- Email: [email protected]
Microwave fast-tracking biodiesel production
Accepted Manuscript Title: MICROWAVE FAST-TRACKING BIODIESEL PRODUCTION Authors: Luiz A. Jermolovicius, Luana C.M. Cantagesso, Renata B. do Nascimento, Edmilson R. de Castro, Eduardo V. dos S. Pouzada, Jos´e T. Senise PII: DOI: Reference:
S0255-2701(16)30209-4 http://dx.doi.org/doi:10.1016/j.cep.2017.03.010 CEP 6946
To appear in:
Chemical Engineering and Processing
Received date: Accepted date:
15-7-2016 21-3-2017
Please cite this article as: Luiz A.Jermolovicius, Luana C.M.Cantagesso, Renata B.do Nascimento, Edmilson R.de Castro, Eduardo V.dos S.Pouzada, Jos´e T.Senise, MICROWAVE FAST-TRACKING BIODIESEL PRODUCTION, Chemical Engineering and Processinghttp://dx.doi.org/10.1016/j.cep.2017.03.010 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
MICROWAVE FAST-TRACKING BIODIESEL PRODUCTION
Luiz A. JERMOLOVICIUS, Luana C. M. CANTAGESSO, Renata B. do NASCIMENTO, Edmilson R. de CASTRO, Eduardo V. dos S. POUZADA, José T. SENISE
Laboratório de Micro-ondas, Centro Universitário do Instituto Mauá de Tecnologia, Praça Mauá, 1 – São Caetano do Sul, SP 09580-900, BRAZIL; e-mail: [email protected], [email protected]
Highlights
Complete kinetic equation justify transesterification Independence from ethanol. Transesterification microwave enhancement is done by changing kinetics parameters. Kinetic equation under microwave irradiation induces operational optimization. Microwave heated process higher productivity was quantified and explained.
Keywords: biodiesel, microwaves, process intensification
1. Introduction Biodiesel is commonly produced by the process of transesterification of triglycerides such as soybean oil, palm oil or any other vegetable oil with ethanol or methanol in the presence of a catalyst. Transesterification can be performed in batch or continuous flow using conventional heating [1]. Biodiesel is a good alternative to be used in vehicles based on Diesel cycle engines because its molecular structure presents a high cetane index [2]. As a fuel based on renewable sources biodiesel can contribute to reduce the intensity of global warming. Its use brings a lot of benefits, say, it is biodegradable, renewable, and capable of promoting sustainable economic development. All these features can reduce the overall dependence on fossil fuels. Nevertheless and despite the high investments in biodiesel plants, like the Brazilian Caramuru Food Co. that opened a 225 million liters manufacturing unit of biodiesel in 2010, corresponding to an investment of 54 million dollars [3], the Secretariat for Planning and Strategic Investments of Brazilian Federal Government Ministry of Planning alerted that the production of biodiesel was economically unfeasible. In fact, the production of biodiesel is more expensive than ordinary diesel, and even today it is not sustainable from an economic point of view [4].
According to Government analysts, the problems plaguing production of biodiesel are focused on two basic issues: (1) the selection of more suitable raw materials for biodiesel production, and (2) infrastructure that enable the commercial exploitation of these raw materials. This point of view is also shared with other countries, where the costs of raw materials become 70 – 95 % of biodiesel production cost [5]. Within this context, a third factor can be introduced: the transesterification technology adopted – predominantly the alkaline catalysis [2, 5, 6, 7, 8, 9] – is not efficient enough to produce a competitive biodiesel. Since 2006 there are more than 650 deposited patents for biodiesel production and only 95 of them are about acid catalyzed processes [10]. In this scenario, it is clear the necessity of biodiesel technology enhancement. The current raw materials costs and their logistic issues escape from the scope of this work, but the technology for biodiesel production is its objective. A first option aiming better processes is the process intensification. It is an innovative and revolutionary focus on chemical processes that allows improvement of productivity and reduction of operation costs [11, 12, 13] and, obviously, it is a nice option to biodiesel production. Analyzing the production of biodiesel under conventional heating, it is observed that the energy necessary to fulfill the reaction is transferred from the hot surface of the reactor by conduction and convection, establishing a thermal gradient between the hot surface and the mass of reactants. This structure makes the velocity of the reaction dependent on heat transfer efficiency and it brings troubles in the process productivity. A second option to improve a chemical process is the adoption of microwave heating instead of conventional heating [14, 15]. Nowadays, application of microwave energy in synthesis processes is a well known alternative to increase the velocity of reactions [16, 17, 18, 19] as it was initially observed in procedures for digestion of samples under microwaves [20]. Basically, microwaves are
electromagnetic energy that can heat polar media. Polar molecules tend to align with the electromagnetic field and this phenomenon produces heat due to the friction caused by the slower reorientation of the molecules compared with the time rate change of the field. There is also a non-thermal specific microwave effect: uncoupling the spin of electrons in the atoms [16, 17, 21], which leads to new mechanisms models generally faster than the conventional ones. Following this approach, an earlier experiment showed a strong synergic effect between microwaves and acid catalysis in esterification maleic anhydride with 2-ethylhexanol-1 [22], lowering the chemical reaction order and the activation energy. It is interesting to observe that since 2006 there were 18 patented processes related to microwave assisted biodiesel production [10]. All these facts induce to think about a faster biodiesel process based in acid catalysis assisted by microwaves. The experiment reported in this work was done aiming at a better understanding of microwave enhanced acid process to biodiesel production from vegetable oils.
2. Microwave batch reactor and its operation As any microwave device the reactor used in this work is modeled at the macromolecular level by Maxwell’s equations. Each particular solution is obtained by applying the electromagnetic boundary conditions which are forced by the chosen geometry and materials that interacts with the electric and magnetic fields. In practice some problems arose that can reduce efficiency of an applicator: (1) good control of geometric aspects, and (2) knowledge of the constitutive parameters that model the materials, say the relative complex permittivity r and, sometimes, the relative complex permeability µr when magnetic materials are involved. The constitutive parameters are frequency and temperature dependents; their values are difficult to be found in the literature, so one
should measure them with some degree of confidence. The geometric issue is also important because of the interaction between materials and electromagnetic fields. In this work a large geometry was used – in the sense that it comprises several wavelengths in all the three spatial dimensions. This framework does not allow closed analytical solutions, so numerical ones must be sought by computer simulation. This approach was used in this work. Fig. 1 shows the microwave batch reactor used to process the transesterification of soya oil with ethanol. It consists of a 500-mL Pyrex kettle vessel with reflux condenser and mechanical stirrer. They are inside a multimodal cavity with a mode stirrer, and coupled to a 2.45 GHz microwave generator with adjustable power up to 3 kW. The setup allows connection of power sensors to measure the forward (irradiated) and reflected microwave power to quantify the value of the effective power transmitted to reacting materials. The reflected power is an inherent loss in microwave applicators and represents how much microwave energy the system was not able to absorb and reflects it back. This reduction is the target design of microwave chemical reactors. Temperature measurement was done with a fiber optic temperature sensor protected by a thermometric well immersed in reacting material. In a typical test the reactor is initially loaded with methane sulfonic acid and ethanol. Then soy oil, previously heated at 70° C, is added under stirring. The reagents quantities are defined for each kind of tests as shown below.
3. Biodiesel yield determination The dosage of the produced biodiesel was done by an indirect method based on the determination of the glycerin formed during transesterification. Numerical results are then justified by stoichiometric relation between biodiesel and glycerin. Although this
reaction is formally reversible, it was assumed as pseudo irreversible when processed with excess of alcohol. The dosage of glycerin is an iodometric method based on its reaction with periodic acid (traditional Malaprade method [23]) and final dosage of iodine with sodium thiosulfate. Fig. 1 shows the sequence of reactions that occur in this procedure. In the usual Malaprade procedure a small portion of sodium lauryl sulfate was introduced with the aim to produce a homogeneous solution, emulsifying oil and biodiesel. The analytical procedure [24] is: weight around 0.5 g of biodiesel production medium (soya oil, ethanol in excess, catalyst, biodiesel, glycerin) with high precision; add 1 mL of 5 % sodium lauryl sulfate solution; add 50 mL of 0.1 N periodic acid solution; rest the reactants for 120 minute under continuous swirling; add 30 mL of 20 % potassium iodide solution and 25 mL of 6 N sulfuric acid; titrate with 0.1 N sodium thiosulfate until pale yellow color is obtained and add 2 mL of soluble starch solution. Then titrate until the blue color disappear. In sequence, use a blank titration without glycerin, to standardize the periodic acid solution.
Knowing that one mol of glycerin represents one mol of biodiesel, then the stoichiometric yield of biodiesel may be determined by Eq. 14. It is based on chemical reaction stoichiometry relations between reactants and products and represents how much of the theoretical amount of desired products was really produced. It is common in scientific literature the use of a technical index based on the relation between the produced mass and the mass of limiting raw material [25, 26, 27] instead of the yield based on chemical stoichiometry. This technical index does not represent what happens in terms of chemical reactions; it only reports a relation of products, but it is much more easy to use than the stoichiometric approach.
𝑀𝐺𝐴 = 𝑀𝐺𝑇 = 𝑀𝐵 = 𝐵𝑌 =
𝑀𝑀𝐺 ∙𝑁𝑇𝑆 ∙10−3 ∙(𝑉𝑇𝑆 𝑂 − 𝑉𝑇𝑆 𝑅 ) 4∙𝑀𝐴 𝑀𝐺𝐴 ∙𝑀𝐴𝑇
𝑀𝐴 3∙𝑀𝑀𝐵 ∙𝑀𝐺𝑇 𝑀𝑀𝐺 𝑀𝐵 ∙𝑀𝑀𝑂 ∙100 𝑀𝑂 ∙𝐼∙3∙𝑀𝑀𝐵
(1) (2) (3) (4)
where MGA stands for the mass of glycerin in a sample, MGT is the mass of total glycerin in an experiment, MB is the prepared mass of biodiesel in an experiment, and BY is the biodiesel stoichiometric yield in an experiment. In (1)(4) MMG is the glycerin molar mass (92,09 gmol-1), MMB is the biodiesel molar mass (248.14 gmol-1), MMO is the soya oil sample molar mass (781.09 gmol-1), MA is the sample mass (g), MGA is the mass (g) of glycerin in the sample mass, MGT is the glycerin total mass (g) of an experiment, MAT is the total mass (g) of an experiment, MB is the biodiesel mass (g) of an experiment, MO is the soya oil mass (g) of an experiment, I is the correction factor for soya oil based on actual saponification, NTS is the sodium thiosulfate normality, VTS 0 is the volume of thiosulfate in standardization titration (mL), VTS R is the volume of thiosulfate in sample titration (mL), and BY is the biodiesel stoichiometric yield (%). In the above calculations, the average molar masses of soya oil and biodiesel were determined by cryoscopy method [28] and taken from literature in the case of glycerin. The purity correction factor for soya oil was determined in function of its actual molar mass (781.09 g.mol-1) and the relation between its actual and theoretical saponification index [29]. The actual value was determined by saponification with excess of alcoholic solution of potassium hydroxide and titration of excess with hydrochloric acid [24]. The correction factor for oil purity is the relation between the actual saponification index (189,59 mg KOH.g-1) and the theoretical saponification index for this molar mass of oil (215,51 mg KOH.g-1).
