Crucibles at a constant weight with a 5 g sample of castor and jatropha seeds were placed in a Lindberg-Blue M oven at 105 °C to determine the moisture content of castor and jatropha seeds. The samples were dried for 4 hours (Cornejo, 2012). The moisture content (% MC) was determined by equation 1.

Where: M _i is the initial mass of the wet sample (g), and M _f is the final mass of the dry sample (g).

A KOMET CA59G mechanical press was used to extract oil from castor and jatropha seeds. The procedure´s conditions of the mechanic press allowed for adjustments in temperature, speed, and nozzle size. In this specific equipment, the extraction pressure or extraction force could not be modified. An experimental design was developed for each seed sample since the proper oil extraction conditions were unknown. A factorial design was selected; variables were chosen by considering the oil´s fluidity gained without generating obstruction in the nozzle cake outlet. The following parameters were chosen for castor seed: temperature of 70 °C and 90 °C, speed 1 and 2 rpm, nozzle of 6 mm and 8 mm. Each treatment was performed in duplicate. The parameters used in the factorial design for jatropha seeds were: temperature of 95 °C and 100 °C, speed of 1 and 1.5 rpm, nozzle of 8 mm and 10 mm. Castor oil has certain advantages over jatropha oil since it requires a lower temperature for extraction by mechanical pressing, which implies a lower energy demand. Due to the limited availability of jatropha seed, a test was carried out for each treatment.

The collected oil was left to settle for 24 hours, the necessary time for the seed cake to settle and separate from the oil. Subsequently, it was vacuum filtered for the removal of impurities. A Kitasato flask, Buchner funnel, vacuum pump, and Whatman #2 filter paper (8 µm) were used. The Jatropha oil filtration time was 30 min. Castor oil was centrifuged twice due to its high viscosity: 1) 4,700 rpm for 30 min and 2) 1,500 rpm for 15 min. The oil yield was calculated by equation 2 (Hernández and Mieres, 2002).

Where M _o is the extracted oil mass (g) and M _s is the processed seed mass (g).

For castor oil, due to the presence of hydratable phospholipids, a degumming process was required. In this process the oil was heated from 50 to 70 °C using a VWR hotplate/stirrer at 250 rpm. Distilled water was added at 3% mass based on oil weight, and the mixture was left under constant stirring for 30 min, while monitoring its temperature (Hernández and Mieres, 2002).

It was necessary to ensure the elimination of impurities to carry out an adequate characterization of the oils. Therefore, a vacuum filtration was done for both oils. The percentage of impurities removed from castor oil by vacuum filtration was lower than that of jatropha oil because its high viscosity made it difficult for the oil to flow in the filtration process. For that reason, it was necessary to subject the castor oil to a centrifugation process where 10.6% of impurities were removed from the oil in the first run and 2.23% in the second one.

Methanol (CH₃OH) was used for both oils, considering an oil/alcohol molar ratio of 1:3, with 31% excess CH₃OH in volume, 1% sodium hydroxide (NaOH) as the catalyst and reaction temperature of 60 °C. The reaction time for castor oil was 30 min (Ferdous et al., 2013), while for jatropha oil it was 120 min (Okullo et al., 2012). 150 mL of previously degummed castor oil and 100 mL of jatropha oil were used for the reaction.

The obtained biodiesel was washed with a water-jet washing method to avoid emulsions. Distilled water was used and heated to 50 °C, the biodiesel: water ratio was 1:3. The number of washes was 5, determined by visual examination of the residual water. The biodiesel was dried at 110 °C for 10 min with constant stirring and stored in a dry environment.

The acid value was determined according to ASTM D974, using potassium hydroxide (KOH), as shown in equation 3:

Where 56.1, N, and V are the molar mass (g/mol), normality, and the volume (mL) of KOH, respectively, and m is the mass of the sample (g).

Due to the high acid value present in jatropha oil, it was necessary to carry out an esterification process for 60 min, at 60 °C, and with constant stirring in a VWR hotplate/stirrer. For every 100 g of jatropha oil, 60 g of methanol (CH₃OH) were added. In addition, 0.27 mL of sulfuric acid (H₂SO₄) at 0.5% w/w were added as catalyst per 100 g of oil.

