Metal Chlorides Grafted on SAPO-5 (MCl x / SAPO-5) as Reusable and Superior Catalysts for Acylation of 2-Methylfuran Under Non-Microwave Instant Heating Condition

: Highly active metal chlorides grafted on silicoaluminophosphate number 5, MCl x / SAPO-5 (M = Cu, Co, Sn, Fe and Zn) catalysts via simple grafting of respective metal chlorides (MCl x ) onto SAPO-5 are reported. The study shows that thermochemical treatment after grafting is essential to ensure the formation of chemical bondings between MCl x and SAPO-5. In addition, the microscopy, XRD and nitrogen adsorption analyses reveal the homogeneous distribution of MCl x species on the SAPO-5 surface. Furthermore, the elemental microanalysis conﬁrms the formation of Si–O–M covalent bonds in ZnCl x / SAPO-5, SnCl x / SAPO-5 and FeCl x / SAPO-5 whereas only dative bondings are formed in CoCl x / SAPO-5 and CuCl x / SAPO-5. The acidity of MCl x / SAPO-5 is also a ﬀ ected by the type of metal chloride grafted. Thus, their catalytic behavior is evaluated in the acid-catalyzed acylation of 2-methylfuran under novel non-microwave instant heating conditions (90–110 ◦ C, 0–20 min). ZnCl x / SAPO-5, which has the largest amount of acidity (mainly Lewis acid sites), exhibits the best catalytic performance (94.5% conversion, 100% selective to 2-acetyl-5-methylfuran) among the MCl x / SAPO-5 solids. Furthermore, the MCl x / SAPO-5 solids, particularly SnCl x / SAPO-5, FeCl x / SAPO-5 and ZnCl x / SAPO-5, also show more superior catalytic performance than common homogeneous acid catalysts (H 2 SO 4 , HNO 3 , CH 3 COOH, FeCl 3 , ZnCl 2 ) with higher reactant conversion and catalyst reusability, thus o ﬀ ering a promising alternative for the replacement of hazardous homogeneous catalysts in Friedel–Crafts reactions.


Introduction
Furan and its acylated derivatives are important aromatic compounds that can be found in plants and microorganisms [1,2]. These heterocyclic compounds have widely been used in drug, food, fragrance, petrochemical and fine chemical industries due to their non-polar characteristics,

Synthesis of SAPO-5 Molecular Sieve
The SAPO-5 microporous solid was synthesized under hydrothermal condition using a hydrogel system with a molar chemical composition of 1Al 2 3 , 3.441 g, 98%, Sigma-Aldrich) and [bzmIm]OH (28.000 g) were mixed in distilled water (6.224 g) and the mixture was magnetically stirred (400 rpm) for 40 min. Phosphoric acid (H 3 PO 4 , 2.670 g, 85.5%, Sigma-Aldrich) was then slowly added into the mixture in a dropwise manner under continuous stirring where the entire process of addition took about 15 min. Subsequently, colloidal HS-40 silica (0.583 g, 40% SiO 2 , Sigma-Aldrich) was added to the hydrogel mixture. The resulting hydrogel was stirred for another 5 min before it was transferred to a 50-mL Teflon-lined stainless-steel autoclave for crystallization (180 • C, 10 h). After hydrothermal treatment, the solid product was recovered by filtration, washed with distilled water until pH 7 and freeze dried. Finally, the as-synthesized SAPO-5 was calcined at 550 • C for 6 h with a heating rate of 1 • C/min in order to remove the organic template. Five types of metal chlorides, namely CuCl 2 , AlCl 3 , CoCl 2 , FeCl 3 , ZnCl 2 , were grafted on SAPO-5 using the following procedure: Typically, the metal chloride in anhydrous form (0.5 mmol, Sigma-Aldrich) was first dissolved in ethanol (20 mL) at room temperature. The calcined SAPO-5 (1.500 g) was then added into the solution and the mixture was refluxed at 80 • C for 2 h. The solids were recovered by rotary evaporation at 80 • C, and heated at 300 • C for 3 h under air flow to produce the respective MCl x /SAPO-5 solids where M = Cu, Co, Sn, Fe and Zn ( Figure 1).

