- Email: [email protected]
Glycerol acetals as diesel additives: Kinetic study of the reaction between glycerol and acetaldehyde
Fuel Processing Technology 116 (2013) 182–188
Contents lists available at SciVerse ScienceDirect
Fuel Processing Technology journal homepage: www.elsevier.com/locate/fuproc
Glycerol acetals as diesel additives: Kinetic study of the reaction between glycerol and acetaldehyde I. Agirre ⁎, M.B. Güemez, A. Ugarte, J. Requies, V.L. Barrio, J.F. Cambra, P.L. Arias Chemical and Environmental Engineering Department, Engineering Faculty of Bilbao, University of the Basque Country (UPV/EHU), Alameda Urquijo s/n, 48013 Bilbao, Spain
a r t i c l e
i n f o
Article history: Received 4 February 2013 Received in revised form 16 April 2013 Accepted 15 May 2013 Available online xxxx Keywords: Glycerol Acetaldehyde Acetals Kinetics Amberlyst
a b s t r a c t Certain acetals can be produced from renewable resources (bioalcohols) and seem to be good candidates for different applications, such as oxygenated diesel additives. This paper addresses the production of acetals (5-hydroxy-2-methyl-1,3 dioxane and 4-hydroxymethyl-2-methyl-1,3 dioxolane) from glycerol and acetaldehyde using Amberlyst 47 acidic ion exchange resin. This ion exchange resin performed well, recording 100% selectivity toward acetal formation at a suitably high initial glycerol concentration. When the initial acetaldehyde concentration was significantly higher than the glycerol concentration, 2,4,6 trimethyl-1,3,5 trioxane was the main reaction product. Unlike other acetalization reactions, the one studied here does not have thermodynamic limitations, and 100% conversion is achieved under different reaction conditions. A kinetic study was performed in a batch stirred tank reactor to study the influence of different process parameters, such as temperature, feed composition and stirring speed. A pseudo-homogeneous kinetic model was developed to describe this reaction kinetics, proving that its rate is just first order on the acetaldehyde concentration under the conditions studied. © 2013 Elsevier B.V. All rights reserved.
1. Introduction There has been a significant increase in recent years in the use of isobutylene and bioethanol to produce oxygenated compounds, such as ethyl-tert-butyl ether (ETBE), as petroleum additives. Nowadays, the use of different biofuels in conventional car engines has become one of the technological goals on the path toward sustainable development. Biodiesel is an alternative fuel obtained from vegetable oils or animal fats, and it has several technical advantages over petro-diesel, such as a reduction in exhaust emissions, improved lubricity and biodegradability, higher flash point and reduced toxicity. There are several other properties, such as cetane number, gross heat of combustion and viscosity, which are very similar in both biodiesels and conventional diesels. However, biodiesels have an inferior performance compared to conventional diesels in terms of oxidation stability, nitrogen oxide emissions, energy content and cold weather operability [1]. Biodiesel is obtained from the chemical reaction between methanol (or ethanol) and animal fats or vegetable oils in the presence of a basic or acid catalyst. This reaction is called transesterification, and apart from methyl (ethyl) esters (biodiesel), glycerol is formed as a by-product (10 wt.%). Small amounts of this tri-alcohol are currently
⁎ Corresponding author. Tel.: +34 946 013 986; fax: +34 946 014 179. E-mail addresses: [email protected] (I. Agirre), [email protected] (M.B. Güemez), [email protected] (A. Ugarte), [email protected] (J. Requies), [email protected] (V.L. Barrio), [email protected] (J.F. Cambra), [email protected] (P.L. Arias). 0378-3820/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.fuproc.2013.05.014
being used in pharmaceutical and personal care products (e.g., cosmetic bonding agent for makeup). In order to avoid its incineration, different alternatives are being investigated in the quest for high value added products. Some potential uses include hydrogen gas production, glycerin acetate and acetal formation as potential fuel additives, as composite additive, and its conversion into citric acid, propylene glycol, acrolein, ethanol and epichlorohydrin [2,3]. A possible solution for the disadvantages biodiesels present is the use of suitable additives. Metal-based additives (manganese, iron, copper, barium …) have so far been the main ones [4], but due to environmental concerns, a number of other additives from renewable sources are being investigated. It is a well-known fact that oxygenated additives reduce HC and CO emissions and provide a high octane and high quality unleaded gasoline. The increase in oxygen content in diesel fuels significantly reduces the emissions of particulate matter. MTBE and ETBE, commonly used as gasoline additives, are not suitable as diesel additives because of their very low cetane numbers (e.g., ETBE cetane number is as low as 2.5). Acetals are more suitable oxygenated additives for diesel fuels [5]. For example 1,1 diethoxy ethane has been proven to reduce exhaust fumes [6], but its low flash point limits its practical use in this area. Heavier acetals derived from glycerol combine positive environmental effects through lower emissions and suitable cetane numbers. As a result, glycerol-derived acetals seem to be good candidates for playing a role in this area [7,8]. However, not all acetals can be used as diesel or biodiesel additives. As previously indicated, some acetals have low flash points, and as a result they are not suitable for use as diesel additives. Glycerol acetals fulfil these diesel specifications, while ethanol acetals require a
I. Agirre et al. / Fuel Processing Technology 116 (2013) 182–188
large aldehyde in order to provide acetals with acceptable flash points. Lower molecular weight acetals are being used as surfactants, flavours and disinfectants [3,7] in cosmetics, foodstuffs, pharmaceuticals and fragrances [2,9]. As explained above, large amounts of glycerol are being produced via biodiesel production, and one possible option for producing value added products involves the reaction of glycerol with aldehydes in order to produce acetals. This study considers glycerol and acetaldehyde acetals. Acetaldehyde may also have a renewable origin, as it can be obtained from bioethanol following a dehydrogenation or partial oxidation process [10,11]. Acetals formed from glycerol and acetaldehyde have been investigated for many years. Aksnes et al. [12,13] have published pioneering papers on these cyclic acetals. More recently, other studies have been carried out on the same products, since they are present in different types of wines [14,15] and have a bearing on their flavour. Acetals can be produced via homogeneous catalytic processes using strong mineral acids as catalysts, such as H2SO4, HF, HCl or p-toluenesulphonic acid [6,16,17]. Kaufhold et al. have patented [16] an industrial process for acetal production. Besides a homogeneous strong acid catalyst, this process uses an entrainer with a normal boiling point between 298.15 K and 348.15 K (hexane, pentane). This entrainer must be water insoluble (b 3% soluble in water), so the water is continuously removed from the reacting phase, shifting the acetalization reversible reaction in the desired direction. However, these processes lead to corrosion problems, are uneconomical and not eco-friendly [6,7]. The use of a heterogeneous catalyst would overcome most of these problems. Therefore, several solid acid catalysts are currently being tested. Capeletti et al. [7] have reported the performance of several solid acid catalysts, from commercial, natural and laboratory sources. They conclude that ion exchange resins perform better than other catalysts, allowing equilibrium values to be reached much faster than with other solid alternatives. Some authors have already verified the good behaviour of these catalysts in acetalization reactions [18–21]. This paper's main objective is to study the kinetics of the reaction between glycerol and acetaldehyde, given their potential as renewable resources for producing acetals with possible applications in the biofuel industry. Furthermore, the estimation of these kinetic parameters can be helpful for future studies on the simulation of industrial-scale production.
