Conversion of Stearic Acid into Bio-Gasoline over Pd / ZSM-5 Catalysts with Enhanced Accessibility

: Palladium supported on nanocrystalline ZSM-5 (n-ZSM-5, Si / Al = 32) and hierarchical ZSM-5 (h-ZSM-5) with di ﬀ erent acidity (Si / Al = 33, 51, 122) were tested in the liquid-phase conversion of stearic acid under nitrogen atmosphere (6 bar). The incorporation of Pd into ZSM-5 zeolite increased signiﬁcantly the share of gasoline in the reaction products due to the promotion by this metal of both decarboxylation and hydrogen transfer reactions. Likewise, the Pd nanoparticles dispersed over the zeolitic support favored the conversion of light oleﬁns formed by end-chain cracking reactions into gasoline-range hydrocarbons according to an oligomerization / cyclization / aromatization pathway. Additionally, Pd / h-ZSM-5 gave rise to higher conversion and selectivity towards gasoline than Pd / n-ZSM-5, due mainly to the enhanced accessibility and improved Pd dispersion achieved when using the hierarchical zeolite. The decrease in the Si / Al atomic ratio in Pd / h-ZSM-5 samples resulted in a rise in the stearic acid conversion, although it was lower than expected. This ﬁnding denotes that, for supports with high acid sites concentration, the Pd availability became the limiting factor as the metal was loaded in similar amounts in all catalysts. Finally, the increase of the reaction temperature with the Pd / h-ZSM-5 (122) catalyst augmented both stearic acid conversion and gasoline selectivity, since it enhanced the conversion of the light oleﬁns, formed as primary cracking products, into liquid hydrocarbons. Therefore, it can be concluded that Pd supported on hierarchical ZSM-5 zeolite is a convenient catalyst for obtaining bio-gasoline from oleaginous feedstock.


Introduction
Oleaginous biomass, consisting mainly of triglycerides and fatty acids, is the major component in different products obtained from plants, animals, and microorganisms. Due to their relatively low oxygen content, oleaginous biomass resources are viewed as promising raw materials for obtaining biofuels and/or bio-based chemicals. Accordingly, a number of routes have been developed and investigated for biofuels production from oleaginous biomass, such as transesterification, catalytic cracking, and hydrodeoxygenation.
Fatty acid methyl esters (FAMEs), generated by transesterification of triglycerides with methanol, are known under the conventional denomination of biodiesel and have already reached the commercial scale [1][2][3]. However, FAMEs present several drawbacks when used in internal combustion engines: high viscosity, low thermal and oxidation stability, and a trend to polymerize, limiting the content of biodiesel up to a maximum of 7 vol% [4,5]. These problems are provoked by the relatively large molecular weight of the triglyceride molecules, as well as by their oxygen content. Therefore, several alternative pathways have been devised to produce fuels from oleaginous biomass

Catalysts Preparation
The hierarchical ZSM-5 supports were synthesized by a protozeolitic units silanization method published in Reference [35], using the following reagents: tetraethylortosilicate (TEOS, 98%, Sigma-Aldrich, St. Louis, LA, USA), tetrapropylammonium hydroxide (TPAOH, Alfa, 40% v/v aqueous solution), aluminum isopropoxide (AIP, Sigma-Aldrich, St. Louis, LA, USA), phenylaminopropyltrimethoxysilane (Sigma-Aldrich, St. Louis, LA, USA), and distilled water. Initially, TPAOH and AIP were dissolved in distilled water and cooled in an ice bath under stirring for several hours until a clear solution was observed. Then, TEOS-previously cooled in ice-was poured dropwise into the solution and left mixing for 1 h. The composition of these mixtures was x Al 2 O 3 :60 SiO 2 :11 TPAOH:1500 H 2 O, where x was varied for obtaining Si/Al molar ratios in the gel of 30, 50, and 100, respectively. The solutions were stirred for 40 h at ambient temperature until TEOS was completely hydrolyzed and then the formed ethanol was evaporated in a rotary evaporator at 60 • C. The obtained gels were pre-crystallized at 90 • C for 20 h under reflux to form protozeolitic units. Crystallization was then interrupted by adding 8 mol%, (on a silicon basis) of the silanization agent, keeping the reflux at 90 • C for 6 h. The suspension of silanized units was finally crystallized at 170 • C in an autoclave reactor for 7 days. The obtained solids were removed from the synthesis medium by centrifugation and rinsed with distilled water. Subsequently, they were dried at 110 ºC overnight and calcined in static air at 550 • C for 5 h. The materials were called as h-ZSM-5 (X), wherein X stands for their actual Si/Al molar ratio.
Palladium was added on the supports by means of impregnation with a PdCl 2 (99.999%, Aldrich) aqueous solution. The Pd concentration in the solution was adjusted to attain 1 wt % metallic Pd on the final catalysts, employing a ratio (pore volume)/(solution volume ratio) of 1/2. After impregnation, the samples were put in an ultrasonic bath for 30 min and subsequently in a rotary evaporator under vacuum at ambient temperature for 5 h. Finally, they were dried at 110 • C for 12 h and then calcined in a muffle oven at 550 • C for 5 h under static air. Calcined samples were subjected to reduction under 30 mL min −1 of H 2 , the temperature increased to 450 • C with a ramp rate of 2 • C min −1 and held for 30 min. Subsequently, the samples were cooled down to ambient temperature under N 2 flow. Then, they were stored in a closed vessel under room conditions until being used in reaction.

