The present invention relates to a method for producing esters from vegetable oils and animal fats by using heterogeneous catalysts, particularly for obtaining biodiesel. Background Art

Biodiesel, used as fuel in Diesel engines, is constituted by a mixture of esters of fatty acids obtained by a reaction of transesterification of vegetable oils and animal fats with methanol and subsequent separation from glycerin.

The transesterification reaction for producing biodiesel is generally performed by using bases of alkaline metals, such as for example NaOH, KOH, NaOCH ₃ and KOCH ₃ [1,2], as catalysts. Currently, the cost of biodiesel is higher than the cost of petroleum- derived diesel fuel and therefore an improvement of the process which leads to lower costs of biodiesel is certainly of interest.

S. Gryglewicz [3] has performed transesterification tests with MgO,

Ca(OH) ₂ , CaO, Ca(CH ₃ O) ₂ , Ba(OH) ₂ , NaOH and has found the following scale of activity NaOH > Ba(OH) ₂ > Ca(CH ₃ O) ₂ > CaO, while magnesium oxide and calcium hydroxide have shown no activity at the boiling point of methanol.

However, barium hydroxide dissolves, producing an essentially homogeneous catalyst, whereas calcium oxide produces fine suspensions which make it difficult to separate the ester and glyceric phases [3].

Stern et al. [4] have proposed catalysts based on zinc oxide and zinc aluminate in the crystalline form of spinel, which have been found to be active at rather high temperatures (T 225°C) in the reaction for transesterification of the oils with methanol. However, these catalysts have been found to be sensitive to the

presence of water and show the best performance for water contents lower than 1000 ppm [5].

Leclercq et al. [6] have found that magnesium oxide and mixed magnesium aluminum oxides are active in the transesterification of oils at the reflux temperature of methanol, but the conditions used (methanol/oil molar ratio of 275-75) and the reaction time (22 hours) required in order to have high conversions do not allow industrial use of the proposed catalysts. Suppes et al. [7] have found that catalysts based on zeolite NaX and ETS-10 are active in the transesterification reaction at temperatures below 125°C. However, the activity of these catalysts is invalidated by the presence of free acidity.

Titanates have been used extensively as homogeneous catalysts in esterification and transesterification reactions due to their high selectivity [8,9]. Generally, titanates are active at temperatures above 180 ⁰ C. Saha et al. [10] have found that tetrabutyl titanate is a homogeneous catalyst for the transesterification of cyclohexyl acrylate with n-butanol and 2-ethylhexanol.

Wang et al. [11] have found instead that catalysts based on titanium oxide supported on silicon oxide and prepared by impregnating samples of SiO ₂ with solutions of tetrabutyl titanate in toluene are active in the transesterification of dimethyl oxalate with phenol to produce methyl phenyl oxalate and dimethyl phenyl oxalate. Disclosure of the Invention

The aim of the present invention is to provide a method for producing esters from vegetable oils and animal fats which allows simultaneously to produce glycerin with high purity.

Within this aim, an object of the present invention is to provide a method for obtaining esters from vegetable oils and animal fats which can be performed effectively even in the presence of water and/or free fatty acid. Further objects of the present invention will become better apparent

from the following detailed description of the invention.

This aim and these and other objects are achieved by the method defined in claim 1, which comprises the following steps:

- reaction between vegetable oils or animal fats and an aliphatic alcohol, in the presence of a catalyst which comprises titanium dioxide supported on silica,

- separation of unreacted alcohol, and

- separation of esters of fatty acids and glycerin obtained by means of said reaction. Ways of carrying out the Invention

Preferably, the aliphatic alcohol that is used is a monoaliphatic alcohol, particularly a C1-C5 aliphatic monoalcohol.

Advantageously, the catalyst used in the method according to the present invention contains titanium dioxide supported on amorphous silica. The catalysts used in the method according to the present invention can be obtained for example by wet grafting, grafting by vapor-phase deposition or impregnation of titanium dioxide precursors.

