Catal Lett (2010) 134:179–183 DOI 10.1007/s10562-009-0228-0
Low Temperature De-acidification Process of Animal Fat as a Pre-Step to Biodiesel Production C. L. Bianchi • D. C. Boffito • C. Pirola V. Ragaini
•
Received: 20 July 2009 / Accepted: 16 November 2009 / Published online: 25 November 2009 Ó Springer Science+Business Media, LLC 2009
Abstract An esterification process is proposed to lower Free Fatty Acids (FFA) content in waste animal fat using solid acid ion exchange resins as catalysts. The final aim is to make this material a suitable feedstock for biodiesel production. Its exploitation does not interfere with the food chain, besides allowing to lower biodiesel production costs. Keywords Animal fat Free fatty acids Deacidification Solid resin
1 Introduction Biodiesel is a non-toxic, biodegradable environmental friendly alternative diesel fuel. The EU 2003/30 directive of the European Commission called for two main goals: 5.75% biodiesel penetration by December 2010 and 10% by December 2020 [1, 2] and therefore the biodiesel market is expected to expand accordingly. Biodiesel can be produced through the transesterification reaction of triglycerides contained in food-grade vegetable oils, in presence of an alcohol (usually methanol) and using an alkaline catalyst, also yielding glycerol as a by-product [2–5], according to the reaction represented by the scheme in Fig. 1.
Electronic supplementary material The online version of this article (doi:10.1007/s10562-009-0228-0) contains supplementary material, which is available to authorized users. C. L. Bianchi (&) D. C. Boffito C. Pirola V. Ragaini Dipartimento di Chimica Fisica ed Elettrochimica, Universita` degli Studi di Milano, Milan, Italy e-mail:
[email protected]
Beside triglycerides, the starting oils also contain Free Fatty Acids (FFA), linear carboxylic acids in the range C14–C22, with varying unsaturation bonds. If the content of FFA in the feedstock exceeds a certain limit (0.5% by weight [2]), a higher amount of the alkaline homogeneous catalysts is required to carry out the process, being partly lost for FFA neutralisation, with the additional drawback of soap formation. Soap locates at the glycerineBD interphase, thus complicating products separation and adversely affecting the achievable yield. This issues can be avoided or minimized if edible and refined oils are used as a feedstock, thanks to the lower acidity percentage, but they raises significantly BD production costs, which makes them not competitive with petroleum-based diesel. In addition, the exploitation of food-grade oils for BD purpose is strongly in competition with the human food requirement. For this reason it is desirable to produce BD from specifically selected cultures or from waste materials, such as exhausted cooking oils or animal fats rejected from slaughtering processes. The availability of multiple and cheap fats sources is deemed to be the key to reduce the overall biodiesel cost, since nowadays the impact of feedstock price on the cost of biodiesel fuel is about more than the 85% [6]. The main drawback of these low-cost raw materials is their high content of FFA, exceeding the mentioned limit of 0.5%, typically ranging from 10 to 25% [2], making it uneconomic the use of alkaline catalyst [7], as discussed. Different approaches have been proposed to get rid of this problem. Among them we can remind the alkali refining method [8], the use of excess catalyst [7], the FFA solvent extraction [9], the distillation refining processes [10] and the pre-esterification, which seems to be the most attractive one. This approach consists in performing a de-acidification process of the oil through the esterification reaction of