analysed by gas chromatography (Agilent 6890 N) using a flame ionization detector (FID) for organic compounds (glycerol, acetaldehyde and acetals) and a thermal conductivity detector (TCD) for water. An Agilent DB-1 60 m × 0.53 mm × 5 μm capillary column was used with Helium as the carrier gas. 2.3. Batch stirred tank reactor (BSTR) The experiments were carried out in a 1 L glass jacketed stirred reactor (Fig. 1). The reaction temperature was controlled by an external thermostat (Lauda RE 304). This thermostat contains an external thermocouple to be placed inside the reacting mixture and allows controlling the reaction temperature with an accuracy of ± 0.02 K. The reactor was also connected to a condenser in order to reflux all the vapours, keeping the reaction volume nearly constant and avoiding emissions by evaporation. The reactants were loaded into the reactor (total initial volume 0.5 L), and after stabilizing the system to the desired temperature the catalyst was added. Samples were taken at specific time intervals for their analysis by GC. A small piece of glass wool was placed in the output-sampling valve to keep the catalyst amount constant in the reactor. Prior to analysis, the samples were diluted (1.5/10 in volume) in dimethyl sulfoxide (DMS) in order to enhance the analysis results by avoiding peak saturation. DMS was selected because all the compounds are soluble in it, and it does not affect the mixture in the way alcoholic organic solvents do. Before adding the catalyst sample, and due to its high moisture content, it was first dried at room temperature. Thus, the catalyst
2. Material and methods 2.1. Materials Glycerol (99 wt.% for synthesis) and acetaldehyde (99.0 wt.%) from Panreac were used as reagents, with the selected catalyst being Amberlyst 47 supplied by Rohm & Haas. This commercial catalyst was chosen because previous studies of similar processes reported its good activity, stability and suitable mechanical strength [19,22]. 5-hydroxy-2-methyl-1,3 dioxane and 4-hydroxymethyl-2methyl-1,3 dioxolane standards were not found, so the reaction progress was followed by measuring reactant concentrations. However, glycerol formal (99 wt.%) (a mixture of two of the isomers formed by reacting glycerol with formaldehyde: 1,3-dioxan-5-ol (55 wt.%) and 1,3-dioxolane-4-methanol (45 wt.%)) from Acros Organics was used to check and compare the response factor of these acetals with the ones obtained in the reaction studied. As these organic compounds are similar, the order of magnitude of the response factors was also assumed to be similar. 2.2. Analysis Both the reactants (glycerol—GLY—and acetaldehyde—AcHO—) and the reaction products (mixtures of acetals—Ac— and water—W—) were
183
Fig. 1. Schematic drawing of the stirred batch reaction system.
184
I. Agirre et al. / Fuel Processing Technology 116 (2013) 182–188
had only the equilibrium humidity with the laboratory atmosphere, and its weight could therefore be controlled. All the tests were repeated at least three times in order to ensure the correct behaviour of the reaction and estimate suitably accurate parameter values within a predetermined tolerance range for a confidence interval of 95% according to the t-student distribution. 3. Results and discussion
commercial sulphuric acid (96 wt.%) was added to 0.5 L of reaction mixture. The behaviour was fully comparable to the behaviour observed using Amberlyst 47 acidic ion exchange resin. What's more, no sub-products were observed when using the sulphuric acid. However, with the exception of these initial experiments, all additional tests were carried out using Amberlyst 47 resin, since its separation is much easier than a homogeneous catalyst and avoids all the corrosion problems that strong liquid acids can generate downstream in an industrial process.
3.1. Preliminary calculations Certain calculations were performed before starting the experiments in order to find the most appropriate range of operating conditions. Glycerol has fairly high viscosity compared to water, which makes it difficult to manage. For this reason, the addition of water in the initial reactant mixture was studied. Glycerol/acetaldehyde/ water mixture properties (viscosity and bubble point) were estimated using the NRTL thermodynamic method in order to find an initial reactant mixture composition with as little water as possible, but with acceptable viscosity values. To this end, glycerol concentrations of 100, 90 and 80 wt.% were added together with acetaldehyde in order to obtain the desired glycerol/acetaldehyde molar ratios. These estimations were carried out for three different glycerol/acetaldehyde molar ratios, as the initial composition was varied to check the order of reaction. The results obtained are shown in Table 1. Water decreases the mixture's viscosity but, on the other hand, as water is one of the reaction products, its presence could imply a decrease in the final conversion. It should be mentioned that according to the data found in the literature, most acetalization reactions are thermodynamically limited reversible reactions, and the presence of one of the reaction products in the initial reactant mixtures would negatively affect the final conversion [3,18–20,23,24]. Nevertheless, the decision was made to use 80 wt.% glycerol concentrations because the production of higher purities would be technically and economically unfeasible (once the biodiesel has been separated from the mixture, glycerol is obtained as a diluted aqueous solution after the washing process). However, as can be seen at the end of this paper, the choice of water as solvent did not affect the final conversion. In addition, the bubble point estimation of the glycerol/acetaldehyde/water mixtures studied allowed setting the maximum temperature for undertaking the reaction in liquid phase and at atmospheric pressure. 3.2. Preliminary experiments Before starting the experiments, a number of tests were performed using H2SO4 as catalyst. These tests were carried out at 313 K and with a 3:1 glycerol to acetaldehyde initial feed molar ratio. 3.5 mL of
3.3. Mass transfer resistance In order to avoid mass transfer resistance, a wide range of stirring speeds (from 700 rpm to 1250 rpm) was tested. There was no observable effect of stirring on the reaction rate (Fig. 2), proving the absence of any external mass transfer resistance even at the lowest stirring speed (700 rpm), so all further experiments were carried out at 700 rpm. These experiments were performed with a 3:1 glycerol to acetaldehyde molar ratio since it is the most viscous initial mixture to work with, and so the stirring speed is more critical. The selected temperature was 313 K, the highest temperature used in this study. The effect of temperature is much higher over the reaction rate than over the mass transfer coefficient. If there was any external mass resistance, it would be observed much more clearly at high temperatures. 3.4. Reaction mechanism Most acetalization reactions involving polyalcohols take place through two reversible steps: a first one, in which alcohol reacts with the aldehyde molecule leading to the formation of the corresponding hemiacetal, and a second one, in which two hydroxyl groups of the hemiacetal join to form the corresponding acetal, releasing a water molecule [23]. da Silva Ferreira et al. [15] have provided a detailed explanation of the reaction mechanism for the reaction between glycerol and acetaldehyde, specifying all the different steps in the reorganization of the hemiacetal molecule. However, a detailed description of the reaction mechanism is beyond the scope of this paper, and the scheme proposed by Sharma & Chopade [23] was used as shown in Fig. 3. The first step in some acetalization reactions, which does not require a catalyst, is much faster than the second step [19,23]. There is also some evidence to indicate a similar behaviour when glycerol and acetaldehyde are used as reactants. Immediately after mixing these compounds at room temperature, the mixture temperature increases sharply due to the exothermic reaction taking place as a non-catalysed process. However, it should be noted that in this case,
Table 1 Estimated properties using NRTL thermodynamic method for different initial reactant mixtures (cases a, b and c correspond to glycerol–water mixtures of 100, 90 and 80 wt.% of glycerol, respectively.). Mass composition (%) GLY a
67.6 80.7 86.2 b 62.9 74.1 78.7 c 16.6 27.4 57.9 67.2 70.9
Gly: AcHO TBubble Viscosity (cPo) Molar ratio (K)
AcHO H2O 32.4 19.3 13.8 30.1 17.7 12.6 79.3 65.7 27.7 16.1 11.3
0.0 0.0 0.0 7.0 8.2 8.7 4.1 6.9 14.5 16.8 17.7
283 K 293 K 303 K 313 K 323 K 1.0 2.0 3.0 1.0 2.0 3.0 0.1 0.2 1.0 2.0 3.0
315 329 339 311 321 328 298 300 310 319 326
29 144 319 15 40 62 0.6 1.2 8.6 16 21
18 76 158 9.4 23 35 0.5 1.0 5.8 10 13
11 42 83 6.1 14 21 – – 4.0 6.8 8.5
7.2 25 45.5 5.3 9.2 13 – – 3.6 4.7 5.8
– – 28.0 – 7.0 8.6 – – – 3.8 4.1
Fig. 2. Effect of the stirring speed. 3:1 glycerol to acetaldehyde molar ratio, 313 K and 2.0 wt.% of Amberlyst 47.
