Highly effective transformation of methyl phenyl carbonate to diphenyl carbonate with recyclable Pb nanocatalyst

Diphenyl carbonate (DPC) is a type of versatile industrial chemical, and the disproportionation of methyl phenyl carbonate (MPC) is a key step to produce DPC. However, the design and formulation of a catalyst for the efficient synthesis of DPC is a major challenge due to its small equilibrium constant. The support material is a critical factor influencing the performance of Pb nanocatalysts. Thus, a series of Pb-based catalysts over MgO, ZrO2, SiO2, TiO2 and Al2O3 were prepared to investigate the effect of the support materials on the physicochemical properties and catalytic performances for the conversion of MPC to effectively synthesize DPC. The catalysts were well characterized by XRD, BET, TEM, XPS, ICP-OES, H2-TPR, Py-IR and NH3-TPD. The results showed that the nature of the support obviously affected the structural properties and catalytic performances, and Pb was dispersed better on SiO2, TiO2, ZrO2 and MgO than on Al2O3, and showed stronger metal-support interaction over MgO and ZrO2. The activity results revealed that PbO/MgO and PbO/ZrO2 exhibited higher catalytic activities because they contained higher Pb dispersion and more Lewis acid sites, and the catalytic activities followed the order PbO/MgO > PbO/ZrO2 > PbO/SiO2 > PbO/Al2O3 > PbO/TiO2. On the contrary, PbO/MgO and PbO/ZrO2 exhibited better reusability due to strong interaction between the highly dispersed Pb and the supports, and the activity decrease in the case of PbO/SiO2, PbO/Al2O3 and PbO/TiO2 mainly resulted from the Pb leaching loss. This work would contribute to exploiting novel catalytic materials in a wide range of applications for the efficient synthesis of organic carbonates.


Introduction
Due to its outstanding electrical, mechanical and heat resistance properties, polycarbonate (PC), as an important engineering thermoplastic for ubiquitous applications in several industries (automobiles, office equipment, medical devices, etc.), has attracted considerable interest in the last decades. 1,2 Diphenyl carbonate (DPC), a building block, was oen employed for the phosgene-free synthesis of PC by melt trans-esterication with bisphenol-A, and was also used for the production of pharmaceuticals, polymer materials, ne chemicals, etc. [3][4][5] It was, therefore, necessary to develop efficient and environmentally benign synthetic routes to replace the highly toxic synthesis via phosgene. Among the alternative routes (phosgene method, 6,7 oxidative carbonylation, 8,9 trans-esterication, 10,11 etc.), the transesterication of phenol and dimethyl carbonate (DMC) was regarded as a most likely pathway to synthesize DPC without employing phosgene since the starting materials presented low toxicity and high biodegradability. This route consisted in two-step reaction where methyl phenyl carbonate (MPC) of the intermediate was rst produced by the transesterication (Scheme 1-(1)), followed by the further transesterication reaction of MPC and phenol Scheme 1 Transesterification process between DMC and phenol into DPC.
(Scheme 1-(2)) or the disproportionation of MPC (Scheme 1-(3)). Concerning the second-step reaction the equilibrium constants K 2 and K 3 indicated that DPC was mainly produced by the second step disproportionation instead of the further trans-esterication. 12,13 On the contrary, the rst step was ratedetermining step since K 1 was much lower than K 3 . Hence, considerable studies on the rst step transesterication have been particularly carried out, [14][15][16][17] but only few efforts have been made to the second step disproportionation. 18,19 Indeed, the disproportionation of MPC is also the key step to produce DPC. Overall, it is very signicant to intensively study on MPC disproportionation, which is advantageous to improving the yield of DPC and the conversion of the starting materials in future industrial applications.
The disproportionation reaction suffered from the low yield and selectivity of DPC due to the relatively low equilibrium constant and reaction rate. Therefore, the development of highly active catalysts is crucial in this reaction. Among various homogeneous catalytic systems reported like organo-tin/ titanium catalysts, even though they exhibited outstanding catalytic activities, 19,20 they had received little consideration because of the separation problems between DPC and the catalysts. Therefore, heterogeneous catalysis currently attracted more attention because these catalysts might be facilely recovered from the reaction mixture by simple ltration and could be recycled, making the procedure more feasible.
