2.1. Characterization of Catalysts
A series of SiO
2, SiO
2/Zr catalysts, and their counterparts were prepared using different methods by employing an EDTA template and chelate agents. Their physicochemical properties were assessed using various characterizations.
Figure 1 represented the diffractograms of SiO
2, SiO
2/Zr, and their modification using EDTA as a template and chelate as well. The SiO
2 diffractogram revealed a broad peak at 23°, which was assigned to the amorphous silica (ICDD No. 39-1425) [
50]. Another study also observed this typical peak when preparing SiO
2 using the sol-gel method [
51]. Notably, the 2θ peaks at 30.11°, 34.76°, 50.30°, and 60.10° which was attributed to the single tetragonal zirconia phase (ICDD No. 80-2155) [
52]. Some studies reported that the zirconia also had monoclinic, cubic, and mixed crystal structure phases [
53,
54], which depend on the catalyst’s preparation. The amorphous silica phase was still presented on SiO
2/Zr, which indicated that the amorphous silica structure was preserved [
55].
As seen in
Figure 1, the EDTA chelate and template-assisted method promoted a significantly different crystal structure towards SiO
2/Zr. These results indicate that the EDTA species employed with different methods, i.e., chelate and template, positively affected the SiO
2/Zr structure. The SiO
2/Zr-CEDTA revealed an amorphous structure with no zirconia phase were discerned, which suggested that the Zr species was finely dispersed towards the silica surface and relatively diminutive to be identified by the XRD [
40]. This result was also reported by Nadia et al. [
56] when preparing NiMo/silica induced by different species such as NaHCO
3, whereby the Mo species was not distinguishable after being loaded onto silica. Rubab et al. [
48] reported that there was no discernible crystal structure in SiO
2/NiO, which might be attributed to the integration of NiO into the SiO
2 matrix, as well as its amorphous nature and slighter crystallite size. This result also suggests that the EDTA chelate-assisted method was changing the tetragonal phase zirconia to the amorphous structure, which was presumably due to the robust bonded interaction between the EDTA and Zr through the complex bond, thus leading to the inefficient Zr crystallization. By contrast, the SiO
2 prepared by the EDTA template-assisted method (SiO
2-TEDTA) enhanced the silica structure to be crystalline, as indicated, and new peaks were formed, which attributed to the mixed phase, such as quartz (ICDD No. 71-5334) and cristobalite (ICDD No. 89-3606). During the calcination process, the EDTA species was removed and presumably generated highly ordered silica, thus leading to crystalline phase structure. Similarly, the SiO
2/Zr prepared by the template-assisted method was also appeared a crystalline structure. Moreover, the additional peaks at 26.76° in SiO
2/Zr-TEDTA suggested the presence of a monoclinic zirconia phase [
52], whereas the unobservable, other monoclinic phase presumably due to the overlapping peaks between crystalline silica. The relative shift of the tetragonal zirconia phase occurred, likely due to the effect of EDTA species as a templating agent.
The FTIR spectra of SiO
2, SiO
2/Zr, and their modification are depicted in
Figure 2. The absorption band at 1003 and 890 cm
−1 on the SiO
2 catalyst (
Figure 2a) were attributed to the asymmetric and symmetric Si-O stretching vibration [
57], whereas the bending mode of Si-O-Si and O-Si-O bonds were observed at 738 and 421 cm
−1, respectively [
58]. These peaks also appeared in SiO
2 prepared by the EDTA template (
Figure 2c) with a relatively shifted towards the lower wavelength, which suggested the local bonding structure change of O and Si atoms promoted by the EDTA template. Meanwhile, the silanol absorption bands were relatively unobserved on the SiO
2/Zr catalyst (
Figure 2b), which was presumably due to the intensive band of silanol groups, thus overlapping the zirconia groups bands.
As shown in
Figure 2, since catalysts involved the same functional groups, there was no appreciable peak change of SiO
2/Zr prepared by the chelate method compared with the parent SiO
2/Zr, whereas there was a slightly different absorption band at a low wavelength when the template method was employed. This condition suggested that the template and chelate methods affected how the functional groups of silanol as well as zirconia groups bonded in the SiO
2/Zr catalyst. The absorption band at 3650 cm
−1 and 1556 cm
−1 corresponded to the Si-OH and -OH groups, respectively [
55,
59], which were maintained for all catalysts.
The SEM micrographs of modified SiO
2, SiO
2/Zr, and their parent catalysts are presented in
Figure 3. SiO
2 catalyst (
Figure 3a) had a tiny particle distributed uniformly, and after being loaded by the Zr species (
Figure 3b), the agglomerated silica was formed which acted as the support for the Zr species.
Remarkably, the SiO
2/Zr prepared by the EDTA template method (
Figure 3c) displayed a high porous structure allocated uniformly, the same as the SiO
2/Zr-TEDTA (
Figure 3d) but with a less uniformly expanded porous structure. Meanwhile, the SiO
2/Zr prepared by the EDTA chelate method revealed an uneven non-porous structure. These results suggested that the EDTA template-assisted method promoted the crystalline porous structure, whereas the EDTA chelate-assisted method led to the non-porous and amorphous structure.
