Synthesis of Al-incorporated sulfated zirconia with improved and stabilized surface sulfur species for removal of trace olefins from aromatics

Abstract

Al-Incorporated sulfated zirconia with improved and stabilized surface sulfur species was synthesized via the co-precipitation method and used as an effective catalyst to remove trace olefins from aromatics in a fixed-bed reactor. The textural and chemical properties of various catalysts were characterized by XRD, ²⁷Al MAS NMR, N₂ adsorption, EA, XPS, H₂-TPR, FT-IR, and pyridine-infrared (IR) spectroscopy techniques to elucidate the Al promoting effect on acid centres and catalytic performance. The results demonstrated that incorporating Al into zirconia could effectively stabilize the tetragonal zirconia with small grain size and improve surface sulfur species, thereby balancing the distribution of Brønsted and Lewis acid sites. Consequently, sulfated zirconia incorporating Al at the level of Al/Zr < 7.5 mol% exhibited more outstanding catalytic activity and stability under a long reaction time compared with conventional sulfated zirconia. Notably, superior reusability with a negligible drop in olefins conversion over five reaction cycles was obtained in the presence of the Al_2.50-SZ. The alkylation reaction process of olefins with aromatics was further investigated by the GC-MS technique when Al_x-SZ was used as the catalyst.

---

1. Introduction

Aromatic hydrocarbons (benzene, toluene and xylene), produced primarily through catalytic reforming and cracking processes, are essential raw materials in petrochemical processes and refineries.¹ However, undesirable olefins as by-products are contained in aromatics owing to manufacturing process technology limitations. These olefins can efficiently react with other substances to form unwanted components because of their active chemical properties, which are destructive to the downstream technological processes and applications of aromatics. Therefore, it is indispensable to remove these impurities from aromatic products.

Clay treatment² is widely adopted in the petrochemical industry to remove olefins. This treatment is mainly based on the strong adsorption of the material toward olefins. Thus, limited by adsorption saturation, the lifetime of clay is quite short. Furthermore, deactivated clay cannot be regenerated due to its poor thermal stability, and it becomes hazardous solid waste and causes environmental pollution problems. Another feasible method is to catalyze the reaction between olefins and aromatics to form long-chain alkylbenzenes. Then, it can be removed by distillation. This reaction is essentially an alkylation reaction, as shown in Scheme 1, which can be effectively promoted by an acid reagent. Thus, clay modified by Lewis acids such as AlCl₃, ZnCl₂ was applied to this reaction and demonstrated excellent catalytic performance.² However, the thermal stability of the modified clay is not improved, which means that deactivated clay still inevitably brings environmental pollution.

Scheme 1 The alkylation reaction between aromatics and olefins.

For the sake of eliminating pollution problems to meet the increasing demands for environmental protection, it is urgent for us to develop environmentally friendly catalysts to replace clay materials. Consequently, various environmentally friendly catalysts, such as modified zeolites,¹ ionic liquids,³ organic–inorganic hybrid silica⁴ and sulfated zirconia,⁵ have been studied in removing trace olefins from aromatics by our previous work. An outstanding olefins conversion was obtained when treated by sulfated zirconia, and the deactivated catalyst could be regenerated by high-temperature calcination. Besides, due to their unique acid properties, environmental friendliness, as well as excellent renewability, sulfated zirconia has been frequently investigated and applied in isomerization,^6,7 alkylation,^5,8 acylation,⁹ condensation¹⁰ and esterification reactions,¹¹ which makes it a promising catalyst for industrial processes. Nevertheless, an unsatisfactory lifetime under the long reaction time is challenging for it available in the industrial process. Previous studies^9,12,13 have confirmed that the deactivation of sulfated zirconia derives from two aspects: (1) the rapid leach of surface sulfur species results in a decrease of acidity. (2) Coke is formed on the surface of the catalyst and eventually covers the active sites. Typically, deactivated sulfated zirconia caused by coking can be regenerated by high-temperature calcination, but the re-sulfonation cannot restore the lost surface sulfur species. The conventional sulfated zirconia,¹⁴ synthesized by post-sulfonation and calcination, has a low sulfate content and relatively weak interaction with surface sulfur species, which in turn results in limited acidity and rapid sulfur species leaching of the catalyst.

It has been found that changing the calcination temperature is a feasible approach to controlling sulfate loading on sulfated zirconia. A. K. Amin et al.¹⁵ reported that high sulfur content of sulfated zirconia could be accomplished by calcination at 400 °C, but another property, this temperature exerts adverse effects on the favourable tetragonal zirconia phase and crystallinity.^16,17 In our recent research,¹⁸ metallic oxides (Fe₂O₃, Co₂O₃, Ni₂O₃) were introduced to sulfated zirconia when it was synthesized by a solvent-free method; results showed that metallic oxides could effectively improve the amounts of S⁶⁺ and adjust active sites on the zirconia surface. Thus, improved catalytic activity was observed when the catalyst was applied to remove olefins from aromatics. In addition, many attempts have been reported to increase the sulfur content of sulfated zirconia and enhance interaction between sulfur species and zirconia by modified sulfated zirconia with Al, thereby improving the catalytic activity and stability of sulfated zirconia. J. H. Wang et al.¹⁹ proposed that the promoting effect of Al on sulfated zirconia in butane isomerization was attributed to increased sulfur retention. M. A. Coelho et al.²⁰ considered that incorporating small amounts of Al into zirconia before sulfation could increase the concentration of intermediate strength acid sites, which lead to enhanced activity and stability for n-butane isomerization. J. H. Wang et al.²¹ claimed that the balanced distribution of acid site strengths with an improved amount of weak Brønsted acid sites is related to the presence of Al in sulfated zirconia. C. L. Chen et al.²² found that the incorporated Al functions to stabilize the tetragonal zirconia phase due to the fact that phase transformation could be retard by Al. Most reports in this field have considered isomerization as a test reaction, but few studies have examined the catalytic performance of alkylation for removing trace olefins from aromatics. Furthermore, the detailed mechanism of acid site formation in the Al-incorporated sulfated zirconia is expected.

In our present study, sulfated zirconia with improved surface sulfur content and sulfur retention was synthesized by incorporating Al into the zirconia with a co-precipitation method. Various characterization techniques, such as XRD, N₂ adsorption, ²⁷Al MAS NMR, EA, XPS, FT-IR, pyridine-infrared (IR) spectroscopy and H₂-TPR, were employed to reveal the nature of the promoting effect on structure and acid centres by Al incorporation. The alkylation of olefins with aromatics catalyzed by the Al_x-SZ catalyst was verified by GC-MS analysis of the reaction products. Meanwhile, the catalytic activity, stability and reusability of Al-incorporated catalyst were systematically evaluated in removing trace olefins from aromatics and compared with conventional sulfated zirconia with no Al incorporation. As a result, it was found that the zirconia crystal structure and surface sulfur species were susceptible to Al incorporation, which in turn contributed to the stabilization of surface sulfur species, improvement of Brønsted acidity, as well as remarkable enhancement of catalytic performance.