Recently, we studied the catalytic performances of various metal oxides including TiO 2 , MoO 3 , PbO and so on, 21-23 and found that PbO was prepared into the supported Pb catalysts or mixed metal oxides signicantly improved the catalytic activity, life time and stability. 24,25 However, the nature and property of the support is oen found to modify the distribution, loading, size, shape and electronic state of the active species as well as their interactions with support with some special physicalchemical properties, which is well known to strongly affect the catalytic activity and stability. [26][27][28][29][30][31] Therefore, the study to investigate the structure-function relationship of the supports on Pb-based catalysts for MPC disproportionation is very worthwhile, which is conducive to screen suitable and promising supports.
In the present work, we formulate Pb-based catalysts by employing potential metal oxides as support phases and focus on studying the effect of various supports on the physicochemical properties of the catalysts and their catalytic performance to provide the role of support materials for developing efficient catalysis materials for MPC disproportionation to effectively synthesize DPC. Herein the Pb-based catalysts are prepared using wet-impregnation process over PbO loading onto SiO 2 , TiO 2 , Al 2 O 3 , MgO and ZrO 2 supports. At the same time, the catalysts are thoroughly characterized by X-ray diffraction (XRD), temperature programmed reduction of H 2 (H 2 -TPR), transmission electron microscopy (TEM), X-ray photoelectron spectrum (XPS), N 2 adsorption/desorption measurement, pyridine-infrared spectroscopy (Py-IR) and temperature programmed desorption of NH 3 (NH 3 -TPD) to clarify the support role on the Pb-based catalysts. In addition, the stabilities of these catalysts are also investigated in detail.

Material preparation
All chemicals adopted in this work were analytical purity, which were purchased from Sinopharm Chemical Reagent Co., Ltd and were directly used as received without further purication. Deionized water was used in all experimental synthesis procedures as needed.
Pure SiO 2 was synthesized by sol-gel process. For this preparation, the desired amounts of tetraethyl orthosilicate, ethanol and distilled water were mixed together at room temperature. Then the ammonium aqueous solution was added gradually into the above solution under stirring at 50 C to generate a transparent gel. The obtained gel was aged at 50 C overnight, dried at 110 C for 12 h and subsequently calcinated 550 C for 5 h.
Pure TiO 2 was prepared by hydrolysis process. Briey, a proper amount of tetrabutyl titanate was dissolved in ethanol, and heated up to 50 C under stirring. Then, appropriate amount of water was slowly added to promote the hydrolysis and was kept for 24 h. Finally, the sediment was ltered, washed with distilled water, and dried at 110 C for 12 h and calcinated at 550 C for 5 h.
Pure Al 2 O 3 was prepared by a precipitation process. A certain amount of aluminum nitrate nonahydrate were dissolved in the deionized water and heated slowly to 90 C. Aerwards, ammonia solution with 25% concentration was added into this solution under stirring until that the pH of 9-10 was reached. The as-prepared solid sediment was aged for 24 h and then vacuum ltered off and washed repeatedly. Finally, the obtained samples were dried at 110 C for 12 h and subsequently calcinated at 550 C for 5 h. Likewise, pure MgO and ZrO 2 were prepared by using the similar process, respectively.
The preparation of supported Pb-based catalysts was carried out by incipient wetness impregnation process. Briey, SiO 2 , TiO 2 , Al 2 O 3 , ZrO 2 and MgO supports were homogeneously impregnated by using the aqueous solution of lead nitrate as precursor salt at ambient temperature for 24 h, respectively. Aer impregnation, the samples were dried in air at 110 C for 12 h, and nally calcinated at 550 C for 5 h. And the PbO loading used was 10 wt% for all the catalysts.

Material characterization
The X-ray powder diffraction (XRD) analysis was carried out on a DX-2700B diffractometer with the Ni-ltered CuKa radiation (1.5418Å). The X-ray tube was operated at 40 kV and 30 mA, and the powder diffractogram was recorded at 0.02 intervals in the range of 5-90 with the scanning rate of 1.2 min À1 .