The grafting technique used, whether wet or by vapor-phase deposition, entails the reaction between a reactive compound of titanium and surface hydroxyls in the substrate or support constituted by a silicon oxide.

The reactive titanium compound used can be for example titanium alkoxide, chloride, or pentadienyl.

The impregnation technique for obtaining catalysts which can be used in the method according to the present invention entails the adsorption of a titanium oxide precursor in the pores of the substrate.

The titanium oxide precursor used can be for example an alkoxide, a chloride or a titanocene et cetera.

XPS measurements performed on the catalysts obtained with these techniques, after calcination at 500 ⁰ C, have shown the presence of a single

species, which is titanium oxide anchored to the silica [14].

The reaction that occurs in the method according to the present invention is a transesterification reaction optionally accompanied by an esterification reaction. The method according to the present invention has proved to be suitable to produce in particular biodiesel in suitable reaction conditions.

The method according to the present invention comprises in particular the steps of mixing vegetable oils or animal fats with an aliphatic alcohol, placing the mixture in contact with the solid catalyst, and then heating to the reaction temperature.

The inventors of the present invention have found that mixed catalysts which comprise titanium oxide supported on silicon oxide, obtained for example by grafting or impregnation technique, are active in the reaction conditions used, for example at a temperature of 150 and 250° C.

Moreover, the inventors of the present invention have found that the catalysts used in the method according to the present invention can also be used in the presence of a certain concentration of water (< 10.000 ppm).

In the method according to the present invention, after the reaction, the catalyst is separated by filtration, the unreacted aliphatic alcohol is separated by distillation, and the glyceric phase is separated by decantation from the ester phase.

The method according to the present invention can also comprise an additional step of transesterification of the unreacted glycerides present in the ester phase.

In the method according to the present invention, the reaction can be performed discontinuously or in continuous, agitated or fixed-bed reactors.

The catalysts used in the method according to the present invention are characterized by a surface area of 80 to 600 m ² /g, a porosity between 0.7 and 1.8 m ² /g, and a silica substrate pore size between 1 and 100 nm.

The following examples are suitable to illustrate the invention and must not be considered as limiting its scope.

All the reagents used were supplied by FLUKA, except for the soybean oil, supplied by Casa Olearia Italiana S. p. A. (Monopoli, province ofBari).

Examples

Example 1 Tests for transesterification of oil without catalyst

Since homogeneous reactions or reactions catalyzed by the steel walls of the reactor are possible at the temperatures being used [12], a reaction test was performed first of all by loading into a small agitated steel reactor 2 g of soybean oil and 0.9 g of methanol. The reactor was placed in a forced- ventilation oven and subjected to the following temperature program: 14 minutes at 50 ⁰ C, heating at 20°

C/min up to the set reaction temperature. The reactors were held at this temperature for 60 minutes. The reactor was then cooled rapidly down to ambient temperature. The resulting conversion was determined by using the H-NMR technique [13]. Tests at two temperatures, 180 and 200 ⁰ C, were conducted. Conversions of 7% and 8% were recorded at 180 and 200 ⁰ C, respectively.

Example 2

Tests for transesterification of acid oil at 180 ⁰ C without catalyst

Since the presence of free acidity also can have an effect on the transesterification reaction, a reaction test was conducted by loading into a small steel reactor 1.9 g of soybean oil, 0.1 g of stearic acid (acid oil 10% by weight) and 0.9 g of methanol.

The reactor was placed in a forced- ventilation oven and subjected to the following temperature program: 14 minutes at 50 ⁰ C, heating at 20° C/min up to 180 ⁰ C; the reactors were held at this temperature for 60 minutes. The reactor was then cooled rapidly down to ambient temperature. The resulting conversion was determined by using the H-NMR technique [14]. The resulting value of the conversion, 25%, highlights the effect of free acidity on the transesterification reaction. The final acidity was also measured and was found equal to 6.5% by weight.

