3.1. N2 isotherm and pore size distribution analysis
Figure 1 displays N
2 adsorption-desorption isotherms of the samples at -195.781°C and the corresponding Barrett-Joyner- Halenda (BJH) pore size distribution based on the adsorption branch.
Figure 2(I) shows that all the curves of the tested samples have similar features and represent a combination of type I and IV isotherms [
27,
28]. The adsorption branch of the curves, which has a type I isotherm, is related to the microporous structures of the catalyst [
29,
30]. The hysteresis loop appears above p/p0 = 0.4, for the fresh catalyst, which can be attributed to IUPAC H4-type, indicating a complex structure with mesopores and micropores[
28,
31]. The hysteresis loops for regenerated and spent catalysts above p/p0 = 0.43 can be attributed to the IUPAC H3-type, which at p/p0 > 0.7 indicates the absence of limiting adsorption due to the accumulation of coke over the mesopores of the catalyst [
32,
33].
At the relative pressure
p/
p0 of about 0, the sharp jump in the amount of N
2 adsorption, located on the fresh catalyst (curve (a)), gradually decreases on the regenerated catalyst (curve (b)) and finally disappears from the spent catalyst (curve (c)), indicating a progressive plugging of the micropore by the deposition of carbon species, as confirmed in
Table 2. The gradual decrease in pore volume from fresh to spent catalyst is also evidence of pore plugging resulting from the deposition of carbon on the catalyst surface. The parallel features of the adsorption-desorption curves in the range of 0.55-0.75 and 0.53-75 for regenerated and spent catalysts, respectively, indicate that these large mesopores are open and may correspond to the catalyst particles seen in the SEM images shown in Figure 4 [
33].
Figure 1(II) displays the distribution of fresh, regenerated, and spent catalyst pore sizes. It can be observed that the fresh catalyst sample has one peak at 46.98 Å, while the regenerated and spent catalyst samples have two peaks at 43.25 Å-318.66 Å and 46.93 Å-317.44 Å, respectively. These second peaks at 318.66 Å and 317.44 Å indicate the presence of mesopores in the samples due to the thermal conditions of reaction and regeneration [
1]. As can be seen in
Figure 1 (II), the uniform structure of the pore distribution of the fresh catalyst is significantly changed during the process. This process can be seen in the curves of the regenerated and spent samples, indicating a relatively large pore blockage in the pore diameter range from 159.98 Å to 318.66 Å.
Table 2 shows that the BET surface area is greatly reduced during the process. The BET surface area of the fresh catalyst of 158.13 m
2g
-1 decreased by more than 50% compared to the regenerated catalyst of 76.78 m
2g
-1 and by more than 60% compared to the spent catalyst of 57.45 m
2g
-1. This event can be referred to the coke deposits on the catalyst surface and metal poisoning or the regeneration temperature above 690°C, which changes the catalyst structure and porosity [
34]
The calculated values of microporosity show that the microporosity of the regenerated catalyst (49.79%) is higher than that of the fresh and spent catalyst (40.39%) and (49.73%), respectively, which is due to the presence of a large mesopore peak of 318.66 Å created by the effect of regeneration gas flow [
1,
35,
36]. However, based on the values for the surface area and porosity of the regenerated catalyst compared to the values of the spent catalyst, it is found that there is a limit to the inverse relationship between surface area and porosity (the surface area of the catalyst decreases while its porosity increases) [
37].
All the texture changes shown in
Table 2, such as pore distribution, BET surface area change, microporosity, and pore volume, were directly related to the catalyst deactivation factor and coke mechanism, including the feedstock and the catalyst reaction-regeneration conditions [
9,
21,
22,
38].
3.2. Crystallography analysis
The XRD patterns of all samples are shown in
Figure 2. The XRD patterns of the fresh sample were compared with those of the regenerated and spent samples to better understand the evolution of catalyst structure due to thermal conditions during the process.
The alumina was present in the fresh sample in the form of Al
2Si
2O
5(OH)
4, kaolinite with a monoclinic system (ICDD: 04-026-7575) [
39]. The characteristic peaks were found at 2θ ≈ 12.40°, 21.23°, 23.13°, 24.96°, 26.51°, 38.36°, and 45.40°, corresponding to the reflections of the (001), (-1-11), (0-21), (002), (111), (-202), and (-203) planes. Other background peaks are associated with the crystalline phase of quartz (ICDD: 98-000-0369, [
40]) with diffraction features at 2θ values of 20.85° and 26.64° corresponding to the reflections of the (100) and (011) planes. In addition, the fresh XRD result confirms the presence of staurolite (ICDD: 98-000-0410) [
41] appearing at 2θ≈10.63°, 31.59°, 35.13, and 37.47°, corresponding to the reflections of planes (020), (002), (-241), and (132).
