As an important chemical intermediate,p-chloroaniline (p-CAN) is widely used in the synthesis of herbicides,dyes,drugs,and pesticides^[1,2]. Until now the best way to synthesize this compound is the direct hydrogenation ofp-chloronitrobenzene (p-CNB) due to its low impact on the environment. The main obstacle to this reaction is the dehalogenation reaction which also occurs. Raney Ni is the most widely used catalyst in this reaction due to its high activity,but in order to maintain the selectivity of the reaction,inhibitors must be added into the reaction system^[3]. Noble metals have also been applied in this reaction,but their selectivity is still far from satisfactory and the high cost of noble metals has prevented large scale application. In recent years nanotubular materials have been widely studied. Our previous report showed that noncrystalline NiPB nanotube can be used as a highly effective heterogeneous catalyst for this reaction^[4]. Other methods to prevent the dehalogenation process include modulating the interaction between the metal and the support^[5,6,7] and using an alloy of different metals^[8,9]. But it still is a challenge in the hydrogenation ofp-CNB top-CAN to achieve high selectivity at high conversion with a good catalyst.

As a unique metal-free polymer material,carbon nitride (CN) has been intensively studied in different areas due to its versatile chemical and physical properties^{[10, 11, 12, 13, 14, 15, 16]}. Its unique electronic structure and outstanding chemical stability make it a promising candidate for applications in catalysis such as photocatalytic water splitting,oxygen reduction,and the selective conversion of organic functional groups^{[17, 18, 19, 20, 21, 22]}. CN has also been used for supporting noble metals ^{[23, 24, 25, 26]} and as an energy storage material^[27]. By different modification methods,the properties of CN can be tuned to make it suitable for different applications. For example,by doping heteroatoms such as S,P,B,and other functional groups into the structure of CN,its band structure and surface property can be modified ^{[28, 29, 30, 31, 32]} for the further improvement of CN. In our previous work,we encapsulated highly dispersed Ni on Al₂O₃ with CN. It was demonstrated that in this catalyst,the Ni was reduced to its metallic state and by donating its electron to CN,a strong interaction was formed between the nickel and CN; the points of the catalyst are shown inFig. 1. By this method,we successfully modified the electronic properties of CN and endowed it with the ability to adsorb and activate hydrogen^[33]. In this work,we use this nanocomposite catalyst as an efficient heterogeneous catalyst for the hydrogenation ofp-CNB top-CAN. The catalyst showed very high selectivity (>99.9%) at high conversion (>95%).

Fig. 1. Schematic showing the principle of the catalyst composed of nickel modified carbon nitride comprising (1) electron donation from nickel to CN,(2) dissociative adsorption of hydrogen on CN,and (3) highly selective reduction ofp-CNB top-CAN by the activated hydrogen.

First,a synthesized nanostructured Al₂O₃ was employed as the support. The Al₂O₃ was prepared by a hydrothermal approach with an alumina sol as precursor,and then calcined in air at 823 K for 12 h. An aqueous solution of Ni(NO₃)₂ was impregnated onto the Al₂O₃ by incipient wetness impregnation,dried in air at 373 K for 12 h and finally calcined in air at 723 K for 4 h. The product was denoted as NiO/Al₂O₃. Before the catalytic test,the NiO/Al₂O₃ was reduced in 5 vol% H₂/N₂ atmosphere at 873 K for 6 h. The product was denoted as Ni/Al₂O₃. The nanocomposite catalyst was synthesized by adding 1 g NiO/Al₂O₃,2 g ethylenediamine,and 4 g carbon tetrachloride into 80 mLm-xylene. The system was heated to 363 K and stirred for 4 h and then heated to 413 K and stirred for another 4 h. After that the dark brown solid mixture was dried in an oven for 12 h and then ground into a fine powder. Finally the composite was treated in N₂ (50 mL/min) at 873 K with a heating rate of 2 K/min and kept at these conditions for 6 h to carbonize the polymer. The product was denoted as CN/Ni/Al₂O₃. The sample composed of CN and Al₂O₃ was synthesized by the same method using Al₂O₃ as the starting material.

X-ray diffraction (XRD) patterns were obtained on a Phillips X’Pro diffractometer using CuK_α radiation (λ = 0.15418 nm) at 40 kV and 25 mA. Fourier transform infrared (FT-IR) spectra of the samples were obtained on a Bruker Vertex 70 spectrophotometer using KBr pellets,recorded with 64 scans with a resolution of 4 cm^‒1. Transmission electron microscopy (TEM) measurements were conducted with a JEOL JEM-200CX instrument at an accelerating voltage of 200 kV. Thermogravimetric (TG) curves were taken on an STA 449C-Thermal Star 300 instrument (NETZSCH,Germany) in air with a heating rate of 10 K/min to 1073 K. The X-ray photoelectron spectroscopy (XPS) measurements were performed on a commercial XPS system (PHI 5000 VersaProbe) equipped with a hemispherical electron analyzer and monochromatic AlK_α X-ray exciting source. H/D exchange was conducted by the following steps. The catalyst sample (200 mg) was reduced in 5 vol% H₂/N₂ (30 mL/min) at a heating rate of 10 K/min to 423 K and kept at this temperature for 1 h. Then the sample was cooled in the same atmosphere to room temperature and the flow was switched to Ar (35 mL/min) to remove physical adsorbed H₂. Deuterium exchanged for proton present on the sample was measured by increasing the temperature to 873 K at a heating rate of 10 K/min under 5 vol% D₂/Ar (30 mL/min). The signal of HD was monitored by mass spectrometry.

