Mesoporous Al-incorporated silica-pillared clay interlayer materials for catalytic hydroxyalkylation of phenol to bisphenol F

Abstract

A series of mesoporous, Al-incorporated, silica-pillared clay (Al-SPCs) interlayer materials with different Al content were prepared in the presence of cationic surfactant by a structure-directing method. The catalysts' structure, texture, and acidic properties were determined using XRD, BET, SEM, TEM, FT-IR, NH₃-TPD and Py-IR, respectively. Characterization results showed that these materials possess mesoporous structures with large specific surface areas. The incorporated Al leads to the increase and redistribution of Brönsted and Lewis acid sites on SPC (silica-pillared clay). The Al-SPCs were used as catalysts for hydroxyalkylation of phenol to bisphenol F and gave a high product yield (95.4%) and selectivity (98.2%) to bisphenol F. Catalytic performance of the catalysts and characterization results proved that the catalytic activity of these catalysts depend on moderate acidity and the textural properties (specific surface areas), and the synergy of Brönsted and Lewis acids is key for the hydroxyalkylation of phenol to bisphenol F. The reusability of the catalysts was studied, and they can be easily recovered and reused at least six times without significant loss of their catalytic activities. Finally, a plausible mechanistic pathway was proposed.

---

1 Introduction

With increasing demand and applicability in the plastics, resins and rubber industries, bisphenol F synthesis has been given extensive attention. Synthesis of bisphenol F via the hydroxyalkylation of phenol with formaldehyde is a typical acid-catalyzed reaction. Classically, hydroxyalkylation of phenol with formaldehyde to bisphenol F can be catalyzed by the use of various conventional mineral acids, such as phosphoric acid, hydrochloric acid, sulfuric acid or other inorganic acids.¹ Even though moderate to high bisphenol F yields have been achieved, sustainable and economically viable routes for bisphenol F production in scalable quantities cause serious challenges due to the fact that these catalysts are toxic, corrosive, and often hard to remove from the reaction solution. Thus, it is keenly desirable to develop new types of catalysts to replace them. Solid catalysts are well recognized these days because of their ease of workup, the separation of products and catalysts, and their economic advantages. More importantly, they have unique properties, including availability, safety, nontoxicity, and insolubility in the vast majority of solvents. Because of these properties, a number of studies have been reported about the design of solid acid catalysts for bisphenol F synthesis, which span over a broad range of catalytic materials, including modified mesoporous silicas,² clay,³ organometallic framework,⁴ and zeolites.⁵ Recently, Chen et al.⁶ reported the metal–organic frameworks of MIL-100 (Fe or Cr) and MIL-101 (Fe or Cr) encapsulated with Keggin phosphotungstic acid as a catalyst for the hydroxyalkylation of phenol to form bisphenol F. The studies indicate that the nature of the transition metal, Fe or Cr, in MIL-100 or MIL-101 materials determine the isomeric distribution of bisphenol F. Even though the materials formed are good catalysts for the hydroxyalkylation of phenol to bisphenol F, the synthesis procedure is somewhat tedious. Garade et al.⁷ reported, for the first time, that DTP/SiO₂ was active for bisphenol F synthesis. However, the yield of bisphenol F was only 34.2%, where the low catalytic activity severely restricted its application. It is therefore desirable to develop a cheaper and easier method to synthesize catalysts for the hydroxyalkylation of phenol with formaldehyde to bisphenol F.

Pillared clay (PILC) is one of the most widely studied interlayer materials, which has potentially wide applications in the areas of adsorption and catalysis.^8–10 The introduction of oxide pillars into the interlayer space results in a significant increase in surface area, thermal stability, and microporosity. More importantly, by changing the pillar oxide and raw clay, the acidity, layer space, and pore size distribution of the PILCs can be regulated over relatively wide ranges. Pillared clay is suitable for use in heterogeneous liquid phase reactions, offering new opportunities for developing environmentally benign and friendly processes. In this article, a series of Al-SPC with different Al content were synthesized, and the catalytic performance of the Al-SPCs for the hydroxyalkylation of phenol to bisphenol F was studied in detail. Al-Pillared montmorillonite (Al-MMT) was also prepared and used as a catalyst for comparative purposes. The experimental results are well explained, based on characterization by X-ray diffraction (XRD), FT-IR using pyridine as a probe (Py-IR), and NH₃ temperature programmed desorption (NH₃-TPD). Also, the reusability of the catalyst was investigated; reaction parameters such as catalyst weight, reaction temperature, reaction time and molar ratio were also optimized.

2 Experimental

2.1 Materials

Commercially available montmorillonite K10, as the raw material, was supplied by Sinopharm Chemical Reagent Co. Ltd., China. Analysis of its mineralogy showed it to be 95% montmorillonite. Its anhydrous structural (layer) formula, which was determined previously¹¹ is [Si_7.86Al_0.14][Al_2.84Fe_0.30Mg_0.86]O₂₀(OH)₄. Montmorillonite K10 is calcium-rich and was converted into the Na-MMT (denoted as MMT) by treatment with NaCl (1 mol L⁻¹ NaCl solution, 100 mL solution per g of clay, 80 °C for 2 h); dodecyl dimethyl benzyl ammonium chloride (C₁₂DMBACl) (A.R.), tetraethoxysilane (TEOS) (A.R.), ammonia (25%) and ethanol (99.7%) were purchased from Beijing Chemical Reagents Company, China. Phenol, formaldehyde (37–40%) and aluminum chloride hexahydrate (AlCl₃·6H₂O) were purchased from Sinopharm Chemical Reagent Co. Ltd., China.