2. Experimental section

2.1. Synthesis of catalysts

Al-Incorporated sulfated zirconia catalysts were synthesized by a precipitation–impregnation method. At the first step, ZrOCl₂·8H₂O and Al(NO₃)₃·9H₂O with different ratios were dissolved in deionized water and then added dropwise into (25%) NH₃·H₂O while stirring well until pH = 9.5. After precipitation, the hydroxide suspension was stirred for another 0.5 h and then aged at room temperature for 12 h. Subsequently, the obtained white colloidal solution was filtered, thoroughly washed with deionized water to remove chloride ions until no chloride ions were detected by the AgNO₃ test, and dried at 120 °C for 12 h. The obtained composite zirconia–alumina hydroxide was then immersed for 1 h in 2 M H₂SO₄ solution and a solid-to-liquid ratio of 1 : 5 was used (by weight). Following filtration without washing, the sample was dried at 120 °C for 24 h and then calcined at 650 °C for 3 h under dry air. The obtained catalysts were named as Al_x-SZ, X represented the molar percentage ratio of Al/Zr. For comparison, conventional sulfated zirconia, marked as SZ, was synthesized using the same route except for a mixture containing ZrOCl₂·8H₂O and NH₃·H₂O during precipitation. The deactivated catalyst was collected and regenerated at 550 °C for 3 h under airflow. The sample was labelled as SZ-RN/Al_x-SZ-RN, N referred to the time of regenerations.

2.2. Characterization of catalysts

2.2.1. X-ray diffraction. Powder X-ray diffraction (XRD) patterns were recorded on a Bruker D8 Advance polycrystalline diffractometer using Cu Kα radiation (λ = 0.15406 nm) within the 2-theta range of 5° to 80°. The crystalline size of the crystal phase could be calculated by Scherrer's equation: D = Kλ/β cos θ.

2.2.2. N₂ adsorption–desorption isotherm. The textural properties were obtained by nitrogen adsorption/desorption isotherms performed on a JW-ZQ200 instrument at −196 °C (Beijing JWGB Sci & Tech Co., Ltd.). Before adsorption measurement, samples were evacuated in situ in a vacuum for 3 h at 300 °C to remove any adsorbed impurities. The surface areas were calculated according to the Brunauer–Emmett–Teller (BET) formula using the adsorption data, ranging from P/P₀ = 0.05 to 0.30. The total pore volume (V_p) was worked out in P/P₀ of about 0.99. The pore size distribution curves and pore diameter (D_a) were determined by the Barrett–Joyner–Halenda (BJH) method.

2.2.3. Elemental analysis. The actual sulfur content of samples was measured on a VARIO ELIII elemental analyzer, and CHNS mode was utilized for detection.

2.2.4. XPS. In order to characterize the element distribution and valence state of the samples, X-ray photoelectron spectroscopy (XPS) measurements were performed using an ESCALAB 250Xi spectrometer equipped with a monochromatic Al Kα X-ray source (1486.6 eV). The scanning resolution was higher than 3 μm, and binding energy of 284.6 eV for the C1s level was used as an internal reference for the samples.

2.2.5. FT-IR spectra. Fourier transform infrared (FT-IR) spectra were collected on a Nicolet IS-10 infrared spectrometer at room temperature by mixing the sample with KBr in a mass ratio (1 : 100). It should be noted that before characterization, all samples and KBr powder were dehydrated in a vacuum oven at 400 °C for 12 h, and the dehydrated samples were pressed into self-supporting wafers (d = 10 mm) without allowing moisture absorption. The relative humidity of ambient conditions was controlled below 25% to suppress the influence of hydration on the characterization experiment. In addition, the accessible surface acid properties were also measured via FT-IR spectroscopy using pyridine as the probe molecule. The powder sample was pressed into a self-supporting wafer with a diameter of 13 mm, degassed in vacuum at 380 °C for 2 h, and then cooled down to 80 °C. After cooling down, the sample was exposed to pyridine vapour, immediately followed by an increase in temperature to desorb the excess pyridine. The spectra were scanned at 200 °C, 450 °C after aping vacuum for 15 min.

2.2.6. H₂-TPR. Temperature programmed reduction of H₂ (H₂-TPR) was used to investigate the interaction between the S component in the catalyst and the support, which was performed on an Auto Chem2910 (Micromeritics) instrument. Pretreat about 50 mg of the powder sample in the quartz tube for 1 hour at 120 °C and He atmosphere to remove water and impurities. Later, the sample was exposed to a reducing gas consisting of 5.0 vol% H₂ in argon, with the temperature rising from 50 °C to 800 °C at a heating rate of 10 °C min⁻¹.

2.2.7. ²⁷Al MAS NMR. ²⁷Al MAS NMR technique was employed to investigate the coordination environment of Al. The spectra were recorded on a Bruker MSL-300 spectrometer at a resonance frequency of 104.34 MHz.

2.2.8. GC-MS. The GC-MS technique was used to analyze the reaction products of aromatics and olefins, which helps us investigate the reaction process of olefin removal from aromatics. The GC-MS analysis was measured on Agilent PY/GC-MS/7890A-5975C [HP-5 MS column, 30 m × 250 μm i.d. × 0.25 μm].

2.3. Catalytic activity tests

The aromatics (Table 1) used in the experiments were provided by Sinopec Zhenhai Refining & Chemical Company, with a bromine index (BI) of about 1200 mg Br/(100 g). Bromine index is an indicator of olefin content in aromatics. Based on ASTM Standard D 2710-92, BI is defined as the number of milligrams of bromine consumed by 100 g of hydrocarbon sample.

Table 1 Components of aromatic materials

Component	Content (wt%)
Non-aromatics	2.53
Benzene	0.01
Toluene	0.04
Ethylbenzene	6.47
p-Xylene	8.75
m-Xylene	18.74
o-Xylene	12.26
C9	33.31
C10^+	17.89

---

The experimental apparatus for catalytic activity test is schematically shown in Fig. 1. The reaction proceeded in a fixed-bed tubular microreactor equipped with a double plunger pump to control the flow rate at 0.5 ml min⁻¹ and a heating system to maintain the temperature at 175 °C. The pressure was carried out at 1.0 MPa by the back pressure valve. The constant temperature zone of the microreactor was filled with dried catalyst (1.72 g, 40–60 mesh), and the liquid hourly space velocity (LHSV) was kept at 30 h⁻¹. Before the reaction, the system was flushed with N₂ for 1 h at the reaction temperature to remove physically adsorbed water from the catalyst and reactor. During the reaction, the reactants and products were collected every hour and analyzed by a bromine index analyzer (LC-6). The conversion of olefins was calculated following the equation below:

BI₀ represents the bromine index of the feedstock, and BI_n refers to the hourly bromine index of the product.

Fig. 1 Schematic diagram of experimental apparatus: 1 – feedstock storage tank, 2 – double plunger pump, 3 – globe valve, 4 – fixed-bed tubular microreactor, 5 – heating furnace, 6 – thermocouple, 7 – pressure gauge, 8 – pressure control valve, 9 – product storage tank, 10 – nitrogen cylinder.

