- Email: [email protected]
Study of correlations between composition and physicochemical properties during methylic and ethylic biodiesel synthesis
Industrial Crops and Products 95 (2017) 18–26
Contents lists available at ScienceDirect
Industrial Crops and Products journal homepage: www.elsevier.com/locate/indcrop
Study of correlations between composition and physicochemical properties during methylic and ethylic biodiesel synthesis Filipe Lins da Silva, Jeilma Rodrigues do Nascimento, Lucas Natã de Melo, José Anderson Silva de Freitas, Janaína Heberle Bortoluzzi, Simoni Margareti Plentz Meneghetti ∗ Grupo de Catálise e Reatividade Química, Instituto de Química e Biotecnologia, Universidade Federal de Alagoas (GCAR/IQB/UFAL), Brazil
a r t i c l e
i n f o
Article history: Received 10 May 2016 Received in revised form 6 September 2016 Accepted 23 September 2016 Keywords: Reaction parameters Viscosity Kinematic viscosity Transesterification
a b s t r a c t In this work, correlations between the physicochemical parameters (density and kinematic viscosity) and the composition of various samples, obtained during the biodiesel production process of methylic and ethylic soybean biodiesel, were investigated. The results point out that the composition of the medium in terms of remaining catalyst and its residues, alcohol and water content, among others variables, has no significant influence on the physicochemical properties studied. These findings are mainly influenced by the composition of the medium in terms of FAAE, TAG, DAG and MAG. © 2016 Elsevier B.V. All rights reserved.
1. Introduction The raw fat materials used for biodiesel production are mostly formed by triacylglycerides (TAG) that are converted to fatty acid alkyl monoesters (FAAE) and glycerol during the transesterification reaction. The transesterification process consists of three consecutive and reversible reactions that form diacylglycerides (DAG) and monoacylglycerides (MAG) as intermediate products (Suarez et al., 2007; Schwab et al., 1987). Therefore, during biodiesel production, it is possible to detect the presence of unconverted TAG, DAG, MAG, alcohol, and the catalyst and its residues in the reaction medium in addition to FAAE and glycerol. Furthermore, if transesterification occurs using hydroxides or alkoxides as catalysts, water molecules are formed, which may also be present in the raw materials that are used. The presence of water causes the hydrolysis of TAG, DAG and MAG and leads to the formation of fatty acids (FA). The determination of the amount
Abbreviations: FAAE, fatty acid alkyl monoesters; FAME, fatty acid methyl monoesters; TAG, triacylglycerides; DAG, diacylglycerides; MAG, monoacylglycerides; HPLC, high performance liquid chromatography; pH, potencial of hydrogen; FA, fatty acids. ∗ Corresponding author at: Instituto de Química e Biotecnologia, Universidade Federal de Alagoas, Av. Lourival de Melo Mota, s/n—Cidade Universitária, 57072-970 Maceió, AL, Brazil. E-mail address: [email protected] (S.M.P. Meneghetti). http://dx.doi.org/10.1016/j.indcrop.2016.09.053 0926-6690/© 2016 Elsevier B.V. All rights reserved.
of each substance during the reaction process and its influence on the physicochemical properties is very important for the development and optimization of processes on both pilot and industrial scales. Two situations can be observed during biodiesel production after a pre-determined reaction time and stirring ends: the formation of two phases or no phases. When two phases form, the top (oily phase) contains unconverted TAG, FAAE, DAG and MAG. In the lower stage (glycerol phase), it is possible to find glycerol, the catalyst and its residues, water formed or assigned to the process by the reactants, excess alcohol, and even smaller portions of esters and fatty acids. When there is no phase separation, all these substances are present as an emulsion, and it is possible to observe the formation of gels and soaps. It has been established that the physicochemical properties of biodiesel are essentially determined by the composition of the reaction medium, and several studies revealed relationships between the properties and compositional features by studying mixtures of pure compounds (Knothe et al., 2005; Chuck et al., 2009; Su et al., 2011; Najafabadi et al., 2012). In this work, physicochemical parameters were determined to provide information that can be applied to the design and improvement of the process mostly used to obtain biodiesel (transesterification) and, consequently, to understanding the reaction medium behavior. To this end, samples that simulated actual reaction conditions were obtained and characterized by determining the amount of FAME (for methylic biodiesel) or FAEE
F.L. da Silva et al. / Industrial Crops and Products 95 (2017) 18–26
19
Table 1 Composition of the medium (FAME, TAG, DAG and MAG) and physicochemical properties (phase separation). FAMEa (wt.%)
MAGa (wt.%)
DAGa (wt.%)
TAGa (wt.%)
Viscosity (mm2 s−1 )
Density (kg m−3 )
Water (mg kg−1 )
Acid number (mg KOH g−1 )
pH
35 45 49 50 51 52 54 55 55 57 59 60 60 79 92 94
6 5 7 4 3 3 4 13 5 6 5 3 6 2 5 5
20 12 12 12 11 11 12 7 10 11 11 11 11 5 1 1
39 34 35 34 35 34 31 25 30 27 26 27 23 14 2 0
17.56 12.02 12.01 10.02 8.50 10.52 8.64 8.79 9.48 8.84 9.17 9.57 9.72 5.71 4.07 4.10
0.920 0.907 0.909 0.897 0.899 0.897 0.896 0.903 0.899 0.904 0.903 0.899 0.902 0.889 0.881 0.880
10098 1866
3.87 0.50
10.3 9.4
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
7857 3409 5789 7048 10312 4726 1355 650 879
0.65 0.45 0.76 3.19 1.60 0.62 0.58 0.22 0.63
10.4 9.3 10.1 9.8 9.8
a b
b
8.2 10.0 10.0
Determined in purified oily phase. No determined.
(for ethylic biodiesel), MAG, DAG and TAG present. The density, kinematic viscosity, pH, acid number, water and alcohol contents were also determined. More specifically, the physicochemical parameters were obtained during the biodiesel production process of methylic and ethylic soybean biodiesel by alkaline transesterification to: (i) simulate the reaction process on a laboratory scale; (ii) analyze the density, kinematic viscosity, pH and acid number; (iii) determine the composition of the medium in terms of water, alcohol, FAAE, TAG, DAG and MAG content, and (iv) define possible correlations between the physicochemical parameters for the various samples obtained during the reaction process and their composition.
2. Material and methods 2.1. Transesterification process To obtain various samples that represent the process, methylic and ethylic transesterifications of soybean oil were performed in several reaction conditions. Sodium hydroxide was used as the catalyst (see Tables S1 and S2 in Supplementary material). In a typical experiment, a 2.0-L glass reactor that was fit to a mechanical stirring, heating and a condenser system was employed. After the established reaction time, the mixture was transferred to a 2.0-L separatory funnel, where it remained at rest for 3 h at 25 ◦ C. After this time, two situations were observed: (i) a two phases formed or (ii) no phases formed. The samples were stored to determine their physicochemical properties. To determine the composition of the TAG, DAG, MAG and FAAE (by HPLC) contents as well as the acid number, a sample was purified from the oily phase (phase separation) or from the mixture (no phase formation) by neutralization with 5% phosphoric acid (v/v), followed by washing with a brine until it reach a pH of 7.0.