The determination of the elemental composition was conducted by Optima 2000 DV inductively coupled plasma-optical emission spectrometer (PerkinElmer Inc., USA). N 2 adsorption-desorption isotherms were measured at liquid nitrogen temperature (À196 C) using a 3H-2000PS2 adsorption instrument (BeiShiDe Instrument). Before the measurements, the catalysts were pretreated at 250 C for 6 h under a high vacuum environment to remove physically adsorbed water. The specic surface area was calculated using the Brunauer-Emmett-Teller (BET) method and the pore volume and porous distribution was derived from Barrett-Joyner-Halenda (BJH) equation.
The morphology of the catalysts was investigated by transmission electron microscopy (JEM-1011) with a eld-emission gun operating at 200 kV. The samples were prepared by dropping an ethanol suspension of the powder particles on a carbon lm supported copper grid.
X-ray photoelectron spectrum (XPS) was employed to detect the chemical composition and states of the atoms on the catalysts, and conducted on a ESCALAB250Xi X-ray photoelectron spectrometer with C as the internal standard (C 1s ¼ 284.7 eV).
H 2 -TPR experiment was carried out on a Micro TP-5076 chemisorption analyzer. Prior to the reduction, the catalyst (about 50 mg) was pretreated in N 2 stream (20 mL min À1 ) at 300 C for 1.5 h and then cooled to room temperature. Aer that, the owing H 2 -N 2 mixture (10% H 2 by volume) was switched on, and the temperature was gradually increased from 50 to 800 C with a ramp of 10 C min À1 . The H 2 consumption was measured with a thermal conduction detector (TCD) that had been calibrated using the standard sample of CuO.
The number of acid sites was determined by NH 3 -TPD using the same instrument as that used for H 2 -TPR. Before the adsorption of NH 3 , the sample (about 100 mg) was rst treated at 300 C for 1.5 h and then decreased to 50 C in a N 2 ow of 20 mL min À1 . Aer the saturated adsorption of NH 3 (10 mL min À1 ) at 50 C for 30 min, the physically adsorbed NH 3 was subsequently degassed from the sample for 1 h under the N 2 ow (20 mL min À1 ). Finally, the sample was heated to 800 C with a rate of 10 C min À1 , the amounts of desorbed NH 3 was monitored by the thermal conduction detector (TCD). The TPD prole was decomposed into Gaussian curves to quantify the weak, medium and strong acid sites.
Py-IR spectra were measured on a Thermo Nicolet 380 FT-IR spectrometer to determine the type of acid sites. Before the measurement, the impurity was removed from the sample under 10 À2 Pa condition at 200 C for 2 h, and then pyridine vapor was introduced into the sample for the saturated adsorption. Aer the evacuation at 200 C for 1 h in owing He atmosphere, the Py-IR spectra were subsequently collected at room temperature.

Catalytic reaction and the product analysis
MPC was synthesized by the reversible disproportionation of DMC and DPC, and its purity was 99.8% determined by high performance liquid chromatography. 20 In a typical experiment, the catalytic performance test of MPC was performed in a threenecked glass beaker at atmospheric pressure. MPC of 150 mmol and a proper amount of the catalysts were placed into the beaker under a owing nitrogen and slowly heated under vigorous stirring. When the reaction temperature was reached, it was kept for the reaction. As the reaction processed, DMC might be distilled out and continuously collected in a ask beaker to promote the equilibrium reaction towards the production of DPC. Upon the completion of the reaction, the reaction was decreased down to ambient temperature. The reaction mixture was analyzed by GC-MS instrument (HP-6890/ 5973) equipped with HP-5 capillary chromatography packed column. Subsequently, the products were quantitatively analyzed with a gas chromatograph (GC-7890A) equipped with HP-5 capillary column (30 m Â 0.32 mm Â 0.25 mm) and a ame ionization detector (FID).