3. Results and discussion

3.1. Structure and composition analysis

The structure and surface properties of as-synthesized catalysts with different ratios of Al/Zr were characterized to elucidate the effect of Al incorporation. Fig. 2 contains the XRD patterns of SZ and Al_x-SZ. These catalysts exhibited diffraction peaks of tetragonal zirconia phase (T-ZrO₂) at around 2θ = 30.2°, 35.3°, 50.3°, 60.2° and monoclinic zirconia phase (M-ZrO₂) at around 2θ = 24.5°, 28.2°, 31.4°. It could be ascertained that all the peaks in Fig. 2 are attributed to T-ZrO₂ or M-ZrO₂. No peaks associated with Al/Al₂O₃/Al₂(SO₄)₃ were observed, demonstrating that the Al was well-dispersed into the ZrO₂. Interestingly, it was observed that the amount of M-ZrO₂ decreased dramatically as aluminium content increased, while the amount of T-ZrO₂ increased and became predominant. This result was consistent with the previous observations that the transformation from the metastable tetragonal phase to the monoclinic phase could be retarded by introducing Al₂O₃ into the ZrO₂.^7,23–25 D. J. Zalewski et al.²⁶ proposed that aluminium incorporated by the co-precipitation method is mainly concentrated in the lattice gap of zirconia and prevents the overgrowth of zirconia particles, thereby zirconia with a small crystallite size was formed. Meanwhile, the small crystallite size of zirconia is beneficial to stabilize metastable T-ZrO₂ during calcination.²⁷Table 2 provides the crystal phase data and the sulfur content of SZ and Al_x-SZ. It was worth noting that the crystallite size of samples obviously decreased with the aluminium incorporation increased, and then more T-ZrO₂ was produced, as shown in Table 2. The previous investigation indicated that the tetragonal zirconia phase is essential for sulfated zirconia appearing super acid and that the stable tetragonal zirconia phase is an indispensable structural condition for catalytic activity.^28–30

Fig. 2 XRD patterns and crystal structure of SZ and Al_x-SZ catalysts with different Al/Zr ratios.

Table 2 The crystal phase data and sulfur content of SZ and Al_x-SZ

Sample	M-ZrO2 (%)	T-ZrO2 (%)	Relative crystallinity (%)	Crystallite size (Å)	Sulfur content (wt%)
SZ	58.6	41.4	87	146	2.7
Al1.25-SZ	41.2	58.8	78	124	2.8
Al2.50-SZ	29.7	70.3	68	117	2.9
Al7.50-SZ	24.9	75.1	64	108	4.8
Al12.5-SZ	23.8	76.2	51	88	5.6
Al50.0-SZ	6.3	93.7	26	67	—

---

Based on the elemental analysis results in Table 2, the order of sulfur content was Al_12.5-SZ > Al_7.50-SZ > Al_2.50-SZ > Al_1.25-SZ > SZ. Interestingly, this order was consistent with the order of the proportion of T-ZrO₂ in catalysts, indicating that the tetragonal zirconia phase was more active and more accessible to react with sulfur species than the monoclinic zirconia phase. P. Z. Wang et al.³¹ also found that the sulfur loading of sulfated zirconia was highly correlated with the content of T-ZrO₂ and DFT calculation results confirmed the higher adsorption energies of SO₃ (−356.5 kJ mol⁻¹) on the surface of T-ZrO₂. It seems that surface sulfur species of the tetragonal zirconia phase is thermal stability favourable. Thus, Al-incorporated catalysts with more T-ZrO₂ containing could better retain surface sulfur species after high-temperature calcination.

Moreover, the incorporated aluminium also significantly affected the relative crystallinity of zirconia when it played a role in retarding the crystal phase transformation. As shown in Table 2, a gradual decrease of zirconia crystallinity appeared from no aluminium to Al/Zr = 50.0%, consistent with the decline of the intensity of the diffraction peaks in the XRD patterns. Significantly, an extremely low crystallinity of the catalyst with high aluminium content (Al/Zr = 50.0%) was observed. Here, to understand this effect caused by incorporated aluminium, ²⁷Al MAS NMR technique was used to analyze the coordination state of aluminium in Al_2.50-SZ, Al_50.0-SZ catalysts and S/Al₂O₃. The recorded spectra are shown in Fig. 3. The sample labelled S/Al₂O₃ was obtained by impregnating Al(OH)₃ with 2 M H₂SO₄ solution and then calcined at 650 °C for 3 h. Compared with S/Al₂O₃, the coordination state of aluminium in Al-incorporated catalysts had changed significantly, particularly only an intense resonance peak was observed at around 0.1 ppm on the spectra of Al_2.50-SZ. As reported by W. M. Hua et al.,³² this Al resonance peak (0.1 ppm) is considered Al³⁺ species in octahedral coordination; therefore, the incorporated aluminium could be introduced into the lattice of the zirconia, forming isomorphs or solid solution with zirconia.^32,33 It confirmed that the Zr–O–Al bonds could be formed during calcination, which would inevitably restrict the formation of Zr–O–Zr bonds. On the other hand, for Al_50.0-SZ with high aluminium content, the aluminium exhibited other different coordination states in addition to octahedral coordination, which was similar to S/Al₂O₃. The peaks at 40 and 70 ppm correspond to the pentacoordinate and tetrahedral Al species in γ-Al₂O₃ crystal structure, respectively.^34,35 Thus, an understanding might be that with the incorporation of excessive aluminium, some of the aluminium would exist on its own besides being present in the lattice of zirconia. These separate aluminium would act as a barrier between the Zr(OH)₄ crystallites and form γ-Al₂O₃ under 650 °C calcination, thereby preventing Zr(OH)₄ from crystallization and growth.³⁶ As crystallinity of zirconia is also a factor indirectly affecting the activity of the catalyst,³⁰ it could assume that not high aluminium content but the appropriate amount of aluminium incorporated might promote an enhancement in catalytic activity.

Fig. 3 ²⁷Al MAS NMR spectra of Al_2.50-SZ and Al_50.0-SZ catalyst. S/Al₂O₃ sample for comparison.

The N₂ adsorption–desorption isotherms and BJH pore size distribution curves of prepared catalysts were displayed in Fig. 4a and b. Obviously, the isotherms of all catalysts could be classified as the type IV isotherm accompanied by the type H2 hysteresis loop, which is assigned to the typical characteristic of mesoporous material. The accumulation of nanocrystals led to the formation of intercrystalline voids, thereby the appearance of mesopores in the catalyst.^36,37 Furthermore, the hysteresis loop was caused by capillary condensation in these mesopores.³⁸Table 3 lists the physical properties, including the surface area, pore volume and pore diameter, of SZ and Al_x-SZ with different aluminium content. It could be seen from data that in a certain range of aluminium incorporating, BET surface areas of samples increased with aluminium content. Especially Al_2.50-SZ had a higher BET surface area of 130.0 m² g⁻¹ as compared to conventional SZ. However, Al_12.5-SZ appeared a noticeable drop in BET surface area (107.7 m² g⁻¹). The BET surface of Al_50.0-SZ was only 65.7 m² g⁻¹, indicating that aluminium incorporated negatively impacted the structure of catalysts when aluminium containing was at a high level.