I. Agirre et al. / Fuel Processing Technology 116 (2013) 182–188
185
Fig. 3. Reaction mechanism based on the one proposed by Sharma & Chopade [23].
the second step of the reaction can be carried out without a catalyst, but the acetal formation rate is extremely slow, as shown in Fig. 4. However, certain difficulties were encountered when analysing these samples, which could be explained by the exothermicity of a non-catalysed reaction. When the sample is injected into the GC (which was set at 498 K due to the low volatility of glycerol) the reverse reaction occurs due to a sharp decrease in the equilibrium constant with the increasing temperature, and as a result most of the hemiacetal disappears, which means the sample analysis is not representative. Similar problems were encountered when the reaction between ethanol and butanal was studied [19]. In this case, HPLC/MS analyses of ethanol–butanal mixtures at room temperature confirmed the presence of the hemiacetal in the samples [25]. In addition to this, some authors [26] have reported the unstable behaviour of hemi-acetals. As indicated in Fig. 3, four different isomers are formed in the acetalization reaction between glycerol and acetaldehyde: 5-hydroxy2-methyl-1,3 dioxane (Ac1) and 4-hydroxymethyl-2-methyl-1,3 dioxolane (Ac2), with both of them being in their cis and trans stereoisomer forms. Due to the lack of standards for GC calibration, each isomer was identified using a GC/MS. Thus, a comparison between the mass spectra obtained and the theoretical ones (Fig. 5) meant each isomer could be identified. Furthermore, it was verified that the mass spectra measured using the GC/MS were in good agreement with the ones reported by da Silva Ferreira et al. [15]. Moreover, according to the literature, Camara et al. (2003) and da Silva Ferreira et al. [14,15] have observed an isomerization reaction of the 1,3 dioxolane toward dioxane formation in the evolution of the acetals under study here. Moreover,
they have reported that 1,3 dioxolane isomers were formed faster, but then the reaction mixture proceeded towards isomerization equilibrium, generating more 1,3 dioxanes. This isomerization reaction was also observed in the case studied, as indicated in Fig. 4. By contrast, Ruiz et al. [27] have observed the isomerization from dioxanes to dioxolanes. In terms of the identification between cis and trans stereoisomers, their identification using mass spectra coincide with the results obtained by Aksnes et al. [12] after carrying out identification tests based on infrared bands and refractive indexes. However, in order to study the reaction kinetics, there is no need to distinguish all four of the acetals because at it will be seen later this acetalization reaction is irreversible. Following Sharma & Chopade's development [23,24], hemiacetal concentration can be considered approximately at equilibrium with glycerol and acetaldehyde at any time. ½HA ¼ K½AcHO½Gly:
ð1Þ
Where K is the equilibrium constant in (mol/L)−1. The formation rate of acetal (Ac) may be expressed as d½Ac ¼ wk3 ½HA−wk4 ½AC½W: dt
ð2Þ
Where w is (gcat)/(reacting volume in L) k3 and k4 are the kinetic constants. k3 in L/(gcat · min) and k4 in L2/(min · gcat) [] indicates concentration in mol/L.
186
I. Agirre et al. / Fuel Processing Technology 116 (2013) 182–188
Fig. 4. Acetal formation as a function of time with and without catalysts (700 rpm, 303 K, glycerol/acetaldehyde molar ratio 1:1).
Substituting [HA] from Eq. (1), d½Ac ¼ wk½AcHO½Gly−wk4 ½AC½W: dt
ð3Þ
Where k = k3K in L2/(gcat · min · mol). According to this theoretical development, the global reaction is reversible and first order with respect to glycerol and acetaldehyde in the forward direction, and also first order with respect to acetal and water in the reversible reaction. In order to check whether this kinetic expression is correct, different experiments were carried out using different feed mole ratios (1:1, 2:1, 3:1, glycerol to acetaldehyde). All the experiments with excess glycerol were performed at 313 K, 700 rpm and with 2.0 wt.% of Amberlyst 47. The experiments with excess acetaldehyde were carried out at 293 K. The overall conversion reaches 100% in all cases when excess glycerol was used. Moreover, the following section, Effect of temperature, shows that the final conversion achieved was also 100% at every temperature. All this evidence proves that the glycerol + acetaldehyde reaction under the operating conditions investigated is irreversible. This reaction kinetics can therefore be studied without the need for an accurate determination of the four product concentrations. All the information found in the literature about acetalization reactions shows that this type of reaction has high thermodynamic limitations. Thus, Sharma & Chopade [23,24] have achieved a maximum of 50% conversion (at 363 K) by reacting formaldehyde with ethylene
Fig. 5. Comparison of the theoretical mass spectra with the experimental one obtained by GC/MS.
glycol, and less than 50% (at 343 K) when reacting ethanol and formaldehyde. Agirre et al. [19] have studied the reaction between ethanol and butanal, and the maximum conversion achieved was also around 50% (at 313 K). Other authors, such as Deutsch et al. [18] and Silva et al. [3] have studied glycerol acetalization reactions with different aldehydes. Deutsch et al. [18] have studied the acetalization reaction of glycerol with formaldehyde using different types of solvents, and the maximum conversions achieved were between 58 and 77% (between 313 and 383 K). On the other hand, Silva et al. [3] have studied the reaction of glycerol with butanal, pentanal, hexanal, octanal and decanal, recording a peak of 80% conversion (at 343 K) with butanal. In general, they have recorded lower conversions using heavier aldehydes.
I. Agirre et al. / Fuel Processing Technology 116 (2013) 182–188
187
Fig. 8. 2,4,6 trimethyl-1,3,5 trioxane formation reaction (side reaction).
different kinetic expressions, it was concluded that the one with the best fit to the experimental data and the best agreement with the observed reaction behaviour is only a first order function of acetaldehyde concentration: d½Ac ¼ wk½AcHO: dt
Fig. 6. Effect of the initial feed ratio (700 rpm, 2 wt.% of Amberlyst 47). a) Excess glycerol (303 K) and b) excess acetaldehyde (393 K).
An integral approach was selected for the kinetic analysis. In order to test the suitability of different kinetic expressions, a fourth order Runge– Kutta integration method was used. Thus, by minimizing the sum of squares of the differences between the experimental concentrations and those predicted by several kinetic expressions, the apparent kinetic constants were calculated observing the adequacy of the fittings obtained. With a view to testing the suitability of different kinetic expressions, use was made of the experimental data obtained in every single test with different glycerol to acetaldehyde ratios. It should be noted that every experiment with a certain ratio was carried out at least three times in order to verify experimental repeatability. After testing
ð4Þ
Thus, after fitting the experimental data to Eq. (4) and estimating the corresponding kinetic parameters, the acetal concentrations measured experimentally were compared to the predicted ones. Fig. 7 shows that the agreement between them is reasonably close and, therefore, this kinetic expression (Eq. (4)) describes the reaction evolution quite accurately. Contrary to published data in the literature, the reaction studied does not have thermodynamic limitations at 313 K. However, certain recent kinetic studies carried out for the acetalization reaction between glycerol and butyraldehyde [28] at 353 K, also using Amberlyst 47 as catalyst, record conversions of more than 90%. As acetalizations are slightly exothermic reactions, if the temperature operating range in this recent study had been in the 293–313 K range, as in the present case, it is quite likely that a similar behaviour would have been recorded. When carried out at low temperatures, glycerol acetalization reactions seem to be either irreversible or nearly irreversible when >C2 aldehydes are used as the other reactant. Another important observation was that low glycerol to acetaldehyde initial molar ratios led to the formation of 2,4,6 trimethyl-1,3,5 trioxane. This compound was identified by GC/MS analysis. When glycerol was depleted in the reactor, trioxane became the main product; the conversion of acetaldehyde therefore continues (see Fig. 6b), but toward the formation of trioxane, as in the reaction scheme shown in Fig. 8. Dintzner et al. [29] and Augé and Gil [30] have explained the production of trioxanes from aldehydes using different heterogeneous acid catalysts. 3.5. Effect of temperature Four different temperatures were tested: 283 K, 293 K, 303 K and 313 K. At least three different experiments were performed for each
Fig. 7. Comparison between experimental acetal concentrations and the ones predicted by the kinetic model (Eq. (4)).
Fig. 9. Effect of temperature on the acetalization reaction. 700 rpm, feed ratio 3:1, 2 wt.% Amberlyst 47.
188
I. Agirre et al. / Fuel Processing Technology 116 (2013) 182–188
Table 2 Kinetic constants at different temperatures and Arrhenius parameters. Ln(A) (pre-exponential factor) Ea (activation energy) T(K) k −1 · g−1 (kJ/mol) (L · min−1 · g−1 cat ) (L · min cat ) 313 303 293 283
(125 ± 2) (75 ± 8) (31 ± 2) (14 ± nd)
· · · ·
10−5 10−5 10−5 10−5
14.51 ± 1.3
55.4 ± 3.2
nd: non-determined.
temperature, with the definitive kinetic parameters being the arithmetic average of the parameters calculated for each experiment. The apparent kinetic parameters were estimated using the integral method explained in Section 3.4. The experimental data obtained in each whole experiment were fitted to Eq. (4), obtaining the corresponding kinetic parameter from each experiment. As in all the previous experiments, all the tests were performed with 2.0 wt.% of Amberlyst 47 resin and at 700 rpm. In terms of the ratio of initial reactants, a 3:1 glycerol to acetaldehyde molar ratio was used, looking for less volatile mixtures. When using the stoichiometric feed ratio, the vaporization effects were significant above 295– 300 K since the bubble point of the initial mixture is 310 K (see Table 1). In the case of the acetalization reaction between glycerol and acetaldehyde, and contrary to other acetalization reactions reported in the open literature [20,23,24], it was observed that 100% conversion is achievable at each temperature. The only change observed when increasing the temperature was the significant increase in the reaction rate (Fig. 9), as is to be expected. In order to apply the kinetics of the reaction studied here to further process developments, the kinetic constants obtained at different temperatures (Table 2) were fitted to the Arrhenius correlation parameters. Table 2 also shows the pre-exponential factor (A) and the activation energy (Ea). It should be noted that the activation energy found is in good agreement with other activation energies of acetalization reactions reported in the literature [19,23,24].