The EDX mapping of all catalysts is presented in
Figure 4. The expected elements of all catalysts, such as Si, O, and Zr, exist in
Figure 4a–e. It can be seen that there was an adequate homogenous zirconium species dispersion towards the surface of the SiO
2, which could promote the metal-support interaction [
60]. However, the distribution of SiO
2 as catalyst support was displayed differently depending on their preparation, which suggested that the EDTA chelate and template method positively affected the catalysts’ morphological surface by promoting the enhancement of Zr dispersion.
The EDX elemental analysis is shown in
Table 1. It can be seen that the Zr content was presented correspondingly over SiO
2/Zr, and their counterparts, with no impurities. These results suggested that the SiO
2/Zr catalysts were successfully prepared.
The N
2 physisorption of SiO
2, SiO
2/Zr, and their modification using chelate and template methods are presented in
Figure 5. Based on the IUPAC categorization, it was apparent that all the catalysts had a type IV isotherm with a hysteresis loop of type H4, which suggested that the catalysts were mesoporous. Furthermore, the H4 hysteresis loop was related to the micropores with narrow slits feature [
56].
The textural features of all catalysts are shown in
Table 2. The surface area of SiO
2 significantly decreased after being loaded by the Zr species, presumably due to the pore blocking of Zr species [
61]. It can be seen that the SiO
2/Zr-KEDTA and SiO
2/Zr-CEDTA had higher surface area compared with the parent SiO
2/Zr catalyst, whereas the SiO
2 prepared by the template method relatively decreased the surface area of the catalyst but increased the total pore volume as well as the average pore radius of the catalysts. This condition indicated that the template and chelate method affected the textural properties of catalysts. The porous structure generated by the EDTA template method enhanced the pore volume of catalysts, whereas the EDTA chelate method reduced the total pore volume of SiO
2/Zr-CEDTA.
The average particle size distribution of all catalysts is presented in
Figure 6. It can be seen that the SiO
2 catalyst (
Figure 6a) had a relatively uniform particle distribution and subsequently led to a non-uniform particle size distribution after being loaded by the Zr species (
Figure 6b). SiO
2-TEDTA catalyst (
Figure 6c) had narrow particle size distribution with better uniformity compared with SiO
2. In contrast, surprising results were obtained on the SiO
2/Zr catalyst, both prepared using the template and the chelate methods (
Figure 6c,d), as it revealed a non-uniform particle size distribution. However, SiO
2/Zr-CEDTA showed a higher non-uniform particle size distribution than SiO
2/Zr-TEDTA.
The average particle size of all catalysts is presented in
Table 3. It can be seen that the average particle size of SiO
2/Zr was higher than the parent SiO
2, presumably due to the effect of Zr impregnation towards the SiO
2 catalyst, which generated aggregate particles. By employing the EDTA template method, the average particle size of SiO
2 tended to decrease, likely due to the porous structure formation, which reduced the average particle size. Regarding SiO
2/Zr catalyst, similarly, employing the template method decreased the average particle size of the catalyst, whereas the chelate method promoted conversely. The strong interaction of metal species with the EDTA through the chelate method likely provided a dense structure, thus leading to the increase in the catalyst’s average particle size.
The total, surface, and pore acidity of all catalysts are presented in
Figure 7. It was distinctly seen that the SiO
2 had a low acidity feature when compared to SiO
2/Zr. This low acidity feature was correlated to the Si
4+ ions, which acted as Lewis acid sites [
55], and gradually increased due to the existence of Zr species that contribute to the increase in SiO
2 acidity [
62]. The catalyst prepared by the template method seemed to decrease the catalysts’ acidity features for both SiO
2 and SiO
2/Zr, whereas SiO
2/Zr prepared by the chelate method promoted an increase in total and surface acidity as well. The low acidity of SiO
2-TEDTA and SiO
2/Zr-TEDTA was presumably due to the porous structure of the catalyst. The interrelatedness of many active sites on the catalyst surface with one another could happen, thus leading to overlapping between active groups in the catalyst, thereby prompting non-optimal probe absorption [
63]. It also can be seen from
Figure 7 that the active catalyst site of SiO
2 and SiO
2/Zr was likely generated from the catalysts’ pores rather than the surface, whereas the catalyst prepared by the chelate method enhanced the catalyst acidity surface, which consequently decreased the pore acidity. This condition suggested that the template and chelate methods generated different acidity features of the prepared catalyst.
The FTIR-pyridine absorbed of SiO
2/Zr and their modification is presented in
Figure 8. It can be seen that the intensity of SiO
2/Zr and SiO
2/Zr-CEDTA at specific wavelengths were increased after pyridine was absorbed, whereas SiO
2/Zr-TEDTA showed no appreciable intensity change, which is consistent with the gravimetric analysis, as reported earlier. The broad absorption band at ca. 1630 cm
−1 was attributed to the Bronsted acid site initiated by the pyridinium ion [
12], whereas the Lewis acid site was observed at ca. 1450 cm
−1 absorption band [
56]. The strong intensity of the Lewis acid site was noticed on the SiO
2/Zr-CEDTA, suggesting that the chelate method promoted a high catalyst acidity.
