2.1. Characterization of the Catalysts Obtained
As can be observed in
Figure 1a, the X-ray diffractogram shows a topography characteristic of a one-dimensional hexagonal structure, of Al-MCM-41 type, as proposed by Beck and colleagues [
25].
The presence of four reflections, which are characteristic to plane diffraction, were identified. Their indices were [(100), (110), (200) and (210)]. According to the literature, these reflections are characteristic of the hexagonal arrangement of the material. The concept of crystallinity cannot be employed since its walls are amorphous silica. The absence of peaks at higher angles indicates that the material is not crystalline; however, it is known that there is an ordered hexagonal network, where the pore is surrounded by the other six, generating the reflections characteristic to MCM-41, which can also be observed by the high intensity of the first reflection (100) [
21,
25,
29,
30].
From the X-ray diffractogram illustrated in
Figure 1b, which represents the HZSM5 sample, an MFI crystalline structure can be observed with five characteristic reflections referring to the Miller indices [(101), (200), (501), (151) and (133)], of high intensity, similar to the standard proposed by IZA, and of satisfactory crystallinity. These confirm that the structure pertinent to ZSM-5 zeolite was not altered due to the thermal treatments conducted. The similarity between the crystalline structures of the HZSM5 zeolites, synthesized in the absence of organic template, with the standard ZSM-5 demonstrates the efficiency of the synthesis methods applied. These are also in accordance with the diffractograms presented in the literature [
26,
27,
31,
32].
This analysis was performed only on the ALMCM sample, since its route of synthesis has a structural template in the synthesis gel. This is contrary to what was proposed for the synthesis of the HZSM5 material.
The thermogravimetric curves (TG/DTG) of the synthesized mesoporous material are shown in
Figure 2. According to data found in the literature [
21,
33], for materials of the ALMCM type in their non-calcined form, two main stages of mass loss are evidenced. The first stage corresponds to the desorption of physically adsorbed water and the second stage to the decomposition of the organic material (surfactant) from the pores of the material. Following these is a less significant mass loss due to the condensation of the silanol groups.
As shown in
Table 1, the first mass loss occurred in the temperature range of 50 to 127 °C. This is possibly attributed to the moisture the material was exposed to before the thermogravimetric analysis. Furthermore, since physically adsorbed water is inherent to silica-based materials, the material is very susceptible to humidity.
The temperature range of 128 to 350 °C, which corresponds to the second mass loss, is characteristic of the removal of the CTMA+ ions that are poorly bound to the surface of the material or located on the external walls of the silica.
In the temperature range of 351 to 653 °C, the mass loss is only perceptible observing the derivative of the thermogravimetric curve. This variation of mass loss may be related to the aluminum impregnation in the ALMCM structure, which might interfere with the condensation of the silanol groups (Si-O-Si and Al-O-Si). It is important to point out that the bonds between silicon, oxygen and aluminum occur with different interactions [
21].
In
Table 1 and in the thermogravimetry curves shown in
Figure 2b, it can be seen that the calcined ALMCM material, at the temperature of 450 °C, presents a single mass loss event in the range of 34 to 130 °C, with 4.88% loss. This is attributed to the water absorbed before the thermogravimetric analysis, since the mesoporous material is silica-based. The mesoporous material was free from an organic template [
21,
34].
Similarities were observed in the spectra for the calcined and non-calcined forms (
Figure 3 and
Table 2). This is due to the absorptions of Si-O-Si and Al-O-Si bonds occurring in the same spectral region, between 1300 and 1000 cm
−1 [
35]. This technique is particularly important because it allows for the identification, through the absorption bands, of the organic functional groups present in the template’s structure, CTMA
+, contained in the ALMCM channels in the non-calcined form. The disappearance of the band related to this functional group, for the calcined sample, shows that the organic material was successfully removed from the pores of the mesoporous molecular sieve [
21].
In the ALMCM spectra, a very wide band is observed in the region of 3467 cm−1. This belongs to possible silanol groups, as well as water adsorbed on the surface of the material. This calcined sample shows no absorption band in the 2931 cm−1 region. This is present in the spectra of the uncalcined sample and in the template, being attributed to the stretching between the C-H of the CH2 and CH3 groups related to the surfactant (CTMA+) molecules. The absence of this band allows for affirming that the organic material present in the pores was removed after the calcination step.
