A general and robust Ni-based nanocatalyst for selective hydrogenation reactions at low temperature and pressure

Abstract

Catalytic hydrogenations are important and widely applied processes for the reduction of organic compounds both in academic laboratories and in industry. To perform these reactions in sustainable and practical manner, the development and applicability of non-noble metal–based heterogeneous catalysts is crucial. Here, we report highly active and air-stable nickel nanoparticles supported on mesoporous silica (MCM-41) as a general and selective hydrogenation catalyst. This catalytic system allows for the hydrogenation of carbonyl compounds, nitroarenes, N-heterocycles, and unsaturated carbon─carbon bonds in good to excellent selectivity under very mild conditions (room temperature to 80°C, 2 to 10 bar H₂). Furthermore, the optimal nickel/meso–silicon dioxide catalyst is reusable (4 cycles) without loss of its catalytic activity.

Highly stable and at the same time active nickel nanoparticles allow for all kinds of hydrogenations.

INTRODUCTION

Catalytic reductions with molecular hydrogen constitute a tremendously important toolbox of methodologies for synthetic organic chemists. In industry, they represent essential steps to produce many fine and chemicals and bulk commodities (1). With the increased availability of green hydrogen, these transformations gain importance in sustainable chemistry (2, 3). Unfortunately, many hydrogenation processes in industry rely on precious metal–based catalysts so far (4, 5). Considering practicability and sustainability aspects, the development of alternative Earth-abundant metal–based heterogeneous catalysts for the hydrogenation of organic compounds is highly desirable. The most prominent example of a non-noble metal hydrogenation catalyst, Raney-Ni, is widely used (6–9). Despite its low cost, commercially available Raney-Ni is not stable and difficult to handle. Furthermore, it exhibits a limited substrate scope, and sensitive functional groups are not tolerated. To solve these problems, in recent years, many non-noble metal–based heterogeneous catalysts based on Ni (10–19), Co (20–22), Fe (23–27), and Cu (28–30) have been developed for the hydrogenation of nitroarenes, carbonyl compounds, nitriles, N-heterocycles, and unsaturated carbon─carbon bonds. However, most of these materials require higher temperature or pressure of hydrogen, which obviously limits practical applications. Hence, there is a demand for the development of general, reusable, and stable as well as highly active and selective non-noble metal–based heterogeneous catalysts, which should ideally work at mild conditions (low temperature and pressure). In this context, we report nickel nanoparticles (NPs) supported on mesoporous silica for selective hydrogenation of ketones/aldehydes to alcohols, nitro compounds to amines, and alkynes/alkenes to alkanes at room temperature and low pressure. In addition, quinolines and related heterocyclic compounds can be hydrogenated too at slightly higher temperature.

RESULTS

Preparation and evaluation of catalysts

In the past decade, many research groups including ours prepared a variety of supported 3d metal-based NPs by combining simple metal salts with organic ligands in the presence of a stable support and subsequent pyrolysis under inert conditions (11, 19, 23, 31–47). Some of the resulting heterogeneous catalysts showed excellent performance in selective hydrogenation (11, 19, 23, 41, 47), oxidation (38, 43, 44), borrowing hydrogen (39, 40), and reductive amination reactions (34, 42) comparable to noble metal catalysts. On the basis of the organic ligands, e.g., amines, amides, and carboxylic acids, applied in the catalyst preparation, the active materials often feature unique structures with carbon-encapsulated metallic or metal oxide NPs (23, 31, 38, 44). On the basis of these works, we questioned the possibility to create these active metal centers without using additional organic ligands. Following this idea, we prepared a series of Ni-NPs supported on various basic, neutral, and acidic supports, for example, commercially available hydroxyapatite (HAP), Vulcan XC72R carbon powder, aerosil silica (SiO₂), γ-Al₂O₃, Nb₂O₅, and synthesized mesoporous silica (meso-SiO₂).

More specifically, nickel(II) nitrate hexahydrate [Ni(NO₃)₂∙6H₂O] was immobilized on the support and calcined in air at 550°C for 5 hours, followed by reduction with 5% H₂/Ar at 550°C for 2 hours. For comparison, Fe- and Mn-based NPs on mesoporous silica were prepared in a similar procedure. The preparation of supported Ni-NPs on meso-SiO₂ is presented in Fig. 1.