4. Conclusions The most pertinent conclusion this research reaches is that the acetalization reaction between glycerol and acetaldehyde is not a thermodynamically limited reversible reaction, as are other acetalization reactions reported in the literature. The great advantage of this finding is that, in principle, the industrial production of these acetals for their use as biodiesel additives can be carried out in conventional reaction systems. Being a non-reversible reaction, it was verified that its rate within the conditions investigated depends solely on the acetaldehyde concentration, and as a result the presence of water in the initial reactant mixture does not have a negative effect on the reaction kinetics because of dilution. This is also an important finding, as water dilution helps to considerably reduce the viscosity of the reacting mixture. All this information offers new insights for glycerol conversion into value added products, such as acetals or similar diesel additives.
Acknowledgements The authors gratefully acknowledge the financial support for this work provided by the Spanish Ministry of Science and Innovation (ENE2009-12743-C04-04), the Basque Government and the University of the Basque Country (UPV/EHU). Furthermore, the authors would like to thank Rohm & Haas for kindly supplying different Amberlyst resins.
References [1] B.R. Moser, S.Z. Erhan, Branched chain derivatives of alkyl oleates: tribological, rheological, oxidation, and low temperature properties, Fuel 87 (2008) 2253–2257. [2] S.B. Umbarkar, T.V. Kotbagi, A.V. Biradar, R. Pasricha, J. Chanale, M.K. Dongare, A.S. Mamede, C. Lancelot, E. Payen, Acetalization of glycerol using mesoporous MoO3/SiO2 solid acid catalyst, Journal of Molecular Catalysis A: Chemical 310 (2009) 150–158. [3] P.H.R. Silva, V.L.C. Gonçalves, C.J.A. Mota, Glycerol acetals as anti-freezing additives for biodiesel, Bioresource Technology 101 (2010) 6225–6229. [4] H. Burtscher, U. Matter, G. Skillas, The effect of fuel additives on diesel engine particulate emissions, Journal of Aerosol Science 30 (1999) S851–S852. [5] D. Naegeli, L. Dodge, Combustion Characterization of Methylal in Reciprocating Engines, Final Report No. SwRI 03-130, Southwest Research Institute, San Antonio, TX, 1992. [6] F. Frusteri, L. Spadaro, C. Beatrice, C. Guido, Oxygenated additives production for diesel engine emission improvement, Chemical Engineering Journal 134 (2007) 239–245, (l). [7] M.R. Capeletti, L. Balzano, G. de la Puente, M. Laborde, U. Sedran, Synthesis of acetal (1,1-diethoxyethane) from ethanol and acetaldehyde over acidic catalysts, Applied Catalysis A: General 198 (2000) L1–L4. [8] N. Rahmat, A.Z. Abdullah, A.R. Mohamed, Recent progress on innovative and potential technologies for glycerol transformation into fuel additives: a critical review, Renewable & Sustainable Energy Reviews 14 (2010) 987–1000. [9] S. Yang, M. Wang, Y. Liang, J. Sun, Eco-friendly one-pot synthesis of acetals and ketals by heterogeneously catalyzed liquid–solid phase reaction, Rare Metals (Beijing, China) 25 (2006) 625–629. [10] J. Requies, M.B. Güemez, P. Maireles, A. Iriondo, V.L. Barrio, J.F. Cambra, P.L. Arias, Zirconia supported Cu systems as catalysts for n-butanol conversion to butyraldehyde, Applied Catalysis A: General 423–424 (2012) 185–191. [11] J. Requies, M.B. Güemez, A. Iriondo, V.L. Barrio, J.F. Cambra, P.L. Arias, Bio n-butanol partial oxidation to butyraldehyde in gas phase on supported Ru and Cu catalysts, Catalysis Letters 142 (2012) 417–426. [12] G. Aksnes, P. Albriktsen, P. Juvvik, Cyclic acetal and ketal isomers of glycerol, Acta Chemica Scandinavica 19 (1965) 920–930, (1947–1973). [13] G. Aksnes, P. Albriktsen, Intramolecular H bonding in cyclic acetal isomers of glycerol, Acta Chemica Scandinavica 20 (1966) 1330–1334. [14] J.S. Camara, J.C. Marques, A. Alves, A.C.S. Ferreira, Heterocyclic acetals in Madeira wines, Analytical and Bioanalytical Chemistry 375 (2003) 1221–1224. [15] A.C. da Silva Ferreira, J. Barbe, A. Bertrand, Heterocyclic acetals from glycerol and acetaldehyde in Port wines: evolution with aging, Journal of Agricultural and Food Chemistry 50 (2002) 2560–2564. [16] M. Kaufhold, M. El-Chahawi, Process for Preparing Acetaldehyde Diethyl Acetal, 5527969, 1996. (Germany). [17] T.W. Green, Protective Groups in Organic Synthesis, Willey, New York, 1981. [18] J. Deutsch, A. Martin, H. Lieske, Investigations on heterogeneously catalysed condensations of glycerol to cyclic acetals, Journal of Catalysis 245 (2007) 428–435. [19] I. Agirre, V.L. Barrio, B. Güemez, J.F. Cambra, P.L. Arias, The development of a reactive distillation process for the production of 1,1 diethoxy butane from bioalcohol: kinetic study and simulation model, International Journal of Chemical Reactor Engineering 8 (2010) A86. [20] A.D. Dhale, L.K. Myrant, S.P. Chopade, J.E. Jackson, D.J. Miller, Propylene glycol and ethylene glycol recovery from aqueous solution via reactive distillation, Chemical Engineering Science 59 (2004) 2881–2890. [21] I. Agirre, V.L. Barrio, B. Güemez, J.F. Cambra, P.L. Arias, Catalytic reactive distillation process development for 1,1 diethoxy butane production from renewable sources, Bioresource Technology 102 (2011) 1289–1297. [22] I. Agirre, I. García, J. Requies, V.L. Barrio, M.B. Güemez, J.F. Cambra, P.L. Arias, Glycerol acetals, kinetic study of the reaction between glycerol and formaldehyde, Biomass and Bioenergy 35 (2011) 3636–3642. [23] S.P. Chopade, M.M. Sharma, Acetalization of ethylene glycol with formaldehyde using cation-exchange resins as catalysts: batch versus reactive distillation, Reactive and Functional Polymers 34 (1997) 37–45. [24] S.P. Chopade, M.M. Sharma, Reaction of ethanol and formaldehyde: use of versatile cation-exchange resins as catalyst in batch reactors and reactive distillation columns, Reactive and Functional Polymers 32 (1997) 53–65. [25] B. Abad, Identificación de la presencia de hemiacetal en una mezcla etanol: butanal (2:1 en moles), Servicio general de análisis UPV/EHU, SGIKER, General Analysis Department of the University of the Basque Country, 2008. [26] G.D. Yadav, A.A. Pujari, Kinetics of acetalization of perfumery aldehydes with alkanols over solid acid catalysts, Canadian Journal of Chemical Engineering 77 (1999) 489–496. [27] V.R. Ruiz, A. Velty, L.L. Santos, A. Leyva-Pérez, M.J. Sabater, S. Iborra, A. Corma, Gold catalysts and solid catalysts for biomass transformations: valorization of glycerol and glycerol–water mixtures through formation of cyclic acetals, Journal of Catalysis 271 (2010) 351–357. [28] B. Güemez, J.M. Requies, I. Agirre, V.L. Barrio, J.F. Cambra, P.L. Arias, Acetalization reaction between glycerol and n-butyraldehyde using an acidic ion exchange resin. Kinetic modelling, Chemical Engineering Journal 228 (2013) 300–307. [29] M.R. Dintzner, Y.A. Mondjinou, D.J. Pileggi, Montmorillonite clay-catalyzed cyclotrimerization and oxidation of aliphatic aldehydes, Tetrahedron Letters 51 (2010) 826–827. [30] J. Augé, R. Gil, A convenient solvent-free preparation of 1,3,5-trioxanes, Tetrahedron Letters 43 (2002) 7919–7920.