Catalysts Characterization
The samples of the reduced catalysts-Pd as Pd 0 -were characterized using conventional techniques: X-ray diffraction, Inductively-coupled plasma atomic emission spectroscopy (ICP-AES), N 2 and Ar adsorption isotherms, NH 3 TPD (Ammonia temperature programme desorption), and Transmission electron microscopy (TEM). High-angle XRD patterns were taken with a Philips X'PERT MPD diffractometer (Philips, Eindhoven, The Netherlands) using Cu-Kα radiation in order to ascertain the crystallinity of the samples. X-ray diffraction patterns were measured within the 5-75 • range employing a step size of 0.1 • and a counting time of 10 s. Inductively-coupled plasma atomic emission spectroscopy (Varian, Palo Alto, CA, USA) was employed to determine the Pd content and Si/Al atomic ratio of the samples with a Varian VISTA-AX CCD spectrophotometer. Previously, the samples were digested in a H 2 SO 4 -HF mixture.
Nitrogen adsorption-desorption isotherms at 77 K were carried out with a Micromeritics TriStar 3000 instrument (Micromeritics, Norcross, GA, USA). The surface areas were calculated by the BET (Brunauer, Emmett, Teller) method using relative pressures in the range 0.05-0.15. Total porosities were determined as the total volume of pores at P/P 0 = 0.98. Argon adsorption isotherms at 87 K were performed with a Quantachrome AUTOSORB instrument (Quantachrome, Boynton Beach, FL, USA). Previously, the samples were outgassed at 300 • C for 3 h under vacuum (P < 10 −5 mmHg).
The cumulative pore volumes and pore size distributions were determined by the non-local density functional theory (NLDFT) model, assuming cylindrical pore geometry.
Ammonia TPD measurements were performed in a Micromeritics Autochem 2910 apparatus (Micromeritics, Norcross, GA, USA). Initially the samples were heated to 550 • C under a 50 Nml min −1 helium flow, holding this temperature for 30 min. Then, the samples were cooled to 180 • C, and subsequently, a 35 Nml min −1 ammonia flow was circulated for 30 min. Then, the ammonia flow was stopped and a He flow of 50 Nml min −1 was passed through the samples for 90 min. Then, the chemisorbed NH 3 was removed by increasing the temperature to 550 • C with a ramp rate of 15 • C min −1 . The ammonia concentration in the effluent gases was monitored by means of a thermal conductivity detector (TCD).
The TEM micrographs were taken in a Philips TECNAI 20 microscope (Philips, Eindhoven, The Netherlands) that operated with a LaB 6 filament under an accelerating voltage of 200 kV. Previously, the samples were prepared by dispersion in acetone, followed by stirring in an ultrasonic bath and final deposition over a carbon-coated copper grid. This technique allows determining the morphology of the support particles as well as the size of the Pd nanoparticles.