However, the XRD patterns of the regenerated catalyst showed that the peak intensities of kaolinite(kaolinite-1A) decreased due dealumination process regardless of the regeneration conditions. Kaolinite-1A (ICDD: 01-083-4643) [
42] appeared essentially at 2θ≈12.19°, 20.04°, and 24.53°corresponding to the reflections of the (001), (110), and (002)-planes, respectively. The silica amorph is also observed as one of the main crystalline phases in the regenerated sample, appearing at 2θ≈ 10.13° and 19.81°, corresponding to the reflections of the (010) and (013) (ICDD: 04-011-9355) [
43] planes of SiO
2 species. The elements Vanadium and nickel were clearly detected in the regenerated sample. The element vanadium is present in the form of lanthanum vanadium oxide hydrate with diffraction peaks at 2θ≈ 12.10°and 13.83° (ICDD: 04-011-1951) [
44], while the element nickel oxidized (NiO) with a diffraction peak at 2θ ≈ 37.76°and 54.48° (ICDD: 04-006-6545) [
45]. These diffraction peaks correspond to the reflections of the (-111), (102) and (200), (220)-planes of LaV
50
14(H
2O)
13 and NiO, respectively. It was also found that the sulfur reacts with the iron and other components to form iron-sulfur carbonyl. This Fe
5S
4(CO)
15 (ICDD: 04-010-2192) [
46] appeared at 2θ≈11.90°, 15.98°, and 20.89°, corresponding to the reflections of the (102), (020), and (220) planes.
The XRD results show that the kaolinite-1A diffraction peaks become very weak and harder to detect in the spent catalyst. This may be related to the loss of the hydroxyl function OH, regardless of the process conditions and residence time of the spent catalyst. It appears that the kaolinite-1A was converted to amorphous y-Al
2.144O
3.2(ICDD: 98-000-0059) [
47], silica (ICDD: 01-073-3406) [
48], and residual kaolinite (kaolin serpentine, 1A) due to dihydroxylation and the presence of metal oxide species reacting with the kaolinite during the process [
32,
49,
50]. Qijun Hu et al. conducted a similar study on the evolution of kaolinite structure and suggested that the temperature range 700-800°C corresponds to the transition of meta-kaolinite into γ-alumina and silica spinel [
51]. The y-Al
2.144O
3.2 diffraction peaks appeared at 2θ ≈ 31.97°, 37.68°, 45.84°, and 66.84°, while the silica peaks appeared at 2θ ≈ 10.32° and 11.92°. These peaks correspond to the reflections of the (220), (311), (440), (400), 440)-planes and (-113), (400), (-113)-planes of y-Al
2.144O
3.2 and silica, respectively. The remains of the kaolin-serpentine (ICDD: 04-025-7033) [
52] appeared essentially at 2θ ≈ 12.36°, 20.48°, 21°, and 24.87°, corresponding to the reflections of the (010), (-101), (111), and (020)-planes. This is due to the fact that certain closed particles require a lot of energy to be dehydroxylated during thermal treatment at high temperatures. A large dispersion of Vanadium, Nickel, and Sulfur is observed in the spent catalyst. In particular, the dispersion of vanadium peaks becomes more intense along the XRD curve. However, due to the reaction with nickel and sulfur, the peaks are more visible than others and can be identified at 2θ ≈ 15.66, 31.26, 34.86°, 45.50°, and 54.81°, corresponding to the reflections of (002), (110), (-112), (114), and (310) planes of V
2NiS
4 species (ICDD: 04-001-6617) [
53]. In addition, the nickel was also associated with sodium to form sodium-nickel oxide with diffraction reflections at 2θ ≈ 15.68° and 37.35°, corresponding to the reflections of the (001) and (110) planes of Na
0.2NiO
1.8 species (ICDD: 04-009-2679 BITYUTSK. PN, et al. (1969)). Three copper sulfide peaks were identified at 2θ ≈ 27.56°, 31.93°, and 66.78°, corresponding to reflections of the (111), (1200), and (400) planes of Cu
2S (ICDD: 03-065-2980 Barth, T. et al. (1926)). A diffraction peak that appeared at 2θ ≈ 45.62° is the characteristic peak of the iron silicon corresponding to the (110) plane of the Fe
0.07Si
0.33 species (ICDD: 04-002-3903) [
54].