Contents lists available at SciVerse ScienceDirect
Fuel Processing Technology journal homepage: www.elsevier.com/locate/fuproc
Glycerol acetals as diesel additives: Kinetic study of the reaction between glycerol and acetaldehyde I. Agirre ⁎, M.B. Güemez, A. Ugarte, J. Requies, V.L. Barrio, J.F. Cambra, P.L. Arias Chemical and Environmental Engineering Department, Engineering Faculty of Bilbao, University of the Basque Country (UPV/EHU), Alameda Urquijo s/n, 48013 Bilbao, Spain
a r t i c l e
i n f o
Article history: Received 4 February 2013 Received in revised form 16 April 2013 Accepted 15 May 2013 Available online xxxx Keywords: Glycerol Acetaldehyde Acetals Kinetics Amberlyst
a b s t r a c t Certain acetals can be produced from renewable resources (bioalcohols) and seem to be good candidates for different applications, such as oxygenated diesel additives. This paper addresses the production of acetals (5-hydroxy-2-methyl-1,3 dioxane and 4-hydroxymethyl-2-methyl-1,3 dioxolane) from glycerol and acetaldehyde using Amberlyst 47 acidic ion exchange resin. This ion exchange resin performed well, recording 100% selectivity toward acetal formation at a suitably high initial glycerol concentration. When the initial acetaldehyde concentration was significantly higher than the glycerol concentration, 2,4,6 trimethyl-1,3,5 trioxane was the main reaction product. Unlike other acetalization reactions, the one studied here does not have thermodynamic limitations, and 100% conversion is achieved under different reaction conditions. A kinetic study was performed in a batch stirred tank reactor to study the influence of different process parameters, such as temperature, feed composition and stirring speed. A pseudo-homogeneous kinetic model was developed to describe this reaction kinetics, proving that its rate is just first order on the acetaldehyde concentration under the conditions studied. © 2013 Elsevier B.V. All rights reserved.
1. Introduction There has been a significant increase in recent years in the use of isobutylene and bioethanol to produce oxygenated compounds, such as ethyl-tert-butyl ether (ETBE), as petroleum additives. Nowadays, the use of different biofuels in conventional car engines has become one of the technological goals on the path toward sustainable development. Biodiesel is an alternative fuel obtained from vegetable oils or animal fats, and it has several technical advantages over petro-diesel, such as a reduction in exhaust emissions, improved lubricity and biodegradability, higher flash point and reduced toxicity. There are several other properties, such as cetane number, gross heat of combustion and viscosity, which are very similar in both biodiesels and conventional diesels. However, biodiesels have an inferior performance compared to conventional diesels in terms of oxidation stability, nitrogen oxide emissions, energy content and cold weather operability [1]. Biodiesel is obtained from the chemical reaction between methanol (or ethanol) and animal fats or vegetable oils in the presence of a basic or acid catalyst. This reaction is called transesterification, and apart from methyl (ethyl) esters (biodiesel), glycerol is formed as a by-product (10 wt.%). Small amounts of this tri-alcohol are currently
⁎ Corresponding author. Tel.: +34 946 013 986; fax: +34 946 014 179. E-mail addresses: [email protected] (I. Agirre), [email protected] (M.B. Güemez), [email protected] (A. Ugarte), [email protected] (J. Requies), [email protected] (V.L. Barrio), [email protected] (J.F. Cambra), [email protected] (P.L. Arias). 0378-3820/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.fuproc.2013.05.014
being used in pharmaceutical and personal care products (e.g., cosmetic bonding agent for makeup). In order to avoid its incineration, different alternatives are being investigated in the quest for high value added products. Some potential uses include hydrogen gas production, glycerin acetate and acetal formation as potential fuel additives, as composite additive, and its conversion into citric acid, propylene glycol, acrolein, ethanol and epichlorohydrin [2,3]. A possible solution for the disadvantages biodiesels present is the use of suitable additives. Metal-based additives (manganese, iron, copper, barium …) have so far been the main ones [4], but due to environmental concerns, a number of other additives from renewable sources are being investigated. It is a well-known fact that oxygenated additives reduce HC and CO emissions and provide a high octane and high quality unleaded gasoline. The increase in oxygen content in diesel fuels significantly reduces the emissions of particulate matter. MTBE and ETBE, commonly used as gasoline additives, are not suitable as diesel additives because of their very low cetane numbers (e.g., ETBE cetane number is as low as 2.5). Acetals are more suitable oxygenated additives for diesel fuels [5]. For example 1,1 diethoxy ethane has been proven to reduce exhaust fumes [6], but its low flash point limits its practical use in this area. Heavier acetals derived from glycerol combine positive environmental effects through lower emissions and suitable cetane numbers. As a result, glycerol-derived acetals seem to be good candidates for playing a role in this area [7,8]. However, not all acetals can be used as diesel or biodiesel additives. As previously indicated, some acetals have low flash points, and as a result they are not suitable for use as diesel additives. Glycerol acetals fulfil these diesel specifications, while ethanol acetals require a
I. Agirre et al. / Fuel Processing Technology 116 (2013) 182–188
large aldehyde in order to provide acetals with acceptable flash points. Lower molecular weight acetals are being used as surfactants, flavours and disinfectants [3,7] in cosmetics, foodstuffs, pharmaceuticals and fragrances [2,9]. As explained above, large amounts of glycerol are being produced via biodiesel production, and one possible option for producing value added products involves the reaction of glycerol with aldehydes in order to produce acetals. This study considers glycerol and acetaldehyde acetals. Acetaldehyde may also have a renewable origin, as it can be obtained from bioethanol following a dehydrogenation or partial oxidation process [10,11]. Acetals formed from glycerol and acetaldehyde have been investigated for many years. Aksnes et al. [12,13] have published pioneering papers on these cyclic acetals. More recently, other studies have been carried out on the same products, since they are present in different types of wines [14,15] and have a bearing on their flavour. Acetals can be produced via homogeneous catalytic processes using strong mineral acids as catalysts, such as H2SO4, HF, HCl or p-toluenesulphonic acid [6,16,17]. Kaufhold et al. have patented [16] an industrial process for acetal production. Besides a homogeneous strong acid catalyst, this process uses an entrainer with a normal boiling point between 298.15 K and 348.15 K (hexane, pentane). This entrainer must be water insoluble (b 3% soluble in water), so the water is continuously removed from the reacting phase, shifting the acetalization reversible reaction in the desired direction. However, these processes lead to corrosion problems, are uneconomical and not eco-friendly [6,7]. The use of a heterogeneous catalyst would overcome most of these problems. Therefore, several solid acid catalysts are currently being tested. Capeletti et al. [7] have reported the performance of several solid acid catalysts, from commercial, natural and laboratory sources. They conclude that ion exchange resins perform better than other catalysts, allowing equilibrium values to be reached much faster than with other solid alternatives. Some authors have already verified the good behaviour of these catalysts in acetalization reactions [18–21]. This paper's main objective is to study the kinetics of the reaction between glycerol and acetaldehyde, given their potential as renewable resources for producing acetals with possible applications in the biofuel industry. Furthermore, the estimation of these kinetic parameters can be helpful for future studies on the simulation of industrial-scale production.
