Precious-metal nanoparticle (NP) catalysts play a vital role in fine chemical production, petroleum refining, and other sectors of the modern chemical industry [1, 2]. The active metal phase is always dispersed on a suitable support such as metal oxides (e.g., Al2O3 [3], ZrO2 [4], and TiO2 [5]), zeolites [6], and carbons [7, 8] to enhance the activity and stability of the catalyst. The support type or properties such as acid-base and redox properties [9] can affect the formation of metal NPs or metal-support interactions [10, 11]; this affects the activity and stability of the solid catalyst, and therefore selection of an appropriate support is one of the most important factors to be considered in the design and synthesis of effective supported catalysts. Among various supports, carbon materials are one of the most common host materials for metal NPs [12, 13] because of the wide availability of carbon sources [14–16] and their excellent wettability by organic compounds [17]. Various carbon-based catalysts have been developed, e.g., Pd/C [18, 19], Pt/C [20, 21], and Ru/C [22]. However, because of the weak interactions between the metal NPs and the carbon surface, it is difficult to control and maintain the size of the deposited noble-metal particles during the catalytic process, and NPs are prone to leach from the support [23]. The chemical and catalytic properties of traditional carbons therefore do not meet the high standards required for catalysis.
Recently, N-doped ordered mesoporous carbons (NOMCs) have emerged as important materials with various advantages [24–28]. The nitrogen atoms incorporated in the carbon architecture can not only change the physicochemical and electronic properties of the support, but also serve as basic or coordination sites that favor metal-support interactions and thereby stabilize small metal NPs [29]. Previous studies have suggested that nitrogen functional groups influence the nucleation and growth kinetics during catalyst NP deposition [30], resulting in smaller catalyst particles and increased dispersion [31]. This is useful because decreasing and controlling the supported metal NP size can significantly increase their surface: bulk atomic ratio. This can affect the total energy and fundamental properties of the NPs and therefore their catalytic performance [29]. Increased interactions (such as coordination and electron transfer) between small metal NPs and nitrogen-functionalized supports can reduce the mobility of the catalytic metal and prevent aggregation or sintering, leading to enhanced durability and stability [9, 29]. Moreover, it has been shown that nitrogen groups in carbon materials can modify the electronic structure of catalyst NPs or serve as bifunctional groups to activate specific substrates, which might enhance the intrinsic catalytic activity [32]. Additionally, compared with amorphous supports, ordered mesoporous carbons with high surface areas, and large pore volumes and pore sizes facilitate mass transfer, and give a confinement effect, which further prevents metal NP growth. These effects all improve the catalytic efficiency [33, 34]. NOMC matrixes are therefore ideal alternative catalyst supports for loading metal NPs. Much research has led to the use of NOMCs as supports for NP catalysts for efficient hydrogenation [35, 36], oxidation [37, 38], coupling [39], and hydrochlorination reactions [40]. However, there are few reports of the synthesis of highly ordered mesoporous carbon materials with high nitrogen contents using simple methods, and their use as catalyst supports [26, 29, 41]. The development of element-doped mesoporous carbons and their corresponding supported small NP catalysts is therefore still needed. Catalysts with more specific properties are also needed.
In this work, we used a NOMC as a support for a Pt-based solid catalyst. The NOMC was synthesized by a soft-templating method using 3-aminophenol as both the nitrogen and carbon source and F127 [poly(ethylene oxide)–poly(propylene oxide)–poly(ethylene oxide) triblock copolymer, EO106PO70EO106] as a template (Fig. 1). Polymerization and successive carbonization steps at high temperatures gave a carbon material with a mesoporous structure and high content of nitrogen functional groups. In situ doping with nitrogen using nitrogen-containing precursors gives homogeneous incorporation of nitrogen into carbons, making such materials excellent model supports for exploring the impact of N-doping on the activities and stabilities of Pt-based catalysts. The effects of the NOMC support on the Pt NPs were investigated, and the catalytic activity and selectivity in the hydrogenation of halogenated nitrobenzenes were compared with those for a commercial Pt/C catalyst and the corresponding nitrogen-free ordered mesoporous carbon (OMC) catalyst, Pt/OMC. Finally, to gain a fundamental understanding of the impact of N-doping on the durability of carbon support materials, recycling tests were performed to evaluate the overall stability of the material.
