1. Introduction Biodiesel (BD) is a liquid biofuel that is defined as a fatty acid methyl ester fulfilling standards such as the ones set by European (EN 14214) and the American (ASTM 6751) regulations. BD is obtained by the transesterification (Scheme 1.1) or alcoholysis of natural triglycerides contained in vegetable oils, animal fats, waste fats and greases, waste cooking oils (WCO) or side-stream products of refined edible oil production with short-chain alcohols, usually methanol or ethanol and using an alkaline homogeneous catalyst (Perego and Ricci, 2012). Scheme 1.1. Transesterification reaction. BD presents several advantages over petroleum-based diesel such as: biodegradability, lower particulate and common air pollutants (CO, SOx emissions, unburned hydrocarbons) emissions, absence of aromatics and a closed CO2 cycle. Refined, low acidity oilseeds (e.g. those derived from sunflower, soy, rapeseed, etc.) may be easily converted into BD, but their exploitation significantly raises the production costs, resulting in a biofuel that is uncompetitive with the petroleum-based diesel (Santori et al., 2012; Lotero et al., 2005). Moreover, the use of the aforementioned oils generated a hot debate about a possible food vs. fuel conflict, i.e. about the risk of diverting farmland or crops at the expense of food supply. It is so highly desirable to produce BD from crops specifically selected for their high productivity and low water requirements (Bianchi et al., 2011; Pirola et al., 2011), or from low-cost feedstock such as used frying oils (Boffito et al., 2012a) and animal fats (Bianchi et al., 2010). The value of these second generation biofuels, i.e. produced from crop and forest residues and from non-food energy crops, is acknowledged by the European Community, which states in its RED directive (European Union, RED Directive 2009/28/EC): ‘‘For the purposes of demonstrating compliance with national renewable energy obligations […], the contribution made by biofuels produced from wastes, residues, non-food cellulosic material, and ligno-cellulosic material shall be considered to be twice that made by other biofuels’’. However, the presence of free fatty acids in the feedstock, occurring in particular in the case of not refined oils, causes the formation of soaps as a consequence of the reaction with the alkaline catalyst. This hinders the contact between reagents and the catalyst and makes difficult the products separation. Many methods have been proposed to eliminate FFA during or prior to transesterification (Pirola et al., 2011; Santori et al., 2012). Among these the FFA pre-esterification method is a very interesting approach to lower the acidity since it allows to lower the acid value as well as to obtain methyl esters already in this preliminary step (Boffito et al., 2012a, 2012b; 2012c Bianchi et al., 2010, 2011; Pirola et al., 2010, 2011). Aims of the work The aims of this work are framed in the context of the entire biodiesel production chain, ranging from the choice of the raw material, through its standardization to the actual biodiesel production. The objectives can be therefore summarized as follows: Assessing the potential of some vegetable or waste oils for biodiesel production by their characterization, deacidification and final transformation into biodiesel; To test different ion exchange resins and sulphated inorganic systems as catalysts in the FFA esterification; To assess the use of ultrasound to assist the sol-gel synthesis of inorganic sulphated oxides to be used as catalysts in the FFA esterification reaction; To assess the use of sonochemical techniques such as ultrasound and microwave to promote both the FFA esterification and transesterification reaction. 2. Experimental details 2.1 Catalysts In this work, three kinds of acid ion exchange resins were used as catalysts for the FFA esterification: Amberlyst®15 (A15), Amberlyst®46 (A46) (Dow Chemical) and Purolite®D5081 (D5081). Their characteristic features are given in Tab. 2.1. Various sulphated inorganic catalysts, namely sulphated zirconia, sulphated zirconia+titania and sulphated tin oxide were synthesized using different techniques. Further details will be given as the results inherent to these catalysts will be presented. Catalyst A15 A46 D5081 Physical form opaque beads Type Macroreticular Matrix Styrene-DVB Cross-linking degree medium medium high Functional group -SO3H Functionalization internal external external external Form dry wet wet Surface area (m2 g-1) 53 75 514a Ave. Dp (Ǻ) 300 235 37a Total Vp (ccg-1) 0.40 0.15 0.47 Declared Acidity (meq H+g-1) 4.7 0.43 0.90-1.1 Measured acidity (meq H+g-1) 4.2 0.60 1.0 Moisture content (%wt) 1.6 26-36 55-59 Shipping weight (g l-1) 610 600 1310a Max. operating temp (K) 393 393 403 Tab. 2.1. Features of the ion exchange resins used as catalysts. The acidity of all the catalysts was determined by ion exchange followed by pH determination as described elsewhere (López et al., 2007; Boffito et al., 2012a; 2012b). Specific surface areas were determined by BET (Brunauer, Emmett and Teller, 1938) and pores sizes distribution with BJH method (Barrett, Joyner and Halenda, 1951). XRD, XPS SEM-EDX and HR-TEM analyses were performed in the case of catalysts obtained with the use of ultrasound (Boffito et al. 2012a). Qualitative analyses of Lewis and Brønsted acid sites by absorption of a basic probe followed by FTIR analyses was also carried out for this class of catalysts (Boffito et al, 2012a). 