Example 3

Synthesis of catalysts with increasing load of HO2 by wet grafting technique and transesterification test

The catalysts were prepared by following the method described by

Santacesaria et al. [14], in which a known quantity of titanium tetraisopropoxide (Ti[OiPr] ₄ , Aldrich 99.999%, d=0.963 g/ml) was dissolved in anhydrous dioxane and the solution was placed in contact with the substrate (SiO ₂ , Grace S432, specific area 320 mVg).

The substrate was calcined for eight hours at 500 ⁰ C, obtaining a silica with a specific area of 282 m ² /g, a pore volume of 1.02 cmVg (B.E.T. analysis) and a surface density of OH = 0.92 mmol/g (thermal analysis).

The grafting reaction was conducted in a glass reactor for 5 hours under constant agitation, and all the operations were performed in an inert environment. After reaction, the solid fraction was recovered by filtration, washed with dioxane and dried in a stove at 120 ⁰ C for an entire night. The resulting solid was finally subjected to a preliminary treatment at 200 ⁰ C for

2 hours and to calcination at 500 ⁰ C for 2 hours, in order to eliminate the residual alkoxide groups. According to this procedure, solids with an increasing load of titanium were prepared.

Table 1 lists the theoretical compositions of the various catalysts prepared and the actual content of titanium anchored to the substrate

determined by UV- Vis spectroscopy (colorimetric method, [14]).

A multilayer catalyst was also prepared by multiple grafting reaction of titanium tetraisopropoxide. If the density of surface hydroxyls (0.92 mmol/g) determined by thermogravimetric means is known, and assuming that each -OH surface group reacts with an alkoxide molecule, the amount of alkoxide to be used so as to obtain a single-layer coating of the silica with a 50% excess in mmol is used. The reaction is performed in a 300-ml jacketed glass reactor in reflux at the boiling point of toluene (115°C) for 6 hours under constant agitation. The product is recovered by filtration in vacuum, washed with toluene and kept in a stove for 12 hours. Hydrolysis with superheated steam is then performed at 150 ⁰ C for 2 hours in a jacketed glass column. The solid is finally calcined at 500 ⁰ C for 3 hours.

Two other operations for grafting, hydrolysis and calcination are performed on the resulting solid and are repeated in the same conditions as the first one.

Table 2 lists the actual content of titanium oxide anchored at each grafting step, determined by UV-Vis spectroscopy (colorimetric method, [14]).

For the various catalysts shown in Tables 1 and 2 and for the silica used as a substrate, the titanium dioxide in anatase form and in rutile form, supplied by Fluka, reaction tests were performed by loading into small steel reactors 2 g of soybean oil, 0.9 g of methanol and 0.1 g of catalyst in powder form.

Before use, the catalysts were kept at 200 ⁰ C for 2 hours. The reactors were placed in a forced-ventilation oven and subjected to the following temperature program: 14 minutes at 50 ⁰ C, heating at 20° C/min up to 180 ⁰ C; the reactors were held at this temperature for 60 minutes. The reactors were then cooled rapidly down to ambient temperature. The resulting conversions were determined by using H-NMR [13].

Table 3 lists the results obtained for the various tests.

As can be seen from the results given in the table, the silicon oxide and titanium oxide in crystalline anatase form exhibit scarce activity, while titanium oxide in crystalline rutile form shows no activity.

A maximum of activity is instead shown for coated silica catalysts for a load of TiO ₂ comprised between 1% and 20% by weight on the silica.

Table 2 - Exam le 3 - Load of TiO ₂ at each raftin ste

Example 4

Test for transesterification of acid oil at 180 ⁰ C with Tiθ2/Siθ2 (3) The catalyst in powder form TiO ₂ /SiO ₂ (3) at 3.09% by weight of

TiO ₂ on the silica, which yielded the best performance in the tests of Example 3, was tested in the presence of an oil acidified with stearic acid. Accordingly, a reaction test was conducted by loading into a small steel reactor 1.9 g of soybean oil, 0.1 g of stearic acid (5% acid oil by weight) and 0.9 g of methanol.