Fig. 1. Synthesis of meso-SiO₂ and its supported Ni-NPs.

To evaluate the hydrogenation performance of these materials, the reduction of acetophenone (1a) at room temperature (27°C) in presence of molecular hydrogen (10 bar) in ⁱPrOH-H₂O solvent mixture was chosen as the benchmark reaction (Fig. 2). Among all the prepared catalysts, meso-SiO₂ supported Ni-NPs (Ni/meso-SiO₂) were found to be the best and produced the desired product 1-phenylethanol (2a) in quantitative yield. Related supported Fe- and Mn-NPs were completely inactive. Similarly, other Ni-NPs supported on HAP and carbon were also not active under these conditions. In contrast, the Ni-NPs immobilized on γ-Al₂O₃ and Nb₂O₅ showed some activities. In addition, meso-SiO₂ and fume SiO₂–supported Ni-NPs exhibited good activity, and product 2a was obtained in up to 80%.

Fig. 2. Hydrogenation of acetophenone: Evaluation of different catalysts.

Reaction conditions: 0.5 mmol of acetophenone, 50 mg of catalyst, 10 bar of H₂, and 2 ml of ⁱPrOH/H₂O (1/1) at room temperature (R.T.; 27°C) for 24 hours. Yields were determined by gas chromatography equipped with flame ionization detection (GC-FID) using n-hexadecane as standard.

Next, the effect of critical reaction parameters was evaluated. Different solvents such as 1,4-dioxane, tetrahydrofuran, and alcoholic solvents (table S1, entries 1 to 6) showed no or less activity for the hydrogenation of acetophenone (1a), while the reaction in water (table S1, entry 7) produced 96% of desired hydrogenated product 2a. Considering the solubility of other substrates, a mixture of ⁱPrOH/H₂O was chosen for further experiments, which led to the desired product in 99% selectivity (table S1, entry 8). Increasing the temperature from room temperature to 40° to 60°C led to the formation of ethylbenzene (2ab) (table S2, entries 2 and 3). This demonstrates the high activity of the catalyst for other hydrogenation reactions (see below) as well. In general, decreasing in reaction time, hydrogen pressure, and catalyst amount also led to the desired product, albeit in lower yield (table S2, entries 4 to 8). Therefore, the hydrogenation of other aromatic ketones was performed at room temperature with 50 mg of catalyst and 10 bar of hydrogen in ⁱPrOH/H₂O to achieve optimal yields.

Characterization of Ni-based catalytic materials

Detailed characterization of several Ni catalysts was performed using x-ray powder diffraction (XRD), transmission electron microscope (TEM), high-resolution transmission electron microscopy (HRTEM), energy-dispersive x-ray spectroscopy (EDS), x-ray photoelectron spectroscopy (XPS) and Brunauer-Emmett-Teller (BET) analysis. According to the small-angle XRD patterns (Fig. 3A) of meso-SiO₂, the peaks at 2.3°, 3.9°, and 4.5° are assigned to the (100), (110), and (200) reflections of MCM-41 (48, 49), indicating that the hexagonal mesoporous structure was well developed in the support. After immobilizing Ni-NPs on the meso-SiO₂ support, the material retained a highly ordered hexagonal structure. For the wide-angle x-ray diffraction (WAXRD) (Fig. 3B), the peaks at 44.4°,51.8°, and 76.3° were attributed to metallic nickel too (50). As expected, at higher reduction temperature, the peaks intensity assigned to metallic Ni increased. When the reduction temperature was less than 450°C, broad peaks at around 37.4°, 43.4°, and 63.1° were observed, which are assigned to NiO (50).

Fig. 3. XRD patterns.

(A) Small-angle powder XRD patterns. (B) Wide-angle powder XRD patterns. References NiO (The International Center for Diffraction Data pdf 00-047-1049) and Ni (00-04-0850). a.u., arbitrary units.