Fig. 4 (a) The N₂ adsorption–desorption isotherms and (b) pore size distribution plots of SZ and Al_x-SZ.

Table 3 Textural properties of SZ and Al_x-SZ

Sample	S BET (m^2 g^−1)	Pore volume^b (cm^3 g^−1)	Pore size^c (nm)
a S BET = multi-point BET total specific surface area was obtained from adsorption data. b Pore volume = pore volume was obtained at P/P0 = 0.99. c Pore size = adsorption average pore diameters measured using Barrett–Joyner–Halenda (BJH) method.
SZ	124.8	0.201	4.9
Al1.25-SZ	125.8	0.195	5.1
Al2.50-SZ	130.0	0.192	5.2
Al7.50-SZ	124.9	0.200	5.1
Al12.5-SZ	107.7	0.147	4.5
Al50.0-SZ	65.7	0.085	4.5

---

Catalysts modified with small amounts of aluminium appeared almost the same pore volume as SZ (0.201 cm³ g⁻¹), while the pore volume of Al_12.5-SZ (0.147 cm³ g⁻¹) and Al_50.0-SZ (0.085 cm³ g⁻¹) was remarkably smaller than that of SZ. It supported the assumption that the pores of catalysts could be blocked by nanocrystals of γ-Al₂O₃, formed by excessive aluminium during calcination, which was consistent with the analysis of ²⁷Al MAS NMR. The average pore size and uniform pore distribution of Al_x-SZ measured by the BJH method were similar to those of SZ, as shown in Fig. 4b and Table 3.

Generally, a high specific surface area is likely conducive to the catalytic reaction on account of more reactants can be adsorbed on active sites. The above results obviously clarified the influence of aluminium incorporation on the textural properties of sulfated zirconia. Specific surface areas could be improved even with a slight aluminium incorporated. On the contrary, excessive aluminium blocked the pores of catalysts and reduced the specific surface area.

3.2. Acidity and active sites analysis

3.2.1. FT-IR spectra analysis. Fig. 5 provides the FT-IR spectra of SZ and Al_x-SZ, obtained at room temperature. A broad peak at around 1635 cm⁻¹ is attributed to the δ_O–H bending frequency of water associated with the surface sulfur groups.^39,40 It could be noticed that the deformation and vibration of the water were weak. The previous study has clarified that approximately 90% of the water adsorbed on sulfated zirconia will be desorbed by activation at 350 °C.³¹ Thus, these samples could be regarded as highly dehydrated due to the pretreatment of catalysts at 400 °C and the careful control over the amount of hydration in the environment, which allowed us to focus on the effect of aluminium modification on surface sulfur groups. The FT-IR spectra of all samples appear the peaks at 994, 1042, 1145 and 1227 cm⁻¹, which are the characteristics of bidentate sulfur species coordinated to zirconia.^25,41–43 The peak at 994 cm⁻¹ is assigned to the typical band for symmetric stretching vibrations of S–O and others are attributed to partially ionized S=O asymmetric stretching vibrations of sulfate groups.^44–47 From the data in Fig. 5, it was apparent that the stretching vibrations of S–O and S=O were enhanced after aluminium incorporation, indicating that more sulfur species onto zirconia was exposed. The result was consistent with the sulfur content given in Table 2. Furthermore, the aluminium incorporated into the zirconia weakened the sulfate absorption at 1042 cm⁻¹, accompanied by the emergence of a new band at 1075 cm⁻¹. Previous studies^45,46 have pointed out that the blueshifting of the bands (1042 cm⁻¹ to 1075 cm⁻¹) suggests the formation of strong bonds between surface sulfur species and zirconia. The split peak at 1075 cm⁻¹ could also be assigned to the typical characteristic of partially ionized S=O bonds in partially ionic nature sulfate species, which plays a crucial role in forming the Brønsted acid site in sulfated zirconia. Similarly, V. Vishwanathan et al.⁴⁸ suggested that the occurrence of Brønsted and Lewis acid sites is significantly associated with the stretching vibration of asymmetric S=O at the higher frequency. The peak at around 1390 cm⁻¹, detected in all samples, can be assigned to the asymmetric stretching band of covalent S=O. The covalent sulfate groups are associated with the Lewis acid sites in sulfated zirconia.^46,49,50 Consequently, acid sites might be affected by incorporating aluminium into the sulfated zirconia catalyst.

Fig. 5 FT-IR spectra of SZ and Al_x-SZ in the ranges from 2000 to 800 cm⁻¹.

3.2.2. XPS analysis. The XPS spectra investigated in the region of the S 2p, Zr 3d and Al 2p of samples, including SZ, Al_2.50-SZ and Al_12.5-SZ, were displayed in Fig. 6. The binding energy of 182.2 eV and 184.6 eV are assigned to Zr 3d doublet in zirconia.^51,52 Nevertheless, it was not hard to notice that the peak locations of Zr 3d exhibited by the three samples all shifted to high binding energy, which was probably since the surface Zr atoms were affected by surface sulfur species with strong electron attraction.^10,53 The spectrum of Al 2p in samples other than SZ could be detected at the binding energy of 74.9 eV, assigned to the aluminium oxidation state of Al³⁺, indicating that surfaces of these Al-incorporated catalysts were abundant in Al. The enrichment of Al corresponded to Zr–O–Al bonds formed by the coordination of Al atoms, present in the lattice of zirconia, with Zr atoms. According to literature,^7,25,32,42 the formation of Zr–O–Al bonds could contribute to accommodating sulfur species and producing more acid sites. Additionally, the spectra corresponding to S 2p for three samples were similar, and each spectrum could be fitted by two peaks. The peaks detected with a binding energy of around 169.0 eV and 170.0 eV are assigned to S⁶⁺.⁵⁴ In this work, sulfur species with two different binding energies were labelled as S^6+,L (sulfur species with low binding energy) and S^6+,H (sulfur species with high binding energy), respectively. As N. W. Liu et al.¹⁸ pointed out, the alkylation activity of sulfated zirconia synthesized by a solvent-free method is highly correlated with the proportion of sulfur species with high binding energy. H. G. Wang et al.⁵⁴ found that the content of sulfur species with higher binding energy in the deactivated catalyst is less than that in the fresh one. These researchers all confirmed that sulfur species with high binding energy play a crucial role in sulfated zirconia catalysts; thereby, S^6+,H instead of total sulfur species (S^T) should be paid more attention.

Fig. 6 X-ray photoelectron spectroscopy (XPS) spectra of (a) SZ, (b) Al_2.50-SZ and (c) Al_12.5-SZ.

The peak area in XPS spectra has a linear relationship with sulfur content when the mass of the sample is the same. The order of S 2p peak area of three catalysts was SZ < Al_2.50-SZ < Al_12.5-SZ, consistent with the analysis result of sulfur content in Table 2. Consequently, a fitted peak area can be regarded as an indicator of absolute content value. In order to clarify the influence of aluminium incorporation on S^6+,H, the ratio of S^6+,H/S^T was calculated and shown in Fig. 6. Notably, the proportion of sulfur species attributed to S^6+,H in Al-incorporated catalysts was evidently higher than that in the aluminium-free catalyst. The two different binding energy of sulfur species indicated that the sulfates were in different chemical environments. Equally, the difference in FT-IR spectra above also confirmed the existence of varying sulfur species.

