- Email: [email protected]
Ethyl ester production by homogeneous alkaline transesterification: Influence of the catalyst
Bioresource Technology 102 (2011) 6385–6391
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Case Study
Ethyl ester production by homogeneous alkaline transesterification: Influence of the catalyst G. Mendow, N.S. Veizaga, C.A. Querini ⇑ Instituto de Investigaciones en Catálisis y Petroquímica-INCAPE-(FIQ-UNL, CONICET), Santiago del Estero 2654, 3000 Santa Fe, Argentina
a r t i c l e
i n f o
Article history: Received 25 November 2010 Received in revised form 19 January 2011 Accepted 20 January 2011 Available online 31 January 2011 Keywords: Biodiesel Ethanol Ethyl ester Transesterification
a b s t r a c t In this work, the process for ethyl ester production is studied using refined sunflower oil, and NaOH, KOH, CH3ONa, and CH3OK, as catalysts. In all cases, the reaction is carried out in a single reaction step. The best conversion is obtained when the catalyst is either sodium methoxide or potassium methoxide. We found that during the transesterification with ethanol, soap formation is more important than in the case of methanol. The saponification reaction consumes an important fraction of the catalyst. The amount of catalyst consumed by this reaction is 100% in the case of using hydroxides as catalyst (KOH or NaOH), and 25%, and 28% when using CH3ONa and CH3OK as catalysts, respectively. Ethanol increases the catalyst solubility in the oil–ethyl ester phase, thus accelerating the saponification reaction. It is possible to obtain high conversions in a one-step reaction, with a total glycerine concentration close to 0.25%. Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Biodiesel is a renewable fuel that can be used pure (B100) or mixed with mineral gasoil in diesel engines. It can be obtained by transesterification of vegetable oils and animal fats (triglycerides) with an alcohol in presence of a catalyst. Typically, the alcohol is methanol, and the catalyst is sodium methoxide. As a result, methyl esters and glycerine are obtained as products. If the alcohol used in the reaction is ethanol, ethyl esters and glycerine are the products. As indicated by Boehman (2005), these fuels have numerous advantages compared to those obtained from petroleum. For example, lower emissions of particulate matter (smoke), carbon monoxide, sulfur dioxide, and hydrocarbons. On the other hand, the biodiesel has better cetane index and lubricity, lower toxicity and higher biodegradability. For these reasons, this fuel has been considered as an alternative in order to partially substitute the mineral gasoil (Freedman et al., 1984, 1986; Schwad et al., 1987; Shay, 1993), gaining increasing acceptance throughout the world. The main component of the biodiesel final cost is the raw material (oil or fat, and alcohol). The largest facilities throughout the world, used soybean oil (USA, Argentina and Brazil), rapeseed oil (Europe), and palm oil (Indonesia). In medium size industries, tallow is also used, such as those found in Paraguay and Brazil. However, the alcohol that has been used almost exclusively is the methanol. Due to the high methanol consumption, the strong dependency of the biodiesel industry on this alcohol implies certain risks, such as lack of supply and high prices. The methanol price follows the crude oil price. For ⇑ Corresponding author. Tel.: +54 342 4533858; fax: +54 342 4531068. E-mail address: querini@fiq.unl.edu.ar (C.A. Querini). 0960-8524/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2011.01.072
example, in Brazil the cost of methanol in January 2008 was 1 u$s/ L, while anhydrous ethanol was below 0.6 u$s/L. At that time the crude oil had a very high price. In April 2010, both alcohols cost approximately 0.52 u$s/L (http://www.unica.com.br, 2010; http:// www.methanex.com, 2010). Nevertheless, the price of methanol is currently lower than that of ethanol, except in countries like Brazil, where the ethanol production is very important. Since the use of ethanol mixed with gasoline is also increasing in different countries, the production capacity is also increasing worldwide and, therefore, it can be expected a decrease in its cost. Ethanol is a good candidate to replace methanol, and has several advantages. One of the most important is the fact that it is a renewable fuel and, consequently, the sustainability of the biodiesel obtained by transesterification of vegetable oils (or animal fats) with ethanol, is enhanced. In addition, the ethyl esters have higher cetane number and heating power (Clark et al., 1984). Another very important advantage of ethyl esters is that they have better cold properties, such as cloud point, cold filter plugging point, or pour point (Encinar et al., 2007). From the environmental point of view, the ethyl esters lead to lower emissions of nitrogen oxides (NOx), carbon monoxide, particulate matter, and has better biodegradability, compared to the methyl esters (Boehman, 2005; Makareviciene and Janulis, 2003). However, there is another important advantage of ethyl ester, which is related to the reaction stoichiometry:
Triglyceride þ 3Ethanol 881g=mol
138g=mol
! 3Ethyl ester þ Glycerine 927g=mol
92g=mol
If the process yield is defined as the ethyl ester/triglyceride mass ratio, the yield in this case is 1.052% (ton of biodiesel/ton of oil). In the
6386
G. Mendow et al. / Bioresource Technology 102 (2011) 6385–6391
case of a production plant with a capacity of 300.000 ton/year, this increase in yield represent and extra production of 12.000 ton/year, as compared to the production with methanol. Therefore, due to the technological advantages of ethyl esters, and the potential economical benefits, it is of interest to study the production process of ethyl esters. Most of the works already reported on the ethyl esters production, did not find a suitable process. In some cases, excess of anhydrous ethanol are used, such as ethanol/triglycerides molar ratio 6:1, which in volume it represents 36 v/v% (Issariyakul et al., 2008; Marjanovic et al., 2010; Zhou et al., 2003). Another alternative presented in the literature, is the use of a cosolvent such as tetrahydrofuran (THF) (Zhou et al., 2003), which has a negative effect in the economy of the process. There are several differences in the physicochemistry of the reacting system based on ethanol compared to methanol. For example, the higher mutual miscibility between the glycerine and the esters in the presence of ethanol, severely complicate the phase separation operation after the reaction. Depending upon the ethanol/oil volume fraction loaded to the reactor, the phase separation may not occur, being necessary to add glycerine (Encinar et al., 2007; Issariyakul et al., 2008) or to evaporate the ethanol (Bouaid et al., 2009) in order to induce phase separation. Another problem is the intensive soap formation that occurs in this system, and leads to the formation of stable emulsions that also complicates the separation of phases. Therefore, the washing procedure requires large volumes of water being necessary to improve this part of the process (Encinar et al., 2007). There are several publications in which the transesterification with ethanol has been addressed (Bouaid et al., 2009; Cˇernoch et al., 2010; Encinar et al., 2007; Marjanovic et al., 2010; Zhou et al., 2003). However, the results are very different among them, not being possible to draw general conclusions regarding the best conditions in order to carry out the reaction with high conversion levels. There are several possible reasons for these discrepancies. On one hand, we have observed that the system is extremely sensitive to minor changes in the experimental conditions, such as water content, reaction temperature fluctuations and deviations, the oil/ ethanol ratio, and catalyst concentration. These variables, even with minor changes, might lead to a system with one or two phases at the end of the reaction, and this has an important effect on the final conversion. On the other hand, in several cases the analytical technique used to follow the conversion may lead to important errors. For example, the GC method, as described in the UNE-EN 14105 (2003) can be applied to methyl esters obtained with soybean, rapeseed, or sunflower oils. Therefore, this can be a source of errors in the analysis due to peak overlapping, mainly in the case of the monoglycerides quantification. In each case, this should be verified. In this work, we present the study that we carried out in order to determine the effect of different homogeneous catalysts and reaction conditions, on the transesterification reaction using ethanol in a one-stage reaction. The formation of soaps in each case is analyzed, and correlated with the catalyst activity. Results obtained with methanol are included as comparison. The total glycerine content is analyzed by a volumetric procedure (Pisarello et al., 2010), which has no limitations regarding the raw material or the alcohol used in the reaction.
2. Experimental 2.1. Biodiesel production process 2.1.1. Transesterification reaction and phase separation The reaction was carried out in a 0.5 L flask, with magnetic stirring, using a 50 mm Teflon-coated magnetic bar, and 800 rpm.