Synthesis of SAPO-5 Molecular Sieve
The SAPO-5 microporous solid was synthesized under hydrothermal condition using a hydrogel system with a molar chemical composition of 1Al2O3:2.5P2O5:2.5[bzmIm]2O:0.47SiO2:180H2O. First, aluminum isopropoxide (Al(OiPr)3,3.441 g, 98%, Sigma-Aldrich) and [bzmIm]OH (28.000 g) were mixed in distilled water (6.224 g) and the mixture was magnetically stirred (400 rpm) for 40 min. Phosphoric acid (H3PO4, 2.670 g, 85.5%, Sigma-Aldrich) was then slowly added into the mixture in a dropwise manner under continuous stirring where the entire process of addition took about 15 min. Subsequently, colloidal HS-40 silica (0.583 g, 40% SiO2, Sigma-Aldrich) was added to the hydrogel mixture. The resulting hydrogel was stirred for another 5 min before it was transferred to a 50-mL Teflon-lined stainless-steel autoclave for crystallization (180 °C, 10 h). After hydrothermal treatment, the solid product was recovered by filtration, washed with distilled water until pH 7 and freeze dried. Finally, the as-synthesized SAPO-5 was calcined at 550 °C for 6 h with a heating rate of 1 °C/min in order to remove the organic template.

Preparation of Metal Chlorides Grafted SAPO-5 (MClx/SAPO-5, M = Cu, Co, Sn, Fe, Zn) Solids
Five types of metal chlorides, namely CuCl2, AlCl3, CoCl2, FeCl3, ZnCl2, were grafted on SAPO-5 using the following procedure: Typically, the metal chloride in anhydrous form (0.5 mmol, Sigma-Aldrich) was first dissolved in ethanol (20 mL) at room temperature. The calcined SAPO-5 (1.500 g) was then added into the solution and the mixture was refluxed at 80 °C for 2 h. The solids were recovered by rotary evaporation at 80 °C, and heated at 300 °C for 3 h under air flow to produce the respective MClx/SAPO-5 solids where M = Cu, Co, Sn, Fe and Zn ( Figure 1).

Characterization
The phase purity of samples was examined by a Bruker Advance D8 XRD diffractometer (CuKα radiation, λ = 1.5418 Å, 40 kV, 10 mA) using a scan speed of 0.2°/min and a step size of 0.02°. The d spacing in (100) plane (d100) and unit cell parameters (ao) of the solid samples were calculated using the Equations (1) and (2), respectively: The surface and morphological properties of the solids were studied using a JEOL JSM-6701F FESEM microscope while the micro-elemental analysis was performed by using an EDX spectrometer (Oxford Instruments Ltd., UK). The porous properties of the samples were studied using the nitrogen

Characterization
The phase purity of samples was examined by a Bruker Advance D8 XRD diffractometer (CuKα radiation, λ = 1.5418 Å, 40 kV, 10 mA) using a scan speed of 0.2 • /min and a step size of 0.02 • . The d spacing in (100) plane (d 100 ) and unit cell parameters (a o ) of the solid samples were calculated using the Equations (1) and (2), respectively: The surface and morphological properties of the solids were studied using a JEOL JSM-6701F FESEM microscope while the micro-elemental analysis was performed by using an EDX spectrometer (Oxford Instruments Ltd., UK). The porous properties of the samples were studied using the nitrogen Brunauer-Emmett-Teller (BET) was then used to determine the surface area while the average pore diameter was estimated using the Barrett-Joyner-Halenda model. The surface acidity of the solids was examined by using pyridine adsorption combined with FTIR spectroscopy. Prior to analysis, a self-supported wafer (ca. 8 mg) was prepared and degassed in an IR vacuum cell at 300 • C under vacuum (4 h, 10 −6 bar). The background spectrum of the sample was acquired after cooling to room temperature. Subsequently, pyridine was adsorbed onto the sample for 1 min. The excessive pyridine was then evacuated, and the IR spectrum of the pellet was recorded (6 cm −1 resolution, 200 scans accumulation). The pellet was heated at 150 • C and maintained for 1 h before the second IR spectrum was recorded after cooling. The pellet was further heated at 300 • C for 1 h before the last IR spectrum was collected. The absorption bands due to Lewis and Brönsted acid sites were detected within the range of 1600-1400 cm −1 and the amount of both acid sites in the solid samples were calculated according to [38].