2.2 Synthesis of SPC

SPC was synthesized in the presence of cationic surfactant by a structure-directing method, according to a previous report.¹² Typically, MMT (1.00 g) was added to 30 mL of water to form a clay suspension. C₁₂DMBACl was dissolved in ethanol, and TEOS was added and stirred for 0.5 h to form a clear solution. The solution was then slowly dropped into the clay suspension. The gel mixture with a molar ratio of clay, surfactant, TEOS, ethanol and water at 1 : 2 : 30 : 1.2 : 250 was stirred for 0.5 h. Subsequently, the mixture was filtered to remove the water and extra TEOS, and the product obtained was C₁₂DMBACl/TEOS-intercalated clay.

The C₁₂DMBACl/TEOS-intercalated clay (2 g) was dispersed in 50 mL of ammonia solution and stirred for 2 h at room temperature. The resultant mixture was subsequently separated from suspension by filtration to obtain powder, dried in air and calcined at 600 °C for 3 h (at a heating rate of 2 °C min⁻¹) to remove C₁₂DMBACl. The prepared sample was denoted as SPC.

2.3 Synthesis of Al-SPC-X

Al-SPC-X were prepared in the presence of cationic surfactant by a structure-directing method as follows: (i) MMT (1.00 g) was suspended in 30 mL of water to form suspension A. (ii) AlCl₃·6H₂O (1.2 g) was added to a solution consisting of 2 mL of ethanol and 8 mL of TEOS to form suspension B. (iii) The suspension B was slowly dropped into suspension A, and stirred for 1 h to obtain a brown sol. Then, 3 mL of ammonia solution were slowly dropped into the sol and stirred for 3 h at room temperature. The resultant mixture was subsequently separated from suspension by filtration to obtain powder, dried in air and calcined at 600 °C for 6 h using a programmed furnace (at a heating rate of 2 °C min⁻¹). The product was denoted as Al-SPC-X, where X represents the number of moles of Al in each gram of MMT.

2.4 Synthesis of Al-MMT

Al-MMT was synthesized by intercalation of Na-MM with aluminum hydroxy oligomeric solution (Al solution), according to our previous reports.¹³ 0.1 mol L⁻¹ solution of NaOH was slowly added to 0.1 mol L⁻¹ AlC1₃ solution prepared from AlCl₃·6H₂O to obtain a final hydrolysis ratio OH⁻/Al³⁺ = 2.4. After aging at 50 °C for 24 h, the AlC1₃ solution was slowly added to the MMT slurry of 2 g/100 mL (10 mmol of Al per gram of MMT). The final suspension was stirred at room temperature for 24 h, then was transferred to dialysis tubing and dialyzed in distilled water for 3 days to remove Cl⁻. The dialysis water was changed every 24 h until the water was tested Cl⁻ free by the AgNO₃ test. The resultant mixture was subsequently separated from suspension by filtration to obtain powder, dried in air and calcined at 400 °C for 4 h.

2.5 Catalytic activity tests

Hydroxyalkylation of phenol with formaldehyde to bisphenol F was performed in a magnetically stirred glass reactor fitted with a reflux condenser and an arrangement for temperature control under nitrogen atmosphere. Briefly, 82.86 mmol of phenol, 5.73 mmol of formaldehyde and catalyst (0.12 g) were added into the glass reactor. The reaction mixture was magnetically stirred and heated to the required temperature. After a definite time interval, the reaction was stopped and 0.03 g of the product was taken and was diluted with 10 mL of methanol. The products were analyzed by a HPLC system with a Shimadzu LC-20AT system, coupled with a SPD-20A UV/Vis detector and a Phenomenex Luna C18 column (250 × 4.6 mm, 5 μm); the column oven temperature was 25 °C, and the mobile phase was methanol : water, with 65 : 35 v/v mobile phase whose flow was 0.6 mL min⁻¹.

The reaction equation for the hydroxylation of phenol with formaldehyde to bisphenol F is shown in Scheme 1.

Scheme 1 Hydroxyalkylation of phenol with formaldehyde to bisphenol F.

The yield and selectivity of bisphenol F were calculated on the basis of formaldehyde. The calculation equations are as follows:

2.6 Catalyst characterization

Powder X-ray diffraction (XRD) patterns were recorded at room temperature on a D8-Advance instrument, with Cu Kα radiation (λ = 0.1541 nm) at 40 kV and 40 mA. Diffraction data were recorded in the 2 h range from 1° to 8° and 10° to 80°, with a scan rate (2θ) of 0.5° s⁻¹ and 2θ = 0.02°. Scanning electron microscopy (SEM) micrographs were obtained on a Hitachi S-4800 microscope operated at 30 kV. The chemical composition of the sample was determined by energy-dispersive X-ray (EDX) analysis using an FEI QUANTA-200 instrument. Transmission electron microscope (TEM) images were obtained using a JEOL JEM-2100 microscope. The TEM samples were prepared by dispersing nanoparticles in acetone to form a suspension. The suspension was sonicated for 20 min and was deposited onto a lacey carbon coated Cu grid. NH₃-TPD analysis was performed with a Micromeritics AutoChem II 2920 V3.05 instrument. Prior to analysis, the catalyst (100 mg) was enclosed in a quartz tube and treated at 300 °C under helium flow of 30 mL min⁻¹ for 1 h. The sample was cooled down to 60 °C under a flow of helium gas and then followed by adsorption of NH₃ for 60 min, which was maintained for 1 h. Subsequently, the catalyst was heated to 700 °C with a ramp of 10 °C min⁻¹ under a helium flow rate of 30 mL min⁻¹. A thermal conductive detector (TCD) was used to measure the desorption amount of NH₃, which was quantified based on the TCD calibration curve that had been obtained by injecting NH₃ pulses of known volume into the helium background flow.

