Modulation of Electronic Structure of Single-Atom Catalysts for Sustainable Production of Full-Spectrum Amines from Renewable Biomass

doi:10.21203/rs.3.rs-2458189/v1

Download PDF

Article

Modulation of Electronic Structure of Single-Atom Catalysts for Sustainable Production of Full-Spectrum Amines from Renewable Biomass

https://doi.org/10.21203/rs.3.rs-2458189/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Using renewable biomass to replace fossil resources for producing low-carbon-footprint but high-value chemicals is a sustainable approach to pursue a carbon-neutral society. Herein, a boron nitrogen doped carbon (BNC) confined Pd single atom catalyst (Pd SAs/BNC) is synthesized and used as a robust catalyst toward the reductive amination reaction to produce low-carbon-footprint amines from renewable biomass. Because of the finely tuned electron structure, the as-synthesized single atom catalyst delivers an ultrahigh turn-over frequency value (max. 1368 h^− 1) in the reductive amination of aldehydes/ketones and demonstrates a great conversion capacity for various aldehyde/ketone and amine/nitro-compound substrates. Extensive characterizations and density functional theory calculations show that the highly polar metal-N site formed between the central Pd single atom and its neighboring N and B atoms favors hydrogen activation from the donor (reductants) and hydrogen transformation to the receptor (C = O group), resulting in a great selectivity. This system could be further extended to directly produce various aromatic and furonic amines from renewable lignocellulosic biomass, and their greenhouse gas emission potentials are negative compared to those of fossil-fuel resource-based amines. This work offers a highly efficient and sustainable approach to construct C-N bonds for the production of numerous amines from carbon-neutral biomass resources.

Physical sciences/Chemistry/Catalysis/Heterogeneous catalysis

Physical sciences/Nanoscience and technology/Nanoscale materials/Two-dimensional materials

Amines represent a category of important compounds that act both as bulk and fine chemicals for polymers, dyes, surfactants, agrochemicals, and pharmaceuticals.^1–4 Reductive amination represents one of the most versatile methodologies widely used for amine production in industry. ^5–7 In the reductive amination process, the C = O groups in aldehyde or ketone compounds react with NH₃ or amines in the presence of reducing agents to construct C-N bonds, forming a variety of amines.^8,9 Currently, most industrially relevant aldehyde or ketone compounds are produced from non-renewable fossil resources, making amine production via reductive amination not sustainable. It would be very attractive if aldehyde and ketone compounds could be produced from renewable biomass resources, as it would create a carbon-neutral opportunity to produce low-carbon-footprint amines without relying on fossil fuel resources.^10–14

However, the diversity and complex structures of the biomass-derived aldehydes and ketones require highly efficient catalysts for reductive amination,^15,16 and the selectivity toward the corresponding amines remains a main challenge due to the presence of several side reactions. Moreover, unlike conventional hydrogenation reactions, reductive amination usually involves excess ammonia or amine reactants, which may be strongly adsorbed on the active sites of the catalysts and become possible poisons to the catalysts. Therefore, an efficient catalyst for reductive amination should survive in the presence of excess ammonia or amine without being poisoned.¹⁷ In addition, such a catalyst should possess moderate hydrogenation activity to avoid over-reduction or weak-hydrogenation.^18–20

Single-atom catalysts (SACs), a category of nano-metal catalysts with tunable electronic structures and all active metal species existing as isolated single atoms, may overcome the limitations of conventional nanoparticle-based catalysts in terms of efficiency (activity and stability) and selectivity.^21–23 The unique electronic and structural properties make SACs especially suitable for the conversion of biomass-derived aldehydes/ketones, as they can achieve preferential adsorption of targeted functional groups in the aldehydes/ketones on single metal sites and facilitate desorption of desired products, leading to enhanced catalytic efficiency and selectivity.^24–27

Specifically, an efficient reductive amination catalyst should possess moderate hydrogenation activity and high ammonia or amine tolerance. Finely tuning the electronic structure of single metal sites is able to creat an efficient reductive amination SAC catalyst. Previous studies have demonstrated that modulating the long-range interactions with tunable functionalities on the substrate of SACs is a feasible approach to tune the electronic structure of single metal centers.^28,29 As a non-metallic heteroatom, the electron-deficient B atom has fewer valence electrons (2s²2p¹) than its valence orbitals, which readily shares electrons donated from the electron-rich N atom (2s²2p³). As a consequence, the formed B–N bonds are partially ionic because of the large difference in the electronegativity of two atoms (B: 2.04, N: 3.04), thus forming a local electric field around the single metal sites. ³⁰ Therefore, a functional B–N region with a local electric field may construct an overall electrostatic surface to interact with the C = O dipole, thus capturing the aldehyde/ketone reactants correctly. Meanwhile, the long-term interaction between B and N may increase the electronic density of Pd active sites and downshift the d-band center, thus increasing the polarity of the metal-N sites.³¹ Such a highly polar metal-N site will weaken the interactions between the metal species and ammonia, thus increasing the SAC tolerance to ammonia and improving the reductive amination reaction efficiency of the SAC.