Reaction temperature was in the range 20–70 °C, and reaction time between 1 and 3 h. Refined sunflower oil with acidity less than 0.1 oleic acid/100 g sample was used as raw material. Preliminar experiments were carried out, in order to determine if the speed of agitation was enough in order to avoid mass transfer limitations. We found that in the above-described reacting system, at 500 rpm or higher speeds, there are no mass transfer limitations, being possible to obtain the same conversion at a given reaction time, using either 500, 800, 1000 or 1200 rpm. The oil was loaded to the reactor, and the temperature adjusted to the desired value. Once the oil reached this value, the alcohol containing the catalyst was added to the reactor. The addition of the ethanol–catalyst mixture was carried out within a time range of 3–5 s, taking as zero time for the soap and catalyst measurement the moment in which all the mixture was transferred to the reaction vessel. Methanol 99.8% purity (water content 350 ppm) from Cicarelli and ethanol 99.5% purity (water content 1150 ppm) from Cicarelli were used as transesterification alcohols. The catalyst was either sodium or potassium hydroxide dissolved in methanol or ethanol, and sodium or potassium methoxide (30 wt.% and 32 wt.% in methanol, respectively), from Evonik. The catalyst concentration was 0.45–1.5 g of NaOH/100 mL of oil, therefore the amount of this compound loaded to the reactor was calculated in order to neutralize the free fatty acids, plus the amount needed to catalyze the reaction. In the case of the other catalysts, equivalent amounts of moles to that used with sodium hydroxide were used. In order to optimize the reaction, different catalyst and alcohol concentrations were used, the temperature was varied between 17 and 65 °C. In all cases, the reaction was carried out in a single stage. Tables 1 and 2 summarize the experiments carried out. In those cases in which no phase separation occurred, two different procedures were used. In one of them, ethanol was evaporated at 60 °C under vacuum (60 mm Hg absolute pressure). The
Table 1 Reaction conditions using sodium hydroxide dissolved in ethanol as catalyst, during 3 h, and total glycerine content in the final biodiesel. The catalyst concentration is referred to the volume of oil. Temperature (°C)
NaOH (g/100 mL oil)
Ethanol (%v/v)
Total glycerine (%) GT
17 50 55 60 65 55 55 65 65 65
1.05 (0.26 mol/L) 1.05 1.05 1.05 1.05 1.05 1.05 1.05 0.45 (0.1137 mol/L) 0.70 (0.175 mol/L)
30 30 30 30 30 35 40 40 40 40
6.35 0.46 0.38 0.40 0.40 0.35 0.43 0.39 0.71 0.64
Table 2 Reaction conditions using sodium methoxide dissolved in ethanol as catalyst, during 3 h, and total glycerine content in the final biodiesel. The catalyst concentration is referred to the volume of oil. Temperature °C
NaCH3O (g/100 mL oil)
Ethanol (%v/v)
Reaction time (h)
Total glycerine % GT
65 65 70 65 65 70 65 55
0.945 (0.172 mol/L) 1.417 (0.257 mol/L) 0.945 1.0125 (0.184 mol/L) 0.662 (0.120 mol/L) 1.08 (0.196 mol/L) 0.945 0.945
40 40 50 50 25 30 25 25
3 3 3 3 2 3 3 3
0.32 0.40 0.44 0.50 0.43 0.37 0.27 0.26
6387
G. Mendow et al. / Bioresource Technology 102 (2011) 6385–6391
other option was to add glycerine, at temperatures between 15 and 60 °C, mixing during 5 min. Phase separation was carried out in a separatory funnel at room temperature. 2.1.2. Biodiesel purification The phase rich in biodiesel was purified using two consecutive extraction steps. In the first one, an aqueous solution of HCl 5 wt.% was used; and in the second one, fresh water. Both extraction stages (commonly called biodiesel washing), were carried out with gentle agitation at 60 °C, during 15 min. The volume of aqueous phase was 30 v/v% relative to the biodiesel phase. Afterwards, the biodiesel was dried by stripping with nitrogen at 80 °C. The total time used in the production procedure followed for the ethanolysis of refined sunflower oil, including the operations of separation and washing, was approximately 3 h. 2.1.3. Soap and catalyst analyses Small samples of the reacting mixture were taken at different times. These samples were used to determine the soap and catalyst content by an acid–base titration method (AOCS Cc17-79 or IRAM 5599). The biodiesel sample was dissolved in a toluene–ethanol mixture, and titrated with 0.1 N HCl solution, using phenolphthalein as indicator to determine the catalyst concentration. Bromophenol blue was used to determine the soap concentration. 2.1.4. Total glycerine analysis by a volumetric procedure In the volumetric procedure, the total glycerine content is obtained after all the glycerides (i.e. mono-, di- and triglycerides) present in the sample, are quantitatively transformed into methyl esters and glycerine by transesterification. Afterwards, the glycerine is extracted first with acidified water, and then with water. Finally, glycerine is titrated according to standard procedures. This method is very sensitive, and it is possible to perform the total glycerine determination with 50 g of sample or less. Free glycerine is determined just by extraction with water, followed by titration as indicated in the ASTM D-1615, IRAM 41089, BS-5711, or AOCS 14-56. Details of this technique can be found elsewhere (Pisarello et al., 2010). Unless otherwise stated, the advancement of the reaction is represented by the total glycerine value (GT), determined by the volumetric procedure. 2.1.5. GC and water analyses Mono-, di-, and tri-glycerides contents (MG, DG, and TG, respectively), were determined according to UNE-EN 14105, and water content was determined by the Karl–Fischer method (UNE-EN ISO 12937, 2001).
EN14105 standard, the GT is calculated using the following equation:
%GT ¼ %GL þ 0:255ð%MGÞ þ 0:146ð%DGÞ þ 0:103ð%TGÞ Where %GT is the total glycerine content, %GL is the free glycerine content, and %MG, %DG, %TG are the content of mono-, di-, and triglycerides, respectively, in all cases, expressed as weight percentage. According to the result mentioned above, the total glycerine content of the biodiesel was 1.09%. We carried out the volumetric analysis, and obtained a value of GT = 0.95%. These values are very similar, and are within the repeatability of the GC technique as reported in EN 14105 or ASTM D-6584. Therefore, the GC analysis can be used to determine the GT when using ethanol for the transesterification reaction with sunflower oil, without introducing appreciable errors. These results are just an example of the difficulties that we found when addressing the transesterification with ethanol, being necessary to control carefully all the process variables in order to have good reproducibility, including humidity control of the raw materials, amount of catalysts, temperature, etc. 3.1. Reaction with sodium and potassium hydroxides 3.1.1. Effect of reaction variables on triglycerides conversion The dissolution of hydroxides in the alcohol leads to the formation of the alkoxide and water, which is the compound that acts as catalyst. The reaction that takes place when using ethanol is the following:
H3 C CH2 OH þ NaOH ! H3 C CH2 O þ Naþ þ H2 O A mol of water is generated by each mol of catalyst that is formed. Using refined sunflower oil, and a hydroxide concentration equivalent to 0.175 mol/L of oil (0.70 g/100 mL oil), 25 v/v% of ethanol, 55 °C, and 3 h reaction time, we found that the total glycerine content was 1.46% and 3.06% for sodium hydroxide and potassium hydroxide, respectively. Therefore, since the sodium hydroxide is more active than the potassium hydroxide, it was chosen to carry out the optimization of the reaction conditions. It has to be mentioned that the EN 14214 standard, specifies as the maximum value of the total glycerine content, 0.25 wt.%. Fig. 1 shows the total glycerine content of the biodiesel obtained at different reaction temperatures. It can be observed that there is a large difference in conversion between the reactions carried out at 17 °C and at 50 °C or higher temperatures, which is due to the large difference in reaction rate. Between 55 and 65 °C, the conversion is practically the same, with a total glycerine content of 0.40 wt.% approximately. At higher temperature, the concentration of alcohol in the liquid phase is lower due to the higher vapor 6.35
1 3. Results and discussion
% G.T.
We carried out experiments in similar conditions to those used by other authors in order to determine the reproducibility of the experiments of transesterification with ethanol. For example, Bouaid et al. (2007) reported an experiment carried out with sunflower oil, 1.5 wt.% KOH, ethanol/oil molar ratio = 5:1, at 32 °C. They obtained that the final product contained 0% MG, 0% DG, 0%TG and 0.003% GT. We obtained under the same conditions the following composition: 1.79% MG, 2.21% DG y 2.96% TG. These values were obtained following the EN14105 standard. Obviously, this is a completely different composition than that obtained by Bouaid et al. (2007). According to the
0.8 0.6
0.46
0.4
0.38
0.4
0.4 0.2 0 17 ºC
50 ºC
55 ºC
60 ºC
65 ºC
Temperature Fig. 1. Total glycerine concentration as a function of reaction temperature. Transesterification of refined sunflower oil using ethanol and sodium hydroxide as catalyst.
6388
G. Mendow et al. / Bioresource Technology 102 (2011) 6385–6391
pressure. Nevertheless, the effect of the reaction temperature on the final conversion is negligible between 55 and 65 °C. The alcohol concentration used in the reaction experiments was between 30% and 40% v/v, referred to the oil (see Table 1). There is an optimum value, which at 55 °C is 35% v/v. At lower alcohol concentration, the reaction rate was slower and therefore, the total glycerine content increased. At higher ethanol concentration (e.g. 40% v/v), the catalyst concentration is smaller. This is because as indicated in Table 1, the catalyst is loaded into the reactor taking into account only the amount of oil. Consequently, as the ethanol amount is increased, the catalyst is diluted, and the effect of alcohol concentration and catalyst concentration are compensated. Table 1 also shows that the catalyst concentration has a positive effect, increasing the triglyceride conversion as the catalyst concentration is increased, as indicated by a lower level of total glycerine. It has to be emphasized, that in all the experiments shown in Table 1, it was not possible to obtain a conversion according to the EN14214 standards, since the total glycerine content was above 0.25 wt.%. It is interesting to remark, that using these same experimental conditions, but with methanol instead of ethanol, it is possible to obtain total glycerine levels well below 0.20 wt.%. 3.1.2. Catalyst and soaps formation The soaps and catalyst concentrations were determined as a function of reaction time. Fig. 2A show the results obtained with sodium hydroxide and potassium hydroxide as catalysts. It can
10
20
Time (min) 30 40
50
60
8
60 50 40
6
NaOH KOH
30
4 20 2
A 0
10
Soap (g Oleate/Kg Sample)
Catalyst (gHydroxide/KgSample)
0
0
50
8 40 6
30
4
20
2
10
B 0
Soap (g Oleate/Kg Sample)
Catalyst (gMethoxide/KgSample)
60 MeONa MeOK
10
0 0
10
20
30
40
50
60
Time (min) Fig. 2. Catalyst and soap evolution during transesterification using ethanol and (A) sodium and potassium hydroxide; (B) sodium and potassium methoxide Reaction conditions: temperature: 55 °C, moles catalyst/Loil: 0.175, ethanol:triglycerides molar ratio 4.25:1 (25% v/v), agitation speed: 800 rpm.