Catalytic Tests of Stearic Acid Conversion
The catalytic experiments were performed in a stainless-steel autoclave furnished with a stirrer and heated by an external oven. In a typical reaction, the corresponding amount of catalyst was loaded (0.4 g) together with 10 wt % stearic acid solution (2 g, 98% Aldrich) in n-dodecane (18 g, ≥99%, Sigma-Aldrich, St. Louis, LA, USA) under 6 bar of nitrogen. Then, the reaction was performed at the desired temperature for a fixed time with stirring. Once the reaction time was reached, the reactor was cooled down to 80 • C. Subsequently, an ice-cooled condenser separated the reaction products into gaseous and liquid fractions which were collected in a gas bag and a vial, respectively. Additionally, the liquid product present in the reactor was collected and separated from the catalyst by filtering under vacuum in a device previously heated to 110 • C.

Analysis of the Reaction Products
A Varian CP-4900 MicroGC (Varian, Palo Alto, CA, USA) equipped with two injection lines with columns Molsieve 5 Å PLOT and ParaPLOT U/Q, respectively, and TCDs was employed for the analyses of the gaseous phase. Both channels allowed detecting and measuring H 2 , CO, CO 2 , and hydrocarbons from C 1 to C 4 . The liquid products were measured in a Brucker SCIEN 456-GC furnished with a fused silica BR 5 column (0.25 mm internal diameter and 30 m length) and a flame ionization detector (FID). This Gas Cromatography (GC) enabled the separation of compounds in the C 4 -C 20 interval. Prior to the analyses, the samples were silylated using a method adapted from the EN-14105 standard. For each 0.86 g of sample, 30 µL of pyridine, 90 µL of N-methyl-N-(trimethylsilyl)-trifluoroacetamide (MSTFA), and 10 µL of 1,2,3,4-tetrahydronaphthalene as internal standard were mixed and strongly shaken. This mixture was held at ambient temperature for 15 min before analysis.

