Utilization of electric arc furnace dust as a solid catalyst in biodiesel production

World’s energy sources like petrochemical oils, natural gas and coal cause global warming and environmental pollution. Therefore, the traditional energy sources must be replaced by the renewable energy resources. Biodiesel has been recognized as one of the effective, green, renewable and sustainable fuels. This paper investigates the production of biodiesel from sunflower oil by using electric arc furnace dust (EAFD) as a heterogeneous solid catalyst. Four reaction variables, i.e., the reaction time, methanol-to-oil (M/O) molar ratio, reaction temperature and EAFD loading, were chosen to determine their effect on biodiesel production. The effect of the all reaction variables on the biodiesel yield was evaluated using response surface methodology. A relation has been developed representing the biodiesel conversion as a function of all the independent variables. Reaction conditions optimization have been studied for the biodiesel yield maximization and the reaction conditions minimization. The optimum biodiesel yield equals 96% at reaction temperature of 57 °C, methanol-to-oil molar ratio of 20:1, reaction time of 1 h and EAFD loading of 5%.


Introduction
Energy is the backbone of the world's life as there is no life on the earth without energy. Energy can define as the ability to do work. Energy can be nonrenewable and renewable energy. Energy that comes from sources that will end or need thousands or millions of years to form it is called nonrenewable energy such as oil, coal and fossil fuels. Energy that can be easily produced and comes from resources like the wind and the sun is called renewable energy (Harrell 2006). Solar energy, wind energy, hydropower, geothermal and biofuels are examples for the renewable energy sources (Thangaraj 2019). Biofuels are fuels that derived from biological carbon fixation as well as from biomass. There are many different types of biofuels such as bioethanol, biodiesel, bio-methanol, and biogas. Biodiesel and bioethanol are considered the most popular biofuels (Grassi 1999).
Biodiesel was defined as mono-alkyl ester of long-chain fatty acids obtained from renewable sources such as animal fats or vegetable oils. Biodiesel is used as an alternative fuel for diesel engines. In simple terms, biodiesel is biofuel produced by chemical processes from animal fats or vegetable oils or recycled cooked oil and an alcohol that can be utilized alone or mixed with diesel oil in diesel engines. The prefix bio represents biological and renewable source, while diesel represents one utilized on diesel engines. The most common way to form biodiesel is through a process named "trans-esterification" that involves using methanol to change the chemical properties of the oil. It is a simple process that produces high conversions with little percentage of glycerol as a byproduct (Hill et al. 2006).
The usage of catalyst in the trans-esterification reaction is very useful and important for higher yield of product and better rate of reaction. Catalysts utilized for transesterification of biodiesel feedstock may be heterogeneous or homogenous according to the solubility of the catalyst in the reaction (Chen et al. 2011). Many researches have been directed to the usage of heterogeneous catalysts for biodiesel-forming reactions to overcome the problems resulted when using homogenous catalysts. Although the homogenous acid catalyst is effectiveness, it can cause contamination issues which need product purification processes and good separation. This will increase the production process cost. Heterogeneous acid catalysts have many advantages such as regenerating and reusing catalyst (Talebian-Kiakalaieh 2013).
Electric arc furnace is considered the core equipment in steel manufacturing. The main function of such equipment is to carry out the melting processes for the given feed. This device produces a dangerous waste besides its main product. This waste is known as the electric arc furnace dust (EAFD). This material is known to be hazardous due to the high content of heavy metals mainly zinc and led. This material is known to have a degree of class 1 hazardous effect. The estimated amounts of waste are very large because it could reach from 15 to 25 kg of waste per a single tone of steel produced which could cause severe harmful for human and the environment (Rizescu and Stoian 2011). Kesic et al. (2012) used a mechano-chemically method to synthesized CaO·ZnO catalyst for biodiesel production. This catalyst gave biodiesel with yield approximately 99 wt% using methanol and sunflower oil (10:1 molar ratio) after 3 h reaction at 60 °C. Rengasamy et al. (2014) studied the biodiesel production from castor oil using synthesized iron nanoparticles. The results of this study revealed that the produced biodiesel using synthesized iron nano-catalyst was considered as prospective alternate fuel to the conventional diesel fuel. Lukic et al. (2014) studied the reaction kinetics of biodiesel production using ZnO supported on alumina/ silica as a solid catalyst. The results proved the kinetics of the methanolysis reaction could be expressed by the firstorder reversible reaction model. Al-Sakkari et al. (2016) studied the possibility of biodiesel production using cement kiln dust (CKD), as heterogeneous catalyst. The resulted optimum reaction conditions for biodiesel production were found to be 6 h reaction time, 2% catalyst loading and 15:1 methanol-to-oil molar ratio at 65 °C reaction temperature and 800 rpm. Ahmed (2017) used zinc oxide nanoparticles as catalyst for biodiesel production by trans-esterification of waste cooking oil with yield 83%. Ali et al. (2017) prepared a nano-catalyst of CaO supported by Fe 3 O 4 magnetic particles by a chemical precipitation method. The results revealed that the highest biodiesel yields for palm seed oil of 69.7% can be obtained under the conditions of (65 °C reaction temperature, 300 min reaction time, 20 methanol/ oil molar ratio and 10 wt% of CaO/Fe 3 O 4 catalyst loading).
The aim of this paper is to investigate the possibility of biodiesel production from sunflower oil using EAFD as heterogeneous catalyst and determining the optimum reaction conditions using response surface methodology (RSM).