be observed that the catalyst concentration (NaOH) decreases very fast at short reaction times. At 5 min, the NaOH concentration dropped from 6.5 g/kg oil to less than 0.5 g/kg of oil. Fig. 2A also shows that the soap concentration rapidly increases at the beginning of the reaction, being at 5 min on oil 52 g sodium oleate/kg. In other words, 95% of the sodium hydroxide initially loaded as catalysts, was converted in 5 min to a sodium soap. These results indicate that under these reaction conditions, saponification reaction is favoured by the presence of water formed during sodium hydroxide dissolution in the alcohol. Consequently, low triglyceride conversion is observed. Fig. 2A shows the evolution of soaps and catalyst in an experiment carried out using KOH under the same experimental conditions as those used in the experiment with NaOH. In the case of the KOH, its concentration decreases even faster and to a lower level than in the case of NaOH. At 5 min reaction time, the catalyst concentration is zero. This result explains why the conversion obtained with KOH is lower than in the case of using NaOH as catalyst. It suggests that there is a faster saponification in the case of the former. In conclusion, when using NaOH and KOH the low triglycerides conversion is due to the transformation of these catalysts into organic salts of sodium and potassium, due to the undesirable saponification reactions that occurs due to the presence of water in the system. This water is formed during the dissolution of the hydroxides in ethanol. 3.1.3. Process yield and process strategy The yield, expressed as the mass of biodiesel obtained per unit mass of oil used as raw material, was below 90% for both catalysts. This is due to the intensive soap formation that occurred during transesterification, as shown above. The soap formation has an impact not only on the process yield, but also on the phase-separation steps downstream the reactor. The presence of large amounts of soaps complicates the biodiesel–glycerine separation due to the formation of emulsions. In addition to this problem, the ethanol increases the mutual solubility of the different components present in this reacting system. We found that depending upon the initial concentration of ethanol, the concentrations at the end of the reaction is such that the system has only one phase. This situation was never observed when using methanol as the transesterification alcohol. This fact leads to a different process design, being necessary to include steps such as ethanol evaporation, or glycerine addition. In the former case, the decrease in the ethanol concentration decreases the mutual solubility and phase separation might occur. In the second case, the addition of an adequate amount of glycerine also leads to a phase separation, with higher concentrations of alcohol, catalyst and glycerine in one phase. Therefore, new process steps are required compared to the process based on methanol. On the other hand, the ethanol evaporation may lead to the reaction reversion, due to the equilibrium displacement when removing the alcohol. 3.2. Reaction with sodium and potassium methoxides Due to the above mentioned inconvenient, it is necessary to use sodium or potassium methoxides as catalysts, since these compounds contain smaller amounts of water and, consequently, lower amount of soaps can be expected to be formed during the reaction. Table 2 shows results obtained with refined sunflower oil and ethanol. For example, using a catalyst concentration of 0.175 mol/L of oil, 25%v/v of ethanol (referred to the oil), 55 °C, and 3 h reaction time, the total glycerine content was 0.258 wt.% with sodium methoxide and 0.286 wt.% with potassium methoxide. These values are better than those obtained with hydroxides. On the other hand, there are no significant differences between
6389
G. Mendow et al. / Bioresource Technology 102 (2011) 6385–6391
3.2.1. Comparison of methanol and ethanol In order to compare the compounds distribution and the conversion, experiments using methanol and ethanol under identical experimental conditions were carried out. Table 3 shows the results. It can be observed that the main difference is the total amount of soaps formed with each alcohol. Using ethanol, there is 3.5 times more soap than when using methanol. This is a big disadvantage, since soaps inhibit the transesterification reaction, and it implies large catalyst consumption in order to be converted into soaps. This is another reason that leads to higher conversion in the system when using methanol. Data in Table 3 shows that most of the soaps are concentrated in the glycerine phase. In the case of the alcohol distribution, it is similar for both of them, with similar concentrations for methanol compared to ethanol in both phases. It is important to remark that the soap concentration in the biodiesel-rich phase is higher when the reaction was carried out with ethanol. This is a major disadvantage, since in a conventional process the decantation is followed by neutralization. In this process stage, the soaps are hydrolyzed forming fatty acids, which are soluble in the biodiesel phase, thus leading to a product with acidity out of specification.
Table 3 Soaps, catalyst, alcohol and glycerine concentration at the end of the reaction. Reaction conditions: Temperature: 65 °C, 0.662 g H3CONa/100 mL oil, ethanol or methanol 25% v/v, agitation speed: 800 rpm, reaction time: 2 h.
Soaps biodiesel phase (goleate/kg) Soaps glycerine phase (goleate/kg) Catalyst biodiesel phase (gMeONa/kg) Catalyst glycerine phase (gMeONa/kg) Total glycerine (%) % Glycerine glycerine phase % Alcohol biodiesel phase % Alcohol glycerine phase (1) Methanol content.
Methanol
Ethanol
0.985 22.96 0.175 44.58 0.163 63 8.28 21.3
3.045 94.58 0.378 30.00 0.43 63 0.19 (1)/6.2 2.15(1)/22.4
Water in the methanol
Soaps in biodiesel (goleate/kg)
Soaps in glycerine (goleate/kg)
Catalyst in biodiesel (gMeONa/kg)
Catalyst in glycerine (gMeONa/kg)
345 ppm 1900 ppm
0.985 0.669
22.96 23.66
0.175 0.336
44.58 41.19
saponification rate observed with ethanol and with methanol, is due to differences between the properties of both alcohols. 3.2.3. Catalyst concentration and soaps formation The concentration of catalyst and soap during the transesterification with ethanol were determined as a function of time, both with sodium and potassium methoxides as catalysts. Results are shown in Fig. 2B. In both cases, the catalyst concentration decreased at the beginning of the reaction. However, this drop in concentration is much smaller than the one observed when using hydroxides as catalysts. The soap concentration obtained with methoxides is approximately 3.5 times smaller than that obtained with hydroxides. Therefore, it can be concluded that sodium and potassium methoxides display higher activity than the corresponding hydroxides, due to a lower extent of saponification that decreases the catalyst concentration. This difference is due to lower water content on the methoxides, as compared to the hydroxides. Fig. 3 shows the results obtained with methanol in an experiment carried out under the same experimental conditions as those used in Fig. 2B. It is very interesting to observe that the catalyst consumption is much lower with methanol, than in the case of using ethanol during the transesterification (compare Figs. 2B and 3). This is in agreement with above discussion, showing that the presence of ethanol favors the saponification reaction. This leads to a higher amount of soaps formed during the reaction, higher catalyst consumption, and a lower conversion when using ethanol. Note that the total glycerine content in the case of using methanol is 0.163 wt.%, while with ethanol under the same experimental conditions is 0.43%. The mass balance for the catalyst was carried out. It was found that in the case of the experiment with methanol, approximately 10% of the catalyst was consumed to produce soaps (Fig. 3), while in the case of the experiment with ethanol, approximately 25% of the catalyst was consumed to produce soaps (Fig. 2B).
60 8
50 40
6 MeONa/MeOH 4
30 20
2
10
Soap (g Oleate/Kg Sample)
3.2.2. Effect of water A possible explanation for the larger saponification rate observed in the process carried out with ethanol might be the higher water content of this alcohol as compared to the methanol. The ethanol used in these experiments contained 1150 ppm of water, while methanol contained 345 ppm. To verify this hypothesis, an experiment with methanol was carried out, adding water to the system in order to have the same amount of water as in the case of ethanol. Results are shown in Table 4. It can be observed that the amounts of soap formed with methanol in both experiments with different amounts of water are practically the same. Therefore, it can be concluded that the difference between the
Table 4 Soap formation during the transesterification with methanol. Effect of water content. Reaction conditions: Temperature: 65 °C, 0.662 g H3CONa/100 mL oil, methanol 25% v/v, agitation speed: 800 rpm, reaction time: 2 h.
Catalyst (gMeONa/KgSample)
the sodium and the potassium methoxides. Therefore, most of the following experiments were carried out using the former, since it is cheaper and of higher availability. In all the experiments shown in Table 2, there was a good and fast phase separation after the reaction. No soapy interphases or emulsions were observed. These are important observations from a technological point of view, since no additional steps have to be added to the process compared to the methanol-based process, in order to induce phase separation. Comparing the results shown in Tables 1 and 2, and the observations related to phase separation; it can be reasonably established that the intensive soap formation observed with the hydroxides as catalysts, is one of the reasons why no phase-separation occurs.