Catalysts Properties
This section discloses the main features of the supports and catalysts employed in this work. Two different types of ZSM-5 zeolites were used as support of the Pd nanoparticles: nanocrystalline and hierarchical ZSM-5 zeolites, named as n-ZSM-5 and h-ZSM-5, respectively. For the hierarchical zeolitic supports, several samples were prepared having different Si/Al atomic ratios (~33-122). Regarding the Pd content, all catalysts presented an amount of this metal slightly lower than the theoretically loaded (1 wt %), with some deviation being observed for the Pd/n-ZSM-5 (32) catalyst, with a Pd amount of 0.84 wt %, which could indicate a lower association of the Pd moieties with the nanocrystalline zeolitic support during the impregnation treatment. Figure 1 displays the high-angle XRD patterns of the reduced Pd catalysts. All the samples exhibited the typical XRD patterns of MFI zeolite, with no trace of amorphous material. Only in the case of the Pd/h-ZSM-5 (33) were less intense reflections appreciated, which suggests the occurrence of smaller crystalline domains in this sample. Despite the low amount of Pd loaded (around 1 wt %), the reflection of Pd metal at 2θ~40.1º (marked with an asterisk) may be observed in all the catalysts in a similar way, bearing out the presence of this metal.
Appl. Sci. 2019, 9, x FOR PEER REVIEW 5 of 16 5 with a Pd amount of 0.84 wt %, which could indicate a lower association of the Pd moieties with the nanocrystalline zeolitic support during the impregnation treatment. Figure 1 displays the high-angle XRD patterns of the reduced Pd catalysts. All the samples exhibited the typical XRD patterns of MFI zeolite, with no trace of amorphous material. Only in the case of the Pd/h-ZSM-5 (33) were less intense reflections appreciated, which suggests the occurrence of smaller crystalline domains in this sample. Despite the low amount of Pd loaded (around 1 wt %), the reflection of Pd metal at 2θ~40.1º (marked with an asterisk) may be observed in all the catalysts in a similar way, bearing out the presence of this metal.  In addition to the adsorption occurring at low relative pressures, arising from the zeolite micropores, the samples also exhibited significant Ar uptake and intermediate relative pressure, which was consistent with the respective nanocrystalline and hierarchical features of the zeolitic supports. In all cases, the incorporation of Pd caused some decreases in the adsorption capacity of the supports. This finding can be assigned to the occurrence of some micro-and mesopore blocking caused by the deposition of the Pd nanoparticles and/or to some contraction of the mesoporosity provoked by the thermal treatment of hydrogen reduction at high temperature (450 °C). On the other hand, almost no hysteresis was appreciated in the desorption branch of the isotherm indicating the absence of bottle-neck pores.  Figure 2 exhibits the Ar adsorption-desorption isotherms of both carriers and Pd-containing materials. In addition to the adsorption occurring at low relative pressures, arising from the zeolite micropores, the samples also exhibited significant Ar uptake and intermediate relative pressure, which was consistent with the respective nanocrystalline and hierarchical features of the zeolitic supports. In all cases, the incorporation of Pd caused some decreases in the adsorption capacity of the supports. This finding can be assigned to the occurrence of some micro-and mesopore blocking caused by the deposition of the Pd nanoparticles and/or to some contraction of the mesoporosity provoked by the thermal treatment of hydrogen reduction at high temperature (450 • C). On the other hand, almost no hysteresis was appreciated in the desorption branch of the isotherm indicating the absence of bottle-neck pores. Appl. Sci. 2019, 9, x FOR PEER REVIEW 6 of 16 6 Figure 2. Ar adsorption-desorption isotherms at 87 K of both the Pd/ZSM-5 catalysts and the corresponding supports.
As inferred from Table 1, the hierarchical Pd/h-ZSM-5 catalysts, irrespective of their Si/Al atomic ratio, showed enhanced textural properties with regard to Pd/n-ZSM-5 (32), with superior BET surface areas (~100 m 2 g -1 above) and pore volumes (>0.04-0.120 cm 3 g -1 ). On the contrary, the sample based on the nanocrystalline ZSM-5 support exhibited the highest micropore surface (274 m 2 g -1 ) and the lowest external/mesopore surface area (103 m 2 g -1 ). For the hierarchical Pd/h-ZSM-5 catalysts, an influence of the Si/Al atomic ratio over the micropore and external/mesopore surface area may be observed. In this regard, the micropore surface area decreased at lower Si/Al atomic ratios, while the contrary was appreciated for the external/mesopore surface area. The great proportion of external/mesopore surface areas in all the samples, and in particular for the catalysts prepared from hierarchical ZSM-5 supports, confirms their high accessibility for the reaction of large molecules, such as stearic acid. As inferred from Table 1, the hierarchical Pd/h-ZSM-5 catalysts, irrespective of their Si/Al atomic ratio, showed enhanced textural properties with regard to Pd/n-ZSM-5 (32), with superior BET surface areas (~100 m 2 g −1 above) and pore volumes (>0.04-0.120 cm 3 g −1 ). On the contrary, the sample based on the nanocrystalline ZSM-5 support exhibited the highest micropore surface (274 m 2 g −1 ) and the lowest external/mesopore surface area (103 m 2 g −1 ). For the hierarchical Pd/h-ZSM-5 catalysts, an influence of the Si/Al atomic ratio over the micropore and external/mesopore surface area may be observed. In this regard, the micropore surface area decreased at lower Si/Al atomic ratios, while the contrary was appreciated for the external/mesopore surface area. The great proportion of external/mesopore surface areas in all the samples, and in particular for the catalysts prepared from hierarchical ZSM-5 supports, confirms their high accessibility for the reaction of large molecules, such as stearic acid. The acid properties of the catalysts were calculated by means of ammonia TPD measurements and their results are summarized in Table 1. Pd/n-ZSM-5 (32) presents a higher content of strong acid sites than Pd/h-ZSM-5 (33). For the hierarchical Pd/h-ZSM-5 catalysts, when the Si/Al atomic ratios increased, there was a parallel abatement in the total acidities, while their strength was fairly similar, according to their respective maximum temperature of ammonia desorption. The acid properties of the Pd-containing catalysts were fairly similar to those of the original supports, which can be found elsewhere [36], with only slight variations in the overall acidity being observed. Figure 3 illustrates TEM micrographs of the Pd-containing catalysts. The Pd/n-ZSM-5 (32) was constituted by agglomerates of nanounits with sizes in the range 40-100 nm. The materials based on hierarchical zeolitic supports presented a sponge-like aspect, being formed by aggregates (200-400 nm) of very small zeolitic nanounits (5-10 nm). In all cases, the presence of Pd nanoparticles deposited over the external surface of the zeolite crystals or located within the aggregates was clearly observed. As inferred from Table 1, the average size of the Pd nanoparticles, determined from the TEM images, was lower for the catalysts prepared from the hierarchical support compared to that of the nanocrystalline ZSM-5, denoting that the mesoporosity present in the former led to a better dispersion of the metal, as it has been earlier reported [36]. Likewise, increasing the Si/Al ratio of the hierarchical zeolitic supports led to progressively larger Pd nanoparticles. Therefore, it can be concluded that the presence of Al in the hierarchical supports afforded a better Pd dispersion.