In the spectra of the uncalcined and calcined ALMCM samples, typical bands corresponding to the asymmetric stretching of the Si-O bond are observed between 1245–1047 cm
−1, as well as to the stretching of the T-O bonds (T = Si or Al) in the region of 802 cm
−1, which are very common in silicates and aluminosilicates [
21].
In both
Figure 4 and
Table 3, the spectrum of the HZSM5 zeolite can be observed. A wide and intense band is located at approximately 3435 cm
−1; this, according to Silverstein and colleagues [
36], refers to the axial deformation vibrations of the OH in the SiOH group, which occur in the region of 3700 to 3200 cm
−1. The band observed in the region of 1641 cm
−1 corresponds to the deformation vibrations of water molecules. The bands at 1224 and 1066 cm
−1 represent the external and internal asymmetric stretching of the Si-O-Si siloxane groups. The band in the region of 788 cm
−1 represents the symmetric stretching of the Si-O-Si siloxane groups [
25,
29,
36].
The internal flexions of the tetrahedrons correspond to the band seen at 457 cm
−1, and the band referring to the length 545 cm
−1 defines the presence of five-membered double rings pertaining to the three-dimensional porous structure of ZSM-5, as reported in the literature [
26,
37,
38].
As can be observed in
Table 3, the absence of bands characteristic of organic molds demonstrates the efficiency of the synthesis in the absence of a template, indicating that its MFI structure was possibly reached due to the action of the Na
+ ions as a structural template. This assumption has not been fully substantiated yet.
Figure 5 shows results of the nitrogen adsorption/desorption isotherms consistent with those presented by the X-ray diffractograms. The ALMCM sample presents a type IV isotherm as shown in
Figure 5a, and the occurrence of capillary condensation at relative pressures between 0.4 and 0.98 can be observed [
21,
25]. Because of the different saturation pressures for condensation (adsorption) and evaporation (desorption), a hysteresis is observed in the same figure, causing the adsorption and desorption isotherms to not coincide.
Also corroborating with the results of the XRD, the HZSM5 material in its protonated form showed a type I isotherm, which is characteristic of microporous materials, as shown in
Figure 5b. However, since the limiting factor for monolayer formation in this material is strongly bound to the micropore volume in the region of 0.43 to 0.94 P/P
0, the multilayer physical adsorption process associated with capillary condensation on the surface of the adsorbate is observed [
39]. The structural properties obtained from the adsorption of N
2 by the BET method are shown in
Table 4.
The similarities between the values of the physical properties of the samples prove the efficiency of the synthesis methods for both the ALMCM and the HZSM5 [
21,
26,
40,
41].
The acidity of the ALMCM and HZSM5 samples was determined by the thermal desorption method (n-butylamine), in order to quantify the density and strength of the acid sites. Based on the amount of n-butylamine mols desorbed per gram of material, the density of the acid sites is quantified and the correlation to the acid strength of the sites is done based on the desorption temperature range. After the adsorption of the base in a continuous-flow fixed bed microreactor, the desorption of n-butylamine was carried out in a thermobalance. The TG/DTG curves of the ALMCM and HZSM5 samples were obtained in the range of 30 to 900 °C, as shown in
Figure 6 for the mesoporous and microporous materials.
For both the ALMCM and HZSM5 samples, the TG/DTG curves showed three mass losses. The mass loss up to 100 °C corresponds to the output of undissolved n-butylamine. It is important to note that the initial temperature above 100 °C exceeds the boiling point (BP) of n-butylamine (77 °C), therefore avoiding physically adsorbed molecules [
29,
30,
31]. The loss of mass relative to n-butylamine desorbed by the material is related to the acid sites. The acidity was measured considering that each mol of n-butylamine adsorbs on one mol of acid sites. The first step in the temperature range observed from 30–100 °C was attributed to physisorption and the loss of n-butylamine adsorbed onto the weak acid sites. The second step in the range 100–300 °C (medium acid sites) was attributed to desorption of the-butylamine chemisorbed and third steps considered in the ranges 300–510 °C (strong acid sites) were attributed. Based on the TG/DTG curves, the acid properties of the active sites of the samples were calculated, as shown in
Table 5.