Catalytic Testing
The catalytic activity of MCl x /SAPO-5 was studied on the Friedel-Crafts acylation of 2-methylfuran under novel non-microwave instant heating conditions, where acetic anhydride was used as an acylating agent. Initially, the activated catalyst (0.200 g, 200 • C, 1 h) was loaded in a glass reaction vessel before acetic anhydride (14.1 mmol) and 2-methylfuran (4.7 mmol) were successively added. The vessel was then sealed with a silicone rubber cap and instantly heated in an Anton Paar's Monowave 50 reactor (110 • C, 20 min, 800 rpm stirring). After the reaction, the catalyst was recovered using high-speed centrifugation (10,000 × g rpm, 5 min) while the reaction solution was quantitatively and qualitatively analyzed using a gas chromatograph (Agilent 7890A, Equity-5 capillary column) and a gas chromatography-mass spectrometer (Agilent 7890A/5975C GCMS System, SPB-1 column), respectively, where toluene (5 µL) was added as an internal standard.

Characterization
The MCl x /SAPO-5 catalysts (M = Cu, Co, Sn, Fe or Zn) were synthesized by using a facile grafting method where the chloride salts were first dissolved in polar solvent before SAPO-5 solid support was added. The solvent was then removed via rotary evaporation before subjecting to thermochemical activation at 300 • C to induce strong chemical bondings formation between the metal chlorides and the SAPO-5 support. In order to confirm this phenomenon, the MCl x /SAPO-5 samples before and after thermochemical activation (0.200 g) were dispersed and sonicated in ethanol solution (5 mL) for 30 min. It was shown that for the non-thermochemical activated MCl x /SAPO-5 samples, colored ethanol solutions were obtained after sonication indicating weak interaction between the metal chloride molecules and the SAPO-5 surface. As a result, a severe leaching problem of colored metal chloride salts was observed, and the results surprisingly were not in line with the works reported in [32,39]. However, the leaching problem was significantly suppressed after thermochemical activation, where the ethanol solutions remain almost transparent after sonication. Thus, it can be concluded that strong chemical bonding between metal chloride and SAPO-5 had been formed after thermochemical activation. Figure 2 shows the XRD patterns of MCl x /SAPO-5 solids and the XRD pattern of pristine SAPO-5 was also recorded for comparison purposes. The phase purity of SAPO-5 in all samples is evidenced from the XRD patterns where all major XRD peaks at 7.  25.84 • (220) and 24.38 • (410) are observed, and they are well indexed according to the standard hexagonal AFI crystal phase with space group p6/mmc [40]. As compared to pristine SAPO-5, the XRD peaks of MCl x /SAPO-5 are slightly right-shifted which can be explained by the reduction of micropore sizes of SAPO-5 as a result of the grafting of metal chloride species [41]. This observation can  The effects of incorporation of metal chlorides on the morphological properties of SAPO-5 were studied using the FESEM microscopy technique. Figure 3 displays the FESEM images of pristine SAPO-5 and MClx/SAPO-5 crystals. As shown, all samples exhibit hexagonal prisms similar to the theoretical shape [42,43], and the morphology of the samples remains intact indicating that the structural integrity of SAPO-5 is preserved after grafting modification and thermochemical activation. The particle size distribution based on the length of the hexagonal prisms of all samples is also plotted by measuring the length of 100 single crystal particles at different spots. The statistical results indicate that the pristine SAPO-5 is sized ranging from 75 nm to 560 nm and has an average crystal size of 228 nm. Upon grafting of metal chloride