Effect of the Pd Incorporation
Initially, tests of stearic acid conversion were performed comparing the parent nanocrystalline n-ZSM-5 and Pd/n-ZSM-5 catalysts to ascertain the role played by the addition of the metal. As illustrated in Figure 4, almost complete conversion of stearic acid into hydrocarbons was attained over both catalysts, showing the high decarboxylation activity of these materials under the employed operation conditions (T = 325 °C, 6 bar of N2). Regarding the hydrocarbon distribution ( Figure 5), the main products obtained were gases and gasoline range compounds. The main components in the gases are C3 and C4 olefins, proceeding from end-chain cracking reactions of the carbocations species formed by stearic acid decarboxylation. On the other hand, two pathways can be envisaged to explain the formation of gasoline range hydrocarbons: i) random cracking reactions at any position in the chain of those carbocations, and ii) oligomerization/cyclization/aromatization (OCA pathway) of the light olefins formed by severe cracking reactions. Both end-chain cracking and OCA pathways are known to occur on the acid sites of the ZSM-5 zeolite [37][38][39]. Interestingly, the addition of Pd to the n-ZSM-5 support results in the production of less gaseous hydrocarbons and more gasoline. Moreover, the carbon atom distribution obtained over the Pd-containing catalyst exhibited a more regular profile, whereas the selectivity towards gasoline-range hydrocarbons increased from 65.5% over the n-ZSM-5 (32) sample up to 77.5% when using the Pd/n-ZSM-5 (32) catalyst. This positive finding can be assigned to a promotion effect of the OCA pathway provoked by the Pd species. In this way, Pd is known to present activity in the oligomerization and polymerization of olefins [40][41][42][43], which would be in agreement with its participation also in olefin oligomerization reactions. Moreover, Pd exhibits hydrogen transfer activity, which could favor the aromatization step in the OCA pathway [44][45][46][47].

Effect of the Pd Incorporation
Initially, tests of stearic acid conversion were performed comparing the parent nanocrystalline n-ZSM-5 and Pd/n-ZSM-5 catalysts to ascertain the role played by the addition of the metal. As illustrated in Figure 4, almost complete conversion of stearic acid into hydrocarbons was attained over both catalysts, showing the high decarboxylation activity of these materials under the employed operation conditions (T = 325 • C, 6 bar of N 2 ). Regarding the hydrocarbon distribution ( Figure 5), the main products obtained were gases and gasoline range compounds. The main components in the gases are C 3 and C 4 olefins, proceeding from end-chain cracking reactions of the carbocations species formed by stearic acid decarboxylation. On the other hand, two pathways can be envisaged to explain the formation of gasoline range hydrocarbons: i) random cracking reactions at any position in the chain of those carbocations, and ii) oligomerization/cyclization/aromatization (OCA pathway) of the light olefins formed by severe cracking reactions. Both end-chain cracking and OCA pathways are known to occur on the acid sites of the ZSM-5 zeolite [37][38][39]. Interestingly, the addition of Pd to the n-ZSM-5 support results in the production of less gaseous hydrocarbons and more gasoline. Moreover, the carbon atom distribution obtained over the Pd-containing catalyst exhibited a more regular profile, whereas the selectivity towards gasoline-range hydrocarbons increased from 65.5% over the n-ZSM-5 (32) sample up to 77.5% when using the Pd/n-ZSM-5 (32) catalyst. This positive finding can be assigned to a promotion effect of the OCA pathway provoked by the Pd species. In this way, Pd is known to present activity in the oligomerization and polymerization of olefins [40][41][42][43], which would be in agreement with its participation also in olefin oligomerization reactions. Moreover, Pd exhibits hydrogen transfer activity, which could favor the aromatization step in the OCA pathway [44][45][46][47].

