Catalytic activity
According to the results obtained from the characterization study, the materials with the best textural and physicochemical properties were chosen for testing in the HDO of guaiacol. The catalytic activity was compared with that of Pt-SBA-15 (mesoporous silica support) and Pt-commercial activated carbon. Figure 3 shows the catalytic activity in the HDO reaction of guaiacol using different mass ratios of guaiacol and catalyst. Among the synthesized catalysts, Pt-ACS 11 and 12 showed the highest conversions in all cases, followed by Pt-ACS 9 and 7. In principle, we can observe a direct correlation with the greater surface area and better characteristics of the support and the catalyst, both structural and physicochemical, such as a greater proportion of reduced platinum species, smaller metallic particle size, and good dispersion. As can be seen in the catalytic activity, the catalysts with higher mesoporosity increase conversion. Therefore, it was necessary to obtain a greater quantity of mesoporous in the synthesis process. In these mesoporous catalysts the reaction probably occurs mainly in the mesopores than inside micropores; this can be understood in terms of the mass diffusion limitations. It’s probably that Pt nanoclusters insert in mesopore framework (due to the size of the crystallite, it would not enter the micropores), therefore, when adding the metal, the mesopore volume and the area decrease mainly than micropore volume (see Table 3). Then the reaction occurs where the active sites are located (inside mesopores).
If we analyze the effect of the substrate/catalyst mass ratio, we can observe that in the case of a low guaiacol/catalyst mass ratio of 2.8, conversions are high, but not much higher than those when a higher ratio is used, for example of 5.6. This suggests that in the first case, there is an excess of catalyst and that it does not translate into greater conversion. Similarly, when we increased the ratio to 8.4, we observed a decrease in conversion in all cases. Industrially, we consider that an average ratio of 5.6 is appropriate for comparing the catalysts. Figure 4 compares the activity of the most active catalyst synthesized in this study and two catalysts with the same loading of platinum but using a commercial activated carbon support and an SBA-15 support synthesized in a previous work (Rivoira et al. 2016). Figure 4 shows that Pt-ACS 11 is more active than Pt-SBA-15, in addition to the fact that the activated carbon contemplates an environmental (eco-friendly) and, consequently, an economical advantage. The difference in activity between Pt-ACS 11 and these two catalysts must be analyzed in greater depth, considering the characteristics of the support, the nature of the surface, and its acidity.
Pt-ACS 11 shows the highest conversion in all cases. It is evident that by increasing the ratio of the mass of guaiacol and the catalyst, greater differences are observed in the conversion between the different catalysts, especially at shorter times. For this reason, the intermediate guaiacol/catalyst ratio of 5.6 is more suitable for comparison. To compare the activity of the synthesized catalysts more precisely, the turnover frequency (TOF) was calculated considering the number of converted molecules per Pt metal site (Eq. 2, Eq. 3 and Eq. 4); the active surface area of Pt was obtained by hydrogen chemisorption (Table 5). Table 8 shows the results. TOF was calculated at two different reaction times, 200 and 400 min, and for the three guaiacol/catalyst ratios. In Table 8, another TOF was calculated using only the acid sites (total and strong) to relate the catalytic activity to the two types of active sites (metal and acids), and thus interpret and correlate the conversion results with the nature of the catalytic surface of the various materials. According to the TOF values for all guaiacol/catalyst ratios, Pt-ACS 11 and Pt-ACS-12 are the most active followed by Pt-ACS-9 and 7. This trend is more evident at higher ratios. The increased activity can be easily attributed to the structural characteristics and properties of the catalysts. Catalysts whose supports have greater surface area and better mesoporosity allowed better dispersion of the platinum species, the active sites of the hydrogenation/hydrogenolysis reaction.
The TOF values at 400 min are almost always lower than those at 200 min, indicating that the reaction rate decreases as the conversion increases, probably due to the formation of products that block the active centers and compete. Pt-ACS 11 is, from the point of view of its physical and chemical properties, the best catalyst: it has the greatest surface area, the best dispersion of Pt, and the smallest metal particle size, in addition to the greatest proportion of reduced Pt, which is known to be required for the hydrogenation of guaiacol. However, in some cases, the TOF of Pt-ACS 12 is greater than that of Pt-ACS 11. This indicates that the reaction is not as sensitive to the dispersion of Pt, or that there may be another factor involved in the mechanism. To elucidate this, we calculated the TOF in relation to the acidic sites of each catalyst. The acidic sites were calculated by Boehm titration; values are listed in Table 7. The TOF of the acid sites were calculated for the total number of acid sites and strong acid sites. If we consider the values in Table 8, the trend is different for both types of sites, i.e., the highest TOF for the total acid sites is found for the most active catalysts Pt-ACS 11 and 12 for the three guaiacol/catalyst ratios. However, this trend is reversed when we calculate it using strong acid sites. This suggests that the active acid sites of this reaction are sites of medium acidity and not sites of strong acidity. Paying special attention to both acidic and metallic TOF values at the highest guaiacol/catalyst ratios (more guaiacol/less catalyst), it can be observed that Pt-ACS 11 is the most active catalyst. In addition, there is a synergistic effect between the platinum sites and the medium acid sites.
