Zirconium-doped porous magadiite heterostructures upon 2D intragallery in situ hydrolysis–condensation–polymerization strategy for liquid-phase benzoylation

Abstract

Novel zirconium-doped porous magadiite heterostructures (PMH-xZr, x = Zr/Si molar ratio) are fabricated by two-dimensional intragallery cosurfactant-directing in situ hydrolysis–condensation–polymerization method of TEOS and Zr-n-propoxide from synthetic Na-magadiite and characterized systematically by XRD, SEM/(HR)TEM, ²⁹Si MAS NMR, BET, UV-vis DRS, NH₃-TPD, pyridine FT-IR, and XPS techniques. The results indicate that the obtained PMH-xZr materials possess high surface area and high thermal stability upon effective assembly of interlayer Zr-doped meso-structural silica and the layers of magadiite. The PMH-xZr samples with x < 0.2 show successful incorporation of Zr into the lattice of interlayer mesostructural silica framework leading to considerably generated Brønsted sites Zr–O(H)–Si and obviously increased Lewis sites Zr–O–Si along with well-kept layered supermicro-mesostructure, while PMH-0.2Zr shows delaminated layers. PMH-0.1Zr exhibits the highest liquid-phase benzoylation activity of anisole with benzoyl chloride (Conv. 99.5%) and yield for 4-methoxybenzophenone (4-MBP) (94.1%) due to the strongest synergy between the high concentration of surface Brønsted sites and supermicro-mesostructure. PMH-0.1Zr can be reused by no further chemical treatment for at least five runs with a slightly reduced 4-MBP yield. These PMH-xZr materials can serve as a promising solid acid catalyst and/or acidic support with high surface area and thermal stability in broad range of catalysis applications.

---

1. Introduction

Porous clay heterostructures (PCHs),¹ derived from cationic layered clays upon 2D clay intragallery co-surfactant directing in situ hydrolysis–condensation–polymerization of TEOS, has aroused gradually increasing interest in adsorption and separation,^2,3 electrochemistry,⁴ as supports⁵ and catalysts^6,7 due to its considerable high surface area, high thermal stability, combined supermicro- and mesoporous structure.^1,8,9 It is noted that the surface acidity of the materials plays a critical role in the catalysis process especially acid-catalytic reactions. However, there's none or a small concentration of acid sites located in the interlayer silica framework apart from those aroused from the parent clay.¹⁰ Several studies have been published aiming at further enhancement of acidity of PCHs, such as acid treatment of clays prior to PCH synthesis,¹¹ incorporating heteroatoms (e.g. Al,⁶ Ti¹²) into the silica pillars, and the use of various clay hosts (e.g. fluorohectorite,¹ montmorillonite,^6,12 synthetic saponite^10,13 and vermiculite¹³). While the acidity modification of PCHs derived from pure siliceous clay magadiite was rarely studied probably due to its higher layer charge density thus difficulties resulting from swelling.¹

Magadiite, a kind of hydrous sodium silicate found in lake beds at Lake Magadi,¹⁴ can be easily prepared by hydrothermal route in lab.^15,16 Upon synthetic silicic magadiite, a good candidate for PCHs can be expected with excellent resistance to acid and hydrothermal stability compared to common clays.¹ Pinnavaia et al.¹ have ever mentioned magadiite in their first report on PCHs from various clay hosts but without detailed study. It is noted that Kwon et al.^7,17 reported a mesoporous silica-pillared H⁺-magadiite via amine-catalyzed hydrolysis of TEOS without pre-swelling step, but the obtained silicic materials yet show very low or no acidity. Our recent study first reported the porous magadiite heterostructures (PMH) and its post-grafting aluminated derivatives (xAl-PMH) from NaAlO₂ showing greatly enhanced Friedel–Crafts alkylation activity.¹⁸

Transition metal oxide ZrO₂ with high corrosion resistance, low thermal conductivity and high melting point has attracted special attention in catalysis,¹⁹ but suffered from low surface area and non-shape selective nature. The common way to overcome these limitations lies in supporting ZrO₂ onto high surface oxides.^20,21 So far, there are a few reports on Zr–Si mixture pillared clays/PCHs. Kosuge et al.²² studied zirconia–silica-based pillared H-ilerite by intercalating octylamine and Zr-butoxide and found the formation of high surface products with uniform micropores and irregular gallery heights upon Zr contents. Han et al.²³ reported microporous Si–Zr-based pillared clays upon the reaction of Na-montmorillonite with colloidal SiO₂–ZrO₂ particles showing limited pore size used for selective adsorption of toluene and mesitylene. Pinto et al.² reported montmorillonite-based PCH modified with Zr by copolymerization of TEOS with Zr-ethoxide for separation of aromatic molecules but without detailed structure and acidity studies. While Cecilia et al.²⁴ prepared natural montmorillonite-based PCHs with varied Zr contents by long time sol–gel process (10 days) of TEOS and Zr-propoxide in 1-propanol and studied their texture, structure and acidity by SEM, BET, XRD, FTIR, TG-DTA, NH₃-TPD and 1-butene isomerization. To the best of our knowledge, there is no report on Zr-doped porous magadiite heterostructures, far from its acidic catalytic applications such as Friedel–Crafts acylation, which is highly desired to be explored.

In the present work, we report novel Zr-doped porous magadiite heterostructures (PMH-xZr) by cosurfactant-directing two-dimensional intragallery in situ hydrolysis–condensation–polymerization strategy of TEOS and Zr-n-propoxide from a synthetic Na-magadiite, which were studied by XRD, SEM/HRTEM, ²⁹Si MAS NMR, BET, UV-vis DRS, XPS, NH₃-TPD and pyridine FT-IR techniques. The effect of Zr-doping on the texture, acidity, and benzoylation performance of anisole with benzoyl chloride of PMH-xZr materials is thoroughly investigated. Moreover, the recycle efficiency of the PMH-xZr catalysts is also studied.

2. Experimental

2.1 Preparation of catalysts

The synthesis of the catalysts follows the strategy shown in Scheme 1. Initial Na-magadiite was prepared via hydrothermal route using silicate gel and NaOH according to the previous work.^18,25 In detail, the starting mixture composed of SiO₂ (Ludox suspension, HS-40, Sigma-Aldrich)/NaOH/H₂O with molar ratio of 9/2/75 was stirred for 3 h and then transferred into a steel autoclave with Teflon lining for hydrothermal treatment at 150 °C for 48 h. The product was washed with deionized water till pH of 7 and air-dried at 60 °C for 24 h giving the Na-magadiite with an empirical formula Na_1.9Si₁₄O₂₉·9.2H₂O (Table S1†), close to previously reported,^7,18 based on TG (Fig. S1†) and EDX data. Then, in step (I), 5.0 g of Na-magadiite and 5.0 g of cetyltrimethylammonium bromide (CTMABr) were suspended in 100 mL of deionized water. The suspension was heated to 80 °C and kept for 4 h and the resultant was washed with deionized water till pH 7 and air-dried at 60 °C for 24 h giving CTMA⁺-magadiite (QM in short). The empirical formula of QM were estimated as (C₁₉H₄₃N)_2.1Si₁₄O₂₉·2.45H₂O (Table S1†) upon the empirical formula of magadiite, assuming that the amount of CTMA⁺ ions exchanged for Na⁺ corresponds to the cation exchange capacity of clay. In step (II, II′), proper amount of QM was added into proper volume of n-octylamine (OA, 99.5%, ∼0.5% H₂O) under magnetic stirring for 1 h forming an intermediate suspension followed by adding a premixed solution of tetraethyl-orthosilicate (TEOS, 99.5%) and zirconium n-propoxide (Zr-nPro, 70 wt% in 1-propanol) with designed Zr/Si ratio to react at 25 °C for 8 h. In above systems, the QM/OA/TEOS/Zr-nPro molar ratios were designed as 1/20/150/0, 1/20/148.2/1.85, 1/20/146.3/3.66, 1/20/142.9/7.14, 1/20/136.4/13.64 and 1/20/125/25, respectively. The resultants were centrifuged, washed by absolute ethanol and dried at 60 °C for 24 h giving the precursors named as QOTM-xZr (x = Zr/Si, refers to Zr/Si molar ratio as 0, 0.0125, 0.025, 0.05, 0.1 and 0.2). The TG plots (Fig. S1†) of the typical precursors QOTM and QOTM-0.1Zr show that the calcination at the temperatures ≥ 550 °C could remove the organic templates. Thus in step (III), the precursors were calcined at 600 °C for 6 h resulting in Zr-doped porous magadiite heterostructure catalysts PMH-xZr.