Pyridine adsorption monitored by in situ infrared FTIR spectra (Py-IR) of the catalyst samples were recorded with a Bruker Vector 22 spectrometer in the absorption mode with a resolution of 4 cm⁻¹. Self-supporting wafers were made and loaded in an IR cell. The wafers were pretreated at 400 °C under flowing oxygen for 2 h. Background spectra were recorded after the sample was cooled to room temperature. Adsorption of pyridine was then conducted until saturation. Py-IR spectra were recorded after degassing for 0.5 h. The specific surface area was estimated according to the Brunauer–Emmett–Teller (BET) equation. The pore size distribution was calculated by the Barett–Joyner–Halenda (BJH) method by the adsorption isotherm branch. Infrared spectra were recorded in the wavelength range of 800–4000 cm⁻¹, using a Bruker vector 22 FT-IR spectrophotometer and the KBr disk technique.

3 Results and discussion

3.1 Catalyst characterization

Fig. 1(A) shows small-angle XRD patterns of SPC, Al-SPCs and Al-MMT. As can be seen in SPC and Al-SPCs, these reflections are very similar, indicating that all of the samples had similar basal spacings and lamellar structures by this synthesis method.¹⁴ However, for Al-MMT, the 001 reflection peak is shifted toward higher diffraction angles, and exhibits lower intensity, illustrating that Al-MMT prepared by Al pillared MMT has a smaller layer spacing than Al-SPCs. The basal spacing values for the montmorillonite component in the MMT, Al-SPCs and Al-MMT catalyst are shown in Table 1. The basal spacing increases from 1.7 nm in the MMT to 3.16 nm in the Al-SPCs. This result was expected because of expansion of the interlayer spacing in the MMT after pillaring treatment, indicating that SiO₂ was successfully intercalated into the clay layers.¹⁵ The basal spacing of Al-MMT is 2.1 nm, which is smaller than Al-SPCs. Interestingly, no significant difference in the intensities among the 001 refraction peaks between SPC and Al-SPCs was observed, suggesting that the amount of Al element used does not affect the layered structure of Al-SPCs.

Fig. 1 XRD patterns of SPC, Al-SPCs and Al-MMT samples.

Table 1 Structural data of the samples

Samples	Surface area (m^2 g^−1)	Basal spacing (nm)	Gallery height (nm)	Pore volume (cm^3 g^−1)	Pore size (nm)	Al content (wt%)
MMT	61	1.7	0.74	0.12	7.4	12.6
SPC	657.8	3.16	2.16	0.76	12.5	10.5
Al-SPC-1	584.6	3.13	2.13	0.72	12.2	15.7
Al-SPC-3	475.2	3.12	2.16	0.63	12.1	21.1
Al-SPC-5	410.9	3.12	2.21	0.54	12.2	27.2
Al-SPC-7	386.3	3.12	2.21	0.32	12.1	32.6
Al-MMT	207.5	2.14	2.20	0.35	10.4	34.9

---

The wide-angle XRD patterns of MMT, SPC, Al-SPCs and Al-MMT are presented in Fig. 1(B). All the samples show the typical peaks of the trioctahedral subgroup of 2 : 1 phyllosilicates, which are ascribed to (110), (020), (004), (130), (200), (330) and (060) diffractions.¹⁶ No diffraction lines are seen in Fig. 1(B), indicating that the Al species are well-dispersed in the interlayered silica oxide framework and there is no long-range change in the crystalline structure of the clay.

The SEM micrographs were used to further study the structural information of the catalysts. As shown in Fig. 2. The MMT and Al-MMT show a closer sheet structure, owing to the MMT and Al-MMT possessing ordered layered structures and small layer spacings. The platelets in the SPC and Al-SPCs were unaffected by hydrolysis, although they swelled slightly more than the MMT and Al-MMT. All of the Al-SPCs derivatives synthesized using cationic surfactants exhibited similar morphologies. Moreover, some small particles can be observed around platelets of SPC and Al-SPCs. This is due to the broken platelets and the forming of some compounds in the extralayer regions of the clays.¹⁷ However, this platelet destruction did not reflect SPC and Al-SPCs gallery structure destruction; the results were confirmed by XRD spectra.

Fig. 2 SEM image of SPC, Al-SPCs and Al-MMT samples.