compounds, the crystal size distribution is nearly intact, and the average crystal size remains almost identical (~230 nm). Furthermore, no big granulate particles of metal chlorides are observed on the surface of MClx/SAPO-5 crystals revealing that no agglomeration of metal chloride salts occurs, and they are evenly distributed on the support surface; this is in accordance with the XRD observation ( Figure 2). The effects of incorporation of metal chlorides on the morphological properties of SAPO-5 were studied using the FESEM microscopy technique. Figure 3 displays the FESEM images of pristine SAPO-5 and MCl x /SAPO-5 crystals. As shown, all samples exhibit hexagonal prisms similar to the theoretical shape [42,43], and the morphology of the samples remains intact indicating that the structural integrity of SAPO-5 is preserved after grafting modification and thermochemical activation. The particle size distribution based on the length of the hexagonal prisms of all samples is also plotted by measuring the length of 100 single crystal particles at different spots. The statistical results indicate that the pristine SAPO-5 is sized ranging from 75 nm to 560 nm and has an average crystal size of 228 nm. Upon grafting of metal chloride compounds, the crystal size distribution is nearly intact, and the average crystal size remains almost identical (~230 nm). Furthermore, no big granulate particles of metal chlorides are observed on the surface of MCl x /SAPO-5 crystals revealing that no agglomeration of metal chloride salts occurs, and they are evenly distributed on the support surface; this is in accordance with the XRD observation ( Figure 2).  The chemical compositions of pristine SAPO-5 and MCl x /SAPO-5 are confirmed by using EDX spectroscopy, micro-elemental analysis and the amount of each element in the samples is summarized in Table 1. As calculated, the molecular formula of pristine SAPO-5 is Si 1.485 Al 12.093 P 10.196 O 47.858 which is quite similar to the theoretical molecular formula of AFI-type molecular sieve (Si n Al 12 P 12-n O 48 ) ( Table 2) [38]. The Si/(P + Al + Si) ratio of pure SAPO-5 is calculated to be 0.062, where the presence of Si atoms is the attribution to the generation of surface Brönsted acidity in SAPO-5 [44]. Meanwhile, the molecular formulae of CuCl x /SAPO-5, CoCl x /SAPO-5, SnCl x /SAPO-5, FeCl x /SAPO-5 and ZnCl x /SAPO-5 are determined to be Cu 3 The amount of metal and chloride atoms present in SAPO-5 is also determined. Interestingly, ZnCl x /SAPO-5, FeCl x /SAPO-5 and SnCl x /SAPO-5 displayed lower Cl:M ratio than their respective pure metal chloride salts. Hence, the elemental analysis suggests that the metal chlorides have been incorporated onto the surface of ZnCl x /SAPO-5, FeCl x /SAPO-5 and SnCl x /SAPO-5 via condensation reaction (AFI-OH + Cl-M → AFI-O-M + HCl). In contrast, CoCl x /SAPO-5 and CuCl x /SAPO-5 show Cl:M ratio ≈ 2, which is nearly similar to their respective metal salts. Hence, the study demonstrates that no condensation reaction between AFI-OH and Cl-M took place in CoCl x /SAPO-5 and CuCl x /SAPO-5. Instead, we believe that the CoCl 2 and CuCl 2 are bound to the SAPO-5 surface via dative covalent bondings, where Cu 2+ and Co 2+ cations are able to form four to six chemical bondings (covalent and dative bonds) [45,46].    