2.2 Characterization of the oils Oils were characterized for what concerns acidity (by acid-base titrations) as reported by Boffito et al. (2012a, 2012b; 2012c), iodine value (Hannus method (EN 14111:2003)), saponification value (ASTM D5558), peroxide value and composition by GC analyses of the methyl ester yielded by the esterification and transesterification. Cetane number and theoretical values of the same properties were determined using equations already reported elsewhere (Winayanuwattikun et al., 2008). 2.3 Esterification and transesterification reactions In Tab. 2.2, the conditions adopted in both the conventional and sonochemically-assisted esterification are reported. For all these experiments a temperature of 336 K was adopted. Vials were used to test the sulphated inorganic oxides, while Carberry reactor (confined catalyst) (Boffito et al., 201c) was used just for the FFA esterification of cooking oil. Rector oil (+ FFA) (g) MeOH (g) catalyst amount vial 21 3.4 5%wt/gFFA sulphated inorganic catalysts slurry 100 16 - 10 g ion exchange resins - 5%wt/gF FA sulphated inorganic catalysts Carberry 300 48 10 g (5 g in each basket) Tab. 2.2. Free fatty acids esterification reaction conditions for conventional and sonochemically-assisted experiments. All the sonochemically-assisted experiments were performed in a slurry reactor. FFA conversions were determined by acid-base titrations of oil samples withdrawn from the reactors at pre-established times and calculated as follows: "FFA conversion (%)=" (〖"FFA" 〗_"t=0" "-" 〖"FFA" 〗_"t" )/〖"FFA" 〗_"t=0" " x 100" In Tab. 2.3, the conditions of both the conventional and ultrasound (US)-assisted transesterification are reported. KOH and CH3ONa were used for conventional experiments, while just KOH for the US-assisted experiments. The BD yield was determined by GC (FID) analysis of the methyl esters. Method Reactor Step gMeOH/100 goil gKOH/100 goil Temp. (K) Time (min) traditional batch step 1 20 1.0 333 90 step 2 5.0 0.50 60 US-assisted batch step 1 20 1.0 313, 333 30 US-assisted continuos step 1 20 1.0 338 30 Tab. 2.3. Transesterification reaction conditions. 3. Results and Discussion 3.1 Characterization and deacidification of different oils by ion exchange resins: assessment of the potential for biodiesel production In Tab. 3.1 the results of the characterization of the oils utilized in this work are displayed. The value in parentheses indicate the theoretical value of the properties, calculated basing on the acidic composition. The acidity of all the oils exceeds 0.5%wt (~0.5 mgKOH/g), i.e. the acidity limit recommended by both the European normative (EN 14214) and American standard ASTM 6751 on biodiesel (BD). The iodine value (IV) is regulated by the EN 14214, which poses an upper limit of 120 gI2/100 g. The number of saturated fatty chains in the fuel determines its behaviour at low temperatures, influencing parameters such as the cloud point, the CFPP (cold filter plugging point) and the freezing point (Winayanuwattikun et al., 2008). The IV are in most of the cases similar to the ones calculated theoretically. When the experimental IV differs from the theoretical one, it is in most of the cases underestimated. This can be explained considering the peroxide numbers (PN), which indicates the concentration of O2 bound to the fatty alkyl chains and is therefore an index of the conservation state of oil. Oils with high IV usually have a high concentration of peroxides, whereas fats with low IV have a relatively low concentration of peroxides at the start of rancidity (King et al., 1933). Moreover, although PN is not specified in the current BD fuel standards, it may affect cetane number (CN), a parameter that is regulated by the standards concerning BD fuel. Increasing PN increases CN, altering the ignition delay time. Saponification number (SN) is an index of the number of the fatty alkyl chains that can be saponified. The long chain fatty acids have a low SN because they have a relatively fewer number of carboxylic functional groups per mass unit of fat compared to short chain fatty acids. In most of the cases the experimental SN are lower than the ones calculated theoretically. This can be explained always considering the PN, indicating a high concentration of oxygen bound to the fatty alkyl chains. Oil Acidity (%wt) IV1 (gI2/ 100 g) PN2 (meqO2 /kg) SN3 (mg KOH/g) CN4 Fatty acids composition (%wt) animal fat (lard)* 5.87 51 2.3 199 62.3 n.d. soybean* 5.24 138 3.8 201 42.4 n.d. tobacco1 1.68 143 (149) 21.9 199 (202) 41.6 (39.8) C14:0 (2.0) C16:0 (8.3) C18:0 (1.5) C18:1 (12.0) C18:2 (75.3) C18:3 (0.6) C20:0 (0.1) C22:0 (0.2) sunflower* 3.79 126 3.7 199 45.4 n.d. WSO5 0.50 118 (129) 71.3 187 (200) 48.9 (44.6) C16:0 (6.9) C18:0 (0.9) C18:1 (40.1) C18:2 (50.9) C18:3 (0,3) C20:0 (0.1) C20:1 (0.4) C22:0 (0.4) palm 2.71 54.0 (53.0) 12.3 201 (208) 61.3 (60.6) 16:0 (43.9) 18:0 (5.6) 18:1 (40.5) 18:2 (8.6) WCO6 2.10 53.9 (50.7) 11.0 212 (196) 59.9 (62.7) C16:0 (38.8) C18:0 (4.1) C18:1 (47.9) C18:2 (4.2) WCO:CRO =3:1 2.12 69.0 (75.5) 30.1 200 (212) 58.1 (55.1) C16:0 (30.1) C18:0 (3.1) C18:1 (51.9) C18:2 (12.0) C18:3 (2.