2.2. Determination of physicochemical parameters and composition of samples The properties were determided according to standard method: kinematic viscosity (mm2 s−1 ) by ASTM D445-15a, acid number (mg KOH g−1 oil) and pH by ASTM D664-11a, density at 20 ◦ C (kg m−3 ) by ASTM D4052-15 and water content using ASTM D630407. The alcohol content was determined by a gravimetric test; 5 g of the sample was placed into a rotaevaporator with a water aspirator (vacuum system) for 45 min at 25 ◦ C, and the mass loss (determined
until weight constant) was calculated. All results were calculated based on the average of three samples. The amount of FAAE, TAG, DAG and MAG was determined by HPLC using the method described in the literature (Carvalho et al., 2012) and one example is presented on Supplementary material (see Fig. S1). The repetibility of the method was assessed by having one operator analyzing ten times the same sample on the same day, using the same apparatus. The standard deviation was 0.02 mm2 s−1 for kinematic viscosity, 0.001 kg m−3 for density, 0.03 mg KOH g−1 oil for acid number, 10 mg kg−1 for water and 0.5% for alcohol content. In the case of FAAE% the value was 4% and for TAG, DAG and MAG was 2%, respectively. The mathematical tool used was Excel – Microsoft Office Professional Plus 2010, version 14.0.4760.1000. In order to outline possible correlations between the physicochemical parameters for the various samples (simulating reaction process) and their composition, regression analysis were calculated using Origin® , version 6.0, Microcal (TM) Software, Inc. The values of coefficient of correlation (r) and coefficient of determination (r2 ) are presented in Tables S3 in Supplementary material. Whenever the coefficient of correlation value is found between ±0.500 and ±0.749 the correlation is considered moderate and, if between ±0.750 and ±1.000, then it is an indicative of strong correlation (Colton et al., 1982). In the case that is possible to establish correlation, the linear regression equations are presented (Table S3). 3. Results and discussion 3.1. Correlations between the composition of the reaction medium and physicochemical parameters during methylic soybean biodiesel (FAME) obtention Several samples that exhibited FAME ratios from 29 to 94% were obtained to represent the reaction process. It is important to highlight that the purpose of this work was not to determine the influence of reaction conditions on the amount of chemical species formed (FAME, TAG, DAG, MAG, water, etc.) because this issue has already been exhaustively discussed in the literature, and many reviews address this theme (Datta and Mandal, 2016; Hoekman et al., 2012; Knothe, 2005). As was mentioned in the experimental section of this paper, some reaction conditions and results in terms of the composition are presented in Supplementary material. Tables 1 and 2 presents the results, which were calculated from the average of a triplicate analysis. These characterizations were
20
F.L. da Silva et al. / Industrial Crops and Products 95 (2017) 18–26
Table 2 Composition of the medium (FAME, TAG, DAG and MAG) and physicochemical properties (no phase separation). FAMEa (wt.%) MAGa (wt.%)
DAGa (wt.%)
TAGa (wt.%)
Viscosity (mm2 s−1 )
Density (kg m−3 )
Water (mg kg−1 )
Alcohol (wt.%) Acid number (mg pH KOH g−1 )
29 29 29 30 32 34 49 56 69
19 20 16 20 19
47 44 53 44 45
18.62 19.14 20.39 17.68 18.18 19.41 16.62 8.08 6.22
0.917 0.918 0.915 0.914 0.915 0.921 0.915 0.894 0.890
15057 11496 5595 6141 10654 2874 2750 8021
10.0 5.0 0.7 0.1 0.2
b
b
a b
4 8 2 6 4 b
b
b
7 3 9
12 10 8
32 31 14
b
0.2 0.6
2.77 2.93 3.24 1.88 2.41 0.78 0.63 0.84 1.035
10.2 10.1 9.3 10.0 10.4 b
9.5 10.7 10.3
Determined in purified oily phase. No determined.
performed in the oily phase, i.e., when there was phase separation (Table 1). Due the formation of persistent emulsion in the glycerol phase, when phase separation was observed, it was not possible to determine the physicochemical parameters in that phase. In the case of no phase separation occurred, the analyzes were performed in the reaction mixture (Table 2). As was already mentioned, the aliquots were purified in all cases (from oily phase or mixture) to determine the FAME, TAG, DAG, MAG content, and acid number.
3.1.1. Methylic soybean biodiesel: composition of a reaction medium (TAG, DAG, MAG and FAME) As previously mentioned, the transesterification process consists of three consecutive and reversible reactions, in which TAG are converted into FAAE, DAG and MAG (Suarez et al., 2007; Schwab et al., 1987). Fig. 1 presents the composition of the reaction medium (unconverted TAG, DAG and MAG) as a function of the FAME content when phase separation (Fig. 1A) or no phase separation (Fig. 1B) were observed. In the both cases, the same tendency was observed, and as expected, as FAME formed, the TAG content decreased; DAG and MAG are continuously formed and consumed, and their quantity tends to decrease during the process. When the FAME amount was between 35 and 94% (Table 1 and Fig. 1A), the formation of the phases was observed after the procedure described in the experimental section. However, in some cases (Table 1), no phase formation was observed (FAME from 29
to 56%). This lack of separation can be related to the formation of a stable emulsion that was not broken during the separation procedure adopted, possibly due to the large amount of DAG and MAG in the medium.
3.1.2. Methylic soybean biodiesel: study of correlations for the oily phase (phase separation) The graphics showing possible correlations between the kinematic viscosity and density with the amount of FAME, TAG, DAG or MAG, are shown in Fig. 2(A–D), for the oily phase samples. From Fig. 2A, it is possible to infer that the evolution of the viscosity and density values was inversely proportional to the amount of FAME (i.e., the higher the conversion to monoesters, the lower the kinematic viscosity of the oily phase), showing a strong degree of correlation (see Table S3 in Supplementary material). This behavior can be related to the intermolecular interactions present in soybean oil and biodiesel, which have a significant influence on the viscosity and density (Knothe et al., 2005). Subsequently, the correlations between the TAG, DAG and MAG contents to the viscosity and density were also investigated (Fig. 2) for the oily phase (phase separation). It is possible to note (Fig. 2B) that the viscosity evolution is directly proportional to the remaining TAG in the medium. A similar behavior was observed for the density values and is consistent with the data presented in Fig. 2A, which show that the viscosity of the FAME was lower than that of the TAG. The DAG content in the
Fig. 1. Composition of the reaction medium (TAG, DAG and MAG) as a function of FAME content: A) phase separation or B) no phase separation.
F.L. da Silva et al. / Industrial Crops and Products 95 (2017) 18–26
21
Fig. 2. A) Kinematic viscosity and density X FAME; B) Kinematic viscosity and density X TAG; C) Kinematic viscosity and density X DAG; and D) Kinematic viscosity and density X MAG for samples acquired during methylic biodiesel obtention process (phase separation).
