Three proton-type ZSM-5 zeolites with different SARs were purchased from Nankai Catalyst Plant, and directly used without any treatments. The SARs of the three types of zeolites, determined by inductively coupled plasma optical emission spectrometry (ICP-OES; Perkin Elmer, Optima 5300DV), are 33, 266, and 487, respectively, and are denoted as ZSM-5-l, ZSM-5-m, and ZSM-5-h in the SAR order from low to high. The ZSM-5 additives were prepared as follows. The slurry comprising ZSM-5 (25 wt%), kaolin matrix (65 wt%), and alumina binder (10 wt%) with a liquid-to-solid mass ratio of 5 was thoroughly mixed and spray-dried to form microspheres with a typical particle size of 70 μm. The additives were identified as Z-l, Z-m, and Z-h, correspondingly. Commercial USY-based FCC catalyst LEO-1000 (denoted as USY), supplied by PetroChina Ltd., was employed as the base catalyst to evaluate the effect of SAR on the catalytic cracking performance. The hybrid catalysts, containing 2.5 wt% ZSM-5 zeolite, were obtained by physical mixing of the USY-based catalyst and respective ZSM-5 additive, and are denoted as USY/Z-l, USY/Z-m, and USY/Z-h, respectively.

X-ray diffraction (XRD) data were collected at ambient temperature in the 2θ range of 5°-80° with a step of 0.02° on a Rigaku D/max diffractometer with Cu K_α radiation and equipped with a graphite monochromator. The crystal morphology was studied by scanning electron microscopy (SEM; Hitachi S4800). IR spectra of the zeolites following adsorption of pyridine (Py-IR) were recorded on a Bruker TENSOR27. NH₃-TPD was performed on a Micromeritics AUTOCHEM II 2920.

The cracking of residue oil was performed at 500 °C for 60 s on a laboratory-designed fixed fluid bed unit. The hybrid catalysts were deactivated at 800 °C for 10 h in the presence of steam prior to the test. The catalyst loading and feedstock input were 200 and 50 g, respectively. The gaseous and liquid effluents were determined by gas chromatography (HP 6890). The paraffins (P), olefins (O), naphthenes (N), and aromatics (A) contents (PONA analysis) of cracked gasoline were analyzed by gas chromatography (Varian CP-3380). The amount of coke deposited on the catalyst was assessed by burning the sample in a carbon analyzer (DF 190). The properties of the residue oil are listed in Table 1.

Table 1 Properties of residue oil. Item MCR a (wt%) Mr b (g/mol) Viscosity c (mm2/s) Metal (μg/g) Saturate (wt%) Aromatic (wt%) Resin (wt%) C (wt%) H (wt%) N (wt%) Ca Cu Pb Ni V Fe Na Value 4.17 473 17.27 6.94 0.27 0.03 6.52 9.59 5.16 200 63.0 33.6 3.4 84.97 12.64 0.11 a Microcarbon residue; b Molecular weight; c Viscosity at 100 °C.

Table 1 Properties of residue oil.

Figure 1 shows the XRD patterns of the ZSM-5 zeolites with SARs of 33, 266, and 487 (ZSM-5-l, ZSM-5-m, and ZSM-5-h, respectively). All patterns were in good agreement with the orthorhombic phase of ZSM-5 zeolite. The relative crystallinity values of ZSM-5-l, ZSM-5-m, and ZSM-5-h were 91, 88, and 90 wt%, respectively, relative to the standard zeolite ZSM-5 reference.

Fig. 1. XRD patterns of the ZSM-5 zeolites with SARs of 33 (ZSM-5-l), 266 (ZSM-5-m), and 487 (ZSM-5-h).

Figure 2 shows the SEM images of the three ZSM-5 samples at different magnifications. The morphology of ZSM-5-m differed considerably from that of ZSM-5-l and ZSM-5-h. In contrast to the spherical features of the ZSM-5-m crystals, ZSM-5-h zeolite crystals, with a regular morphology, were primarily oriented along the b-axis ^{[10, 11, 12]}. Though the size of the crystals in ZSM-5-l varied, the morphology of the ZSM-5-l crystals was comparable with that of ZSM-5-h crystals.