4. Defining the operational conditions Here the objective is the identification of the region defined by the range of the controlled variables that can produce the highest stoichiometric yield in biodiesel. They were selected as (1) the molar relation between ethanol and soya oil, (2) catalyst concentration (mass of catalyst per mass of oil), (3) specific absorption rate (W/g) which is directly associated with the processing temperature, and (4) reaction time (min). The catalyst adopted was methane sulfonic acid. The Simplex method [30] was used for optimization. All tests were done in duplicate. Table 1 shows the initial Simplex (columns 1 to 5). Analyzing the results form tests 1 to 5, the worst result (column 4) was eliminated and substituted by its symmetrical point (column 6). Proceeding with the analysis of the new result and comparing it to the former, the new worst result (column 2) was eliminated and substituted by its symmetrical counterpart (column 7). Once again, the worst (point 5) was substituted by column 8, and so on. When column 6 was substituted, the new result became the worst. Then, columns 1, 3, 6, 7, and 8 define the optimum region and its central point was adopted as the optimum point for this microwave heated acid biodiesel process.
5. Comparison of conventional and microwave processes The objective of this comparison between microwave assisted biodiesel process and a conventional heating alkaline biodiesel process is to confirm which of them has higher productivity. Analysis of variance [31] was used for this purpose. Considering that all processes may produce high stoichiometric yield but not exactly at the same time, then a comparison of results from a conventional heating alkaline process and the microwave heated acid process was done at unique establishing time of operation, specifically 7.5 min.
Excluding a natural doubt if the effect on biodiesel production is due to the acid catalyst or to the microwave energy, another universe was introduced in this comparison: a conventional heating acid process performed at the same conditions of the microwave irradiated process but with conventional electric heating. Usual laboratory equipment (cf. Fig. 2) was used for this process. The data set for conventional heating alkaline process was prepared in usual laboratory equipment, as shown in Fig. 2, and supported by published procedure [32; 33]. This process used 120 g of soybean oil with 781.09 average molar mass and average 189.59 mg KOH/g saponification index, ethanol and sodium methoxide as catalyst, with a 20minute heating period at 45 C by an electric blanket. The reaction control was the same as adopted for the microwave-assisted process described above. It is important to observe that the running time for these tests was reduced to 7.5 min; that is the time necessary for the microwave process to get its maximum yield. Results are shown in Table 2. It can be observed that the reduction of processing time also reduced the yield of biodiesel in the conventional heating processes, because they are slower than the microwave one which achieves more than 99 % of yield. Analysis of variance applied to Table 2 data and F distribution [31] adoption resulted in a critical value Fcrit = 9.34, with 99.9 % of confidence and Φ1 = 2 and Φ2 = 21 degrees of freedom, against a calculated factor Fcalc = 21.00. This means that one of the universes analyzed is different with 99.9 % of confidence. Observing the data, it is clear that the different universe is the microwave heated process. Completing the analysis, a t Student’s Test was applied to conventional heating processes data. It resulted in a critical factor t crit = 1.721 with degree of freedom Φ = 21 and 10 % of confidence and a calculated factor t calc = 0.875. This means that is not a significant difference between conventional heating under alkaline or acid processes.
Longer running time experiments (20 minutes) were also performed in the same conditions as described above. The results of biodiesel stoichiometric yield of this new series are plotted in Fig. 3, where a normalized time scale was adopted, and defined by the relation between effective running time and conventional alkaline process running time to reach their maximum yield. This make a nice confirmation that the microwave heated process needs less time to reach its maximum yield in comparison with the conventional processes described above. Effectively, the process velocity increased 42.5 %.
These data confirm that the increase of biodiesel yield may be attributed to a microwave effect, not to a kind of acid or alkaline catalyst. Moreover, one can think about doubling a rector’s productivity or half-fold its original volume. The experimental mean of stoichiometric yield was 99.24 % (cf. Table 2), a higher value compared to published biodiesel yields of 95.5 % [34] and 97.1 % [35] for microwave irradiated transesterification of waste cooking oil. Electrical energy consumption showed interesting figures too: 30.9 Wh, 64.5 Wh, and 75.0 Wh for the microwave, the conventional acid, and the conventional alkaline process, respectively. These values were measured with a wattmeter. Consequently, the microwave process shows an economy of 58.8 % or 52.1 % in comparison with the conventional alkaline or the conventional acid process, respectively.
6. Trying to understand the microwave effect Among the three processes considered in this work, the microwave heated has the highest efficiency, but there is no explanation about it, yet.
Considering that the yield of a process is partially controlled by the reaction chemical kinetic, then it is possible to suppose that microwave energy impacts the original reaction mechanism anyway, thus accelerating the process. It stands to reason that a complete kinetic equation determination for the microwave assisted acid process was done. The procedure for complete chemical equation determination [22; 36] was based on initial rate methods [37] combined with a multilinear regression [38]. It starts with knowledge of two empirical facts: a) that the triglycerides transesterification with ethanol may be represented by an equilibrium reaction; b) that the reversibility of this reaction is not strong from the experimental point of view, so it can be conceived as practically irreversible in laboratories, then it is possible to assume it as a pseudo irreversible reaction. This leads to W = K0·exp[E/(R.T)]·CAn·CBmCCp
(5)
In (5) W denotes the reaction rate (mol/kgmin), K0 is the pre-exponential Arrhenius factor, E is the activation energy (J/mol), R is the ideal gas constant (8.31446 J/Kmol), T is the temperature (K), n and m are the reaction orders in respect to triglycerides and ethanol, respectively, CA, CB and CC are the concentrations of triglycerides, ethanol and catalyst, respectively (usually measured in mol/L, but in this case adopted as mol/kg), and exp is the base of natural logarithms ( 2.71828). The unit for concentration adopted was the molal (mol/kg) to avoid troubles with reproducibility with sampling hot reacting material and incomplete physical separation of glycerol from biodiesel. It is also important to observe that no solvents were used; the reaction solution was just soya oil and ethanol plus a catalyst. The catalyst methane sulfonic acid concentration was optimized at the start of each experiment and was adopted to be constant, then the product K0CCp becomes constant (In this paper it is denoted as a pseudo constant k0).
Eq. (5) has its intricacies in modeling the phenomenon: it has two variables in potential form and a third variable in the exponential form. It is usual to plot experimental kinetic data in graphics which represents a first, second or third order [37]. This is not an exact approach, but it helps the determination of a complete chemical kinetic equation. Levenspiel presents a method based on initial velocities of reactions that can be adopted for a multilinear regression by linearizing Eq. (5) with the help of logarithms. The linearized equation becomes: log(W) = log(k0) – [E/(R.T)]·log(e) + n·log(CA) + m·log(CB)
(6)
Eq. (6) is linear. It is easily observed after renaming their variables as below: Y = a0 + a1·z1 + a2·z2 a3·z3,
(7)
where Y = log(W), z1 = log(CA), z2 = log(CB), z3 = 1/T, a0 = log(k0), a1 = n, a2 = m, and a3 = (E/R)log(e). In order to apply a statistical regression method, Eq. (7) was transformed into reduced parameters (xi) Y = b0 + b1·x1 + b2·x2 + b3·x3,
(8)
where b0, b1, b2, b3 are the regression coefficients, and x1, x2, x3 are the reduced parameters calculated by: xi = (zi – zi0) / Δzi ,
(9)
where zi is the natural value of a variable, zi0 is its natural central value, and Δzi is the associated incremental. The temperatures adopted for this experiment were the equilibrium temperatures when the reactants were irradiated with 300 W of microwave (347.6 oC) and 400 W (384.0 oC). These microwave effective powers were selected to attend equipment operational requirements. The central value and increment change for soya oil were 3.82 and 0.080
mol/kg respectively, and for ethanol they were 1,12 and 0,43 mol/kg respectively. Table 3 shows typical values in natural and reduced form. The order of execution of duplicated tests was random without restrictions. They produced sixteen kinetic curves as, for example, the curves for the points of maximum and minimum velocities shown in Fig. 4. The results of initial velocity of reaction (Y1 and Y2) are shown in Table 5. Their values were determined by numerical derivative taken at time zero of the experimental kinetics curves obtained with Table 3 reaction conditions.
It is very interesting to observe that curves for low level of soya oil concentration plus another variable in low level produced a kinetic curve similar to Fig. 4b and was the slower reactions. Points with high level of soya oil concentration and point 2 (lower soya oil concentration, but higher value for ethanol and temperature) produced curves similar to 4a and they were faster. There are two differences between curves 4a and 4b; one is the initial rate of reaction and the second is an inflexion at 3 min. instant time. In a first view, this inflexion may be related to partial substitution of carboxylic acids from glycerin residue in oil molecule. It is an induction that the first substitution of fatty acid in oil molecule is faster than the others. Multilinear regression applied to data taken from Table 5 results in Y = 0.986 0.0143·x1 + 0.000681·x2 0.0618·x3.
(10)
The significance of this regression was verified by using the Student’s t-distribution. The standard error was calculated with the worst replicate test, getting a value of se = 0.2231 and sbj = se.(m.N)-1/2 = 0.000139, where m is the number of data in one replication (m = 8) and N is the number of replications (N = 2). Supposing that all sbj are equal, we have
the values for tcalculated for each parameter tj = |bj|/sbj: t0 = 1768, t1 = 26, t2 = 1.2 and t3 = 111. For two degrees of freedom and 99.9 % of confidence the ‘t’ critical factor is 9.925, and for 95 % the critical factor is 4.303. The regression factor b2 has no significance with 95 % of confidence and all others have significance within 99.9 %. Consequently, Eq. (10) represents adequately the experimental data and can be rewritten in the following form: Y = 0.986 0.0143·x1 0.0618·x3.
(11)
The lack of fit of Eq. (11) was tested with Fischer’s variance rate. F = sg2/se2
(12)
where, sg2 is the mean square goodness-of-fit and se2 is the mean square error. Their values were 0.004897 and 0.002749, respectively, and Fischer’s factor was 1.78 which value is less than the critical factor (with 90 % of confidence and degree of freedom 3 and 8, Fcrit = 2.92). This means that the Eq. (11) fits well the experimental data. Using them, Eq. (11) and Eq. (7), it is possible to establish the values of the parameters of the chemical kinetic equation for the microwave heated acid biodiesel; they are: reaction order for soya oil, n1 = 1.57; energy of activation, E = 26526,71 J/mol; pseudo Arrhenius pre-exponential factor considering a constant catalyst concentration, k0 = 8.5 105. The complete chemical kinetic for soya oil transesterification with ethanol and methane sulfonic acid under microwaves is: W = 8.5105·exp[26527/(RT)]·CA1.6
(13)
The reaction order with respect to ethanol resulted in n2 = 0.0039, which is practically a zero-order reaction. It is interesting to observe that the regression coefficient b2 does not have significance because its parameter is of zero order, having no influence on the chemical reaction velocity. Summarizing, the biodiesel production under microwaves is
independent of ethanol concentration considering the experimental range of parameters adopted. Eq. (13) has two uncommon features: the activation energy and the reaction order related to soya oil concentration have negative values. The former means that the reaction rate will decrease with temperature increase [39; 40; 41; 42;43] and the latter means that the reaction velocity will decrease with the increase of concentration of this specific reagent [44; 45; 46]. Both facts induce the necessity to optimize the microwave enhanced transesterification conditions to achieve a better productivity, just to find the more effective temperature and soya oil concentration. The optimization done at beginning of this work is now justified by the kinetic behavior of microwave heated biodiesel process.