Scheme 1 Schematic of the synthesis strategy of the PMH-xZr catalysts.

2.2 Characterization

X-ray diffraction (XRD) patterns were recorded on a Rigaku D/Max 2500 VB2+/PC diffractometer using Cu Kα radiation (40 kV, 50 mA) in 2θ range of 1–10° (small angle) and Cu Kα radiation (40 kV, 200 mA) in 2θ range of 15–60° (wide angle). The FT-IR spectra were recorded on a Bruker Vector22 FT-IR spectrometer in the range of 400–4000 cm⁻¹ with 4 cm⁻¹ resolution and 60 scans by the standard KBr disk technique (sample/KBr = 1/100) with almost identical mass of samples. Thermogravimetric analyses were taken on a Mettler Toledo TGA/DSC 1/1100 SF thermo analyzer at a heating rate of 10 °C min⁻¹ in 25–1000 °C in N₂ flow (25 mL min⁻¹). ²⁹Si MAS NMR spectra were obtained at 79.5 MHz using a 7 mm zirconia rotor and a spinning frequency of 5 kHz. A pulse duration of 4.00 μs and a pulse delay of 60 s were used. The ²⁹Si chemical shift is reported with respect to Kaolin standard (δ = −91 ppm). Textural analysis was done by N₂ adsorption–desorption at 77 K using a Quadrasorb SI automated gas adsorption system. Prior to the analysis, the samples were degassed under vacuum at 300 °C overnight. The specific surface area was estimated using the BET method, micropore volume by using the t-method of DeBoer, and the pore size distribution upon both the BJH and density functional theory (DFT) models. X-ray photoelectron spectra (XPS) were recorded on a VG ESCALAB250 X-ray photoelectron spectrometer at a base pressure in the analysis chamber of 2 × 10⁻⁹ Pa using Al Kα source (1486.6 eV) and the binding energies corrected using C 1s line of aliphatic carbon contamination at 285.0 eV. SEM/EDX results were obtained on an Oxford Instruments INCAx-act EDX detector attached to a Zeiss Supra 55 field emission scanning electron microscopy using a 20 kV electron beam and 60 s acquisition time. Transmission electron microscope (TEM) and HRTEM images were obtained with Hitachi-800 and JEM-2010, operated at an accelerating voltage of 200 kV. Temperature-programmed desorption of NH₃ was performed on a Thermo Electron Corporation TPD/R/O 1100 series instrument. Prior to NH₃ sorption, 0.1 g sample was out-gassed at 300 °C in pure N₂ flow (20 mL min⁻¹) for 1 h. Then, the temperature was naturally cooled to 100 °C, and NH₃ was adsorbed by exposing sample to a stream of pure NH₃ for 0.5 h. It was then flushed at 100 °C with N₂ for another 0.5 h to remove physisorbed NH₃. The desorption of NH₃ was carried out in pure N₂ flow (20 mL min⁻¹) by increasing the temperature to 600 °C at a heating rate of 10 °C min⁻¹. The pyridine adsorption FT-IR studies were carried out on a Nicolet 380 FT-IR spectrometer. The IR spectra (in absorbance mode) were recorded using self-supporting pellets of the sample powder. The pellet (diameter 13 mm) of 70 mg was placed in an IR cell designed to carry out spectroscopic measurement at varied temperatures and equipped with CaF₂ windows. The samples were first heated to 400 °C at a heating rate of 10 °C min⁻¹ in a pure N₂ flow and kept for 2 h and then cooled to room temperature. The pyridine vapor was introduced under N₂ flow for 1 h; subsequently weakly adsorbed pyridine was flushed for 0.5 h under N₂ flow. The FT-IR spectra were recorded at 100, 200, 300, 400 °C and a resolution of 8 cm⁻¹ is attained after averaging over 64 scans for all the FT-IR spectra reported here. The UV-vis diffuse reflectance spectra (DRS) were obtained on a Hitachi U-3900 VIS spectrophotometer equipped with the Varian diffuse reflectance accessory in 200–800 nm. Barium sulfate was used as a reference material.

2.3 Catalytic studies

The liquid phase benzoylation reactions of anisole with benzoyl chloride (BzCl) was performed in a 50 mL three-necked glass flask, equipped with a magnetic stirrer, a reflux condenser, a thermometer and a CaCl₂ guard tube. Reaction conditions: anisole (12 mL, 108 mmol), BzCl (324 μL, 2.65 mmol), the catalyst (0.2 g), n-tetradecane (as a GC internal standard, 0.1 g, 99.5% purity), reaction temperature, 110–140 °C, reaction time, 3–7 h. The liquid phase (0.2 μL) was analyzed before and after reaction by an Agilent 7890A GC, equipped with an Agilent J & W HP-5 capillary column (5% phenyl polysiloxane, 30 m × 0.32 mm × 0.25 μm) and a flame ionization detector (FID: 340 °C, injector: 320 °C, temperature program: 120 (3 min) to 220 °C (1 min) by 20 °C min⁻¹). Conversion is given as the conversion into 4-methoxybenzophenone (4-MBP) and 2-methoxybenzophenone (2-MBP); benzoic acid (BA) and phenyl benzoate (PB) were the only by-products. The selectivity to the desired product is given as the amount of 4-MBP divided by the amount of total products. The control experiment confirmed that the reaction does not take place in the absence of catalyst. The used catalyst was separated from reaction system by centrifugation (5000 rpm), washed thoroughly with absolute ethanol, dried at 120 °C overnight, and subjected to next run. Commercial AlCl₃ was used for comparison. The product was identified by ¹H NMR (Brucker Avance 600 MHz). As shown in Fig. S2,† the main product was confirmed as 4-MBP (small amount of 2-MBP may be evaporated in evaporation step due to its similar boiling point to anisole).

3. Results and discussion

3.1 Crystal structure and textural properties

Fig. 1 shows the XRD patterns of the PMH-xZr catalysts in small and wide 2θ range. The XRD of the intermediates after each step and pristine magadiite are shown in Fig. S3A.† The XRD of the magadiite shows a strong peak at 2θ 5.72° (d₀₀₁ = 1.54 nm) and five-finger-shaped peaks in 20–30° indicating the formation of well-defined magadiite (JCPDS 42-1350).^1,15,16 The QM shows an enhanced d₀₀₁ of 2.96 nm on intercalation of CTMA⁺. Most other peaks of the magadiite layers seem to suffer from coalescence into broad bands, a phenomenon reminiscent of the turbostraticulate stacking occurred for many clay minerals. Then, a large excess of co-surfactant OA was added to stable the swelling gallery favoring the 2D intragallery hydrolysis–condensation–polymerization of subsequently added TEOS/Zr-nPro upon a hydrated silica structures formed around a monolayer of micellar CTMA⁺ and OA assemblies.¹ Then the QOTM precursor shows a similar d₀₀₁ of 2.94 nm to QM, while the Zr-doped samples QOTM-xZr show gradually increased interlayer space from 3.09 to 3.36 nm with Zr-doping, implying the incorporation of Zr into the intragallery silica framework given the thickness 1.12 nm of magadiite layer.^8,15

Fig. 1 XRD patterns of PMH (a) and PMH-xZr (b–f: x = 0.0125, 0.025, 0.05, 0.1, 0.2) at small (A) and wide (B) angle range.

After calcination, the obtained pure PMH shows further expanded d₀₀₁ as 3.38 nm, together with broadened peak shape, which is probably due to the stabilization role of the co-templates though with somewhat disturbance of layer ordering owing to the removal of the gallery cosurfactants during calcinations for the completion of dehydroxylation and cross-linking of intragallery assembled silica structure.^1,12 This expanding effect is well-retained in the final PMH-xZr samples as revealed by a clear broad (001) peak in low 2θ angle though with reduced intensity probably due to the partial loss of crystallinity upon the thermal decomposition of interlayer organic moieties. With increasing Zr-doping (x = 0.0125–0.1), the (001) lines remain clear and the d₀₀₁ value increases and reaches the maximum 3.82 nm for PMH-0.05Zr, and then slightly reduced to 3.69 nm for PMH-0.1Zr, implying the uniformly ordered incorporation of zirconium into the interlayer mesostructure of PMH. While PMH-0.2Zr shows almost completely disappeared (001) line, implying the delaminated structure due to its maximal Zr-doping, similar to previously reported PCHs with titanium and zirconium.^12,24 These results indicate that the incorporation of Zr into the lattice of interlayer meso-structural silica strongly affects the order of the stacking layers along the c-axis, causing a dramatic decrease of the coherent diffraction domain along this direction with increased Zr-doping. Note that silica-based PMH with proper Zr-doping can hold well-retained layered structure assembled with intragallery mesostructure such as PMH-0.1Zr and PMH-0.05Zr.

