Hui Liao,
Peiqi Weng,
Haigen Huang*,
Ruoxia Tan,
Rui Zhu,
Yao Liu and
Zhijun Wang*
Key Laboratory of Coordination Chemistry of Jiangxi Province, School of Chemistry and Chemical Engineering, Jinggangshan University, Ji'an, Jiangxi 343009, China. E-mail: haigenhuang@126.com; wangzhijun@jgsu.edu.cn
First published on 12th July 2023
The hydrogenation of nitroarenes to aromatic amines with H2 under solvent-free conditions in place of organic solvents is a crucial process for the synthesis of fine chemicals. However, the catalysts that have been identified so far are relatively inactive with primary nitroarenes under solvent-free conditions. Herein, ordered mesoporous carbon (OMC)-supported highly dispersed Ru nanoparticles (Ru NPs) were easily prepared by using a coordination-assisted solvent evaporation induced co-assembly method, where 8-hydroxyquinoline exerts a significant influence on the mesostructure ordering and specific surface area of the Ru/OMC composites. The ultrasmall-sized Ru NPs (∼2.0 nm) supported on ordered mesoporous carbon were then applied in the hydrogenation of nitroarenes, exhibiting high activity and selectivity for numerous structurally diverse nitroarenes under solvent-free conditions. Compared with Ru/OMC-imp and Ru/AC-imp, Ru/OMC-800 exhibited a much higher activity and selectivity for hydrogenation of nitroarenes, suggesting its overwhelmingly better performance than OMC or AC supporting noble NPs by simple impregnation method. The strong metal–support interaction is capable of stabilizing the Ru NPs and achieving good recyclability as well as high selectivity.
The metal/OMC composites with a variety of metal nanoparticles that fabricated using traditional impregnation or precipitation techniques have exhibited extraordinary potential applications in hydrogenation catalysis, like Pt/OMC,28 Pd/OMC,29,30 Rh/OMC,31 Au/OMC.32 With respect to the catalytic performance, supported ruthenium catalysts are proved to be efficient for selective hydrogenation with higher selectivity especially for halogenated nitroarenes but lower activity than palladium or platinum catalysts.33–36 Ordered mesoporous carbon supported ruthenium (Ru/OMC) catalysts are highly favorable due to the combined advantages of the active ruthenium species and a unique OMC support with well controlled porosity and structures, as well as superior acid-base tolerance and hydrothermal stability,37–39 which exhibited unusual catalytic properties compared to ruthenium catalysts on the classical supports (Ru/Al2O3,40 Ru/AC,41–43 Ru/CNFs44 RuP/SiO2 (ref. 45) and so on) for hydrogenation reactions. To date, fabricating highly stable Ru/OMC catalysts with ruthenium species homogeneously dispersed on ordered mesoporous carbons with extremely high stability using a facile method remains a challenge, as the Ru/OMC catalysts usually exhibit poor stability due to their property toward aggregation and severe leaching of the ruthenium species, which was caused by their poor dispersion and weak interfacial contact with the OMC.46 The reported Ru/OMC catalysts were usually synthesized by traditional impregnation or precipitation on OMC prepared in advance, followed by reduction with hydrogen or other reducing agents. As expected, poor control over the ruthenium nanoparticles dispersion, size distribution and weak attachment onto the surface of OMC were inevitable in this tedious multi-step synthesis method.47 From the latest research on the carbon supported metal catalysts, we know that direct pyrolysis of metal organic frameworks (MOFs) or metal coordination polymers under an inert atmosphere can afford an alternative approach that deposit metal nanoparticles firmly attached to a carbonaceous framework.48–50 Consequently, it appears particularly significant to directly assemble metal particles onto the OMC frameworks with a plenitudinous interfacial contact and strong interaction, which are formed simultaneously by in situ pyrolysis of metal and carbon precursors in one strategy.
Herein, we first develop a facile one-pot coordination compounds assisted co-assembly strategy to synthesize well-defined ordered mesoporous Ru/carbon (Ru/OMC) composites with highly dispersed ultrafine Ru nanoclusters homogeneously-embedded in the carbon matrices via the soft-templating method with the assistance of 8-hydroxyquinoline (8-HQ). This facile route renders homogeneously dispersed Ru NPs inside the ordered mesoporous carbons without using any stabilizer and additional reductant and was confirmed by various characterization techniques. Moreover, the obtained Ru/OMC materials showed high efficiency and reusability in solvent-free selective hydrogenation of nitrobenzene to aniline with H2 under mild conditions with extremely high stability.