All chemicals were used as received unless otherwise stated. Pluronic F127 was purchased from Sigma Aldrich. 3-aminophenol, phenol, formaldehyde (37 wt%) and hexamethylenetetramine (HMT) were purchased from the Shanghai Chemical Reagent Company of the Chinese Medicine Group. H2PtCl6·6H2O, a commercial Pt/C (5 wt%) catalyst, and halogenated nitrobenzenes and other nitroarenes were purchased from the Aladdin Reagent Company.
The NOMC support was synthesized using a previously reported soft-templating method [26]. 3-Aminophenol (0.654 g) and HMT (0.42 g) were dissolved in water (75 mL) under vigorous stirring. Pluronic F127 (0.47 g) dissolved in water (5 mL) was added. The mixture was stirred at 35 ℃ for 24 h; the color of the solution changed from colorless to saffron yellow. The solution was aged at 95 ℃ under static conditions for 8 h. The solid product was recovered by filtration, washed with deionized water, and air-dried at 50 ℃ under vacuum overnight. The obtained polymer was carbonized at 600 ℃ for 3 h in a N2 atmosphere at a heating rate of 1 ℃/min to give the NOMC support.
A nitrogen-free OMC was prepared using a solvent-evaporation-induced aggregate assembly approach. In a typical procedure, 20 wt% NaOH aqueous solution (0.13 g) was added to melted phenol (0.61 g) at 42 ℃ in a flask under stirring for 10 min. Formalin (37 wt% formaldehyde, 1.05 g) was added. The mixture was stirred for 1 h at 70 ℃ and then cooled to room temperature. The pH was adjusted to 7.0 using HCl solution. Water was removed by vacuum evaporation and ethanol was added to form a 20 wt% ethanolic solution. An ethanolic resol solution (5.0 g) and tetraethyl orthosilicate (2.08 g) were added to a mixture of ethanol (8.0 g), HCl solution (0.2 mol/L, 1.0 g), and Pluronic F127 (1.6 g). The mixture was stirred at 40 ℃ for 2 h and transferred to a dish to allow ethanol evaporation at room temperature for 24 h. Thermopolymerization was performed at 100 ℃ for 24 h. Calcination was performed in a tubular furnace at 350 ℃ for 2 h and 900 ℃ for 2 h under a N2 flow to give a mesoporous carbon-silica nanocomposite. The carbon-silica nanocomposite was immersed in 10 wt% HF solution for 24 h to remove the silica and give mesoporous carbon.
An aqueous solution of the metal precursor was prepared by dissolving H2PtCl6·6H2O (1.00 g) in water (100 mL) at room temperature. The prepared NOMC powder (0.282 g) was impregnated with H2PtCl6 solution (1.5 mL). After stirring for 2 h at room temperature, water was removed by rotary evaporation and the solid was dried in a vacuum at 50 ℃. The solid was reduced in a 5% H2-95% N2 mixed atmosphere at 200 ℃ for 2 h at a heating rate of 1 ℃/min to give the solid Pt/NOMC catalyst.
Transmission electron microscopy (TEM) was performed using an FEI Tecnai G2-TF30 electron microscope operated at an accelerating voltage of 300 kV. N2 adsorption-desorption isotherms were recorded using an ASAP2020 volumetric adsorption analyzer. The surface area was calculated from the adsorption branch in the relative pressure range 0.05–0.15 using the Brunauer-Emmett-Teller (BET) equation. The total pore volume was calculated at a relative pressure of p/p0 > 0.99. Pore size distribution curves were obtained by analysis of the adsorption branch of the isotherm. X-ray diffraction (XRD) patterns were obtained using a desktop powder X-ray diffractometer at 40 kV and 30 mA with Cu Kα radiation. X-ray photoelectron spectroscopy (XPS) was performed using an Axis Ultra Dld spectrometer, with an Al Kα radiation (1486.6 eV) incident beam in a N2 atmosphere.