Fig. 2 depicts the FT-IR spectra of PMH and PMH-xZr samples, while those of the intermediates and magadiite shown in Fig. S3B.† It can be seen from Fig. S3B† that magadiite shows a broad band at ∼3440 cm⁻¹ assigned to ν(OH) and a small sharp one at 3663 cm⁻¹ to free OH group,^25,26 implying the presence of terminal silanol groups.²⁷ The intense band at 1083 cm⁻¹ can be assigned to ν_as(Si–O–Si), whereas two overlapped peaks at 820 and 781 cm⁻¹ to ν_s(Si–O–Si) involving silicon motions, two ones at 620 and 573 cm⁻¹ to the bending modes of single and double Si–O–Si rings of SiO₄ unit,^26,27 and a sharp one at 462 cm⁻¹ to ν_s(Si–O–Si).²⁶ For QM, two sharp peaks at 2852 and 2921 cm⁻¹ are due to ν_s(C–H) and ν_as(C–H) of –CH₂– groups, respectively, and a sharp one at 1469 cm⁻¹ to CH₃ antisymmetric deformation modes and CH₂ scissoring modes, implying the intercalation of CTMA⁺. For QOTM and QOTM-xZr samples, strong IR bands for organic groups are clearly seen due to the introduction of cosurfactant and inorganic precursors. While the strong ν_as(Si–O–Si) at 1083 cm⁻¹ of magadiite clearly reduces and slightly downshifts to ∼1079 cm⁻¹, and two bands at 820 and 781 cm⁻¹ to ν_s(Si–O–Si) show similar decrease in intensity and merge in one peak with gradually reduced wavenumbers from 808 cm⁻¹ of QOTM to 803 cm⁻¹ of QOTM-0.2Zr with Zr-doping. These phenomena may be due to the formation of Si–O–Zr linkages, similar to those occurred in aluminated LAPONITE®-derived PCHs²⁸ or Zr-doped silica,²⁹ upon the intercalation of silica or Zr-doped silica molecular sieve-like clusters between the magadiite layers.

Fig. 2 FT-IR spectra of PMH (a) and PMH-xZr samples (b–f: x = 0.0125, 0.025, 0.05, 0.1, 0.2) (inset refers to the amplified IR spectra between 4000–3200 cm⁻¹).

As for PMH and PMH-xZr (Fig. 2), a weak broad band in 3420–3450 cm⁻¹ can be assigned to ν(OH) and a quite small sharp one at 3745 cm⁻¹ (inset in Fig. 2) to the terminal Si–OH groups.^27,30 While a strong band at 1086 cm⁻¹ accompanied by a shoulder at 1206 cm⁻¹ of PMH can be ascribed to ν_as(Si–O–Si) and Si–O–Si stretching modes of five-membered rings in calcined samples,^25–27 respectively, implying considerable structure stability of magadiite sheets in PMH materials. Interestingly, the sharp Si–OH signals at ∼3745 cm⁻¹ of PMH are variously reduced for the PMH-xZr samples with increasing Zr-doping, implying the incorporation of Zr into the interlayer space of PMH which may cause varied amount of Si/Zr–OH groups probably upon the formation of Si–O–Zr linkages. Moreover, the intensity of the band at 1083 cm⁻¹ and its fair-defined shoulder (1237 cm⁻¹) due to the Si–O–Si framework is greatly reduced, and gradually becomes narrow with slight downshift as Zr-doping is increased especially for the samples with high Zr content (x > 0.05). Typically for PMH-0.1Zr and PMH-0.2Zr, the ν_as(Si–O–Si) occurs at ∼1077 cm⁻¹. Two other bands, at 811 and 454 cm⁻¹ for PMH, characteristic of Si–O bonds in a ring structure,^26,27,29 exhibit similar band downshifts and intensity decrease for PMH-xZr with Zr-doping, also reflecting a gradual increase of Zr in the Si–O network in forming Zr–O–Si hetero-linkages. While PMH-0.2Zr shows the minimum absorption intensity and the maximum band downshift in the ring structural Si–O bands (799 and 449 cm⁻¹) corresponding to the delaminated structure of the sample. These results are probably due to the superimposition of stretching and rocking vibrations of Si–O–Si and Si–O–Zr bonds.^20,31

Fig. 3 shows the ²⁹Si MAS-NMR spectra of pure PMH and PMH-xZr samples. All the spectra present a broad signal between −80 and −140 ppm, which can be deconvoluted to three main components with chemical shifts at ca. −92, −102, and −111 ppm resulted from Q², Q³ and Q⁴ silicon nuclei, respectively, where the Qⁱ corresponds to silicon nuclei with i siloxane linkages, i.e., Q² to disilanol Si(OSi)₂(OR)₂, where R is H or Zr, Q³ to silanol Si(OSi)₃(OR), and Q⁴ to Si(OSi)₄ in the framework. The pure PMH exhibits a prominent Q⁴ Si(OSi)₄ signal near −111.1 ppm, a moderate Q³ Si(OSi)₃(OH) one near −102.0 ppm and a weak Q² Si(OSi)₂(OH)₂ near −92.3 ppm.^17,18,32 The ²⁹Si MAS NMR spectra of the PMH-xZr samples are quite similar to that of pure PMH. One may see that the samples modified with low Zr-doping (x < 0.2) show a larger population of the Q² and Q³ coordination, thus a bigger value of (Q³ + Q²)/Q⁴, that strongly indicates more Si ions being replaced by Zr ions in interlayer silica framework. While the value of (Q³ + Q²)/Q⁴ of PMH-0.2Zr reduces to 0.33 (Table S2†), similar to the value of PMH, being attributed to the higher Zr-doping leading to small amount of extra-framework Zr oxides (also verified by later UV-vis DRS). Therefore, it can be deduced upon the ²⁹Si MAS-NMR and FT-IR data that the zirconium atoms are incorporated into the silica framework via Si–O–Zr linkages in the PMH-xZr samples. This is the first report on the fabrication of Zr-doped PMH via in situ 2D intragallery cosurfactant-directing TEOS/Zr-nPro hydrolysis–condensation–polymerization method.

Fig. 3 ²⁹Si MAS NMR spectra for PMH (a), PMH-0.0125Zr (b), PMH-0.05Zr (c), PMH-0.1Zr (d) and PMH-0.2Zr (e).

The N₂ adsorption–desorption isotherms (Fig. S4a†) of the PMH-xZr samples clearly display a type IV isotherm with a H4 hysteresis loop, implying a certain mesoporosity related to the slit-shaped pores (IUPAC).³³ Interestingly, a gradual increase in N₂ sorption observed in the relative pressure (p/p₀) of 0.05–0.3 suggests the presence of supermicropores and small mesopores.^1,5,33 The BJH pore size distributions (PSD) indicate that the six distributions all mainly center at ∼1.89 nm (Fig. S4b†). Considering the tensile strength effect,³⁴ that is, the current pore data ∼1.89 nm might be the combined effect of the micropores and mesopores, BJH model is not completely suitable for the samples. It is noted that all the desorption inflection points (Fig. S4a†) are very smooth and desorption hysteresis loops at p/p₀ < 0.25 are not closed, strongly indicating the possible nitrogen fills in micropores.^33,35 Therefore, another PSD calculation model, density functional theory (DFT) model usually applied to characterize mesoporous materials containing micropores such as mesoporous carbon,³⁶ is used to illustrate the PSD. The DFT distributions (Fig. S4c†) show obvious existence of the multi-modal pore structure, involving some supermicropores with optimal size of ∼0.84, 0.92, and 1.15 nm, and small mesopores at ∼1.89–1.97 nm near to the gallery height of samples upon XRD. The quite similar average pore size implies the uniform Zr-incorporation into the interlayer meso-silica framework except PMH-0.2Zr with a little larger one possibly due to its delaminated layers.