0
0 0
20
40
60
80 100 120 140 160 180 Time (min)
Fig. 3. Catalyst and soap evolution during transesterification using methanol and sodium methoxide.
6390
G. Mendow et al. / Bioresource Technology 102 (2011) 6385–6391
2- 40 g Methanol (25%) 3- 4.41 g MeONa
1- 180 g Oil
7- Acid Water
9- Water
Biodiesel 190.8 g
11- Biodiesel 179.5 g
4 W-1
D-1 R-1
Glycerol 28.1 g
W-2
8- Water/ Methanol
10- Acid Water
S-1
Fig. 4. Mass balance during transesterification of sunflower oil, using methanol and sodium methoxide as catalyst. Reaction conditions: temperature: 65 °C, 0.662 g H3CONa/ 100 mL oil, methanol 25% v/v, agitation speed: 800 rpm, reaction time: 2 h.
1- 180 g Oil
2- 39.5 g Ethanol (25%) 3- 4.41 g MeONa 7- Acid Water 9- Water Biodiesel 196.9 g
4 D-1
11- Biodiesel 179.5 g
W-1
W-2
8- Water/ Ethanol
R-1 Glycerol: 24.54 g
10-Acid Water
S-1
Fig. 5. Mass balance during transesterification of sunflower oil, using ethanol and sodium methoxide as catalyst. Reaction conditions: Temperature: 65 °C, 0.662 g H3CONa/ 100 mL oil, ethanol 25% v/v, agitation speed: 800 rpm, reaction time: 2 h.
Consequently, the lower conversion observed with ethanol is at least partly, due to this catalyst transformation in soaps. 3.2.4. Process yields The mass balances were carried out both with methanol and with ethanol, in order to determine the process yield, expressed as (biodiesel mass/oil mass). Figs. 4 and 5 show the results. In the case of the methyl ester production, the maximum theoretical yield is 100.45%, while in the case of the ethyl esters it is 105.22%. This is due to the difference in the molecular weight of the alcohols and, consequently, the higher molecular weight of the ethyl ester molecule, as compared to that of the methyl ester. Figs. 4 and 5 show that the experimental values were very similar in both cases, approximately 99.5%. Therefore, in the case of the ethanol, there is a larger loss of yield as compared to the maximum possible value, due to the higher amount of soaps formed when using ethanol. 3.2.5. Effect of reaction variables on triglycerides conversion The transesterification reaction using ethanol and sodium methoxide as catalyst was carried out under several experimental conditions. The results are shown in Table 2. The best conversion level obtained in these experiments corresponds to a total glycerine content of 0.258 wt.%, and was obtained using a high catalyst concentration. Even this, it was not possible to obtain the value required by the EN14214 standard, which is 0.25 wt.%. In the temperature range analyzed in this study (55–70 °C), the best result was obtained at the lowest temperature, i.e. 55 °C. The conversion shows an optimum value as a function of the catalyst concentration. At low concentration (0.662 g MeONa/ 100 mL oil) the total glycerine concentration was 0.43 wt.%, and at high catalyst concentration (1.417 g MeONa/100 mL oil) the total glycerine concentration was 0.396 wt.%. At low catalyst doses, the reaction rate is too low. At high catalyst concentration, the high rate of soaps formation leads to a reaction rate inhibition by these compounds and, consequently, a lower conversion was obtained,
as above discussed. The best conversion was obtained using a sodium methoxide concentration of 0.945 g/100 mL oil. In the case of the ethanol volume loaded into the reactor, it can be seen that as the ratio ethanol/oil (volume) is decreased, the conversion is better. As explained in the case of the results obtained with sodium hydroxide, this is due to a catalyst dilution effect.
3.3. Ethanol vs. methanol effects on the physico-chemical properties The data shown and discussed above can be summarized as follows. In all cases, the result obtained with ethanol is compared with the corresponding result obtained with methanol: - Saponification reactions occurs faster with ethanol. - The soap concentration is higher in the biodiesel phase with ethanol. - The catalyst concentration is higher in the biodiesel phase with ethanol. These results are in agreement with the liquid–liquid equilibria reported by other authors. It has been shown that the partial miscibility zone in the ternary diagram alcohol–glycerine–ester is considerably smaller in the case of ethanol compared to methanol (Andreatta et al., 2008; Liu et al., 2008; Zhou et al., 2006). This means that the concentration of glycerine in the biodiesel phase is higher in the case of using ethanol. There are no reports regarding the concentration of water, catalyst, or soaps in each phase after the reaction. Our results indicate that the higher mutual solubility of the compounds that form the system when using ethanol, also increases the solubility of catalyst, water, and soaps in the biodiesel phase. The consequence of this phenomenon is that the presence of higher concentration of catalyst and water in this phase accelerates the saponification reaction and the catalyst consumption. In the case of using methanol, the catalyst and the water are concentrated in the glycerine phase, and due to mass transfer limitations, the saponification occurs at a lower rate.
G. Mendow et al. / Bioresource Technology 102 (2011) 6385–6391
4. Conclusions In this work, we showed that the saponification reaction is much faster in the case of using ethanol during the transesterification, and because of this, higher catalyst consumption takes place and consequently, a lower reaction rate with a lower final conversion is observed. On the other hand, in the presence of methanol, the amount of soaps formed during the reaction is considerably lower, and this is another reason for the much higher activity observed when using this alcohol. Using methoxides as catalysts, phase separation is fast and clear and there is no formation of stable emulsions. References Andreatta, A.E., Casa´s, L.M., Hegel, P., Bottini, S.B., Brignole, E.A., 2008. Phase equilibria in ternary mixtures of methyl oleate, glycerol, and methanol. Ind. Eng. Chem. Res. 47, 5157–5164. Boehman, L., 2005. Biodiesel production and processing. Fuel Process. Technol. 86, 1057–1058. Bouaid, A., Martínez, M., Aracil, J., 2007. A comparative study of the production of ethyl esters from vegetable oils as a biodiesel fuel optimization by factorial design. Chem. Eng. J. 134, 93–99. Bouaid, A., Martinez, M., Aracil, J., 2009. Production of biodiesel from bioethanol and Brassica carinata oil: oxidation stability study. Biores. Technol. 100, 2234–2239. ˇ ernoch, M., Hájek, M., Skopal, F., 2010. Ethanolysis of rapeseed oil – distribution of C ethyl esters, glycerides and glycerol between ester and glycerol phases. Biores. Technol. 101, 2071–2075. Clark, S.J., Wagner, L., Schrock, M.D., Pienaar, P.G., 1984. Methyl and ethyl soybean esters as renewable fuels for diesel engines. J. Am. Oil Chem. Soc. 61, 1632– 1638.
6391
Encinar, J.M., González, J.F., Rodríguez-Reinares, A., 2007. Ethanolysis of used frying oil. Biodiesel preparation and characterization. Fuel Proc. Technol. 88, 513–522. Freedman, B., Pryde, E.H., Mounts, T.L., 1984. Variables affecting the yields of fatty esters from transesterified vegetable oils. J. Am. Oil Chem. Soc. 61, 1638–1643. Freedman, B., Butterfield, R.O., Pryde, E.H., 1986. Transesterification kinetics of soybean oil. J. Am. Oil Chem. Soc. 63, 1375–1380. Issariyakul, T., Kulkarni, M.G., Meher, L.C., Dalai, A.K., Bakhshi, N.N., 2008. Biodiesel production from mixtures of canola oil and used cooking oil. Chem. Eng. J. 140, 77. Liu, X., Piao, X., Wang, Y., Zhu, S., 2008. Liquid–liquid equilibrium for systems of (fatty acid ethyl esters + ethanol + soybean oil and fatty acid ethyl esters + ethanol + glycerol). J. Chem. Eng. Data 53, 359–362. Makareviciene, V., Janulis, P., 2003. Environmental effect of rapseed oil ethyl ester. Renewable Energy 28, 2395–2403. Marjanovic, A.V., Stamenkovic, O.S., Todorovic, Z.B., Lazic, M.L., Veljkovic, V.B., 2010. Kinetics of the base-catalyzed sunflower oil ethanolysis. Fuel 89, 665– 671. Pisarello, M.L., Dalla Costa, B.O., Veizaga, N.S., Querini, C.A., 2010. Volumetric method for free and total glycerin determination in biodiesel. Ind. Eng. Chem. Res. 49, 8935–8941. Schwad, A.W., Bagby, M.O., Freedman, B., 1987. Preparation and properties of diesel fuels from vegetable oils. Fuel 66, 1372–1378. Shay, E.G., 1993. Diesel fuel from vegetable oils: status and opportunities. Biomass and Bioenergy 4, 227–242. UNE-EN 14105, 2003. Determination of free and total glycerol and mono-, di-, and tri-glyceride contents. UNE-EN ISO 12937, 2001. Coulometric Karl Fischer titration method. Zhou, H., Lu, H., Liang, B., 2006. Solubility of multicomponent systems in the biodiesel production by transesterification of Jatropha Curcas L. Oil with methanol. J. Chem. Eng. Data 51, 1130–1135. Zhou, W., Konar, S.K., Boocock, D.G.B., 2003. Ethyl esters from the single-phase base-catalyzed ethanolysis of vegetable oils. J. Am. Oil Chem. Soc. 80, 367– 371.