analysed by gas chromatography (Agilent 6890 N) using a flame ionization detector (FID) for organic compounds (glycerol, acetaldehyde and acetals) and a thermal conductivity detector (TCD) for water. An Agilent DB-1 60 m × 0.53 mm × 5 μm capillary column was used with Helium as the carrier gas. 2.3. Batch stirred tank reactor (BSTR) The experiments were carried out in a 1 L glass jacketed stirred reactor (Fig. 1). The reaction temperature was controlled by an external thermostat (Lauda RE 304). This thermostat contains an external thermocouple to be placed inside the reacting mixture and allows controlling the reaction temperature with an accuracy of ± 0.02 K. The reactor was also connected to a condenser in order to reflux all the vapours, keeping the reaction volume nearly constant and avoiding emissions by evaporation. The reactants were loaded into the reactor (total initial volume 0.5 L), and after stabilizing the system to the desired temperature the catalyst was added. Samples were taken at specific time intervals for their analysis by GC. A small piece of glass wool was placed in the output-sampling valve to keep the catalyst amount constant in the reactor. Prior to analysis, the samples were diluted (1.5/10 in volume) in dimethyl sulfoxide (DMS) in order to enhance the analysis results by avoiding peak saturation. DMS was selected because all the compounds are soluble in it, and it does not affect the mixture in the way alcoholic organic solvents do. Before adding the catalyst sample, and due to its high moisture content, it was first dried at room temperature. Thus, the catalyst
2. Material and methods 2.1. Materials Glycerol (99 wt.% for synthesis) and acetaldehyde (99.0 wt.%) from Panreac were used as reagents, with the selected catalyst being Amberlyst 47 supplied by Rohm & Haas. This commercial catalyst was chosen because previous studies of similar processes reported its good activity, stability and suitable mechanical strength [19,22]. 5-hydroxy-2-methyl-1,3 dioxane and 4-hydroxymethyl-2methyl-1,3 dioxolane standards were not found, so the reaction progress was followed by measuring reactant concentrations. However, glycerol formal (99 wt.%) (a mixture of two of the isomers formed by reacting glycerol with formaldehyde: 1,3-dioxan-5-ol (55 wt.%) and 1,3-dioxolane-4-methanol (45 wt.%)) from Acros Organics was used to check and compare the response factor of these acetals with the ones obtained in the reaction studied. As these organic compounds are similar, the order of magnitude of the response factors was also assumed to be similar. 2.2. Analysis Both the reactants (glycerol—GLY—and acetaldehyde—AcHO—) and the reaction products (mixtures of acetals—Ac— and water—W—) were
183
Fig. 1. Schematic drawing of the stirred batch reaction system.
184
I. Agirre et al. / Fuel Processing Technology 116 (2013) 182–188
had only the equilibrium humidity with the laboratory atmosphere, and its weight could therefore be controlled. All the tests were repeated at least three times in order to ensure the correct behaviour of the reaction and estimate suitably accurate parameter values within a predetermined tolerance range for a confidence interval of 95% according to the t-student distribution. 3. Results and discussion
commercial sulphuric acid (96 wt.%) was added to 0.5 L of reaction mixture. The behaviour was fully comparable to the behaviour observed using Amberlyst 47 acidic ion exchange resin. What's more, no sub-products were observed when using the sulphuric acid. However, with the exception of these initial experiments, all additional tests were carried out using Amberlyst 47 resin, since its separation is much easier than a homogeneous catalyst and avoids all the corrosion problems that strong liquid acids can generate downstream in an industrial process.
3.1. Preliminary calculations Certain calculations were performed before starting the experiments in order to find the most appropriate range of operating conditions. Glycerol has fairly high viscosity compared to water, which makes it difficult to manage. For this reason, the addition of water in the initial reactant mixture was studied. Glycerol/acetaldehyde/ water mixture properties (viscosity and bubble point) were estimated using the NRTL thermodynamic method in order to find an initial reactant mixture composition with as little water as possible, but with acceptable viscosity values. To this end, glycerol concentrations of 100, 90 and 80 wt.% were added together with acetaldehyde in order to obtain the desired glycerol/acetaldehyde molar ratios. These estimations were carried out for three different glycerol/acetaldehyde molar ratios, as the initial composition was varied to check the order of reaction. The results obtained are shown in Table 1. Water decreases the mixture's viscosity but, on the other hand, as water is one of the reaction products, its presence could imply a decrease in the final conversion. It should be mentioned that according to the data found in the literature, most acetalization reactions are thermodynamically limited reversible reactions, and the presence of one of the reaction products in the initial reactant mixtures would negatively affect the final conversion [3,18–20,23,24]. Nevertheless, the decision was made to use 80 wt.% glycerol concentrations because the production of higher purities would be technically and economically unfeasible (once the biodiesel has been separated from the mixture, glycerol is obtained as a diluted aqueous solution after the washing process). However, as can be seen at the end of this paper, the choice of water as solvent did not affect the final conversion. In addition, the bubble point estimation of the glycerol/acetaldehyde/water mixtures studied allowed setting the maximum temperature for undertaking the reaction in liquid phase and at atmospheric pressure. 3.2. Preliminary experiments Before starting the experiments, a number of tests were performed using H2SO4 as catalyst. These tests were carried out at 313 K and with a 3:1 glycerol to acetaldehyde initial feed molar ratio. 3.5 mL of
3.3. Mass transfer resistance In order to avoid mass transfer resistance, a wide range of stirring speeds (from 700 rpm to 1250 rpm) was tested. There was no observable effect of stirring on the reaction rate (Fig. 2), proving the absence of any external mass transfer resistance even at the lowest stirring speed (700 rpm), so all further experiments were carried out at 700 rpm. These experiments were performed with a 3:1 glycerol to acetaldehyde molar ratio since it is the most viscous initial mixture to work with, and so the stirring speed is more critical. The selected temperature was 313 K, the highest temperature used in this study. The effect of temperature is much higher over the reaction rate than over the mass transfer coefficient. If there was any external mass resistance, it would be observed much more clearly at high temperatures. 3.4. Reaction mechanism Most acetalization reactions involving polyalcohols take place through two reversible steps: a first one, in which alcohol reacts with the aldehyde molecule leading to the formation of the corresponding hemiacetal, and a second one, in which two hydroxyl groups of the hemiacetal join to form the corresponding acetal, releasing a water molecule [23]. da Silva Ferreira et al. [15] have provided a detailed explanation of the reaction mechanism for the reaction between glycerol and acetaldehyde, specifying all the different steps in the reorganization of the hemiacetal molecule. However, a detailed description of the reaction mechanism is beyond the scope of this paper, and the scheme proposed by Sharma & Chopade [23] was used as shown in Fig. 3. The first step in some acetalization reactions, which does not require a catalyst, is much faster than the second step [19,23]. There is also some evidence to indicate a similar behaviour when glycerol and acetaldehyde are used as reactants. Immediately after mixing these compounds at room temperature, the mixture temperature increases sharply due to the exothermic reaction taking place as a non-catalysed process. However, it should be noted that in this case,
Table 1 Estimated properties using NRTL thermodynamic method for different initial reactant mixtures (cases a, b and c correspond to glycerol–water mixtures of 100, 90 and 80 wt.% of glycerol, respectively.). Mass composition (%) GLY a
67.6 80.7 86.2 b 62.9 74.1 78.7 c 16.6 27.4 57.9 67.2 70.9
Gly: AcHO TBubble Viscosity (cPo) Molar ratio (K)
AcHO H2O 32.4 19.3 13.8 30.1 17.7 12.6 79.3 65.7 27.7 16.1 11.3
0.0 0.0 0.0 7.0 8.2 8.7 4.1 6.9 14.5 16.8 17.7
283 K 293 K 303 K 313 K 323 K 1.0 2.0 3.0 1.0 2.0 3.0 0.1 0.2 1.0 2.0 3.0
315 329 339 311 321 328 298 300 310 319 326
29 144 319 15 40 62 0.6 1.2 8.6 16 21
18 76 158 9.4 23 35 0.5 1.0 5.8 10 13
11 42 83 6.1 14 21 – – 4.0 6.8 8.5
7.2 25 45.5 5.3 9.2 13 – – 3.6 4.7 5.8
– – 28.0 – 7.0 8.6 – – – 3.8 4.1
Fig. 2. Effect of the stirring speed. 3:1 glycerol to acetaldehyde molar ratio, 313 K and 2.0 wt.% of Amberlyst 47.