%) C20:0 (0.2) C22:0 (0.1) WCO:CRO =1:1 2.19 76.8 (90.7) 51.3 188 (203) 58.1 (52.8) C16:0 (21.5) C18:0 (2.1) C18:1 (55.8) C18:2 (14.7) C18:3 (5.1) C20:0 (0.8) C22:0 (0.1) WCO:CRO =1:3 2.24 84.5 (104) 62.4 177 (202) 58.1 (49.9) 14:0 (0.1) 16:0 (14.7) 16:1 (0.7) 18:0 (6.85) 18:1 (40.0) 18:2 (37.0) 18:3 (0.25) 20:0 (0.25) 22:0 (0.15) rapeseed (CRO7) 2.20 118 (123) 71.6 165 (200) 52.8 (45.9) C16:0 (4.1) C18:0 (0.1) C18:1 (63.7) C18:2 (20.2) C18:3 (10.2) C20:0 (1.5) C22:0 (0.2) rapeseed* 4.17-5.12 108 (107) 3.5 203 (200) 48.9 (49.5) C16:0 (7.6) C18:0 (1.3) C18:1 (64.5) C18:2 (23.7) C18:3 (2.4) C20:0 (0.5) Brassica juncea 0.74 109 (110) 178 (185) 52.4 (51.1) C16:0 (2.4) C18:0 (1.1) C18:1 (19.9) C18:2 (19.2) C18:3 (10.9) C20:0 (7.2) C20:1 (1.7) C22:0 (0.9) C22:1 (34.8) 24:0 (1.9) safflower 1.75 139 48.9 170 47.1 n.d. WCO: tobacco2 =1:1 4.34 119 (112) 56.0 191 (203) 48.1 (48.0) C16:0 (22.5) C18:0 (3.2) C18:1 (32.0) C18:2 (42.1) C18:3 (0.2) tobacco2 6.17 141 (151) 33.4 183 (201) 44.4 (39.5) C16:0 (8.7) C18:0 (1.6) C18:1 (12.8) C18:2 (76.0) C18:3 (0.7) C20:0 (0.1) C22:0 (0.1) 1Iodine value; 2Peroxide number; 3Saponification number; 4Cetane number; 5Winterized sunflower oil, 6Waste cooking oil; 7Crude rapeseed oil; * refined, commercial oils acidified with pure oleic acid up to the indicated value. Tab. 3.1. Results of the characterization of the oils. The results of the FFA esterification performed on the different oils are given in Fig. 3.1. Fig. 3.1. Results of the FFA esterification reaction on different oils. The dotted line represents a FFA concentration equal to 0.5%wt, i.e. the limit required by both the European and American directives on BD fuel and to perform the transesterification reaction avoiding excessive soaps formation. The FFA esterification method is able to lower the acidity of most of the oils using the ion exchange resins A46 and D5081 as catalysts in the adopted reaction conditions. High conversion was obtained with A15 at the first use of the catalyst, but then its catalytic activity drastically drops after each cycle. The total loss of activity was estimated to be around 30% within the 5 cycles (results not shown for the sake of brevity). A possible explanation concerning this loss of activity may be related to the adsorption of the H2O yielded by the esterification on the internal active sites, which makes them unavailable for catalysis. When H2O molecules are formed inside the pores, they are unable to give internal retro-diffusion due to their strong interaction with H+ sites and form an aqueous phase inside the pores. The formation of this phase prevents FFA from reaching internal active sites due to repulsive effects. What appears to influence the FFA conversion is the refinement degree of the oil. WCO is in fact harder to process in comparison to refined oils (Bianchi et al., 2010; Boffito et al., 2012c), probably due to its higher viscosity which results in limitations to the mass transfer of the reagents towards the catalyst. Indeed, the required acidity limit is not achieved within 6 hours of reaction. A FFA concentration lower than 0.5%wt is not achieved also in the case of WCO mixture 3:1 with CRO and 1:1 with tobacco oil and in the case of the second stock of tobacco oil (tobacco2). This is attributable to the very low quality of these feedstocks due to the waste nature of the oil itself, in the case of WCO, or to the poor conservation conditions in the case of tobacco oilseed. In this latter case, the low FFA conversion was also ascribed to the presence of phospholipids, responsible for the deactivation of the catalyst. BD yields ranging from 90.0 to 95.0 and from 95.0 to 99.9% were obtained from deacidified raw oils using KOH and NaOCH3 as a catalyst, respectively. In Fig. 3.2, the comparison between A46 and D5081 at different temperatures and in absence of drying pretreatment (wet catalyst) is displayed. As expected, D5081 performs better than A46 in all the adopted conditions. Nevertheless, the maximum conversion within a reaction time of 6 hours is not achieved by any of the catalysts both operating at 318 K and in the absence of drying pretreatment. A more detailed study on the FFA esterification of WCO and its blends with rapeseed oil and gasoline was carried out. In Tab. 3.2 a list of all the experiments performed with WCO is reported