7. Analysis of kinetical data There was a strong increase in the number of publications about several kinetical aspects of biodiesel production after 2014. From 2001 to 2009 there were 31 papers (circa 3 per year), and from 2010 to 2016 there were 135 papers published (circa 20 per year) [47]. Table 6 shows some of the most recent papers about chemical kinetic equation. Also a review [48] with anterior data was analyzed. All papers report a kinetic equation determination by graphical agreement of experimental data to formal first or second order, as seen in chemical reactor design courses [37]. This method is not the most precise; it is good for an initial study, although insufficient for process design. In spite of having its operational difficulties the method described in Session 6 was adopted in this work because it can determine kinetic parameters without any previous definition of reaction order. Moreover, with this regression method it is possible to verify the significance of chemical kinetics equation parameters and the level of its fit to experimental data.
It was observed that first order reactions is predominant for different triglycerides, circa 48 % of the published papers. There are some articles reporting results of second order and even fractional order (between one and two). The papers report global orders or related to triglycerides concentration, but do not explain why alcohol concentration was not considered. The regression method showed that, under microwave irradiation, the alcohol concentration is not significant. The reason becomes clear when the reaction order value with respect to ethanol was calculated; it was practically zero which means independence of the reactant concentration. The order with respect to soya oil was determined as 1.6, which is greater than the orders of conventional biodiesel process, independently of the triglycerides source and the catalyst used. High order reactions imply higher velocities; this may justify the observed higher velocity of the microwave assisted transesterification. To the best knowledge of the authors it seems that there is no published information in the state of art about the negative order with respect to soya oil. This means that the increase of oil concentration results in reaction rate decrease. In other words, if the concentration of ethanol is lowered in a system oil / ethanol without solvent, then the velocity will decrease. The activation energy observed under microwave irradiation is lower than the conventional processes as can be seen in Table 6 and in Verma’s review [48]. It is the case for soya oil. This fact justify the higher reaction rate when the transesterification is done with microwave irradiation. There are some exceptions relating lower values of activation energy than that of the microwave enhanced process which may be attributed to natural reactivity of the specific triglycerides of these cases; they were rice, cottonseed, palm, rapeseed, and peanut oils.
Another information not available in publications of the state of the art is the negative activation energy. This means that the increase in reaction temperature will decrease the velocity for production of biodiesel. This fact may be attributed to temperature stimulation of reversible path of transesterification. Note that one premise usually adopted for this kinetic study is that the reversible path does not represent a significant influence on the process. There are some others papers that report experiments for transesterification of triglycerides under microwave irradiation and give kinetical data, as shown in table 7. As expected, all global orders were higher than one (except for one paper) and all activation energies were lower than that of conventional heating (again except for one paper). The high value of activation energy is for Ceiba pentandra seed oil [49], which probably react under elevated level of energy. The first order reported [57] may be explained by the fact that this value was assumed to determine kinetic parameters. In the case of soya oil, there is one publication for conventional heating [58] and other for microwave heating [57]. The former ratify that microwave heated process is superior than the conventional one. Unfortunately, the order of reaction was not determined empirically, but previously assumed in the latter work. Finally, the microwave equipment setup of Terigar’s work did not allow measurements of incident and reflected powers.
8. Conclusions This work confirms the higher productivity of microwave heated acid biodiesel production process, as shown in Fig. 3 and Table 2. From a process time point of view, a microwave process can increase the biodiesel production in 42 % (cf. Fig. 3). Microwave heated acid process is capable to reach the complete conversion of triglycerides in less operation time than the conventional heated process, as shown in
Table 2 and Fig. 3. In other words, microwave process is faster than conventional processes and for this reason, it is 58.8 % more economic in terms of electrical energy consumption than the conventional heated alkaline process, as shown in Session 5. With this information one may visualize a possible increase in biodiesel industry productivity by applying microwave energy as an alternative heat source. A complete chemical kinetic equation for transesterification of soya oil with ethanol under microwave irradiation was also determined as a result of this work (cf. Eq. (12)). It shows that the biodiesel microwave enhanced acid process is independent of ethanol concentration and it requires an optimization of soya oil and ethanol concentration and operation temperature, because of the negative value in the order of reaction with respect to soya oil, the zero order with respect to ethanol and a negative activation energy. These facts show that there is an optimal set of operational parameters for economical operation, as discussed in Session 6. Comparison
between
experimental kinetic
equation
for
microwave-enhanced
transesterification of soya oil with ethanol and published kinetic equations shows a lower energy activation than that of conventional heated processes and strong differences in reaction order values, as discussed in Session 7. Considering that a kinetic equation represents the reaction mechanism, it is possible to accept that microwave energy altered it in such a way that permits a new series of intermediate paths resulting in a faster reaction. Summarizing, switching from conventional heating to microwave heating changes the reaction mechanism and accelerates the transesterification, as shown, and promotes an operational time reduction to get the maximum yield with considerably energy economy. This is an example of a non-thermal microwave effect.
Acknowledgement The authors would like to thank D. Z. Passeti and C. Tognela. Supports from Instituto Mauá de Tecnologia – IMT and Fundação de Amparo à Pesquisa do Estado de São Paulo – FAPESP (Grant no. 2011/50154-9) are gratefully acknowledged.
References [1] M.U. Nascimento, S.C.A. Vasconcelos, B.E. Azevedo, C.F. Silva, Otimização da produção de biodiesel a partir de óleo de babaçu com aquecimento por micro-ondas, Eclética Química 34 (2009) 4. [2] G. Knothe, V.J. Gerpen, J. Krahl, Manual de Biodiesel, Edgard Blücher, São Paulo, Brazil, 2007. [3] F. Lopes, 3.9.10, Caramuru amplia sua produção de biodiesel. Valor Econômico. Available http://www.valoreconomico.com.br. [4] W.S. Dália, A produção do biodiesel: uma perspectiva para a agroenergia no nordeste brasileiro, O Futuro da Indústria: biodiesel – coletânea de artigos, Indústria e Comércio Exterior: IEL, 27-36, 2006. [5] P.P. Oh, H.L.N. Lau, J. Chen, M.F. Chong, Y.M. Choo, A Review on Conventional Technologies and Emerging Process Intensification (PI) Methods for Biodiesel Production, Renewable and Sustainable Energy Review 16 (2012) 5131-5245. [6] A. Pandey, Handbook of plant-based biofuels, CRC Press, Florida, United States, 2009. [7] Produção de combustíveis líquidos a partir de óleos vegetais, Fundação Centro Tecnológico de Minas Gerais, Programa de Energia, vol.2 ch. 1, Belo Horizonte, Minas Gerias, Brasil, 1983. [8] C.V. Ferreira, Otimização do Processo de Produção de Biodiesel a Partir de Óleos de Mamona e Algodão (Master thesis), Universidade Federal de Pernambuco, Brasil, 2007. [9] C.N. Khalil et al., “Biodiesel purification method,” U.S. Patent 20110209981, 1 st September, 2011. [10] European Patent Office. https://worldwide.espacenet.com/search, 2016 (accessed 25.6.16). [11] S. Becht, R. Franke, A. Geisselmann, H. Hahn, An industrial view of process intensification, Chemical Engineering Processing: Process Intensification 48 (2009) 329-332. [12] A. Stankiewicz, J.A. Moulin, Process intensification: transforming chemical engineering, Chemical Engineering Progress 50 (2000) 22-34. [13] D. Reay, C. Ramshaw, A. Harvey, A Process intensification. Engineering for efficiency. Sustainability and flexibility, Elsevier, United Kingdom, 2008. [14] A.C. Metaxas, R.J. Meredith, Industrial Microwave Heating, Peter Peregrinus, London, Great Britain, 1983. [15] G. Roussy, J.A. Pearce, Foundations and industrial applications of microwave and radio frequency fields (1995) India, John Wiley, 1995. [16] P. Lindström, J. Tierney, B. Wathey, J. Westman, Microwave assisted organic synthesis: a Review, Tetrahedron, 57 (2001) 9225-9283.
[17] J.T. Senise, L.A. Jermolovicius, Microwave chemistry: a fertile field for scientific research and industrial applications, (2004) Journal of Microwaves and Optoelectronics 3 (2004) 97-112. [18] C.O. Kappe, Microwave dielectric heating in synthetic organic chemistry, Chem. Soc. Rev. 37 (2008) 1127-1139. [19] C.R. Strauss, A Strategic, “Green” Approach to organic chemistry with microwave assistance and predictive yield optimization as core, enabling technologies, Australian J. Chem. 62 (2009) 3-15. [20] F.E. Smith, E.A. Arsenault, Microwave-assisted sample preparation in analytical chemistry, Talanta 45 (1996) 1207-1268. [21] A.L. Buchachenko, Recent advances in spin chemistry, Pure Appl. Chem. 72 (2000) 2243-2258. [22] L.A. Jermolovicius, B. Schneiderman, J.T. Senise, E.R. Castro, “Microwave synergic effect on maleic anhydride catalytic esterification with 2-Ethylhexanol-1,” presented at the International Microwave and Optoelectronics Conference, Foz do Iguaçu, Paraná, Brasil, 2003. [23] A.I. Vogel, Elementary practical organic chemistry, part III: quantitative organic analysis, Longman, United States, 1958. [24] L.C.M. Cantagesso, L.A. Jermolovicius, R.B. do Nascimento, E.R. Castro, J.T. Senise, “Processo irradiado por micro-ondas para produção de biodiesel,” presented at the 2º Seminário de Pesquisa do Grande ABC, São Bernardo do Campo, São Paulo, Brasil, 2012. [25] D.D. Bala, D. Chidamdaram, Enhancing kinetics of biodiesel production using Morpholine, Catalysis Communications 83 (2016) 48-52. [26] P. Nautiyal, K.A. Subramanian, M.G. Dastidar, Kinetic and thermodynamic studies on biodiesel production from Spirulina platensis algae biomass using single stage extraction-transesterification process, (2014) Fuel 135 228-234. [27] J. Li, X. Peng, M. Luo, C. Zhao, C. Gu, Y. Zu, Y. Fu, Biodiesel Production from Camtotheca acuminate seed oil catalyzed by novel Brönsted-Lewis acidic ionic liquid, Applied Energy 115 (2014) 438-414. [28] R.N. Rangel, Práticas de Físico-Química, 3rd ed, Edgard Blücher, São Paulo, Brazil, 2006. [29] ABNT (Brazilian Association of Technical Norms) MB-75, Determinação de índice de saponificação para óleos e gorduras, 1951. [30] R.E. Bruns, I.S. Scarminio, B.B. Barros Neto, Planejamento e Otimização de Experimentos. Editora da Universidade de Campinas, São Paulo, Brazil, 1995. [31] A.B. Ara, A.V. Musetti, B. Schneiderman, Introdução à Estatística, Edgard Blücher, São Paulo, Brasil, 2003. [32] P.R.C. Neto, L.F.S. Rossi, G.F. Zagonel, L.P. Ramos, Produção de biocombustível alternativo ao óleo diesel através da transesterificação de óleo de soja usado em fritura, Química Nova 4 (2000) 531-537. [33] R. Geris, A.N. Santos, A.B. Amaral, S.I. Maia, D.V. Castro, R.J. Carvalho (2007) Biodiesel de soja: reação de transesterificação para aula práticas de química orgânica, Química Nova 30 (2007) 1369-1373. [34] K. Chen, Y. Lin, K. Hsu, H. Wang, Improving biodiesel yields from waste cooking oil by using sodium methoxide and a microwave system, Energy 38 (2012) 151-156. [35] Y. Lin, P. Yang, S. Chen, J. Lin, Improving biodiesel yields from waste cooking oil using ionic liquids as catalysts with a microwave heating system, Fuel Processing Technology 115 (2013) 57-62.