The porous properties of pristine SAPO-5 and MClx/SAPO-5 solids were studied using XRD and N2 adsorption-desorption analyses, and the data are summarized in Table 3. The study shows that the pristine SAPO-5 initially has a specific BET surface area (SBET) and an average pore size (D) of 257 m 2 /g and 37.9 nm, respectively. It also exhibits a considerably high total pore volume (0.35 cm 3 /g) due to its highly porous nature. After incorporating various metal chlorides, all grafted solid samples show a decrease in the surface area, pore size and pore volume where ZnClx/SAPO-5 experiences the most significant effects with those parameters reduce to 89 m 2 /g, 19.3 nm and 0.04 cm 3 /g, respectively. Hence, this phenomenon reveals that the metal chloride particles are partially filled and occupied the pores in MClx/SAPO-5, which is in agreement with the XRD results that show a decrease of d100 spacing and unit cell parameters upon grafting modification (Figure 2 and Table 3). The porous properties of pristine SAPO-5 and MCl x /SAPO-5 solids were studied using XRD and N 2 adsorption-desorption analyses, and the data are summarized in Table 3. The study shows that the pristine SAPO-5 initially has a specific BET surface area (S BET ) and an average pore size (D) of 257 m 2 /g and 37.9 nm, respectively. It also exhibits a considerably high total pore volume (0.35 cm 3 /g) due to its highly porous nature. After incorporating various metal chlorides, all grafted solid samples show a decrease in the surface area, pore size and pore volume where ZnCl x /SAPO-5 experiences the most significant effects with those parameters reduce to 89 m 2 /g, 19.3 nm and 0.04 cm 3 /g, respectively. Hence, this phenomenon reveals that the metal chloride particles are partially filled and occupied the pores in MCl x /SAPO-5, which is in agreement with the XRD results that show a decrease of d 100 spacing and unit cell parameters upon grafting modification (Figure 2 and Table 3). The surface acidity of MCl x /SAPO-5 was characterized by pyridine adsorption coupled with FTIR spectroscopy, viz. a very reliable analysis for surface acidity study (type, strength and amount of acid sites). By using FTIR spectroscopy, the Lewis acid sites and Brönsted acid sites of a solid acid catalyst can be detected at 1454 and 1545 cm −1 , respectively [47]. The pyridine adsorption-FTIR spectra after desorption at 25 • C, 150 • C and 300 • C are shown in Table 4 where these three desorption temperatures can be used to identify weak, mild and strong acid sites. For pristine SAPO-5, the Brönsted acid sites with strong acid strength (~25 µmol/g) are observed, while mild Lewis acidity (~72 µmol/g) due to surface defect sites is also detected.
Upon grafting of metal chlorides, the Lewis acidity of MCl x /SAPO-5 is significantly enhanced while the Brönsted acidity remains almost unaffected. For instance, the mild and strong Lewis acidities of FeCl x /SAPO-5 have increased about 4 and 25 folds after surface modification, respectively. Among the MCl x /SAPO-5 samples prepared, ZnCl x /SAPO-5, which has the lowest surface area, possesses the largest number of Lewis acid sites especially with mild acid strength (444.9 µmol/g) whereas FeCl x /SAPO-5 has the largest amount of strong Lewis acid sites (219.1 µmol/g). On the other hand, CuCl x /SAPO-5 possesses the lowest number of mild and strong Lewis acid sites; only 126.6 and 38.9 µmol/g were measured, respectively.
The Lewis-Brönsted acids ratios (L/B) are also calculated in order to reveal the acid nature of the solids. For pristine SAPO-5, the L/B ratio with mild acid strength is very low (2.7) while the L/B ratio for strong acid strength is merely 0.3. Hence, the results agree with the previous work that SAPO-5 is a mild-to-strong Brönsted acid solid catalyst [48]. Interestingly, the L/B ratio increases significantly after incorporation of metal chlorides, particularly ZnCl x /SAPO-5 and FeCl x /SAPO-5, where the Lewis acidity with mild and strong acid strength dominates the total acid sites of the samples. Table 4. Surface acidity of MCl x /SAPO-5 measured using pyridine (Py) adsorption coupled with FTIR spectroscopy technique.