Raw materials
The raw materials used in this work were as follows: (a) Electric arc furnace dust (EAFD) supplied from Egyptian steel factory at Ain El-Sokhna which was used to catalyze the reaction between sunflower oil and methanol. (b) Sunflower oil that was purchased from Egyptian local market used as liquid source in this work. (c) Methanol (MeOH) 99% that was supplied by Morgan Chemical Ltd., Egypt.

Assessment of EAFD
X-ray fluorescence is used for elemental analysis of EAFD to determine the concentration of various elements. X-ray diffraction (XRD) shows the phases present in the material. Standard sieving procedure is used to determine the particle size distribution as described by ASTM D422 (2007). The sieves used are in compliance with ASTM E11 (2009).

Biodiesel production
The trans-esterification reaction took place in a round glass batch reactor fitted with reflux condenser. Sunflower oil was heated using magnetic heat stirrer then mixed with EAFD and methanol once the oil reached the targeted temperature. A reflux condenser was fitted at one of the reactor necks to condense the vaporized methanol during the reaction. Temperature was determined using a thermocouple as shown in Fig. 1. The mixture was filtered at the end the required reaction time using filter paper to remove the waste powder then transferred to a separating funnel after cooling to separate glycerol from biodiesel after 2 h settling for sufficient separation of the products mixture. The biodiesel was heated for 30 min at 80 °C for removal of unreacted methanol. The resulted biodiesel was weighted for conversion calculations by the following relation.

Experimental design
In this study, experimental design using response surface methodology (RSM) has been used to develop and construct a full analysis of the effect of reaction variables on reaction responses which is the biodiesel yield. Four independent variables have been selected which are reaction time, methanol-to-oil molar ratio (M/O), EAFD catalyst loading concentration and reaction temperature as shown in Table 1.
According to the experimental work done at the maximum limit of the range, the resulted biodiesel conversion was found to be equal to 99.7%. Increasing the values for the reaction conditions more than that will not make large change to the value of conversion. According to the experimental work done at the minimum limit of the range, the resulted biodiesel conversion was found to be equal to 70.2%. Decreasing the values less than that will decrease will result in unacceptable biodiesel conversion. The reaction conditions and ranges were also chosen based on the review papers done by Refaat (2011), El-Sheltawy and Al-Sakkari (2016), Sulaiman (2016), and Ling et al. (2019). Stirring rate was equal to 750 rpm in all experiments. Thirty experimental runs were done as an approach to minimize the number of experiments using central composite design technique (CCD) as shown in Table 2. The experimental runs from 25 to 30 have the same conditions called the center point of design. The proposed optimization targets have been selected based on economic and environmental considerations: minimizing the cost of biodiesel production by minimizing the reaction conditions specially reaction time and temperature which means minimizing energy cost while maintaining (1) Biodiesel Conversion% = Weight of biodeisel produced Weight of sunflower oil × 100.  maximum production of biodiesel. Design Expert 12 software was used to investigate the experimental runs, its order, regression analysis, graphical analysis and numerical optimization. Eco-friendly biodiesel production was done using solid waste as a heterogeneous catalyst.