Based on the preceding analysis, an envisioned structural relation between sulfur species and zirconia was proposed, as displayed in Scheme 2. The model reveals that the electron shift of S atoms in sulfate groups would occur since the electron withdrawing effect of hydroxyl group adjacent to the sulfate group, resulting in the appearance of ionic nature sulfates.^54–56 The sulfur species of ionic nature sulfates were assigned to S^6+,H. Moreover, the sulfate groups without being affected by hydroxyl groups were anchored to the surface of zirconia as covalent sulfate groups, and they could be considered S^6+,L. On the other hand, the sulfur species in Al-incorporated sulfated zirconia would not only be anchored to Zr sites but also to Al sites, because of the dehydration condensation between –Al–OH and HSO₄⁻ during calcination would lead to the formation of Al–O–S.³³ Guided by the principle of electronegativity equalization, as proposed by Sanderson, it could be known that the electronegativity of Al³⁺ (1.61 eV) is higher than that of Zr⁴⁺ (1.33 eV). Thus, the positively charged electricity on S atoms in Al–O–S bonds could be enhanced because the strong electron-withdrawing of Al³⁺ decreased the charges transferred from Al atoms to S atoms. As a result, the surface sulfur species anchored to Al sites were in a state of high binding energy.

Scheme 2 The surface structural models of a) sulfated zirconia and b) Al-incorporated sulfated zirconia.

3.2.3. FT-IR spectra of adsorbed pyridine analysis. The acid site is a vital factor affecting the reaction activity of the catalyst. Therefore, pyridine-FTIR spectroscopy was used to characterize Brønsted acid sites and Lewis acid sites. Fig. 7 provides the FT-IR spectra of SZ and Al_x-SZ after pyridine adsorption and evacuation. Generally, the spectra recorded at 200 °C were related to the total acidity, while after evacuation at 450 °C the obtained spectra were attributed to strong acidity. The band observed at approximately 1445 cm⁻¹ is assigned to the C–C vibration of pyridine adsorbed at the Lewis acid sites. The characteristic band at approximately 1540 cm⁻¹ is due to the pyridine interacting with Brønsted acid sites.⁵⁷ Moreover, another strong adsorption band appeared at around 1490 cm⁻¹ owing to the contributions of Lewis and Brønsted acid sites.⁵⁸ As demonstrated in Fig. 7a, various samples were enriched in Lewis and Brønsted acid sites after the evacuation of pyridine at 200 °C. What made a difference was that the band attributable to Lewis acid sites gradually shrank along with the increasing aluminium incorporation. In contrast, the band corresponding to Brønsted acid sites steadily enhanced. In addition, the spectra of various samples recorded after heating at 450 °C are shown in Fig. 7b. Only a narrow band could be observed around 1445 cm⁻¹, indicating that they generate slight strong Lewis acid and almost no strong Brønsted acid. In order to quantitatively analyze the acid types, acid strengths, the results, summarized in Table 4, were obtained by utilizing the following empirical equation guided by Lambert–Beer law.⁵⁹

C(pyridine on L site) = 1.42 × IA(L)R²/W

C(pyridine on B site) = 1.88 × IA(B)R²/W

where C is the concentration (μmol g⁻¹ catalyst); IA (L or B) is the peak area of Lewis or Brønsted acid (cm⁻¹); R is the radius of the wafer (cm); W is the weight of the wafer (g).

Fig. 7 FT-IR spectra of SZ and Al_x-SZ after pyridine adsorption at 80 °C and desorption at (a) 200 °C; and (b) 450 °C.

Table 4 The amount and strength of Lewis and Brønsted acid sites on SZ and Al_x-SZ

Sample	T	TL	TB	WL	WB	SL	SB	Sulfate concentration (S atoms per nm^2)
T, total acidity, evacuation at 200 °C; S, strong acidity, evacuation at 450 °C; W, weak acidity; B, Brønsted acidity; L, Lewis acidity. Units = μmol g^−1.
SZ	107	81	26	77	26	4	—	4.0
Al1.25-SZ	92	52	40	51	40	1	—	4.1
Al2.50-SZ	90	45	45	42	45	3	—	4.2
Al7.50-SZ	88	39	48	39	48	—	—	7.2
Al12.5-SZ	80	31	49	30	49	1	—	9.8

---

Table 4 shows that a trace amount of Al incorporation (Al/Zr = 1.25%) caused a 14% reduction in the total acidity (decreased from 107 μmol g⁻¹ to 92 μmol g⁻¹). Fortunately, this unexpected sharp decline did not continue with the doubling of Al content but became smooth. Further analysis on Lewis and Brønsted acidity was noteworthy that Brønsted acidity increased with the ratio of Al/Zr. It demonstrated that aluminium incorporation was conductive to generate more Brønsted acid sites for sulfated zirconia. However, Lewis acidity appeared the opposite trend. The following statement could better explain the changes in the acid type and acidity. The improvement of acid sites was associated with sufficient sulfate species loading on sulfated zirconia. Data have testified in Fig. 2 and Table 2 that Al promoted the growth of tetragonal zirconia with small grain size, resulting in more sulfur species being well retained on the surface of catalysts. Meanwhile, the incorporated Al could enhance the partially ionic nature sulfur species, which could be responsible for promoting Brønsted acid.^44,45,60,61 Moreover, as reported in the literature, Lewis sites are dominant in low-sulfate-content sulfated zirconia catalysts³⁶ and the surface concentration of the sulfate largely determines the amounts of Brønsted acid sites.⁶² The Al incorporated also contributed to improving the surface sulfate concentration of the catalysis, as displayed in Table 4. Thus, the Brønsted acid sites were enhanced (Fig. S1†), while accumulated sulfur species might cover part of the Lewis acid sites. On the other hand, the strong one in various samples was insignificant compared to the total acidity, leading to the regular of the weak acidity obtained by subtraction consistent with that of total acidity.