together with the FFA conversion achieved in each case, while in Fig. 3.3 the influence of the viscosity of the blends of WCO is shown. Fig. 3.2. Comparison between the catalysts. D5081 and A46 at a) different catalysts amounts and b) temperatures and treatments. The results show that Carberry reactor is unsuitable for FFA esterification since a good contact between reagents and catalyst is not achieved due to its confinement. A15 deactivated very rapidly, while A46 and D5081 maintained their excellent performance during all the cycles of use due to the reasons already highlighted previously. The blends of WCO and CRO show an increase of the reaction rate proportional to the content of the CRO, that is attributable to the decreases viscosity (Fig. 3.3), being all the blend characterized by the same initial acidity. Also the use of diesel as a solvent resulted in a beneficial effect for the FFA esterification reaction, contributing to the higher reaction rate. Feedstock %wtFFAt=0 Reactor Cat. gcat/100 goil gcat/100 g feedstock Number of cat. re-uses FFA conv. (%), 1st use, 6 hr 1 WCO 2.10 Carberry A15 3.3 3.3 6 15.4 2 WCO 2.10 slurry A15 10 10 6 71.7 3 WCO 2.10 Carberry A46 3.3 3.3 6 7.7 4 WCO 2.10 slurry A46 10 10 6 62.0 5 WCO 2.10 slurry D5081 10 10 6 63.7 6 CRO 2.20 slurry A46 10 10 10 95.9 7 CRO 2.20 slurry D5081 10 10 10 93.7 8 WCO 2.10 slurry A46 10 10 0 62.0 9 WCO 75 CRO 25 2.12 7.5 71.3 10 WCO 50 CRO 50 2.19 5.0 79.9 11 WCO 25 CRO 75 2.24 2.5 86.1 12 CRO 2.20 10 95.9 13 WCO 75 DIESEL 25 1.74 7.5 76.8 14 WCO 50 DIESEL 50 1.17 5.0 58.7 15 WCO 25 DIESEL 75 0.65 2.5 40.4 16 WCO 25 DIESEL 75 (higher FFA input) 2.44 2.5 63.5 Tab. 3.2. Experiments performed with waste cooking oil. . Fig. 3.3. FFA conversions and viscosities of the blend of WCO with rapeseed oil. 3.2. Sulphated inorganic oxides as catalysts for the free fatty acid esterification: conventional and ultrasound assisted synthesis Conventional syntheses In Tab. 3.3, the list of all the catalyst synthesized with conventional techniques is reported together with the results of the characterization. Catalyst Composition Prep. method precursors T calc. SSA (m2g-1) Vp (cm3g-1) meq H+g-1 1 SZ1 SO42-/ZrO2 one-pot sol-gel ZTNP1, (NH4)2SO4 893 K O2 107 0.09 0.90 2a SZ2a SO42-/ZrO2 two-pots sol-gel ZTNP, H2SO4 893 K 102 0.10 0.11 2b SZ2b SO42-/ZrO2 two-pots sol-gel ZTNP, H2SO4 653 K 110 0.10 0.12 3 SZ3 SO42-/ZrO2 Physical mixing ZrOCl2.8H2O (NH4)2SO4 873 K 81 0.11 1.3 4 SZ4 Zr(SO4)2/SiO2 Impregnation Zr(SO4)2.4H2O SiO2 873 K 331 0.08 1.4 5 SZ5 Zr(SO4)2/Al2O3 Impregnation Zr(SO4)2.4H2O Al2O3 873 K 151 0.09 0.67 6 ZS Zr(SO4)2.4H2O (commercial) - - - 13 0.12 9.6 7 STTO_0 SO42-/SnO2 Physical mixing + impregnation SnO2 TiO2 P25 H2SO4 773 K 16.8 0.10 3.15 8 STTO_5 SO42-/95%SnO2-5%TiO2 773 K 15.9 0.11 3.43 9 STTO_10 SO42-/ 90%SnO2-10%TiO2 773 K 16.5 0.09 5.07 10 STTO_15 SO42-/ 85%SnO2-15%TiO2 773 K 14.9 0.11 7.13 11 STTO_20 SO42-/ 80%SnO2-20%TiO2 773 K 16.9 0.09 7.33 Tab. 3.3. Sulphated inorganic catalysts synthesized with conventional techniques. The FFA conversions of the sulphated Zr-based systems are provided in Fig. 3.4a and show that Zr-based sulphated systems do not provide a satisfactory performance in the FFA esterification, probably due to their low acid sites concentration related to their high SSA. Even if catalysts such as SZ3 and SZ4 exhibit higher acidity compared to other catalysts, it is essential that this acidity is located mainly on the catalyst surface to be effectively reached by the FFA molecules, as in the case of ZS. In Figure 3.4b, the results of the FFA esterification tests of the sulphated Sn-Ti systems are shown. Other conditions being equal, these catalysts perform better than the sulphated Zr-based systems just described. This is more likely due to the higher acidity along with a lower surface area. With increasing the TiO2 content, the acidity increases as well. This might be ascribable to the charge imbalance resulting from the heteroatoms linkage for the generation of acid centres, (Kataota and Dumesic, 1988). As a consequence, the activity increases with the TiO2 content along with the acidity of the samples. For the sake of clarity, in Fig. 3.4c the FFA esterification conversion is represented as a function of the number of active sites per unit of surface area of the samples. Ultrasound- assisted synthesis In Tab. 3.4, the list of all the catalyst synthesized with conventional techniques is reported together with the results of the characterization. Samples SZ and SZT refer to catalysts obtained with traditional sol-gel method, while samples termed USZT refer to US-obtained sulphated 80%ZrO2-20%TiO2. The name is followed by the US power, by the length of US pulses and by the molar ratio of water over precursors. For example, USZT_40_0.1_30 indicates a sample obtained with 40% of the maximum US power, on for 0.1 seconds (pulse length) and off for 0.9 seconds, using a water/ZTNP+TTIP molar ratio equal to 30. SZT was also calcined at 