I. Agirre et al. / Fuel Processing Technology 116 (2013) 182–188
185
Fig. 3. Reaction mechanism based on the one proposed by Sharma & Chopade [23].
the second step of the reaction can be carried out without a catalyst, but the acetal formation rate is extremely slow, as shown in Fig. 4. However, certain difficulties were encountered when analysing these samples, which could be explained by the exothermicity of a non-catalysed reaction. When the sample is injected into the GC (which was set at 498 K due to the low volatility of glycerol) the reverse reaction occurs due to a sharp decrease in the equilibrium constant with the increasing temperature, and as a result most of the hemiacetal disappears, which means the sample analysis is not representative. Similar problems were encountered when the reaction between ethanol and butanal was studied [19]. In this case, HPLC/MS analyses of ethanol–butanal mixtures at room temperature confirmed the presence of the hemiacetal in the samples [25]. In addition to this, some authors [26] have reported the unstable behaviour of hemi-acetals. As indicated in Fig. 3, four different isomers are formed in the acetalization reaction between glycerol and acetaldehyde: 5-hydroxy2-methyl-1,3 dioxane (Ac1) and 4-hydroxymethyl-2-methyl-1,3 dioxolane (Ac2), with both of them being in their cis and trans stereoisomer forms. Due to the lack of standards for GC calibration, each isomer was identified using a GC/MS. Thus, a comparison between the mass spectra obtained and the theoretical ones (Fig. 5) meant each isomer could be identified. Furthermore, it was verified that the mass spectra measured using the GC/MS were in good agreement with the ones reported by da Silva Ferreira et al. [15]. Moreover, according to the literature, Camara et al. (2003) and da Silva Ferreira et al. [14,15] have observed an isomerization reaction of the 1,3 dioxolane toward dioxane formation in the evolution of the acetals under study here. Moreover,
they have reported that 1,3 dioxolane isomers were formed faster, but then the reaction mixture proceeded towards isomerization equilibrium, generating more 1,3 dioxanes. This isomerization reaction was also observed in the case studied, as indicated in Fig. 4. By contrast, Ruiz et al. [27] have observed the isomerization from dioxanes to dioxolanes. In terms of the identification between cis and trans stereoisomers, their identification using mass spectra coincide with the results obtained by Aksnes et al. [12] after carrying out identification tests based on infrared bands and refractive indexes. However, in order to study the reaction kinetics, there is no need to distinguish all four of the acetals because at it will be seen later this acetalization reaction is irreversible. Following Sharma & Chopade's development [23,24], hemiacetal concentration can be considered approximately at equilibrium with glycerol and acetaldehyde at any time. ½HA ¼ K½AcHO½Gly:
ð1Þ
Where K is the equilibrium constant in (mol/L)−1. The formation rate of acetal (Ac) may be expressed as d½Ac ¼ wk3 ½HA−wk4 ½AC½W: dt
ð2Þ
Where w is (gcat)/(reacting volume in L) k3 and k4 are the kinetic constants. k3 in L/(gcat · min) and k4 in L2/(min · gcat) [] indicates concentration in mol/L.
186
I. Agirre et al. / Fuel Processing Technology 116 (2013) 182–188
Fig. 4. Acetal formation as a function of time with and without catalysts (700 rpm, 303 K, glycerol/acetaldehyde molar ratio 1:1).
Substituting [HA] from Eq. (1), d½Ac ¼ wk½AcHO½Gly−wk4 ½AC½W: dt
ð3Þ
Where k = k3K in L2/(gcat · min · mol). According to this theoretical development, the global reaction is reversible and first order with respect to glycerol and acetaldehyde in the forward direction, and also first order with respect to acetal and water in the reversible reaction. In order to check whether this kinetic expression is correct, different experiments were carried out using different feed mole ratios (1:1, 2:1, 3:1, glycerol to acetaldehyde). All the experiments with excess glycerol were performed at 313 K, 700 rpm and with 2.0 wt.% of Amberlyst 47. The experiments with excess acetaldehyde were carried out at 293 K. The overall conversion reaches 100% in all cases when excess glycerol was used. Moreover, the following section, Effect of temperature, shows that the final conversion achieved was also 100% at every temperature. All this evidence proves that the glycerol + acetaldehyde reaction under the operating conditions investigated is irreversible. This reaction kinetics can therefore be studied without the need for an accurate determination of the four product concentrations. All the information found in the literature about acetalization reactions shows that this type of reaction has high thermodynamic limitations. Thus, Sharma & Chopade [23,24] have achieved a maximum of 50% conversion (at 363 K) by reacting formaldehyde with ethylene
Fig. 5. Comparison of the theoretical mass spectra with the experimental one obtained by GC/MS.
glycol, and less than 50% (at 343 K) when reacting ethanol and formaldehyde. Agirre et al. [19] have studied the reaction between ethanol and butanal, and the maximum conversion achieved was also around 50% (at 313 K). Other authors, such as Deutsch et al. [18] and Silva et al. [3] have studied glycerol acetalization reactions with different aldehydes. Deutsch et al. [18] have studied the acetalization reaction of glycerol with formaldehyde using different types of solvents, and the maximum conversions achieved were between 58 and 77% (between 313 and 383 K). On the other hand, Silva et al. [3] have studied the reaction of glycerol with butanal, pentanal, hexanal, octanal and decanal, recording a peak of 80% conversion (at 343 K) with butanal. In general, they have recorded lower conversions using heavier aldehydes.
I. Agirre et al. / Fuel Processing Technology 116 (2013) 182–188
187
Fig. 8. 2,4,6 trimethyl-1,3,5 trioxane formation reaction (side reaction).
different kinetic expressions, it was concluded that the one with the best fit to the experimental data and the best agreement with the observed reaction behaviour is only a first order function of acetaldehyde concentration: d½Ac ¼ wk½AcHO: dt
Fig. 6. Effect of the initial feed ratio (700 rpm, 2 wt.% of Amberlyst 47). a) Excess glycerol (303 K) and b) excess acetaldehyde (393 K).
An integral approach was selected for the kinetic analysis. In order to test the suitability of different kinetic expressions, a fourth order Runge– Kutta integration method was used. Thus, by minimizing the sum of squares of the differences between the experimental concentrations and those predicted by several kinetic expressions, the apparent kinetic constants were calculated observing the adequacy of the fittings obtained. With a view to testing the suitability of different kinetic expressions, use was made of the experimental data obtained in every single test with different glycerol to acetaldehyde ratios. It should be noted that every experiment with a certain ratio was carried out at least three times in order to verify experimental repeatability. After testing
ð4Þ
Thus, after fitting the experimental data to Eq. (4) and estimating the corresponding kinetic parameters, the acetal concentrations measured experimentally were compared to the predicted ones. Fig. 7 shows that the agreement between them is reasonably close and, therefore, this kinetic expression (Eq. (4)) describes the reaction evolution quite accurately. Contrary to published data in the literature, the reaction studied does not have thermodynamic limitations at 313 K. However, certain recent kinetic studies carried out for the acetalization reaction between glycerol and butyraldehyde [28] at 353 K, also using Amberlyst 47 as catalyst, record conversions of more than 90%. As acetalizations are slightly exothermic reactions, if the temperature operating range in this recent study had been in the 293–313 K range, as in the present case, it is quite likely that a similar behaviour would have been recorded. When carried out at low temperatures, glycerol acetalization reactions seem to be either irreversible or nearly irreversible when >C2 aldehydes are used as the other reactant. Another important observation was that low glycerol to acetaldehyde initial molar ratios led to the formation of 2,4,6 trimethyl-1,3,5 trioxane. This compound was identified by GC/MS analysis. When glycerol was depleted in the reactor, trioxane became the main product; the conversion of acetaldehyde therefore continues (see Fig. 6b), but toward the formation of trioxane, as in the reaction scheme shown in Fig. 8. Dintzner et al. [29] and Augé and Gil [30] have explained the production of trioxanes from aldehydes using different heterogeneous acid catalysts. 3.5. Effect of temperature Four different temperatures were tested: 283 K, 293 K, 303 K and 313 K. At least three different experiments were performed for each
Fig. 7. Comparison between experimental acetal concentrations and the ones predicted by the kinetic model (Eq. (4)).
Fig. 9. Effect of temperature on the acetalization reaction. 700 rpm, feed ratio 3:1, 2 wt.% Amberlyst 47.
188
I. Agirre et al. / Fuel Processing Technology 116 (2013) 182–188
Table 2 Kinetic constants at different temperatures and Arrhenius parameters. Ln(A) (pre-exponential factor) Ea (activation energy) T(K) k −1 · g−1 (kJ/mol) (L · min−1 · g−1 cat ) (L · min cat ) 313 303 293 283
(125 ± 2) (75 ± 8) (31 ± 2) (14 ± nd)
· · · ·
10−5 10−5 10−5 10−5
14.51 ± 1.3
55.4 ± 3.2
nd: non-determined.
temperature, with the definitive kinetic parameters being the arithmetic average of the parameters calculated for each experiment. The apparent kinetic parameters were estimated using the integral method explained in Section 3.4. The experimental data obtained in each whole experiment were fitted to Eq. (4), obtaining the corresponding kinetic parameter from each experiment. As in all the previous experiments, all the tests were performed with 2.0 wt.% of Amberlyst 47 resin and at 700 rpm. In terms of the ratio of initial reactants, a 3:1 glycerol to acetaldehyde molar ratio was used, looking for less volatile mixtures. When using the stoichiometric feed ratio, the vaporization effects were significant above 295– 300 K since the bubble point of the initial mixture is 310 K (see Table 1). In the case of the acetalization reaction between glycerol and acetaldehyde, and contrary to other acetalization reactions reported in the open literature [20,23,24], it was observed that 100% conversion is achievable at each temperature. The only change observed when increasing the temperature was the significant increase in the reaction rate (Fig. 9), as is to be expected. In order to apply the kinetics of the reaction studied here to further process developments, the kinetic constants obtained at different temperatures (Table 2) were fitted to the Arrhenius correlation parameters. Table 2 also shows the pre-exponential factor (A) and the activation energy (Ea). It should be noted that the activation energy found is in good agreement with other activation energies of acetalization reactions reported in the literature [19,23,24].