Fig. 3. A) Water amount (mg/kg) as a function of FAME (%) content and B) Acid number (mg KOH g−1 and pH as a function of FAME (%) content.
22
F.L. da Silva et al. / Industrial Crops and Products 95 (2017) 18–26
Fig. 4. A) Viscosity and density X FAME; B) Viscosity and density X TAG; C) Viscosity and density X DAG; and D) Viscosity and density X MAG for samples acquired during methylic biodiesel obtention process (no phase separation).
Table 3 Molar distribution of MeOH in the reaction medium (phase formation). Molar ratio oil:MeOH
FAMEa (wt.%)
MeOH linked (esters)b (wt.%)
Free MeOH in oily phasec (wt.%)
Free MeOH in glycerol phased (wt.%)
1:2 1:2 1:3 1:3 1:6 1:6
45 55 69 79 92 94
56.4 69.0 57.7 66.1 38.5 39.3
0.5 0.4 0.8 0.2 1.3 2.2
43.1 30.6 41.5 33.7 60.2 58.5
a b c d
Determined in the purified oily phase. Calculated considering FAME, MAG and DAG amount. Determined experimentally. Calculated by the difference between MeOH linked in the esters and free MeOH in the oily phase.
medium (Fig. 2C) has an important influence on the development of the viscosity and density, which is analogous to that observed when analyzing the TAG correlation with the same properties. The same argument, with respect to intermolecular interactions between the species present, is valid here. In all cases a high degree of correlation was detected. However, it is not possible to establish a reasonable trend of the viscosity and density to the MAG content (Fig. 2D) and it can be verified in Table S3 (Supplementary material), by the low values
of coefficient of correlation. The reason for such behavior can be associated with the fact that the MAG molecule has two hydroxyl groups and a slight steric hindrance. As a consequence, hydrogen bonds are favored, and the importance of Van der Waals interactions (that are prevalent for other compounds, such as DAG and TAG) are minimized. Therefore, a small variation in the levels of MAG will not cause a substantial variation in the physicochemical properties.
F.L. da Silva et al. / Industrial Crops and Products 95 (2017) 18–26
23
Fig. 5. Acid number (mg KOH g−1 ) and pH as function of FAME content.
Table 4 Composition of the medium (FAEE, TAG, DAG and MAG) and physicochemical properties (no phase separation). FAEEa (wt.%)
MAGa (wt.%)
DAGa (wt.%)
TAGa (wt.%)
EtOH (wt.%)
Viscosity (mm2 s−1 )
Density (kg m−3 )
Acid number (mg KOH g−1 )
pH
10 12 16 19 19 20 23 23 26 28 30 32 36 38 40 42 43 49 51 52 54 56 62 63 66 67 68 68 68 75 77 78
2 3
12 12
76 74
b
b
b
b
b
b
b
b
b
b
b
b
b
6
19
52
b
b
b
b
b
b
6 10 10 10 12 12 22 11 15 12 11 12 26
18 22 22 18 22 22 21 17 18 19 19 20 13
48 38 36 36 28 26 16 29 18 18 17 14 5
1.3 5.1 34 0.5 – 0.4 0.7 30.9
b
b
b
b
12 20 12 10 13 9 13 13
16 11 14 14 13 11 9 7
9 3 7 8 6 12 3 3
1.6
b
b
b
1.5
19.4 19.5 24.4 23.3 23.1 20.9 18.9 23.1 22.3 21.6 13.3 13.1 14.1 12.5 11.3 11.4 11.3 10.6 12.7 11.5 10.7 12.8 11.2 10.1 11.9 6.3 10.7 8.3 9.0 6.2 6.9 7.1
914 913 917 915 913 911 909 917 916 919 905 905 907 902 904 904 907 907 907 908 905 906 905 891 891 890 902 890 901 893 875 892
6.12 9.85 7.43 6.44 4.90 1.17 10.81 6.48 7.19 19.93 16.90 11.51 1.52 20.34 17.25 22.34 20.04 17.16 24.58 43.05 30.06 21.24 10.51 8.57 29.10 36.84 13.19 15.40 23.91 11.07 11.61 6.80
8.17 8.00 10.03 9.90 9.86 9.74 7.91 9.40 9.70 7.42 8.13 9.19 7.63 9.27 7.17 9.24 7.85 9.40 9.39 8.00 8.08 7.43 9.24 7.53 8.38 8.28 8.11 7.95 8.65 9.79 9.49 9.33
a b
b
b b
2.9 b b
1.1 b b
2.6 b
0.6 b
b b b b b
3.2 b
Determined in purified oily phase. No determined.
Complementarily, correlations to the water content, acid number and pH, and FAME (%) were studied (Fig. 3A and B, respectively). Regarding the water content, it is possible to observe a decrease in the value as the FAME content increases in the medium (Fig. 3A),
but a it not possible to stablish a correlation (see Table S4 in Supplementary material). This behavior can be easily explained because when FAME is being formed, more glycerol is generated. During
24
F.L. da Silva et al. / Industrial Crops and Products 95 (2017) 18–26
Fig. 6. Composition of a reaction medium (TAG, DAG and MAG) as a function of FAEE content (no phase separation).
the phase separation step, due to the polarity of the medium, it is expected that the water remains in the glycerol phase. From Fig. 3B, we can see that there is no a relationship between the acid number and pH with the FAME content present in oily phase. This result demonstrates that these parameters indicate the efficiency of the biodiesel purification process. Regarding the determination of the alcohol content, the main objective was to evaluate the distribution after phase separation (glycerol and oily). Thus, from its experimental determination in the oily phase, it was possible to estimate their distribution, considering the various oil:alcohol molar ratios employed, as shown in Table 3. From these results, it is possible to observe that a large amount of free alcohol (unreacted) remains in the glycerol phase. This trend was expected due to the chemical nature of the species involved because there is a high affinity between glycerol and methanol, as they are both alcohols and with a similar polarity (Carvalho et al., 2012). As already mentioned, due to persistent suspension formation, the samples of glycerol (in which there was phase separation) were not analyzed. 3.1.3. Methylic soybean biodiesel: study of correlations of the mixture (no phase separation) The graphics exposing probable correlations regarding the kinematic viscosity and density with the amount of FAME, TAG, DAG or MAG are shown in Fig. 4(A–D). As shown in Fig. 4, the correlation between the density and kinematic viscosity with the formed FAME content was similar to that observed for the oily phase when the methylic biodiesel process was investigated. In this case, it should be noted that even though the physicochemical properties were determined in the mixture without any purification, the same trends observed for the samples in which there was phase separation can be observed, and similar interpretations are valid (Table S3, in Supplementary material, to examine coefficients of correlation and determination). Regarding the investigation of the correlation between the amount of FAME to the acid number and pH (Fig. 5), it was not
possible to establish trend, as was observed in the case of the oily phase (phase separation). The analysis of the glycerol content by the European Standard EN 14105, which uses gas chromatography, was not used in this study because the samples that represented the reaction process contained high amounts of glycerol. 3.2. Correlations between the composition of the reaction medium and physicochemical parameters, during ethylic soybean biodiesel obtention In the study of ethylic biodiesel, which was similar to the study of methyl biodiesel, several samples were obtained that exhibited different compositions to represent the process. Tables 4 and 5 present the medium composition (in terms of TAG, DAG, MAG and FAEE) and the respective physicochemical characterizations. In this case, no phase separation was observed for the majority of reactions. This result was expected because the formation of stable emulsions during ethanolysis is related to the presence of high amounts of DAG and MAG (Zhou et al., 2003). An accurate analysis of the results show that, in fact, the phase separation occurred when the FAEE content exceeded 76%, a situation in which there is less TAG, DAG and MAG in the reaction medium. 3.2.1. Ethylic soybean biodiesel: composition of a reaction medium (TAG, DAG, MAG and FAEE) and a study of the mixture (no phase separation) Due the small number of samples that showed phase separation (Table 5), it was not possible to establish any trends. The results and discussion will be focused on the samples in which no phase separation was observed (Table 4). Fig. 6 presents the composition of the reaction medium (unconverted TAG, DAG and MAG) as function of the FAME content when no phase separation was observed. As was discussed for methylic biodiesel, as FAME formed, the TAG content diminished. DAG and MAG were formed and consumed, and their amounts tended to decrease as the process proceeded.