3.3. Catalytic activity and reusability for alkylation

According to our previous study,^{5,18,59,63,64} during the SZ-catalyzed alkylation of olefins and aromatics, the continuous decrease with the reaction time of olefin conversion was observed, demonstrating the deactivation of the SZ catalyst. In this study, to better understand the promoting effect of aluminium incorporation on the catalytic activity and stability of sulfated zirconia, SZ and Al_x-SZ catalysts with different ratios of Al/Zr were evaluated in alkylation for the removal of trace olefins from aromatics. In order to accelerate the evaluation of the stability of a catalyst, the liquid hourly space velocity in our test was amplified to 30 h⁻¹ while it was generally 1 h⁻¹ adopted by the industrial process. The results of the continuous 8 hours of reaction were displayed in Fig. 8. It was worth mentioning that the catalytic activity and stability were enhanced due to the Al incorporation (the molar ratio of Al/Zr = 1.25 to 7.50%) and then was dropped at a higher Al content. The poor catalytic performance of the Al_2.50-Z without being sulfated by H₂SO₄ solution was obtained, indicating that the modification of sulfate groups is essential for the excellent catalytic activity of the catalysts. The SZ catalyst performed excellent catalytic activity within the first 2 hours of the reaction. However, it was disappointing that the activity decline trend continued from the 3rd hour to the end of the reaction. The conversion of olefin at the 8th hour was only 64.67%. Simultaneously, what was striking about the data in Fig. 8 was that Al_1.25-SZ, Al_2.50-SZ and Al_7.5-SZ all performed higher activity than sulfated zirconia. The final olefins conversion of the Al_2.50-SZ catalyst was improved to 84.10%. On the contrary, the Al_12.5-SZ catalyst exhibited rapid deactivation. It could hence claim that the low-aluminium-incorporated sulfated zirconia has outstanding performance concerning their catalytic activity and stability, while the high aluminium content is detrimental to the lifetime of the catalyst.

Fig. 8 Hourly conversion of olefins from aromatics upon SZ, Al_x-SZ and Al_2.50-Z.

In order to validate the alkylation process for removing olefins from aromatics using Al_x-SZ as the catalyst, the model oil was used as feedstock, and the reaction was carried out under the same conditions as the catalytic activity test. The model oil was prepared by adding a certain amount of p-xylene (19 wt%) and 1-octene (1 wt%). The GC-MS technique was applied to analyze the reacted model oil as well as fresh oil, and the results were presented in Fig. 9. It could be seen from the graph above that the new peaks appeared at 16.4, 16.5 and 16.8 min. These new components detected in the reacted oil were identified to be reaction products. On the other hand, the mass spectrometry analysis (Fig. S2†) confirmed that these products were C₁₆H₂₆ (m/z = 218) formed by the reaction of 1-octene (C₈H₁₆) with p-xylene (C₈H₁₀), and the structures of the reaction products were shown in Fig. 9, respectively. No dimerization products were found; therefore, the reaction using Al_x-SZ as the catalyst in removing trace olefins in aromatics belongs to the Friedel–Crafts alkylation reaction.

Fig. 9 The GC-MS analysis of fresh model oil and reacted model oil.

In addition, different commercial solid acid catalysts, such as zeolite (USY), activated clay and modified clay (ROC), were also evaluated in removing trace olefins from aromatics at the same reaction conditions. The catalytic performances (Fig. S3†) demonstrated that Al-incorporated sulfated zirconia (Al/Zr < 7.50 mol%) appeared more excellent catalytic activities over the whole 8 hours compared with these commercial solid acid catalysts. Moreover, the lifetime of Al-incorporated sulfated zirconia was significantly extended, and the industrial application is promising.

Interestingly, the catalytic activity of as-synthesized catalysts was not positively correlated with the surface sulfur loading. It could be considered that the surface sulfur loading and sulfur species significantly affect the acid sites according to the analysis of acid properties in section 3.2.2, which, in turn, determined the catalytic performance. Thus, to gain a better understanding of the correlation between the acid type, acidity, and activity of the catalyst, we plotted relevant data in Fig. 10. It was noteworthy that whether initial reaction activity or final reaction activity of the catalyst was considered, the correlations between reactivity and acidity were similar. Even these catalysts with different acidity performed relatively high catalytic activity under the initial reaction condition. Still, as the reaction continued, the negative influences of inadequate acidity and plentiful acidity in the catalyst on catalytic activity were highlighted. It can also be inferred from Fig. 10 that both Lewis acid and Brønsted acid contributed to catalytic activity, which was consistent with prior studies^1,4,59,65 that both Lewis acid and Brønsted acid play a vital role in the alkylation.

Fig. 10 (a) The correlation between total Lewis acidity and catalytic activity. (b) Correlation between total Brønsted acidity and catalytic activity.

Guided by GC-MS analysis and the mechanism proposed by S. Narayanan et al.^66,67 for aniline alkylation reaction, the mechanism of olefin alkylation on both Lewis and Brønsted acid sites could be considered to follow the process shown in Scheme 3. Firstly, olefins are adsorbed on Brønsted acid sites to form stable carbocations, while benzene is adsorbed on Lewis acid sites due to its strong electron donating properties. Then, the carbocation reacted with benzene, adsorbed on Lewis acid sites to accomplish the alkylation reaction; meanwhile, benzene could be desorbed to release Lewis acid sites. Moreover, the proton, lost by benzene adsorbed on Lewis acid sites to balance the positive charge, migrate to the Brønsted acid sites to recover the hydroxyl group. Finally, Lewis acid and Brønsted acid sites are both recovered and continue with the next cycle. However, when the acidity in catalysts is excessive, especially for Brønsted acid, the product adsorbed on the acid site is challenging to desorb, resulting in final coking. Overall, the appropriate acidity in the catalyst was also a factor that allowed sulfated zirconia with excellent activity and high stability when applied in the alkylation of aromatics with olefins, while incorporating Al into zirconia properly contributed to balancing the distribution of Brønsted and Lewis acid sites of the catalysts.

Scheme 3 The mechanism of olefins alkylation on both Lewis and Brønsted acid sites (benzene as a representative for aromatics).

Apart from catalytic activity and stability, reusability is also a vital factor affecting the performance of the catalyst in the actual process of removing trace olefins from aromatics. In our present study, the reusability of catalysts was estimated by successive reaction cycles, and Al_2.50-SZ and SZ were selected for comparison. After each catalytic reaction, the catalyst was separated, dried at 110 °C overnight and regenerated at 550 °C in air for 3 h. Then the reaction was performed under the same reaction conditions as before. As seen in Fig. 11, it was evident that the activity of Al_2.50-SZ was almost fully recovered during the first four runs and only a 7.90% decline of the final conversion after five runs. Instead, the activity of SZ could not entirely be recovered after regenerating only once. Although the reactivity recovered to 95% ∼ 98% of the fresh SZ catalyst within four hours, the conversion of olefin plunged to 79.3% at the fifth hour. The continuous effect of deactivation finally resulted in the conversion at the eighth hour was only 51.4%, equivalent to 79.4% of the reactivity of fresh SZ under the equal reaction conditions. A sample calcination treatment can re-expose the active sites covered by the coke deposit to recover the catalytic activity, but this treatment cannot compensate for the loss of surface sulfate species. The activity results showed that Al_2.50-SZ performed more excellent stability and reusability than SZ, which was likely caused by Al_2.50-SZ losing fewer sulfur species during the reaction and regeneration.

Fig. 11 Hourly conversion of olefins from aromatics upon Al_2.50-SZ-RX and SZ-RX.