4. Conclusions The most pertinent conclusion this research reaches is that the acetalization reaction between glycerol and acetaldehyde is not a thermodynamically limited reversible reaction, as are other acetalization reactions reported in the literature. The great advantage of this finding is that, in principle, the industrial production of these acetals for their use as biodiesel additives can be carried out in conventional reaction systems. Being a non-reversible reaction, it was verified that its rate within the conditions investigated depends solely on the acetaldehyde concentration, and as a result the presence of water in the initial reactant mixture does not have a negative effect on the reaction kinetics because of dilution. This is also an important finding, as water dilution helps to considerably reduce the viscosity of the reacting mixture. All this information offers new insights for glycerol conversion into value added products, such as acetals or similar diesel additives.
Acknowledgements The authors gratefully acknowledge the financial support for this work provided by the Spanish Ministry of Science and Innovation (ENE2009-12743-C04-04), the Basque Government and the University of the Basque Country (UPV/EHU). Furthermore, the authors would like to thank Rohm & Haas for kindly supplying different Amberlyst resins.
References [1] B.R. Moser, S.Z. Erhan, Branched chain derivatives of alkyl oleates: tribological, rheological, oxidation, and low temperature properties, Fuel 87 (2008) 2253–2257. [2] S.B. Umbarkar, T.V. Kotbagi, A.V. Biradar, R. Pasricha, J. Chanale, M.K. Dongare, A.S. Mamede, C. Lancelot, E. Payen, Acetalization of glycerol using mesoporous MoO3/SiO2 solid acid catalyst, Journal of Molecular Catalysis A: Chemical 310 (2009) 150–158. [3] P.H.R. Silva, V.L.C. Gonçalves, C.J.A. Mota, Glycerol acetals as anti-freezing additives for biodiesel, Bioresource Technology 101 (2010) 6225–6229. [4] H. Burtscher, U. Matter, G. Skillas, The effect of fuel additives on diesel engine particulate emissions, Journal of Aerosol Science 30 (1999) S851–S852. [5] D. Naegeli, L. Dodge, Combustion Characterization of Methylal in Reciprocating Engines, Final Report No. SwRI 03-130, Southwest Research Institute, San Antonio, TX, 1992. [6] F. Frusteri, L. Spadaro, C. Beatrice, C. Guido, Oxygenated additives production for diesel engine emission improvement, Chemical Engineering Journal 134 (2007) 239–245, (l). [7] M.R. Capeletti, L. Balzano, G. de la Puente, M. Laborde, U. Sedran, Synthesis of acetal (1,1-diethoxyethane) from ethanol and acetaldehyde over acidic catalysts, Applied Catalysis A: General 198 (2000) L1–L4. [8] N. Rahmat, A.Z. Abdullah, A.R. Mohamed, Recent progress on innovative and potential technologies for glycerol transformation into fuel additives: a critical review, Renewable & Sustainable Energy Reviews 14 (2010) 987–1000. [9] S. Yang, M. Wang, Y. Liang, J. Sun, Eco-friendly one-pot synthesis of acetals and ketals by heterogeneously catalyzed liquid–solid phase reaction, Rare Metals (Beijing, China) 25 (2006) 625–629. [10] J. Requies, M.B. Güemez, P. Maireles, A. Iriondo, V.L. Barrio, J.F. Cambra, P.L. Arias, Zirconia supported Cu systems as catalysts for n-butanol conversion to butyraldehyde, Applied Catalysis A: General 423–424 (2012) 185–191. [11] J. Requies, M.B. Güemez, A. Iriondo, V.L. Barrio, J.F. Cambra, P.L. Arias, Bio n-butanol partial oxidation to butyraldehyde in gas phase on supported Ru and Cu catalysts, Catalysis Letters 142 (2012) 417–426. [12] G. Aksnes, P. Albriktsen, P. Juvvik, Cyclic acetal and ketal isomers of glycerol, Acta Chemica Scandinavica 19 (1965) 920–930, (1947–1973). [13] G. Aksnes, P. Albriktsen, Intramolecular H bonding in cyclic acetal isomers of glycerol, Acta Chemica Scandinavica 20 (1966) 1330–1334. [14] J.S. Camara, J.C. Marques, A. Alves, A.C.S. Ferreira, Heterocyclic acetals in Madeira wines, Analytical and Bioanalytical Chemistry 375 (2003) 1221–1224. [15] A.C. da Silva Ferreira, J. Barbe, A. Bertrand, Heterocyclic acetals from glycerol and acetaldehyde in Port wines: evolution with aging, Journal of Agricultural and Food Chemistry 50 (2002) 2560–2564. [16] M. Kaufhold, M. El-Chahawi, Process for Preparing Acetaldehyde Diethyl Acetal, 5527969, 1996. (Germany). [17] T.W. Green, Protective Groups in Organic Synthesis, Willey, New York, 1981. [18] J. Deutsch, A. Martin, H. Lieske, Investigations on heterogeneously catalysed condensations of glycerol to cyclic acetals, Journal of Catalysis 245 (2007) 428–435. [19] I. Agirre, V.L. Barrio, B. Güemez, J.F. Cambra, P.L. Arias, The development of a reactive distillation process for the production of 1,1 diethoxy butane from bioalcohol: kinetic study and simulation model, International Journal of Chemical Reactor Engineering 8 (2010) A86. [20] A.D. Dhale, L.K. Myrant, S.P. Chopade, J.E. Jackson, D.J. Miller, Propylene glycol and ethylene glycol recovery from aqueous solution via reactive distillation, Chemical Engineering Science 59 (2004) 2881–2890. [21] I. Agirre, V.L. Barrio, B. Güemez, J.F. Cambra, P.L. Arias, Catalytic reactive distillation process development for 1,1 diethoxy butane production from renewable sources, Bioresource Technology 102 (2011) 1289–1297. [22] I. Agirre, I. García, J. Requies, V.L. Barrio, M.B. Güemez, J.F. Cambra, P.L. Arias, Glycerol acetals, kinetic study of the reaction between glycerol and formaldehyde, Biomass and Bioenergy 35 (2011) 3636–3642. [23] S.P. Chopade, M.M. Sharma, Acetalization of ethylene glycol with formaldehyde using cation-exchange resins as catalysts: batch versus reactive distillation, Reactive and Functional Polymers 34 (1997) 37–45. [24] S.P. Chopade, M.M. Sharma, Reaction of ethanol and formaldehyde: use of versatile cation-exchange resins as catalyst in batch reactors and reactive distillation columns, Reactive and Functional Polymers 32 (1997) 53–65. [25] B. Abad, Identificación de la presencia de hemiacetal en una mezcla etanol: butanal (2:1 en moles), Servicio general de análisis UPV/EHU, SGIKER, General Analysis Department of the University of the Basque Country, 2008. [26] G.D. Yadav, A.A. Pujari, Kinetics of acetalization of perfumery aldehydes with alkanols over solid acid catalysts, Canadian Journal of Chemical Engineering 77 (1999) 489–496. [27] V.R. Ruiz, A. Velty, L.L. Santos, A. Leyva-Pérez, M.J. Sabater, S. Iborra, A. Corma, Gold catalysts and solid catalysts for biomass transformations: valorization of glycerol and glycerol–water mixtures through formation of cyclic acetals, Journal of Catalysis 271 (2010) 351–357. [28] B. Güemez, J.M. Requies, I. Agirre, V.L. Barrio, J.F. Cambra, P.L. Arias, Acetalization reaction between glycerol and n-butyraldehyde using an acidic ion exchange resin. Kinetic modelling, Chemical Engineering Journal 228 (2013) 300–307. [29] M.R. Dintzner, Y.A. Mondjinou, D.J. Pileggi, Montmorillonite clay-catalyzed cyclotrimerization and oxidation of aliphatic aldehydes, Tetrahedron Letters 51 (2010) 826–827. [30] J. Augé, R. Gil, A convenient solvent-free preparation of 1,3,5-trioxanes, Tetrahedron Letters 43 (2002) 7919–7920.