Results and discussion
3.1 Characterization of the catalysts 3.1.1 BET characterization. The BET surface area, pore volume and pore size distribution determined by nitrogen adsorption-desorption measurement for all the catalysts and their corresponding supports were listed in Table 1. The average pore diameters for these materials were in the mesoporous range, and the surface area and pore volume of the catalysts decreased in comparison to the corresponding supports. It was noted that the SiO 2 support displayed the highest BET surface area among all support materials, moreover, aer impregnating, the PbO/SiO 2 catalyst also exhibited a larger BET specic surface area and pore volume in contrast to PbO/MgO, PbO/ Al 2 O 3 , PbO/ZrO 2 and PbO/TiO 2 , and the orders of both the surface area and pore volume for them were as follows: PbO/ SiO 2 > PbO/ZrO 2 > PbO/MgO > PbO/TiO 2 > PbO/Al 2 O 3 , in the same orders as that for the pure supports. The average pore diameters of the catalysts also decreased than those of the pure supports. This was probably due to PbO being dispersed on the surface of these supports or lled in the pore structure aer the lead impregnating, which might result in the decrease of these textural properties.
3.1.2 XRD characterization. To detect the structures of Pb species on various supports, XRD measurement were performed for both the catalysts and the pure supports. As could be seen in Fig. 1, XRD patterns of pure supports exhibited their respective characteristic peaks, while there were some differences with their corresponding Pb-based catalysts among the characteristic peaks. For instance, SiO 2 exhibited a broad peak at around 22 , and aer PbO loading on its surface no PbO crystalline phase diffraction peak was observed for PbO/SiO 2 . The diffraction peak at 25.4, 37.9, 48.1, 54.1 and 62.5 were attributed to  Fig. 2. Firstly, for SiO 2 , ZrO 2 and MgO supports, there were no evident reduction peak appearances in the temperature ranges of 100-800 C. By contrast, the reduction characteristic peaks at 730 and 781 C were observed for pure TiO 2 and Al 2 O 3 , respectively. Secondly, for various Pbbased catalysts, only one distinct reduction characteristic peak was observed in PbO/SiO 2 , PbO/TiO 2 , PbO/ZrO 2 and PbO/ MgO respectively, suggesting that mainly PbO was reduced. As reported in the literatures, 25,34 pure PbO revealed a clear reduction peak of 588 C, which was assigned to PbO species reduction. Therefore, these peaks in PbO/SiO 2 (552 C), PbO/ TiO 2 (591 C), PbO/ZrO 2 (348 C) and PbO/MgO (410 C) were typically ascribed to the reduction of highly dispersed surface PbO species, nevertheless, it was noteworthy that there were also obvious differences in the peak positions for them, suggesting that PbO loading onto different supports had a great inuence on the reduction temperature of PbO species. Notably, the reduction temperature of PbO species on SiO 2 , ZrO 2 and MgO supports were shied to lower temperatures. This was mainly due to the strong interaction between PbO species and the supports which was favorable for the stability of the catalysts during reaction. Whereas in the case of PbO/TiO 2 the reduction peak shied to a higher temperature and this was mainly due to the poor interaction between PbO and the TiO 2 support. Thus, the results suggested that the interaction strengths between the active component and the different supports were remarkably different.
Thirdly, for PbO/Al 2 O 3 , ve reduction characteristic peaks appeared. The rst reduction peak at 595 C corresponded to the surface PbO specie reduction, and the second reduction peak at 676 C might result from the reduction of the bulk PbO, as the XRD results displayed the existence of PbO crystalline diffraction peaks, which was primarily assigned to the weak interaction between PbO and the Al 2 O 3 support so that the PbO species was difficult to reduce at a higher reduction temperature. The following three peaks with reduction temperatures at 702, 739 and 760 C were probably associated with the reduction of the Pb-Al mixed species formed via the interaction at high temperature (>700 C) of partially reduced Pb-species and the support. 35,36 Overall, according to the reduction temperatures of PbO species, it could be inferred that the interaction strengths between PbO and the supports were in the following order of PbO/ZrO 2 > PbO/MgO > PbO/SiO 2 > PbO/TiO 2 > PbO/Al 2 O 3 , which might lead to different catalytic properties.