2.2. Hydrocracking Test
The catalytic activity of all catalysts was evaluated for CPO hydrocracking by using a catalyst weight of 0.5 g, a hydrogen flow rate of 30 mL/min, a feed flow rate of 20 mL/h, and a hydrocracking temperature of 500 °C for 1 h. The conversion and product yield of the hydrocracking process are presented in
Table 4.
It can be seen that the SiO
2/Zr generated a much higher conversion compared to SiO
2. The presence of Lewis acid from Zr species presumably promoted the increase of hydrocracking conversion. Similar findings were also reported by Alisha et al. [
64], which showed that the addition of metal species on the support (Mo/SBA-15) promoted the conversion of waste palm cooking oil to hydrocarbon. The low conversion of triglycerides through the hydrocracking process catalyzed by silica was also reported by other studies [
55]. Meanwhile, the catalyst prepared by the EDTA template method likely decreases the conversion of both, SiO
2 and SiO
2/Zr catalysts. On the contrary, SiO
2/Zr prepared by the chelate method promoted high conversion up to 87.73%. These results indicated that the acidity of the catalyst affected the hydrocracking conversion, whereby the chelate method preparation promoted the synergetic effect towards high conversion. Based on
Table 3, all catalysts relatively generated high yields due to high-temperature hydrocracking. The SiO
2/Zr had the lowest liquid yield with high gas yield compared with other catalysts. The high gas produced over the SiO
2/Zr catalyst was presumably due to high pore acidity, which promoted the excessive hydrocracking process, thereby increasing the gas yield. The gas product commonly consists of the uncondensable gas such as CO, CO
2, and other C
1-C
5 hydrocarbons. It seemed that the SiO
2/Zr prepared by the template and chelate method higher the liquid yield and reduced the gas yield as well, with pronounced activity observed by the SiO
2/Zr -CEDTA catalyst. Similarly, the SiO
2-TEDTA had a higher liquid yield compared with the parent SiO
2. The porous structure of the catalyst prepared by this method assumably promoted the high liquid yield rather than its acidity. According to these results, the SIO
2/Zr-CEDTA generated the highest liquid compared with another catalyst. SiO
2 catalyst had low coke due to the low acidity of the catalyst, whereas the SiO
2/Zr-CEDTA generated high coke yield compared with others. Wijaya et al. [
55] stated that high catalyst acidity could induce coke formation. Based on
Table 4, the residue yield of the hydrocracking process for all catalysts ranges from 10.04–28.58%. The residue product consists of the unreacted triglyceride, and the lowest residue yield (10.04%) was achieved by the SIO
2/Zr-CEDTA, which suggested that the SIO
2/Zr prepared by the chelate method was the most effective for CPO hydrocracking compared with other prepared catalysts.
Table 5 shows the selectivity product towards bio-gasoline and bio-aviation with different catalysts. It can be seen that the SiO
2 catalyst provided high selectivity towards the bio-diesel product, with the lowest bio-gasoline fraction. The lowest bio-gasoline fraction was generated, presumably due to low catalyst acidity [
55]. Compared with the parent SiO
2, the SiO
2/Zr catalyst increased the bio-gasoline and bio-jet selectivity and decreased the bio-diesel selectivity. Furthermore, the bio-jet selectivity increased after the SiO
2 catalyst was prepared by the template method. This condition suggested that the porous structure of SiO
2-TEDTA provided effective diffusion towards the bio-jet product. During the CPO hydrocracking process, triglycerides and hydrogen gas diffused to the surface or pores of the catalyst and were followed by adsorption to the active site of the catalyst. This step would crack the triglycerides into carbon atoms with lower chains forming the biofuel fraction and some gases. After that, the product was desorbed from the surface of the catalyst and subjected to diffusion into the gas phase. Meanwhile, the selectivity of bio-jet, as well as bio-diesel, is relatively the same when employing the SiO
2/Zr-TEDTA catalyst. Moreover, the SiO
2/Zr prepared by the template and chelate method generated no significant change in the bio-gasoline selectivity.
The reusability study of catalysts was evaluated to understand the catalyst’s stability towards CPO conversion.
Table 6 compares the reusability performance of SiO
2/Zr-TEDTA and SiO
2/Zr-CEDTA toward CPO conversion. As can be seen in
Table 2, the catalytic performance of SiO
2/Zr-TEDA after the second run decreased up to 8.73%, and continually decreased up to 21.28% compared with the first run. Meanwhile, only 11.79% of the decrease in catalytic performance was observed when employing the SiO
2/Zr-KEDTA, which suggested that the chelate method provides a relatively stable conversion towards CPO. The decrease in the conversion during the consecutive run was presumably due to coke deposition [
7]. It has been known that the catalyst could deactivate through coke formation. This coke blocked the catalyst’s active site to initiate the reaction, thereby reducing the catalyst’s performance [
62].