The hydrogenation reaction was carried out in a bath-type autoclave reactor. In a typical reaction,the catalyst (50 mg) andp-CNB (50 mg) were added into the C₂H₅OH solvent (50 mL). Hydrogen was introduced into the reactor at room temperature to the pressure of 0.6,0.8,1.0,1.2,or 1.4 MPa. The reactor was heated to the required temperature and kept at this temperature for 2 h. After the reaction,the reactor was cooled to room temperature naturally and the catalyst was filtered off. The resulting solution was analyzed by gas chromatography with a flame ionization detector using an HP-5 column.

Figure 2(a) shows the XRD patterns of the samples. The characteristic peaks at 37.4°,39.5°,45.7°,and 67.2° can be attributed to the (311),(222),(400),and (522) planes of Al₂O₃,respectively. All the samples showed the characteristic XRD peaks indexed to Al₂O₃ and there was no peak which can be attributed to nickel oxide,metal nickel or NiAl₂O₄,which indicated that the nickel was highly dispersed. CN/Al₂O₃ and CN/Ni/Al₂O₃ did not show the typical (002) peak of CN. This was attributed to its relatively low intensity compared to that of Al₂O₃. After the encapsulation of CN,Fig. 2(b) shows the TEM image of CN/Ni/Al₂O₃ after heating in N₂. We can see the layered CN surrounding the surface of the catalyst with a thickness of 2‒3 nm. This was consistent with TG results (Fig. 2(d)).Fig. 2(e) shows the schematic structure of the catalyst,which is similar to the structure in our previous work.

Fig. 2. (a) XRD patterns of the samples; (b) TEM image of CN/Ni/Al2O3; (c) IR profiles of CN/Al2O3 and CN/Ni/Al2O3; (d) TG profiles of CN/Al2O3 and CN/Ni/Al2O3; (e) Schematic show of the structure of the catalyst based on the data of characterization.

Figure 2(c) shows the FT-IR profiles of the samples with CN. CN/Al₂O₃ and CN/Ni/Al₂O₃ showed similar absorption peaks: the C‒N stretching modes between 1120 and 1620 cm^‒1 and the N‒H stretching vibration modes at 3440 cm^‒1,which were ascribed to the CN surrounding the surface of the catalyst^[34].Fig. 2(d) shows the TG profiles of CN/Ni/Al₂O₃ and CN/Al₂O₃. The amounts of CN in CN/Ni/Al₂O₃ and CN/Al₂O₃ were 22 wt% and 23 wt%,respectively,indicating that the CN content in the sample was not changed by the existence of Ni. The DSC profiles demonstrated that the CN in CN/Ni/Al₂O₃ started to decompose at a lower temperature as compared to the CN in CN/Al₂O₃. This feature demonstrated an interaction between the nickel and the CN.

Figure 3 shows the XPS spectra of the CN/Ni/Al₂O₃ catalyst. The atomic ratios of the elements are shown in the figure and the remaining atomic proportion was attributed to Al and O in the catalyst. The C 1s core level spectrum revealed the appearance of two additional carbon components to the typicalsp² hybridized graphitic carbon at 284.6 eV. The peak at 285.4 eV was attributed to carbon atoms covalently bonded with N atoms in an aromatic structure. The peak at 286.6 eV was attributed to thesp² hybridized carbon bonded to the ‒NH₂ group in the aromatic ring. The N 1s core level spectra can be deconvoluted into four peaks located at 398.5,399.1,400.2,and 401.1 eV,which were assigned to pyridinic nitrogen,amine nitrogen,pyrrol nitrogen,and graphitic nitrogen,respectively^[35,36]. The C/N ratio was 5.5,which was nearly the same as the result of the elemental analysis with a value of 5.8. The Ni 2p_3/2 core level spectra comprised two peaks located at 855.1 and 856.8 eV,which were assigned to the Ni which has a strong interaction with the CN and alumina,thus keeping Ni itself in an electron deficient metal state.

Fig. 3. XPS spectra of CN/Ni/Al2O3. (a) C; (b) N; (c) Ni.