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Case Study
Ethyl ester production by homogeneous alkaline transesterification: Influence of the catalyst G. Mendow, N.S. Veizaga, C.A. Querini ⇑ Instituto de Investigaciones en Catálisis y Petroquímica-INCAPE-(FIQ-UNL, CONICET), Santiago del Estero 2654, 3000 Santa Fe, Argentina
a r t i c l e
i n f o
Article history: Received 25 November 2010 Received in revised form 19 January 2011 Accepted 20 January 2011 Available online 31 January 2011 Keywords: Biodiesel Ethanol Ethyl ester Transesterification
a b s t r a c t In this work, the process for ethyl ester production is studied using refined sunflower oil, and NaOH, KOH, CH3ONa, and CH3OK, as catalysts. In all cases, the reaction is carried out in a single reaction step. The best conversion is obtained when the catalyst is either sodium methoxide or potassium methoxide. We found that during the transesterification with ethanol, soap formation is more important than in the case of methanol. The saponification reaction consumes an important fraction of the catalyst. The amount of catalyst consumed by this reaction is 100% in the case of using hydroxides as catalyst (KOH or NaOH), and 25%, and 28% when using CH3ONa and CH3OK as catalysts, respectively. Ethanol increases the catalyst solubility in the oil–ethyl ester phase, thus accelerating the saponification reaction. It is possible to obtain high conversions in a one-step reaction, with a total glycerine concentration close to 0.25%. Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Biodiesel is a renewable fuel that can be used pure (B100) or mixed with mineral gasoil in diesel engines. It can be obtained by transesterification of vegetable oils and animal fats (triglycerides) with an alcohol in presence of a catalyst. Typically, the alcohol is methanol, and the catalyst is sodium methoxide. As a result, methyl esters and glycerine are obtained as products. If the alcohol used in the reaction is ethanol, ethyl esters and glycerine are the products. As indicated by Boehman (2005), these fuels have numerous advantages compared to those obtained from petroleum. For example, lower emissions of particulate matter (smoke), carbon monoxide, sulfur dioxide, and hydrocarbons. On the other hand, the biodiesel has better cetane index and lubricity, lower toxicity and higher biodegradability. For these reasons, this fuel has been considered as an alternative in order to partially substitute the mineral gasoil (Freedman et al., 1984, 1986; Schwad et al., 1987; Shay, 1993), gaining increasing acceptance throughout the world. The main component of the biodiesel final cost is the raw material (oil or fat, and alcohol). The largest facilities throughout the world, used soybean oil (USA, Argentina and Brazil), rapeseed oil (Europe), and palm oil (Indonesia). In medium size industries, tallow is also used, such as those found in Paraguay and Brazil. However, the alcohol that has been used almost exclusively is the methanol. Due to the high methanol consumption, the strong dependency of the biodiesel industry on this alcohol implies certain risks, such as lack of supply and high prices. The methanol price follows the crude oil price. For ⇑ Corresponding author. Tel.: +54 342 4533858; fax: +54 342 4531068. E-mail address: querini@fiq.unl.edu.ar (C.A. Querini). 0960-8524/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2011.01.072
example, in Brazil the cost of methanol in January 2008 was 1 u$s/ L, while anhydrous ethanol was below 0.6 u$s/L. At that time the crude oil had a very high price. In April 2010, both alcohols cost approximately 0.52 u$s/L (http://www.unica.com.br, 2010; http:// www.methanex.com, 2010). Nevertheless, the price of methanol is currently lower than that of ethanol, except in countries like Brazil, where the ethanol production is very important. Since the use of ethanol mixed with gasoline is also increasing in different countries, the production capacity is also increasing worldwide and, therefore, it can be expected a decrease in its cost. Ethanol is a good candidate to replace methanol, and has several advantages. One of the most important is the fact that it is a renewable fuel and, consequently, the sustainability of the biodiesel obtained by transesterification of vegetable oils (or animal fats) with ethanol, is enhanced. In addition, the ethyl esters have higher cetane number and heating power (Clark et al., 1984). Another very important advantage of ethyl esters is that they have better cold properties, such as cloud point, cold filter plugging point, or pour point (Encinar et al., 2007). From the environmental point of view, the ethyl esters lead to lower emissions of nitrogen oxides (NOx), carbon monoxide, particulate matter, and has better biodegradability, compared to the methyl esters (Boehman, 2005; Makareviciene and Janulis, 2003). However, there is another important advantage of ethyl ester, which is related to the reaction stoichiometry:
Triglyceride þ 3Ethanol 881g=mol
138g=mol
! 3Ethyl ester þ Glycerine 927g=mol
92g=mol
If the process yield is defined as the ethyl ester/triglyceride mass ratio, the yield in this case is 1.052% (ton of biodiesel/ton of oil). In the
6386
G. Mendow et al. / Bioresource Technology 102 (2011) 6385–6391
case of a production plant with a capacity of 300.000 ton/year, this increase in yield represent and extra production of 12.000 ton/year, as compared to the production with methanol. Therefore, due to the technological advantages of ethyl esters, and the potential economical benefits, it is of interest to study the production process of ethyl esters. Most of the works already reported on the ethyl esters production, did not find a suitable process. In some cases, excess of anhydrous ethanol are used, such as ethanol/triglycerides molar ratio 6:1, which in volume it represents 36 v/v% (Issariyakul et al., 2008; Marjanovic et al., 2010; Zhou et al., 2003). Another alternative presented in the literature, is the use of a cosolvent such as tetrahydrofuran (THF) (Zhou et al., 2003), which has a negative effect in the economy of the process. There are several differences in the physicochemistry of the reacting system based on ethanol compared to methanol. For example, the higher mutual miscibility between the glycerine and the esters in the presence of ethanol, severely complicate the phase separation operation after the reaction. Depending upon the ethanol/oil volume fraction loaded to the reactor, the phase separation may not occur, being necessary to add glycerine (Encinar et al., 2007; Issariyakul et al., 2008) or to evaporate the ethanol (Bouaid et al., 2009) in order to induce phase separation. Another problem is the intensive soap formation that occurs in this system, and leads to the formation of stable emulsions that also complicates the separation of phases. Therefore, the washing procedure requires large volumes of water being necessary to improve this part of the process (Encinar et al., 2007). There are several publications in which the transesterification with ethanol has been addressed (Bouaid et al., 2009; Cˇernoch et al., 2010; Encinar et al., 2007; Marjanovic et al., 2010; Zhou et al., 2003). However, the results are very different among them, not being possible to draw general conclusions regarding the best conditions in order to carry out the reaction with high conversion levels. There are several possible reasons for these discrepancies. On one hand, we have observed that the system is extremely sensitive to minor changes in the experimental conditions, such as water content, reaction temperature fluctuations and deviations, the oil/ ethanol ratio, and catalyst concentration. These variables, even with minor changes, might lead to a system with one or two phases at the end of the reaction, and this has an important effect on the final conversion. On the other hand, in several cases the analytical technique used to follow the conversion may lead to important errors. For example, the GC method, as described in the UNE-EN 14105 (2003) can be applied to methyl esters obtained with soybean, rapeseed, or sunflower oils. Therefore, this can be a source of errors in the analysis due to peak overlapping, mainly in the case of the monoglycerides quantification. In each case, this should be verified. In this work, we present the study that we carried out in order to determine the effect of different homogeneous catalysts and reaction conditions, on the transesterification reaction using ethanol in a one-stage reaction. The formation of soaps in each case is analyzed, and correlated with the catalyst activity. Results obtained with methanol are included as comparison. The total glycerine content is analyzed by a volumetric procedure (Pisarello et al., 2010), which has no limitations regarding the raw material or the alcohol used in the reaction.
2. Experimental 2.1. Biodiesel production process 2.1.1. Transesterification reaction and phase separation The reaction was carried out in a 0.5 L flask, with magnetic stirring, using a 50 mm Teflon-coated magnetic bar, and 800 rpm.