[36] L.A. Jermolovicius, J.T. Senise, E.R. Castro, “Microwave enhanced kinetics of pseudo-irreversible esterification of maleic anhydride with 2-Ethylhexanol,” presented at the 10th International Conference on Microwave and High Frequency Heating, Modena, Italy, 2005. [37] O. Levenspiel, Engenharia das Reações Químicas, third ed., Edgard Blücher, São Paulo, Brazil, 2000. [38] W.G. Cochran, G. M. Cox, Experimental designs, second ed., John Wiley, New York, United States, 1966. [39] S.W. Benson, O. Dobis, Existence of negative activation energies in simple bimolecular metathesis reactions and some observations on too-fast reactions, J. Phys. Chem. A 102 (1998) 5175-5181. [40] T. Shimomura, K.J. Tolle, J. Smid, Energy and entropy of activation of propagation by the free polystyryl anions and their ion pairs. The Phenomenon of "Negative" Activation Energy, J. Amer. Chem. Soc. 89 (1967) 796-803. [41] H. Langhals, S. Christian, A. Hofer, Substitution of aromatics by amines at room temperature with negative energy of activation: amino peri-arylenes as metal-free components for dye-sensitized solar cells, The Journal of Organic Chemistry 78 (2013) 9883-9891. [42] K.N. Houk, N.G. Rondan, Origin of negative energies and entropy control of halocarbene cycloadditions and related fast reactions, J. Amer. Chem. Soc. 106 (1984) 4293-4294. [43] M. Mozurkewich, J.L. Lamb, S.W. Benson, Negative activation energies and curved Arrhenius plots. 2. Hydroxyl + carbon monoxide, J. Phys. Chem. 88 (1984) 64356441. [44] F.C. Meunier, G. Yalonsky, D. Reid, Negative Apparent Kinetic Order in SteadyState Kinetics of the Water-Gas Shift Reaction over a Pt–CeO2 Catalyst, Catalysis Today 138 (2008) 216-221. [45] V.R. Gangwal, J. Van Der Schaff, B.F.M. Kuster, The effect of mass transport limitation on the estimation of intrinsic kinetic parameters for negative order reactions, Applied Catalysis A: General 274 (2004) 275-283. [46] M. Morbidelli, A. Varma, Isothermal Diffusion-Reaction in a Slab Catalyst with Bimolecular Langmuir-Hinshelwood Kinetics Connections with Negative FirstOrder Kinetics, Chem. Eng. Science 36 (1983) 289-296. [47] http://scopus.com, 2017 (accessed 17.1.17). [48] P. Verma, M.P. Sharma, Review of Process Parameters for Biodiesel Production from Different Feedstocks, Renewable and Sustainable Energy Reviews 62 (2016) 1062-1071. [48] G. Baskar, A. Gurulladevi, T. Nishanthini, R. Aiswarya, K. Tamilarasan, Optimization and kinetics of biodiesel production from mahua oil using manganese doped zinc oxide nanocatalyst, Renewable Energy 103 (2017) 641-646. [49] T. Lieu, S. Yusup, M. Moniruzzaman, Kinetic study on microwave-assisted esterification of free fatty acids derived from Ceiba pentandra seed oil, Bioresource Technology 211 (2016) 248-256. [50] M. Kumar, M.P. Sharma, Kinetics of transesterification of Chlorella protothecoides microalgal oil to biodiesel, Waste Biomass Valor, 7 (2016) 1123-1130. [51] I. Sreedhar, Y. Kishan, Process standardization and kinetics of ethanol driven biodiesel production by transesterification of Rice bran Oil, Int. J. Ind. Chem. 7 (2016) 121-129.
[52] I. Reyero, G. Arzamendi, S. Zabala, L. Gandía, Kinetics of the NaOH – Catalyzed transesterification of sunflower oil with ethanol to produce biodiesel, Fuel Processing Technology 129 (2015) 147-256. [53] J. Chen, X. Hu, X. Wang, Y. Xu, E. Hu, Kinetic investigations of biodiesel from cottonseed oil and ethanol by transesterification in biomaterial and its application, Advanced Materials Research 578 (2012) 73-77. [54] L.J. Konwae, J. Wärna, P. Mäki-Arvela, N. Kumar, J. Mikkola, Reaction kinetics with catalyst deactivation in simultaneous esterification and transesterification of acid oils to biodiesel (FAMA) over a mesoporous sulphonated carbon catalyst, Fuel 166 (2016) 1-11. [55] P. Patil, V.G. Gude, S. Pinappu, S. Deng, Transesterification Kinetics of Camelina sativa Oil on Metal Oxide Catalysts Under Conventional and Microwave Heating Conditions, Chemical Engineering Journal 168 (2011) 1296-1300. [56] V. Cole, P.R. Cogate, Intensification of glycerolysis reaction of higher free fatty acid containing feedstock using microwave irradiation, Fuel Processing Technology 118 (2014) 110-116. [57] B.G. Terigar, S. Balasubramanian, M. Lima, D. Boldor, D. Transesterification of soybean and rice bran oil with ethanol in a continuous-flow microwave-assisted system: Yields, quality and reaction kinetics. Energy & Fuels 24 (2010) 6609-6615. [58] L. Wu, T. Wei, Z. Lin, Y. Zou, Z. Tonge, J. Sun, Bentonite enhanced biodiesel production by NaOH – catalyzed transesterification: Process optimization and kinetics and thermodynamic analysis, Fuel 182 (2016), 920-927. [59] W. Ye, Y. Gao, H. Ding, M. Liu, S. Liu, X. Han, J. Qi, Kinetics of transesterification of palm oil under conventional heating and microwave irradiation using CaO as heterogeneous catalyst, Fuel 180 (2016) 574-579.
Fig. 1. Preparation of biodiesel using microwaves.
Fig. 1. Sequence of reactions for indirect dosage of biodiesel and its yield determination.
Fig. 2. Preparation of biodiesel by conventional process.
Fig. 3. Biodiesel stoichiometric yields for different processes at long operation time.
(a)
(b)
Fig. 4. Examples of empirical kinetics curves for production of biodiesel under microwave irradiation. (a) point 5 (+ + ), maximum velocity. (b) point 8 ( ), minimum velocity.
Table 1: Simplex evolution towards an optimum operation point. CONTROLLED TEST # VARIABLES 1
2
3
4
5
6
7
8
Optimum point
relation ethanol
3.08
3.08
3.08
3.08
2.68
2.88
2.78
3.23
3.08
Catalyst concentration (g/g)
3.20
3.20
3.20
2.39
3.00
3.92
3.46
3.89
3.46
Specific microwave power (W/g)
1.96
1.96
1.69
1.87
1.87
1.87
1.74
1.76
1.76
Reaction time (min)
7.50
2.50
5.00
5.00
5.00
5.00
8.75
8.13
7.50
Mean biodiesel yield (%)
98.83
96.90
98.59
94.98
95.94
98.12
97.95
99.28
99.20
Standard Deviation
0.85
0.75
1.66
1.39
1.73
0.35
1.06
0.87
1.52
Molar between and oil
RESULTS
Table 2: Biodiesel yields for 7.5 min running different processes. Biodiesel stoichiometric yield (%) Microwave Conventional heating heating Acid process Alkaline Acid process process 99.43 57.99 61.80 99.47 58.47 62.10 99.66 58.93 63.10 99.34 60.99 62.58 98.82 60.08 61.42 98.54 58.97 60.84 99.60 62.20 63.58 99.02 62.20 62.42 Mean (and Standard Deviation) 99.24 (0.40) 59.98 (1.66) 62.23 (0.89)
Table 3: Parameters for chemical kinetic determination. Natural parameters for equation (7) Test
1 2 3 4 5 6 7 8
Soya oil Ethanol Temperature concentration concentration inverse CA (molkg-1)
CB (molkg-1)
1/T (K-1)
z1
z2
z3
3.90 3.74 3.90 3.74 3.90 3.74 3.90 3.74
1.56 1.56 0.69 0.69 1.56 1.56 0.69 0.69
0.0016 0.0016 0.0016 0.0016 0.0015 0.0015 0.0015 0.0015
Table 5: Results of reaction initial velocity. reduced parameters Test
Soya oil Ethanol Temperature concentration concentration inverse
Results of reaction initial velocity (mol/kgmin)
1 2 3 4 5 6 7 8
x1
x2
x3
Y1
Y2
+ + + + -
+ + + + -
+ + + + -
0.0868 0.1076 0.1127 0.0631 0.1210 0.0888 0.0867 0.0577
0.0898 0.1056 0.1108 0.6007 0.1234 0.0864 0.0836 0.0548
Table 6: Published data for biodiesel production kinetics.
Triglyceride source
Alcohol
Catalyst phase
Catalyst active compound
Reaction order Global
Mahua oil Soyabean oil
methanol heterog. methanol
both
Mn doped ZnO
Annona squamosa Chlorella protothecoides Rie bran oil Canola oil and corn oil Sunflower
metahnol heterog. methanol homog. ethanol heterog. methanol homog. ethanol homog.
Bentonite and NaOH Char from Ceiga pentranda KOH CaO NaOH and morpholine NaOH
Cottonseed oil Mesua ferrea Linn oil
ethanol homog. methanol heterog.
KOH Sulfonated carbon
Velocity constant (*)
Preexponential factor
Triglyceride
1 pseudo 1
Activation energy
Kinetic method adopted
Reference
181.91
graphical
[48]
31.03
graphical
[58]
25,723
[40] [50] [51] [25] [52] [53] [54]
(kJ/mol) 0.090.21
1 1 pseudo 1 pseudo 1 2 1.291.54 2
0.111
48.7
graphical graphical graphical graphical graphical
0.1687
21,607 39
graphical graphical
0.034 0.0103
86,091
(*) min-1 for first order; Lmol-1min-1 for second order
Table 7: Published data for biodiesel production kinetics under microwave irradiation. Triglyceride source
Alcohol
Catalys t phase
Catalyst active compound
Reaction order
Preexponenti al factor
Global
Triglycerid e
Alcoho l
BaO
3
2
1
5,195
SrO
2
2
0
1,584
2
1
0.00058
Camelina sativa oil
methan ol
heterog .
Palm oil
methan ol
heterog .
CaO
3
methan ol
homog.
Metal sulfates and ionic liquid
1
0.00217
ethanol
homog.
KOH
2
0.012
methan ol
homog.
H2SO4
pseud o2
3.98E9
Camptothec a acuminata seed oil Nagchampa oil Ceiba pentandra seed oil
Soyabean oil
4.48 ethanol
Rice bran oil
homog.