123
180
C. L. Bianchi et al.
OCOR 1 OCOR2
OH 3 CH 3OH
OCOR3 Triglyceride
R1OCOCH3
basic catalysis
OH OH
Methanol
Glycerine
R2OCOCH3 R3OCOCH3
FAME (Fatty alkyl methyl ester) BIODIESEL
Fig. 1 Base-catalyzed transesterification of a triglyceride
FFA, employing homogeneous [11–13] or heterogeneous [14–16] acid catalysis, so as to obtain monoesters compatible with final biodiesel. The reaction can be represented with the scheme in Fig. 2. Heterogeneous catalysis is preferred as catalyst can be easily separated and recovered from reaction mixture. For this reason many efforts regarding the use of solid acid catalysts for the FFA esterification were recently carried out [14–16]. In the present work, the direct esterification reaction of FFA in animal fat was studied in presence of methanol and using solid acid resins as heterogeneous catalysts. Six different types of AmberlystÒ (commercially available from Dow Advanced Materials), with different physical and chemical properties, as for surface area, porosity and acidity, were tested varying catalyst/fat weigh ratio, in order to find the most catalytically active one for the studied reaction. Each test was carried out for 6 h, withdrawing samples from the reactor at pre-established times and analyzing them through titration with KOH. Pure oleic acid was added to each stock in order to have a similar starting acidity value for all the tests. Karl-Fischer analysis was also carried out to check the presence of water in the reaction medium after the esterification. A remarkable aspect of the proposed process is represented by the mild operative conditions, i.e. low temperature (between 303 and 338 K) and atmospheric pressure.
2 Experimental 2.1 Materials The employed reactants were methanol ([99%), pure oleic acid (Fluka products) and commercially available lard. Titrations were carried out using KOH 0.1 M in ethanol and dissolving the sample in a diethylether/ ethanol acid catalysis
RCOOH
CH3 OH
RCOOCH3
Fig. 2 Acid-catalyzed esterification of a carboxylic acid
123
H2 O
solution (9:1 volume ratio). KOH, diethylether ([99%), ethanol ([99,8%) and phenolphthalein (2 wt% solution in ethanol) used as indicator for titrations are Fluka products, too. For Karl Fischer analysis a Karl-Fisher reagent was used (Fluka product). The catalysts are macroreticular strongly acidic cation exchanger (solid resins): Amberlyst 15d (A15), 36d (A36), 39w (A39), 40w (A40), 46w (A46) and 70w (A70) from Dow Advanced Materials. Amberlysts have a gel type structure of spheres forming a macroporous polymer functionalised with strong –SO3H acid site (Brønsted acidity). Amberlysts are provided both in the dry form (symbol d) and in the wet form (w). The former does not require any pre-treatment as the active sites are already free and available on the resin’s surface and inside the pores; in the latter, on the contrary, active sites are surrounded by water molecules. In this case a pre-treatment was carried out for all samples in an oven for 16 h at 383 K (not exceeding the recommended maximum operating temperature) to completely remove water, so making all the active sites available. The choice of drying the catalysts in an oven rather than pre-conditioning them in methanol (easy to recover as it does not form any azeotrope with water) has been dictated by the need to grant the same starting conditions for both dry and wet Amberlysts. 2.2 Experimental Procedure and Apparatus The experimental runs were performed in a lab-scale pyrex batch reactor with a capacity of 0.250 L, equipped with a thermometer and a coil condenser. The stirring was kept with a magnetic stirrer at 300 rpm and the desired temperature was reached and maintained through a thermostatic bath. Due to the large variability of the FFA content in the starting materials, suitable amounts of oleic acid were added to all the lard samples in order to obtain an initial value of acidity between 5 and 7%. The acidified sample was preheated in the reactor to the selected reaction temperature thus obtaining a homogeneous mixture of oleic acid and fat, to which methanol was added. The catalyst was then introduced and the reaction started and carried out for 6 h, taking samples every two hours. Catalyst/fat ratio was varied in a range between 1.25 to 10% by weight. A higher amount of alcohol than the stoichiometric one (1:1 molar ratio with the acid) was used to shift the reaction towards the desired products. In addition it was necessary to take into account the high viscosity of the reaction medium, which can affect mass transfer of the reactants towards the catalyst. For these two reasons the employed methanol quantity was calculated based on the actual oil
Low Temperature De-acidification Process of Animal Fat as a Pre-Step to Biodiesel Production
mass inside the reactor rather than on the reaction stoichiometry. Consequently a 4:1 alcohol/oil molar ratio, i.e. 30:1 alcohol/FFA molar ratio, was used (16 g of methanol were added per 100 g of oil). The acidity of the alcohol/fat mixture was assumed as the initial acidity value, i.e. the one at time zero. The withdrawn samples were analyzed by acid-base titration to evaluate the FFA concentration. A weighted amount of the sample was dissolved in a mixture of diethyl ether/ethanol (9:1 wt. ratio), using phenolphtalein as an indicator and 0.1 N KOH solution as titrant. As usual in similar works [14], the acidity value was calculated as follows: a¼
V MW C 100 W
where a is FFA concentration in weight [%], V is the volume of solution needed for titration (mL), MW is the molecular weight of oleic acid (mg/mmol), C is the concentration of KOH titrating solution (mmol ml-1) and W is the weight of the analyzed sample (mg). FFA percent conversion was determined as follows: ai at FAAconv ¼ 100 ai where ai is the initial acidity value and at is the acidity value at reaction time t. At the end of each test the reaction mixture was let stand in order to allow separation between fat and methanol, so obtaining two liquid phases. Water content in each phase was determined through Karl-Fischer coulometric titration, using anhydrous methanol as solvent. Karl-Fischer determination was also performed to quantify the amount of water present inside the Amberlysts pores after the reaction.