The detailed textural data are listed in Table 1. The specific surface areas (S_BET) of PMH-xZr samples are greatly increased with Zr-doping. The S_BET reaches to the maximum 603.3 m² g⁻¹ with the relatively larger pore volume of 0.40 mL g⁻¹ for PMH-0.05Zr, implying more terminal oxygen atoms from the bridging ones, leading to structural stabilization of SiO₄ tetrahedral in an interlayer 3D silica network.²⁹ Then the S_BET is contrarily reduced with further increased Zr-doping probably due to the increased incorporation of larger Zr⁴⁺ ion (radius 0.084 nm) than Si⁴⁺ (0.026 nm) into the gallery SiO₂ framework leading to obviously enhanced interlayer packing density of SiO₂–ZrO₂ molecule sieve-like clusters. The great S_BET drop for PMH-0.2Zr, consistent with its lower amount of super-micropores than others, suggests the presence of some amorphous ZrO₂-like species (verified by later UV-vis DRS) causes delamination of the layered PMH material as XRD showed. Clearly, PMH-0.05Zr and PMH-0.1Zr simultaneously possesses relatively larger S_BET and Zr-doping and well-kept layered supermicro-mesoporous heterostructures.

Table 1 Textural data and related structure parameters of the catalysts

Samples	SBET/m^2 g^−1	Smico^a/m^2 g^−1	VT/cm^3 g^−1	Vmicro^a/cm^3 g^−1	RBJH/nm	RDFT/nm	GH^b/nm	Zr/Si^c
a Microporous area and volume obtained by t-plot method.b Gallery height (GH) obtained by subtracting the thickness 1.12 nm of magadiite layer from d001 upon XRD analysis.c Based on EDX analysis.
PMH-0.2Zr	242.8	179.1	0.241	0.093	1.92	1.97	—	0.130
PMH-0.1Zr	401.6	340.8	0.288	0.174	1.89	0.92, 1.15, 1.89	2.571	0.056
PMH-0.05Zr	603.3	458.8	0.400	0.227	1.89	0.84, 0.92, 1.15, 1.89	2.705	0.038
PMH-0.025Zr	557.7	487.5	0.388	0.246	1.89	0.84, 0.92, 1.15, 1.89	2.492	0.021
PMH-0.0125Zr	533.6	457.8	0.395	0.231	1.89	0.84, 0.92, 1.16, 1.89	2.460	0.013
PMH	442.8	353.3	0.354	0.177	1.91	0.84, 0.92, 1.15, 1.90	2.260	—

---

3.2 Morphology

Fig. 3 shows the SEM/EDX images of PMH-xZr samples. The morphologies of magadiite and intermediates are shown in Fig. S5 and S6.† Clearly, magadiite exhibits rosette morphology of ∼15 μm (inset in Fig. S5†) constructed by contact packing of well-defined plates, similar to previously reported.³² The QM shows loose discrete rectangle plates with inhomogeneous sizes originated from broken rosette morphology of magadiite due to the intercalation of CTMA⁺. The HRTEM images further reveal the (001) lattice fringes as 1.54 and 2.72 nm for magadiite and QM, respectively, in line with the XRD data. Then, the QOTM shows considerably swelling plate-like morphology, though the plates are quite irregular and some show an edge-curved shape due to the hydrolysis–condensation–polymerization of TEOS in 2D gallery of the magadiite. The morphological similarity of QOTM to QM implies that the intercalation of cosurfactants and TEOS occurs in a topotactic fashion. The SEM of Zr-doped precursors QOTM-xZr exhibit similar platy morphology to QOTM except QOTM-0.2Zr showing very thin broken layers due to its delaminated layers upon its largest Zr-doping content.

On calcinations, the PMH (Fig. 4a) shows flake like layers with a smooth surface and regular packing probably due to the decomposition of interlayer organic groups and proper layer rearrangement during the calcinations. The EDX of PMH (inset in Fig. 4a) indicates the existence of Si and O with Si/O ratio of 4/8.1, higher than that of pristine magadiite (4/8.2), implying the intercalation of ordered meso-silica into the interlayer space of magadiite. However, the SEM images of PMH-xZr samples (Fig. 4(b)–(f)) clearly describe an irregularly shaped aggregate of rugged and irregular platy particles. Differently, PMH-0.2Zr shows obviously isolated irregularly shaped small platy particles, in line with its possible delaminated layer structure as the XRD shown. These morphological changes of the PMH-xZr samples are more likely due to the incorporation of Zr as silica–zirconia molecular sieve-like clusters into the interlayer space of the magadiite. A typical EDX of PMH-0.1Zr (inset in Fig. 4e) clearly indicates the incorporation of Zr in PMH.

Fig. 4 SEM/EDX results of PMH (a) and PMH-xZr (b–f: x = 0.0125, 0.025, 0.05, 0.1, 0.2).

Fig. 5 presents the TEM and HRTEM images of PMH-0.05Zr and PMH-0.1Zr. The HRTEM images indicate that the gallery space appears in light in contrast to the dark layers and the mean thickness of the dark layer is 1.25 and 1.23 nm for PMH-0.05Zr and PMH-0.1Zr, respectively, similar to ∼1.18 nm of magadiite layer (Fig. S5†) and literature value 1.12 nm,^14,32 strongly implying the transformation of the precursors into PMH-0.05Zr and PMH-0.1Zr with no change of original layer structure of the starting magadiite. The (001) lattice fringes of 3.25 and 3.12 nm for PMH-0.05Zr and PMH-0.1Zr, respectively, are in good agreement with the XRD data.

Fig. 5 (HR)TEM images of PMH-0.05Zr (a and b) and PMH-0.1Zr (c and d) and their corresponding FFT diffractograms (insets).

3.3 Surface acidity

NH₃-TPD was used to study the acid strength distribution of the samples and, furthermore, to obtain the quantitative amount of acid sites in the specified temperature range. The pyridine FT-IR was performed to determine the contribution of Brønsted (B) and Lewis (L) acid sites.

It is generally accepted that the acid strength depends on the NH₃ desorption temperature: weak (100–200 °C), medium (200–400 °C) and strong (>400 °C).³⁷ The NH₃-TPD profile of PMH (Fig. 6) shows almost undetectable desorption peak, implying its very small amount (0.071 mmol g⁻¹), similar to previously reported PCHs from montmorillonite and saponite et al.^1,12 The removal of surfactant from the as-prepared PMH by calcination leads to the formation of protons, which are necessary to balance the clay layer charge, resulting in pure PMH intrinsically acid. However, all PMH-xZr samples (Fig. 6) exhibit clearly enhanced acidity. The low Zr-doped samples PMH-xZr (x = 0.0125 and 0.025) show only one maximum at ∼160 °C with clearly enhanced TCD signal, implying the formation of weaker acid sites. With increased Zr-doping, PMH-0.05Zr, PMH-0.1Zr and PMH-0.2Zr present appropriate two-maxima type of TPD-profile, i.e. two maxima at ∼190 °C and 360 °C, implying the formation of weak and medium-strong acid sites. These curves are broad and the complete NH₃-desorption is not reached even at 500 °C, suggesting a heterogeneous distribution of acid strengths, though the weak-to-medium acid sites are majority considering the ammonia desorption mainly occurred between 100 and 450 °C. It is noted that the acidity is increased with the amount of Zr incorporated to the framework, being the PMH-0.1Zr the most acidic among the PMH-xZr samples (Table 2), and the lower acidity of PMH-0.2Zr with the most Zr-doping may be due to the delaminated layer structure. Clearly, all PMH-xZr samples exhibit a total acidity much higher than PMH, pure mesoporous silica,³⁸ Zr-doped mesoporous silica,³⁹ a hexagonal ordered mesoporous silica grafted with 20 wt% ZrO₂ (ref. 40) and a ZrO₂-pillared montmorillonite.⁴¹ The presence of acid sites is mainly due to the incorporation of Zr into the interlayer mesoporous silica network where superficial Zr⁴⁺ species with low coordination exist. Generally, the high concentration of acid sites exposed on the catalyst can be related to the high surface area. However, in the present system, the highest acid amount is found for PMH-0.1Zr (0.968 mmol g⁻¹), though it only shows relatively larger S_BET (401 m² g⁻¹). This phenomenon can be explained as that PMH-0.1Zr simultaneously holds higher Zr-doping and relatively higher S_BET closely related to its ordered packing of interlayer SiO₂–ZrO₂ clusters in largely extended gallery space as FT-IR, XRD and HRTEM showed, thus the maximum concentration of acid sites. It should be noted that chemisorbed NH₃ (assuming 0.15 nm² occupied per NH₃ molecule¹²) covers the fraction of the sample surface lower than 22% except delaminated PMH-0.2Zr assuming the formation of monolayer (Table 2).