After the reaction, the mixture was transferred to a separatory funnel and allowed to settle for two hours. The resulting oil was dried with vigorous stirring at 110 °C for 10 min. The CH₃OH from the reaction was recovered by a rotary evaporator DLAB RE100-PRO at 55 °C and 160 rpm.

The cloud point was determined according to ASTM D2500. The pour point was determined according to ASTM D97. Both determinations were made in triplicate.

The dynamic viscosity was determined using a BROOKFIELD CAP 2000+ viscometer. The measurements were made in triplicate, with a temperature range of 50 to 100 °C. A #6 cone was used for castor oil, a #1 cone for jatropha oil, and a #2 cone for biodiesel.

Tests were carried out using an Agilent 7890A GC chromatograph attached to a 5975C mass detector, equipped with an HP-5MS capillary column (30 m x 0.25 mm x 0.25 µm). An automatic sampler was used to inject 1 μL of solution. The ionization energy was 70 eV with a mass range of 30 to 800 m/z. The initial temperature on the column was 125 °C for up to 0.5 min, the ramp set at 25 °C/min to 150 °C for up to 2 min, then to 200 °C with a speed of 50 °C/min. The injector temperature was set at 255 °C and the detector at 270 °C. The flow rate of the carrier gas (helium) was 1.0 mL/min injected with a 1:50 gas dilution. The identification of individual components was based on comparison with the NIST98 mass spectral library. All determinations were carried out in triplicate.

The HHV for castor and jatropha biodiesel was estimated using the mass fraction and the molar fraction of the fatty acid methyl esters (FAME) present in the biodiesel. Equations 4 and 5 were used to calculate the HHV.

Where w _i , x _i , and HHV _i are the mass fraction, the molar fraction, and the higher heating value of a given FAME, respectively.

The mass fraction considered was obtained by gas chromatography, while the molar fraction was determined by direct conversion from its mass fraction. To estimate the HHV _i equation 6 was applied (Ramírez et al., 2012):

Where: M _i and N are the molecular weight and the number of double bonds in a given FAME, respectively. For this equation, an average absolute deviation of 1.92% is reported between the experimental and calculated HHV’s when using the mass and molar fractions. This equation was used to estimate the HHV of beef tallow oil, soybean oil, sunflower oil, corn oil, and cottonseed oil (Ramírez et al., 2012). The estimated HHV was compared to the methodology proposed by Fassinou (2012).

The reported moisture content of castor seeds varies from 1.84-5.4% (Omari et al., 2015) compared to that obtained of 3.74%. For the jatropha seeds, the reported moisture content ranges between 4.75-19.57% (Garnayak et al., 2008) compared to that obtained of 4.7%. Castor and jatropha seeds have a moisture content within the reported range, making them less susceptible to deterioration by microorganisms to be stored without affecting their viability.

Table 1 shows the ANOVA corresponding to the factorial design of castor and jatropha seeds, respectively.

It was concluded that the significant factors were B and C for castor oil and A and C for jatropha oil.

The Minitab^® version 18 software was used for the analysis of the results, using residuals to verify the assumptions of normality, constant variance, and independence of the model, which are characteristics that corroborate the validity of the model, indicating that the response variable is normally distributed, with the same variance in each treatment and the measurements were independent. Equations 7 and 8 are the equations obtained from the experimental designs for castor (Y _C ) and jatropha (Y _J ) mechanical extraction, respectively.

Where S, N, and T are the speed (rpm), nozzle size (mm), and temperature (°C), respectively.

Based on the statistical analysis, the optimal conditions for castor oil extraction were determined: 90 °C, 2 rpm, and a nozzle of 6 mm, achieving a yield of 36.97%. For the jatropha oil, the optimal conditions were: 100 °C, 1.5 rpm, and a nozzle of 10 mm, achieving a yield of 20.11%. The castor oil yield obtained was below the range of 38-48% by warm pressing (> 70 °C) reported by Scholz and Nogueira da Silva (2008), while jatropha oil was similar to the values achieved by other researchers where similar extraction equipment was used (Yate et al., 2020).