181
Table 1 Catalysts features Catalyst
A15d A36d A39w A40w A46w A70w
Surface areaa (m2 g-1)
53
33
32
33
75
36
˚ ) 300 Ave. pore diameter (A
240
230
170
235
220
Total pore volume (cc g-1) Acidityb (meq H? g-1)
0.40
0.20
0.20
0.15
0.15
0.20
4.7
5.4
5.0
2.2
0.43
2.55
393
423
403
413
393
463
Max. operating temp. (K) a
Nitrogen BET;
b
Dry weight
the actual active sites necessary for FFA esterification reaction. In particular, A 46 distinguishes itself for being sulfonated only on its surface and not inside pores. The differences among all the samples concern both the physical and the chemical properties: in particular the surface area varies in the range between 32 and 75 m2g-1, ˚ and the the average pores diameter between 170 and 300 A ? -1 acidity between 0.43 and 5.4 meq. H g . 3.2 Deacidification Tests Before starting catalytic tests, a blank test at 338 K was performed with just the fat/methanol acidified mixture, confirming that in absence of catalyst no FFA conversion occurs. The results of the deacidification tests are summarized in Figs. 3 to 5. Figure 3 shows the comparison among the results obtained by all the tested resins. Both the effect of reaction
3 Results and Discussion 3.1 Catalysts The main features of the employed catalysts are reported in Table 1 [17]. These resins have the appearance of gel structured spheres made up of a macroporous copolymer of styrene crosslinked with divinylbenzene [18]. The amount of crosslinking component determines the main physical characteristics (surface area and pore size distribution) of these catalysts [19]. The higher DVB’s quantity added during the copolimeration, the higher the reticulation degree. Indeed, the amount of divinylbenzene plays the major role in controlling the physical features of these catalysts. The polymer so obtained is functionalised with the addition of strong –SO3H acid groups, which represent
Fig. 3 FFA conversion (%) after 6 h. Comparison of Amberlysts in different operative conditions. Black bar T = 338 K, 10% per weight (wt) of catalyst vs. fat; grey bar: T = 323 K, 10% wt of catalyst vs. fat (cat/fat); white bar: T = 338 K, 5% wt cat/fat. The dotted line represents the value of FFA conversion necessary to obtain a lard with a FFA content \0.5% per weight, i.e. suitable for industrial applications
123
182
Fig. 4 FFA conversion (%) vs reaction time in presence of A70, catalyst/fat ratio per weight 10% of the catalyst and different reaction temperatures: square = 338 K, diamond = 323 K, filled triangle = 313 K, filled circle = 303 K. The dotted line represents the value of FFA conversion necessary to obtain a lard with a FFA content \0.5% per weight, i.e. suitable for industrial applications
Fig. 5 FFA conversion (%) vs. time in presence of A70 with different amounts of the catalysts: square = 10%, diamond = 5%, filled triangle = 2.5%, filled circle = 1.25%, at 338 K. The dotted line represents the value of FFA conversion necessary to obtain a lard with a FFA content \0.5% per weight, i.e. suitable for industrial applications
temperature (T = 338 or 323 K) and catalyst amount vs. fat (10% or 5% per weight) was investigated. As expected, the final conversion increases with the increase of the temperature, being esterification an endothermic reaction [15]. At 338 K (methanol incipient boiling temperature) all the catalysts show excellent results giving FFA conversions higher than 90%; this means we obtain a lard with a FFA content lower than 0.5% and then suitable for the industrial subsequent transesterification process. This limit for industrial application is represented in Fig. 3, 4 and 5 by the dotted line corresponding to a FFA conversion of 90%. The catalyst A 40 is the only sample not leading to this result; its unsatisfactory performance can be explained taking into account the chemical and physical properties of this catalyst, showing both a lower specific