In a typical procedure, a halogenated nitrobenzene (2.56 mmol) and ethanol (5 mL) were added to a test tube containing the desired amount of Pt/NOMC (25 mg, 2.56 μmol Pt). The test tube was transferred to a stainless-steel autoclave and the autoclave was sealed. After purging several times with H2, the pressure was adjusted to 1 MPa and the system was stirred at 25 ℃ in a constant-temperature bath. After the reaction, the H2 pressure was released and the solid catalyst was separated by centrifugation. The conversion and selectivity were determined by gas chromatography using an HP-5 capillary column (30 m × 0.32 mm × 0.5 μm). In the recycling tests, the solid catalyst obtained by centrifugation was added to fresh substrate and solvent, and directly used to perform the next catalytic cycle.
The ordered mesoporous properties of the synthesized N-doped carbon support and corresponding solid Pt/NOMC catalyst were investigated using TEM. Fig. 2(a) and (b) show that the NOMC sample had a high degree of periodicity viewed from different directions. The low-angle XRD pattern of NOMC shows an intense peak corresponding to the (110) reflection along with weak shoulder peaks assigned to the (200) and (211) reflections (Fig. 2(e)). These results suggest that an ordered mesoporous support with cubic Im3m symmetry was obtained [26, 42]. The XRD pattern after immobilization of Pt NPs, i.e., of the solid catalyst Pt/NOMC, is similar to that of the support, indicating that the periodic mesoporous structure was retained after Pt NP loading. TEM was used to determine the size distribution of the supported Pt NPs on the NOMC; almost no Pt NPs were observed on the external surface of the support (Fig. 2(c) and (d)), possibly because the Pt NPs were mainly loaded into the channels of the mesoporous carbon. The images show that ultrasmall and uniform Pt NPs were homogeneously dispersed on the surfaces of the NOMC pores. The particle size distribution of the Pt NPs is shown in the inset in Fig. 2(c); the mean particle diameter was about 1.0 ± 0.5 nm. The well-dispersed ultra-small Pt NPs, without aggregation, can be attributed to the stabilizing effect of the incorporated nitrogen and the confinement effect of the ordered mesoporous structure.
The N2 adsorption isotherms of NOMC and Pt/NOMC are type Ⅳ with a type-H2 hysteresis loop, indicating a mesoporous structure and spherical pores with narrow pore windows (Fig. 2(f)). This result is consistent with the TEM and XRD results. The isotherm shape was preserved after Pt NP loading, indicating that the Pt NPs did not block or alter the pore system. It is worth noting that the adsorption and desorption branches are not closed, a phenomenon that is often observed for porous polymers and sometimes for porous carbons, and which is not yet fully understood [26]. The BET surface area and total pore volume of the NOMC support were calculated to be 418 m2/g and 0.22 cm3/g, respectively, which might be beneficial for metal particle dispersion and fast diffusion of substrates and products during the catalytic process. After Pt NP loading, the specific surface area (344 m2/g) and pore volume (0.18 cm3/g) decreased slightly; this might be related to Pt loading in the mesochannels. The pore size of Pt/NOMC (Fig. 2(g)) differed only slightly from that of NOMC, suggesting that the ordered mesostructure was well preserved after Pt NP loading, and the Pt/NOMC mesostructure was not blocked by Pt NPs.