773 K for 6 hours, employing the same heating rate. This sample is reported as SZT_773_6h in entry 2a. Further details about the preparation can be found in a recent study (Boffito et al., 2012b). Entry Catalyst Acid capacity (meq H+/g) SSA (m2g-1) Vp (cm3g-1) Ave. BJH Dp (nm) Zr:Ti weight ratio S/(Zr+Ti) atomic ratio 1 SZ 0.30 107 0.20 6.0 100 0.090 2 SZT 0.79 152 0.19 5.0 79:21 0.085 2a SZT_773_6h 0.21 131 0.20 5.0 n.d.1 n.d 3 USZT_20_1_30 0.92 41.7 0.12 12.5 80:20 0.095 4 USZT_40_0.1_30 1.03 47.9 0.11 9.5 81:19 0.067 5 USZT_40_0.3_30 1.99 232 0.27 4.5 81:19 0.11 6 USZT_40_0.5_7.5 1.70 210 0.20 5.0 78:22 0.086 7 USZT_40_0.5_15 2.02 220 0.20 5.0 80:20 0.13 8 USZT_40_0.5_30 2.17 153 0.20 5.0 78:22 0.12 9 USZT_40_0.5_60 0.36 28.1 0.10 10 79:21 0.092 10 USZT_40_0.7_30 1.86 151 0.16 5.0 78:22 0.11 11 USZT_40_1_15 3.06 211 0.09 7.0 80:20 0.15 12 USZT_40_1_30 1.56 44.1 0.09 7.0 80:20 0.17 Tab. 3.4. Sulphated inorganic Zr-Ti systems synthesized with ultrasound-assisted sol-gel technique. Some of the results of the characterizations are displayed in Tab. 3.4. The results of the catalytic tests are shown in Fig. 3.5 a, b and c. In Fig. 3.5a and 3.5b the FFA conversions are reported for the samples synthesized using the same or different H2O/precursors ratio, respectively. Fig. 3.5. FFA conversions of sulphated inorganic Zr-Ti systems synthesized with ultrasound-assisted sol-gel for a) the same amount of H2O, b) different amount of H2O used in the sol-gel synthesis, c) in function of the meq of H+/g of catalyst Both the addition of TiO2 and the use of US during the synthesis are able to improve the properties of the catalysts and therefore the catalytic performance in the FFA esterification. The addition of TiO2 is able to increase the Brønsted acidity and, as a consequence, the catalytic activity (compare entries 1 and 2 in Tab. 3.4). The improvement in the properties of the catalysts due the use of US is probably caused by the effects generated by acoustic cavitation. Acoustic cavitation is the growth of bubble nuclei followed by the implosive collapse of bubbles in solution as a consequence of the applied sound field. This collapse generates transient hot-spots with local temperatures and pressures of several thousand K and hundreds of atmospheres, respectively (Sehgal et al., 1979). Very high speed jets (up to 100 m/s) are also formed. As documented by Suslick and Doktycz (Suslick and Doktycz, 1990), in the presence of an extended surface, such as the surface of a catalyst, the formation of the bubbles occurs at the liquid-solid interface and, as a consequence of their implosion, the high speed jets are directed towards the surface. The use of sonication in the synthesis of catalysts can therefore improve the nucleation production rate (i.e. sol-gel reaction production rate) and the production of surface defects and deformations with the formation of brittle powders (Suslick and Doktycz, 1990). For the samples obtained with the US pulses with on/off ratio from 0.3/0.7 on, the conversion does not increase much more compared to the one achieved with the sample obtained via traditional sol-gel synthesis. Their conversion is in fact comparable (see samples USZ_40_0.3_30, USZ_40_0.5_30, USZ_40_0.7_30 and SZT in Fig. 3.5a. The similarity in the catalytic performance of these catalysts may be ascribable to the fact that they are characterized by comparable values of SSA (entries 2, 5, 8, 10 in Tab. 3.4) and, in the case of the catalysts obtained with pulses, also by comparable acidities (entries 5, 8, 10 in Tab. 3.4). A high SSA may in fact be disadvantageous for the catalysis of the reaction here studied for the reasons already highlighted in the previous sections. The best catalytic performance is reached by the sample USZT_40_1_30, i.e. the one obtained using continuous US at higher power. This catalyst results in fact in a doubled catalytic activity with respect to the samples prepared either with the traditional synthesis or with the use of pulsed US. In spite the acidity of this catalyst is lower than that of the samples obtained with the US pulses, it is characterized by a rather low surface area (entry 12 in Tab. 3.4) that can be associated with a localization of the active sites mainly on its outer surface. As evidenced by the FTIR measurements (not reported for the sake of brevity), it is also important to highlight, that only in the case of the USZT_40_1_30 sample, a not negligible number of medium-strong Lewis acid sites is present at the surface, together with a high number of strong Brønsted acid centres. The XRD patterns of the samples were typical of amorphous systems, due to the low calcination temperatures. Samples calcined for a long time (SZT_773_6h) exhibit almost no catalytic activity (results not reported for the sake of brevity). This catalytic behaviour might be ascribable to the loss of part of the sulphates occurred during the calcinations step that result also in a very low acid capacity (see Tab. 