Table 8
Catalyst | TOF2.8 | TOF5.6 | TOF8.4 | TOF2.8 Acid sites, 200 min | TOF5.6 Acid sites, 200 min | TOF8.4 Acid sites, 200 min | % HDO |
min | min | min |
200 | 400 | 200 | 400 | 200 | 400 | Total | Strong | Total | Strong | Total | Strong | |
Pt-ACComm. | | | 440 | 370 | | | | | | | | | 146 |
Pt-SBA-15 | | | 495 | 374 | | | | | | | | | 191 |
Pt-ACS7 | 310 | 196 | 475 | 376 | 329 | 342 | 0.29 | 0.82 | 0.45 | 1.25 | 0.31 | 0.87 | 218 |
Pt-ACS8 | 208 | 154 | 275 | 208 | 103 | 125 | 0.25 | 0.68 | 0.33 | 0.90 | 0.12 | 0.34 | 110 |
Pt-ACS9 | 308 | 190 | 527 | 383 | 477 | 441 | 0.34 | 0.88 | 0.58 | 1.51 | 0.53 | 1.36 | 205 |
Pt-ACS10 | 245 | 169 | 226 | 155 | 155 | 156 | 0.37 | 0.86 | 0.42 | 0.99 | 0.23 | 0.55 | 109 |
Pt-ACS11 | 329 | 204 | 387 | 683 | 683 | 491 | 0.39 | 0.73 | 0.70 | 1.32 | 0.81 | 1.52 | 300 |
Pt-ACS12 | 345 | 213 | 407 | 654 | 654 | 522 | 0.40 | 0.77 | 0.68 | 1.31 | 0.76 | 1.46 | 292 |
TOF calculated considering the number of converted molecules per platinum metal site (col 1–3) at 200 and 400 min of reaction time, and TOF calculated considering the number of converted molecules per total and strong acid sites (col. 4–6). In all cases for the three guaiacol/catalyst ratio. Col.8 shows TOF for HDO products.
According to the main reaction products identified by GC chromatography in this study, a simplified scheme for the HDO of guaiacol is shown in Scheme 1. Phenol, anisole, catechol, benzene and cyclohexane were identified. Many other peaks corresponding to heavy products and/or oxygenated products were observed at high retention time in the chromatography, they were grouped as “other products” since they appeared in very small quantities or were difficult to identify. Toluene and methyl cyclohexene could be formed according to Ghampson et.al (2012) but were not detected in this study.
The evolution of the reaction explained by the mechanism proposed in Scheme 1 arises from the identified main products mentioned above and the understanding of three possible pathways: demethylation (DME), demethoxylation (DMO), and dehydroxylation (DHY) of guaiacol. For instance, C. A. Teles et al. (2022) reported that the HDO reaction of guaiacol over different supported Pd catalysts may occur as follows: the methoxy group removal from guaiacol to produce phenol can take place by direct DMO producing methanol, or indirectly by DME of anisole (formed by DHY of guaiacol) producing methane, or by DHY of catechol (formed by DME of guaiacol) releasing water. Thus, phenol can be considered a main product or an intermediate for further deoxygenation in guaiacol HDO. Lai et al. (2016) proposed a complex mechanism for the HDO of guaiacol, including these three pathways over Ni@Pd and Ni@Pt bimetallic catalysts involving formation of xylenols and xylenes. Methyl transfer (conversion of methoxy to a methyl group instead of methane) may occur in anisole to produce cresol followed by subsequent methylation, resulting in xylenol. This pathway can be explained in terms of the acidic properties of the support and the noble metal catalyst (Gutierrez et al 2009; Bui et al 2011; Nimmanwudipong et al. 2011). Anisole can also be deoxygenated by DMO to form benzene, but direct transalkylation may occur, also affording xylenol, as proposed by Lai et al. (2016). These products were not favored under our reaction conditions. The latest stage of the mechanism occurs specifically in consequence of the presence of platinum active sites involving the appearance of the two deoxygenated and desired products (benzene and cyclohexane), in which phenol can be further hydrodeoxygenated, forming benzene followed by the hydrogenation of the aromatic ring to yield cyclohexane (He et al. 2018; Gao et al. 2015). These consecutive reactions agree with those proposed by Zhu et al. (2011). They reported that acid sites catalyze methyl transfer, while metals catalyze demethylation and hydrodeoxygenation.