Lewis Acid
Sites (

Effect of Reaction Time and Temperature
Friedel-Crafts acylation of 2-methylfuran with acetic anhydride as an acylation agent was selected as a model reaction (Scheme 1) to study the catalytic behavior of MCl x /SAPO-5 catalysts. The reaction was conducted at 90-110 • C within 20 min by using an instant heating technique that mimics microwave fast heating [49]. Of note, the reaction condition used is much gentle than those reported in [50][51][52][53].
The catalytic reaction was also tested with pristine SAPO-5. However, only 45.1% conversion was afforded under similar reaction conditions due to its low Lewis acidity.

Effect of Reaction Time and Temperature
Friedel-Crafts acylation of 2-methylfuran with acetic anhydride as an acylation agent was selected as a model reaction (Scheme 1) to study the catalytic behavior of MClx/SAPO-5 catalysts. The reaction was conducted at 90-110 °C within 20 min by using an instant heating technique that mimics microwave fast heating [49]. Of note, the reaction condition used is much gentle than those reported in [50][51][52][53].
The reaction is inactive in the absence of a catalyst (110 °C, 20 min) where nearly no reaction conversion (0.2%) was recorded ( Figure 5). When MClx/SAPO-5 catalysts are added, the conversion increases tremendously with 100% selective to 2-acetyl-5-methylfuran, viz. a valuable biofuel compound and intermediate in the pharmaceutical industry [54], under similar reaction conditions (110 °C, 20 min). Among the solid catalysts prepared, ZnClx/SAPO-5 is the most reactive, giving 94.5% conversion, followed by FeClx/SAPO-5 (87.8%), SnClx/SAPO-5 (81.5%), CoClx/SAPO-5 (72.3%) and CuClx/SAPO-5 (66.7%). Hence, the catalytic reaction results further support the FTIR-pyridine adsorption study that the acylation of 2-methylfuran requires mild-to-strong Lewis acidity to accelerate the reaction. The catalytic reaction was also tested with pristine SAPO-5. However, only 45.1% conversion was afforded under similar reaction conditions due to its low Lewis acidity. The activation energies of acylation of a 2-methylfuran reaction catalyzed with and without MClx/SAPO-5 solid catalysts are also determined using the Arrhenius plots in order to understand the chemical kinetics of this reaction. The reaction was found to follow second-order rate law where the reaction kinetics depend on the concentrations of 2-methylfuran and acetic anhydride reactants [55]. The second-order rate constants, k2nd, at 90 °C, 100 °C and 110 °C are obtained and used to plot the Arrhenius linear plot: ln k versus 1/T. The activation energy of non-catalyzed acylation of 2methylfuran is very high (181.3 kJ mol −1 ) [49]. Nevertheless, the activation energy is reduced when pristine SAPO-5 or MClx/SAPO-5 solid catalysts are added ( Figure 6). This phenomenon occurs because the solid acid catalyst has increased the rate of reaction by offering an alternative route to the reaction product which is more feasibly occurred (lower activation energy) than the reaction pathway not mediated by the catalyst [56]. Among the solid acid catalysts studied, ZnClx/SAPO-5 gives the lowest activation energy (36.4 kJ mol −1 ), followed by FeClx/SAPO-5 (39.2 kJ mol −1 ), SnClx/SAPO-5 (40.1 kJ mol −1 ), CoClx/SAPO-5 (63.9 kJ mol −1 ), CuClx/SAPO-5 (82.3 kJ mol −1 ) and SAPO-5 (96.4 kJ mol −1 ). Based on the catalytic study results, the best catalyst is ZnClx/SAPO-5 and hence it is chosen for further catalytic study. The activation energies of acylation of a 2-methylfuran reaction catalyzed with and without MCl x /SAPO-5 solid catalysts are also determined using the Arrhenius plots in order to understand the chemical kinetics of this reaction. The reaction was found to follow second-order rate law where the reaction kinetics depend on the concentrations of 2-methylfuran and acetic anhydride reactants [55]. The second-order rate constants, k 2nd , at 90 • C, 100 • C and 110 • C are obtained and used to plot the Arrhenius linear plot: ln k versus 1/T. The activation energy of non-catalyzed acylation of 2-methylfuran is very high (181.3 kJ mol −1 ) [49]. Nevertheless, the activation energy is reduced when pristine SAPO-5 or MCl x /SAPO-5 solid catalysts are added ( Figure 6). This phenomenon occurs because the solid acid catalyst has increased the rate of reaction by offering an alternative route to the reaction product which is more feasibly occurred (lower activation energy) than the reaction pathway not mediated by the catalyst [56]. Among the solid acid catalysts studied, ZnCl x /SAPO-5 gives the lowest activation energy (36.4 kJ mol −1 ), followed by FeCl x /SAPO-5 (39.2 kJ mol −1 ), SnCl x /SAPO-5 (40.1 kJ mol −1 ), CoCl x /SAPO-5 (63.9 kJ mol −1 ), CuCl x /SAPO-5 (82.3 kJ mol −1 ) and SAPO-5 (96.4 kJ mol −1 ). Based on the catalytic study results, the best catalyst is ZnCl x /SAPO-5 and hence it is chosen for further catalytic study.