Experimental work done on optimum biodiesel sample
Physicochemical properties were determined for the optimum biodiesel sample and compared to Biodiesel International Standards ASTM D6751 (2020) and the European Biodiesel Standard, EN 14214 (2019). The final optimum biodiesel sample derived from sunflower oil has been analyzed through gas chromatography. Analysis of total FAME and methyl linoleate in the produced biodiesel has been conducted according to the standard EN 14103 (2020) method. In addition, analysis of total and free glycerol and triglycerides content in B100 biodiesel was performed according to the standard EN 14105 (2011) method.

Reusability of EAFD
The optimum biodiesel sample which resulted from using M/O molar ratio of 20:1, catalyst loading of 5%, reaction temperature of 57 °C, reaction time of 1 h and stirring rate of 750 rpm which gave 96% for biodiesel conversion undergone filtration for solid particles removal; then, these solid particles were washed with methanol for glycerol removal and then finally dried to remove any traces of methanol on the particles. This EAFD was reused to catalyze trans-esterification reaction of sunflower oil under the same optimum reaction conditions and without adding any new solid catalyst to reaction medium. After conducting trans-esterification reaction, the conversion was calculated for each experiment to test the catalyst efficiency and strength. Table 3 shows the results of X-ray fluorescence (XRF) analysis for EAFD. The solid waste consists mainly of Fe 2 O 3 , ZnO, CaO and SiO 2 oxides with other oxides. These oxides are good biodiesel catalyst according to previous research work, so this is a good indication that this solid waste can be used as biodiesel catalyst.

Mineralogical analysis of EAFD
The mineralogical analysis of the waste powder in Fig. 2 shows that it mainly consists of zinc oxide, zinc iron oxide, sodium manganese silicate hydroxide and potassium manganese sulfide phases.   Figure 3 shows the cumulative screen analysis curve of EAFD waste. The vertical axis represents the fraction retained on each screen with certain diameter. This figure shows that the grind waste is very fine. The mean particle size of waste powder equals 4.5 μm.

Analysis on produced biodiesel
The biodiesel was generated, and its conversion was calculated and   Expert program suggests the quadratic model to be the best model using ANOVA analysis technique. Some terms in the quadratic model were neglected as they are insignificant as their P-values are greater than 0.1, so the model become a reduced quadratic model. Figure 4 shows the results of ANOVA analysis that was done by the design expert program. The following equation represents the relationship between the biodiesel conversion and the reaction conditions as a reduced quadratic model.
where Y is the biodiesel conversion or the reaction response, A is the reaction time in hrs, B is methanol-to-oil ratio, C is Catalyst loading as wt%, and D is the reaction temperature in o C. Coefficient values, R 2 adj and R 2 were determined to measure the validity of the fitting model which was found to be 0.9369 and 0.9521 and, respectively, which insure the high significance of the predicted model.
The predicted values were plotted versus experimental results of the biodiesel conversion. This plot shows reasonable agreement and high correlation as shown in Fig. 5. Figure 6 shows the effect of methanol/oil molar ratio, reaction time, reaction temperature and catalyst loading on the biodiesel conversion. The reaction temperature and M/O ratio have the highest effect of the biodiesel conversion, while the catalyst loading and reaction time have a very small effect on the biodiesel conversion