XPS was used to investigate the interactions between the small Pt NPs and the nitrogen-functionalized support. Fig. 3(a) shows that the NOMC gave strong signals from carbon, nitrogen, and oxygen elements. The N 1s XPS spectrum shows two peaks; the lower energy peak, at 398.58 eV, can be assigned to pyridinic nitrogen, and the peak centered at 400.76 eV is attributed to pyrrolic nitrogen (Fig. 3(b)) [26]. Elemental analysis based on sample combustion showed that the bulk nitrogen content of the NOMC support was about 9.58 wt%. The Pt XPS spectrum of Pt/NOMC shows a doublet corresponding to Pt0 4f7/2 (71.41 eV) and 4f5/2 (74.91 eV). The weak peaks at 72.52 and 76.8 eV correspond to Pt2+ species, and no peaks from Pt4+ were observed, which suggests that all the Pt4+ precursor had been reduced in situ by treatment with H2 (Fig. 3(c)). The XPS data suggest that metallic Pt0 was the major phase formed on the catalyst surface (~76%). Notably, the binding energy of Pt0 4f7/2 in Pt/OMC, i.e., without incorporated nitrogen, differs from the corresponding value for Pt/NOMC (71.41 versus 71.82 eV). This slight deviation could be related to weak interactions between the catalyst and the support.
p-Chloroaniline is an industrially important organic intermediate used in the synthesis of fine chemicals and pharmaceuticals [7]. The direct hydrogenation of halogenated nitrobenzenes is an attractive route for the formation of p-chloroaniline. However, hydrodehalogenation often occurs as a side reaction during the catalytic process, resulting in low catalytic selectivity. The development of functional catalysts with specific properties that enable control of the catalytic activity and selectivity is therefore important. Solid catalysts that have both high activity and selectivity are highly desirable. In this study, we selected the hydrogenation of halogenated nitrobenzenes as a model reaction for evaluating our solid catalyst, i.e., Pt/NOMC, which has a high surface area, high nitrogen content, and ultrasmall Pt NPs.
The reaction was performed with ethanol as the solvent under mild conditions (25 ℃, 10 bar H2). Fig. 4(a) shows the catalytic kinetic curve for the hydrogenation of p-chloronitrobenzene over Pt/NOMC. Clearly, Pt/NOMC efficiently catalyzed hydrogenation and the conversion increased smoothly with reaction time. The reactant was fully converted in 0.5 h, and the selectivity for p-chloroaniline (99%) was high. These results are encouraging because hydrodehalogenation reactions were almost completely avoided. The relationship between conversion and the reaction time was almost linear, indicating that the hydrogenation was zero order. The turnover frequency (TOF) calculated from the kinetic plot (Fig. 4(a)) was high, i.e., 2508 h−1. The Pt/NOMC properties suggest that the high activity and selectivity of the supported solid catalyst could be related to its high surface area, nitrogen functional groups, and ultrasmall NPs.
The performance of the Pt/NOMC catalyst was compared with those of other types of catalyst, i.e., Pt/OMC, a commercial Pt/C catalyst, and metal-oxide-supported catalysts (Pt/SiO2 and Pt/TiO2), in the hydrogenation of p-chloronitrobenzene under the same reaction conditions. The Pt loadings of the supported catalysts were all about 2 wt% (except in the case of the commercial Pt/C catalyst, which was 5 wt%), similar to the loading on the Pt/NOMC catalyst. The particle sizes and distributions differ depending on the support, therefore these solid catalysts were characterized using TEM. Fig. 4(c) shows that without the stabilizing effect of doped nitrogen functionalities, many Pt NPs were distributed on the external surface of the ordered mesoporous carbon but not inside the channels, and the particles were larger (4.4 ± 1.7 nm). These results indicate that the nitrogen species incorporated into the carbon support play a key role in preventing uncontrolled growth of the Pt NPs and stabilize the Pt NPs formed mainly inside the mesoporous channels. The catalytic activity of Pt/OMC in the hydrogenation of p-chloronitrobenzene was similar to that of Pt/NOMC, and 100% conversion was achieved in the same reaction time (Fig. 4(b)). However, the catalytic selectivity was a little lower (97.3% vs 99%), possibly because of the weak interactions between Pt and the NOMC support, shown by the XPS results. For the commercial Pt/C catalyst, only 89% conversion and 90% selectivity were achieved under similar reaction conditions. The higher activities of Pt/NOMC and Pt/OMC can be attributed to the ordered mesoporous structure, which could promote homogeneous distribution of Pt NPs and facilitate mass transfer. The activity of the Pt/SiO2 catalyst, which is a commonly used solid catalyst, was low (conversion 46% and selectivity 93%). The low catalytic activity might be related to its poor oil wettability compared with those of other carbon supports. Furthermore, the particles were much larger than those on the other supported catalysts (Fig. 4(e)). The Pt/TiO2 catalyst had small particles (Fig. 4(f)) and full conversion was achieved, but the selectivity was only 91%. These results show that the NOMC performed better than other common supports, especially in terms of reaction selectivity.