Influence of the Hierarchical Porosity
With the goal of assessing the effect of the presence of hierarchical porosity in the zeolitic support, Pd/n-ZSM-5 (32) and Pd/h-ZSM-5 (33) catalysts were tested and compared in the conversion of stearic acid. Figure 6 shows the conversion and selectivities by groups to the different products so obtained over both catalysts. A quite higher activity was exhibited by Pd/h-ZSM-5 (33) with regards to Pd/n-ZSM-5 (32), with values of the stearic acid conversion of 67 and 33%, respectively. Since both materials present similar acid properties, this remarkable difference can be related to the presence of the hierarchical porosity, which increases the accessibility of the zeolite acid sites and enhances the dispersion of the Pd nanoparticles over the zeolitic support.

Influence of the Hierarchical Porosity
With the goal of assessing the effect of the presence of hierarchical porosity in the zeolitic support, Pd/n-ZSM-5 (32) and Pd/h-ZSM-5 (33) catalysts were tested and compared in the conversion of stearic acid. Figure 6 shows the conversion and selectivities by groups to the different products so obtained over both catalysts. A quite higher activity was exhibited by Pd/h-ZSM-5 (33) with regards to Pd/n-ZSM-5 (32), with values of the stearic acid conversion of 67 and 33%, respectively. Since both materials present similar acid properties, this remarkable difference can be related to the presence of the hierarchical porosity, which increases the accessibility of the zeolite acid sites and enhances the dispersion of the Pd nanoparticles over the zeolitic support.

Influence of the Hierarchical Porosity
With the goal of assessing the effect of the presence of hierarchical porosity in the zeolitic support, Pd/n-ZSM-5 (32) and Pd/h-ZSM-5 (33) catalysts were tested and compared in the conversion of stearic acid. Figure 6 shows the conversion and selectivities by groups to the different products so obtained over both catalysts. A quite higher activity was exhibited by Pd/h-ZSM-5 (33) with regards to Pd/n-ZSM-5 (32), with values of the stearic acid conversion of 67 and 33%, respectively. Since both materials present similar acid properties, this remarkable difference can be related to the presence of the hierarchical porosity, which increases the accessibility of the zeolite acid sites and enhances the dispersion of the Pd nanoparticles over the zeolitic support. Regarding the product distribution, and under the conditions employed in these experiments, diesel range hydrocarbons with selectivity around 15% were also produced over both catalysts, in addition to the gases and gasoline fractions. Interestingly, the use of the hierarchical zeolitic support also led to a remarkable improvement in the gasoline selectivity (72 over Pd/h-ZSM-5 (33) versus 56% when using the Pd/n-ZSM-5 (32) catalyst). This result can be mainly related to the improved Pd dispersion and enhanced accessibility through its mesopores over the hierarchical ZSM-5 sample, which favors the OCA pathway by increasing the closeness between acid and metallic sites and by enhancing hydrogen transfer reactions.
Therefore, the addition of a secondary porosity within the mesopore range in the ZSM-5 support had a paramount influence on the behavior of the corresponding catalyst in stearic acid conversion, leading to remarkable enhancements in both catalytic activity and selectivity towards gasoline range hydrocarbons.