F.L. da Silva et al. / Industrial Crops and Products 95 (2017) 18–26
25
Table 5 Composition of the medium (FAEE, TAG, DAG and MAG) and physicochemical properties (phase separation).
b
FAEEa (wt.%)
MAGa (wt.%)
DAGa (wt.%)
TAGa (wt.%)
EtOH (wt.%)
Viscosity (mm2 s−1 )
Density (kg m−3 )
Acid number (mg KOH g−1 )
pH
76 79 80 83 84 86
15 10 9 7 3 4
8 11 11 9 10 9
1 0 1 1 3 1
2.3 7.0 12.5 7.7 2.5 3.0
4.3 5.7 5.5 5.3 4.7 4.3
894 885 883 882 876 880
15.96 3.57 3.78 4.34 17.73 15.12
8.56 9.90 9.70 9.40 8.56 7.98
No determined. a Determined in purified oily phase.
Fig. 7. A) Kinematic viscosity and density X FAEE; B) Kinematic viscosity and density X TAG; C) Kinematic viscosity and density X DAG; and D) Kinematic viscosity and density X MAG, for samples acquired during ethylic biodiesel obtention process (no phase separation).
To understand the physicochemical parameters that are related to the medium composition, possible correlations were investigated (Fig. 7). The correlations of the FAEE and TAG amounts with the kinematic viscosity and density values (Fig. 7A and B) followed the same trend observed in the case of methylic biodiesel (Figs. 2 and 4) and, high degree of correlation could be stablished in most cases. However, it was not possible to establish correlations or trends between the DAG and MAG levels and the evaluated prop-
erties (Fig. 7C and D). Most likely, this behavior can be related to the complexity of the mixture. As was with observed with the methylic biodiesel, it was not possible to establish correlations between the FAEE amount and the acid number or pH (Fig. 8), which confirmed that these parameters indicate whether the purification step was effective.
26
F.L. da Silva et al. / Industrial Crops and Products 95 (2017) 18–26
Fig. 8. Acid number and pH as a function of FAEE content.
4. Conclusions
References
In this study, it was possible to establish some correlations between physicochemical parameters (density and kinematic viscosity) and the composition of the medium during the reaction of the transesterification of soybean oil with methanol and ethanol. An analysis of the results allows us to infer that the composition of the medium in terms of remaining catalyst and its residues, alcohol and water content, among others variables, has no significant influence on the physicochemical properties studied. These findings are mainly influenced by the composition of the medium in terms of FAAE, TAG, DAG and MAG.
Carvalho, M.S., Mendonc¸a, M.A., Pinho, D.M.M., Resck, I.S., Suarez, P.A.Z., 2012. Chromatographic analyses of fatty acid methyl esters by HPLC-UV and GC-FID. J. Braz. Chem. Soc. 23, 763–769. Chuck, C.J., Bannister, C.D., Hawley, J.G., Davidson, M.G., La Bruna, I., Paine, A., 2009. Predictive model to assess the molecular structure of biodiesel fuel. Energy Fuels 23 (4), 2290–2294. Colton, T., Freedman, L.S., Johnson, A.L. (Eds.), 1982. John Wiley and Sons, New York. Datta, A., Mandal, B.K., 2016. A comprehensive review of biodiesel as analternative fuel for compression ignition engine. Renew. Sustain. Energy Rev. 57, 799–821. Hoekman, S.K., Broch, A., Robbins, C., Ceniceros, E., Natarajan, M., 2012. Review of biodiesel composition, properties, and specifications. Renew. Sustain. Energy Rev. 16, 143–169. Knothe, G., Van Gerpen, J., Krahl, J., In: Knothe, G., Van Gerpen, J., Krahl, J. (Eds.), 2005. AOCS Press, Urbana, IL. Knothe, G., 2005. Dependence of biodiesel fuel properties on thestructure of fatty acid alkyl esters. Fuel Process. Technol. 86, 1059–1070. Najafabadi, H.A., Pazuki, G., Vossoughi, M., 2012. Estimation of biodiesel physical properties using local composition based models. Ind. Eng. Chem. Res. 51, 13518–13526. Schwab, A.W., Bagby, M.O., Freedman, B., 1987. Preparation and properties of diesel fuels from vegetable oils. Fuel 66, 1372–1378. Su, Y.C., Liu, Y.A., Diaz Tovar, C.A., Gani, R., 2011. Selection of prediction methods for thermophysical properties for process modeling and product design of biodiesel manufacturing. Ind. Eng. Chem. Res. 50, 6809–6836. Suarez, P.A.Z., Meneghetti, S.M.P., Meneghetti, M.R., Wolf, C.R., 2007. Transformation of triglycerides into fuels, polymers and chemicals: some applications of catalysis in oleochemistry. Quim. Nova 30, 667–676. Zhou, W., Konar, S.K., Boocock, D.G.B., 2003. Ethyl esters from the single phase base-catalyzed ethanolysis of vegetable oils. JAOCS 80 (4), 367–371.
Acknowledgements Financial support from Brazilian research founding agencies, such as Research and Projects Financing (FINEP), National Counsel of Technological and Scientific Development (CNPq), Brazilian Federal Agency for the Support and Evaluation of Graduate Education (CAPES), and Alagoas Research Support Foundation (FAPEAL) are grateful acknowledge. SMPM thank CNPq for research fellowships and FLS, JRN, LNM, JASF thanks CNPq for fellowship. The authors are indebt with Ministry of Science, Technology and Innovation (MCTI) and CNPq for FISQUIBIODIESEL project. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.indcrop.2016.09. 053.