Figure 4 shows the quantitative H/D exchange results of CN/Ni/Al₂O₃ and CN/Al₂O₃. For CN/Ni/Al₂O₃,the main peakα of HD evolution was at 516 K with an amount of 156 μmol/g. This exchange temperature was higher than that of Ni/Al₂O₃ in our previous report. This phenomenon probably indicated that the activity of the hydrogen chemisorbed on the surface of CN/Ni/Al₂O₃ was inferior to that on Ni/Al₂O₃. Theβ species at the higher temperature was attributed to the residual H in the CN matrix. The density of adsorbed H atom was 0.65/nm²,and from the N content in the sample,the molar ratio of the chemisorbed H and N was 1/27. For the sample composed of CN encapsulated Al₂O₃,the peak temperature for H-D exchange was delayed to 770 K. Theβ species mainly represented hydrogen tightly bonded to the CN matrix,which is inert compared to the hydrogen chemisorbed on the surface of the Ni modified CN. The amount of this kind of hydrogen was more than 256 μmol/g. This observation demonstrated that theα species hydrogen adsorbed on the CN in CN/Ni/Al₂O₃ and then exchanged with deuterium at a rather mild temperature was the chemisorbed hydrogen active in the hydrogenation reaction.

Fig. 4. H/D exchange profiles of CN/Ni/Al2O3 (a) and CN/Al2O3 (b).

Table 1 shows the hydrogenation ofp-CNB catalyzed by the different samples. The Ni/Al₂O₃ catalyst possessed much higher activity than CN/Ni/Al₂O₃. After 4 h of reaction,the conversion ofp-CNB over Ni/Al₂O₃ reached 100%,but the selectivity top-CAN dropped from 75.6% to 63.1%. In contrast,the CN/Ni/Al₂O₃ catalyst possessed a relative low hydrogenation activity,but its selectivity towardsp-CAN was very high (>99.9%). This can be attributed to the relatively low activity of the chemisorbed hydrogen on the catalyst surface shown in the H/D exchange. As the reaction time was increased to 10 h,the conversion of thep-CNB reached 96.8% while the selectivity towardsp-CNB was kept above 99.9%. The CN/Al₂O₃ catalyst was completely inactive for the hydrogenation ofp-CNB,indicating that the hydrogen in CN itself is a nonreactive hydrogen species.

Table 1 The catalytic performance of the catalysts for hydrogenation ofp-CNB top-CAN. Entry Catalyst Time (h) p-CNB conversion (%) p-CAN selectivity (%) 1 Ni/Al2O3 2 82.3 75.6 2 Ni/Al2O3 4 100 63.1 3 CN/Al2O3 2 0 — 4 CN/Al2O3 4 0 — 5 CN/Ni/Al2O3 2 31.2 >99.9 6 CN/Ni/Al2O3 4 57.5 >99.9 7 CN/Ni/Al2O3 10 96.8 >99.9 Reaction conditions: Catalyst 50 mg,p-CNB 50 mg,ethanol 50 mL,initial hydrogen pressure 1 MPa,393 K.

Table 1 The catalytic performance of the catalysts for hydrogenation ofp-CNB top-CAN.

Figure 5(a) shows the catalytic performance of CN/Ni/Al₂O₃ under 393 K and different initial hydrogen pressure. As the hydrogen pressure increased,the conversion ofp-CNB increased accordingly. But after the initial hydrogen pressure reached 1.4 MPa,the selectivity top-CAN started to decrease.Figure 5(b) shows the catalytic activity of CN/Ni/Al₂O₃ at 1 MPa of initial H₂ pressure and different reaction temperatures. At 353 K the catalyst was almost inactive,and as the temperature increased,the conversion ofp-CNB increased. When the temperature reached 413 K,the selectivity began to decrease and at 433 K the selectivity had dropped to 80.6%. We concluded that the catalyst possessed very high selectivity top-CAN even at high reaction temperature and hydrogen pressure.

Fig. 5. Catalytic activity of CN/Ni/Al2O3 catalyst under different initial hydrogen pressures (a) and reaction temperatures (b).

Figure 6 shows that as the conversion ofp-CNB increased,the selectivity top-CAN was maintained at nearly 100%,indicating that dehalgenation did not occur as the reaction proceeded. Moreover,the catalyst can be reused several times without activity loss. This result demonstrated that the catalyst possessed high stability under the reaction conditions.

Fig. 6. Catalytic activity of CN/Ni/Al2O3 at 393 K with an initial hydrogen pressure of 1 MPa (a) and reuse test of the catalyst (b).

A CN/Ni/Al₂O₃ catalyst was synthesized and applied in the hydrogenation ofp-CNB top-CAN. The catalyst showed very high selectivity top-CAN. This was attributed to the relatively low activity of the chemisorbed hydrogen on the catalyst surface. Also,because the nickel surface was surrounded by CN,it lacked a site for the adsorption ofp-CNB,which contributed to its high selectivity. This nanocomposite composed of Ni modified CN showed promise for application in selective hydrogenation reactions and other related processes under mild conditions.