromate reagent method (William & Reese, 1950).

Statistical analysis
Kinetic model prediction, simulation, estimation of kinetic parameters, and correlations were carried out in MATLAB version R 2015a software.

Prediction of kinetic parameters
Prediction of the kinetic parameters was carried out by fitting the experimental data into empirical equations. Fermentation kinetic parameters such as maximum specific growth rate (µmax), half-saturation concentration (K s ), maximum ethanol concentration (P m ), maximum ethanol production rate (r pm ), and lag phase (t L ) were predicted reference to the theoretical graph relevant to the highest correlation between the experimental and theoretical graphs.
Monod equation (equation 1) was used to predict µmax and K s (Tussanee et al., 2015). (1) Where µ is the specific growth rate per h, and S e is the substrate concentration in the effluent in g/L.
Modified Gompertz model (equation 2) was carried used to predict the kinetic parameters; Pm., rpm , and t L (Tussanee et al., 2015). ( Where, P is the ethanol concentration in g/L, and t is the fermentation time in h.

Data
Sugar concentration and ethanol concentration of the filtrates at 24-h intervals were measured.

RESULTS AND DISCUSSION
The highest R2 values of each fruit type and microorganism are observed and tabulated in Table 2 and Table 3 respectively.

Analytical results
The calculated results on sugar utilization, ethanol yield, ethanol productivity, and fermentation efficiency for the three types of fruits using equations 3, 4, 5, and 6 are tabulated in Table 4. The data in Table 4 verifies that a higher percentage of sugar of all three types of fruits, ranging from 76 to 88%, has been utilized for bioethanol production. Minimum percentage of sugar utilization was observed from papaya fruit when ferment with PM microorganism. The maximum percentage of sugar utilization of 88% was observed with banana fruit when ferment with PM. The relatively higher percentage of sugar utilization by each fruit type is due to the higher ethanol tolerance of PM microorganism. Experimental and theoretical bioethanol yield for all types of fruits is relatively close, and actual ethanol yield is always lower than the theoretical bioethanol yield and the experimental results on ethanol concentrations are in an acceptable range comparative to the reported bioethanol concentrations by Akin-Osaniye et al., 2008, Suhas et al., 2013and Brent et al., 1996. The actual and theoretical bioethanol yield of 13.0 g/L and 15.9 g/L, respectively, were observed using banana (embul variety) with PM which reveals that PM is a new collection to the bioethanol industry with a higher ethanol yield. From the confirmation run, carried out by fermentation banana fruit with PM at 32 oC temperature and 5 pH by maintaining fruit to water concentration of 1:1, a maximum ethanol yield of around 13.0 g/L was obtained. This amount of bioethanol yield is considerably high using a natural microorganism. Reference to the same table, ethanol yield of jackfruit with yeast is about 7.4 g/L, which is increased up to 7.5 g/L with PM. A study by Akin-Osaniye et al., in 2008, reported that the ethanol yield of jackfruit is 5.4 g/L with same fermentation conditions with Saccharomyces cerevisiae, and it has been increased by 2 g/L with PM (Akin-Osaniye et al., 2008). Considering the fermentation time, for all types of fruits, fermentation with PM has resulted in a long fermentation time of 120 hours with increasing bioethanol yield. This suggests that PM is a highly ethanol tolerant microorganism. The highest ethanol productivity of 0.107 g/L.h was observed with banana fruit with yeast and the highest bioethanol yield in grams ethanol over grams glucose was observes as 0.168 for banana fruit with PM microorganism (Table 4). These results appear to consistent with the results published in the literature by Tussanee et al., in 2015. The observed fermentation efficiencies in Table  4 are in a considerably higher range of 81.8% to 98.1%, which proves the maximum bioethanol productivity during the fermentation processes.