NaOH
Activatio n energy
Kinetic method adopted
Microwave furnace
Referenc e
graphical
Domestic
[55]
graphical
Scientific in domestic style
[59]
graphical
Scientific in domestic style
[27]
graphical
Domestic
[56]
53,717
graphical
Scientific in domestic style
[49]
11,147
assumed / graphical
Scientific in domestic style adapted for continuous flow (CSTR) Scientific in domestic style adapted for continuous flow (CSTR)
[57]
(kJ/mol)
37.6
1 4.8
6,334
assumed / graphical
S0255-2701(16)30209-4 http://dx.doi.org/doi:10.1016/j.cep.2017.03.010 CEP 6946
To appear in:
Chemical Engineering and Processing
Received date: Accepted date:
15-7-2016 21-3-2017
Please cite this article as: Luiz A.Jermolovicius, Luana C.M.Cantagesso, Renata B.do Nascimento, Edmilson R.de Castro, Eduardo V.dos S.Pouzada, Jos´e T.Senise, MICROWAVE FAST-TRACKING BIODIESEL PRODUCTION, Chemical Engineering and Processinghttp://dx.doi.org/10.1016/j.cep.2017.03.010 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
MICROWAVE FAST-TRACKING BIODIESEL PRODUCTION
Luiz A. JERMOLOVICIUS, Luana C. M. CANTAGESSO, Renata B. do NASCIMENTO, Edmilson R. de CASTRO, Eduardo V. dos S. POUZADA, José T. SENISE
Laboratório de Micro-ondas, Centro Universitário do Instituto Mauá de Tecnologia, Praça Mauá, 1 – São Caetano do Sul, SP 09580-900, BRAZIL; e-mail: [email protected], [email protected]
Highlights
Complete kinetic equation justify transesterification Independence from ethanol. Transesterification microwave enhancement is done by changing kinetics parameters. Kinetic equation under microwave irradiation induces operational optimization. Microwave heated process higher productivity was quantified and explained.
Keywords: biodiesel, microwaves, process intensification
1. Introduction Biodiesel is commonly produced by the process of transesterification of triglycerides such as soybean oil, palm oil or any other vegetable oil with ethanol or methanol in the presence of a catalyst. Transesterification can be performed in batch or continuous flow using conventional heating [1]. Biodiesel is a good alternative to be used in vehicles based on Diesel cycle engines because its molecular structure presents a high cetane index [2]. As a fuel based on renewable sources biodiesel can contribute to reduce the intensity of global warming. Its use brings a lot of benefits, say, it is biodegradable, renewable, and capable of promoting sustainable economic development. All these features can reduce the overall dependence on fossil fuels. Nevertheless and despite the high investments in biodiesel plants, like the Brazilian Caramuru Food Co. that opened a 225 million liters manufacturing unit of biodiesel in 2010, corresponding to an investment of 54 million dollars [3], the Secretariat for Planning and Strategic Investments of Brazilian Federal Government Ministry of Planning alerted that the production of biodiesel was economically unfeasible. In fact, the production of biodiesel is more expensive than ordinary diesel, and even today it is not sustainable from an economic point of view [4].
According to Government analysts, the problems plaguing production of biodiesel are focused on two basic issues: (1) the selection of more suitable raw materials for biodiesel production, and (2) infrastructure that enable the commercial exploitation of these raw materials. This point of view is also shared with other countries, where the costs of raw materials become 70 – 95 % of biodiesel production cost [5]. Within this context, a third factor can be introduced: the transesterification technology adopted – predominantly the alkaline catalysis [2, 5, 6, 7, 8, 9] – is not efficient enough to produce a competitive biodiesel. Since 2006 there are more than 650 deposited patents for biodiesel production and only 95 of them are about acid catalyzed processes [10]. In this scenario, it is clear the necessity of biodiesel technology enhancement. The current raw materials costs and their logistic issues escape from the scope of this work, but the technology for biodiesel production is its objective. A first option aiming better processes is the process intensification. It is an innovative and revolutionary focus on chemical processes that allows improvement of productivity and reduction of operation costs [11, 12, 13] and, obviously, it is a nice option to biodiesel production. Analyzing the production of biodiesel under conventional heating, it is observed that the energy necessary to fulfill the reaction is transferred from the hot surface of the reactor by conduction and convection, establishing a thermal gradient between the hot surface and the mass of reactants. This structure makes the velocity of the reaction dependent on heat transfer efficiency and it brings troubles in the process productivity. A second option to improve a chemical process is the adoption of microwave heating instead of conventional heating [14, 15]. Nowadays, application of microwave energy in synthesis processes is a well known alternative to increase the velocity of reactions [16, 17, 18, 19] as it was initially observed in procedures for digestion of samples under microwaves [20]. Basically, microwaves are
electromagnetic energy that can heat polar media. Polar molecules tend to align with the electromagnetic field and this phenomenon produces heat due to the friction caused by the slower reorientation of the molecules compared with the time rate change of the field. There is also a non-thermal specific microwave effect: uncoupling the spin of electrons in the atoms [16, 17, 21], which leads to new mechanisms models generally faster than the conventional ones. Following this approach, an earlier experiment showed a strong synergic effect between microwaves and acid catalysis in esterification maleic anhydride with 2-ethylhexanol-1 [22], lowering the chemical reaction order and the activation energy. It is interesting to observe that since 2006 there were 18 patented processes related to microwave assisted biodiesel production [10]. All these facts induce to think about a faster biodiesel process based in acid catalysis assisted by microwaves. The experiment reported in this work was done aiming at a better understanding of microwave enhanced acid process to biodiesel production from vegetable oils.
2. Microwave batch reactor and its operation As any microwave device the reactor used in this work is modeled at the macromolecular level by Maxwell’s equations. Each particular solution is obtained by applying the electromagnetic boundary conditions which are forced by the chosen geometry and materials that interacts with the electric and magnetic fields. In practice some problems arose that can reduce efficiency of an applicator: (1) good control of geometric aspects, and (2) knowledge of the constitutive parameters that model the materials, say the relative complex permittivity r and, sometimes, the relative complex permeability µr when magnetic materials are involved. The constitutive parameters are frequency and temperature dependents; their values are difficult to be found in the literature, so one
should measure them with some degree of confidence. The geometric issue is also important because of the interaction between materials and electromagnetic fields. In this work a large geometry was used – in the sense that it comprises several wavelengths in all the three spatial dimensions. This framework does not allow closed analytical solutions, so numerical ones must be sought by computer simulation. This approach was used in this work. Fig. 1 shows the microwave batch reactor used to process the transesterification of soya oil with ethanol. It consists of a 500-mL Pyrex kettle vessel with reflux condenser and mechanical stirrer. They are inside a multimodal cavity with a mode stirrer, and coupled to a 2.45 GHz microwave generator with adjustable power up to 3 kW. The setup allows connection of power sensors to measure the forward (irradiated) and reflected microwave power to quantify the value of the effective power transmitted to reacting materials. The reflected power is an inherent loss in microwave applicators and represents how much microwave energy the system was not able to absorb and reflects it back. This reduction is the target design of microwave chemical reactors. Temperature measurement was done with a fiber optic temperature sensor protected by a thermometric well immersed in reacting material. In a typical test the reactor is initially loaded with methane sulfonic acid and ethanol. Then soy oil, previously heated at 70° C, is added under stirring. The reagents quantities are defined for each kind of tests as shown below.
3. Biodiesel yield determination The dosage of the produced biodiesel was done by an indirect method based on the determination of the glycerin formed during transesterification. Numerical results are then justified by stoichiometric relation between biodiesel and glycerin. Although this
reaction is formally reversible, it was assumed as pseudo irreversible when processed with excess of alcohol. The dosage of glycerin is an iodometric method based on its reaction with periodic acid (traditional Malaprade method [23]) and final dosage of iodine with sodium thiosulfate. Fig. 1 shows the sequence of reactions that occur in this procedure. In the usual Malaprade procedure a small portion of sodium lauryl sulfate was introduced with the aim to produce a homogeneous solution, emulsifying oil and biodiesel. The analytical procedure [24] is: weight around 0.5 g of biodiesel production medium (soya oil, ethanol in excess, catalyst, biodiesel, glycerin) with high precision; add 1 mL of 5 % sodium lauryl sulfate solution; add 50 mL of 0.1 N periodic acid solution; rest the reactants for 120 minute under continuous swirling; add 30 mL of 20 % potassium iodide solution and 25 mL of 6 N sulfuric acid; titrate with 0.1 N sodium thiosulfate until pale yellow color is obtained and add 2 mL of soluble starch solution. Then titrate until the blue color disappear. In sequence, use a blank titration without glycerin, to standardize the periodic acid solution.
Knowing that one mol of glycerin represents one mol of biodiesel, then the stoichiometric yield of biodiesel may be determined by Eq. 14. It is based on chemical reaction stoichiometry relations between reactants and products and represents how much of the theoretical amount of desired products was really produced. It is common in scientific literature the use of a technical index based on the relation between the produced mass and the mass of limiting raw material [25, 26, 27] instead of the yield based on chemical stoichiometry. This technical index does not represent what happens in terms of chemical reactions; it only reports a relation of products, but it is much more easy to use than the stoichiometric approach.
𝑀𝐺𝐴 = 𝑀𝐺𝑇 = 𝑀𝐵 = 𝐵𝑌 =
𝑀𝑀𝐺 ∙𝑁𝑇𝑆 ∙10−3 ∙(𝑉𝑇𝑆 𝑂 − 𝑉𝑇𝑆 𝑅 ) 4∙𝑀𝐴 𝑀𝐺𝐴 ∙𝑀𝐴𝑇
𝑀𝐴 3∙𝑀𝑀𝐵 ∙𝑀𝐺𝑇 𝑀𝑀𝐺 𝑀𝐵 ∙𝑀𝑀𝑂 ∙100 𝑀𝑂 ∙𝐼∙3∙𝑀𝑀𝐵
(1) (2) (3) (4)
where MGA stands for the mass of glycerin in a sample, MGT is the mass of total glycerin in an experiment, MB is the prepared mass of biodiesel in an experiment, and BY is the biodiesel stoichiometric yield in an experiment. In (1)(4) MMG is the glycerin molar mass (92,09 gmol-1), MMB is the biodiesel molar mass (248.14 gmol-1), MMO is the soya oil sample molar mass (781.09 gmol-1), MA is the sample mass (g), MGA is the mass (g) of glycerin in the sample mass, MGT is the glycerin total mass (g) of an experiment, MAT is the total mass (g) of an experiment, MB is the biodiesel mass (g) of an experiment, MO is the soya oil mass (g) of an experiment, I is the correction factor for soya oil based on actual saponification, NTS is the sodium thiosulfate normality, VTS 0 is the volume of thiosulfate in standardization titration (mL), VTS R is the volume of thiosulfate in sample titration (mL), and BY is the biodiesel stoichiometric yield (%). In the above calculations, the average molar masses of soya oil and biodiesel were determined by cryoscopy method [28] and taken from literature in the case of glycerin. The purity correction factor for soya oil was determined in function of its actual molar mass (781.09 g.mol-1) and the relation between its actual and theoretical saponification index [29]. The actual value was determined by saponification with excess of alcoholic solution of potassium hydroxide and titration of excess with hydrochloric acid [24]. The correction factor for oil purity is the relation between the actual saponification index (189,59 mg KOH.g-1) and the theoretical saponification index for this molar mass of oil (215,51 mg KOH.g-1).