Table 5 shows the concentrations of total acid sites in mmol of base per grams of catalyst. This demonstrates that the HZSM5 zeolite has a higher value than the one of the amorphous ALMCM material. The values of total acid sites are referenced in terms of NH
3 and not of n-butylamine since ammonia is one of the species resulting from the desorption of n-butylamine at the strong acid sites.
2.2. Catalytic Test
The catalytic tests were carried out with pure catalysts, HZSM5 zeolite and mesoporous ALMCM, as well as with mechanical mixtures between these materials in the proportions of 75% of HZSM5 and 25% of ALMCM; 50% of HZSM5 and 50% of ALMCM; 25% of HZSM5 and 75% of ALMCM.
Figure 7 shows the pyrograms of the products obtained in the thermal and thermo-catalytic pyrolysis of sunflower oil.
Bio-oil obtained from the thermal pyrolysis of sunflower oil is a complex mixture of organic compounds with carbon atom number from C15 to C37 (
n-alkanes e branched alkenes) and oxygenated products such as aldehydes, ketones, carboxylic acids, alcohols and esters. The most abundant chemical substances are as follows: diethyl phthalate (ester), heptadecane (
n-C17), heneicosane (
n-C21), n-Nonadecanol-1(alcohols), 9-Octadecenamide, (Z)-, tetratriacontane (
n-C34), squalene (C30, branched alkenes) and hexatriacontane (
n-C37). As the pyrolysis temperature used in this study is low (500 °C), the formation of carbon dioxide, carbon monoxide and aromatic, aliphatic and cyclic hydrocarbons was not observed. Generally, triglyceride pyrolysis mechanisms are difficult due to their complex chemical structures [
42].
Figure 8 [
42] shows the decomposition of a triglyceride reaction.
The data indicate a difference in the composition of the products obtained in the thermal cracking and in the thermo-catalysts. This is contrary to the idea that the presence of catalysts does not significantly alter the composition of the products of a vegetable oil pyrolysis, as reported in the literature [
21,
24]. On the other hand, it shows that the distribution of the products is modified by the action of the catalysts.
The intensity of the peaks related to the products with higher molar mass (greater retention time) in the pyrograms from thermo-catalytic processes is higher when compared to the same peaks in pyrograms from the thermal process.
Mesoporous materials are known for their ability to break down various high molecular weight compounds [
43]. Microporous zeolites such as HY, HBeta and HZSM-5 have been studied for the cracking of several compounds for many decades because of their attributes, such as high acidity and good thermal stability [
26,
44]. However, due to their small pore openings, bulky molecules, such as triglycerides, face a more difficult time to enter, adding a disadvantage to their use. This led to the development of catalysts with larger pore openings, which facilitate the entry of larger molecules. These, however, lack the thermal stability suitable for catalysis because of their low acidic nature [
45]. The combination of these micro and mesoporous catalytic materials results in the formation of composites that exhibit the advantages of both. The micro and mesoporous reorganization of these materials aimed at evaluating their catalytic activity based on the variety and selectivity of the products obtained through the thermo-catalytic pyrolysis reaction.
The pyrolysis of sunflower oil in the presence of catalyst (AlMCM-41 and HZSM-5) presents as a product the carbon monoxide, which provides evidence of the decarbonylation reaction. The decarboxylation and decarbonylation reaction is important for the reduction of the oxygenates product. Equations (1) and (2) show decarboxylation and decarbonylation reaction [
42]:
The products of pyrolysis of oil sunflower in the presence of the AlMCM catalyst have products with carbon atom number from C6 to C23, HZSM5 from C2 to C19 and the blend (AlMCM and HZSM5) from C3 to C26 and oxygenated products. The presence of the catalyst AlMCM directs the product to linear saturated and unsaturated hydrocarbons such as n-alkane (pentane, 2, 3-dimethyl- and heptadecane) e n-alkenes (1-hepteno, 2-octene, (E)-, 7-tetradecene and 5-eicosyne), cyclic hydrocarbons (cyclooctene, cyclopropane, 1-hexyl-2-propyl-, cis-and nonylcyclohexane) and oxygenated compounds (2-propenal, 1-Nonanol, octanoic acid and 6-octadecenoic acid). However, the catalyst HZSM5 directs the product to aromatic (benzene, toluene, o-xylene, ethylbezene and naphthalene), aliphatic (1-Hexene and 3-heptene), cyclic hydrocarbons (cyclopropane, 1-ethyl-2-propyl-, cis-and cyclohexene, 3-pentyl) and different oxygenated compounds such as fatty acids (hexanoic acid, heptanoic acid, n-decanoic acid and undecanoic acid), ketone (2-cyclohexen-1-one,4,4,5-trimethyl-), and alcohol (n-pentadecanol and 1-heptacosanol).