Product yield and selectivity were calculated considering 100% of the products identified in the chromatography, including "other" products. Guaiacol HDO products observed in this work seem to correspond to a series of consecutive hydrodeoxygenations, as proposed in the literature [Teles et al. 2018, Ghampson et al. 2012) (see Scheme 1). Demethylation occurs to yield catechol, but it is likely to occur in a smaller proportion.
The selectivity of the synthesized catalysts for the different products of guaiacol HDO is presented in Fig. 5. Phenol was the main product in all cases, followed by anisole and catechol, according to the reaction steps described in Scheme 1. Then, the desired deoxygenated products such as benzene and cyclohexane were observed for all catalysts. However, the most active catalysts, Pt-ACS 11 and 12, obtained the highest selectivity of 18% and 12%, respectively, followed by Pt-ACS 9 and 7 (10%).
Figure 6 compares the selectivity of Pt-ACS 11 with Pt-SBA-15 and Pt-AC Commercial. The main products of the three catalysts were phenol, anisole, and catechol; phenol and anisole are direct products of HDO with consecutive formation of benzene and cyclohexane. All these products (phenol, anisole, benzene, and cyclohexane) were considered to calculate the HDO ratio, which is a percentage of the guaiacol conversion to deoxygenated products, since there is a portion of the reagent that goes to the undesired pathway, producing catechol and other products. HDO % was calculated as follows: HDO % = (selectivity of phenol + anisole + benzene + cyclohexane) x guaiacol conversion/100. It is important to clarify that phenol and anisole are products from partial dexoygenation of guaiacol while benzene and cyclohexane from total deoxygenation. The selectivity to benzene and cyclohexane for Pt-ACS 11 was much higher than that for Pt-SBA-15 and Pt-AC Commercial catalysts. The TOF of HDO % with respect to the Pt active sites was calculated; the results are shown in the last column of Table 8. As expected, the Pt-ACS 11 catalyst shows the highest TOF, followed by Pt-ACS 12. The TOF values obtained for Pt-SBA-15 and the commercial catalyst are lower. The last two catalysts present a large proportion of catechol, indicating that the demethylation pathway is favored in these catalysts.
Comparing our results, we see that they are in good agreement with those found in the literature similar systems and conditions. Teles et al. (2022) investigated the HDO of guaiacol over supported Pd catalysts at atmospheric pressure and a temperature of 300°C. Demethoxylation which afforded phenol was the major reaction pathway for all catalysts, showing only a slight contribution from the demethylation reaction. Yet, a significant dehydroxylation reaction was still found in catalysts with Pd supported on ZrO2 and TiO2. Author reported that the hydroxyl group was strongly adsorbed on the catalyst surface, which could block the catalytic sites and hinder further conversion of phenol, ultimately leading to reduced deoxygenation rates. The conclusion drawn was that demethylation with the production of catechol as a reaction intermediate was the most straightforward reaction. In this work, however, phenol and anisole were the main products yielded, which suggests that phenol was directly formed through cleavage of CAr–OCH3 bond in all catalysts. Following Teles et al. (2022) anisole was afforded by dehydroxylation of guaiacol. As the cleavage of CAr–OH bond requires higher energy that that of CAr–OCH3 bond (414 and 356 KJ/mol, respectively) (Bui et al. 2011), demethoxylation is expected to produce phenol over dehydroxylation. This result was consistent with ours. Ghampson et al. (2012) researched the HDO of guaiacol in a batch reactor over SBA-15 silica-supported molybdenum nitride catalysts at 300°C and 49 atm of H2 pressure. SBA-15 silica-supported catalysts transformed guaiacol directly to phenol through demethoxylation with no production of catechol. The lower catechol production with the SBA-15 silica support was significant to minimize coking reactions and reduce consumption of hydrogen. They obtained 44% conversion with 30% yield of phenol. All catalysts in this study yielded more phenol than catechol. SBA-15 support shows Si-OH groups, which act as Lewis acid sites, allowing demethoxylation. In this study, with the carbons activated with phosphoric acid, the acidity obtained and determined by Bohem titration showed that direct demethoxylation is also favored. The commercial carbon support showed greater ability for demethylation, yielding catechol, transformed into phenol through hydrogenolysis (Saidi et al 2014; Sepulveda et al 2011; Yang et al. 2014). Pt/C catalyst was studied in the HDO of guaiacol in a fixed-bed reactor at different temperatures and atmospheric pressure, affording high kinetic constants in the HDO of guaiacol. The primary products of the liquid-phase reaction were phenol, catechol and cyclopentanone (Zhu et al.2011). The HDO of guaiacol was also studied on La-modified Pt/Al2O3 batch reactor at 215°C and 30 atm, obtaining 80% of HDO at 100% conversion (Escobar et al. 2023).