TEM analysis of the Ni/meso-SiO₂-550 NPs at different magnifications is presented in Fig. 4. The synthesized meso-SiO₂ has extremely regular channels and Ni-NPs are highly dispersed with ultrasmall particles ranging from 2 to 4 nm. The average diameter of these NPs is 2.7 nm. In addition, the Ni-NPs are distributed evenly on meso-SiO₂, which is also determined by high-angle annular dark-field (HAADF) and elemental mapping (EDS) (Fig. 5) analysis. We assume that the highly dispersed nature of Ni-NPs inside the channels inhibits their agglomeration toward larger particles, resulting in the excellent hydrogenation activity of Ni/meso-SiO₂ (51, 52).

Fig. 4. HRTEM images of Ni/meso-SiO₂-550 nanocatalyst.

(A to C) HRTEM images, with (B) highlighting a Ni-NP and (C) a mean diameter of 2.7 nm. Lattice spacing of 0.204 nm corresponds to Ni (111) (66).

Fig. 5. HAADF and EDS images of Ni/meso-SiO₂-500.

(A to E) HAADF images. (F to I) EDS images of Ni (F), Si (G), O (H), and overlay of Ni with Si (I).

The surface chemical composition of the optimal Ni nanocatalyst was analyzed by XPS. As shown in Fig. 6, Ni 2p_3/2 and 2p_1/2 peaks are observed at 856.2 and 874.0 eV along with multiple components at 858.6 and 876.3 eV. In addition, satellite peaks are pronounced at 862.3, 880.1, 866.9, and 882.9 eV. These signals are assigned to Ni²⁺ (53, 54), which can be explained by well-known fast surface oxidation of small Ni(0) particles exposed to air before the XPS analysis. Notably, no metallic Ni peak (around at 852.6 eV) (53) is observed on the surface of these materials. Considering the XRD and TEM results, it indicates that “protected” Ni(0) NPs inside the hexagonal channels of the support constitute the active catalyst species.

Fig. 6. XPS spectra of Ni/meso-SiO₂-550 nanocatalyst.

The porous texture of several different Ni materials was also investigated using N₂-adsorption experiments, and results are summarized in Table 1. With the relative pressure increasing N₂ adsorption increased to 0.3 to 0.4, an obvious adsorption shape jump was observed. After that, it increased slowly (fig. S1). It exhibits typical type IV adsorption isotherms, which demonstrate the mesoporous structure of these nanomaterials. In general, the meso-SiO₂–based catalysts have high surface area. The introduction of Ni did not damage the porous skeleton of the support but decreased the BET surface area and increased the pore size diameter. In addition, with increasing Ni loading, the surface area decreased, but the pore size diameter increased. The Ni specific surface area were determined by H₂-adsorption tests (table S3). The optimal catalyst (Ni/meso-SiO₂-550) exhibited the highest metallic surface area in 84.3 m²/g_metal, while other catalysts showed lower Ni-specific surface area ranging from 0.3 to 4.5 m²/g_metal. It confirmed that Ni-NPs were highly dispersed in the optimal catalyst. In agreement with the other characterization results, we conclude that metallic Ni species are distributed in the hexagonal channels of mesoporous silica with high surface area.

Table 1. Textural properties of different Ni materials.

Entry	Catalyst	SBET (m2/g)	Average pore diameter (nm)
1	Ni/C	199	9.2
2	Ni/com-fume-SiO2	40	15.6
3	Ni/HAP	53	2.3
4	Ni/γ-Al2O3	73	2.1
5	Ni/Nb2O5	48	12.2
6	Meso-SiO2	948	3.1
7	Ni/meso-SiO2-spent	797	3.3
8	Ni/meso-SiO2–550	758	4.3
9	3% Ni/meso-SiO2	795	3.6