Reaction variables interactions effect on biodiesel conversion
The following surface plots represent the effect of reaction variables interactions on biodiesel conversion. Figure 7 shows the contour plot that represents the relation between M/O ratio and temperature interactions (BD) and biodiesel conversion, while Fig. 8 shows its surface or 3D plot. Figure 7 and Fig. 8 show that the combination of M/O ratio and temperature has a big effect on the biodiesel conversion. Figure 9 shows the surface plot that represents the relation between effects of methanol/oil and reaction time on biodiesel conversion and indicates that the combination between methanol/oil and reaction time has a little effect on the biodiesel conversion compared with the combination of M/O ratio and temperature.

Optimization of reaction variables
The Design Expert software developed optimum 100 solutions within the required targets and the optimum solution was the solution with the highest desirability (74%). The optimum conditions using numerical optimization have been concluded at M/O molar ratio of 20:1, catalyst loading of 5%, reaction temperature of 57 °C, reaction time of 1 h and stirring rate of 750 rpm resulting in 96% for biodiesel conversion. Figure 10 shows the summary for  Table 5 shows the physicochemical properties and its standard limits. All of the properties agree with both EN 14214 (2019) and ASTM D6751 (2020).

Analysis on optimum produced biodiesel sample
The final optimum biodiesel sample derived from sunflower oil has been analyzed through. Table 6 illustrates the results concluded from gas chromatography (GC) analyses for the optimum produced sample. It is clearly shown that the produced biodiesel has agreed with the specified specification for both European standards EN 14103 (2020) and EN 14105 (2011). Both free glycerol and total glycerol in the produced biodiesel samples did not exceed the specification limits of EN 14105 (2011) by recording 0.01 and 0.019% (m/m), respectively. Similarly, monoglycerides, diglycerides and triglycerides concentration in the final pure biodiesel recorded acceptable results in comparison with EN 14105 (2011) by recording 0.0043, 0.0078, 0.0856% (w/w), respectively. Finally, total FAME concentration has recorded using EN 1403 (2020). It has an acceptable result with concentration of 97.6%. (m/m).  Figure 11 shows that the conversion dropped from 96% after the first use and reached 85% at after the second run then decreased to about 70% by the end of third use and finally decrease to 50% at the end of fourth use. There are many reasons for decreasing the activity of the catalyst such as the contamination of glycerol on the active center of the catalyst. The second reason is leaching of catalyst which means active species loss from the solid because of its transformation into the liquid medium. The results showed that EAFD can be reused only two or three times as a maximum; after that, a new catalyst will be used.

Conclusion
An economic biodiesel production from sunflower oil has been examined using electric arc furnace dust (EAFD) solid waste as a heterogeneous catalyst. The XRF analysis shows that EAFD solid waste consists mainly of Fe 2 O 3 , ZnO, CaO and SiO 2 oxides, so this is a good indication that this solid waste can be used as biodiesel catalyst. Four independent reactions conditions, i.e., the reaction temperature, methanol-to-oil (M/O) molar ratio, reaction time and catalyst loading have been chosen to detect their effect on biodiesel production. Thirty experimental runs were done as an approach to minimize the number  the biodiesel production cost and maximization of its produced amount. The optimum conditions using numerical optimization have been concluded at M/O molar ratio of 20:1, catalyst loading of 5%, reaction temperature of 57 °C, reaction time of 1 h and stirring rate of 750 rpm resulting in 96% for biodiesel conversion. Gas chromatography analysis showed that the produced biodiesel has agreed with the specified specification for both European standards EN 14103 (2020) and EN 14105 (2020). The reusability proved that EAFD can be reused only two or three times as a maximum; after that, a new catalyst will be used.