Effect of the Si/Al Atomic Ratio in Hierarchical ZSM-5
The Si/Al atomic ratio of the zeolitic ZSM-5 materials is a key variable, since it determines the acid concentration, which is meant to play a decisive role in the stearic acid conversion and the obtained product distribution. To determine the effect of this parameter, three Pd/h-ZSM-5 samples, with Si/Al atomic ratios of 33, 51, and 122, were assessed as catalysts of this reaction. The attained conversions and selectivity by groups are shown in Figure 7, while the selectivity by atom carbon number is illustrated in Figure 8. It can be appreciated that decreasing the Si/Al ratio provoked an improvement of the stearic acid conversion, but this was not as significant as it might be expected. Thus, a three-fold increase in the aluminum content did not even double the attained conversion. Moreover, it was remarkable the high conversion of stearic acid obtained over Pd/h-ZSM-5 (122) (47%), despite its low acidity. Additionally, if the textural properties of these three catalysts are compared, the catalyst with the highest accessibility (higher mesopore surface area/ lowest micropore surface area) is also the sample with the lowest Si/Al atomic ratio (Pd/h-ZSM-5 (33)). Therefore, considering its enhanced properties, it should provide a much higher conversion than that observed (68%), if only accessibility and aluminum content were responsible for the attained conversions. This fact bears out the relevant role also played by the Pd species, contributing significantly to the decarboxylation activity. Moreover, as all the samples contained similar Pd amounts, the availability Regarding the product distribution, and under the conditions employed in these experiments, diesel range hydrocarbons with selectivity around 15% were also produced over both catalysts, in addition to the gases and gasoline fractions. Interestingly, the use of the hierarchical zeolitic support also led to a remarkable improvement in the gasoline selectivity (72 over Pd/h-ZSM-5 (33) versus 56% when using the Pd/n-ZSM-5 (32) catalyst). This result can be mainly related to the improved Pd dispersion and enhanced accessibility through its mesopores over the hierarchical ZSM-5 sample, which favors the OCA pathway by increasing the closeness between acid and metallic sites and by enhancing hydrogen transfer reactions.
Therefore, the addition of a secondary porosity within the mesopore range in the ZSM-5 support had a paramount influence on the behavior of the corresponding catalyst in stearic acid conversion, leading to remarkable enhancements in both catalytic activity and selectivity towards gasoline range hydrocarbons.

Effect of the Si/Al Atomic Ratio in Hierarchical ZSM-5
The Si/Al atomic ratio of the zeolitic ZSM-5 materials is a key variable, since it determines the acid concentration, which is meant to play a decisive role in the stearic acid conversion and the obtained product distribution. To determine the effect of this parameter, three Pd/h-ZSM-5 samples, with Si/Al atomic ratios of 33, 51, and 122, were assessed as catalysts of this reaction. The attained conversions and selectivity by groups are shown in Figure 7, while the selectivity by atom carbon number is illustrated in Figure 8. It can be appreciated that decreasing the Si/Al ratio provoked an improvement of the stearic acid conversion, but this was not as significant as it might be expected. Thus, a three-fold increase in the aluminum content did not even double the attained conversion. Moreover, it was remarkable the high conversion of stearic acid obtained over Pd/h-ZSM-5 (122) (47%), despite its low acidity. Additionally, if the textural properties of these three catalysts are compared, the catalyst with the highest accessibility (higher mesopore surface area/ lowest micropore surface area) is also the sample with the lowest Si/Al atomic ratio (Pd/h-ZSM-5 (33)). Therefore, considering its enhanced properties, it should provide a much higher conversion than that observed (68%), if only accessibility and aluminum content were responsible for the attained conversions. This fact bears out the relevant role also played by the Pd species, contributing significantly to the decarboxylation activity. Moreover, as all the samples contained similar Pd amounts, the availability of the Pd centers may become the limiting factor for the conversion of stearic acid when using catalysts with high acid sites concentration. 11 of the Pd centers may become the limiting factor for the conversion of stearic acid when using catalysts with high acid sites concentration.  Regarding the product distribution, the three catalysts gave rise to selectivities towards gasoline-range hydrocarbons above 70%, with a slight increase being observed as the acidity of the support was reduced. On the other hand, the selectivity to n-paraffins was lower than 30%, regardless of the used Si/Al atomic ratio. Regarding the selectivity by carbon atom number, it was interesting to observe how the main hydrocarbon fraction obtained at the lowest Si/Al atomic ratios (33 and 51) was C 5 , while for the highest Si/Al atomic ratio, it was C 8 , and to a lesser extent, C 11 . This fact suggests also that the OCA pathway is promoted by the presence of Pd, this effect becoming more evident when using zeolite catalysts with low Al contents, i.e., high Pd/Al ratios. In contrast, at low Pd/Al ratios, the end-chain cracking reactions take place in a higher extent than the OCA pathway, explaining that lighter hydrocarbons are predominant in the product distribution  The effect of the temperature on the stearic acid conversion was determined over the Pd/h-ZSM-5 (122) catalyst by operating at 250, 275, 300, and 325 • C. Figure 9 illustrates the conversion and selectivities by groups so obtained, whereas Figure 10 shows the corresponding selectivities by carbon atom.
At the lowest temperature (250 • C), the conversion was very low (6%) and the product comprised solely C 2 -C 4 hydrocarbons. With the increase in temperature, a strong enhancement of the stearic acid conversion took place, reaching almost total conversion at 325 • C. This confirms the high activity of the catalytic system here investigated, although some contribution of the thermal reaction cannot be discarded when working at the highest temperature.
It is noteworthy that the selectivity towards C 1 -C 4 hydrocarbons dropped steadily with temperature, until reaching a minimum value of 12% at 325 • C. In the same way, the selectivity to products within the diesel was reduced when working at higher temperature. As a consequence, the selectivity towards the gasoline fraction was significantly improved with the temperature, rising from 46% at 275 • C to a remarkable value of 86% at 325 • C. The effect of the temperature on the stearic acid conversion was determined over the Pd/h-ZSM-5 (122) catalyst by operating at 250, 275, 300, and 325 °C. Figure 9 illustrates the conversion and selectivities by groups so obtained, whereas Figure 10 shows the corresponding selectivities by carbon atom.  At the lowest temperature (250 °C), the conversion was very low (6%) and the product comprised solely C2-C4 hydrocarbons. With the increase in temperature, a strong enhancement of the stearic acid conversion took place, reaching almost total conversion at 325 °C. This confirms the high activity of the catalytic system here investigated, although some contribution of the thermal reaction cannot be discarded when working at the highest temperature.
It is noteworthy that the selectivity towards C1-C4 hydrocarbons dropped steadily with Regarding the selectivity by carbon atom, shown in Figure 9, the main hydrocarbon obtained was C 3 at 250 and 275 • C, changing to C 8 and C 11 at 300 • C. The selectivity towards n-paraffins in the C 5 -C 12 fraction was mostly 26% at 325 • C and lower at 250 and 275 • C, pointing out that the main components of this bio-gasoline were branched, naphthenes, and aromatic hydrocarbons formed as a consequence of isomerization, cyclization, and aromatization reactions. The enhanced relevance of the OCA pathway was revealed by the relatively plain product distribution obtained in terms of atom carbon number when increasing the reaction temperature up to 300 • C.
In summary, these results indicate that by adjusting the properties of the ZSM-5 support, as well as the operation conditions, it is possible to achieve almost total conversion of stearic acid with a high selectivity towards gasoline-range hydrocarbons.