Selection of optimum fermentation parameters
The optimum fermentation parameters were obtained by considering the highest correlation between the experimental graph and theoretical graphs for each type of fruit with two different microorganisms. Monod equation (Equation 1) was used to obtain the theoretical graph of the variation of sugar concentration of the substrate with time for each fruit variety with each microorganism for estimating optimum fermentation process parameters and relevant fermentation kinetic parameters. Modified Gompertz equation (Equation 2) was used to obtain theoretical graphs of the variation of ethanol concentration in the filtrate with time for each fruit variety with each microorganism for estimating the ethanol production kinetic parameters. Corresponding graphs for the highest correlation with the experimental data are shown in Figures 1 and 2 respectively.
The highest correlation coefficients for the fermentation and ethanol production are given in Table  5. The best fermentation process parameters such as temperature, pH, and concentration of the fruit pulp: fruit to water ratio (w:w) for each fruit type with two different microorganisms are given in the same table.
Fermentation and ethanol production kinetic parameters, obtained from Monod model and Modified Gompertz equation are given in Table 6. Reference to Tables 4 and 5, the highest ethanol yield of 15.9 g/L has been predicted for the fermentation of banana fruit with PM at 35 o C temperature, pH of 5, and fruit to water concentration of 1:1. The lowest fermentation temperature and pH of 32 oC and 4.3, respectively, have been predicted for papaya fruit with fruit to water concentration of 1:1.5, resulting in maximum ethanol yield of around 5.0 g/L.

Concentration
Correlation (   The obtained best pH value of 4.5 for papayais in compatible with the data reported in the literature by Akin-Osaniye et al., in 2008. The best fermentation parameters for jackfruit using Yeast are 32 o C temperature, pH 5.0, and fruit to water concentration of 1:1.5. On the other hand, with PM microorganisms, the best fermentation parameters/ conditions of 32 oC temperature, pH of 4.3 and fruit to water concentration of 1:2 have been observed. The fermentation process parameters using yeast, tabulated in Table 5, are in compatible with the data reported by Akin-Osaniye et al., 2008. It was noticed that the bioethanol yield of all the fruits has been increased with PM microorganism.
Considering to the correlations in Table 5, all the correlations are in a higher range, above 98.50, which indicates that the bioethanol production process is well explained by the Monod equation and modified Gompertz equations. Besides, the received graphs in this study are quite similar to the graphs obtained by a previous study reported Tussanee et al., in 2015.

Prediction of kinetic parameters
The kinetic parameters in Table 6 were obtained by performing kinetic analysis with Monod and modified Gompertz equations in MATLAB by considering the graphs obtained with the highest correlations. Values for the maximum specific growth rate (µmax) and saturation concentration (K s ) were obtained from Monad equation while maximum ethanol concentration (Pm), maximum ethanol production rate (r pm ) and lag time (from the beginning of fermentation to exponential ethanol production) (t L ) were obtained using modified Gompertz equation. According to Table 5, the values for µmax were found in the range of 0.039 to 0.309. These values seem to be similar to the values, reported in the literature (Ariyanti & Hadiyanto, 2013;Farias et al., 2014). Usually halfsaturation concentration, K s is a low value but the results of K s, obtained in this study are relatively high due to higher substrate concentration, which is commonly found in ethanol production (Nadya et al., 2012). The calculated values for P m for fermentation of the three fruits with yeast are in agreement with the values reported in previous studies on ethanol production (Nadya et al., 2012). Values for r pm and t L among the three different types of fruits studied were found in the range of 0.06-0.19 g/L.h and 9.21-20.24 h respectively and these values, validate the results of previous studies by Tussanee et al. in 2015. These kinetic parameters could be useful to design the bioreactor for commercial-scale ethanol production with these overripe fruit masses.

CONCLUSIONS
Fermentation of overripe banana, papaya and jackfruit was successfully carried out in batch fermentation and the fermentation results were used for the optimization and kinetic parameter estimation of the fermentation process. Banana was identified as the most potential freely available fruit in Sri Lanka for the bioethanol production using a novel microorganism, PM. The highest bioethanol yield of about 13.0 g/L was obtained by fermentation banana fruit mass with a concentration of fruit mass to water ratio of 1:1 using PM microorganism by fermentation at 35 oC temperature and pH of 5. The fermentation process was well explained by the Monod equation and modified Gompertz equations with high correlations of 99.61 and 99.13 respectively.
Considering the experimental bioethanol yield and the available banana waste data of Sri Lanka in 2017, approximately 2x106 litters of anhydrous bioethanol could be produced annually by utilising whole available banana waste in Sri Lanka. Further, the residual materials of the bioethanol production processes could be used as an ecofriendly, non-toxic and biodegradable fertilizer.

CONFLICTS OF INTEREST
The authors declare that they have no conflict of interest.