Textural parameters and mesoscale ordering of the materials were further analyzed using nitrogen adsorption–desorption. Nitrogen sorption isotherms and pore size distribution curves are shown in Fig. 3. It can be seen that all the samples show typical type IV isotherms with a sharp capillary condensation step at P/P0 = 0.4–0.8 and a well-defined H1-type hysteresis loop, indicating the mesoporous structures of materials with cylindrical channels. The pore size distributions are very narrow with peaks arranging from 3.5 nm to 4 nm. Some typical textural parameters of the samples are listed in Table 1. The BET specific surface area of the samples is calculated to be 530–640 m2 g−1 with a narrow fluctuation for the xRu/OMC-800 (x = 0, 1, 3, 5, 7, 10). The decrease of the surface area without adding 8-HQ indicates the structural shrinkage caused by high-temperature pyrolysis.
Fig. 2 (a) Low-angle XRD and (b) wide-angle XRD pattern of 3Ru/OMC-700, 3Ru/OMC-750, 3Ru/OMC-850 and 3Ru/OMC-900. |
Fig. 3 (a) N2 sorption isotherms, and (b) BJH pore size distributions of the OMC-800, xRu/OMC-800 and Ru/OMC-800-I. |
Sample | Rua (wt%) | a0b (nm) | SBET (m2 g−1) | Vp (cm3 g−1) | Dp (nm) | Wtc (nm) |
---|---|---|---|---|---|---|
a Ru contents were determined by ICP.b Calculated from SXRD patterns (a0 = 2d100/√3).c Wall thickness values calculated from Wt = a0 – Dp. | ||||||
OMC-800-3 | — | 12.3 | 531.1 | 0.36 | 3.5 | 8.8 |
3Ru/OMC-800-I | 2.11 | — | 559 | 0.41 | 3.8 | — |
1Ru/OMC-800 | 1.06 | 15.8 | 572.3 | 0.49 | 3.8 | 12 |
3Ru/OMC-800 | 2.68 | 12.6 | 613 | 0.40 | 3.5 | 7.5 |
5Ru/OMC-800 | 4.41 | 12.9 | 639.3 | 0.47 | 3.5 | 9.4 |
7Ru/OMC-800 | 6.43 | 12.3 | 627.7 | 0.51 | 3.8 | 8.5 |
10Ru/OMC-800 | 9.72 | 12.3 | 586.1 | 0.44 | 3.8 | 8.5 |
3Ru/OMC-700 | 2.39 | 11.3 | 587.2 | 0.37 | 3.5 | 7.8 |
3Ru/OMC-750 | 2.52 | 12.3 | 642.1 | 0.43 | 3.5 | 8.8 |
3Ru/OMC-850 | 2.44 | 12.3 | 635.2 | 0.44 | 3.5 | 8.8 |
3Ru/OMC-900 | 2.40 | 12.3 | 495.2 | 0.31 | 3.2 | 9.1 |
TEM images of the Ru/OMC (Fig. 5a–h) show stripe-like and hexagonally arranged patterns in large domains, further confirming an ordered 2-D hexagonal mesostructured, except for the sample prepared without adding 8-HQ, which show disordered structure. As shown in TEM images, highly ordered hexagonal arrangements of pores along the [001] direction is observed clearly. The uniform, highly ordered stripe-like and hexagonally arranged images in large domains indicate a highly ordered mesostructured, which was consistent with the result of low-angle XRD patterns and made the ordered mesoporous carbon structure of samples conformed. In addition, the superfine uniform metal nanoparticles and small pore diameter were apparent in the TEM images. Interestingly, the average particle size of the Ru NPs kept a slight change at approximately 1.8–2.5 nm when the ruthenium concentration varies; while the dispersion of uniformly distributed Ru NPs improves as the molar percentage of Ru decreases. This could be attributed to the carbon in situ generated by carbonization of resol acts as a building block to confine and enclose the growth of Ru nanoparticles during carbonization process. Owing to the strong coordination or dispersion effect of 8-HQ species during the coassembly process and effectively restrain the crystal growth through effectively coordinating Ru preventing the aggregation during pyrolysis even under high temperature. So in this way, well-dispersed and homogeneous metallic Ru nanoparticles were incorporated into mesoporous carbon matrices through a simple one-pot synthesis procedure with very accessible hydrothermal conditions.