4. Defining the operational conditions Here the objective is the identification of the region defined by the range of the controlled variables that can produce the highest stoichiometric yield in biodiesel. They were selected as (1) the molar relation between ethanol and soya oil, (2) catalyst concentration (mass of catalyst per mass of oil), (3) specific absorption rate (W/g) which is directly associated with the processing temperature, and (4) reaction time (min). The catalyst adopted was methane sulfonic acid. The Simplex method [30] was used for optimization. All tests were done in duplicate. Table 1 shows the initial Simplex (columns 1 to 5). Analyzing the results form tests 1 to 5, the worst result (column 4) was eliminated and substituted by its symmetrical point (column 6). Proceeding with the analysis of the new result and comparing it to the former, the new worst result (column 2) was eliminated and substituted by its symmetrical counterpart (column 7). Once again, the worst (point 5) was substituted by column 8, and so on. When column 6 was substituted, the new result became the worst. Then, columns 1, 3, 6, 7, and 8 define the optimum region and its central point was adopted as the optimum point for this microwave heated acid biodiesel process.
5. Comparison of conventional and microwave processes The objective of this comparison between microwave assisted biodiesel process and a conventional heating alkaline biodiesel process is to confirm which of them has higher productivity. Analysis of variance [31] was used for this purpose. Considering that all processes may produce high stoichiometric yield but not exactly at the same time, then a comparison of results from a conventional heating alkaline process and the microwave heated acid process was done at unique establishing time of operation, specifically 7.5 min.
Excluding a natural doubt if the effect on biodiesel production is due to the acid catalyst or to the microwave energy, another universe was introduced in this comparison: a conventional heating acid process performed at the same conditions of the microwave irradiated process but with conventional electric heating. Usual laboratory equipment (cf. Fig. 2) was used for this process. The data set for conventional heating alkaline process was prepared in usual laboratory equipment, as shown in Fig. 2, and supported by published procedure [32; 33]. This process used 120 g of soybean oil with 781.09 average molar mass and average 189.59 mg KOH/g saponification index, ethanol and sodium methoxide as catalyst, with a 20minute heating period at 45 C by an electric blanket. The reaction control was the same as adopted for the microwave-assisted process described above. It is important to observe that the running time for these tests was reduced to 7.5 min; that is the time necessary for the microwave process to get its maximum yield. Results are shown in Table 2. It can be observed that the reduction of processing time also reduced the yield of biodiesel in the conventional heating processes, because they are slower than the microwave one which achieves more than 99 % of yield. Analysis of variance applied to Table 2 data and F distribution [31] adoption resulted in a critical value Fcrit = 9.34, with 99.9 % of confidence and Φ1 = 2 and Φ2 = 21 degrees of freedom, against a calculated factor Fcalc = 21.00. This means that one of the universes analyzed is different with 99.9 % of confidence. Observing the data, it is clear that the different universe is the microwave heated process. Completing the analysis, a t Student’s Test was applied to conventional heating processes data. It resulted in a critical factor t crit = 1.721 with degree of freedom Φ = 21 and 10 % of confidence and a calculated factor t calc = 0.875. This means that is not a significant difference between conventional heating under alkaline or acid processes.
Longer running time experiments (20 minutes) were also performed in the same conditions as described above. The results of biodiesel stoichiometric yield of this new series are plotted in Fig. 3, where a normalized time scale was adopted, and defined by the relation between effective running time and conventional alkaline process running time to reach their maximum yield. This make a nice confirmation that the microwave heated process needs less time to reach its maximum yield in comparison with the conventional processes described above. Effectively, the process velocity increased 42.5 %.
These data confirm that the increase of biodiesel yield may be attributed to a microwave effect, not to a kind of acid or alkaline catalyst. Moreover, one can think about doubling a rector’s productivity or half-fold its original volume. The experimental mean of stoichiometric yield was 99.24 % (cf. Table 2), a higher value compared to published biodiesel yields of 95.5 % [34] and 97.1 % [35] for microwave irradiated transesterification of waste cooking oil. Electrical energy consumption showed interesting figures too: 30.9 Wh, 64.5 Wh, and 75.0 Wh for the microwave, the conventional acid, and the conventional alkaline process, respectively. These values were measured with a wattmeter. Consequently, the microwave process shows an economy of 58.8 % or 52.1 % in comparison with the conventional alkaline or the conventional acid process, respectively.
6. Trying to understand the microwave effect Among the three processes considered in this work, the microwave heated has the highest efficiency, but there is no explanation about it, yet.
Considering that the yield of a process is partially controlled by the reaction chemical kinetic, then it is possible to suppose that microwave energy impacts the original reaction mechanism anyway, thus accelerating the process. It stands to reason that a complete kinetic equation determination for the microwave assisted acid process was done. The procedure for complete chemical equation determination [22; 36] was based on initial rate methods [37] combined with a multilinear regression [38]. It starts with knowledge of two empirical facts: a) that the triglycerides transesterification with ethanol may be represented by an equilibrium reaction; b) that the reversibility of this reaction is not strong from the experimental point of view, so it can be conceived as practically irreversible in laboratories, then it is possible to assume it as a pseudo irreversible reaction. This leads to W = K0·exp[E/(R.T)]·CAn·CBmCCp
(5)
In (5) W denotes the reaction rate (mol/kgmin), K0 is the pre-exponential Arrhenius factor, E is the activation energy (J/mol), R is the ideal gas constant (8.31446 J/Kmol), T is the temperature (K), n and m are the reaction orders in respect to triglycerides and ethanol, respectively, CA, CB and CC are the concentrations of triglycerides, ethanol and catalyst, respectively (usually measured in mol/L, but in this case adopted as mol/kg), and exp is the base of natural logarithms ( 2.71828). The unit for concentration adopted was the molal (mol/kg) to avoid troubles with reproducibility with sampling hot reacting material and incomplete physical separation of glycerol from biodiesel. It is also important to observe that no solvents were used; the reaction solution was just soya oil and ethanol plus a catalyst. The catalyst methane sulfonic acid concentration was optimized at the start of each experiment and was adopted to be constant, then the product K0CCp becomes constant (In this paper it is denoted as a pseudo constant k0).
Eq. (5) has its intricacies in modeling the phenomenon: it has two variables in potential form and a third variable in the exponential form. It is usual to plot experimental kinetic data in graphics which represents a first, second or third order [37]. This is not an exact approach, but it helps the determination of a complete chemical kinetic equation. Levenspiel presents a method based on initial velocities of reactions that can be adopted for a multilinear regression by linearizing Eq. (5) with the help of logarithms. The linearized equation becomes: log(W) = log(k0) – [E/(R.T)]·log(e) + n·log(CA) + m·log(CB)
(6)
Eq. (6) is linear. It is easily observed after renaming their variables as below: Y = a0 + a1·z1 + a2·z2 a3·z3,
(7)
where Y = log(W), z1 = log(CA), z2 = log(CB), z3 = 1/T, a0 = log(k0), a1 = n, a2 = m, and a3 = (E/R)log(e). In order to apply a statistical regression method, Eq. (7) was transformed into reduced parameters (xi) Y = b0 + b1·x1 + b2·x2 + b3·x3,
(8)
where b0, b1, b2, b3 are the regression coefficients, and x1, x2, x3 are the reduced parameters calculated by: xi = (zi – zi0) / Δzi ,
(9)
where zi is the natural value of a variable, zi0 is its natural central value, and Δzi is the associated incremental. The temperatures adopted for this experiment were the equilibrium temperatures when the reactants were irradiated with 300 W of microwave (347.6 oC) and 400 W (384.0 oC). These microwave effective powers were selected to attend equipment operational requirements. The central value and increment change for soya oil were 3.82 and 0.080
mol/kg respectively, and for ethanol they were 1,12 and 0,43 mol/kg respectively. Table 3 shows typical values in natural and reduced form. The order of execution of duplicated tests was random without restrictions. They produced sixteen kinetic curves as, for example, the curves for the points of maximum and minimum velocities shown in Fig. 4. The results of initial velocity of reaction (Y1 and Y2) are shown in Table 5. Their values were determined by numerical derivative taken at time zero of the experimental kinetics curves obtained with Table 3 reaction conditions.
It is very interesting to observe that curves for low level of soya oil concentration plus another variable in low level produced a kinetic curve similar to Fig. 4b and was the slower reactions. Points with high level of soya oil concentration and point 2 (lower soya oil concentration, but higher value for ethanol and temperature) produced curves similar to 4a and they were faster. There are two differences between curves 4a and 4b; one is the initial rate of reaction and the second is an inflexion at 3 min. instant time. In a first view, this inflexion may be related to partial substitution of carboxylic acids from glycerin residue in oil molecule. It is an induction that the first substitution of fatty acid in oil molecule is faster than the others. Multilinear regression applied to data taken from Table 5 results in Y = 0.986 0.0143·x1 + 0.000681·x2 0.0618·x3.
(10)
The significance of this regression was verified by using the Student’s t-distribution. The standard error was calculated with the worst replicate test, getting a value of se = 0.2231 and sbj = se.(m.N)-1/2 = 0.000139, where m is the number of data in one replication (m = 8) and N is the number of replications (N = 2). Supposing that all sbj are equal, we have
the values for tcalculated for each parameter tj = |bj|/sbj: t0 = 1768, t1 = 26, t2 = 1.2 and t3 = 111. For two degrees of freedom and 99.9 % of confidence the ‘t’ critical factor is 9.925, and for 95 % the critical factor is 4.303. The regression factor b2 has no significance with 95 % of confidence and all others have significance within 99.9 %. Consequently, Eq. (10) represents adequately the experimental data and can be rewritten in the following form: Y = 0.986 0.0143·x1 0.0618·x3.
(11)
The lack of fit of Eq. (11) was tested with Fischer’s variance rate. F = sg2/se2
(12)
where, sg2 is the mean square goodness-of-fit and se2 is the mean square error. Their values were 0.004897 and 0.002749, respectively, and Fischer’s factor was 1.78 which value is less than the critical factor (with 90 % of confidence and degree of freedom 3 and 8, Fcrit = 2.92). This means that the Eq. (11) fits well the experimental data. Using them, Eq. (11) and Eq. (7), it is possible to establish the values of the parameters of the chemical kinetic equation for the microwave heated acid biodiesel; they are: reaction order for soya oil, n1 = 1.57; energy of activation, E = 26526,71 J/mol; pseudo Arrhenius pre-exponential factor considering a constant catalyst concentration, k0 = 8.5 105. The complete chemical kinetic for soya oil transesterification with ethanol and methane sulfonic acid under microwaves is: W = 8.5105·exp[26527/(RT)]·CA1.6
(13)
The reaction order with respect to ethanol resulted in n2 = 0.0039, which is practically a zero-order reaction. It is interesting to observe that the regression coefficient b2 does not have significance because its parameter is of zero order, having no influence on the chemical reaction velocity. Summarizing, the biodiesel production under microwaves is
independent of ethanol concentration considering the experimental range of parameters adopted. Eq. (13) has two uncommon features: the activation energy and the reaction order related to soya oil concentration have negative values. The former means that the reaction rate will decrease with temperature increase [39; 40; 41; 42;43] and the latter means that the reaction velocity will decrease with the increase of concentration of this specific reagent [44; 45; 46]. Both facts induce the necessity to optimize the microwave enhanced transesterification conditions to achieve a better productivity, just to find the more effective temperature and soya oil concentration. The optimization done at beginning of this work is now justified by the kinetic behavior of microwave heated biodiesel process.