Fig. 6 NH₃-TPD profiles of PMH (a) and PMH-xZr samples (b–f: x = 0.0125, 0.025, 0.05, 0.1, 0.2).

Table 2 Acid concentration and acid distribution of the samples upon NH₃-TPD and pyridine-IR

Samples	Weak-medium/mmol g^−1	Strong/mmol g^−1	Total acidity/mmol g^−1	Acidity density^a/NH3/nm^2	Coveragesurf. ^bby NH3/%	TDesorp./°C	B sites/mmol g^−1	L sites/mmol g^−1	B/L ratio
a Acidity density (NH3 molecules per nm^2) = [acidity (mol g^−1) × NA/SBET (nm^2 g^−1)].b Percentage of the surface covered by a monolayer of chemisorbed ammonia. It was assumed that the surface occupied by one NH3 molecule is 0.15 nm^2.^12
PMH	0.071	—	0.071	0.097	1.4	200	—	—	—
PMH-0.0125Zr	0.480	—	0.480	0.542	8.1	200	—	0.223	—
PMH-0.025Zr	0.552	—	0.552	0.596	9.0	200	0.054	0.284	0.19
						300	—	0.081	—
PMH-0.05Zr	0.506	0.238	0.744	0.743	11.1	200	0.065	0.358	0.18
						300	—	0.155	—
						400	—	0.127	—
PMH-0.1Zr	0.722	0.246	0.968	1.452	21.8	200	0.071	0.373	0.19
						300	0.056	0.205	0.27
						400	—	0.161	—
PMH-0.2Zr	0.565	0.23	0.795	1.972	29.6	200	0.070	0.351	0.20
						300	0.040	0.180	0.22

---

Pyridine FT-IR was used to further probe the nature of the acidity of the samples (Fig. 7). The amount of both B and L sites was derived from the intensity of the band at 1540 cm⁻¹ (PyH⁺ pyridinium ions) and 1445 cm⁻¹ (pyridine bonded onto Lewis sites, PyL) with extinction coefficient 0.73 and 1.11 cm mol⁻¹,^39,42 respectively. The PMH-xZr samples with x in 0.025–0.2 contains both Brønsted and Lewis sites, however the former are much weaker than the latter. Particularly, PMH-0.1Zr shows the maximum amount and strength of acid sites of the PMH-xZr catalysts, in accordance with its NH₃-TPD result.

Fig. 7 Concentration of surface acid sites determined for PMH-xZr (x = 0.025, 0.05, 0.1, 0.2). Insets, IR spectra of the samples pre-adsorbed with pyridine.

Thermodesorption of pyridine was followed by FT-IR to estimate the acid strength of both Brønsted and Lewis sites. The position of the pyridine IR bands is independent on the strength of the site, and depends only on the type of bonding, therefore the relative intensities of the bands of pyridine were taken to estimate the relative acid strength of acid sites in the examined samples. The FT-IR of PMH (Fig. S7†), after pyridine adsorption and outgassing at 100 °C shows two weak absorptions at 1444 and 1592 cm⁻¹, assigned to pyridine adsorbed on weak Lewis sites or to weakly hydrogen-bonded pyridine,³⁹ which are nearly vanished after outgassing at 200 °C, indicating the trace or no acid sites on pure PMH. However, for PMH-xZr samples with x in 0.025–0.2 (Fig. 7), the IR spectra present not only greatly enhanced absorptions at 1445 and 1592 cm⁻¹, but also a gradually strengthened band at ∼1538 cm⁻¹ due to pyridine bonded to B sites (pyridinium ions), and a band at 1489 cm⁻¹ to a combined contribution from both B and L sites,^39,43 implying the formation of both B and L acid sites after Zr-doping. These bands can be clearly seen after evacuation at 200 °C. With further increased temperatures, the intensities of the characteristics of both B and L sites reduce, quite similar to the Zr-modified silica reported by Teo and Zeng.²⁹

It is well-known that pure silica framework of MCM-41 and ZrO₂ are Lewis solids without B acid site,^44,45 and the fact that the pure PMH nearly does not hold B and L sites; therefore, the B and L sites of the PMH-xZr samples are caused by the presence of Zr atoms replaced the Si ones. Given that the Zr⁴⁺ ions has much larger radius (0.084 nm) than Si⁴⁺ (0.026 nm); when the smaller Si⁴⁺ ions are replaced by the larger Zr⁴⁺ ions in the framework, the bond length of Zr–O–Si clearly differs from that of Si–O–Si, and this phenomenon must leads to structural microstrain within the lattice cell. Changes in the electron density around Si, due to charge unbalance, or differences in electronegativity or local structure deformation resulting from the introduction of Zr⁴⁺ ion into the vicinity of the hydroxyls carrying Si, may weaken the SiO–H bond; this is one of the possible origins resulting in the B acid sites on the Zr-modified PMH materials. Table 2 gives the contributions of B and L sites of the PMH-xZr samples. The considerably created B sites and greatly increased L sites of the PMH-xZr are highly desirable for extended solid acid catalytic applications.

3.4 Nature of the acidity

UV-vis DRS is commonly used to detect the framework and extra-framework Zr species in hybrid solids, particularly dispersed oxides and metal ions in constrained environment such as silica,⁴⁶ zeolites and alumosilicate materials⁴⁷ to obtain information on coordination environment, oxidation state of the embedded transition metal ions. As shown in Fig. 8, PMH-0.1Zr and PMH-0.2Zr exhibit a strong absorption at 207 nm and a broad band at 275 nm with increased strength in the latter. The former can be attributed to the ligand-to-metal charge transfer (LMCT) from an O²⁻ (2p) to an isolated Zr⁴⁺ (4d) ion in a tetrahedral configuration,^46,48 while the later assigned to small Zr_xO_y clusters in the extra-silica framework (probably seven-coordinate Zr species like those of monoclinic ZrO₂) specially for PMH-0.2Zr.⁴⁸ While PMH-xZr with low Zr-doping (x < 0.05) only show a broad and quite weak band covering 200–350 nm compatible with that of nearly isolated eight- or seven-fold Zr⁴⁺ ions in oxides,^48,49 and approximately detected band near 200 nm especially for PMH-0.05Zr due to the LMCT transitions. These results clearly indicate that Zr species in PMH-xZr are mainly incorporated into the interlayer silica–zirconia framework and the higher Zr-doping leading to small amount of extra-framework zirconium oxides, in line with the above ²⁹Si MAS NMR results.

Fig. 8 UV-vis DRS spectra of the PMH-xZr catalysts (a–e: x = 0.0125, 0.025, 0.05, 0.1, 0.2).

As a powerful technique to distinguish the nature of Zr ions presented in micro- and mesoporous Zr substituted molecular sieves and similar materials,^19,50 XPS analysis was employed to further explore the nature of the acid sites of the catalysts. Fig. 9 shows the Zr 3d XPS spectra of PMH-xZr samples, while the corresponding Si 2p and O 1s XPS shown in Fig. S8.† With Zr-doping, Si 2p and O 1s binding energy (BE) maintain constant for low Zr-doped samples (x < 0.05) and slightly shift to lower BE for high Zr-doped samples compared to pure PMH, and Zr 3d exhibits similar changing trend, implying the slight change of electron density around Si, O and Zr cores with Zr-doping. In detail, obviously asymmetric O 1s peak can be deconvoluted into two components, implying varied environments for oxygen, i.e. Si–O–Si (∼533.04 eV) and Si–O–Zr (∼530.76 eV) (Table S3†). The intensity/area ratios of these two peaks closely match to the surface Si/Zr ratio. In such Zr-substituted analogues of siliceous PMH materials, the question arises as to the true insertion of Zr into the silica framework, as opposed to its presence as a segregated oxide phase. Here, the BE of O 1s for PMH-0.1Zr at ∼530.76 eV (Table S3†) is at least 0.56 eV higher than that of pure ZrO₂ (530.2 eV),⁵⁰ strongly evidencing the presence of framework Zr via Si–O–Zr linkage. Moreover, the Zr 3d_5/2 peak shifts to high BE (183.47 eV) compared to ZrO₂ (Zr 3d_5/2 182.6 eV),⁵¹ also implying the formation of Si–O–Zr bond. Meantime, the higher Zr 3d_5/2 BE of PMH-0.1Zr indicates a higher ionic character of the bonding Zr–O^39,52 which benefits for creating a stronger acidity in the catalyst as pyridine-IR revealed.