3.4. Stability of surface sulfur species analysis

SZ-R2 (regenerate) and Al_2.50-SZ-R5 (regenerate) were characterized by elemental analysis and pyridine-FTIR spectroscopy (FT-IR spectra are given in Fig. S4†). As Table 5 shows, the sulfur content of SZ-R2 decreased by 44.4%, while that of Al_2.50-SZ only dropped by 27.6% after regenerated five times, which was consistent with the conjecture that Al_2.50-SZ lost fewer sulfur species. In the meanwhile, both Lewis acidity and Brønsted acidity in the two samples were changed. For SZ-R2, the plummet of Lewis acidity could be observed; thereby, the total acidity was reduced considerably. It was undoubtedly related to the disappearance of active sites caused by the loss of sulfur species, a typical characteristic of conventional sulfated zirconia deactivation. Additionally, the total acidity of Al_2.50-SZ-R5 was slightly decreased, and the transformation of the Brønsted acid site to Lewis acid site could be observed from Table 5. This transformation may be explained by the fact that the drop of surface sulfur loading, which in turn decreased the surface sulfur concentration of Al-incorporated sulfated zirconia and Lewis sites, previously covered by excess sulfur species, were re-exposed.

Table 5 The amount and strength of Lewis and Brønsted acid sites on Al_2.50-SZ-R5 (regenerate) and SZ-R2 (regenerate)

Sample	T	TL	TB	WL	WB	SL	SB	Sulfur (wt%)
T, total acidity, evacuation at 200 °C; S, strong acidity, evacuation at 450 °C; W, weak acidity; B, Brønsted acidity; L, Lewis acidity. Units = μmol g^−1.
SZ	107	81	26	77	26	4	—	2.7
SZ-R2	77	52	25	48	25	4	—	1.5
Al2.50-SZ	90	45	45	42	45	3	—	2.9
Al2.50-SZ-R5	85	52	33	50	33	2	—	2.1

---

Fig. 12 presents the characterization data for H₂-TPR. It could be found that a sharp peak was observed in the patterns of various samples, which was attributed to the reduction of surface sulfur species.⁶⁸ The most significant aspect of this figure was that the peaks of the Al-incorporated catalysts shifted to high reduction temperature and exhibited larger areas than that of SZ. It was likely that the Al–O–S bond had higher bond energy, making it hard for the catalyst to release sulfur species. Under the previous analysis of XPS, aluminium incorporation allowed the catalyst to gain more sulfur with high binding energy, which was assigned to the sulfur that interacted with the Al–O bond. Therefore, the Al–O–S bond dominated a stronger “anchor” role to sulfur species than the Zr–O–S bond. On the other hand, as mentioned in section 3.1, sulfur species on the surface of T-ZrO₂ exhibited superior thermal stability, while aluminium incorporation could enhance the crystallization of T-ZrO₂. Taken together, it seemed that aluminium incorporation could improve the strength and stability of surface sulfate species. And this promotion resulted in the catalyst performing outstanding stability and reusability.

Fig. 12 H₂-TPR patterns of SZ and Al_x-SZ in the range from 50 to 800 °C.

4. Conclusions

Al-Incorporated sulfated zirconia was successfully synthesized by incorporating Al into the zirconia with a co-precipitation method and evaluated in alkylation for removing trace olefins from aromatics. Incorporation of Al at the level of Al/Zr <7.5 mol% enhanced catalytic activity and reusability, attributed to the superior acid site loading and increased sulfur contents retained in ZrO₂, while higher Al concentrations reduced the specific surface area and weakened Lewis acid sites, resulting in lower activity and lifetime. In particular, the most outstanding catalytic activity was obtained by Al_2.50-SZ, which also performed a negligible drop in olefins conversion over five reaction cycles when it came to the reusability of catalysts. According to the characterization results, it was ascertained that the structural and surface sulfate species of the catalyst were significantly affected by the presence of Al. Incorporating small amounts of Al into zirconia can effectively promote forming abundant Zr–O–Al bonds and stabilising the tetragonal zirconia with small grain size, both of which contribute to improving and stabilizing surface sulfur species. Furthermore, Al promoters play a significant role in increasing the surface sulfur concentration and enhancing the partially ionic nature surface sulfur species, resulting in the increase of Brønsted acid sites with the decrease of Lewis acid sites.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

We gratefully acknowledge financial support from the National Natural Science Foundation of China (No. 21808054).