Ni/meso-SiO₂-550 catalyzed hydrogenation of ketones and aldehydes

Apart from acetophenone (1a), diverse aromatic and aliphatic ketones can be successfully hydrogenated in the presence of the optimal Ni/meso-SiO₂-550 catalyst at low temperature (Fig. 7). For example, a series of acetophenones with electron-donating and -withdrawing substituents (1b to 1g) were hydrogenated to their corresponding secondary alcohols in excellent yields up to 95%. Other arylalkyl ketones (1h and 1i) reacted smoothly to afford the desired alcohols too. 2-Acetonaphthone (1j) and benzophenone (1k) were also converted to the corresponding alcohols in excellent yields. Cyclic ketones (1l and 1m) gave 88 and 92% yields of the respective alcohols. Similarly, α-ester-(1n) and α-trifluoro-acetophenone (1o) were hydrogenated well. At this point, it should be noted that all hydrogenations reported so far proceeded without any additives at low temperature. However, in case of aliphatic ketones, improved results were obtained in the presence of base at 60°C. As shown in table S4, there was no conversion of 2-octanone in absence of KOH. However, using simple and inexpensive KOH for the Ni/meso-SiO₂ catalyzed hydrogenation of 2-octanone achieved 84% yield of desired alcohol. It indicates that the important role of KOH in the hydrogenation of aliphatic ketones. The presence of KOH can facilitate the formation of an enolate ion, which easily adsorb on the catalyst surface and promote hydrogenation to alcohols (55). Next, other aliphatic ketones were also hydrogenated under these conditions. Both linear (1p to 1s, 1t, and 1v) and cyclic ketones (1u) were efficiently converted to the corresponding alcohols. Notably, a natural product (56), α-ionone (1v), was selectively reduced too. In this case, the cyclic C═C bond was untouched, while the linear keto and allylic C═C bond were hydrogenated. In addition to ketones, several aldehydes including biomass-based ones (57, 58) (1w to 1y and 1aa) were also hydrogenated smoothly in the presence of this Ni catalyst.

Fig. 7. Ni/meso-SiO₂-550 catalyzed hydrogenation of ketones and aldehydes.

(A) Aromatic ketones and (B) aliphatic ketones. Reaction conditions: 0.5 mmol of substrates, 50 mg of catalyst (8.1 mol % Ni), and 2 ml of ⁱPrOH/H₂O (1/1) for 24 hours at room temperature; ^b at 40°C, 10 bar of H₂; ^c at room temperature, 30 bar H₂; ^d with 5% KOH at 60°C, 10 bar of H₂; ^e with 5% KOH at 80°C, 10 bar of H₂; and ^f with 5% KOH, at 40°C, 10 bar of H₂; isolated yields. ^g Five-gram–scale reaction: 41.75 mmol (5.01 g) of substrate, 4.20 g of catalyst,100 ml of ⁱPrOH/H₂O (1/1), at room temperature, 10 bar of H₂, for 24 hours, isolated yield.

Ni/meso-SiO₂-550 catalyzed hydrogenation of nitroarenes

Next, we performed the hydrogenation of aromatic and aliphatic nitro compounds at room temperature in the presence of Ni/meso-SiO₂ and 2 bar of molecular hydrogen. As shown in Fig. 8, a series of nitroarenes bearing both electron-donating and -withdrawing groups were converted to the corresponding anilines in high yields. Substituents and functional groups such as fluoro, chloro, bromo, iodo, ester, and amide (3d to 3i) were all well tolerated. Nitro-substituted heterocycles (3j and 3k) were also hydrogenated to give the desired products in high yields too. Furthermore, biologically active nitro compounds (3o and 3p) were tested, and, in these cases, the nitro group was also selectively reduced. Aliphatic nitro compounds (3r and 3s) were also successfully hydrogenated and produced corresponding primary amines in up to 84% yields under similar reaction conditions. It should be noted that these reactions are more challenging compared to nitroarenes.

Fig. 8. Ni/meso-SiO₂-550 catalyzed hydrogenation of nitroarenes.

Reaction conditions: 0.5 mmol of substrates, 50 mg of catalyst (8.1 mol % Ni), 2 ml of ⁱPrOH/H₂O (1/1), and 2 bar of H₂ at room temperature for 24 hours, isolated yields. ^a Two-gram–scale reaction: 12.0 mmol (1.99 g) of substrate, 1.20 g of catalyst, and 100 ml of ⁱPrOH/H₂O (1/1) at room temperature, 2 bar of H₂, for 24 hours, isolated yield. ^b Yields were determined by GC-FID using n-hexadecane as standard.