Fig. 9 Zr 3d XPS of the PMH-xZr catalysts (a–e: x = 0.0125, 0.025, 0.05, 0.1, 0.2).

Comparing with bulk composition Zr/Si ratio of the samples from EDX, the surface Zr/Si ratios (Table S3,† XPS peak area ratios normalized by the atomic sensitivity factors) upon XPS data suggests that Si-rich on the surface is observed for the low Zr-doping samples as PMH-0.0125Zr and PMH-0.025Zr, while Zr-rich on the surface for the highest Zr-doping sample PMH-0.2Zr, and PMH-0.05Zr and PMH-0.1Zr shows similar bulk and surface Zr/Si ratio, implying their best interlayer Zr-doping silica molecular sieve-like heterogeneous structure. Specially, the surface Zr/Si ratio of PMH-0.1Zr is 0.077, close to the theoretical bulk value 0.056, revealing no significant elemental enrichment on the surface and the homogeneous doping of Zr into the silica matrix. This result is quite similar to an XPS observation for the ZrO₂–SiO₂ system prepared using Zr-n-propoxide and TEOS in n-propanol,²⁹ but different from the oxide gel system upon ZrOCl₂ and TEOS in ethanol, in which a great enrichment of Si ions on the surface is found for the sample with Zr/Si ratio 0.1 (17.0 wt% as ZrO₂), whereas a surface Zr enrichment is observed for the sample with Zr/Si ratio 0.05 (9.3 wt% as ZrO₂).⁵³

Based on above analysis, it can be deduced that the proper Zr-doping into interlayer silica framework renders the PMH-xZr materials unique layered supermicro-mesoporous structure and greatly enhanced B acid sites related to Si–O(H)–Zr linkage and L acid sites (Si–O–Zr), strongly implying the efficient acidic tunability of transition metal-doped PMH materials on in situ 2D-magadiite interlayer co-surfactant directing hydrolysis–condensation–polymerization method for acid catalysis process.

3.5 Catalytic activity

The Friedel–Crafts acylation reaction is of great importance in fine chemical industry owing to create functionalized aromatic ketones such as 4-MBP (para-methoxybenzophenone) from the benzoylation of anisole with BzCl (Scheme 2) mainly used as a valuable perfumery intermediate and a precursor for antioxidants often used in cosmetics, unsaturated polyesters, PVC, and alkylated resins.^54–56 The catalytic activity of the catalysts was tested in liquid phase benzoylation of anisole. Primarily, the dependency of the conversion degree of the acylation on the reaction time and temperature and reactants ratio were performed over PMH-0.025Zr. It can be seen from Fig. S9† that the benzoylation activity shows pronounced dependent on reaction temperature and mole ratio of reactants. With constant anisole/BzCl ratio of 40 : 1, the BzCl conversion is increased from 75.6% at 110 °C to 93.5% at 140 °C after 3 h. The 4-MBP yield reaches ∼93.5% at 140 °C after 5 h, and slightly increases after 7 h of reaction. The BzCl conversion increases substantially with reaction temperature, suggesting that the reaction was intrinsically kinetically controlled. If one takes into account the higher conversion and shorter reaction time, the reaction temperature of 140 °C and the reaction time of 3 h should be the better option. Then, with a varied anisole/BzCl ratio (7 : 1), a significant decrease in both BzCl conversion and reaction rate is observed at 140 °C, and the conversion of BzCl only reaches 42.4% after 3 h, much lower than those with anisole/BzCl ratio of 40 : 1. This can be explained as that the lower anisole/BzCl ratio, i.e. the higher concentration of BzCl make the active sites of the catalyst become saturated with BzCl, leading to quantitatively low formation of active electrophilic benzoylinium to react with anisole existed in bulk phase, thus showing a decrease in conversion of BzCl and rate of reaction. Therefore, the reaction temperature of 140 °C, reaction time of 3 h and anisole/BzCl ratio of 40 : 1 is selected as the optimal reaction conditions.

Scheme 2 Friedel–Crafts benzoylation reaction of anisole with benzoyl chloride (BzCl).

A series of novel PMH-xZr catalysts was evaluated at optimal reaction conditions and the detailed activity data are summarized in Table 3 (entries 1–6). Beside the ketones, BA and PB were also formed as side-products. The pure PMH shows quite low BzCl conversion of 58.4% probably due to its quite low intrinsic B acidity. Then for the Zr-doped sample PMH-0.0125Zr, greatly enhanced BzCl conversion of 80.4% can be attributed to its greatly increased total acidity especially B sites resulting from the formation of Zr–OH and Zr–O–Si–OH linkages though pyridine IR fail to detect the B acid site due to steric hindrance for larger pyridine molecules. Then for PMH-0.025Zr, further increased BzCl conversion of 93.5% can be attributed to its further increased acidity especially detectable B sites. While further increasing Zr-doping, PMH-0.05Zr and PMH-0.1Zr show BzCl conversion of 98.5% and 99.5%, respectively. The comparable BzCl conversion of the former to the latter can be ascribed to the higher specific surface area thus more available B acid site to reactants compensated to its lower acidity than the latter. As for the contrarily reduced BzCl conversion of PMH-0.2Zr (89.1%), it can be attributed to its reduced available acid sites upon delaminated layer structures as XRD indicated. It is also noted hat the PMH-0.05Zr and PMH-0.1Zr even present BzCl conversion of 98.5% and 99.5%, quite near to conventional AlCl₃ (99.9%), implying excellent distribution of acid sites on these two catalysts, in line with their NH₃-TPD data. It is generally believed that benzoylation activity of anisole with BzCl is dependent on both typical Lewis acids (e.g. AlCl₃) and Brønsted acids (e.g. H₂SO₄) in homogeneous phase. While in heterogeneous system, Quaschning et al.⁵² reported benzoylation activity of anisole with BzCl over SO₄²⁻/ZrO₂ and pointed out that reaction exclusively proceeds at Brønsted acid sites, and Sakthivel et al.⁵⁷ reported that sulfation of ZrO₂ predominantly generates B acidity responsible for catalytic activity in the benzoylation reaction over SO₄/Zr_1−xSn_xO₂. Given that pure ZrO₂ has no activity in the present reaction system, the excellent benzoylation activity of anisole with BzCl over the novel PMH-xZr catalysts can be attributed predominantly to the Brønsted acid sites generated upon Zr-doping to PMH, while the Lewis acids show less influence on conversion of the benzoylation agents into the desired aromatic ketones under the applied conditions.

Table 3 Friedel–Crafts benzoylation reaction results of anisole with benzoyl chloride on the catalysts^a

No.	Catalysts	Conv./%	Sel./%				P/O	Yield4-MBP/%	Activity^b/mmol g^−1 h^−1	Refs
			4-MBP	2-MBP	BA	PB
a Reaction conditions: anisole 108 mmol (11.7 g), BzCl 2.65 mmol (0.373 g), catalysts loading (Wcat.) 0.2 g (0.016 g cm^−3), reaction temperature 140 °C, reaction time 3 h.b Activity (mmol g^−1 h^−1) = [BzCl (mmol) × conversion]/[mass of catalyst (g) × reaction time (h)].c Anisole 108 mmol (11.7 g), BzCl 15.88 mmol (2.24 g), Wcat. 0.4 g, reaction temperature 110 °C, reaction time 1 h.d From ref. 52 reaction time 7 h, others same as this work.e From ref. 57 reaction temperature 60 °C, others same as this work.f From ref. 55, anisole 270 mmol (29.3 g), BzCl 39.7 mmol (5.59 g), Wcat. 1.05 g (0.03 g cm^−3), 37% Conv. obtained at 70 °C and 2 h, 100% Conv. at 110 °C and 4 h. Activity was estimated upon 37% BzCl Conv. at 70 °C and 2 h.
1	PMH	58.4	92.3	5.6	1.8	0.3	16.5	53.9	2.6	This work
2	PMH-0.0125Zr	80.4	92.2	5.6	1.9	0.3	16.5	74.1	3.6	This work
3	PMH-0.025Zr	93.5	93.7	4.6	1.5	0.2	20.4	87.6	4.1	This work
4	PMH-0.05Zr	98.5	94.3	4.0	1.6	0.1	23.6	92.9	4.4	This work
5	PMH-0.1Zr	99.5	94.6	3.7	1.5	0.2	25.6	94.1	—	This work
6	PMH-0.2Zr	89.1	93.1	5.9	0.8	0.2	15.8	83.0	3.9	This work
7	PMH-0.05Zr^c	37.2	92.6	4.7	2.6	0.1	19.7	34.4	14.8	This work
8	AlCl3	99.9	75.2	4.7	4.1	26.0	16	75.1	—	This work
9	SO4^2−/ZrO2^d	86	96	4	—		24	85.6	1.6	52
10	SO4^2−/Zr0.95Sn0.05O2^e	38	78.9	—	—		—	30	1.7	57
11	20% w/w Cs2.5H0.5PW12O40/K-10^f	37/100	100	—	—		—	37/100	7.0/—	55