3.1.4 TEM characterization. Fig. 3 displayed the TEM images of the catalysts. As shown in Fig. 3(A and a), PbO/SiO 2 mainly represented the spherical-like structure, and the PbO particles might be clearly observed, indicating that they were highly dispersed on the surface of the SiO 2 . For PbO/TiO 2 , PbO/ ZrO 2 , PbO/Al 2 O 3 and PbO/MgO displayed in Fig. 3(B, b)-(E, e), their morphology were remarkably different from each other.  The diameters of all these catalysts were also not uniform, and the dissociation of PbO particles and the supports could not be detected by visual inspection. Comparatively speaking, PbO/ ZrO 2 exhibited a smaller particle size than the other catalysts.
3.1.5 XPS characterization. XPS analyses were employed to identify the chemical composition and states of the catalysts. The XPS survey spectrum veried the presence of Pb in the supported catalysts. As shown in Fig. 4(A), the denitive binding energy peak located at approximately 142.9-144.0 eV and 137.9-139.2 eV corresponded to Pb 4f 5/2 and Pb 4f 7/2 , 23,37 indicating that Pb 2+ species existed on the surface of these catalysts. Fig. 4(B) showed that the O 1s spectra of the catalysts (except PbO/Al 2 O 3 catalyst) could be deconvoluted into two Gaussian peaks (O I and O II ). In general, O I was ascribed to the lattice oxygen in the literature, 28,34 and O II was related to the adsorbed oxygen species on the catalyst surface. The binding energy of the O 1s spectra was different in each catalyst, which was probably due to the different support properties for the supported Pb-based catalysts.
3.1.6 NH 3 -TPD characterization. The process of MPC disproportionation to produce DPC was mainly dependent on the acidity of the catalyst. Herein NH 3 -TPD was carried out for measuring the acidic properties of the catalysts. The NH 3 -TPD proles were displayed in Fig. 5. According to the literatures, 38,39   Paper the acid sites of the catalysts were usually classied based on desorption temperature range such as weak (around 100-200 C), medium (around 200-400 C), and strong (above 400 C) acidic sites, respectively. As shown in Fig. 3, the acidity of the catalysts depended on the nature of the supports, and PbO/SiO 2 , PbO/TiO 2 and PbO/ZrO 2 exhibited remarkable and broad NH 3 desorption peaks, suggesting the wide distribution of acid strengths from weak, medium and strong acid sites respectively. For PbO/MgO catalyst, only one strong high-temperature NH 3 desorption peak at the range of 250 to 700 C was detected, suggesting a strong acid property. PbO/Al 2 O 3 showed much stronger NH 3 desorption peak in the high-temperature region, indicating more strong acidic sites existed on the surface. Besides, the acidity data was also calculated and the results were given in Table 2. Clearly, the total acid amount in PbO/ MgO was remarkably higher than those of other catalysts, and followed the sequence of PbO/MgO > PbO/ZrO 2 > PbO/SiO 2 > PbO/TiO 2 > PbO/Al 2 O 3 . According to previous studies, the acidity of the catalysts was correlated with the activity, thus the high acidity of the catalyst was benecial for the activity.
3.1.7 Py-IR characterization. The type of acid sites was performed by pyridine adsorption FT-IR spectroscopy. Fig. 6 showed the Py-IR spectra of the catalysts. As shown in Fig. 6, PbO/SiO 2 , PbO/TiO 2 , PbO/ZrO 2 and PbO/MgO exhibited two remarkable absorption peaks at around 1438-1446 and 1584-1596 cm À1 , which were attributed to Lewis and Brønsted acidic sites, respectively. 40,41 However, for PbO/Al 2 O 3 , only one absorption peak was displayed at 1446 cm À1 that suggested the presence of only Lewis acidic sites. Meanwhile, the concentration ratio of Lewis to Brønsted acid sites was calculated by semiquantitative analysis based on the relative peak area corresponding to Lewis and Brønsted acidity (A L /A B ) in Table 2. Apparently, the A L /A B ratios were higher for PbO/ZrO 2 and PbO/ MgO compared to others, and combined with the acid amount results determined by NH 3 -TPD, which indicated that PbO/ZrO 2 and PbO/MgO contained much more Lewis acidic sites.