FT-IR patterns of MMT, SPC, Al-SPCs and Al-MMT are shown in Fig. 3. The absorption bands at about 3450 and 1658 cm⁻¹ were assigned to the stretching and bending vibrations of the O–H bonds in the clay and water molecules present in the interlayer.¹⁸ In all curves, the band at around 3652 cm⁻¹ is due to the vibration of Al–OH. Typical bands of the silicate components appear between 1300 and 800 cm⁻¹, and the absorption at 852 cm⁻¹ is due to the Si–OH in the tetrahedral layer.¹⁹ The band located at 1028 cm⁻¹ was indexed to the asymmetric stretching bonds of Si–O–Si.²⁰ Compared to the FT-IR spectra of pure MMT and SPC, a significant increase in the peaks intensity at 3652 cm⁻¹ with increasing content of Al in Al-SPCs is attributable to the appearance of Al–OH bending vibrations. The above results revealed that Al atoms were incorporated into the interlayered framework of clay.

Fig. 3 FT-IR spectra of SPC, Al-SPCs and Al-MMT samples.

To get further insight into the layered structure of Al-SPCs, HRTEM investigations were then conducted and the corresponding images are shown in Fig. 4. The HRTEM image (Fig. 4(A)) of MMT shows that the sheets remain closely together. From the image (Fig. 4(B) and (D)) of the SPC and Al-SPC-7, clear solid dark lines and the pores can be observed, which is attributable to the clay layers.²¹ Furthermore, the uniform gallery pores can be clearly observed between the dark layers. Obviously, the Al-SPC-5 (Fig. 4(C)) shows the uniform layered structure, and has relatively higher gallery height than MMT.

Fig. 4 HRTEM images of (A) MMT, (B) SPC, (C) Al-SPC-5 and (D) Al-SPC-7 samples.

The N₂ adsorption–desorption isotherms of the SPC, Al-SPCs and Al-MMT are presented in Fig. 5(A) and average pore dimensions and volumes obtained from the maxima (BJH) are presented in Table 1. Irrespective of treatment method, the N₂ isotherms of the catalysts can be assigned to be a combination of types I and IV,²² according to the BDDT (Brunauer, Deming, Deming, and Teller) classification, and the features of the hysteresis loop correspond to type B in Boer's five types, suggesting the presence of the open slit-shaped capillaries with very wide bodies and narrow short necks in the region between interlayers.²³ Compared with MMT and Al-MMT, the SPC and Al-SPCs have a significant increase in porosity and surface area, which is attributable to the formation of rigid intercalated porous structures when TEOS, the pillaring precursor, is converted into firm enough silica pillars. As shown in Fig. 5(A), the slight increase in the amount of N₂ adsorbed with increasing relative pressure from a low to a medium value (P/P_o = 0.01–0.2) indicate that these materials possess supermicropores and small mesopores.²⁴ In the P/P_o region from 0.0 to 0.4, the adsorption isotherm gives a good fit with the BET equation, as well as with the Langmuir equation; the above results indicate that the pores formed between parallel layers are quite open.

Fig. 5 N₂ adsorption–desorption curves (A) and pore size distributions (B) of SPC, Al-SPCs and Al-MMT samples.

The BJH-adsorption pore size distribution curves of the samples are shown in Fig. 5(B). All the Al-SPCs studied displayed different mesopore size distributions. The average pore size of Al-SPCs was 12.1–12.2 nm, which was slightly less than the average pore size of SPC, indicating that the interlayer frameworks are changed by the incorporation of Al. This might be owing to the extragallery Al₂O₃ occupying the spaces of the gallery channels and increasing the thickness of the pore walls, and resulting in a slight decrease in the average pore size. These results are in great agreement with the N₂ adsorption–desorption isotherms. The average pore size of SPC is closely related to the interlayer space height. This implies that the interlayer surfactant plays an important structure-directing role in the synthesis of the intercalated silica. The surface area and pore volume of the Al-SPCs decreased with increase in the content of Al (Table 1), which is owing to the packing of small metal particles on the surfaces and pores.²⁵

NH₃-TPD analysis was performed in order to evaluate the quantification of the catalysts' acid density and the results are shown in Fig. 6. The amount of surface acid sites based on NH₃ desorption are presented in Table 2. From Fig. 6, three types of desorption peaks from the 99 to 530 °C region were observed, which could be denoted as weak, moderate and strong acid sites, respectively.²⁶ The weak and moderate acid sites are in the form of the OH groups bonded to the pillars' Al ions (Al–OH), and the strong acid site are associated with the OH groups bonded to tetrahedrally coordinated Al ions (Al_Td–OH).²⁷ As the Al content increased from 1 to 7, the total amount of acid sites of Al-SPCs significantly increased. Meanwhile, the amount of moderate acid sites of Al-SPCs was enhanced considerably from 1.90 to 3.19 mmol g⁻¹. Interestingly, the amount of weak and strong acid sites only showed a minor change. In addition, the highest amount of moderate acid sites was found in Al-MMT, which may due to the high Al content for Al-MMT (Table 1).

Fig. 6 NH₃-TPD profiles of SPC, Al-SPCs and Al-MMT samples.