---

The selectivity to 4-MBP of 92.2–94.6% over all the catalysts with varied Zr-doping seems to be owing to the similar PSD and optimal pore size. However, PMH-0.1Zr and PMH-0.05Zr with high yield of 4-MBP show much higher/comparable ratio of para- to ortho-MBP (P/O) as 26 and 24 than in AlCl₃ system (16) and previously reported superacid SO₄²⁻/ZrO₂ (24),⁵² reflecting the less diffusion resistance of the 4-MBP suffered in these catalysts. This is because the open gallery framework of PMH-xZr (x < 0.2) samples with considerable amounts of layered supermicropores near 0.84, 0.92 and 1.15 nm (Fig. S4†) affords to smaller diffusion resistance for narrow-shaped 4-MBP molecules (0.51 × 1.21 nm estimated from ChemOffice Ultra 2004 in Fig. S10†) than for near square-shaped 2-MBP (0.72 × 0.98 nm). Deutsch et al.⁵⁸ have previously reported obvious pore size dependent performance of the sulfated-ZrO₂ with larger mesopores (∼6.8 nm) for the acylation of anisole with benzoic anhydride compared to commercial microporous H-BEA (0.76 nm) and H-mordenite (0.70 nm) upon steric hindrance effect.

In order to further compare the activity of the present PMH-xZr catalysts with previous report, PMH-0.05Zr was also tested in the benzoylation of anisole with BzCl, under the reaction conditions of anisole/BzCl = 7 : 1, W_cat. 0.4 g, and 110 °C, similar to previous report,⁵⁵ and shows extremely high activity of 14.8 mmol g⁻¹ h⁻¹ with much faster kinetics probably due the unique PMH structure (Table 3, entry 7). This value is much higher than those of previously reported superacid catalysts SO₄²⁻/ZrO₂ (ref. 52) and SO₄²⁻/Zr_0.95Sn_0.05O₂,⁵⁷ and similar to WO₃/Zr-SBA-15 (11.9 mmol g⁻¹ h⁻¹),⁵⁹ though lower than heteropoly acid-loaded montmorillonite 20% Cs_2.5H_0.5PW₁₂O₄₀/K-10 (28.4 mmol g⁻¹ h⁻¹).⁵⁵ These results strongly suggest the remarkably enhanced benzoylation activity of the present PMH-xZr catalyst, which can be mainly ascribed to considerably created Brønsted acidity along with the considerably high surface area.

Combined the characterization analyses, we can deduce that the Brønsted acid sites affect the benzoylation activity of anisole using PMH-xZr as the heterogeneous catalysts, along with the well-kept layered supermicro-mesoporosity. In view of these results and discussion, the schematic illustration of the PMH-xZr catalyst and the nature of the acid sites of the catalyst affording to the benzoylation mechanism of anisole with BzCl are tentatively proposed and presented in Fig. 10.

Fig. 10 Schematic structure of the PMH-xZr catalysts and the nature of Brønsted and Lewis acid sites on PMH-xZr.

3.6 Recyclability

One advantage of heterogeneous catalysts over homogeneous catalysts is their potential reusability. The optimal catalyst PMH-0.1Zr was recovered and used for five consecutive benzoylation of anisole with BzCl (Fig. S11†). After each run, the used catalyst was separated from reaction system by centrifugation, washed thoroughly with absolute ethanol, dried at 120 °C overnight, and subjected to next run. Some loss in activity occurred during the reaction, and the conversion was lowered to 93.2% after five cycles; however, the selectivity was still as high as 93.2%. The loss of activity might be due to the deposition of carbonaceous species (namely, benzoate species) at the active site of the catalyst after the reaction. Indeed, the originally white catalyst became a little pink after the reaction, indicative of some adsorbate on it. The XRD and IR data of the used catalyst PMH-0.1Zr-reused5 well support above analyses. As shown in Fig. S12,† the XRD of PMH-0.1Zr-reused5 shows a broad (001) line with slightly reduced intensity and the corresponding IR spectrum exhibits very weak but distinguishable absorptions at 1605 and 1385 cm⁻¹ ascribing to ν_as(COO⁻) and ν_s(COO⁻), respectively, implying the deposition of quite small amount of benzoate species on the catalyst. Therefore, the PMH-0.1Zr-reused5 was recalcined at 600 °C for 1 h, and the obtained PMH-0.1Zr-recalcined becomes white again and displays almost restored initial activity. The XRD of PMH-0.1Zr-recalcined shows clear (001) line with similar d₀₀₁ to initial PMH-0.1Zr and the corresponding IR describes completely disappeared absorptions due to ν_as(COO⁻) and ν_s(COO⁻), indicating the complete removal of the adsorbed carbonaceous species. These results strongly demonstrate the high thermal stability and supermicro-mesoporous structure stability of the PMH-xZr catalysts upon the protection of the layers of the staring magadiite thus easy recovery and high cycling efficiency, fundamentally originated from the uniformly incorporated Zr species in interlayer of PMH strongly reducing Zr leaching during the repeated tests.

4. Conclusions

It is very interesting that novel group of the porous magadiite heterostructures (PMH) intercalated with SiO₂–ZrO₂ molecular sieve-like clusters were gropingly fabricated via a 2D gallery in situ hydrolysis–condensation–polymerization strategy of TEOS and Zr-n-propoxide from synthetic Na-magadiite. The obtained PMH-xZr catalysts are characterized with respect to their textural parameters, chemical nature of Zr-incorporated into the PMH structure as well as surface acidity. Both the pure silicic PMH and PMH-xZr samples intercalated with silica and silica–zirconia molecular sieve-like clusters, respectively, is found to be thermally stable up to 600 °C. Photoemission peak shapes and binding energy values suggest the formation of the Zr–O(H)–Si bonds in the PMH-xZr catalysts. The in situ incorporation of Zr into the PMH significantly increases surface acidity of PMH by generation of both considerably more Brønsted acid sites as well as obviously increased Lewis acid sites upon Zr-doping. The liquid-phase Friedel–Craft acylation study of the PMH-xZr catalysts for benzoylation of anisole with benzoyl chloride shows that PMH-0.1Zr exhibits the highest 4-MBP yield owing to its most Brønsted acid sites and optimal layered supermicro-mesoporous structure. The simply recovered sample exhibits some loss in activity owing to the reduced acid sites of the catalyst due to the adsorption of carbonaceous species, while the recalcined catalyst almost restored the initial activity even after five cycles, implying a highly stable and efficient acid catalytic performance of PMH-xZr heterostructures by an in situ 2D gallery-template methodology. The present work may open a new avenue for preparing sustainable, recyclable and eco-friendly heterogeneous catalysts with high acidity and large surface area associated with transition metal doped layered supermicro-mesostructures for wide catalysis applications.

Acknowledgements

This work was supported by the 973 Program (No. 2011CBA00508), National Natural Science Foundation of China (No. 21276015), the Fundamental Research Funds for the Central Universities (YS1406), Beijing Engineering Center for Hierarchical Catalysts and IRT1205.