Contents lists available at ScienceDirect
Industrial Crops and Products journal homepage: www.elsevier.com/locate/indcrop
Study of correlations between composition and physicochemical properties during methylic and ethylic biodiesel synthesis Filipe Lins da Silva, Jeilma Rodrigues do Nascimento, Lucas Natã de Melo, José Anderson Silva de Freitas, Janaína Heberle Bortoluzzi, Simoni Margareti Plentz Meneghetti ∗ Grupo de Catálise e Reatividade Química, Instituto de Química e Biotecnologia, Universidade Federal de Alagoas (GCAR/IQB/UFAL), Brazil
a r t i c l e
i n f o
Article history: Received 10 May 2016 Received in revised form 6 September 2016 Accepted 23 September 2016 Keywords: Reaction parameters Viscosity Kinematic viscosity Transesterification
a b s t r a c t In this work, correlations between the physicochemical parameters (density and kinematic viscosity) and the composition of various samples, obtained during the biodiesel production process of methylic and ethylic soybean biodiesel, were investigated. The results point out that the composition of the medium in terms of remaining catalyst and its residues, alcohol and water content, among others variables, has no significant influence on the physicochemical properties studied. These findings are mainly influenced by the composition of the medium in terms of FAAE, TAG, DAG and MAG. © 2016 Elsevier B.V. All rights reserved.
1. Introduction The raw fat materials used for biodiesel production are mostly formed by triacylglycerides (TAG) that are converted to fatty acid alkyl monoesters (FAAE) and glycerol during the transesterification reaction. The transesterification process consists of three consecutive and reversible reactions that form diacylglycerides (DAG) and monoacylglycerides (MAG) as intermediate products (Suarez et al., 2007; Schwab et al., 1987). Therefore, during biodiesel production, it is possible to detect the presence of unconverted TAG, DAG, MAG, alcohol, and the catalyst and its residues in the reaction medium in addition to FAAE and glycerol. Furthermore, if transesterification occurs using hydroxides or alkoxides as catalysts, water molecules are formed, which may also be present in the raw materials that are used. The presence of water causes the hydrolysis of TAG, DAG and MAG and leads to the formation of fatty acids (FA). The determination of the amount
Abbreviations: FAAE, fatty acid alkyl monoesters; FAME, fatty acid methyl monoesters; TAG, triacylglycerides; DAG, diacylglycerides; MAG, monoacylglycerides; HPLC, high performance liquid chromatography; pH, potencial of hydrogen; FA, fatty acids. ∗ Corresponding author at: Instituto de Química e Biotecnologia, Universidade Federal de Alagoas, Av. Lourival de Melo Mota, s/n—Cidade Universitária, 57072-970 Maceió, AL, Brazil. E-mail address: [email protected] (S.M.P. Meneghetti). http://dx.doi.org/10.1016/j.indcrop.2016.09.053 0926-6690/© 2016 Elsevier B.V. All rights reserved.
of each substance during the reaction process and its influence on the physicochemical properties is very important for the development and optimization of processes on both pilot and industrial scales. Two situations can be observed during biodiesel production after a pre-determined reaction time and stirring ends: the formation of two phases or no phases. When two phases form, the top (oily phase) contains unconverted TAG, FAAE, DAG and MAG. In the lower stage (glycerol phase), it is possible to find glycerol, the catalyst and its residues, water formed or assigned to the process by the reactants, excess alcohol, and even smaller portions of esters and fatty acids. When there is no phase separation, all these substances are present as an emulsion, and it is possible to observe the formation of gels and soaps. It has been established that the physicochemical properties of biodiesel are essentially determined by the composition of the reaction medium, and several studies revealed relationships between the properties and compositional features by studying mixtures of pure compounds (Knothe et al., 2005; Chuck et al., 2009; Su et al., 2011; Najafabadi et al., 2012). In this work, physicochemical parameters were determined to provide information that can be applied to the design and improvement of the process mostly used to obtain biodiesel (transesterification) and, consequently, to understanding the reaction medium behavior. To this end, samples that simulated actual reaction conditions were obtained and characterized by determining the amount of FAME (for methylic biodiesel) or FAEE
F.L. da Silva et al. / Industrial Crops and Products 95 (2017) 18–26
19
Table 1 Composition of the medium (FAME, TAG, DAG and MAG) and physicochemical properties (phase separation). FAMEa (wt.%)
MAGa (wt.%)
DAGa (wt.%)
TAGa (wt.%)
Viscosity (mm2 s−1 )
Density (kg m−3 )
Water (mg kg−1 )
Acid number (mg KOH g−1 )
pH
35 45 49 50 51 52 54 55 55 57 59 60 60 79 92 94
6 5 7 4 3 3 4 13 5 6 5 3 6 2 5 5
20 12 12 12 11 11 12 7 10 11 11 11 11 5 1 1
39 34 35 34 35 34 31 25 30 27 26 27 23 14 2 0
17.56 12.02 12.01 10.02 8.50 10.52 8.64 8.79 9.48 8.84 9.17 9.57 9.72 5.71 4.07 4.10
0.920 0.907 0.909 0.897 0.899 0.897 0.896 0.903 0.899 0.904 0.903 0.899 0.902 0.889 0.881 0.880
10098 1866
3.87 0.50
10.3 9.4
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
7857 3409 5789 7048 10312 4726 1355 650 879
0.65 0.45 0.76 3.19 1.60 0.62 0.58 0.22 0.63
10.4 9.3 10.1 9.8 9.8
a b
b
8.2 10.0 10.0
Determined in purified oily phase. No determined.
(for ethylic biodiesel), MAG, DAG and TAG present. The density, kinematic viscosity, pH, acid number, water and alcohol contents were also determined. More specifically, the physicochemical parameters were obtained during the biodiesel production process of methylic and ethylic soybean biodiesel by alkaline transesterification to: (i) simulate the reaction process on a laboratory scale; (ii) analyze the density, kinematic viscosity, pH and acid number; (iii) determine the composition of the medium in terms of water, alcohol, FAAE, TAG, DAG and MAG content, and (iv) define possible correlations between the physicochemical parameters for the various samples obtained during the reaction process and their composition.
2. Material and methods 2.1. Transesterification process To obtain various samples that represent the process, methylic and ethylic transesterifications of soybean oil were performed in several reaction conditions. Sodium hydroxide was used as the catalyst (see Tables S1 and S2 in Supplementary material). In a typical experiment, a 2.0-L glass reactor that was fit to a mechanical stirring, heating and a condenser system was employed. After the established reaction time, the mixture was transferred to a 2.0-L separatory funnel, where it remained at rest for 3 h at 25 ◦ C. After this time, two situations were observed: (i) a two phases formed or (ii) no phases formed. The samples were stored to determine their physicochemical properties. To determine the composition of the TAG, DAG, MAG and FAAE (by HPLC) contents as well as the acid number, a sample was purified from the oily phase (phase separation) or from the mixture (no phase formation) by neutralization with 5% phosphoric acid (v/v), followed by washing with a brine until it reach a pH of 7.0.