Table 2 Acidic properties of various solid acid catalysts

Catalyst	NH3-TPD amount of acidic sites (mmol g^−1)				Acidity by type (mmol g^−1)
	Weak	Moderate	Strong	Total	Brönsted	Lewis	Lewis/Brönsted
SPC	0.49	1.90	1.11	3.50	0.18	0.13	0.72
Al-SPC-1	0.50	2.05	1.09	3.64	0.25	0.23	0.92
Al-SPC-3	0.51	2.27	1.06	3.84	0.27	0.26	0.96
Al-SPC-5	0.54	2.92	1.10	4.56	0.29	0.41	1.41
Al-SPC-7	0.56	3.19	1.93	5.68	0.32	0.53	1.66
Al-MMT	0.62	4.60	1.02	6.24	—	—	—

---

The acid properties of SPC and Al-SPCs were evaluated by FT-IR measurement for samples with adsorbed pyridine, and the results are shown in Fig. 7. The amounts of Brönsted and Lewis acid sites are listed in Table 2. These peaks could be assigned to the chemisorption of molecular pyridine at different types of surface acidic sites. The band at 1447 cm⁻¹ in the different FT-IR spectra in Fig. 7 is assignable to coordinated pyridine species with Lewis acid sites.²⁸ The band at 1544 cm⁻¹ is due to pyridinium ions bonded to Brönsted acid sites. The band at 1490 cm⁻¹ is attributed to pyridine associated with both Lewis and Brönsted acid sites.^29,30 It can clearly found that only small peaks corresponding to pyridine adsorbed on acidic sites were observed for SPC, indicating that the surface acidities of SPC were weak. Meanwhile, the amounts of Brönsted and Lewis acid sites on Al-SPCs increased from 0.18 to 0.32 mmol g⁻¹, and 0.13 to 0.53 mmol g⁻¹, respectively, with the increase in Al content from 1 to 7. Note that Al significantly affected the acidity of the MMT materials and the addition of Al resulted in the increase and redistribution of the two types of acid sites. This result shows that Al-SPCs were highly active for the reaction from a higher L/B ratio; therefore, the hydroxyalkylation of phenol to bisphenol F is catalyzed by the synergy of Brönsted acid and Lewis acid, rather than by a single acid. This result was confirmed by Py-IR results and is consistent with the above NH₃-TPD result.

Fig. 7 Py-IR spectra of SPC and Al-SPCs samples.

As shown in the Scheme 2, Al was incorporated into the silica-framework between the interlayer regions of MMT.

Scheme 2 Schematic for the preparation of Al-containing mesoporous silica-pillared clay.

3.2 Activity measurement

The catalytic activities of the prepared SPC, Al-SPCs and Al-MMT were evaluated for the hydroxyalkylation of phenol. The results are summarized in Table 3. As compared to MMT and SPC, Al-SPCs showed the presence of the aluminum species in the gallery silica framework. Obviously, Al-SPC-5 given the highest product yield (95.4%), higher than that very high catalytic activity, indicating that catalytic activity is a result of the Al-SPC-3 catalyst (72.8%). With the increase in Al content from 1 to 5, the bisphenol F yield increased from 36.5 to 95.4%. This is attributed to the enhanced amount of moderate acid sites and higher L/B ratio, which resulted in higher catalytic activity, confirming the results of NH₃-TPD and Py-IR (Table 2). With further increase in Al content up to 7, there was a decrease of bisphenol F yield (94.7%). This may be due to the Al-SPCs derivatives prepared with high Al content containing mainly isolated tetracoordinated Al species incorporated into the gallery silica structure, which leads to a gradual reduction of the quality of the pore structure and of the pore volumes, and results in the decline of specific surface areas. Interestingly, the Al-MMT showed lower catalytic activity than Al-SPC-5 and Al-SPC-7, although it had the higher amount of moderate acid sites. This is attributed to the low specific surface areas of Al-MMT. These experimental results are in good agreement with the NH₃-TPD analysis and the N₂ adsorption–desorption isotherms, indicating that the catalytic activity is dependent on the amount of moderate acid sites and the specific surface area of catalyst for the hydroxyalkylation of phenol. Al-SPC-5 exhibited higher catalytic performance than the others; consequently, Al-SPC-5 was chosen as the catalyst for the following study.

Table 3 Catalytic activities of MMT, SPC, Al-SPCs and Al-MMT for bisphenol F^a

Catalyst	Yield/%	Selectivity/%	Isomer distribution/%
			4,4′-Isomer	2,4′-Isomer	2,2′-Isomer
a Reaction conditions: phenol/formaldehyde molar ratio, 15 : 1; catalyst concentration, 0.003 g g^−1; reaction temperature, 353 K; reaction time, 40 min.
MMT	10.2	90.5	18.3	51.2	30.5
SPC	12.3	91.2	20.7	51.1	28.2
Al-SPC-1	36.5	97.6	27.4	51.3	21.3
Al-SPC-3	72.8	97.2	29.1	50.8	20.1
Al-SPC-5	95.4	98.2	29.6	50.2	20.2
Al-SPC-7	94.7	98.1	29.3	50.1	21.6
Al-MMT	87.3	90.4	32.2	48.5	19.3