Conclusions
Palladium supported over either nanocrystalline or hierarchical ZSM-5 samples showed remarkable properties for liquid-phase conversion of stearic acid into liquid fuels using mild reaction temperatures under nitrogen atmosphere. These catalysts exhibited high selectivity towards gasoline-range hydrocarbons, plus an additional share of gaseous components produced by end-chain cracking reactions. The occurrence of Pd in the catalyst composition led towards an increase in the gasoline selectivity, due to the higher extent of the oligomerization/cyclization/aromatization pathway (OCA), which promoted the conversion of the light olefins, formed as primary end-chain cracking products.
When using ZSM-5 supports with hierarchical porosity, both stearic acid conversion and gasoline selectivity were significantly improved. These facts were caused by the enhanced accessibility and better Pd dispersion attained over the hierarchical zeolites in respect to the Pd/n-ZSM-5 catalyst. It is also noteworthy that the increase of the Al content in the hierarchical Pd/h-ZSM-5 catalysts brought about a lower increase in conversion than the expected one, considering the rise of the zeolite acid site concentration. This fact was indicative of the critical role played by the Pd species in the activity of the catalysts, suggesting that the availability of the Pd centers may become a limiting factor in the conversion of stearic acid.
The enhancement of the reaction temperature increased the stearic acid conversion, reaching virtually complete conversion at 325 • C. Interestingly, the selectivity towards gasoline underwent a high increase with this variable, from 46% at 275 • C to a superb value of 86% at 325 • C. This phenomenon was explained in terms of a larger extent of the OCA pathway.
Accordingly, by adjusting the zeolitic support properties and operation conditions, it is possible to achieve Pd/ZSM-5 catalysts showing both high catalytic activity and selectivity for the conversion of fatty acids (stearic acid) into bio-gasoline.