The crystalline phases of GeO
2 (ICDD: 04-021-5362 Li Y.F., et al. (2014)) and TiO
2 (ICDD: 98-000-0375) [
55] are also detected at 2θ ≈ 11.76°and 27.44°, 54.32°, respectively. These peaks correspond to the reflections of the (211) and (110), (211) planes of GeO
2 and TiO
2, respectively.
The crystalline phases of Quartz stay unchanging at all steps of the process (regenerated and spent catalysts) and keep their diffraction peaks at 2θ of 20.85° and 26.64°.
According to the XRD analysis, during the thermal treatment of the catalyst, it was clearly observed that the texture changed due to the intermolecular reaction. This clearly indicates that some components lose their crystallinity and crosslink to form new structures, as shown in
Figure 3 [
1,
31]. Under the same thermal conditions, there are no significant peaks of gamma alumina on the regeneration XRD curve in contrast to the spent catalyst. This occurrence of gamma alumina peaks in the spent catalyst shows that the residence time of the catalyst in the process has an important influence on the kaolinite transformation process. Furthermore, this means that at a constant thermal condition, the linkage and interaction of catalyst components depend on the residence time of the catalyst in the process. The longer the residence time, the deeper the reaction between the components.
The presence of various forms of nickel, sulfur, and vanadium as major impurities in the regenerated and spent catalysts contributes to catalyst deactivation and attrition factors [
10,
11,
56,
57,
58,
59,
60,
61]. These poisons occur naturally in the atmospheric residue (feedstock) and are associated with various structures during the process. Marris D. et al. studied the effect of sulfur in catalyst reaction and showed that sulfur on the catalyst surface affects the performance of catalyst activity[
17]. U.J. Etim et al. concluded that vanadium and nickel have an important influence on the FCC catalyst behavior. The first one has an influence on the catalyst activity, while the second one favors the coke formation through the mechanism of dehydrogenation reaction [
11].
Table 3 presents the main tested sample’s chemical oxide components, which were determined using a nondestructive analytical X-ray fluorescence technique to complete the XRD study. The XRF results show a gradual increase of iron content in catalyst samples during the catalytic cracking process of the atmospheric residual oil as feedstock (3.2 ppm of Iron)[
26], which may have some important implications, including a direct poisoning effect on the catalyst active sites, an effect on the selectivity of some reactions, a physical effect of iron deposition (pore plugging), and physical deconstruction [
57]. The presence of lanthanum in the form of lanthanum oxide (La2O3) in the catalyst sample is related to improving catalyst resistance, assisting in the degradation of heavy molecular chains, and promoting good product stability [
62,
63,
64,
65,
66]. In addition, the amount of alumina and silica, which are the main components of all tested samples, increased from fresh to regenerated catalyst, which can be attributed to the dealumination process [
22].
3.3. UV-visible near-infrared spectra and Raman spectra analysis
Figure 4 shows the Raman spectra of fresh(I), regenerated (II), and spent (III) catalysts. The fresh catalyst exhibited Raman bands of 339.14, 1363.69,1594.94, and 2915.96 cm
-1, while the regenerated catalyst showed bands of 143.93, 394.55, 508.68, 633.35, 1368.14, 1568.47, 2473.42, and 3104.40 cm
-1. The spectra bands of 215.37, 267.81, 655.64, 1093.30, 2513.65, and 3102.48 cm
-1 were identified from the spent catalyst.
The fresh Raman bands of 1363.69 and 2915.96 cm
-1 were identified and corresponded to the stretching vibration of bond P=0 (P
20
5) and bond OH, respectively, while the band of 339.14 cm
-1 is assigned to the bending mode of a 6-membered ring in the aluminosilicate zeolite structure [
67].
The bands of the regenerated Raman spectrum of 143.93, 394.55, 508.68, and 636.35 cm
-1 can be assigned to the lattice vibrations in crystals (LA modes), the Si-O-Si bond, the asymmetric stretching vibration of GeS
2 and the stretching vibration of TiO
2 bond, respectively. The bands of 1368.47 cm
-1 and 1568.47 cm
-1 refer to the ring stretches of polyaromatic species and the ring stretches of Aliphatic azo respectively. It is important to note that the regenerated Raman bands of 1368.47 and 1568.47 cm
-1, indicative of the carbon structures, are located near the overlapping peaks of the D1 and G bands, which are disordered graphite lattice (graphene layer edges) and ideal graphite lattice [
31,
68,
69].