As wide-angle XRD (Fig. 1b) was not informative to determine the electronic state of the Ru element due to the very small size of the NPs, which only exhibited two remarkable broad diffraction peaks at 2θ values of approximately 23.5° and 43.5°, corresponding to the (002) and (101) diffractions of typical amorphous carbon, respectively. Further insight into the catalyst was obtained by XPS (Fig. 7): The corresponding spectra of carbon, ruthenium and oxygen for the surface of the 3Ru/OMC-800 composite as a case were directly and distinctly detected by the XPS elemental survey scans. For the sake of determining the electronic state of Ru in the 3Ru/OMC-800 catalyst, what we used is the high-resolution of the XPS measurement, which is high enough to make a distinction between them and confirm the electronic state of Ru in spite of the sectional germination between the peaks of C1s and Ru 3d. As shown in Fig. 7, the peaks at 284.6, 285.2 and 281.1 eV suggest the presence of C1s, Ru3d3/2 and Ru3d5/2, respectively. The binding energy of the Ru3d5/2 level in the 3Ru/OMC-800 catalyst is about 0.9 eV higher than that of the standard zero-valent state of Ru (280.2 eV), suggesting that the Ru3+ ion has been successfully reduced.52 The shift indicates that the electron transfer may occur between the oxygen and the Ru nanoparticles of the 3Ru/OMC-800 composites, and as expected, the electron density of the Ru atoms is attracted and reduced by the high electronegativity of oxygen, which lead to an electron-deficient state of electron deficiency of the Ru nanoparticles for the 3Ru/OMC-800 composites.53 In addition, the other C1s spectra of 3Ru/OMC-800 composites could be estimated to be C–OH, C (epoxy/alkoxy), and CO groups and their peaks were at 285.8, 287.7 and 289.6 eV, respectively.54
The key idea of our method for synthesis of the xRu/OMC-800 was to exploit the efficient growth of ultrafine Ru nanoparticles embedded in the carbon frame with high dispersion to provide more active sites. Obviously, smaller Ru nanoparticles with higher active surface should be beneficial for the deposition of ultrathin Ru nanoparticles and thereby for enhanced catalytic activity. Thus, we subsequently studied the effect of pyrolysis temperature on the properties of resultant products, as pyrolysis temperature plays a decisive role in particle size. The dispersion of the Ru/OMC catalyst was determined with XRD and electron microscopy techniques. The wide-angle patterns of 3Ru/OMC-T (T = 700, 750, 850, 900) in Fig. 2b also reveal that there are two broad peaks lied in around 23.5° and 43.5°, indication of (002) and (101) diffractions of typical amorphous carbon correspondingly, and because there were also no obvious peaks of Ru, it is deduced that the Ru nanoparticles in the samples of 3Ru/OMC-T were very small. As expected, the surface areas and pore volumes of the 3Ru/OMC-T composites decreased with increasing calcination temperature probably due to the blocking of the mesostructure assembling, although a partial pore blocking cannot be ruled out (Table 1). The morphology of all 3Ru/OMC-T (T = 700, 750, 850, 900) samples was observed by TEM images (Fig. 6a–d) and the last three pictures demonstrated stripe-geometry and hexagonally arranged pore morphology, which was consistent with the result of low-angle XRD patterns (Fig. 2a) and made the ordered mesoporous carbon structure of samples conformed. In addition, the superfine uniform metal nanoparticles (∼2 nm) and small pore diameter were apparent in the TEM images. A more detailed pore structural characterization was revealed by N2 sorption–desorption isotherms in Fig. 4, and all 3Ru/OMC-T samples show typical type-IV carves with a typical hysteresis loop, revealing the appearance of apparent mesoporous structure in synthesis. Structural and textual properties of 3Ru/OMC-T (T = 700, 750, 850, 900) composites were listed in Table 1, which showed the similar surface areas and pore parameters. The decrease of the pore diameters and the unclosed hysteresis loop of 3Ru/OMC-900 resulted from the structural shrinkage caused by high-temperature pyrolysis. These N2 adsorption/desorption data are in good agreement with the small angle XRD and TEM results.