---

3.3 Recyclability of catalysts

Since Al-SPCs showed rather high catalytic activity and green chemistry principles, its reusability was much more important. After completion of the reaction, the catalyst was recovered by filtration, washed with ethanol and acetone, and air-dried. The recovered catalyst was reused in subsequent runs and the results are listed in Table 4. It can be clearly seen that only a slight decrease in bisphenol F yield was observed after the sixth run, indicating that the acid sites were not largely lost during the repeated process. In order to confirm the above speculation, the recovered Al-SPC-5 was explored using NH₃-TPD, XRD, and FT-IR. XRD of the fresh and recovered Al-SPC-5 catalyst after six runs are listed in Fig. 8(A). It is clearly seen that the intensity of the 001 reflection peak decreased. Moreover, the 001 reflection peak shifted toward higher diffraction angles after the sixth cycle. The reason for this result may be that reactants and trimer were adsorbed onto the clay layers after the sixth cycle, and resulted in reduced basal spacings. NH₃-TPD spectra of the sixth cycle were repeated using Al-SPC-5 and were analyzed to further confirm the acidic properties of the recycled catalysts; the results are presented in Fig. 8(B). As compared with the fresh catalyst, a slight decrease in the moderate acid sites peak intensities of the recovered catalyst Al-SPC-5 after six cycles was observed, which was attributed to the acid site adsorption of some compounds, resulting in a reduction in the amount of acid sites after the sixth cycle.

Table 4 Catalytic activities of Al-SPC-5 recycled different numbers of times

Cycle no	Yield/%	Selectivity/%	Isomer distribution/%
			4,4′-Isomer	2,4′-Isomer	2,2′-Isomer
0	95.4	98.2	27.3	50.6	22.1
1	94.1	97.5	27.5	50.2	22.3
2	93.3	97.1	28.2	51.3	20.5
3	93.1	97.0	28.4	50.5	21.1
4	92.0	97.0	29.1	50.8	20.1
5	91.4	96.4	28.6	50.1	21.3
6	91.2	96.4	28.1	50.7	21.2

---

Fig. 8 Small XRD patterns (A) and NH₃-TPD profile (B) of fresh and recovered Al-SPC-5 after six cycles.

In order to further explain the superior recyclability of Al-SPCs, the FT-IR spectrum of the sixth cycle of repeatedly used Al-SPC-5 was compared with that of the fresh Al-SPC-5, and the results are shown in Fig. 9. As compared with the spectrum of the fresh Al-SPC-5, there is a slight decrease in the peak intensity at 3652 cm⁻¹ (corresponding to the Al–OH, which is an acid site) in the spectrum of the recovered Al-SPC-5. This is well in agreement with the result where the yield shows a slight decrease after the sixth run. These results clearly show that the Al-SPCs have the great potential to be effectively separated and reused for the hydroxyalkylation of phenol.

Fig. 9 FT-IR spectra of fresh and recovered Al-SPC-5 after six cycles.

3.4 Optimization of reaction parameters

The effects of the phenol/formaldehyde molar ratio (3, 5, 10, 15, and 20) on the yield, selectivity, and the isomers' distribution of bisphenol F over Al-SPC-5 are illustrated in Fig. 10. The product yield is increased significantly from 48 to 95%, with a change in the phenol/formaldehyde ratio from 3 to 15; with further increase in the mole ratio of the phenol/formaldehyde to 20, both the yield and selectivity of bisphenol F do not show any significant changes. This was mainly because of the blockage of active sites of the catalyst by more phenol molecules present in excess amounts for a mole ratio of 20.³² Moreover, when increasing phenol concentration, the formaldehyde molecules can be surrounded by phenol molecules, which would result in the initially formed bisphenol F molecules almost not being in contact with the formaldehyde molecules to form higher homologues.³³ The selectivities of the 4,4′ and 2,4′-isomers increase from 26 to 28% and from 46 to 49%, whereas the selectivity of the 2,2′-isomer decreases from 29 to 23%. Thus, the optimal molar ratio was 15.

Fig. 10 Effect of phenol/formaldehyde molar ratio on yield, selectivity and isomer distribution of bisphenol F. Reaction conditions: catalyst, Al-SPC-5; reaction time, 40 min; reaction temperature, 343 K; catalyst concentration, 0.003 g g⁻¹.

Fig. 11 shows the effect of catalyst concentration on the yield, selectivity, and isomer distribution of bisphenol F; the product yield dramatically increased from 75 to 95% upon increasing the catalyst concentration from 0.0005 to 0.003 g g⁻¹. On further increase in catalyst concentration, however, the product yield remained almost unchanged. This may be attributed to the overused catalysts being also beneficial for accelerating the side-reaction to produce more byproducts. Interestingly, no significant change was observed in the selectivity to bisphenol F and isomer distribution, as the catalyst concentration increased from 0.0005 to 0.003 g g⁻¹. Thus, 0.003 g g⁻¹ was selected as the suitable catalyst dose.

Fig. 11 Effect of catalyst concentration on the yield, selectivity and isomer distribution of bisphenol F. Reaction conditions: catalyst, Al-SPC-5; mole ratio of phenol/formaldehyde, 15; catalyst concentration, 0.0005–0.004 g g⁻¹; reaction time, 40 min; reaction temperature, 343 K.

The effects of the reaction temperature on product yield, selectivity to bisphenol F and isomer distribution over Al-SPC-5 were investigated and are presented in Fig. 12. It was clearly observed that the product yield significantly increased from 52 to 95%, while the bisphenol F selectivity just changed slightly, ranging from 98 to 97% with increased temperature from 313 to 343 K. However, on further increase in temperature to 353 K, the product yield showed a contradictory decreasing trend. It has been previously proved that an increase in the reaction temperature can result in the formation of some undesired by-products, such as trimer. In addition, the selectivity of the 4,4′-isomer decreased from 49% to 32%. However, the selectivity of 2,4′- and 2,2′-isomers increased from 42 to 47% and 9 to 21%, respectively; thus, 343 K was chosen as the optimal reaction temperature.