123
C. L. Bianchi et al.
surface area and a lower acid site concentration, if compared to other Amberlysts. Being these two parameters directly connected to catalytic activity, their simultaneous deficiency is clearly the cause of the unsatisfactory performance shown by A40 in all the working condition, which confirms that this catalyst is not suitable for the FFA’s esterification. On the contrary, the catalytic performances of the sample A46 appear to be remarkable, in spite of the low concentration of active acid sites (the lowest among all the tested Amberlyst). This result can be explained considering the particular configuration of the catalytic particles, characterised by an innovative egg-shell structure where acid sites are located only on its surface, thus being immediately and easily available for the reaction [17], overcoming the internal diffusional limitations, occurring with catalytic sites distributed along internal pores. The reduction of both the working temperature to 323 K and of the catalyst amount to 5% (at 338 K) leads to a decrease in the catalytic performance, with the exception of the A70, which maintains an adequate FFA conversion ([90%). A70 shows the best results in all the operation conditions. Surprisingly, this catalyst is characterized by a high maximum operating temperature but by a low surface area and a medium value of average size pore distribution and acidity (see Table 1). Therefore, it is difficult to find a simple correlation between its activity and its physicochemical characteristics, apart from stating that its characteristics, intermediate vs. the other catalyst samples, represent the best trade-off for the applied test conditions. Given its good catalytic activity, A70 was further tested to evaluate its performances in milder operating conditions, i.e. lower temperatures and lower catalyst/fat ratio. The results thus obtained are displayed in Figs. 4 and 5, showing that the activity of the catalyst decreases as the reaction temperature or its concentration decrease. However it is worth remarking that even at room temperature (T = 303 K) or with a very little amount of the catalyst, as low as 1.25 wt% vs. fat, the FFA conversion, after 6 h of reaction, is still interesting (60 and 30%, respectively). Comparable studies carried out at low reaction temperatures [15] or with little amounts of catalyst [14, 15, 19] show lower FFA conversion than those obtained with A70. It is interesting to note that the trend of FFA conversion is analogous to the one observed by the authors in the esterification of water-diluted acetic acid streams [20–22]. The acid/triglyceride system hereby described has a more complex matrix if compared to the acetic acid/water one. In the former, in fact, the sites to be esterificated (FFA) are part of the triglyceride substrate, differently from acetic acid in water. Regardless to this, the two systems seem to have a similar behaviour.
Low Temperature De-acidification Process of Animal Fat as a Pre-Step to Biodiesel Production
Karl- Fischer analysis revealed that water developed from the esterification reaction was mainly located inside methanol (*80%) and inside Amberlysts (*20%). This can be explained considering that water molecules, due to their polarity, can be adsorbed on the SO3- H? active group, with dipole-dipole interaction, so surrounding H? acid sites and making them unavailable for catalytic action. This cause of deactivation can also explain the flattening typically observed in all the curves (Figs. 4 and 5): as the reaction proceeds, water develops in major quantity and plugs the active sites of Amberlysts causing a decrease of their catalytic activity. Another possible cause of deactivation is the mechanical stress to which catalyst is subject during the stirring. In fact, the collisions of catalyst’s particles, each against the other and against the inner walls of the reactor, cause fragmentation. The debris generated, not catalytically active (inner part of the particles), can collapse over the active sites hindering their action. Actually A70 is characterised by the highest maximum operating temperature among all the Amberlysts and this can be also ascribed to its higher mechanical resistance. As well as the other Amberlysts catalysts, also A70’s shows with time a decrease of FFA conversion (see Fig. 4), but it reaches a plateau at higher yields.