Fig. 2. SEM images of the ZSM-5 zeolites at different magnifications. (a), (a’), ZSM-5-l; (b), (b’) ZSM-5-m; (c), (c’) ZSM-5-h.

Figure 3 displays the Py-IR spectra of the ZSM-5 zeolites. The peaks at 1452 and 1542 cm⁻¹ were ascribed to pyridine attached to Lewis and Brönsted acid sites, respectively ^{[13, 14]}. With increasing SARs, the density of both the Brönsted acid sites and Lewis acid sites declined markedly as observed in Fig. 3 and Table 2. ZSM-5-l, which featured a low SAR, displayed a higher content of Brönsted acid sites when compared with the other zeolite samples. Accordingly, it is expected that reactions, such as olefin cracking, that are catalyzed by Brönsted acid sites will occur predominantly.

Fig. 3. Py-IR spectra of the ZSM-5 zeolites with different SARs.

Table 2 Acidity of the ZSM-5 zeolites determined by Py-IR spectroscopy at 200 °C. Zeolite Brönsted sites (μmol/g) Lewis sites (μmol/g) Brönsted/Lewis ZSM-5-l 535.4 42.1 12.7 ZSM-5-m 193.6 25.5 7.6 ZSM-5-h 28.3 5.5 5.2

Table 2 Acidity of the ZSM-5 zeolites determined by Py-IR spectroscopy at 200 °C.

Figure 4 shows the NH₃-TPD profiles of the ZSM-5 zeolites with varying SARs. The ZSM-5 zeolites displayed two prominent desorption peaks at 232 and 442 °C, which were assigned to weak acid sites and strong acid sites, respectively ^[15]. An increase in SAR resulted in a marked decrease in the density of both acid sites, consistent with the Py-IR analysis of the zeolites.

Fig. 4. NH3-TPD profiles of the ZSM-5 zeolites with different SARs.

The fixed fluid bed data for residue cracking are given in Table 3. Under the given conditions, the USY-based catalyst achieved a gasoline yield of 49.60% and a LPG (sum of C_3s and C_4s) yield of 16.52%. Hybrid catalyst USY/Z-l achieved a significantly higher LPG yield (up to 22.44%) and a considerably lower C₅⁺ gasoline yield of 42.76%. With increasing SARs, LPG yields decreased gradually (19.98% for USY/Z-m and 19.10% for USY/Z-h), and the loss of gasoline was considerably inhibited, achieving 44.10% and 46.40% yields for USY/Z-m and USY/Z-h, respectively.

Table 3 Fixed fluid bed data for residue cracking over USY and hybrid catalysts. Item USY/Z-l USY/Z-m USY/Z-h USY Dry gas (wt%) 2.78 2.80 2.57 2.58 H2 0.09 0.12 0.10 0.09 H2S 0.15 0.16 0.16 0.16 CH4 1.05 1.14 1.06 1.06 C2H6 0.53 0.55 0.53 0.56 C2H4 0.96 0.83 0.72 0.71 LPG a (wt%) 22.44 19.98 19.10 16.52 C3H8 2.26 2.11 1.80 1.55 C3H6 6.80 5.65 5.53 4.42 n-C4H10 1.29 1.35 1.24 1.20 i-C4H10 8.00 7.16 6.68 5.91 C4H8 4.09 3.71 3.85 3.44 C5+ Gasoline b (wt%) 42.76 44.10 46.40 49.60 LCO c (wt%) 16.56 17.25 16.59 16.81 HCO d (wt%) 6.56 6.84 6.57 5.92 Coke (wt%) 8.12 8.27 8.00 7.79 Total (wt%) 99.22 99.24 99.23 99.22 Conversion (wt%) 76.11 75.14 76.07 76.50 a Liquid petroleum gas, combined total C3 and C4; b Cracked gasoline with the cut point (C5-200 °C); c Light cycle oil with the cut point (200-340 °C); d Heavy cycle oil above 340 °C.

Table 3 Fixed fluid bed data for residue cracking over USY and hybrid catalysts.

Furthermore, the introduction of ZSM-5 additives attenuated the conversion level of feedstock because the pores of ZSM-5 were only 5-6 Å wide, insufficiently large for big molecules to enter. Compared with the USY-based catalyst, a slight decrease in the conversion degree of residue oil was detected in the presence of the hybrid catalysts. Moreover, the introduction of ZSM-5 additives had little influence on the formation of dry gas and coke, although relatively high gas and coke selectivity were observed over hybrid catalyst USY/Z-l.