3.4). For the sake of clarity, in Fig. 3.5c the FFA conversions as a function of the concentration of the acid sites normalized to the surface area are reported for the most significant samples. For what concerns how the water/precursors ratio affects the catalysts acidity, some general observations can be made: increasing it up to a certain amount increases the H+ concentration (compare entries from 6 to 9 and 11 to 12 in Tab. 3.4) because the rate of the hydrolysis and the number of H2O molecules that can be chemically bounded increases. Nevertheless, increasing the water/precursor ratio over a certain amount (30 for pulsed and 15 for continuous US, entries 8 and 11 in Tab. 3.4, respectively), seems to have a negative effect on the acidity concentration. In fact, the risk of the extraction of acid groups by the excess of water increases as well and the US power density decreases. 3.3 Sonochemically-assisted esterification and transesterification Esterification In Tab. 3.5 a list of the sonochemically-assisted esterification experiments is displayed together with the final acidities achieved after 4 hours of reaction. The reactor used for these experiments, provided with both an US horn (20 kHz) and a MW emitter (2450 MHz) is described elsewhere in detail (Ragaini et al., 2012). Standard calorimetric measurements were carried out to measure the actual emitted power (Suslick and Lorimer, 1989). Considering entries from 1 to 6 (rapeseed oil with high acidity), a final acidity lower than 0.5%wt is achieved within 4 hours operating at the conventional temperature of 336 K with all the methods, while this does not happen operating at lower temperatures. In particular, the lowest acidity is achieved at 336 K with MW. Considering entries from 7 to 12, inherent to the raw tobacco oilseed, final acidities lower than 0.5%wt are achieved only with the use of US. It is remarkable that at the temperature of 293 K the FFA esterification reaction rate results 6X faster than the conventional process at the same temperature. In the case of the rapeseed oil with low acidity (entries from 13 to 20), the use of MW increases the FFA conversion at 293 K and 313 K but not at 336 K. Moreover, the higher the applied power, the higher the FFA conversion. Oil Initial acidity (%wt) Cat. Technique Temp. (K) Emitted power (W) Tthermostat (K) Final acidity (%wt), 4 hr 1 Rapeseed oil (5)* 4.2-5.0 A46 conventional 313 - 315 1.18 2 336 338 0.50 3 ultrasound 313 38.5 293 0.55 4 336 313 0.48 5 microwaves 313 61.4 293 0.69 6 336 313 0.32 7 Tobacco 1.17 A46 conventional 293 - 293 0.97 8 313 315 0.55 9 336 338 0.45 10 ultrasound 293 38.5 277 0.48 11 313 293 0.46 12 336 313 0.30 13 Rapeseed oil (2)* 2.0-2.3 D5081 conventional 293 - 277 0.82 14 313 315 0.44 15 336 338 0.25 16 microwaves 293 31.7 277 0.73 17 313 31.7 293 0.34 18 61.4 293 0.37 19 336 31.7 313 0.29 20 61.4 313 0.25 Tab. 3.5. Sonochemically-assisted esterification experiments. The positive effects of acoustic-cavitation in liquid-solid systems are ascribable to the asymmetric collapse of the bubbles in the vicinity of the solid surface. When a cavitation bubble collapses violently near a solid surface, liquid jets are produced and high-speed jets of liquid are driven into the surface of a particle. These jets and shock waves improve both the liquid–solid and liquid-liquid mass transfer (Mason and Lorimer, 1988). MW is considered as a non-conventional heating system: when MW pass through a material with a dipole moment, the molecules composing the material try to align with the electric field (Mingos et al., 1997). Polar molecules have stronger interactions with the electric field. Polar ends of the molecules tend in fact to align themselves and oscillate in step with the oscillating electric field. Collisions and friction between the moving molecules results in heating (Toukoniitty et al., 2005). The increase of the FFA conversion as the power increases may be attributed to the fact that more power is delivered to the system and, therefore, the enhanced temperature effects caused by electromagnetic irradiation are increased with respect to lower powers. Differently the reason why a too high power was detrimental at the temperature of 336 K could be accounted for by two factors: i) the acoustic cavitation is enhanced at lower temperatures due to the higher amount of gas dissolved; ii) possible generation of too high temperatures inside the reaction medium that could have caused the removal of methanol from the system through constant evaporation or pyrolysis. Transesterification Transesterification experiments were performed on rapeseed oil both in batch and continuous mode. For the batch experiments two kinds of reactors were used: a traditional reaction vessel and a Rosett cell reactor, both with two ultrasound horns with different tip diameters (13 and 20 mm), and operating powers. A Rosett cell is a reactor designed to promote hydrodynamic cavitation through its typical loops placed at