Fig. 4 (a) N2 sorption isotherms, and (b) BJH pore size distributions of the 3Ru/OMC-700, 3Ru/OMC-750, 3Ru/OMC-850 and 3Ru/OMC-900. |
Fig. 5 TEM images of the OMC-800, xRu/OMC-800 and Ru/OMC-800-I: (a) OMC-800, (b) Ru/OMC-800-I, (c) x = 1, (d) x = 3, (e) x = 5, (f) n = 7, (g) n = 10 and (h) HRTEM images of the 3Ru/OMC-800. |
Entry | Catalyst | Conv. (%) | Sel. (%) | TOF |
---|---|---|---|---|
a Reaction conditions: 20 mg catalyst, NB/Ru = 1000/1 (mol mol−1), H2 pressure = 4.0 MPa, 50 °C, 120 min. | ||||
1 | No catalyst | 0 | 0 | 0 |
2 | OMC | 0 | 0 | 0 |
3 | 3Ru/OMC–800-I | 41.1 | 92.0 | 206 |
4 | 1Ru/OMC–800 | 15.4 | 91.7 | 77 |
5 | 3Ru/OMC–800 | 86.6 | 95.7 | 633 |
6 | 5Ru/OMC–800 | 99.7 | 99.0 | 498 |
7 | 7Ru/OMC–800 | 99.6 | 98.6 | 454 |
8 | 10Ru/OMC–800 | 99.3 | 99.1 | 437 |
9 | 3Ru/OMC–700 | 51.3 | 94.7 | 406 |
10 | 3Ru/OMC–750 | 66.8 | 89.3 | 488 |
11 | 3Ru/OMC–850 | 70.4 | 90.5 | 517 |
12 | 3Ru/OMC–900 | 67.8 | 92.4 | 489 |
13 | 3Ru/OMC–800-imp | 44.1 | 73.2 | 314 |
14 | 3Ru/OMC–800-imp | 62.9 | 57.4 | 476 |
15 | 3Ru/AC–800-imp | 23.6 | 52.8 | 202 |
16 | 3Ru/TiO2-imp | 19.7 | 63.1 | 198 |
17 | 3Ru/Al2O3-imp | 34.2 | 48.3 | 248 |
To understand the benefits of highly dispersed Ru/OMC catalyst with Ru nanoparticles strongly attached to the well-defined carbonaceous framework for the hydrogenation of NB, the 3Ru/OMC-imp catalysts with 3.0 wt% loading were also prepared by simply impregnation method and used for the hydrogenation of NB under the same conditions, which was also baked under the temperature of 800 °C. Compared to the sample synthesized by one-pot method, the Ru/OMC-imp catalyst showed much lower activity, only 44.1% (Table 2, entry 13) conversion was obtained under same reaction conditions. Significantly, under the same impregnation method, a certain degree of improvement of activity for the catalyst was found by the introduction of 8-HQ (Table 2, entry 14). It is further explained that 8-HQ plays an important role in promoting catalyst activity. As the common method for maximizing the number of active sites and efficiently utilizing expensive metals is to reduce the catalyst particle domain size as far as possible, while further investigation have in fact shown that noble metal deposited on carbons by simply impregnation method can easily leach during catalytic processes because the interaction between the metal NPs and the carriers surface is weak.58
Besides, supports can play an important role in the catalytic process by providing new active sites and may strongly affect both the physical and chemical properties of metal nanoparticles. The most-used activated carbon (AC) supporting Ru (3Ru/AC-imp), 3Ru/TiO2-imp, and 3Ru/Al2O3-imp sharing the same mass percentage of 3 wt% Ru were prepared with the same approach, using the hydrogenation of NB as a model reaction. The catalytic activities of the Ru nanoparticles supported on different carriers were tested and compared with the Ru/OMC prepared by the one pot 8-HQ assisted method, On the basis of our experimental data showed in Table 1, the 3Ru/AC, 3Ru/TiO2 and 3Ru/Al2O3 show much lower activity and selectivity to the NB hydrogenation with H2 in the absence of solvent (Table 2, entry 15–17). Compared with these supports, the ordered mesoporous carbon with high surface area and large pore volume can favor mass transfer and show better adsorption property. Furthermore, the catalytic activity per surface metal atom and selectivity can depend strongly on the choice of support, and the extent of metal reduction. Besides, the dispersion states and the surrounding environment of central metal or metal oxides can significantly influence the reactivity and selectivity of nanostructured supported catalysts.59 On the other hand, the functional group of the carbon matrix also worth considering as previously research have found that the carbonyl group, the carboxylic group, or anhydride has remarkable effects on nitrobenzene hydrogenation.60,61 Among these groups which took great parts in the reaction, carbonyl group was highly active, which can be detected, while carboxylic group as well as anhydride can inhibit catalytic activity significantly, which were undetectable form the XPS result in Fig. 7. Modern views suggest that the reaction occurs mainly between reactants chemisorbed on the support and hydrogen activated on the metal sites.