Fig. 12 Effect of reaction temperature on the yield, selectivity and isomer distribution of bisphenol F. Reaction conditions: catalyst, Al-SPC-5; phenol/formaldehyde, 15; reaction time, 40 min; reaction temperature, 313–353 K; catalyst concentration, 0.003 g g⁻¹.

The effects of reaction time on the product yields, bisphenol F selectivity and isomer distribution were also studied and the results are shown in Fig. 13. It was clearly observed that the product yield increased rapidly from 65 to 95%, while bisphenol F selectivity remained almost unchanged with increased reaction time from 10 to 40 min. When reaction time was further prolonged to 50 min, the product yield and selectivity to bisphenol F only showed a minor change. This was expected due to the bisphenol F conversion into a trimer, with longer reaction time. The selectivity of 4,4′- and 2,4′-isomers decreased from 30 to 28% and 53 to 51%, respectively; however, the selectivity of the 2,2′-isomers increased from 17 to 20%. Based on the results, we consider that 40 min is the reaction time for optimal performance.

Fig. 13 Effect of reaction time on the yield, selectivity and isomer distribution of bisphenol F. Reaction conditions: catalyst, Al-SPC-5; mole ratio of phenol/formaldehyde, 15; reaction time, 10–50 min; catalyst concentration, 0.003 g g⁻¹.

3.5 Plausible mechanistic pathway

On the basis of our group's previous studies¹³ and some related literature,³³ a plausible catalytic reaction mechanism was proposed and is shown in Scheme 3. The synthesis of bisphenol F via the condensation reaction of phenol with formaldehyde is a typical acid-catalyzed reaction,^34,35 following the adsorption of the formaldehyde on the Brönsted acid sites, and formation of a carbenium ion intermediate (A) by eliminating water. The carbenium ion is produced by the abstraction of H⁻ over Lewis acid sites of the catalyst. Subsequently, the carbenium ion could attack the phenol to form para or ortho hydroxy benzyl alcohol (B or C). B or C are activated in the presence of Brönsted acid or Lewis acid to form para or ortho hydroxy benzyl carbonium ion (D or E) and simultaneously release a water molecule. Species D or E, with strong electrophilicity, then attacks the para or ortho carbon atom of a phenol molecule to form 2,2′-isomers, 2,4′-isomers, and 4,4′-isomers of bisphenol F.

Scheme 3 Proposed plausible mechanism for hydroxyalkylation of phenol and formaldehyde to bisphenol F.

It can be found from the entire catalytic process that the catalytic active sites of Al-SPCs are mainly from the hydrogen proton (Brönsted acid) and coordinatively unsaturated Al^m+ sites (Lewis acid) in the pillars.³¹ Rataboul et al.³⁶ measured the minimum-activation energy required for catalysis by Brönsted and Lewis acid sites, and manifested that the protonation of formaldehyde is activated by Lewis acid to capture H⁻ from the reaction and requires a greater activation energy than that required for Brönsted acid. Catalysts with high Lewis to Brönsted acid site ratios result in higher catalytic activity, which is confirmed by the results of Py-IR. This means that the hydroxyalkylation of phenol to bisphenol F is catalyzed by the synergy of Brönsted acid and Lewis acid, rather than by a single acid. These results are also in qualitative agreement with the experimental work of Weingarten and co-workers.³⁷

4 Conclusions

Mesoporous Al-SPCs with different Al content were prepared in the presence of cationic surfactant by a structure-directing method. These materials possess mesoporous structures with large specific surface areas. The incorporated Al leads to the increase and redistribution of Brönsted and Lewis acid sites on SPC. These catalysts show excellent catalytic activity and selectivity to bisphenol F. The sufficient moderate acid sites and high surface areas are critical for achieving the highest yield and selectivity of bisphenol F in the hydroxyalkylation of phenol. Moreover, the reusability of the catalysts was studied and the results show that the catalysts can be recovered for at least six cycles, without significant loss of their catalytic activities. In addition, the influences of various reaction parameters like mole ratio, catalyst concentration, reaction temperature and reaction time on the product yield and selectivity to bisphenol F were investigated and a plausible mechanistic pathway was proposed.

Acknowledgements

This research was financially supported by the National Natural Science Foundation of China (no. 51378187, J1210040), Innovative Research Team Development Plan-Ministry of Education of China (no. IRT1238), the Key Project of Hunan Provincial Education Department (no. 13CY001), and Hunan Provincial International Cooperation Project of China (no. 2014WK3030).