Notes and references

1. X. Pu, N. Liu and L. Shi, Microporous Mesoporous Mater., 2015, 201, 17–23.
2. X. Pu, N. Liu, Z. Jiang and L. Shi, Ind. Eng. Chem. Res., 2012, 51, 13891–13896.
3. Y. Tian, X. Meng and L. Shi, Ind. Eng. Chem. Res., 2013, 52, 6655–6661.
4. N. Liu, X. Pu and L. Shi, Chem. Eng. Sci., 2014, 119, 114–123.
5. J. Yao, N. Liu, L. Shi and X. Wang, Catal. Commun., 2015, 66, 126–129.
6. G. X. Yu, X. L. Zhou, C. Tang, C. L. Li, J. A. Wang and O. Novaro, Catal. Commun., 2008, 9, 1770–1774.
7. A. I. Rabee, L. J. Durndell, N. E. Fouad, L. Frattini, M. A. Isaacs, A. F. Lee, G. A. Mekhemer, V. C. dos Santos, K. Wilson and M. I. Zaki, Mol. Catal., 2018, 458, 206–212.
8. V. S. Marakatti, S. Marappa and E. M. Gaigneaux, New J. Chem., 2019, 43, 7733–7742.
9. F. Cavani, S. Guidetti, C. Trevisanut, E. Ghedini and M. Signoretto, Appl. Catal., A, 2011, 409, 267–278.
10. I. Dosuna-Rodríguez, C. Adriany and E. M. Gaigneaux, Catal. Today, 2011, 167, 56–63.
11. N. Essayem, V. Martin, A. Riondel and J. C. Védrine, Appl. Catal., A, 2007, 326, 74–81.
12. M. A. Abedin, S. Kanitkar, S. Bhattar and J. J. Spivey, Catal. Today, 2021, 365, 71–79.
13. P. Wang, J. Zhang, G. Wang, C. Li and C. Yang, J. Catal., 2016, 338, 124–134.
14. Y. Zhang, T. Chen, G. Zhang, G. Wang and H. Zhang, Appl. Catal., A, 2018, 562, 258–266.
15. A. K. Amin, W. Trisunaryanti and K. Wijaya, J. Nano Res., 2019, 57, 31–39.
16. P. Nascimento, C. Akratopoulou, M. Oszagyan, G. Coudurier, C. Travers, J. F. Joly and J. C. Vedrine, in New Frontiers in Catalysis - Proceedings of the 10th International Congress on Catalysis, Budapest, 1993, pp. 1185–1197,  DOI:10.1016/s0167-2991(08)64443-2.
17. S. Ohyama, Top. Catal., 2003, 22, 337–343.
18. N. Liu, X. Wang, L. Shi and X. Meng, New J. Chem., 2019, 43, 3625–3632.
19. J. Wang and C. Mou, Microporous Mesoporous Mater., 2008, 110, 260–270.
20. M. A. Coelho, D. E. Resasco, E. C. Sikabwe and R. L. White, Catal. Lett., 1995, 32, 253–262.
21. J. Wang and C. Mou, Appl. Catal., A, 2005, 286, 128–136.
22. C. Chen, S. Cheng, H. Lin, S. Wong and C. Mou, Appl. Catal., A, 2001, 215, 21–30.
23. J. Wang and C. Mou, Catal. Today, 2008, 131, 162–172.
24. Y. Yang and H. Weng, Appl. Catal., A, 2010, 384, 94–100.
25. I. Ullah, T. Taha, A. M. Alenad, I. Uddin, A. Hayat, A. Hayat, M. Sohail, A. Irfan, J. Khan and A. Palamanit, Interfaces, Surf. Interfaces, 2021, 25, 101227.
26. D. J. Zalewski, S. Alerasool and P. K. Doolin, Catal. Today, 1999, 53, 419–432.
27. P. D. L. Mercera, J. G. V. Ommen, E. B. M. Doesburg, A. J. Burggraaf and J. R. H. Ross, Appl. Catal., 1990, 57, 127–148.
28. C. Morterra, C. G. F. Pinna and M. Signoretto, J. Catal., 1995, 157, 109–123.
29. T. Y. Toshio Ishida and K. Tanabe, Chem. Lett., 1988, 1869–1872.
30. X. Zhang, A. I. Rabee, M. Isaacs, A. F. Lee and K. Wilson, Engineering, ACS Sustainable Chem. Eng., 2018, 6, 14704–14712.
31. P. Wang, S. Wang, C. Yang, C. Li and X. Bao, Ind. Eng. Chem. Res., 2019, 58, 14638–14645.
32. W. Hua, Y. Xia, Y. Yue and Z. Gao, J. Catal., 2000, 196, 104–114.
33. Z. Ma, X. Meng, C. Yang, N. Liu, Y. Zhang and L. Shi, Ind. Eng. Chem. Res., 2017, 56, 5598–5606.
34. R. Iwamoto, C. Fernandez, J. P. Amoureux and J. Grimblot, J. Phys. Chem. B, 1998, 102, 4342–4349.
35. S. C. Shen, Q. Chen, P. S. Chow, G. H. Tan, X. T. Zeng, Z. Wang and R. B. H. Tan, J. Phys. Chem. C, 2007, 111, 700–707.
36. R. O. Patrizia Canton, F. Pinna, G. Strukul, P. Riello, M. Meneghetti, G. Cerrato, C. Morterra and A. Benedetti, Chem. Mater., 2001, 13, 1634–1641.
37. R. Beerthuis, L. Huang, N. R. Shiju, G. Rothenberg, W. Shen and H. Xu, ChemCatChem, 2018, 10, 211–221.
38. H. Tüysüz, Y. J. Hwang, S. B. Khan, A. M. Asiri and P. Yang, Nano Res., 2012, 6, 47–54.
39. S. Ardizzone, C. L. Bianchi, G. Cappelletti and F. Porta, J. Catal., 2004, 227, 470–478.
40. T. Witoon, T. Permsirivanich, N. Kanjanasoontorn, C. Akkaraphataworn, A. Seubsai, K. Faungnawakij, C. Warakulwit, M. Chareonpanich and J. Limtrakul, Catal. Sci. Technol., 2015, 5, 2347–2357.
41. J. Ryczkowski, Catal. Today, 2001, 68, 263–381.
42. E. Vlasov, S. Myakin, M. Sychov, A. Aho, A. Y. Postnov, N. Mal'Tseva, A. Dolgashev, S. O. Omarov and D. Y. Murzin, Catal. Lett., 2015, 145, 1651–1659.
43. S. Zhou, Y. Song, J. Zhao, X. Zhou and L. Chen, Fuels, Energy Fuels, 2021, 35, 14860–14867.
44. M. Abdollahi-Alibeik and E. Shabani, J. Iran. Chem. Soc., 2013, 11, 351–359.
45. B. Tyagi, M. K. Mishra and R. V. Jasra, J. Mol. Catal. A: Chem., 2007, 276, 47–56.
46. B. T. Manish, K. Mishra and R. V. Jasra, Ind. Eng. Chem. Res., 2003, 42, 5727–5736.
47. J. L. Ropero-Vega, A. Aldana-Pérez, R. Gómez and M. E. Niño-Gómez, Appl. Catal., A, 2010, 379, 24–29.
48. V. Vishwanathan, G. Balakrishna, B. Rajesh, V. Jayasri, L. M. Sikhwivhilu and N. J. Coville, Catal. Commun., 2008, 9, 2422–2427.
49. T. Yamaguchi, T. Jin and K. Tanabe, J. Phys. Chem., 1986, 90, 3148–3152.
50. M. Bensitel, O. Saur, J.-C. Lavalley and B. A. Morrow, Mater. Chem. Phys., 1988, 19, 147–156.
51. X. Li, K. Nagaoka, R. Olindo and J. Lercher, J. Catal., 2006, 238, 39–45.
52. V. K. Booramurthy, R. Kasimani, S. Pandian, B. J. E. S. Ragunathan and P. Research, Environ. Sci. Pollut. Res., 2020, 1–8.
53. S. Ardizzone, C. L. Bianchi, V. Ragaini and B. Vercelli, Catal. Lett., 1999, 62, 59–65.
54. H. Wang, Y. Li, F. Yu, Q. Wang, B. Xing, D. Li and R. Li, Chem. Eng. J., 2019, 364, 111–122.
55. K. Arata and M. J. A. C. Hino, Appl. Catal., 1990, 59, 197–204.
56. T. Yamaguchi, T. Jin and K. Tanabe, J. Phys. Chem., 1986, 90, 3148–3152.
57. C. A. Emeis, J. Catal., 1993, 141, 347–354.
58. P. Kalita, N. Gupta and R. Kumar, J. Catal., 2007, 245, 338–347.
59. K. Ren, D. Kong, X. Meng, X. Wang, L. Shi and N. Liu, J. Saudi Chem. Soc., 2019, 23, 198–204.
60. M. K. Mishra, B. Tyagi and R. V. Jasra, J. Mol. Catal. A: Chem., 2004, 223, 61–65.
61. Y. Sun, S. Ma, Y. Du, L. Yuan, S. Wang, J. Yang, F. Deng and F.-S. Xiao, J. Phys. Chem. B, 2005, 109, 2567–2572.
62. K. Föttinger, K. Zorn and H. Vinek, Appl. Catal., A, 2005, 284, 69–75.
63. Q. Zhao, J. Yao, L. Shi and X. Wang, RSC Adv., 2016, 6, 84553–84561.
64. J. Liu, N. Liu, K. Ren, L. Shi and X. Meng, Ind. Eng. Chem. Res., 2017, 56, 7693–7699.
65. N. Liu, X. Pu, X. Wang and L. Shi, J. Ind. Eng. Chem., 2014, 20, 2848–2857.
66. S. Narayanan and K. Deshpande, J. Mol. Catal. A: Chem., 1995, 104, 109–113.
67. S. Narayanan and K. Deshpande, Appl. Catal., A, 2000, 199, 1–31.
68. H. Song, N. Wang, H. Song, F. Li and Z. Jin, Chin. J. Chem. Eng., 2014, 22, 1226–1231.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/d1cy01443a

---