The bands of 2473.47 cm-1 and 3104.40 cm-1 of the regenerated catalyst correspond to P-H and 0-H bonds, respectively. Interestingly, all the bands disappear in the fresh and regenerated catalyst or become very weak in the spent catalysts due to the higher carbon content. However, the Raman spectrum of the spent catalyst shows a very low-frequency vibration that can be associated with the final stages of the crystallization process, such as the bonds at 215.37 and 267.81cm-1. The stretching vibration of the P=S bond at 655.64 cm-1 was also identified as the main peak in the Raman spectra of the spent catalyst.
Figure 4 IV displays the UV-visible near-infrared spectra of fresh, regenerated, and spent catalysts. All curves have similar features and exhibit two important peaks between 214 nm and 297 nm. The UV-vis band of 214 nm is presented on both curves of the tested samples with low absorbance in the spent catalyst, which is due to the reduction of Al content during the process and is assigned to the πp-πd transition between O
2- and Al
3+ [
49,
70]. The band of the fresh UV-Vis spectrum of 297
λmax (nm) is related to the band of the regenerated and spent catalyst of 267
λmax (nm) and can be attributed to the C=O bond (π→π*).
3.5. Catalyst morphology changes during the FCC Process.
The images of crystal morphology of the tested fresh, regenerated, and spent catalyst samples with sizes of 200 μm and 50μm are shown in
Figure 6 using field emission scanning electron microscopy (SEM). The images of the fresh catalyst SEM shown in
Figure 5a indicate that the particles are uniformly spherically charged and have a clean surface and boundaries. It is found that there are smaller spherical particles attached to the largest particle surfaces of the fresh catalyst, which may happen during the preparation of the catalyst. As can be seen in the images of the fresh catalyst, the lightest area of the particles has crystallites, while the darkest area is support.
The regenerated SEM images (
Figure 6b) show a significant physical texture change, the mesoporous structure almost disappears, and the particles become dark. This could be related to the metal poison [
38,
56] and coke deposited on the catalyst (0.04-0.05 wt.%), which was confirmed by using the
LECO carbon analysis. It can also be observed that the small particles attached to the large particles of the fresh sample were broken during the process and the shape of the particles became more surrounded due to process temperature and running time as reported by Xinzhuang Zhang et al [
73,
74].
Figure 6c shows the images of the spent catalyst SEM after deactivation with a carbon content of 1.26 wt% on the catalyst. These images show a complete disappearance of the mesopores, physical abrasion of the particles, and fragmentation into smaller pieces. These phenomena may relate to collisions between particles or particles against the tube [
6,
11,
75], lateral cracks on the particle surface, and surface attrition due to stresses caused by fluid dynamics [
12], as well as poisonous metals such as iron, nickel, vanadium, and sulfur, or the reaction between iron and calcium [
57,
60]. The catalyst’s attrition such as catalyst abrasion and fragmentation, is an important factor affecting particle properties [
10,
74,
76,
77,
78,
79,
80,
81].
The profiles of the weight percentages of the main chemical elements confirm the thermal behavior of the catalyst components during the catalytic fluidized bed cracking and give a clear picture of it. From the Si/Al ratio shown in the figures of SEM (a, b, and c), it can be observed that the regenerated catalyst cannot be recovered as fresh, and the more the catalyst is treated, the more the Si/Al ratio approaches 1, which is characteristic of spent catalyst [
82].
3.6. Thermal Analysis
The sample’s thermal events were analyzed through a thermo-gravimetric (TGA) profile along with a derivative thermo-gravimetric (DTG).
Figure 7 shows the behavior of uncalcined catalysts from fresh, regenerated, and spent catalysts as presented in
Table 3. The TGA results show that five main thermal events were identified as mass loss zones across all curves. For all samples, the temperature range RT (room temperature) to 100°C was not considered because of the influence of moisture.
The first mass loss (zone 1) was measured in the temperature range of 100-300 °C. In this range, it is observed that the main mass losses for fresh, regenerated, and spent catalyst samples are estimated to be 4.06 wt.%, 2.73 wt.%, and 1.03 wt.%, respectively. This first event can be attributed to the release of water bonds from the tested catalysts as a dehydration phase on the surface of SiO
2 in the tested samples. In this step, it is interesting to note that the mass loss values indicate a gradual decrease from fresh to spent catalyst, which is due to the chemical change of the catalyst during the process. This process can be explained by the fact that the density of hydroxyl groups (OH) is higher in the fresh catalyst and decreases during the thermal treatment [
74,
83].