The analyses show that hydrocarbons make up for most of the products formed and the percentages of oxygenates decreased with the use of the catalysts. As can be seen in
Table 6, the HZSM5 catalyzed pyrolysis showed a better response to the reduction of oxygenates, a result attributed to the Brönsted and Lewis acid sites found in the zeolite, which were effective for the deoxygenation of the oil under the pre-conditions established. In accordance with this aspect, the physical mixture with a greater percentage of HZSM5 (75HZSM5-25ALMCM) presented the same behavior in terms of deoxygenation. As the percentage of HZSM5 in the catalyst samples decreases, so does the deoxygenation potential. The deoxygenation in the sample catalyzed by ALMCM, as well as in the sample with a higher percentage of ALMCM (25HZSM5-75ALMCM) was not as satisfactory, due to the lower acidity of the material, as described previously. Nevertheless, all the samples from the thermo-catalytic processes had a lower oxygen content than those in the thermal process.
Table 6 shows that the selectivity of hydrocarbons is directly related to the acid sites and to the pore diameter of the catalyst used. When assessing the thermal process, it can be seen that 87% of the products obtained are of longer chains (above 16 carbons), therefore being unsuitable for being used as fuel.
In the catalytic processes, the catalysts proved to offer better selectivity caused by the mesoporous pores. The ALMCM catalyzed pyrolysis resulted in satisfactory percentages of products with added-value for the petrochemical industry, including higher percentages in the fraction of kerosene (C11-C16).
The sample catalyzed by HZSM5 presented selectivity to light hydrocarbons in the gas range, and a posterior reduction in the range above C16, increasing production in the gasoline range (C5–C10).
When it comes to the samples from the physical mixtures, the products formed are similar to those obtained in the processes catalyzed by the major catalyst of the blend. Considering that in the OILSUN/25HZSM5-75ALMCM sample the ALMCM catalyst is in the majority, and is of mesoporous nature, the products formed are C5-C16 hydrocarbons. The OILSUN/75HZSM5-25ALMCM sample, on the other hand, resulted in products in the range of light hydrocarbons, with selectivity for products in the gasoline range.
The sample catalyzed by the OILSUN/50HZSM5-50ALMCM blend showed a balance between the products formed in the gasoline and diesel ranges, proving that the two catalysts acted together, providing a further reduction of the fractions above C16.
The good results attributed to the thermo-catalytic pyrolyzes are due to the ease of diffusion of the compounds originating from the primary cracking of sunflower oil, particularly the fatty acids, by the micro and mesoporous channels of the catalysts used. Such diffusion attenuates the fragmentation of the carbon chains of these compounds, thus making it possible to obtain hydrocarbons with a higher number of carbon atoms [
46].
Another factor, which greatly influences the composition of the products, is the composition of the raw material, in this case vegetable oil. Oils and fats with a high amount of unsaturated fatty acids and with relatively small carbon chains favor the gasoline fraction with high aromatic content [
21,
47]. The unsaturation facilitates the carbon chain cracking process and the reactions of cyclization and subsequent aromatization [
21,
47]. Contrastively, those with a high content of saturated fatty acids and with large carbon chains favor diesel fractions with low aromatic content.
Considering that the interest of this work was in obtaining biofuel with characteristics similar to those of the petroleum-derived gasoline, kerosene and diesel, it was observed that the thermo-catalytic samples from the physical mixtures 50HZSM5-50ALMCM and 25HZSM5-75ALMCM present the most adequate results to the gasoline, kerosene, diesel range, when compared to the thermocatalytic sample, ALMCM, HZSM5, 75HZSM5-25ALMCM, and the thermal pyrolysis. This indicates that the acid sites found in the HZSM5 catalyst and the pore size of the ALMCM catalyst were highly efficient for the oil cracking.