the bottom of vessel. Sonicators used in this work were provided by Synetude Company (Chambery, France). In Fig. 3.6, results from the conventional and the US-assisted batch experiments are compared. The US methods allows to attain very high yields in much shorter times than the traditional method and using less reagents (see Tab. 2.3) in just one step. The beneficial effects given by the US are attributable to the generation of acoustic cavitation inside the reaction medium leading to the phenomena already described in the case of esterification reaction. In particular, with the use of the Rosett cell reactor, BD yields of 96.5% (dotted lined) are achieved after 10 minutes of reaction. This is likely due to the combined approach exploiting acoustic cavitation along with hydrodynamic cavitation, which is able to provide a very efficient mixing inside the system. The use of the Rosett cell reactor provided transesterification reaction rates up to 15X faster than the conventional process. Continuous experiments were performed using two tubular reactors with different volumes (0.070 L at 35 KHz and 0.700 L at 20 kHz) and different US powers (19.3 and 68.3 W, respectively). The volume of the treated reagents was varied to obtain the same power density in both the reactors. Results are presented in Fig. 3.7. BD yields higher than 96.5% were obtained in the case of the small reactor within a reaction time of ~5 minutes. It is remarkable that BD yields higher than 90% were obtained using pulsed US (2 seconds on, 2 seconds off) after only 18 seconds, corresponding to just one passage in the reactor. In this case the transesterification reaction rate was 300X faster than the conventional process. The beneficial effects of pulses for the reactivity of the transesterification have been extensively reported (Chand et al., 2010; Kumar et al., 2010). In particular, as reported by Chand, when pulses are adopted, excessive heating of the reaction medium is not promoted, so preventing the loss of the gases dissolved in the system that are necessary for the acoustic cavitation to occur. Moreover, excessive heating during the transesterification reaction might lead to evaporation followed by pyrolysis of methanol and its subsequent removal from the reaction environment. 4. Conclusions As a conclusion to this work, some final remarks can be claimed: Feedstocks with a high potential for biodiesel (BD) production are Brassica juncea oilseed, which can be used as feedstock for BD100, Carthamus tinctorus, tobacco, animal fat and waste cooking oil to be used in BD blends with other oils or in diesel blends. However, blending different oils among them or with diesel already during the free fatty acids (FFA) esterification reaction may increase the reaction rate due to the lowered viscosity. Free fatty acids esterification over acid ion exchange resins in slurry reactors remains the preferred method of oils deacidification due to the optimal contact between the reagents and the catalyst and the good durability over time. The final high BD yields obtained for the oils de-acidified with the pre-esterification method over sulphonic ion exchange resins demonstrate its effectiveness in lowering the acidity and the possibility of obtaining high quality biodiesel from the selected feedstocks. Surface acidity and specific surface area of sulphated inorganic systems can be increased by both adding TiO2 and using ultrasound (US) in precise experimental conditions to assist the sol-gel synthesis of the catalysts. Changing the experimental conditions of US during the sol-gel synthesis makes also possible to tune the properties of the catalysts. In spite of not satisfying FFA conversions were obtained, US-assisted sol-gel synthesis turns out to be an extremely interesting method to obtain catalysts with high acidity and surface area. Both US and microwaves (MW) enhanced the FFA esterification reaction rate at temperatures lower than the one used conventionally (336 K). The positive effects of US are attributable to the phenomena generated inside the reaction medium by the acoustic cavitation, while MW are able to generate temperature effects localized in the proximity of the catalyst surface and to increase MeOH-oil solubility. US-assisted transesterification reaction is much faster than conventional transesterification: BD yields higher than 96.5% were achieved in most of the cases within 10 minutes of reaction, whereas the conventional method requires 150 minutes, besides higher reagents amount and higher temperatures. In particular, BD yields higher than 90% were obtained using a continuous reactor and pulsed US within 18 seconds, corresponding to just one passage in the reactor. In this case the transesterification reaction rate resulted to be 300X faster than the conventional process. Suggestions for the continuations of the work concern the further study of the synthesis of sulphated inorganic systems such as SO42-/ZrO2 or SnO2 or TiO2 with US and