7. Analysis of kinetical data There was a strong increase in the number of publications about several kinetical aspects of biodiesel production after 2014. From 2001 to 2009 there were 31 papers (circa 3 per year), and from 2010 to 2016 there were 135 papers published (circa 20 per year) [47]. Table 6 shows some of the most recent papers about chemical kinetic equation. Also a review [48] with anterior data was analyzed. All papers report a kinetic equation determination by graphical agreement of experimental data to formal first or second order, as seen in chemical reactor design courses [37]. This method is not the most precise; it is good for an initial study, although insufficient for process design. In spite of having its operational difficulties the method described in Session 6 was adopted in this work because it can determine kinetic parameters without any previous definition of reaction order. Moreover, with this regression method it is possible to verify the significance of chemical kinetics equation parameters and the level of its fit to experimental data.
It was observed that first order reactions is predominant for different triglycerides, circa 48 % of the published papers. There are some articles reporting results of second order and even fractional order (between one and two). The papers report global orders or related to triglycerides concentration, but do not explain why alcohol concentration was not considered. The regression method showed that, under microwave irradiation, the alcohol concentration is not significant. The reason becomes clear when the reaction order value with respect to ethanol was calculated; it was practically zero which means independence of the reactant concentration. The order with respect to soya oil was determined as 1.6, which is greater than the orders of conventional biodiesel process, independently of the triglycerides source and the catalyst used. High order reactions imply higher velocities; this may justify the observed higher velocity of the microwave assisted transesterification. To the best knowledge of the authors it seems that there is no published information in the state of art about the negative order with respect to soya oil. This means that the increase of oil concentration results in reaction rate decrease. In other words, if the concentration of ethanol is lowered in a system oil / ethanol without solvent, then the velocity will decrease. The activation energy observed under microwave irradiation is lower than the conventional processes as can be seen in Table 6 and in Verma’s review [48]. It is the case for soya oil. This fact justify the higher reaction rate when the transesterification is done with microwave irradiation. There are some exceptions relating lower values of activation energy than that of the microwave enhanced process which may be attributed to natural reactivity of the specific triglycerides of these cases; they were rice, cottonseed, palm, rapeseed, and peanut oils.
Another information not available in publications of the state of the art is the negative activation energy. This means that the increase in reaction temperature will decrease the velocity for production of biodiesel. This fact may be attributed to temperature stimulation of reversible path of transesterification. Note that one premise usually adopted for this kinetic study is that the reversible path does not represent a significant influence on the process. There are some others papers that report experiments for transesterification of triglycerides under microwave irradiation and give kinetical data, as shown in table 7. As expected, all global orders were higher than one (except for one paper) and all activation energies were lower than that of conventional heating (again except for one paper). The high value of activation energy is for Ceiba pentandra seed oil [49], which probably react under elevated level of energy. The first order reported [57] may be explained by the fact that this value was assumed to determine kinetic parameters. In the case of soya oil, there is one publication for conventional heating [58] and other for microwave heating [57]. The former ratify that microwave heated process is superior than the conventional one. Unfortunately, the order of reaction was not determined empirically, but previously assumed in the latter work. Finally, the microwave equipment setup of Terigar’s work did not allow measurements of incident and reflected powers.
8. Conclusions This work confirms the higher productivity of microwave heated acid biodiesel production process, as shown in Fig. 3 and Table 2. From a process time point of view, a microwave process can increase the biodiesel production in 42 % (cf. Fig. 3). Microwave heated acid process is capable to reach the complete conversion of triglycerides in less operation time than the conventional heated process, as shown in
Table 2 and Fig. 3. In other words, microwave process is faster than conventional processes and for this reason, it is 58.8 % more economic in terms of electrical energy consumption than the conventional heated alkaline process, as shown in Session 5. With this information one may visualize a possible increase in biodiesel industry productivity by applying microwave energy as an alternative heat source. A complete chemical kinetic equation for transesterification of soya oil with ethanol under microwave irradiation was also determined as a result of this work (cf. Eq. (12)). It shows that the biodiesel microwave enhanced acid process is independent of ethanol concentration and it requires an optimization of soya oil and ethanol concentration and operation temperature, because of the negative value in the order of reaction with respect to soya oil, the zero order with respect to ethanol and a negative activation energy. These facts show that there is an optimal set of operational parameters for economical operation, as discussed in Session 6. Comparison
between
experimental kinetic
equation
for
microwave-enhanced
transesterification of soya oil with ethanol and published kinetic equations shows a lower energy activation than that of conventional heated processes and strong differences in reaction order values, as discussed in Session 7. Considering that a kinetic equation represents the reaction mechanism, it is possible to accept that microwave energy altered it in such a way that permits a new series of intermediate paths resulting in a faster reaction. Summarizing, switching from conventional heating to microwave heating changes the reaction mechanism and accelerates the transesterification, as shown, and promotes an operational time reduction to get the maximum yield with considerably energy economy. This is an example of a non-thermal microwave effect.
Acknowledgement The authors would like to thank D. Z. Passeti and C. Tognela. Supports from Instituto Mauá de Tecnologia – IMT and Fundação de Amparo à Pesquisa do Estado de São Paulo – FAPESP (Grant no. 2011/50154-9) are gratefully acknowledged.
References [1] M.U. Nascimento, S.C.A. Vasconcelos, B.E. Azevedo, C.F. Silva, Otimização da produção de biodiesel a partir de óleo de babaçu com aquecimento por micro-ondas, Eclética Química 34 (2009) 4. [2] G. Knothe, V.J. Gerpen, J. Krahl, Manual de Biodiesel, Edgard Blücher, São Paulo, Brazil, 2007. [3] F. Lopes, 3.9.10, Caramuru amplia sua produção de biodiesel. Valor Econômico. Available http://www.valoreconomico.com.br. [4] W.S. Dália, A produção do biodiesel: uma perspectiva para a agroenergia no nordeste brasileiro, O Futuro da Indústria: biodiesel – coletânea de artigos, Indústria e Comércio Exterior: IEL, 27-36, 2006. [5] P.P. Oh, H.L.N. Lau, J. Chen, M.F. Chong, Y.M. Choo, A Review on Conventional Technologies and Emerging Process Intensification (PI) Methods for Biodiesel Production, Renewable and Sustainable Energy Review 16 (2012) 5131-5245. [6] A. Pandey, Handbook of plant-based biofuels, CRC Press, Florida, United States, 2009. [7] Produção de combustíveis líquidos a partir de óleos vegetais, Fundação Centro Tecnológico de Minas Gerais, Programa de Energia, vol.2 ch. 1, Belo Horizonte, Minas Gerias, Brasil, 1983. [8] C.V. Ferreira, Otimização do Processo de Produção de Biodiesel a Partir de Óleos de Mamona e Algodão (Master thesis), Universidade Federal de Pernambuco, Brasil, 2007. [9] C.N. Khalil et al., “Biodiesel purification method,” U.S. Patent 20110209981, 1 st September, 2011. [10] European Patent Office. https://worldwide.espacenet.com/search, 2016 (accessed 25.6.16). [11] S. Becht, R. Franke, A. Geisselmann, H. Hahn, An industrial view of process intensification, Chemical Engineering Processing: Process Intensification 48 (2009) 329-332. [12] A. Stankiewicz, J.A. Moulin, Process intensification: transforming chemical engineering, Chemical Engineering Progress 50 (2000) 22-34. [13] D. Reay, C. Ramshaw, A. Harvey, A Process intensification. Engineering for efficiency. Sustainability and flexibility, Elsevier, United Kingdom, 2008. [14] A.C. Metaxas, R.J. Meredith, Industrial Microwave Heating, Peter Peregrinus, London, Great Britain, 1983. [15] G. Roussy, J.A. Pearce, Foundations and industrial applications of microwave and radio frequency fields (1995) India, John Wiley, 1995. [16] P. Lindström, J. Tierney, B. Wathey, J. Westman, Microwave assisted organic synthesis: a Review, Tetrahedron, 57 (2001) 9225-9283.
[17] J.T. Senise, L.A. Jermolovicius, Microwave chemistry: a fertile field for scientific research and industrial applications, (2004) Journal of Microwaves and Optoelectronics 3 (2004) 97-112. [18] C.O. Kappe, Microwave dielectric heating in synthetic organic chemistry, Chem. Soc. Rev. 37 (2008) 1127-1139. [19] C.R. Strauss, A Strategic, “Green” Approach to organic chemistry with microwave assistance and predictive yield optimization as core, enabling technologies, Australian J. Chem. 62 (2009) 3-15. [20] F.E. Smith, E.A. Arsenault, Microwave-assisted sample preparation in analytical chemistry, Talanta 45 (1996) 1207-1268. [21] A.L. Buchachenko, Recent advances in spin chemistry, Pure Appl. Chem. 72 (2000) 2243-2258. [22] L.A. Jermolovicius, B. Schneiderman, J.T. Senise, E.R. Castro, “Microwave synergic effect on maleic anhydride catalytic esterification with 2-Ethylhexanol-1,” presented at the International Microwave and Optoelectronics Conference, Foz do Iguaçu, Paraná, Brasil, 2003. [23] A.I. Vogel, Elementary practical organic chemistry, part III: quantitative organic analysis, Longman, United States, 1958. [24] L.C.M. Cantagesso, L.A. Jermolovicius, R.B. do Nascimento, E.R. Castro, J.T. Senise, “Processo irradiado por micro-ondas para produção de biodiesel,” presented at the 2º Seminário de Pesquisa do Grande ABC, São Bernardo do Campo, São Paulo, Brasil, 2012. [25] D.D. Bala, D. Chidamdaram, Enhancing kinetics of biodiesel production using Morpholine, Catalysis Communications 83 (2016) 48-52. [26] P. Nautiyal, K.A. Subramanian, M.G. Dastidar, Kinetic and thermodynamic studies on biodiesel production from Spirulina platensis algae biomass using single stage extraction-transesterification process, (2014) Fuel 135 228-234. [27] J. Li, X. Peng, M. Luo, C. Zhao, C. Gu, Y. Zu, Y. Fu, Biodiesel Production from Camtotheca acuminate seed oil catalyzed by novel Brönsted-Lewis acidic ionic liquid, Applied Energy 115 (2014) 438-414. [28] R.N. Rangel, Práticas de Físico-Química, 3rd ed, Edgard Blücher, São Paulo, Brazil, 2006. [29] ABNT (Brazilian Association of Technical Norms) MB-75, Determinação de índice de saponificação para óleos e gorduras, 1951. [30] R.E. Bruns, I.S. Scarminio, B.B. Barros Neto, Planejamento e Otimização de Experimentos. Editora da Universidade de Campinas, São Paulo, Brazil, 1995. [31] A.B. Ara, A.V. Musetti, B. Schneiderman, Introdução à Estatística, Edgard Blücher, São Paulo, Brasil, 2003. [32] P.R.C. Neto, L.F.S. Rossi, G.F. Zagonel, L.P. Ramos, Produção de biocombustível alternativo ao óleo diesel através da transesterificação de óleo de soja usado em fritura, Química Nova 4 (2000) 531-537. [33] R. Geris, A.N. Santos, A.B. Amaral, S.I. Maia, D.V. Castro, R.J. Carvalho (2007) Biodiesel de soja: reação de transesterificação para aula práticas de química orgânica, Química Nova 30 (2007) 1369-1373. [34] K. Chen, Y. Lin, K. Hsu, H. Wang, Improving biodiesel yields from waste cooking oil by using sodium methoxide and a microwave system, Energy 38 (2012) 151-156. [35] Y. Lin, P. Yang, S. Chen, J. Lin, Improving biodiesel yields from waste cooking oil using ionic liquids as catalysts with a microwave heating system, Fuel Processing Technology 115 (2013) 57-62.