It is widely recognized that the olefins in gasoline have a much higher reactivity when compared with the paraffins generated over acidic ZSM-5 catalysts ^{[16, 17]}. Upon adsorption onto the acid sites of ZSM-5, especially on Brönsted acid sites, the olefins initially interact with the acid sites to form the intermediate products, carbenium ions. These intermediate products react subsequently to yield smaller olefin molecules, including propylene and butenes, via a β-session mechanism ^[17]. Consequently, the cracking of gasoline olefins declines with decreasing densities of Brönsted acid sites (increase in SAR), and more gasoline olefins are saturated by hydrogen atoms via intermolecular hydrogen transfer reactions. Hence, we may deduce that the significant changes in the catalytic properties of the three hybrid catalysts are mainly related to the change in the acidity of the ZSM-5 additives due to the change in the SAR.

Other factors, such as the particle size and morphology of ZSM-5 zeolite, may also influence the selectivity of the catalysts. It was previously reported that compared with the FCC catalyst containing typical ZSM-5 zeolite (average particle size of 5.48 μm), the catalyst containing small particles (average particle size of 1.99 μm) could lead to increases in the LPG and propylene yields by 0.41% and 0.08%, respectively ^[6]. By studying the effects of the morphology and the particle size of ZSM-5 on the catalytic performance for heptane, Baba and co-workers found that ZSM-5 nanosheets showed a slightly lower catalytic activity than particulate ZSM-5 (17.6% vs. 17.8%), and the distribution of the hydrocarbon products was independent of the morphology of ZSM-5 ^[9]. According to these literature results, we can conclude that differences in the particle size and morphology of ZSM-5 zeolites can also influence the selectivity of the catalysts to a certain extent, however, such differences would not entirely account for the significant changes in the catalytic performance of the hybrid catalysts.

To further understand the effect of SAR of the ZSM-5 zeolites on the selectivity of the hybrid catalysts, the SAR dependency of ∆C_ns/C_ns(USY) (n = 3, 4) of the hybrid catalysts is shown in Fig. 5. The high ∆C_3s/C_3s(USY) values suggested that ZSM-5 zeolite had a higher C_3s selectivity relative to C_4s, and the highest ∆C_3s/C_3s(USY) value of 50% was attained over the hybrid catalyst USY/Z-l. The ∆C_ns/C_ns(USY) values decreased with increasing SARs. The cracking of gasoline occurring in the pores of ZSM-5 zeolite was acid-catalyzed, accordingly, the decline in ∆C_ns/C_ns(USY) was mainly due to a decrease in the acid density of the ZSM-5 zeolites. The sharper decline in ∆C_3s/C_3s(USY) with increasing SARs, relative to ∆C_4s/C_4s(USY), may result from a decrease in the extent of cracking of C₆ olefins that is believed to have a higher C_3s selectivity over C_4s, and declines more drastically when compared with gasoline olefins with higher carbon numbers ^{[16, 17, 18]}. Additionally, other factors, such as the difference in the morphology of ZSM-5 crystals, may contribute to the change in ∆C_ns/C_ns(USY) because the morphology of ZSM-5 crystals has been reported to influence product selectivity ^{[19, 20]}.

Fig. 5. SAR dependence of ∆Cns/Cns(USY) (n = 3, 4) of the ZSM-5 zeolites. (C3s, combined propane and propylene, wt%; C4s, combined n-butane, i-butane and all butenes, wt%).