Leaching of metal NPs from supports and the growth of metal NPs during reactions are major problems for supported ultrasmall metal NP catalysts, and lead to a sharp decrease in the catalytic activity. The stability of Pt/NOMC was investigated by performing recycling tests in the hydrogenation of p-chloronitrobenzene. Fig. 5(a) shows that the solid catalyst was easily recovered by simple centrifugation and reused more than 10 times without loss of catalytic activity and selectivity. The Pt concentration in the liquid reaction solution, determined using inductively coupled plasma atomic emission spectroscopy, was less than 0.1 ppm, indicating that Pt leaching into the liquid phase was negligible. The Pt/NOMC catalyst recovered after the 10th run was examined using TEM (Fig. 5(c)). There were no obvious differences between fresh Pt/NOMC and the recycled catalyst, although Fig. 5(d) shows that the particle size had increased slightly (1.0 ± 0.5 nm vs 2.0 ± 1.5 nm). The recyclability of Pt/OMC was also investigated for comparison (Fig. 5(b)). The conversion decreased with increasing cycle number. After 10 cycles, the conversion and selectivity had decreased to 57% and 88%, respectively. These results confirm that the interactions between small Pt NPs and the NOMC functional support are crucial in maintaining the activity and acting as an inhibitor during recycling.
Based on the superior performance of the NOMC-supported catalyst, the reactions of other halogenated nitrobenzenes and nitroarenes were investigated. The data in Table 1 shows that o-chloronitrobenzene and m-chloronitrobenzene were fully converted, without hydrodehalogenation reactions, under the test conditions (99.5% and 98.8% selectivity, respectively, entries 1 and 2). p-Bromonitrobenzene also reacted smoothly to give the corresponding amine with 95.5% selectivity (entry 3). Other nitrobenzene compounds with various substituent groups, i.e., hydroxy, methoxy, formyl, and ester groups, were also investigated. Excellent catalytic activities were achieved with p-nitrophenol, o-nitrophenol, p-nitroanisole, and methyl 4-nitrobenzoate as substrates, and > 99% selectivities were achieved under the reaction conditions (entries 4–7). However, with 4-nitrobenzaldehyde and 4-nitrobenzyl alcohol as substrates, the selectivities were lower (85.4% and 70.0%, entries 8 and 9) because side reactions occurred more easily.
A N-doped ordered mesoporous carbon material with a high nitrogen content (9.58 wt%) was successfully prepared using a soft-templating method. Ultrasmall Pt NPs were uniformly distributed on the nitrogen-functionalized support to give a Pt/NOMC catalyst. The N-doped carbon gave well-dispersed and stable small NPs, and enhanced catalytic hydrogenation activity and selectivity compared with those of conventional supports, namely Pt/OMC, commercial Pt/C, and metal-oxide-supported catalysts (Pt/SiO2 and Pt/TiO2). The Pt/NOMC catalyst showed good stability and was reused 10 times without loss of catalytic activity.
2017,
Vol. 38 Issue (7): 1252-1260
PDF
2017,
Vol. 38 Issue (7): 1252-1260 PDF