The performance for reduction of halogenated nitroarenes is an indicator that cannot be ignored to evaluate the catalyst. As for halogenated nitroarenes selective hydrogenation, the hydrogenolysis of the C–X, in which X represents halogen, always happens, especially for Pd, Pt, Ni -based catalysts, while ruthenium catalysts were considered to be potential catalysts for restraining dehalogenation, while maintaining high efficiency for reduction of the nitro group. In order to understand the performance of the catalyst for hydrogenation of halogenated nitro compounds, the catalytic properties of the 3Ru/OMC-800 catalyst were probed using the selective hydrogenation of 3-nitrochlorobenzene (m-CNB) as a model reaction. As illustrated in Fig. 8, the m-CNB conversion increased almost linearly in the former part of the reaction, then gradually slowed down and finally m-CNB was completely converted to the corresponding m-CAN either in the presence or the absence of a solvent. During the entire reactions, there was only a certain amount of m-CNSB accompanied in products, and disappeared with a totally conversion of m-CNB into m-CAN. Interesting, the possible byproducts reported in chlorine substituted nitroarenes reduction, which include the dehalogenated product–aniline, or the coupling product such as N-ethylchloroaniline, bischlorophenyldiazene and dichloroazoxybenzene were undetectable in the whole process.62 This phenomenon indicated that the nitro group in the aromatic ring was first reduced to the nitroso group, and further was transformed into the corresponding amine in two consecutive steps. During the reaction, undesired hydrodechlorination were avoided, as the –Cl bond remained intact. It is well known that highly dispersed Ru nanoparticles in the channel of OMC would offer more active sites for the hydrogenation, which not only increases the adsorption probability of the reactant molecules for speeding up the reaction but also greatly shortens the residence time of the reaction species and thus avoids over hydrogenation. Moreover, it is generally recognized that the side reaction of hydrodehalogenation of aromatic halides could be inhibited when the metal nanoparticles are situated in electron-deficient states, the extent of electron feedback from the Ru particles to the aromatic ring in m-CAN were weakened.60 Take a note of the XPS result the 3Ru/OMC-800 catalyst, when compared to the standard zero-valent state of Ru (the binding energy of the Ru3d5/2 level: 280.2 eV), a much higher value (281.1 eV) was obtained in the 3Ru/OMC-800 catalyst, suggesting that the Ru particles in the 3Ru/OMC-800 catalyst presented a much more electron-deficient state. Therefore, it is worth pointing out the high selectivity for hydrogenation of m-CNB over the 3Ru/OMC-800 catalyst mainly attributed to the electronic effect caused by the electron-deficient state of the Ru nanoparticles.