Notes and references

1. M. Okihama and J. Kunitake, Patent JP 09067287, 1997.
2. S. K. Jana, T. Kugita and S. Namba, Catal. Lett., 2003, 90, 3–4.
3. R. Liu, X. Niu, X. Xia, Z. Zeng, G. Zhang and Y. Lu, RSC Adv., 2015, 5, 62394–62401.
4. N. Maksimchuk, M. Timofeeva, M. Melgunov, A. Shmakov, Y. Chesalov, D. Dybtsev, V. Fedin and O. Kholdeeva, J. Catal., 2008, 257, 315–323.
5. K. Nowinska and W. Kaleta, Appl. Catal., A, 2000, 203, 91–100.
6. M. Chen, J. Yan, Y. Tan, Y. Li, Z. Wu, L. Pan and Y. Liu, Ind. Eng. Chem. Res., 2015, 54, 11804–11813.
7. A. C. Garade, V. S. Kshirsagar and C. V. Rode, Appl. Catal., A, 2009, 354, 176–182.
8. M. A. De Leon, M. Sergio, J. Bussi, G. Ortiz de la Plata, A. E. Cassano and O. M. Alfano, Ind. Eng. Chem. Res., 2015, 54, 1228–1235.
9. J. Virkutyte and R. S. Varma, ACS Sustainable Chem. Eng., 2014, 2, 1545–1550.
10. C. D. Keenan, M. M. Herling, R. Siegel, N. Petzold, C. R. Bowers, E. A. Rossler, J. Breu and J. Senker, Langmuir, 2013, 29, 643–652.
11. H. Mao, B. Li, X. Li, L. Yue, J. Xu, B. Ding, X. Gao and Z. Zhou, Microporous Mesoporous Mater., 2010, 130, 314–321.
12. S. Yang, G. Liang, A. Gu and H. Mao, Ind. Eng. Chem. Res., 2012, 51, 15593–15600.
13. X. Wu, X. Xia, R. Liu and Y. Chen, RSC Adv., 2016, 6, 34625–34632.
14. Q. Hao, G. Wang, Z. Liu, J. Lu and Z. Liu, Ind. Eng. Chem. Res., 2010, 49, 9004–9011.
15. A. Gil, R. Trujillano, M. A. Vicente and S. A. Korili, Ind. Eng. Chem. Res., 2008, 47, 7226–7235.
16. H. Mao, B. Li, X. Li, L. Yue, Z. Liu and W. Ma, Ind. Eng. Chem. Res., 2010, 49, 583–591.
17. Y. Gao, Y. Guo and H. Zhang, J. Hazard. Mater., 2016, 302, 105–113.
18. F. Tomul, Ind. Eng. Chem. Res., 2011, 50, 7228–7240.
19. H. Mao, B. Li, X. Li, Z. Liu and W. Ma, Mater. Res. Bull., 2009, 44, 1569–1575.
20. Y. Han, H. Matsumoto and S. Yamanaka, Chem. Mater., 1997, 9, 2013–2018.
21. H. Mao, B. Li, X. Li, Z. Liu and W. Ma, Mater. Res. Bull., 2009, 44, 1569–1575.
22. S. G. Hur, T. W. Kim, S. J. Hwang, S. H. Hwang, J. Yang and J. H. Choy, J. Phys. Chem. B, 2006, 110, 1599–1604.
23. J. H. Choy, H. Jung, Y. S. Han, J. B. Yoon, Y. G. Shul and H. J. Kim, Chem. Mater., 2002, 14, 3823–3828.
24. M. A. Vicente, C. Belver, M. Sychev, R. Prihod'ko and A. Gil, Ind. Eng. Chem. Res., 2009, 48, 406–414.
25. J. Valverde, A. Lucas, F. Dorado, A. Romero and P. García, Ind. Eng. Chem. Res., 2005, 44, 2955–2965.
26. M. Balaraju, V. Rekha, P. S. S. Prasad, B. L. A. P. Devi, R. B. N. Prasad and N. Lingaiah, Appl. Catal., A, 2009, 354, 82–87.
27. G. Centi, S. Perathoner and F. Trifirb, J. Phys. Chem., 1992, 96, 2617–2629.
28. F. Chambon, F. Rataboul, C. Pinel, A. Cabiac, E. Guillon and N. Essayem, Appl. Catal., B, 2011, 105, 171–181.
29. B. Devassy, F. Lefebvre and S. Halligudi, J. Catal., 2005, 231, 1–10.
30. A. M. Alsalme, P. V. Wiper, Y. Z. Khimyak, E. F. Kozhevnikova and I. V. Kozhevnikov, J. Catal., 2010, 276, 181–189.
31. H. Mao, B. Li, L. Yue, L. Wang, J. Yang and X. Gao, Appl. Clay Sci., 2011, 53, 676–683.
32. R. Liu, X. Xia, X. Niu, G. Zhang, Y. Lu, R. Jiang and S. He, Appl. Clay Sci., 2015, 105–106, 71–77.
33. D. Wang, Z. He, Z. Wu, Y. Tan, Y. Li and Y. Liu, Chin. J. Chem. Eng., 2016 DOI:10.1016/j.cjche.2016.02.005.
34. J. Tian, H. An, X. Cheng, X. Zhao and Y. Wang, Ind. Eng. Chem. Res., 2015, 54, 7571–7579.
35. Q. Wang, Z. M. Wu, Y. Li, Y. Tan, N. Liu and Y. Liu, RSC Adv., 2014, 4, 33466–33473.
36. F. Chambon, F. Rataboul, C. Pinel, A. Cabiac, E. Guillon and N. Essayem, Appl. Catal., B, 2011, 105, 171–181.
37. R. Weingarten, G. A. Tompsett, W. C. Conner and G. W. Huber, J. Catal., 2011, 279, 174–182.

---