Reaction temperature was in the range 20–70 °C, and reaction time between 1 and 3 h. Refined sunflower oil with acidity less than 0.1 oleic acid/100 g sample was used as raw material. Preliminar experiments were carried out, in order to determine if the speed of agitation was enough in order to avoid mass transfer limitations. We found that in the above-described reacting system, at 500 rpm or higher speeds, there are no mass transfer limitations, being possible to obtain the same conversion at a given reaction time, using either 500, 800, 1000 or 1200 rpm. The oil was loaded to the reactor, and the temperature adjusted to the desired value. Once the oil reached this value, the alcohol containing the catalyst was added to the reactor. The addition of the ethanol–catalyst mixture was carried out within a time range of 3–5 s, taking as zero time for the soap and catalyst measurement the moment in which all the mixture was transferred to the reaction vessel. Methanol 99.8% purity (water content 350 ppm) from Cicarelli and ethanol 99.5% purity (water content 1150 ppm) from Cicarelli were used as transesterification alcohols. The catalyst was either sodium or potassium hydroxide dissolved in methanol or ethanol, and sodium or potassium methoxide (30 wt.% and 32 wt.% in methanol, respectively), from Evonik. The catalyst concentration was 0.45–1.5 g of NaOH/100 mL of oil, therefore the amount of this compound loaded to the reactor was calculated in order to neutralize the free fatty acids, plus the amount needed to catalyze the reaction. In the case of the other catalysts, equivalent amounts of moles to that used with sodium hydroxide were used. In order to optimize the reaction, different catalyst and alcohol concentrations were used, the temperature was varied between 17 and 65 °C. In all cases, the reaction was carried out in a single stage. Tables 1 and 2 summarize the experiments carried out. In those cases in which no phase separation occurred, two different procedures were used. In one of them, ethanol was evaporated at 60 °C under vacuum (60 mm Hg absolute pressure). The
Table 1 Reaction conditions using sodium hydroxide dissolved in ethanol as catalyst, during 3 h, and total glycerine content in the final biodiesel. The catalyst concentration is referred to the volume of oil. Temperature (°C)
NaOH (g/100 mL oil)
Ethanol (%v/v)
Total glycerine (%) GT
17 50 55 60 65 55 55 65 65 65
1.05 (0.26 mol/L) 1.05 1.05 1.05 1.05 1.05 1.05 1.05 0.45 (0.1137 mol/L) 0.70 (0.175 mol/L)
30 30 30 30 30 35 40 40 40 40
6.35 0.46 0.38 0.40 0.40 0.35 0.43 0.39 0.71 0.64
Table 2 Reaction conditions using sodium methoxide dissolved in ethanol as catalyst, during 3 h, and total glycerine content in the final biodiesel. The catalyst concentration is referred to the volume of oil. Temperature °C
NaCH3O (g/100 mL oil)
Ethanol (%v/v)
Reaction time (h)
Total glycerine % GT
65 65 70 65 65 70 65 55
0.945 (0.172 mol/L) 1.417 (0.257 mol/L) 0.945 1.0125 (0.184 mol/L) 0.662 (0.120 mol/L) 1.08 (0.196 mol/L) 0.945 0.945
40 40 50 50 25 30 25 25
3 3 3 3 2 3 3 3
0.32 0.40 0.44 0.50 0.43 0.37 0.27 0.26
6387
G. Mendow et al. / Bioresource Technology 102 (2011) 6385–6391
other option was to add glycerine, at temperatures between 15 and 60 °C, mixing during 5 min. Phase separation was carried out in a separatory funnel at room temperature. 2.1.2. Biodiesel purification The phase rich in biodiesel was purified using two consecutive extraction steps. In the first one, an aqueous solution of HCl 5 wt.% was used; and in the second one, fresh water. Both extraction stages (commonly called biodiesel washing), were carried out with gentle agitation at 60 °C, during 15 min. The volume of aqueous phase was 30 v/v% relative to the biodiesel phase. Afterwards, the biodiesel was dried by stripping with nitrogen at 80 °C. The total time used in the production procedure followed for the ethanolysis of refined sunflower oil, including the operations of separation and washing, was approximately 3 h. 2.1.3. Soap and catalyst analyses Small samples of the reacting mixture were taken at different times. These samples were used to determine the soap and catalyst content by an acid–base titration method (AOCS Cc17-79 or IRAM 5599). The biodiesel sample was dissolved in a toluene–ethanol mixture, and titrated with 0.1 N HCl solution, using phenolphthalein as indicator to determine the catalyst concentration. Bromophenol blue was used to determine the soap concentration. 2.1.4. Total glycerine analysis by a volumetric procedure In the volumetric procedure, the total glycerine content is obtained after all the glycerides (i.e. mono-, di- and triglycerides) present in the sample, are quantitatively transformed into methyl esters and glycerine by transesterification. Afterwards, the glycerine is extracted first with acidified water, and then with water. Finally, glycerine is titrated according to standard procedures. This method is very sensitive, and it is possible to perform the total glycerine determination with 50 g of sample or less. Free glycerine is determined just by extraction with water, followed by titration as indicated in the ASTM D-1615, IRAM 41089, BS-5711, or AOCS 14-56. Details of this technique can be found elsewhere (Pisarello et al., 2010). Unless otherwise stated, the advancement of the reaction is represented by the total glycerine value (GT), determined by the volumetric procedure. 2.1.5. GC and water analyses Mono-, di-, and tri-glycerides contents (MG, DG, and TG, respectively), were determined according to UNE-EN 14105, and water content was determined by the Karl–Fischer method (UNE-EN ISO 12937, 2001).
EN14105 standard, the GT is calculated using the following equation:
%GT ¼ %GL þ 0:255ð%MGÞ þ 0:146ð%DGÞ þ 0:103ð%TGÞ Where %GT is the total glycerine content, %GL is the free glycerine content, and %MG, %DG, %TG are the content of mono-, di-, and triglycerides, respectively, in all cases, expressed as weight percentage. According to the result mentioned above, the total glycerine content of the biodiesel was 1.09%. We carried out the volumetric analysis, and obtained a value of GT = 0.95%. These values are very similar, and are within the repeatability of the GC technique as reported in EN 14105 or ASTM D-6584. Therefore, the GC analysis can be used to determine the GT when using ethanol for the transesterification reaction with sunflower oil, without introducing appreciable errors. These results are just an example of the difficulties that we found when addressing the transesterification with ethanol, being necessary to control carefully all the process variables in order to have good reproducibility, including humidity control of the raw materials, amount of catalysts, temperature, etc. 3.1. Reaction with sodium and potassium hydroxides 3.1.1. Effect of reaction variables on triglycerides conversion The dissolution of hydroxides in the alcohol leads to the formation of the alkoxide and water, which is the compound that acts as catalyst. The reaction that takes place when using ethanol is the following:
H3 C CH2 OH þ NaOH ! H3 C CH2 O þ Naþ þ H2 O A mol of water is generated by each mol of catalyst that is formed. Using refined sunflower oil, and a hydroxide concentration equivalent to 0.175 mol/L of oil (0.70 g/100 mL oil), 25 v/v% of ethanol, 55 °C, and 3 h reaction time, we found that the total glycerine content was 1.46% and 3.06% for sodium hydroxide and potassium hydroxide, respectively. Therefore, since the sodium hydroxide is more active than the potassium hydroxide, it was chosen to carry out the optimization of the reaction conditions. It has to be mentioned that the EN 14214 standard, specifies as the maximum value of the total glycerine content, 0.25 wt.%. Fig. 1 shows the total glycerine content of the biodiesel obtained at different reaction temperatures. It can be observed that there is a large difference in conversion between the reactions carried out at 17 °C and at 50 °C or higher temperatures, which is due to the large difference in reaction rate. Between 55 and 65 °C, the conversion is practically the same, with a total glycerine content of 0.40 wt.% approximately. At higher temperature, the concentration of alcohol in the liquid phase is lower due to the higher vapor 6.35
1 3. Results and discussion
% G.T.
We carried out experiments in similar conditions to those used by other authors in order to determine the reproducibility of the experiments of transesterification with ethanol. For example, Bouaid et al. (2007) reported an experiment carried out with sunflower oil, 1.5 wt.% KOH, ethanol/oil molar ratio = 5:1, at 32 °C. They obtained that the final product contained 0% MG, 0% DG, 0%TG and 0.003% GT. We obtained under the same conditions the following composition: 1.79% MG, 2.21% DG y 2.96% TG. These values were obtained following the EN14105 standard. Obviously, this is a completely different composition than that obtained by Bouaid et al. (2007). According to the
0.8 0.6
0.46
0.4
0.38
0.4
0.4 0.2 0 17 ºC
50 ºC
55 ºC
60 ºC
65 ºC
Temperature Fig. 1. Total glycerine concentration as a function of reaction temperature. Transesterification of refined sunflower oil using ethanol and sodium hydroxide as catalyst.
6388
G. Mendow et al. / Bioresource Technology 102 (2011) 6385–6391
pressure. Nevertheless, the effect of the reaction temperature on the final conversion is negligible between 55 and 65 °C. The alcohol concentration used in the reaction experiments was between 30% and 40% v/v, referred to the oil (see Table 1). There is an optimum value, which at 55 °C is 35% v/v. At lower alcohol concentration, the reaction rate was slower and therefore, the total glycerine content increased. At higher ethanol concentration (e.g. 40% v/v), the catalyst concentration is smaller. This is because as indicated in Table 1, the catalyst is loaded into the reactor taking into account only the amount of oil. Consequently, as the ethanol amount is increased, the catalyst is diluted, and the effect of alcohol concentration and catalyst concentration are compensated. Table 1 also shows that the catalyst concentration has a positive effect, increasing the triglyceride conversion as the catalyst concentration is increased, as indicated by a lower level of total glycerine. It has to be emphasized, that in all the experiments shown in Table 1, it was not possible to obtain a conversion according to the EN14214 standards, since the total glycerine content was above 0.25 wt.%. It is interesting to remark, that using these same experimental conditions, but with methanol instead of ethanol, it is possible to obtain total glycerine levels well below 0.20 wt.%. 3.1.2. Catalyst and soaps formation The soaps and catalyst concentrations were determined as a function of reaction time. Fig. 2A show the results obtained with sodium hydroxide and potassium hydroxide as catalysts. It can
10
20
Time (min) 30 40
50
60
8
60 50 40
6
NaOH KOH
30
4 20 2
A 0
10
Soap (g Oleate/Kg Sample)
Catalyst (gHydroxide/KgSample)
0
0
50
8 40 6
30
4
20
2
10
B 0
Soap (g Oleate/Kg Sample)
Catalyst (gMethoxide/KgSample)
60 MeONa MeOK
10
0 0
10
20
30
40
50
60
Time (min) Fig. 2. Catalyst and soap evolution during transesterification using ethanol and (A) sodium and potassium hydroxide; (B) sodium and potassium methoxide Reaction conditions: temperature: 55 °C, moles catalyst/Loil: 0.175, ethanol:triglycerides molar ratio 4.25:1 (25% v/v), agitation speed: 800 rpm.