MW. Future work should also be devoted to the optimization of the experimental variables related to the use of MW and US to promote both FFA esterification and transesterification reactions. References Barrett E.P., Joyner L.G., Halenda P.P., “The determination of pore volume and area distributions in porous substances. I. Computations from nitrogen isotherms”, J. Am. Chem. Soc. 1951, 73, 373. Bianchi C.L., Boffito D.C., Pirola C., Ragaini V., “Low temperature de-acidification process of animal fat as a pre-step to biodiesel production”, Catal. Lett., 2010, 134, 179. Bianchi C.L., Pirola C., Boffito D.C., Di Fronzo A., Carvoli G., Barnabè D., A. Rispoli, R. Bucchi, “Non edible oils: raw materials for sustainable biodiesel”, in Stoytcheva M., Montero G. (Eds.): Biodiesel Feedstocks and Processing Technologies, Intech, 2011, pp. 3-22. Boffito D.C., Pirola C., Galli F., Di Michele A., Bianchi C.L., “Free Fatty Acids Esterification of Waste Cooking Oil and its mixtures with Rapeseed Oil and Diesel”, Fuel, 2012a, accepted on 19th October 2012, DOI: 10.1016/j.fuel.2012.10.069. Boffito D.C., Crocellà V., Pirola C., Neppolian B., Cerrato G., Ashokkumar M., Bianchi C.L., “Ultrasonic enhancement of the acidity, surface area and free fatty acids esterification catalytic activity of sulphated ZrO2-TiO2 systems”, J. Catal., 2012b, http://dx.doi.org/10.1016/j.jcat.2012.09.013 Boffito D.C., Pirola C., Bianchi C.L., “Heterogeneous catalysis for free fatty acids esterification rea.ction as a first step towards biodiesel production”, Chem, Today, 2012c, 30, 14. Brunauer S., Hemmett P., Teller E., “Adsorption of Gases in Multimolecular Layers”, J. Am. Chem. Soc. 1938, 60, 309. López D. E., Suwannakarn K., Bruce D. A., Goodwin JG. “Esterification and transesterification on tungstated zirconia: Effect of calcination temperature”, J Catal 2007, 247, 43. Mason T.J., Lorimer J.P., “Sonochemistry, Theory, Applications and Uses of Ultrasound in Chemistry“, Efford, J. Wiley, New York, 1988. Mingos D.M.P.,Baghurst D.R., “Applications of Microwave Dielectric Heating Effects to Synthetic Problems in Chemistry“, Microwave-Enhanced Chemistry, American Chemical Society,Washington, DC, USA, 1997. Perego C., Ricci, M., “Diesel fuel from biomass”, Catal. Sci. Technol., 2012, 1, 1776. Pirola C., Boffito D.C., Carvoli G., Di Fronzo A., Ragaini V., Bianchi C.L., “Soybean oil deacidification as a first step towards biodiesel production”, in D. Krezhova (Ed.): Recent Trends for Enhancing the Diversity and Quality of Soybean Products, Intech, 2011, pp. 321-44. Pirola C., Bianchi C.L., Boffito D.C., Carvoli G., Ragaini V., “Vegetable oil deacidification by Amberlyst : study of catalyst lifetime and a suitable reactor configuration”, Ind. Eng. Chem. Res., 2010, 49, 4601. Ragaini V., Pirola C., Borrelli S., Ferrari C., Longo I., “Simultaneous ultrasound and microwave new reactor: Detailed description and energetic considerations”, Ultrasonics Sonochemistry 2012, 19, 872 Sehgal C., Steer R.P., Sutherland R.G., Verrall R.E., “Sonoluminescence of argon saturated alkali metal salt solutions as a probe of acoustic cavitation”, J. Chem. Phys., 1979, 70, 2242. Suslick K. S., Doktycz, S. J., "The Effects of Ultrasound on Solids" in Mason, T.J.: Advances in Sonochemistry, JAI Press: New York, 1990, vol.1, pp. 197-230. Toukoniitty B., Mikkola J.P., Murzin D.Yu., Salmi T., “Utilization of electromagnetic and acoustic irradiation in enhancing heterogeneous catalytic reactions”, Appl. Catal. A 2005, 279, 1 Winayanuwattikun P., Kaewpiboon C., Piriyakananon K., Tantong S., Thakernkarnkit W., Chulalaksananukul W. et al. “Potential plant oil feedstock for lipase-catalyzed biodiesel production in Thailand”, Biomass. and Bioen. 2008, 32, 1279.
BIODIESEL PRODUCTION FROM NON-FOODSTUFF: CHEMISTRY, CATALYSIS AND ENGINEERING / D.c. Boffito ; tutor: C. Bianchi ; co-tutor: C. Pirola ; coordinator: D. Roberto. - Milano : Università degli studi di Milano. UNIVERSITA' DEGLI STUDI DI MILANO, 2013 Jan 11. ((25. ciclo, Anno Accademico 2012.
|Titolo:||BIODIESEL PRODUCTION FROM NON-FOODSTUFF: CHEMISTRY, CATALYSIS AND ENGINEERING|
|Supervisori e coordinatori interni:||ROBERTO, DOMINIQUE MARIE|
|Data di pubblicazione:||11-gen-2013|
|Parole Chiave:||biodiesel ; heterogeneous catalysis ; sonochemistry ; ultrasound ; microwaves ; ion exchange resins ; sulphated catalysts ; transesterification ; free fatty acids ; esterification ; vegetable oils|
|Settore Scientifico Disciplinare:||Settore CHIM/04 - Chimica Industriale|
Settore ING-IND/25 - Impianti Chimici
|Citazione:||BIODIESEL PRODUCTION FROM NON-FOODSTUFF: CHEMISTRY, CATALYSIS AND ENGINEERING / D.c. Boffito ; tutor: C. Bianchi ; co-tutor: C. Pirola ; coordinator: D. Roberto. - Milano : Università degli studi di Milano. UNIVERSITA' DEGLI STUDI DI MILANO, 2013 Jan 11. ((25. ciclo, Anno Accademico 2012.|
|Digital Object Identifier (DOI):||http://dx.doi.org/10.13130/boffito-daria-camilla_phd2013-01-11|
|Appare nelle tipologie:||Tesi di dottorato|