Ni/meso-SiO₂-550 catalyzed hydrogenation of N-heterocycles

Next, we turned our attention to the hydrogenation of N-heterocycles, which allows us to prepare a variety of interesting scaffolds for bioactive compounds in a green and straightforward manner (59). In addition, these hydrogenations are considered as a key step in the development of potential candidates for both on-board and off-board hydrogen storage systems (60). As shown in Fig. 9, the hydrogenation of quinolines and other N-heterocycles proceeded smoothly in presence of this Ni/meso-SiO₂ catalyst. Notably, in case of 6-quinolinecarboxaldehyde (5j), not only the N-heterocyclic ring was reduced but also deoxygenation of the aromatic aldehyde group occurred and gave 6-methyl-1,2,3,4-tetrahydroquinoline in 92% yield. The reduction of other N-heterocycles such as quinoxaline (5k) and acridine (5l) was also performed, and the reduced N-heterocycles were selectively achieved in up to 97% yield.

Fig. 9. Ni/meso-SiO₂-500 catalyzed hydrogenation of N-heterocycles.

Reaction conditions: 0.5 mmol of substrates, 50 mg of catalyst (8.1 mol % Ni), 2 ml of ⁱPrOH/H₂O (1/1), 10 bar H₂, at 60°C for 24 hours, isolated yields. ^a Two-gram–scale reaction. 13.8 mmol (1.97 g) of substrate, 1.38 g of catalyst, 100 ml of ⁱPrOH/H₂O (1/1), at 60°C, 10 bar of H₂, for 24 hours, isolated yield.

Ni/meso-SiO₂-500 catalyzed hydrogenation of alkenes and alkynes

The hydrogenation of carbon─carbon double or triple bonds represents essential processes in the (petro)chemical industry, giving access to valuable intermediates for polymers (61, 62) and fine and bulk chemicals (63–65). Notably, the hydrogenation of nonactivated olefins under mild conditions is a challenging task for most heterogeneous non-noble metal-based catalysts. Nevertheless, the presented Ni/meso-SiO₂ material here can be generally applied for the reduction of several aromatic and aliphatic alkenes and alkynes at room temperature and 2 bar of H₂ (7a to 7m). As displayed in Fig. 10, chemoselective hydrogenation of terminal and internal C─C double bonds is also achieved in presence of amide, nitrile, ketone, ester, and aldehyde (7i to 7m) groups. Gratifyingly, under the mild standard conditions vide supra, all these groups were tolerated. It is worthwhile mentioning that the general applicability of our catalyst is showcased further by the hydrogenation of neutral (e.g., 7g), electron-rich (e.g., 7i), and electron-poor (e.g., 7j) olefins.

Fig. 10. Ni/meso-SiO₂-550 catalyzed hydrogenation of alkenes and alkynes.

Reaction conditions: 0.5 mmol of substrates, 50 mg of catalyst (8.1 mol % Ni), 2 ml of ⁱPrOH/H₂O (1/1), 2 bar of H₂, at room temperature, for 24 hours, isolated yields. ^aYields were determined by GC-FID using n-hexadecane as standard.

Preforming all these catalytic reactions in ⁱPrOH or mixture of ⁱPrOH-water in absence of hydrogen revealed that there were no reactions (fig. S2). These results indicated that this Ni-catalyzed reduction reactions occurred in the presence of molecular hydrogen.

To further prove the utility and practicability of the optimal Ni catalyst, scale-up reactions were conducted. For this purpose, three different substrates (Figs. 7, 1a; 8, 3i; and 9, 5c) were hydrogenated on multigram scale. In all these cases, similar conversions and yields were obtained to that of the benchmark milligram-scale reactions.

Last, recycling and reusability of Ni/meso-SiO₂-550 was carried out for the model reaction at half and full conversions (fig. S3). From the third run onwards, a decrease in yield was observed. However, the activity and selectivity of the recycled catalyst can be regenerated by pyrolyzing the spent catalyst at 550°C for 2 hours under 5% H₂/Ar. The XRD patterns of the recycled catalyst are shown in fig. S5. WAXRD (fig. S5B), which illustrates that small amounts of Ni particles were oxidized to NiO in the recycled catalyst, resulting in the lower yield of 2a in the third run. However, HRTEM (fig. S6), HAADF (fig. S7), and EDS (fig. S8) images showed that most of the highly dispersed Ni particles did not change after the reaction. The spectral surface features in the Ni 2p region (fig. S9) of the recycled catalyst stayed intact, and only Ni²⁺ could be observed. The porous structure (fig. S10) was not damaged after recycling experiments. Nevertheless, after the reaction, the BET surface area (Table 1) of the catalyst material slightly increased from 758 to 797 m²/g. The inductively coupled plasma optical emission spectroscopy analysis (fig. S11) of fresh (4.87 wt %) and recycled (4.42 wt %) catalyst revealed that there is slightly metal leaching occurred.