Figure 1 illustrates the influence of significant factors on the performance of castor and jatropha oil extraction processes. In the case of castor seed, the maximum oil extraction was 110.9 g from a 300 g sample. For jatropha seed, the higher oil extraction was 100.55 g from a 500 g sample.

The conversion percentages of castor and jatropha oils to biodiesel were 96.66 and 98.88% respectively, which is a great advantage when used as raw material. The conversion percentage will depend mainly on the type and quality of the seeds, the pre-treatment, and the transesterification conditions applied. Other authors report conversion percentages for castor oil of 76% and jatropha oil of 91% (Okullo et al., 2012). Likewise, castor oil requires a lower cost pre-treatment for its conversion into biodiesel.

Table 2 displays the results of the analyzed properties for castor and jatropha oils, as well as the biodiesel obtained from them.

For biodiesel production, an acid value of less than 2 mg KOH/g is considered an essential condition for carrying out the transesterification reaction through alkaline catalysis, to avoid saponification (Mashad et al., 2008). The acid values for castor and jatropha oils were within the ranges reported (Piloto et al., 2011; Pradhan et al., 2012; Omari et al., 2015; Ahmad et al., 2016); however, for jatropha oil, the acid value was higher than the value required to perform the transesterification reaction. Therefore, the percentage of free fatty acids was reduced by an esterification process using H₂SO₄ before transesterification (Mashad et al., 2008). The high acid value for jatropha oil is attributed to the presence of oleic (C18:1) and linoleic (C18:2) acids since they are the two components with the highest proportion and present one and two unsaturations in their structure chemistry, respectively. Also, the high acid value for jatropha oil can be due to different factors such as the type of seed, the origin, and even the storage time (De Oliveira et al., 2009). After esterification, the free fatty acid content decreased by 98.13% at 60 °C, 60% w/w methanol, and 0.5% w/w of sulfuric acid, achieving an acid value of 0.45 mg KOH/g. The maximum acid value for biodiesel allowed by the ASTM D6751 standard is 0.50 mg KOH/g, so biodiesel from castor and jatropha oils is within the values reported in the literature (Lu et al., 2009; Pradhan et al., 2012; Banerjee et al., 2017).

The cloud points measured for castor and jatropha oils are similar to those reported in the literature (Okullo et al., 2012; Ahmad et al., 2016). Castor biodiesel has a lower cloud point than jatropha biodiesel, making castor biodiesel more suitable for use in cold locations. Jatropha biodiesel may require the use of additives to improve its cloud point.

The pour points for castor and jatropha oil were reached at the same temperature, which was similar to those reported for jatropha oil (De Oliveira et al., 2009; Koh et al., 2011); while for castor oil, it ranges from -13 to -15 °C (Tunio et al., 2016; Kumar et al., 2020). However, it does not represent complications for its use in tropical areas and favors its ability to work at low temperatures. The measured pour points for jatropha oil biodiesel were within the reported ranges (Koh et al., 2011). However, the pour point was not achieved for castor biodiesel due to equipment limitations, where the temperature reached was -20 °C, exceeding the measurement capacity. Temperatures of -12 to -30 °C are reported for the pour point of castor biodiesel (Tunio et al., 2016; Chidambaranathan et al., 2020). The difference between the values obtained and those reported in the literature may be due to factors such as the geographical area to which the seeds belong, climatic conditions, and the state of the crop.

The viscosity of the oil used as raw material directly affects the viscosity of the biodiesel produced. Figure 2 exhibits the results of the tests to measure the viscosity of castor and jatropha oils and biodiesel.