Catalytic Comparative Study
The catalytic performance of MCl x /SAPO-5 is also compared with some common homogeneous Lewis and Brönsted acid catalysts under the same reaction condition where an equivalent amount of catalyst is applied. As shown in Table 5, homogeneous Lewis acids, such as ZnCl 2 (75.5%) and FeCl 3 (68.1%), are found to be more reactive than the homogeneous Brönsted acids, such as H 2 SO 4 (68.4%), HNO 3 (56.2%) and CH 3 COOH (32.7%), where the former class of catalysts undergoes different catalytic reaction route but kinetically more favored by first dissociating the C-O-C bond of the anhydride into the acetoxy anion (limiting step) instead of acylium cation (by the latter ones) prior to forming 2-acetyl-5-methylfuran [54]. Nevertheless, these homogeneous catalysts are not reusable due to separation difficulty and surprisingly, their catalytic reactivity is much lower than those of SnCl x /SAPO-5, FeCl x /SAPO-5 and FeCl x /SAPO-5. Hence, it is speculated that the presence of both Brönsted and Lewis acid sites in the MCl x /SAPO-5 might contribute to the synergistic cooperative effect on the catalytic enhancement of the Friedel-Crafts reaction [57].

Catalyst Reusability Test
Catalyst reusability of MCl x /SAPO-5 catalysts is also studied for four consecutive cycles as it represents a significant advantage for green chemical production [57]. As shown in Figure 7, SnCl x /SAPO-5, FeCl x /SAPO-5 and ZnCl x /SAPO-5, which have stronger covalently-bonded metal chlorides, exhibit fewer steep slopes than the CoCl x /SAPO-5, and CuCl x /SAPO-5, which have weaker dative bondings bound to the metal chloride compounds. As a result, the previous three catalysts have higher catalyst reusability than the latter two. Among the catalysts prepared, ZnCl x /SAPO-5 is the most reactive and its reaction conversion remains considerably high (78.4%) even after the fourth consecutive run, whereas CoCl x /SAPO-5 and CuCl x /SAPO-5 show the least catalytic activity, where, their reaction conversion is almost comparable as that of pristine SAPO-5 after the fourth consecutive cycle. Thus, ZnCl x /SAPO-5 is considerably stable in terms of catalyst reusability and hence, it can be a recyclable and active heterogeneous catalyst for various acid-catalyzed reactions.
Furthermore, the catalytic activity of metal chlorides leached in reaction solutions was also examined. The MClx/SAPO-5 (0.20 g) was first stirred in acetic anhydride (14.1 mmol) and 2-methyl furan (4.7 mmol) for 5 h at room temperature before the solution (after separated from the solid catalyst (10,000 rpm, 5 min)) was heated at 110 • C for 20 min. It is shown that the reaction conversions of the solutions after separating from ZnCl x /SAPO-5, FeCl x /SAPO-5 and SnCl x /SAPO-5 (<5%) are lower than those of CoCl x /SAPO-5 and CuCl x /SAPO-5 (~10%). Thus, it indicates that the respective metal chlorides are strongly bound to the former three catalysts as compared to the latter two catalysts, in line with the XRD and ICP-OES spectroscopy data.
are strongly bound to the former three catalysts as compared to the latter two catalysts, in line with the XRD and ICP-OES spectroscopy data.