2.2. Determination of physicochemical parameters and composition of samples The properties were determided according to standard method: kinematic viscosity (mm2 s−1 ) by ASTM D445-15a, acid number (mg KOH g−1 oil) and pH by ASTM D664-11a, density at 20 ◦ C (kg m−3 ) by ASTM D4052-15 and water content using ASTM D630407. The alcohol content was determined by a gravimetric test; 5 g of the sample was placed into a rotaevaporator with a water aspirator (vacuum system) for 45 min at 25 ◦ C, and the mass loss (determined
until weight constant) was calculated. All results were calculated based on the average of three samples. The amount of FAAE, TAG, DAG and MAG was determined by HPLC using the method described in the literature (Carvalho et al., 2012) and one example is presented on Supplementary material (see Fig. S1). The repetibility of the method was assessed by having one operator analyzing ten times the same sample on the same day, using the same apparatus. The standard deviation was 0.02 mm2 s−1 for kinematic viscosity, 0.001 kg m−3 for density, 0.03 mg KOH g−1 oil for acid number, 10 mg kg−1 for water and 0.5% for alcohol content. In the case of FAAE% the value was 4% and for TAG, DAG and MAG was 2%, respectively. The mathematical tool used was Excel – Microsoft Office Professional Plus 2010, version 14.0.4760.1000. In order to outline possible correlations between the physicochemical parameters for the various samples (simulating reaction process) and their composition, regression analysis were calculated using Origin® , version 6.0, Microcal (TM) Software, Inc. The values of coefficient of correlation (r) and coefficient of determination (r2 ) are presented in Tables S3 in Supplementary material. Whenever the coefficient of correlation value is found between ±0.500 and ±0.749 the correlation is considered moderate and, if between ±0.750 and ±1.000, then it is an indicative of strong correlation (Colton et al., 1982). In the case that is possible to establish correlation, the linear regression equations are presented (Table S3). 3. Results and discussion 3.1. Correlations between the composition of the reaction medium and physicochemical parameters during methylic soybean biodiesel (FAME) obtention Several samples that exhibited FAME ratios from 29 to 94% were obtained to represent the reaction process. It is important to highlight that the purpose of this work was not to determine the influence of reaction conditions on the amount of chemical species formed (FAME, TAG, DAG, MAG, water, etc.) because this issue has already been exhaustively discussed in the literature, and many reviews address this theme (Datta and Mandal, 2016; Hoekman et al., 2012; Knothe, 2005). As was mentioned in the experimental section of this paper, some reaction conditions and results in terms of the composition are presented in Supplementary material. Tables 1 and 2 presents the results, which were calculated from the average of a triplicate analysis. These characterizations were
20
F.L. da Silva et al. / Industrial Crops and Products 95 (2017) 18–26
Table 2 Composition of the medium (FAME, TAG, DAG and MAG) and physicochemical properties (no phase separation). FAMEa (wt.%) MAGa (wt.%)
DAGa (wt.%)
TAGa (wt.%)
Viscosity (mm2 s−1 )
Density (kg m−3 )
Water (mg kg−1 )
Alcohol (wt.%) Acid number (mg pH KOH g−1 )
29 29 29 30 32 34 49 56 69
19 20 16 20 19
47 44 53 44 45
18.62 19.14 20.39 17.68 18.18 19.41 16.62 8.08 6.22
0.917 0.918 0.915 0.914 0.915 0.921 0.915 0.894 0.890
15057 11496 5595 6141 10654 2874 2750 8021
10.0 5.0 0.7 0.1 0.2
b
b
a b
4 8 2 6 4 b
b
b
7 3 9
12 10 8
32 31 14
b
0.2 0.6
2.77 2.93 3.24 1.88 2.41 0.78 0.63 0.84 1.035
10.2 10.1 9.3 10.0 10.4 b
9.5 10.7 10.3
Determined in purified oily phase. No determined.
performed in the oily phase, i.e., when there was phase separation (Table 1). Due the formation of persistent emulsion in the glycerol phase, when phase separation was observed, it was not possible to determine the physicochemical parameters in that phase. In the case of no phase separation occurred, the analyzes were performed in the reaction mixture (Table 2). As was already mentioned, the aliquots were purified in all cases (from oily phase or mixture) to determine the FAME, TAG, DAG, MAG content, and acid number.
3.1.1. Methylic soybean biodiesel: composition of a reaction medium (TAG, DAG, MAG and FAME) As previously mentioned, the transesterification process consists of three consecutive and reversible reactions, in which TAG are converted into FAAE, DAG and MAG (Suarez et al., 2007; Schwab et al., 1987). Fig. 1 presents the composition of the reaction medium (unconverted TAG, DAG and MAG) as a function of the FAME content when phase separation (Fig. 1A) or no phase separation (Fig. 1B) were observed. In the both cases, the same tendency was observed, and as expected, as FAME formed, the TAG content decreased; DAG and MAG are continuously formed and consumed, and their quantity tends to decrease during the process. When the FAME amount was between 35 and 94% (Table 1 and Fig. 1A), the formation of the phases was observed after the procedure described in the experimental section. However, in some cases (Table 1), no phase formation was observed (FAME from 29
to 56%). This lack of separation can be related to the formation of a stable emulsion that was not broken during the separation procedure adopted, possibly due to the large amount of DAG and MAG in the medium.
3.1.2. Methylic soybean biodiesel: study of correlations for the oily phase (phase separation) The graphics showing possible correlations between the kinematic viscosity and density with the amount of FAME, TAG, DAG or MAG, are shown in Fig. 2(A–D), for the oily phase samples. From Fig. 2A, it is possible to infer that the evolution of the viscosity and density values was inversely proportional to the amount of FAME (i.e., the higher the conversion to monoesters, the lower the kinematic viscosity of the oily phase), showing a strong degree of correlation (see Table S3 in Supplementary material). This behavior can be related to the intermolecular interactions present in soybean oil and biodiesel, which have a significant influence on the viscosity and density (Knothe et al., 2005). Subsequently, the correlations between the TAG, DAG and MAG contents to the viscosity and density were also investigated (Fig. 2) for the oily phase (phase separation). It is possible to note (Fig. 2B) that the viscosity evolution is directly proportional to the remaining TAG in the medium. A similar behavior was observed for the density values and is consistent with the data presented in Fig. 2A, which show that the viscosity of the FAME was lower than that of the TAG. The DAG content in the
Fig. 1. Composition of the reaction medium (TAG, DAG and MAG) as a function of FAME content: A) phase separation or B) no phase separation.
F.L. da Silva et al. / Industrial Crops and Products 95 (2017) 18–26
21
Fig. 2. A) Kinematic viscosity and density X FAME; B) Kinematic viscosity and density X TAG; C) Kinematic viscosity and density X DAG; and D) Kinematic viscosity and density X MAG for samples acquired during methylic biodiesel obtention process (phase separation).