In Figure 2 it can be seen that the viscosity of castor oil is much higher than that of jatropha oil. This is due to the hydroxyls present in the triglyceride molecule of castor oil, which give it this high viscosity characteristic. Viscosity increases with increasing chain length and decreases with the number of double bonds (level of unsaturation in the chain) (Silitonga et al., 2013). The presence of linoleic acid (C18:2) contributes to the decrease in the viscosity of jatropha oil since it represents the second compound with the highest proportion and presents two unsaturations in its chemical structure. The majority presence of ricinoleic acid (C18:1) in castor oil gives it a higher viscosity because this acid has one unsaturation in its chemical structure.

The viscosity of esters obtained from castor and jatropha oils through the transesterification reaction was significantly lower, achieving reductions of up to 85 and 80%, respectively, compared to the viscosity of the raw materials. Furthermore, castor biodiesel has a higher viscosity than jatropha biodiesel since castor oil has the highest viscosity of all known vegetable oils; however, its complete solubility in alcohol makes it suitable for conversion to biodiesel (Valderrama et al., 1994).

The resulting chromatograms of biodiesel from castor and jatropha oil are exposed in Figures 3 and 4.

Table 3 displays the type and percentage of fatty acid methyl esters present in the extracted castor and jatropha oil.

The high presence of nitrogen compounds in biofuels is not desirable because when they burn, they become sources of NO_x in the exhaust gases (Kaewpengkrow et al., 2013). Due to the chemical structure of FAME’s, biodiesel can be more susceptible to oxidation than mineral diesel. Saturated FAME’s increase cloud point, cetane number, and improve storage stability; while polyunsaturated FAME’s decrease these properties (Das et al., 2009). However, it has been reported that unsaturated FAME’s increase fuel lubricity and improve cold flow properties (Hoekman et al., 2012). Therefore, the presence of saturated and unsaturated FAME’s is desirable for biodiesel production to obtain a fuel with better properties.

The central methyl esters in castor biodiesel were methyl ricinoleate (66.38%), methyl oleate (13.79%), and methyl linoleate (7.54%). These values were within the range of those in the literature (Okullo et al., 2012).

The chromatographic profile of jatropha biodiesel indicates that the most representative compounds were methyl oleate (31.64%), methyl linoleate (27.13%), and methyl palmitate (13.75%), similar to literature values (Piloto et al., 2011; Okullo et al., 2012).

The HHV of biodiesel can be predicted by the FAME composition of biodiesel (Fassinou, 2012; Ramírez et al., 2012). Table 4 reveals the values of w _i and x _i calculated with the relations taken from Table 3, and the HHV _i was calculated with equation 6.

Table 4 depicts that the compound with the highest contribution to HHV in castor biodiesel is methyl ricinoleate with 26.49-26.82 MJ/kg. The HHV reported for methyl ricinoleate is 40.37 MJ/kg (Fassinou, 2012), so comparing this value with that obtained in equation 6, there is a difference of 0.08%. For jatropha biodiesel, the two compounds with the highest contribution to HHV are methyl oleate with 12.68-15.36 MJ/kg and methyl linoleate with 10.81-13.18 MJ/kg. The reported values for the HHV of methyl oleate and methyl linoleate are 39.90 MJ/kg and 39.85 MJ/kg, respectively (Fassinou, 2012). Comparing these values with those obtained through equation 6 there is a difference of 0.45% for methyl oleate and 0.05% for methyl linoleate. Given the composition of the methyl esters, the HHV for castor biodiesel of 39.74 to 40.25 MJ/kg can be predicted; while for jatropha biodiesel, an HHV between the ranges of 32.37 to 39.94 MJ/kg can be expected. Ramírez et al., (2012) report that the methodology for estimating the HHV of biodiesel should not be used for castor biodiesel, so it was compared to a methodology proposed by Fassinou, (2012). The difference between both methodologies was 0.27%.

When using blend biodiesel with diesel there is a decrease in HHV with an increasing mixing ratio. The HHV from B5 to B100 represents a range of 39.74 to 41.80 MJ/kg for castor biodiesel and 32.37 to 41.79 MJ/kg for jatropha biodiesel. The HHV of biodiesels from the current work are slightly lower than those of gasoline (40.43 MJ/kg), and diesel (41.40 MJ/kg) (SENER, 2019).