Previously, Madon proposed that ZSM-5 zeolites could catalyze both normal and branched olefin cracking to give propylene, butenes, 2-methyl 1-butene, and 2-methyl 2-butene ^[21]. In the present work, the following interesting phenomenon was noted: besides these light olefins, the yields of isobutane and isopentane increased markedly in the presence of ZSM-5 zeolite catalysts. The highest yields of isobutane and isopentane were obtained in the presence of hybrid catalyst USY/Z-l as shown in Table 3 and Fig. 6. According to the related literature reports ^{[16, 17, 18]}, the cracking of normal octene over Brönsted acid sites mainly involves the intermediate formation of carbenium C₈⁺. As shown in Scheme 1, carbenium C₈⁺ can undergo β-scission (cleavage of C-C bond at the β-position to the positively charged atom) to form smaller olefins and a smaller carbenium (not shown in Scheme 1), or react with other paraffin molecules via intermolecular hydrogen transfer. The smaller carbenium may further release a hydrogen cation to form another olefin, or react with paraffin in close proximity. Because of the narrow pore and low acid density of ZSM-5 zeolites, the carbenium C₈⁺ reacts to predominantly form smaller olefins, including propylene, butenes, and pentylenes. As the shift of the double bond in the olefins occurs easily at elevated temperatures, large amounts of isobutene and isopentene are obtained during the cracking reactions. When these olefins are subsequently adsorbed onto the USY-based catalyst with a high density of acid sites, they are saturated by hydrogen atoms through intermolecular hydrogen transfer to form isobutane and isopentane.

Fig. 6. SAR dependence of isobutane and isopentane yields of the ZSM-5 zeolites.

Scheme 1. Possible cracking routes of normal octene molecules over ZSM-5 zeolites.

The results of the PONA analysis and the octane number of cracked gasoline are shown in Table 4. The presence of ZSM-5 additives had pronounced effects on the gasoline composition, especially paraffins and aromatics. The reduction of paraffins appears to be inconsistent with the fact that the reactivity of olefins over ZSM-5 zeolites is much higher than that of paraffins ^[17]. To rationalize this finding, it is necessary to understand the interaction mechanism between the USY-based catalyst and ZSM-5 additives. Primary cracking of the feeds produces large quantities of gasoline olefins, which would be saturated via intermolecular hydrogen transfer over the USY-based catalyst. Upon introduction of ZSM-5 additives with low SAR (i.e., 33), the gasoline olefins with a high reactivity could be easily converted into light olefins, thus consequently leading to reduced yields of gasoline paraffins. With increasing SARs, the cracking of gasoline olefins over ZSM-5 additives declines, and more olefins are preserved and then saturated by hydrogen atoms over the USY-based catalyst. Furthermore, the extent of cracking of gasoline aromatics over the hybrid catalysts is relatively low, which could lead to increased concentrations of aromatics. The aromatization of hydrocarbon molecules over the hybrid catalysts may also contribute to the increase in the aromatics concentration. As a result, the motor octane number (MON) values of gasoline, which are mainly determined by the aromatics concentration, increased upon introduction of the ZSM-5 additives. These values decreased gradually with increasing SARs (Table 4).

Table 4 PONA analysis and octane number of cracked gasoline. Item USY/Z-l USY/Z-m USY/Z-h USY n-Paraffins (%) 2.82 2.88 3.09 3.45 Isoparaffins (%) 34.86 37.68 40.05 43.14 Cyclo-olefins (%) 2.02 1.88 1.78 1.52 Iso-olefins (%) 7.02 7.02 7.74 7.22 n-olefins (%) 3.12 2.98 3.59 3.38 Naphthenes (%) 10.89 10.03 10.11 10.00 Aromatics (%) 39.26 37.53 33.65 31.29 RON a 91.1 91.7 90.2 88.6 MON b 84.2 83.9 83.2 81.4 a Research octane number; b Motor octane number.

Table 4 PONA analysis and octane number of cracked gasoline.

Notably, the research octane number (RON), which is one of the most important indexes of gasoline, could be improved upon introduction of the ZSM-5 additives (Table 4). The use of ZSM-5 additives increased the RON value of gasoline from 88.6 (over the USY-based catalyst) to 91.1 (USY/Z-l), 91.7 (USY/Z-m), and 90.2 (USY/Z-h). However, as mentioned before, a substantial loss in gasoline was observed in the presence of hybrid catalyst USY/Z-l. In contrast, the loss of gasoline was considerably inhibited by increasing the SAR to 266 or 487. These results suggest that changing the SAR of ZSM-5 zeolites enables the formation of efficient hybrid FCC catalysts, which can improve the octane number of gasoline without substantial loss of gasoline. In our study, the enhancement in the RON value of gasoline was mainly attributed to the moderate aromatization and isomerization reactivity of the ZSM-5 additives that are mainly attributed to the relatively small pores and suitable acidic properties of the ZSM-5 zeolites with higher SARs.