Fig. 3. A) Water amount (mg/kg) as a function of FAME (%) content and B) Acid number (mg KOH g−1 and pH as a function of FAME (%) content.
22
F.L. da Silva et al. / Industrial Crops and Products 95 (2017) 18–26
Fig. 4. A) Viscosity and density X FAME; B) Viscosity and density X TAG; C) Viscosity and density X DAG; and D) Viscosity and density X MAG for samples acquired during methylic biodiesel obtention process (no phase separation).
Table 3 Molar distribution of MeOH in the reaction medium (phase formation). Molar ratio oil:MeOH
FAMEa (wt.%)
MeOH linked (esters)b (wt.%)
Free MeOH in oily phasec (wt.%)
Free MeOH in glycerol phased (wt.%)
1:2 1:2 1:3 1:3 1:6 1:6
45 55 69 79 92 94
56.4 69.0 57.7 66.1 38.5 39.3
0.5 0.4 0.8 0.2 1.3 2.2
43.1 30.6 41.5 33.7 60.2 58.5
a b c d
Determined in the purified oily phase. Calculated considering FAME, MAG and DAG amount. Determined experimentally. Calculated by the difference between MeOH linked in the esters and free MeOH in the oily phase.
medium (Fig. 2C) has an important influence on the development of the viscosity and density, which is analogous to that observed when analyzing the TAG correlation with the same properties. The same argument, with respect to intermolecular interactions between the species present, is valid here. In all cases a high degree of correlation was detected. However, it is not possible to establish a reasonable trend of the viscosity and density to the MAG content (Fig. 2D) and it can be verified in Table S3 (Supplementary material), by the low values
of coefficient of correlation. The reason for such behavior can be associated with the fact that the MAG molecule has two hydroxyl groups and a slight steric hindrance. As a consequence, hydrogen bonds are favored, and the importance of Van der Waals interactions (that are prevalent for other compounds, such as DAG and TAG) are minimized. Therefore, a small variation in the levels of MAG will not cause a substantial variation in the physicochemical properties.
F.L. da Silva et al. / Industrial Crops and Products 95 (2017) 18–26
23
Fig. 5. Acid number (mg KOH g−1 ) and pH as function of FAME content.
Table 4 Composition of the medium (FAEE, TAG, DAG and MAG) and physicochemical properties (no phase separation). FAEEa (wt.%)
MAGa (wt.%)
DAGa (wt.%)
TAGa (wt.%)
EtOH (wt.%)
Viscosity (mm2 s−1 )
Density (kg m−3 )
Acid number (mg KOH g−1 )
pH
10 12 16 19 19 20 23 23 26 28 30 32 36 38 40 42 43 49 51 52 54 56 62 63 66 67 68 68 68 75 77 78
2 3
12 12
76 74
b
b
b
b
b
b
b
b
b
b
b
b
b
6
19
52
b
b
b
b
b
b
6 10 10 10 12 12 22 11 15 12 11 12 26
18 22 22 18 22 22 21 17 18 19 19 20 13
48 38 36 36 28 26 16 29 18 18 17 14 5
1.3 5.1 34 0.5 – 0.4 0.7 30.9
b
b
b
b
12 20 12 10 13 9 13 13
16 11 14 14 13 11 9 7
9 3 7 8 6 12 3 3
1.6
b
b
b
1.5
19.4 19.5 24.4 23.3 23.1 20.9 18.9 23.1 22.3 21.6 13.3 13.1 14.1 12.5 11.3 11.4 11.3 10.6 12.7 11.5 10.7 12.8 11.2 10.1 11.9 6.3 10.7 8.3 9.0 6.2 6.9 7.1
914 913 917 915 913 911 909 917 916 919 905 905 907 902 904 904 907 907 907 908 905 906 905 891 891 890 902 890 901 893 875 892
6.12 9.85 7.43 6.44 4.90 1.17 10.81 6.48 7.19 19.93 16.90 11.51 1.52 20.34 17.25 22.34 20.04 17.16 24.58 43.05 30.06 21.24 10.51 8.57 29.10 36.84 13.19 15.40 23.91 11.07 11.61 6.80
8.17 8.00 10.03 9.90 9.86 9.74 7.91 9.40 9.70 7.42 8.13 9.19 7.63 9.27 7.17 9.24 7.85 9.40 9.39 8.00 8.08 7.43 9.24 7.53 8.38 8.28 8.11 7.95 8.65 9.79 9.49 9.33
a b
b
b b
2.9 b b
1.1 b b
2.6 b
0.6 b
b b b b b
3.2 b
Determined in purified oily phase. No determined.
Complementarily, correlations to the water content, acid number and pH, and FAME (%) were studied (Fig. 3A and B, respectively). Regarding the water content, it is possible to observe a decrease in the value as the FAME content increases in the medium (Fig. 3A),
but a it not possible to stablish a correlation (see Table S4 in Supplementary material). This behavior can be easily explained because when FAME is being formed, more glycerol is generated. During
24
F.L. da Silva et al. / Industrial Crops and Products 95 (2017) 18–26
Fig. 6. Composition of a reaction medium (TAG, DAG and MAG) as a function of FAEE content (no phase separation).
the phase separation step, due to the polarity of the medium, it is expected that the water remains in the glycerol phase. From Fig. 3B, we can see that there is no a relationship between the acid number and pH with the FAME content present in oily phase. This result demonstrates that these parameters indicate the efficiency of the biodiesel purification process. Regarding the determination of the alcohol content, the main objective was to evaluate the distribution after phase separation (glycerol and oily). Thus, from its experimental determination in the oily phase, it was possible to estimate their distribution, considering the various oil:alcohol molar ratios employed, as shown in Table 3. From these results, it is possible to observe that a large amount of free alcohol (unreacted) remains in the glycerol phase. This trend was expected due to the chemical nature of the species involved because there is a high affinity between glycerol and methanol, as they are both alcohols and with a similar polarity (Carvalho et al., 2012). As already mentioned, due to persistent suspension formation, the samples of glycerol (in which there was phase separation) were not analyzed. 3.1.3. Methylic soybean biodiesel: study of correlations of the mixture (no phase separation) The graphics exposing probable correlations regarding the kinematic viscosity and density with the amount of FAME, TAG, DAG or MAG are shown in Fig. 4(A–D). As shown in Fig. 4, the correlation between the density and kinematic viscosity with the formed FAME content was similar to that observed for the oily phase when the methylic biodiesel process was investigated. In this case, it should be noted that even though the physicochemical properties were determined in the mixture without any purification, the same trends observed for the samples in which there was phase separation can be observed, and similar interpretations are valid (Table S3, in Supplementary material, to examine coefficients of correlation and determination). Regarding the investigation of the correlation between the amount of FAME to the acid number and pH (Fig. 5), it was not
possible to establish trend, as was observed in the case of the oily phase (phase separation). The analysis of the glycerol content by the European Standard EN 14105, which uses gas chromatography, was not used in this study because the samples that represented the reaction process contained high amounts of glycerol. 3.2. Correlations between the composition of the reaction medium and physicochemical parameters, during ethylic soybean biodiesel obtention In the study of ethylic biodiesel, which was similar to the study of methyl biodiesel, several samples were obtained that exhibited different compositions to represent the process. Tables 4 and 5 present the medium composition (in terms of TAG, DAG, MAG and FAEE) and the respective physicochemical characterizations. In this case, no phase separation was observed for the majority of reactions. This result was expected because the formation of stable emulsions during ethanolysis is related to the presence of high amounts of DAG and MAG (Zhou et al., 2003). An accurate analysis of the results show that, in fact, the phase separation occurred when the FAEE content exceeded 76%, a situation in which there is less TAG, DAG and MAG in the reaction medium. 3.2.1. Ethylic soybean biodiesel: composition of a reaction medium (TAG, DAG, MAG and FAEE) and a study of the mixture (no phase separation) Due the small number of samples that showed phase separation (Table 5), it was not possible to establish any trends. The results and discussion will be focused on the samples in which no phase separation was observed (Table 4). Fig. 6 presents the composition of the reaction medium (unconverted TAG, DAG and MAG) as function of the FAME content when no phase separation was observed. As was discussed for methylic biodiesel, as FAME formed, the TAG content diminished. DAG and MAG were formed and consumed, and their amounts tended to decrease as the process proceeded.