Fig. 8 Reaction profiles of 3Ru/OMC-800 as a function of time for the hydrogenation of m-CNB. Reaction conditions: 20 mg catalyst, m-CNB/Ru = 1000/1 (mol mol−1), H2 pressure = 4.0 MPa, 70 °C. |
The reaction protocol was further extended to various substituted nitroaromatics to verify the scope, activity, chemoselectivity and the general applicability of hydrogenation reactions catalysed by 3Ru/OMC-800. It could be seen that the 3Ru/OMC-800 showed high activities and excellent selectivity for the hydrogenation of the studied substrates containing electron-donating groups or electron-withdrawing groups like –CH3, –CF3, –F, –Br, CH3O–, –OH, –NH2, –NO2 and so on, which can be smoothly reduced to the corresponding aromatic amines in the absence of solvent (Table 3, entries 1–19). For example, fluoronitrobenzene or bromonitrobenzene can be smoothly hydrogenated to the desired product quantitatively and the side reaction of dehalogenation reaction was suppressed at complete conversions. Significantly, nitro group in the presence of a broad spectrum of substitutes containing the easily reducible –CHO, –COOH and –CN functional group also remained unaffected during the reaction (Table 3, entries 20–22). For all substrates contain easily reducible groups, the selectivity to functionalized anilines was ≥99% to produce the fully hydrogenated amine. Notably, the catalyst could also promote the heterocyclic nitroaromatic reduction in the yields of >99% without being deactivated by N heterocyclic (Table 3, entries 23–25).
Entry | Substrate | Tb (°C) | Time (h) | Conv. (%) | Sel. (%) |
---|---|---|---|---|---|
a Reaction conditions: 20 mg catalyst, substrates/Ru = 1000/1 (mol mol−1), H2 pressure = 4.0 MPa.b Actual reaction temperatures and melting point of substrates in parentheses. | |||||
1 | 75 (52) | 2 | >99 | >99 | |
2 | 45 (2) | 9 | >99 | >99 | |
3 | 70 (54) | 4.5 | >99 | >99 | |
4 | 130 (112) | 5.5 | >99 | >99 | |
5 | 60 (43) | 6 | >99 | >99 | |
6 | 140 (126) | 15 | >99 | >99 | |
7 | 85 (81) | 3.5 | >99 | >99 | |
8 | 100 (76) | 8.5 | >99 | >99 | |
9 | 70 (57) | 9 | >99 | >99 | |
10 | 100 (69.7) | 3.5 | >99 | >99 | |
11 | 140 (126) | 14 | >99 | >99 | |
12 | 140 (119) | 8 | >99 | >99 | |
13 | 100 (89) | 5 | >99 | >99 | |
14 | 70 (40) | 2.5 | >99 | >99 | |
15 | 50 (21) | 3 | >99 | >99 | |
16 | 60 (9) | 6.5 | >99 | >99 | |
17 | 75 (31) | 13 | >99 | >99 | |
18 | 150 (127) | 4.5 | >99 | >99 | |
19 | 70 (19) | 15.5 | >99 | >99 | |
20 | 120 (116) | 4.5 | >99 | >99 | |
21 | 145 (142) | 6.5 | >99 | >99 | |
22 | 60 (20) | 10 | >99 | >99 | |
23 | 85 (57) | 9 | >99 | 91.5 | |
24 | 95 (92) | 18 | >99 | >99 | |
25 | 90 (73) | 13.5 | >99 | >99 |
Reusability is an important Index to evaluate the performance of heterogeneous catalysts. At last, we selected the reaction conditions with a conversion of ∼86% in the first cycle, at which the NB hydrogenation was controlled by chemical kinetics, to investigate the potential reusability of the 3Ru/OMC-800 catalyst for the hydrogenation of NB in the absence of solvent. The catalyst after each run was recovered by a simple filtration and drying without any other post-treatment. Ten reaction runs were performed and the NB conversion in each run was almost constant at ∼86%, as shown in Fig. 9. After the tenth run, the spent catalyst had no discernible variations in textural properties (SBET: 598.4 m2 g−1; Vp: 0.39 cm3 g−1; Dp: 3.5 nm), phase structure and Ru particle size distribution analyzed by XRD, BET and TEM as shown in Fig. 10. The final reaction solution and the spent catalyst was tested by ICP and it is found that there was no any Ru detected in the product and the metal content of the spent catalyst was almost the same with the fresh catalyst. This result demonstrated that the 3Ru/OMC-800 catalyst was highly stable and reusable for the solvent-free hydrogenation of nitroarenes with H2.
Fig. 10 (a) Low-angle XRD, (b) wide-angle XRD patterns, (c) N2 sorption isotherms and pore size distribution and (d) TEM image of the spent catalyst for ten cycles. |
This journal is © The Royal Society of Chemistry 2023 |