It is widely agreed that the HDO reaction requires a bifunctional catalyst, i.e., a metal associated with a specific support close to an acid site. Some studies suggest that the reaction occurs on the metal sites close to an acid site.
Figure 7 summarizes the comparison between the catalysts and correlates the guaiacol conversion with the HDO ratio and the total acidity of the catalysts synthesized, using the activated carbon supports. The higher HDO ratio correlates well with the higher conversion and with total acidity.
The good performance of activated carbon could be linked to the incorporation of phosphorus species in the structure. Li et al. (2018) have explored the impact of phosphorus on the hydrogenation process, employing a P-doped Ni/Al2O3 catalyst. The results indicate that the incorporation of phosphorus species could act as a catalytic promoter, enhancing the activity and selectivity of the catalyst. Chen et al. (2020) also investigated the hydrogenation of nitrobenzene to aniline using phosphorus-doped carbon nanotubes and discovered that phosphorus could act as a catalytic promoter. Subsequently, Yangcheng et al. (2023) found that the presence of active P species in Pd/PHS led to the selective conversion of vanillin and, consequently, improved the yield of hydrodeoxygenated compounds. These studies have demonstrated the benefits of the presence of phosphorus in HDO reactions. Dieu-Phuong et al. (2020) studied the hydrodeoxygenation of oleic acid over P-modified catalyst and observed that the HDO reaction is highly dependent on the potency and arrangement of acid sites. Therefore, the introduction of phosphoric acid enhances the reactivity of the electrophilic acid center, effectively activating the oxygen atom of some of the functional groups present on the carbon surface, e.g. the acyl bond, making it susceptible to attack by hydrogen atoms (Scheme 2). Hence, we could explain the better performance of our catalysts due to the synergy between the acid sites and metal sites in the guaiacol HDO reaction. In the current investigation, phosphorus was incorporated during the synthesis of carbon as an activating agent, and P compounds were incorporated into the material without the need for post-synthesis incorporation. The presence of phosphorous, responsible for the acidity of the surface materials, could contribute to the improved performance of the HDO reaction. The presence of oxygenated surface groups, such as phenolic compounds and carboxylic acids, on the AC surface facilitates interaction with the oxygen unshared pairs of electrons in phosphoric acid and with phosphate anions. They probably react to produce an acyl group (R-CO-) generated by the elimination of hydroxyl group from a surface carboxylic acid by phosphorous action (Scheme 2). The acyl group generally exists in its cationic form because its radical form is particularly unstable and decomposes quickly.
In the mechanism proposed, the cationic acyl generated (Lewis acid site) can accommodate the unshared electrons present in the guaiacol hydroxyl group (–OH) or in its methoxy group (–O-CH3) due to its electrophilic nature. Nucleophilic attraction (ion–dipole) keeps the substrate molecule adsorbed over the carbon surface. The –OH o –O-CH3 guaiacol bond is susceptible cleavage by the entry of hydrogen adsorbed on the neighboring Pt metal site. Analogous mechanisms may occur with other carbon groups that, in a similar way, interact with the phosphorous species, attracting the guaiacol molecule and its intermediates.
On the other hand, chemisorption of H2 showed that the Pt nanoparticles were well dispersed on the support material. Yangcheng et al. (2023) studied the dispersion of Pd metal particles on a zeolite, with and without phosphorus species. They observed that in their nanoparticle-supported catalyst, there was a shift towards a higher dispersion, achieving nanoparticles with a size around 1.5 nm. They found that the presence of phosphorus enhanced Pd dispersion on the support. Thus, we believe that the acidity incorporated in the activation stage of the catalyst through phosphoric acid, has a positive influence on the dispersion of the metallic particles. Finally, the joint effect of the higher particle dispersion (which could be attributed to acidity and the large surface area obtained for the synthesized catalyst) and the contribution of acidity to HDO reactions make the Pt particles over the active carbon a particularly efficient catalyst. Considering the low manufacturing cost, we could affirm that it is a promising material for the HDO reaction of biofuels.