F.L. da Silva et al. / Industrial Crops and Products 95 (2017) 18–26
25
Table 5 Composition of the medium (FAEE, TAG, DAG and MAG) and physicochemical properties (phase separation).
b
FAEEa (wt.%)
MAGa (wt.%)
DAGa (wt.%)
TAGa (wt.%)
EtOH (wt.%)
Viscosity (mm2 s−1 )
Density (kg m−3 )
Acid number (mg KOH g−1 )
pH
76 79 80 83 84 86
15 10 9 7 3 4
8 11 11 9 10 9
1 0 1 1 3 1
2.3 7.0 12.5 7.7 2.5 3.0
4.3 5.7 5.5 5.3 4.7 4.3
894 885 883 882 876 880
15.96 3.57 3.78 4.34 17.73 15.12
8.56 9.90 9.70 9.40 8.56 7.98
No determined. a Determined in purified oily phase.
Fig. 7. A) Kinematic viscosity and density X FAEE; B) Kinematic viscosity and density X TAG; C) Kinematic viscosity and density X DAG; and D) Kinematic viscosity and density X MAG, for samples acquired during ethylic biodiesel obtention process (no phase separation).
To understand the physicochemical parameters that are related to the medium composition, possible correlations were investigated (Fig. 7). The correlations of the FAEE and TAG amounts with the kinematic viscosity and density values (Fig. 7A and B) followed the same trend observed in the case of methylic biodiesel (Figs. 2 and 4) and, high degree of correlation could be stablished in most cases. However, it was not possible to establish correlations or trends between the DAG and MAG levels and the evaluated prop-
erties (Fig. 7C and D). Most likely, this behavior can be related to the complexity of the mixture. As was with observed with the methylic biodiesel, it was not possible to establish correlations between the FAEE amount and the acid number or pH (Fig. 8), which confirmed that these parameters indicate whether the purification step was effective.
26
F.L. da Silva et al. / Industrial Crops and Products 95 (2017) 18–26
Fig. 8. Acid number and pH as a function of FAEE content.
4. Conclusions
References
In this study, it was possible to establish some correlations between physicochemical parameters (density and kinematic viscosity) and the composition of the medium during the reaction of the transesterification of soybean oil with methanol and ethanol. An analysis of the results allows us to infer that the composition of the medium in terms of remaining catalyst and its residues, alcohol and water content, among others variables, has no significant influence on the physicochemical properties studied. These findings are mainly influenced by the composition of the medium in terms of FAAE, TAG, DAG and MAG.
Carvalho, M.S., Mendonc¸a, M.A., Pinho, D.M.M., Resck, I.S., Suarez, P.A.Z., 2012. Chromatographic analyses of fatty acid methyl esters by HPLC-UV and GC-FID. J. Braz. Chem. Soc. 23, 763–769. Chuck, C.J., Bannister, C.D., Hawley, J.G., Davidson, M.G., La Bruna, I., Paine, A., 2009. Predictive model to assess the molecular structure of biodiesel fuel. Energy Fuels 23 (4), 2290–2294. Colton, T., Freedman, L.S., Johnson, A.L. (Eds.), 1982. John Wiley and Sons, New York. Datta, A., Mandal, B.K., 2016. A comprehensive review of biodiesel as analternative fuel for compression ignition engine. Renew. Sustain. Energy Rev. 57, 799–821. Hoekman, S.K., Broch, A., Robbins, C., Ceniceros, E., Natarajan, M., 2012. Review of biodiesel composition, properties, and specifications. Renew. Sustain. Energy Rev. 16, 143–169. Knothe, G., Van Gerpen, J., Krahl, J., In: Knothe, G., Van Gerpen, J., Krahl, J. (Eds.), 2005. AOCS Press, Urbana, IL. Knothe, G., 2005. Dependence of biodiesel fuel properties on thestructure of fatty acid alkyl esters. Fuel Process. Technol. 86, 1059–1070. Najafabadi, H.A., Pazuki, G., Vossoughi, M., 2012. Estimation of biodiesel physical properties using local composition based models. Ind. Eng. Chem. Res. 51, 13518–13526. Schwab, A.W., Bagby, M.O., Freedman, B., 1987. Preparation and properties of diesel fuels from vegetable oils. Fuel 66, 1372–1378. Su, Y.C., Liu, Y.A., Diaz Tovar, C.A., Gani, R., 2011. Selection of prediction methods for thermophysical properties for process modeling and product design of biodiesel manufacturing. Ind. Eng. Chem. Res. 50, 6809–6836. Suarez, P.A.Z., Meneghetti, S.M.P., Meneghetti, M.R., Wolf, C.R., 2007. Transformation of triglycerides into fuels, polymers and chemicals: some applications of catalysis in oleochemistry. Quim. Nova 30, 667–676. Zhou, W., Konar, S.K., Boocock, D.G.B., 2003. Ethyl esters from the single phase base-catalyzed ethanolysis of vegetable oils. JAOCS 80 (4), 367–371.
Acknowledgements Financial support from Brazilian research founding agencies, such as Research and Projects Financing (FINEP), National Counsel of Technological and Scientific Development (CNPq), Brazilian Federal Agency for the Support and Evaluation of Graduate Education (CAPES), and Alagoas Research Support Foundation (FAPEAL) are grateful acknowledge. SMPM thank CNPq for research fellowships and FLS, JRN, LNM, JASF thanks CNPq for fellowship. The authors are indebt with Ministry of Science, Technology and Innovation (MCTI) and CNPq for FISQUIBIODIESEL project. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.indcrop.2016.09. 053.