[36] L.A. Jermolovicius, J.T. Senise, E.R. Castro, “Microwave enhanced kinetics of pseudo-irreversible esterification of maleic anhydride with 2-Ethylhexanol,” presented at the 10th International Conference on Microwave and High Frequency Heating, Modena, Italy, 2005. [37] O. Levenspiel, Engenharia das Reações Químicas, third ed., Edgard Blücher, São Paulo, Brazil, 2000. [38] W.G. Cochran, G. M. Cox, Experimental designs, second ed., John Wiley, New York, United States, 1966. [39] S.W. Benson, O. Dobis, Existence of negative activation energies in simple bimolecular metathesis reactions and some observations on too-fast reactions, J. Phys. Chem. A 102 (1998) 5175-5181. [40] T. Shimomura, K.J. Tolle, J. Smid, Energy and entropy of activation of propagation by the free polystyryl anions and their ion pairs. The Phenomenon of "Negative" Activation Energy, J. Amer. Chem. Soc. 89 (1967) 796-803. [41] H. Langhals, S. Christian, A. Hofer, Substitution of aromatics by amines at room temperature with negative energy of activation: amino peri-arylenes as metal-free components for dye-sensitized solar cells, The Journal of Organic Chemistry 78 (2013) 9883-9891. [42] K.N. Houk, N.G. Rondan, Origin of negative energies and entropy control of halocarbene cycloadditions and related fast reactions, J. Amer. Chem. Soc. 106 (1984) 4293-4294. [43] M. Mozurkewich, J.L. Lamb, S.W. Benson, Negative activation energies and curved Arrhenius plots. 2. Hydroxyl + carbon monoxide, J. Phys. Chem. 88 (1984) 64356441. [44] F.C. Meunier, G. Yalonsky, D. Reid, Negative Apparent Kinetic Order in SteadyState Kinetics of the Water-Gas Shift Reaction over a Pt–CeO2 Catalyst, Catalysis Today 138 (2008) 216-221. [45] V.R. Gangwal, J. Van Der Schaff, B.F.M. Kuster, The effect of mass transport limitation on the estimation of intrinsic kinetic parameters for negative order reactions, Applied Catalysis A: General 274 (2004) 275-283. [46] M. Morbidelli, A. Varma, Isothermal Diffusion-Reaction in a Slab Catalyst with Bimolecular Langmuir-Hinshelwood Kinetics Connections with Negative FirstOrder Kinetics, Chem. Eng. Science 36 (1983) 289-296. [47] http://scopus.com, 2017 (accessed 17.1.17). [48] P. Verma, M.P. Sharma, Review of Process Parameters for Biodiesel Production from Different Feedstocks, Renewable and Sustainable Energy Reviews 62 (2016) 1062-1071. [48] G. Baskar, A. Gurulladevi, T. Nishanthini, R. Aiswarya, K. Tamilarasan, Optimization and kinetics of biodiesel production from mahua oil using manganese doped zinc oxide nanocatalyst, Renewable Energy 103 (2017) 641-646. [49] T. Lieu, S. Yusup, M. Moniruzzaman, Kinetic study on microwave-assisted esterification of free fatty acids derived from Ceiba pentandra seed oil, Bioresource Technology 211 (2016) 248-256. [50] M. Kumar, M.P. Sharma, Kinetics of transesterification of Chlorella protothecoides microalgal oil to biodiesel, Waste Biomass Valor, 7 (2016) 1123-1130. [51] I. Sreedhar, Y. Kishan, Process standardization and kinetics of ethanol driven biodiesel production by transesterification of Rice bran Oil, Int. J. Ind. Chem. 7 (2016) 121-129.
[52] I. Reyero, G. Arzamendi, S. Zabala, L. Gandía, Kinetics of the NaOH – Catalyzed transesterification of sunflower oil with ethanol to produce biodiesel, Fuel Processing Technology 129 (2015) 147-256. [53] J. Chen, X. Hu, X. Wang, Y. Xu, E. Hu, Kinetic investigations of biodiesel from cottonseed oil and ethanol by transesterification in biomaterial and its application, Advanced Materials Research 578 (2012) 73-77. [54] L.J. Konwae, J. Wärna, P. Mäki-Arvela, N. Kumar, J. Mikkola, Reaction kinetics with catalyst deactivation in simultaneous esterification and transesterification of acid oils to biodiesel (FAMA) over a mesoporous sulphonated carbon catalyst, Fuel 166 (2016) 1-11. [55] P. Patil, V.G. Gude, S. Pinappu, S. Deng, Transesterification Kinetics of Camelina sativa Oil on Metal Oxide Catalysts Under Conventional and Microwave Heating Conditions, Chemical Engineering Journal 168 (2011) 1296-1300. [56] V. Cole, P.R. Cogate, Intensification of glycerolysis reaction of higher free fatty acid containing feedstock using microwave irradiation, Fuel Processing Technology 118 (2014) 110-116. [57] B.G. Terigar, S. Balasubramanian, M. Lima, D. Boldor, D. Transesterification of soybean and rice bran oil with ethanol in a continuous-flow microwave-assisted system: Yields, quality and reaction kinetics. Energy & Fuels 24 (2010) 6609-6615. [58] L. Wu, T. Wei, Z. Lin, Y. Zou, Z. Tonge, J. Sun, Bentonite enhanced biodiesel production by NaOH – catalyzed transesterification: Process optimization and kinetics and thermodynamic analysis, Fuel 182 (2016), 920-927. [59] W. Ye, Y. Gao, H. Ding, M. Liu, S. Liu, X. Han, J. Qi, Kinetics of transesterification of palm oil under conventional heating and microwave irradiation using CaO as heterogeneous catalyst, Fuel 180 (2016) 574-579.
Fig. 1. Preparation of biodiesel using microwaves.
Fig. 1. Sequence of reactions for indirect dosage of biodiesel and its yield determination.
Fig. 2. Preparation of biodiesel by conventional process.
Fig. 3. Biodiesel stoichiometric yields for different processes at long operation time.
(a)
(b)
Fig. 4. Examples of empirical kinetics curves for production of biodiesel under microwave irradiation. (a) point 5 (+ + ), maximum velocity. (b) point 8 ( ), minimum velocity.
Table 1: Simplex evolution towards an optimum operation point. CONTROLLED TEST # VARIABLES 1
2
3
4
5
6
7
8
Optimum point
relation ethanol
3.08
3.08
3.08
3.08
2.68
2.88
2.78
3.23
3.08
Catalyst concentration (g/g)
3.20
3.20
3.20
2.39
3.00
3.92
3.46
3.89
3.46
Specific microwave power (W/g)
1.96
1.96
1.69
1.87
1.87
1.87
1.74
1.76
1.76
Reaction time (min)
7.50
2.50
5.00
5.00
5.00
5.00
8.75
8.13
7.50
Mean biodiesel yield (%)
98.83
96.90
98.59
94.98
95.94
98.12
97.95
99.28
99.20
Standard Deviation
0.85
0.75
1.66
1.39
1.73
0.35
1.06
0.87
1.52
Molar between and oil
RESULTS
Table 2: Biodiesel yields for 7.5 min running different processes. Biodiesel stoichiometric yield (%) Microwave Conventional heating heating Acid process Alkaline Acid process process 99.43 57.99 61.80 99.47 58.47 62.10 99.66 58.93 63.10 99.34 60.99 62.58 98.82 60.08 61.42 98.54 58.97 60.84 99.60 62.20 63.58 99.02 62.20 62.42 Mean (and Standard Deviation) 99.24 (0.40) 59.98 (1.66) 62.23 (0.89)
Table 3: Parameters for chemical kinetic determination. Natural parameters for equation (7) Test
1 2 3 4 5 6 7 8
Soya oil Ethanol Temperature concentration concentration inverse CA (molkg-1)
CB (molkg-1)
1/T (K-1)
z1
z2
z3
3.90 3.74 3.90 3.74 3.90 3.74 3.90 3.74
1.56 1.56 0.69 0.69 1.56 1.56 0.69 0.69
0.0016 0.0016 0.0016 0.0016 0.0015 0.0015 0.0015 0.0015
Table 5: Results of reaction initial velocity. reduced parameters Test
Soya oil Ethanol Temperature concentration concentration inverse
Results of reaction initial velocity (mol/kgmin)
1 2 3 4 5 6 7 8
x1
x2
x3
Y1
Y2
+ + + + -
+ + + + -
+ + + + -
0.0868 0.1076 0.1127 0.0631 0.1210 0.0888 0.0867 0.0577
0.0898 0.1056 0.1108 0.6007 0.1234 0.0864 0.0836 0.0548
Table 6: Published data for biodiesel production kinetics.
Triglyceride source
Alcohol
Catalyst phase
Catalyst active compound
Reaction order Global
Mahua oil Soyabean oil
methanol heterog. methanol
both
Mn doped ZnO
Annona squamosa Chlorella protothecoides Rie bran oil Canola oil and corn oil Sunflower
metahnol heterog. methanol homog. ethanol heterog. methanol homog. ethanol homog.
Bentonite and NaOH Char from Ceiga pentranda KOH CaO NaOH and morpholine NaOH
Cottonseed oil Mesua ferrea Linn oil
ethanol homog. methanol heterog.
KOH Sulfonated carbon
Velocity constant (*)
Preexponential factor
Triglyceride
1 pseudo 1
Activation energy
Kinetic method adopted
Reference
181.91
graphical
[48]
31.03
graphical
[58]
25,723
[40] [50] [51] [25] [52] [53] [54]
(kJ/mol) 0.090.21
1 1 pseudo 1 pseudo 1 2 1.291.54 2
0.111
48.7
graphical graphical graphical graphical graphical
0.1687
21,607 39
graphical graphical
0.034 0.0103
86,091
(*) min-1 for first order; Lmol-1min-1 for second order
Table 7: Published data for biodiesel production kinetics under microwave irradiation. Triglyceride source
Alcohol
Catalys t phase
Catalyst active compound
Reaction order
Preexponenti al factor
Global
Triglycerid e
Alcoho l
BaO
3
2
1
5,195
SrO
2
2
0
1,584
2
1
0.00058
Camelina sativa oil
methan ol
heterog .
Palm oil
methan ol
heterog .
CaO
3
methan ol
homog.
Metal sulfates and ionic liquid
1
0.00217
ethanol
homog.
KOH
2
0.012
methan ol
homog.
H2SO4
pseud o2
3.98E9
Camptothec a acuminata seed oil Nagchampa oil Ceiba pentandra seed oil
Soyabean oil
4.48 ethanol
Rice bran oil
homog.
NaOH
Activatio n energy
Kinetic method adopted
Microwave furnace
Referenc e
graphical
Domestic
[55]
graphical
Scientific in domestic style
[59]
graphical
Scientific in domestic style
[27]
graphical
Domestic
[56]
53,717
graphical
Scientific in domestic style
[49]
11,147
assumed / graphical
Scientific in domestic style adapted for continuous flow (CSTR) Scientific in domestic style adapted for continuous flow (CSTR)
[57]
(kJ/mol)
37.6
1 4.8
6,334
assumed / graphical