Notes and references

1. A. Galarneau, A. Barodawalla and T. J. Pinnavaia, Nature, 1995, 374, 529–531.
2. M. L. Pinto, V. K. Saini, J. M. Guil and J. Pires, Sep. Purif. Technol., 2012, 98, 337–343.
3. F. Qu, L. Zhu and K. Yang, J. Hazard. Mater., 2009, 170, 7–12.
4. A. J. Tchinda, E. Ngameni, I. T. Kenfack and A. Walcarius, Chem. Mater., 2009, 21, 4111–4121.
5. Q. Hao, Z. Liu, B. Zhang, G. Wang, C. Ma, W. Frandsen, J. Li, Z. Liu, Z. Hao and D. Su, Chem. Mater., 2012, 24, 972–974.
6. C. Zhou, X. Li, Z. Ge, Q. Li and D. Tong, Catal. Today, 2004, 93, 607–613.
7. K. W. Park, J. H. Jung, H. J. Seo and O. Y. Kwon, Microporous Mesoporous Mater., 2009, 121, 219–225.
8. P. Cool and E. F. Vansant, Pillared clays and porous clay heterostructures, in Handbook of layered materials, ed. S. M. Auerbach, K. A. Carrado and P. K. Dutta, Marcel Dekker Inc., New York, 2004, pp. 261–311.
9. A. C. Perdigón, D. Li, C. Pesquera, F. González, B. Ortiz, F. Aguado and C. B. Perdigón, J. Mater. Chem. A, 2013, 1, 1213–1219.
10. M. Polverejan, Y. Liu and T. J. Pinnavaia, Chem. Mater., 2002, 14, 2283–2288.
11. M. Pichowicz and R. Mokaya, Chem. Commun., 2001, 2100–2101.
12. L. Chmielarz, B. Gil, P. Kuśtrowski, Z. Piwowarska, B. Dudek and M. Michalik, J. Solid State Chem., 2009, 182, 1094–1104.
13. L. Chmielarz, P. Kuśtrowski, Z. Piwowarska, B. Dudek, B. Gil and M. Michalik, Appl. Catal., B, 2009, 88, 331–340.
14. H. P. Eugster, Science, 1967, 157, 1177–1180.
15. S. Okutomo, K. Kuroda and M. Ogawa, Appl. Clay Sci., 1999, 15, 253–264.
16. C. Eypert-Blaison, E. Sauzéat, M. Pelletier, L. J. Michot, F. Villiéras and B. Humbert, Chem. Mater., 2001, 13, 1480–1486.
17. O. Y. Kwon, H. S. Shin and S. W. Choi, Chem. Mater., 2000, 12, 1273–1278.
18. Q. Sun, C. Zhang, H. Sun and H. Zhang, Ind. Eng. Chem. Res., 2014, 53, 12224–12237.
19. B. M. Devassy and S. B. Halligudi, J. Catal., 2005, 236, 313–323.
20. H. J. M. Bosman, E. C. Kruissink, J. Vanderspoel and F. Vandenbrink, J. Catal., 1994, 148, 660–672.
21. E. Rodréguez-Castellón, A. Jiménez-López, P. Maireles-Torres, D. J. Jones, J. Rozière, M. Trombetta, G. Busca, M. Lenarda and L. Storaro, J. Solid State Chem., 2003, 175, 159–169.
22. K. Kosuge and P. S. Singh, Chem. Mater., 2000, 12, 421–427.
23. Y. S. Han and S. Yamanaka, J. Solid State Chem., 2006, 179, 1146–1153.
24. J. A. Cecilia, C. García-Sancho and F. Franco, Microporous Mesoporous Mater., 2013, 176, 95–102.
25. B. Zebib, J. F. Lambert, J. Blanchard and M. Breysse, Chem. Mater., 2006, 18, 34–40.
26. Y. Huang, Z. Jiang and W. Schwieger, Chem. Mater., 1999, 11, 1210–1217.
27. G. B. Superti, E. C. Oliveira and H. O. Pastore, Chem. Mater., 2007, 19, 4300–4315.
28. M. Zimowska, H. Pálková, J. Madejová, R. Dula, K. Pamin, Z. Olejniczak, B. Gil and E. W. Serwicka, Microporous Mesoporous Mater., 2013, 175, 67–75.
29. S. H. Teo and H. C. Zeng, J. Phys. Chem. B, 2001, 105, 9093–9100.
30. X. S. Zhao, G. Q. Lu, A. K. Whittaker, G. J. Millar and H. Y. Zhu, J. Phys. Chem. B, 1997, 101, 6525–6531.
31. P. Biswas, P. Narayanasarma, C. M. Kotikalapudi, A. K. Dalai and J. Adjaye, Ind. Eng. Chem. Res., 2011, 50, 7882–7895.
32. J. S. Dailey and T. J. Pinnavaia, Chem. Mater., 1992, 4, 855–863.
33. K. S. W. Sing, D. H. Everett, R. A. W. Haul, L. Moscou, R. A. Pierotti, J. Rouquerol and T. Siemieniewska, Pure Appl. Chem., 1985, 57, 603–619.
34. J. C. Groen, L. A. A. Peffer and J. Pérez-Ramérez, Microporous Mesoporous Mater., 2003, 60, 1–17.
35. D. Yang, J. Li, Y. Xu, D. Wu, Y. Sun, H. Zhu and F. Deng, Microporous Mesoporous Mater., 2006, 95, 180–186.
36. J. Jagiello and M. Thommes, Carbon, 2004, 42, 1227–1232.
37. P. Berteau and B. Delmon, Catal. Today, 1989, 5, 121–137.
38. C. García-Sancho, R. Moreno-Tost, J. Mérida-Robles, J. Santamaría-González, A. Jiménez-López and P. Maireles-Torres, Appl. Catal., B, 2011, 108, 161–167.
39. C. García-Sancho, R. Moreno-Tost, J. Mérida-Robles, J. Santamaría-González, A. Jiménez-López and P. Maireles-Torres, Appl. Catal., A, 2012, 433, 179–187.
40. B. Katryniok, S. Paul, V. Bellière-Baca, P. Rey and F. Dumeignil, Green Chem., 2010, 12, 2079–2098.
41. M. Knapczyk, M. Motek, W. Suprun, H. Papp and T. Grybeck, Catal. Today, 2011, 176, 331–335.
42. J. Datka, A. M. Turek, J. M. Jehng and I. E. Wachs, J. Catal., 1992, 135, 186–199.
43. B. Bachiller-Baeza and J. A. Anderson, J. Catal., 2004, 228, 225–233.
44. F. Cavani, S. Guidetti, L. Marinelli, M. Piccinini, E. Ghedini and M. Signoretto, Appl. Catal., B, 2010, 100, 197–204.
45. D. Carriazo, C. Domingo, C. Martín and V. Rives, J. Solid State Chem., 2008, 181, 2046–2057.
46. S. Y. Chen, J. F. Lee and S. Cheng, J. Catal., 2010, 270, 196–205.
47. K. V. Bineesh, D. K. Kim, D. W. Kim, H. J. Cho and D. W. Park, Energ. Environ. Sci., 2010, 3, 302–310.
48. M. P. Pachamuthu, V. V. Srinivasan, R. Maheswari, K. Shanthi and A. Ramanathan, Appl. Catal., A, 2013, 462, 143–149.
49. J. M. Liu, S. J. Liao, G. D. Jiang, X. L. Zhang and L. Petrik, Microporous Mesoporous Mater., 2006, 95, 306–311.
50. D. J. Jones, J. Jiménez-Jiménez, A. Jiménez-López, P. Maireles-Torres, P. Olivera-Pastor, E. Rodriguez-Castellonb and J. Rozière, Chem. Commun., 1997, 431–432.
51. M. L. Zheludkevich, R. Serra, M. F. Montemor, S. I. M. Miranda and M. G. S. Ferreir, Surf. Coat. Technol., 2006, 200, 3084–3094.
52. V. Quaschning, J. Deutsch, P. Druska, H. J. Niclas and E. Kemnitz, J. Catal., 1998, 177, 164–174.
53. S. C. Moon, M. Fujino, H. Yamashita and M. Anpo, J. Phys. Chem. B, 1997, 101, 369–373.
54. G. A. Olah, Friedel-Crafts and Related Reactions, Wiley-Interscience, New York, 1964.
55. G. D. Yadav, N. S. Asthana and V. S. Kamble, J. Catal., 2003, 217, 88–99.
56. G. Sartori and R. Maggi, Chem. Rev., 2011, 111, 181–214.
57. R. Sakthivel, H. A. Prescott, J. Deutsch, H. Lieske and E. Kemnitz, Appl. Catal., A, 2003, 253, 237–247.
58. J. Deutsch, A. Trunschke, D. Müller, V. Quaschning, E. Kemnitz and H. Lieske, J. Mol. Catal. A: Chem., 2004, 207, 51–57.
59. Y. Q. Zhang, S. J. Wang, J. W. Wang, L. L. Lou, C. Zhang and S. Liu, Solid State Sci., 2009, 11, 1412–1418.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra10911f

---