Results of the catalytic performance evaluation
The catalytic performance evaluation results for MPC disproportionation over the catalysts were revealed in Table 3. Pure supports were basically inactive under the current studies, while PbO loading on these supports exhibited considerably different catalytic performances. The catalytic activities of PbO/MgO and PbO/ZrO 2 were close and obviously higher than those of PbO/ SiO 2 , PbO/Al 2 O 3 and PbO/TiO 2 , and the sequence of the activities for the MPC conversion was as follows: PbO/MgO > PbO/ ZrO 2 > PbO/SiO 2 > PbO/Al 2 O 3 > PbO/TiO 2 . Previous studies displayed that the BET surface area and acidity of catalyst played signicant role in the catalytic activity and product selectivity. Nevertheless, this order of the activity was not correlated with the sequence of BET surface area, indicating that the surface area might be not a main factor determining the activity. Therefore, the catalytic activity might be closely related to the acidity. Combining with the characteristic results of Py-IR and NH 3 -TPD, the relationship between the activity and Lewis acidity was displayed in Fig. 7. It was clear that the activity depended on the acidic site, and the more Lewis acid site was,  the higher the activity was. And the decrease sequence of MPC conversion was in accordance with the order of the Lewis acid site, suggesting that the Lewis acid was the critical factor to determine the catalytic activity. What's more, although the highest DPC yield was obtained over PbO/MgO due to the acidity of this catalyst being the highest among all the catalysts, yet it was also noteworthy that a strong acid strength could induce the decarboxylation decomposition of MPC into the byproduct anisole. Therefore, it could be ascertained that the higher anisole content was mainly due to its excessively strong acidity. 20 Moreover, PbO/Al 2 O 3 with more strong acid strength exhibited the higher anisole content, which further testied the reason. Thus, the results also suggested that medium and moderately strong acidic sites were favorable for the higher DPC selectivity.

Recyclability of the catalysts
Recycling experiments of the catalysts were performed to check the stability. Aer each run, the catalysts were separated by ltration, washed thoroughly with DMC and dried in vacuum at 110 C overnight, and then reused for next cycle reaction. As displayed in Fig. 8    In heterogeneous catalysis, the leaching of active component was oen the main reason which decreased the activity for further reuse. Therefore, the recovered catalysts were detected by ICP-OES analysis (Table S1 †), and the content of Pb in the fresh and the third reused catalyst showed that Pb content had no signicant loss over PbO/MgO and PbO/ZrO 2 , but dropped markedly over PbO/SiO 2 , PbO/Al 2 O 3 and PbO/TiO 2 , respectively. Hence, the leaching loss of Pb might lead to the decrease of catalytic activities for the recovered PbO/SiO 2 , PbO/Al 2 O 3 and PbO/TiO 2 catalysts. Combined with the characteristic results of H 2 -TPR, the excellent recyclability of PbO/MgO and PbO/ZrO 2 was possibly attributed to their strong interactions of the supports with highly dispersed Pb, which was potentially important for practical applications. Besides, the recovered catalysts aer the third use were also detected by XRD characterization (Fig. S1 †). The result indicated that the diffraction peaks of the third reused catalysts were basically identical to those of the fresh catalysts, suggesting that with the exception of PbO/Al 2 O 3 , PbO was still amorphous or microcrystalline states in the recovered catalysts. Therefore, MgO and ZrO 2 might be suitable and promising support materials for the better preparation of Pb-based catalysts to transform MPC disproportionation to efficiently synthesize DPC in comparison with SiO 2 , Al 2 O 3 and TiO 2 .

Conclusions
In summary, the support nature modied the structural properties of Pb catalysts and exerted marked inuence on their catalytic performances. The catalysts were characterized by physicochemical techniques, and the results showed that PbO was dispersed better on SiO 2 , TiO 2 , ZrO 2 and MgO than Al 2 O 3 , and the stronger metal-support interaction was obtained over MgO and ZrO 2 . The active results showed that the catalytic performances over MgO and ZrO 2 to load Pb catalysts were signicantly superior to those of SiO 2 , Al 2 O 3 and TiO 2 , because of the higher active component dispersion and more Lewis acid active sites. Furthermore, the excellent recyclability in the case of PbO/MgO and PbO/ZrO 2 was potentially important in industrial application. Thus, from this study it could be noted that the more Lewis acid site was present, the more the activity was, and the nature of the support was a key parameter for MPC disproportionation.

Conflicts of interest
There are no conicts to declare.