The reactor was placed in a forced- ventilation oven and subjected to the following temperature program: 14 minutes at 50 ⁰ C, heating at 20° C/min up to 180 ⁰ C; the reactors were held at this temperature for 60 minutes. The reactor was then cooled rapidly down to ambient temperature. The resulting conversion was determined by using the H-NMR technique [13]. The resulting conversion value, 46.33%, highlights that the free acidity has only a partial effect on the catalyst, reducing its activity by

approximately 27%. The final acidity of the oil after the reaction was also measured and was found to be equal to 3.7% by weight, indicating that there is also a catalytic effect on the conversion of the free acid. Indeed, in Example 2, where a test was conducted in the same conditions but without the catalyst, the concentration of residual acid reached 6.5%.

Example 5

Tests for transesterification of oil containing different concentrations of water at 180 ⁰ C (activity comparison) The catalyst in powder form TiO ₂ /SiO ₂ (3) at 3.09% by weight of

TiO ₂ on the silica which yielded the best performance in the tests of Example 3 was tested in the presence of different amounts of water.

Three reaction tests were conducted by loading into small steel reactors 2 g of soybean oil, 0.9 g of methanol, 0.1 g of catalyst, and respectively 5000 ppm, 1000 ppm and 500 ppm of water.

The reactors were placed in a forced-ventilation oven and subjected to the following temperature program: 14 minutes at 50 ⁰ C, heating at 20° C/min up to 180 ⁰ C; the reactors were kept at this temperature for 60 minutes. The reactors were then cooled rapidly down to ambient temperature.

The ester phase was analyzed by gas chromatography [15] for tests in the presence of water, while the test performed in the absence of water was analyzed by H-NMR spectroscopy [13].

The results of Table 4 show that the concentration of methyl esters is influenced by the presence of water: activity decreases as the amount of water increases. In particular, the catalyst is not affected by the water, if present up to a concentration of 500 ppm, but the concentration of methyl esters decreases by approximately 10% and is therefore scarcely affected by the water at least up to a concentration of 1000 ppm and decreases only by 20% if it is present at a concentration of 5000 ppm.

Table 4 - Exam le 5 - Tests at 180 ⁰ C in the resence of water

Example 6 Test for esterification of oil with Tiθ2/Siθ2 catalyst (S) at 3.09% in an autoclave at 225°C

A reaction test was conducted by loading into a 1 -liter agitated autoclave 250 g of soybean oil with 114 g of methanol, 6.25 g of catalyst in powder (TiO ₂ /SiO ₂ (3) at 3.09% of TiO ₂ on SiO ₂ ). The autoclave was heated to 225°C. After 159 minutes, a sample was taken and H-NMR analysis [13] yielded a conversion of 75.20%. As time progressed, the conversion increased up to 88.37% after 339 minutes.

The autoclave was then cooled to ambient temperature. The product unloaded from the autoclave was filtered. The methanol was distilled and the glyceric phase was separated from the ester phase by means of a separation funnel.

The ester phase was analyzed by H-NMR [13] and confirmed a conversion of 92.06%.

Example 7

Synthesis of a T1O2 catalyst supported on S1O2 pellets by means of an impregnation technique and catalytic test

The substrate used in this preparation is constituted by silica pellets (DEGUSSA AG, Aerolyst™ SiO ₂ 3038, specific area 270 m7g); the substrate was pretreated by keeping it at 200 ⁰ C in a stove for an entire night. A catalyst was prepared with a theoretical load of TiO ₂ equal to 3.1% by placing in contact, in an inert environment, a solution of titanium tetraisopropoxide in toluene with the silica pellets. The reaction was performed in a jacketed glass reactor, in which thermostatically-controlled oil at the boiling point of toluene (110 ⁰ C) was made to circulate. The reactor was also provided with a refrigeration column and the reaction occurred under reflux of the solvent. After 3 hours of reaction, the product was recovered by filtration and washed with toluene. Finally, the recovered solid is placed for one night in a stove at 120 ⁰ C and calcined first at 200 ⁰ C for 2 hours and then at 500 ⁰ C for 3.5 hours.