be observed that the catalyst concentration (NaOH) decreases very fast at short reaction times. At 5 min, the NaOH concentration dropped from 6.5 g/kg oil to less than 0.5 g/kg of oil. Fig. 2A also shows that the soap concentration rapidly increases at the beginning of the reaction, being at 5 min on oil 52 g sodium oleate/kg. In other words, 95% of the sodium hydroxide initially loaded as catalysts, was converted in 5 min to a sodium soap. These results indicate that under these reaction conditions, saponification reaction is favoured by the presence of water formed during sodium hydroxide dissolution in the alcohol. Consequently, low triglyceride conversion is observed. Fig. 2A shows the evolution of soaps and catalyst in an experiment carried out using KOH under the same experimental conditions as those used in the experiment with NaOH. In the case of the KOH, its concentration decreases even faster and to a lower level than in the case of NaOH. At 5 min reaction time, the catalyst concentration is zero. This result explains why the conversion obtained with KOH is lower than in the case of using NaOH as catalyst. It suggests that there is a faster saponification in the case of the former. In conclusion, when using NaOH and KOH the low triglycerides conversion is due to the transformation of these catalysts into organic salts of sodium and potassium, due to the undesirable saponification reactions that occurs due to the presence of water in the system. This water is formed during the dissolution of the hydroxides in ethanol. 3.1.3. Process yield and process strategy The yield, expressed as the mass of biodiesel obtained per unit mass of oil used as raw material, was below 90% for both catalysts. This is due to the intensive soap formation that occurred during transesterification, as shown above. The soap formation has an impact not only on the process yield, but also on the phase-separation steps downstream the reactor. The presence of large amounts of soaps complicates the biodiesel–glycerine separation due to the formation of emulsions. In addition to this problem, the ethanol increases the mutual solubility of the different components present in this reacting system. We found that depending upon the initial concentration of ethanol, the concentrations at the end of the reaction is such that the system has only one phase. This situation was never observed when using methanol as the transesterification alcohol. This fact leads to a different process design, being necessary to include steps such as ethanol evaporation, or glycerine addition. In the former case, the decrease in the ethanol concentration decreases the mutual solubility and phase separation might occur. In the second case, the addition of an adequate amount of glycerine also leads to a phase separation, with higher concentrations of alcohol, catalyst and glycerine in one phase. Therefore, new process steps are required compared to the process based on methanol. On the other hand, the ethanol evaporation may lead to the reaction reversion, due to the equilibrium displacement when removing the alcohol. 3.2. Reaction with sodium and potassium methoxides Due to the above mentioned inconvenient, it is necessary to use sodium or potassium methoxides as catalysts, since these compounds contain smaller amounts of water and, consequently, lower amount of soaps can be expected to be formed during the reaction. Table 2 shows results obtained with refined sunflower oil and ethanol. For example, using a catalyst concentration of 0.175 mol/L of oil, 25%v/v of ethanol (referred to the oil), 55 °C, and 3 h reaction time, the total glycerine content was 0.258 wt.% with sodium methoxide and 0.286 wt.% with potassium methoxide. These values are better than those obtained with hydroxides. On the other hand, there are no significant differences between
6389
G. Mendow et al. / Bioresource Technology 102 (2011) 6385–6391
3.2.1. Comparison of methanol and ethanol In order to compare the compounds distribution and the conversion, experiments using methanol and ethanol under identical experimental conditions were carried out. Table 3 shows the results. It can be observed that the main difference is the total amount of soaps formed with each alcohol. Using ethanol, there is 3.5 times more soap than when using methanol. This is a big disadvantage, since soaps inhibit the transesterification reaction, and it implies large catalyst consumption in order to be converted into soaps. This is another reason that leads to higher conversion in the system when using methanol. Data in Table 3 shows that most of the soaps are concentrated in the glycerine phase. In the case of the alcohol distribution, it is similar for both of them, with similar concentrations for methanol compared to ethanol in both phases. It is important to remark that the soap concentration in the biodiesel-rich phase is higher when the reaction was carried out with ethanol. This is a major disadvantage, since in a conventional process the decantation is followed by neutralization. In this process stage, the soaps are hydrolyzed forming fatty acids, which are soluble in the biodiesel phase, thus leading to a product with acidity out of specification.
Table 3 Soaps, catalyst, alcohol and glycerine concentration at the end of the reaction. Reaction conditions: Temperature: 65 °C, 0.662 g H3CONa/100 mL oil, ethanol or methanol 25% v/v, agitation speed: 800 rpm, reaction time: 2 h.
Soaps biodiesel phase (goleate/kg) Soaps glycerine phase (goleate/kg) Catalyst biodiesel phase (gMeONa/kg) Catalyst glycerine phase (gMeONa/kg) Total glycerine (%) % Glycerine glycerine phase % Alcohol biodiesel phase % Alcohol glycerine phase (1) Methanol content.
Methanol
Ethanol
0.985 22.96 0.175 44.58 0.163 63 8.28 21.3
3.045 94.58 0.378 30.00 0.43 63 0.19 (1)/6.2 2.15(1)/22.4
Water in the methanol
Soaps in biodiesel (goleate/kg)
Soaps in glycerine (goleate/kg)
Catalyst in biodiesel (gMeONa/kg)
Catalyst in glycerine (gMeONa/kg)
345 ppm 1900 ppm
0.985 0.669
22.96 23.66
0.175 0.336
44.58 41.19
saponification rate observed with ethanol and with methanol, is due to differences between the properties of both alcohols. 3.2.3. Catalyst concentration and soaps formation The concentration of catalyst and soap during the transesterification with ethanol were determined as a function of time, both with sodium and potassium methoxides as catalysts. Results are shown in Fig. 2B. In both cases, the catalyst concentration decreased at the beginning of the reaction. However, this drop in concentration is much smaller than the one observed when using hydroxides as catalysts. The soap concentration obtained with methoxides is approximately 3.5 times smaller than that obtained with hydroxides. Therefore, it can be concluded that sodium and potassium methoxides display higher activity than the corresponding hydroxides, due to a lower extent of saponification that decreases the catalyst concentration. This difference is due to lower water content on the methoxides, as compared to the hydroxides. Fig. 3 shows the results obtained with methanol in an experiment carried out under the same experimental conditions as those used in Fig. 2B. It is very interesting to observe that the catalyst consumption is much lower with methanol, than in the case of using ethanol during the transesterification (compare Figs. 2B and 3). This is in agreement with above discussion, showing that the presence of ethanol favors the saponification reaction. This leads to a higher amount of soaps formed during the reaction, higher catalyst consumption, and a lower conversion when using ethanol. Note that the total glycerine content in the case of using methanol is 0.163 wt.%, while with ethanol under the same experimental conditions is 0.43%. The mass balance for the catalyst was carried out. It was found that in the case of the experiment with methanol, approximately 10% of the catalyst was consumed to produce soaps (Fig. 3), while in the case of the experiment with ethanol, approximately 25% of the catalyst was consumed to produce soaps (Fig. 2B).
60 8
50 40
6 MeONa/MeOH 4
30 20
2
10
Soap (g Oleate/Kg Sample)
3.2.2. Effect of water A possible explanation for the larger saponification rate observed in the process carried out with ethanol might be the higher water content of this alcohol as compared to the methanol. The ethanol used in these experiments contained 1150 ppm of water, while methanol contained 345 ppm. To verify this hypothesis, an experiment with methanol was carried out, adding water to the system in order to have the same amount of water as in the case of ethanol. Results are shown in Table 4. It can be observed that the amounts of soap formed with methanol in both experiments with different amounts of water are practically the same. Therefore, it can be concluded that the difference between the
Table 4 Soap formation during the transesterification with methanol. Effect of water content. Reaction conditions: Temperature: 65 °C, 0.662 g H3CONa/100 mL oil, methanol 25% v/v, agitation speed: 800 rpm, reaction time: 2 h.
Catalyst (gMeONa/KgSample)
the sodium and the potassium methoxides. Therefore, most of the following experiments were carried out using the former, since it is cheaper and of higher availability. In all the experiments shown in Table 2, there was a good and fast phase separation after the reaction. No soapy interphases or emulsions were observed. These are important observations from a technological point of view, since no additional steps have to be added to the process compared to the methanol-based process, in order to induce phase separation. Comparing the results shown in Tables 1 and 2, and the observations related to phase separation; it can be reasonably established that the intensive soap formation observed with the hydroxides as catalysts, is one of the reasons why no phase-separation occurs.