For all zones, the mass losses of the fresh catalyst are associated with the progressive removal of its components, including the removal of co-precipitated nitrate, NiO, GeO2, Rb2O, and ZrO2 over the temperature zone ranges. However, in the process reaction that occurred at a temperature of 500°C, the mass loss profiles and the endothermic DTG peak at 499°C showed that the fresh catalyst lost significant weight before the reaction took place in the riser-reactor.
The second and third mass loss in the regenerated and spent samples occurred between 300 °C and 550 °C and was the most important phase due to the combustion of coke on the catalyst surfaces.
In the third zone, the DTG curves of the fresh and spent catalysts exhibited strong endothermic peaks at 499 °C and 540 °C, respectively. These peaks can be attributed to the decomposition phase, i.e., the penetration of heat between the layers of decomposed materials, which led to the decomposition of hydrocarbons.
The fourth zone was particularly characterized by the fine and small exothermic peak at 609.90 °C from the regenerated DTG curve. Possibly, this peak is related to the inert carbon or graphitic carbon in the catalysts, as confirmed in the first part of the Raman spectra analysis.
The mass loss in zone 4 (600-700 °C) and above for all samples may be related to the oxidation of the metallic components and crystalline alteration of the catalyst supports, as reported in several studies in the literature [
84,
85,
86].
The equilibrium catalyst behavior at different stages of the FCC process (fresh, regenerated, and spent catalyst) shows a significant change in its composition, which can be seen from the mass loss profiles presented in
Table 3.
Accordingly, in the fifth mass loss zone 1.87 wt.%, 1.78 wt.%, and 0.32 wt.% for fresh, regenerated, and spent catalysts, respectively, there are no other significant weight losses up to 850 °C. This event could be referred to the final hydrocarbon oxidation and the conversion of the trapped metal into metal oxides [
84].
3.7. Kinetics and Thermodynamics Parameters calculation using TGA
Assuming that all zone conversions are first-order reactions, the model of Coats and Redfern was developed to calculate the kinetic and thermodynamic parameters for each zone using the following equation and the constants of Plank and Boltzmann[
87,
88,
89,
90]:
Where
wi: initial weight, wt: weight of sample at particular temperature T, wf: final weight.
A: pre-exponential factor, : heating rate (10 °C/min), R: gas constant (8.3143 Jmol-1K-1), Ea: activation energy, T: temperature (K) at the peak of the DTG curves.
The slope and intercept values of each zone were obtained by plotting
versus 1000/T, then we used these values to calculate apparent activation energy and other thermodynamic parameters listed in
Table 4, such as pre-exponential factor (A), Gibbs free energy
, Entropy Changes
, and Enthalpy changes
[
23,
90,
91,
92].
Where K is Boltzmann's constant (1.381 × 10−23 J.K−1), and h is Planck's constant (6.626 × 10−34 J s).
The obtained positive values of entropy Changes
, enthalpy changes
, and negative values of
proved that all zones are non-spontaneous endothermic reactions [
93,
94,
95,
96,
97]. In zone 1, lower enthalpy values are observed in all the samples studied, which is due to physical adsorption due to hydrogen bonding and other complex reactions. The high enthalpy and energy values of zone 5 in the regenerated and spent catalysts imply that all the deposited coke on the catalysts in these temperature ranges has been removed and the remaining structural bonds of the catalyst require more energy to be broken.
According to this work, it was demonstrated that activation energies calculated from slope values gradually increase with the thermal treatment of the catalyst. This is because of the progressive degradation of the catalyst structure, such as the removal of components like Nitrogen, H
2O, and others in the respective temperature ranges. It seems that the deeper the catalyst is treated, the more the catalyst loses the hydroxyl bonds (OH) and other degradation products, and the more energy is required [
83].
As shown in
Table 4, all R
2 values are close to 1, indicating good agreement between the model and the experimental data used to estimate the kinetic parameters [
24]. The variation of activation energies in different zones is sometimes related to the H/C ratio in the sample [
98]. For example, the lower activation energy observed in zone 5 (134.56.10
2 Jmol
-1) of the spent catalyst compared to zone 4 (190.16.10
2 Jmol
-1) may relate to the temperature range of 700-850°C to remove the hard coke in the sample with a significant mass loss (1.78 wt.%)[
99].