Two reaction tests were conducted by loading into small steel reactors 2 g of soybean oil, 0.9 g of methanol and 0.1 g of catalyst.

The reactors were placed in a ventilated oven and subjected to different temperature programs. In one case, the test was conducted at 180° C and therefore the temperature program was as follows: 14 minutes at 50° C, heating at 20°C/min to 180 ⁰ C; the reactor was kept at this temperature for 60 minutes. In the other case, the test was conducted at 200 ⁰ C and therefore the temperature program was as follows: 2 minutes at 50 ⁰ C, heating at 15°C/min up to 200 ⁰ C; the reactor was held at this temperature for 1.7 hours. At the end of the reaction, the reactors were then cooled rapidly down to ambient temperature.

The ester phase obtained in the two cases was analyzed by gas chromatography, obtaining the compositions listed in Table 5. As can be seen, the conversion to esters in the test at 180 ⁰ C is lower than the

conversion obtained with a catalyst in powder form having a similar titanium concentration (see Example 3). This effect is due mostly to the diffusion limitations that occur when using pellets rather than powder as a substrate. However, this limitation can be attenuated by increasing the temperature. If the temperature is increased from 180 ⁰ C to 200 ⁰ C, the contraction of the esters shifts from 35% to 71%.

Table 5 - Example 7 - Tests at 180 and 200 ⁰ C for the catalyst supported on silica pellets

The method according to the present invention allows a reduction in the cost of biodiesel, linked mostly to the intensification of the process and therefore to the possibility to work in continuous plants.

Another economic advantage of the method according to the present invention occurs by producing glycerin, the co-product of biodiesel, with a higher degree of purity, eliminating or reducing purification costs. This last advantage is due substantially to the absence of soaps, which due to their surfactant properties can also form stable emulsions which require onerous treatments for separation. The possibilities of improvement are linked to the use of a heterogeneous catalyst.

The heterogeneous catalyst used in the method according to the

present invention in fact limits the formation of surfactant species in solution (soaps), which by facilitating the formation of emulsions slow the step of separation of the glycerin and of the biodiesel, avoiding the need to neutralize the products, and also avoids the need for an operation to separate the glycerin from the residues of the homogeneous catalyst. Moreover, with traditional alkaline catalysts, the presence of humidity caused for example by non-anhydrous alcohol, increases all the mentioned negative phenomena. The disclosures in Italian Patent Application No. MI2005A000361 from which this application claims priority are incorporated herein by reference.

References

[1] US-4,164,506 (1979)

[2] US-5,730,029 (1998)

[3] S. Gryglewicz, Bioresource Technology 70 (1999) 249-253 [4] US-5,908,946 (1999)

[5] EP-1352893 Al (2003)

[6] E. Leclerq, A. Finiels, C. Moreau, JAOCS 78 (2001) 1161

[7] GJ. Suppes, M.A. Dasari, EJ. Doskocil, PJ. Mankidy, MJ. Goff, App.

Catalysis A: General 257 (2004) 213 [8] S. P. Wang, X.B. Ma, Z.H. Li, G.H. Xu, Nat. Gas Chem. Eng. (China) 27

(2001) 1

[9] X.B. Ma, H.L. Guo, S. P. Wang, Y.L. Sun, Fuel Process Technol. 83

(2003) 275

[10] B. Saha, M. Streatm, Reactive & Functional Polymers 40 (1999) 13-27 [11] S. Wang, X. Ma, H. Guo, J. Gong, X. Yang, G. Xu, Journal of

Molecular Catalysis A: Chemical 214 (2004) 273-279

[12] Dasari, M.A., Goff, MJ., Suppes, GJ. JAOCS 80 (2003) 189-192

[13] G. Gelbard, O. Bres, R.M. Vargas, F. Vielfaure, U.F. Schuchardt,

JAOCS, 1995, 72, 1239 [14] E. Santacesaria, M. Cozzolino, M. Di Serio, A.M. Venezia, R. Tesser,

Applied Catalysis A: General 270 (2004) 177-192

[15] UNI 10946:2001