0
0 0
20
40
60
80 100 120 140 160 180 Time (min)
Fig. 3. Catalyst and soap evolution during transesterification using methanol and sodium methoxide.
6390
G. Mendow et al. / Bioresource Technology 102 (2011) 6385–6391
2- 40 g Methanol (25%) 3- 4.41 g MeONa
1- 180 g Oil
7- Acid Water
9- Water
Biodiesel 190.8 g
11- Biodiesel 179.5 g
4 W-1
D-1 R-1
Glycerol 28.1 g
W-2
8- Water/ Methanol
10- Acid Water
S-1
Fig. 4. Mass balance during transesterification of sunflower oil, using methanol and sodium methoxide as catalyst. Reaction conditions: temperature: 65 °C, 0.662 g H3CONa/ 100 mL oil, methanol 25% v/v, agitation speed: 800 rpm, reaction time: 2 h.
1- 180 g Oil
2- 39.5 g Ethanol (25%) 3- 4.41 g MeONa 7- Acid Water 9- Water Biodiesel 196.9 g
4 D-1
11- Biodiesel 179.5 g
W-1
W-2
8- Water/ Ethanol
R-1 Glycerol: 24.54 g
10-Acid Water
S-1
Fig. 5. Mass balance during transesterification of sunflower oil, using ethanol and sodium methoxide as catalyst. Reaction conditions: Temperature: 65 °C, 0.662 g H3CONa/ 100 mL oil, ethanol 25% v/v, agitation speed: 800 rpm, reaction time: 2 h.
Consequently, the lower conversion observed with ethanol is at least partly, due to this catalyst transformation in soaps. 3.2.4. Process yields The mass balances were carried out both with methanol and with ethanol, in order to determine the process yield, expressed as (biodiesel mass/oil mass). Figs. 4 and 5 show the results. In the case of the methyl ester production, the maximum theoretical yield is 100.45%, while in the case of the ethyl esters it is 105.22%. This is due to the difference in the molecular weight of the alcohols and, consequently, the higher molecular weight of the ethyl ester molecule, as compared to that of the methyl ester. Figs. 4 and 5 show that the experimental values were very similar in both cases, approximately 99.5%. Therefore, in the case of the ethanol, there is a larger loss of yield as compared to the maximum possible value, due to the higher amount of soaps formed when using ethanol. 3.2.5. Effect of reaction variables on triglycerides conversion The transesterification reaction using ethanol and sodium methoxide as catalyst was carried out under several experimental conditions. The results are shown in Table 2. The best conversion level obtained in these experiments corresponds to a total glycerine content of 0.258 wt.%, and was obtained using a high catalyst concentration. Even this, it was not possible to obtain the value required by the EN14214 standard, which is 0.25 wt.%. In the temperature range analyzed in this study (55–70 °C), the best result was obtained at the lowest temperature, i.e. 55 °C. The conversion shows an optimum value as a function of the catalyst concentration. At low concentration (0.662 g MeONa/ 100 mL oil) the total glycerine concentration was 0.43 wt.%, and at high catalyst concentration (1.417 g MeONa/100 mL oil) the total glycerine concentration was 0.396 wt.%. At low catalyst doses, the reaction rate is too low. At high catalyst concentration, the high rate of soaps formation leads to a reaction rate inhibition by these compounds and, consequently, a lower conversion was obtained,
as above discussed. The best conversion was obtained using a sodium methoxide concentration of 0.945 g/100 mL oil. In the case of the ethanol volume loaded into the reactor, it can be seen that as the ratio ethanol/oil (volume) is decreased, the conversion is better. As explained in the case of the results obtained with sodium hydroxide, this is due to a catalyst dilution effect.
3.3. Ethanol vs. methanol effects on the physico-chemical properties The data shown and discussed above can be summarized as follows. In all cases, the result obtained with ethanol is compared with the corresponding result obtained with methanol: - Saponification reactions occurs faster with ethanol. - The soap concentration is higher in the biodiesel phase with ethanol. - The catalyst concentration is higher in the biodiesel phase with ethanol. These results are in agreement with the liquid–liquid equilibria reported by other authors. It has been shown that the partial miscibility zone in the ternary diagram alcohol–glycerine–ester is considerably smaller in the case of ethanol compared to methanol (Andreatta et al., 2008; Liu et al., 2008; Zhou et al., 2006). This means that the concentration of glycerine in the biodiesel phase is higher in the case of using ethanol. There are no reports regarding the concentration of water, catalyst, or soaps in each phase after the reaction. Our results indicate that the higher mutual solubility of the compounds that form the system when using ethanol, also increases the solubility of catalyst, water, and soaps in the biodiesel phase. The consequence of this phenomenon is that the presence of higher concentration of catalyst and water in this phase accelerates the saponification reaction and the catalyst consumption. In the case of using methanol, the catalyst and the water are concentrated in the glycerine phase, and due to mass transfer limitations, the saponification occurs at a lower rate.
G. Mendow et al. / Bioresource Technology 102 (2011) 6385–6391
4. Conclusions In this work, we showed that the saponification reaction is much faster in the case of using ethanol during the transesterification, and because of this, higher catalyst consumption takes place and consequently, a lower reaction rate with a lower final conversion is observed. On the other hand, in the presence of methanol, the amount of soaps formed during the reaction is considerably lower, and this is another reason for the much higher activity observed when using this alcohol. Using methoxides as catalysts, phase separation is fast and clear and there is no formation of stable emulsions. References Andreatta, A.E., Casa´s, L.M., Hegel, P., Bottini, S.B., Brignole, E.A., 2008. Phase equilibria in ternary mixtures of methyl oleate, glycerol, and methanol. Ind. Eng. Chem. Res. 47, 5157–5164. Boehman, L., 2005. Biodiesel production and processing. Fuel Process. Technol. 86, 1057–1058. Bouaid, A., Martínez, M., Aracil, J., 2007. A comparative study of the production of ethyl esters from vegetable oils as a biodiesel fuel optimization by factorial design. Chem. Eng. J. 134, 93–99. Bouaid, A., Martinez, M., Aracil, J., 2009. Production of biodiesel from bioethanol and Brassica carinata oil: oxidation stability study. Biores. Technol. 100, 2234–2239. ˇ ernoch, M., Hájek, M., Skopal, F., 2010. Ethanolysis of rapeseed oil – distribution of C ethyl esters, glycerides and glycerol between ester and glycerol phases. Biores. Technol. 101, 2071–2075. Clark, S.J., Wagner, L., Schrock, M.D., Pienaar, P.G., 1984. Methyl and ethyl soybean esters as renewable fuels for diesel engines. J. Am. Oil Chem. Soc. 61, 1632– 1638.
6391
Encinar, J.M., González, J.F., Rodríguez-Reinares, A., 2007. Ethanolysis of used frying oil. Biodiesel preparation and characterization. Fuel Proc. Technol. 88, 513–522. Freedman, B., Pryde, E.H., Mounts, T.L., 1984. Variables affecting the yields of fatty esters from transesterified vegetable oils. J. Am. Oil Chem. Soc. 61, 1638–1643. Freedman, B., Butterfield, R.O., Pryde, E.H., 1986. Transesterification kinetics of soybean oil. J. Am. Oil Chem. Soc. 63, 1375–1380. Issariyakul, T., Kulkarni, M.G., Meher, L.C., Dalai, A.K., Bakhshi, N.N., 2008. Biodiesel production from mixtures of canola oil and used cooking oil. Chem. Eng. J. 140, 77. Liu, X., Piao, X., Wang, Y., Zhu, S., 2008. Liquid–liquid equilibrium for systems of (fatty acid ethyl esters + ethanol + soybean oil and fatty acid ethyl esters + ethanol + glycerol). J. Chem. Eng. Data 53, 359–362. Makareviciene, V., Janulis, P., 2003. Environmental effect of rapseed oil ethyl ester. Renewable Energy 28, 2395–2403. Marjanovic, A.V., Stamenkovic, O.S., Todorovic, Z.B., Lazic, M.L., Veljkovic, V.B., 2010. Kinetics of the base-catalyzed sunflower oil ethanolysis. Fuel 89, 665– 671. Pisarello, M.L., Dalla Costa, B.O., Veizaga, N.S., Querini, C.A., 2010. Volumetric method for free and total glycerin determination in biodiesel. Ind. Eng. Chem. Res. 49, 8935–8941. Schwad, A.W., Bagby, M.O., Freedman, B., 1987. Preparation and properties of diesel fuels from vegetable oils. Fuel 66, 1372–1378. Shay, E.G., 1993. Diesel fuel from vegetable oils: status and opportunities. Biomass and Bioenergy 4, 227–242. UNE-EN 14105, 2003. Determination of free and total glycerol and mono-, di-, and tri-glyceride contents. UNE-EN ISO 12937, 2001. Coulometric Karl Fischer titration method. Zhou, H., Lu, H., Liang, B., 2006. Solubility of multicomponent systems in the biodiesel production by transesterification of Jatropha Curcas L. Oil with methanol. J. Chem. Eng. Data 51, 1130–1135. Zhou, W., Konar, S.K., Boocock, D.G.B., 2003. Ethyl esters from the single-phase base-catalyzed ethanolysis of vegetable oils. J. Am. Oil Chem. Soc. 80, 367– 371.