4 Conclusions A de-acidification process of low-cost animal fat was performed using six different commercial catalysts, Amberlysts, solid acid resins suitable for esterification processes. A FFA conversion greater than 90%, corresponding to a FFA concentration in the lard lower than 0.5%, is of practical interest for direct biodiesel production. This important result, from this type of feedstock, was reached with all the tested catalysts at a temperature of 338 K and a catalyst vs. fat concentration of 10 wt%. Only sample A40 showed unsatisfactory results. Decreasing both the reaction temperature and the amount of catalyst, the conversion drops below 90% for all the tested catalysts, but A70, which keeps providing good performances even in less severe operating conditions. In particular, good FFA conversions are observed even
183
working at 303 K or with a catalyst concentration vs. fat of only 1.25 wt%. Is it expected that a further optimization of A70 catalysts and a better tuning with the reaction conditions can lead to a relevant decrease of the whole cost of biodiesel production, combining the use of low cost feedstock based on not-refined waste materials, such as animal fats, with a more economical process, carried out in mild conditions. Acknowledgment The authors are grateful to Dow Advanced Materials for kindly providing the catalysts employed in the present work.
References 1. Bozbas K (2008) Renew Sustain Energy Rev 12:542 2. Canakci M, Sanli H (2008) J Ind Microbiol Biotechnol 35:431 3. Pinto AC, Guariero LLN, Rezende MJC, Ribeiro NM, Torres EA, Lopes WA, de P Pereira PA, de Andrade JB (2005) J Braz Chem Soc 16B:1313 4. Agarwal AK (2007) Prog Energy Comb Sci 33:233 5. Ma F, Hanna MA (1999) Bioresour Technol 70:1 6. Haas MJ, McAllon AJ, Yee WC, Foglia TA (2005) Bioresour Technol 97:671 7. Jeromin L, PeuKert E, Wollomann G (1979) US Patent 4, 164, 506 8. Suwannakarn K, Loreto E, Goodwin JG Jr, Lu C (2008) J Catal 255:279 9. Rao KV, Krishnasamy S, Penumarthy V (2009) WO 2009047793 10. Xingzhong Y, Jia L, Guanming Z, Jingang S, Jingyi T, Guohe H (2008) Renew Energy 33:1678 11. Dube´ MA, Zheng S, Mclean DD, Kates M (2004) J Am Oil Chem Soc 81:599 12. Zheng S, Kates M, Dube` MA, McLean DD (2006) Biomass Bioenergy 30:267 13. Zullaikah S, Lai CC, Vali SR, Ju YH (2005) Bioresour Technol 96:1889 14. Marchetti JM, Errazu AF (2008) Fuel 87:3477 15. Marchetti JM, Miguel VU, Errazu AF (2007) Fuel 86:906 16. Di Serio M, Cozzolino M, Giordano M, Tesser R, Patrono P, Santacesaria E (2007) Ind Eng Chem Res 46:6379 17. Dow Advanced Materials, Products Information Sheets available at the site www.rohmhaas.com 18. Chakrabarti A, Sharma MM (1993) React Polym 20:1 19. Oezbay N, Oktar N, Tapan NA (2008) Fuel 87:1789 20. Bianchi CL, Ragaini V, Pirola C, Carvoli G (2003) Appl Catal B 40:93 21. Ragaini V, Bianchi CL, Pirola C, Carvoli G (2006) Appl Catal B 64:66 22. Ragaini V, Bianchi CL, Pirola C (2007) Chem Eng J 131:257
123