DISCUSSION

Here, we presented the synthesis of Ni-NPs supported on mesoporous silica with high surface area. The optimal material Ni/meso-SiO₂ allows for general hydrogenation of important organic functional groups including carbonyl compounds, nitroarenes, and N-heterocycles, as well as alkenes and alkynes, under very mild conditions. Compared to currently applied heterogeneous nickel catalysts (Raney-Ni), this catalyst is bench-stable and robust and does not need any specific precautions. It is conveniently prepared by simple impregnation of nickel nitrate on meso-SiO₂ and subsequent calcination and reduction.

MATERIALS AND METHODS

Synthesis of mesoporous silica

The mesoporous silica was synthesized using cetyltrimethylammonium bromide (CTAB) as the template and tetraethyl orthosilicate (TEOS) as silica source. In a typical procedure, 4.85 g of CTAB was added to 250 ml of deionized water in round-bottom flask, and the mixture was stirred at 35°C for 30 min. Then, 20.5 ml of NH₃∙H₂O (25% solution) was added and continued the stirring at room temperature for another 15 min. After the addition of 20 g of TEOS, the mixture was again stirred for 2 hours. The obtained mixture was subsequently aged at 90°C for 3 days. The precipitated white solid was washed with deionized water and dried at 100°C overnight. Last, the resulted product was calcined at 550°C for 5 hours with ramping 2°C/min under air.

Preparation of Ni/meso-SiO₂

The Ni/meso-SiO₂ was synthesized using impregnation method. In a 100-ml round-bottom flask, 0.4937 g of nickel(II) nitrate hexahydrate was stirred in 40 ml of methanol at room temperature for 15 min. Then, 2 g of prepared meso-SiO₂ was added, and the round-bottom flask containing mixture was placed into an aluminum block preheated at 65°C and stirred until the solvent was evaporated. The obtained solid was calcined at 550°C for 5 hours with ramping 5°C/min under air. Then, the calcined samples were transferred into a tube furnace and reduced under hydrogen (5% H₂/Ar) at 550°C for 2 hours with ramping 5°C/min. Same procedure was applied for the preparation of other materials such as Ni/HAP, Ni/C, Ni/γ-Al₂O₃, Ni/com-SiO₂, Ni/Nb₂O₅, Fe/meso-SiO₂, and Mn/meso-SiO₂.

General procedure for the catalytic hydrogenation reactions

For aromatic ketones/aldehydes, nitroarenes, N-heterocycles, alkenes/alkynes hydrogenation reactions, 0.5 mmol of substrate, a magnetic stirring bar, and 2 ml of ⁱPrOH/H₂O (1/1) were transferred to an 8-ml glass vial. Then, 50 mg of catalyst was added, and the glass vial was fitted with cap, septum, and needle. The glass vials containing reaction mixtures (seven vials with different substrates at one time) were placed into a 300-ml autoclave. Next, the autoclave was flushed with 10 bar of H₂ twice and charged with required H₂ pressure. The pressurized autoclave was placed in a preheated aluminum block at required temperature. After completion of reactions, the autoclave was cooled, the remaining H₂ was slowly discharged, and reaction vials were taken out. Then, the reaction mixture containing products was added to brine (3 ml), and products were extraction with ethyl acetate (3 × 2 ml). Then organic layer containing products was filtered and dried with anhydrous Na₂SO₄. The products were purified by column chromatography (silica, n-pentane/ethyl acetate mixture). The resulting products were analyzed by nuclear magnetic resonance and gas chromatography–mass spectrometry. For the hydrogenation of aliphatic ketones, same procedure was applied with the addition of 5% KOH.

Supplementary Materials

This PDF file includes:

Sections S1 to S10

Figs. S1 to S11

Tables S1 to S5

References