Therefore, inspired by the abovementioned ideas, we develop a versatile approach to systematically modulate the electronic structure of Pd single atoms by integrating the single-atom sites with a highly polarized boron nitrogen doped carbon (BNC) support. The projected density of states (PDOS) and synchrotron X-ray adsorption spectroscopy analysis were performed for validate the modulation of the electronic configuration of the Pd single atoms via B and N functionalities. Due to the finely tuned electron structure, the as-synthesized Pd single atoms/BNC catalyst exhibited a wide adaptability and great efficiency for the conversion of various aldehyde/ketone and amine/nitro-compound substrates.

Synthesis and structural characteristics of the Pd₁/BNC SACs

The single-atom palladium catalyst (Pd₁/BNC) was prepared by adopting a supermolecular controlled pyrolysis strategy (Fig. 1s). Supramolecular frameworks were fabricated by the assembly of urea and boric acid, which were driven by H-bonding interactions. Then, the atomically dispersed Pd catalyst was obtained by pyrolyzing a mixture of the palladium (II) complex of bipyridine and urea-boric acid supramolecular compounds. In this approach, the nitrogen content in Pd₁/BNC increased substantially by forming h-BN nanostructures, which played a key role in anchoring and stabilizing Pd single sites. It should be noted that no acid leaching treatment was required in the preparation procedure, making the synthesis process more sustainable and environmentally friendly.

The morphology of the catalyst was characterized by transmission electron microscopy (TEM) (Fig. 1a and Supplementary Figs. 4 and 5). As revealed by Fig. 1a, Pd₁/BNC possessed a thin graphene-like layered structure without the formation of obvious particles. The X-ray diffraction (XRD) pattern of Pd₁/BNC displays a broad peak at approximately 25° (2θ) related to the (002) plane of graphitized carbon, and no obvious signals for metallic Pd species were observed (Supplementary Fig. 1). Moreover, the Raman spectrum shows two bands characteristic of carbon materials with I_D/I_G = 1.1 (Supplementary Fig. 2), indicating the poor crystalline nature of the Pd₁/BNC sample, which is in good agreement with the XRD results. The pore features of Pd₁/BNC were analyzed with nitrogen sorption isotherms (Supplementary Fig. 3), showing a surface area of 127 m²/g.

To characterize the dispersion of Pd in Pd₁/BNC, aberration-corrected high-angle annular dark-field scanning transmission electron microscopy (AC HAADF-STEM) was adopted. As shown in Fig. 1b and c, high-density bright dots (highlighted by white dashed circles) could be observed, corresponding to the isolated Pd single atoms. Furthermore, 3D isolines and atom-overlapping Gaussian-function fitting mapping (Fig. 1d) of the square from Fig. 1c verify the atomic dispersion of Pd sites. Pd single atoms were well separated, as depicted in the intensity profile along X-Y (Fig. 1d). Additionally, the HAADF-STEM image shows no observable Pd NPs in Pd₁/BNC (Fig. 2e), in line with the ring-like selected-area electron diffraction (SAED) pattern (Fig. 1e, inset). The homogeneous spatial distribution of C, N, B and Pd over the entire architecture was evidenced by EDS mapping analysis of Pd₁/BNC (Fig. 1f). The content of Pd measured via inductively coupled plasma atomic emission spectroscopy (ICP‒AES) analysis was approximately 1.24 wt.%.

Electronic structure and coordination micro-environment of the Pd₁/BNC SACs

X-ray absorption fine structure (XAFS) analysis was carried out to clarify the local fine electron structure of the single atomic Pd sites in Pd₁/BNC. As displayed in Fig. 2a, the Fourier transferred extended X-ray absorption fine structure (FT-EXAFS) spectrum of Pd₁/BNC displays one dominant peak at approximately 1.5 Å, which could be attributed to Pd-N/O bonds.³² Additionally, no Pd-Pd coordination was detected in comparison with the WT contour plots of Pd foil, validating the isolated feature of the atomically dispersed Pd species in Pd₁/BNC. Wavelet transform (WT) analysis was performed to further investigate the atomic configuration of the Pd₁/BNC catalyst (Fig. 2b), aiming to provide powerful resolution in both the R and k spaces of the Pd atoms and discriminate the back scattering atoms. Similar to the FT-EXAFS spectrum, the WT contour plots of the Pd₁/BNC catalyst present only one maximum intensity at ~ 5 Å⁻¹, which could be attributed to the Pd-N interaction. For comparison, both Pd foil and PdO counterparts show higher intensity maxima at 9.3 and 8.4 Å⁻¹, corresponding to Pd-Pd and Pd-O-Pd coordination, respectively. The structural parameters of single Pd atoms were analyzed via EXAFS fitting. As the results show (Supplementary Fig. 6), the Pd species in the Pd₁/BNC catalyst exist in the form of isolated Pd-N₃ sites.

The chemical state of Pd in the Pd₁/BNC catalyst was further investigated with K-edge X-ray absorption near-edge structure (XANES) analysis. The energy absorption threshold of the Pd₁/BNC catalyst was located between those of Pd foil and PdO, suggesting that the atomically dispersed Pd atoms carry partially positive charges (Fig. 2c). The chemical state of Pd in Pd₁/BNC was further investigated with XPS. As the results show (Fig. 2d), the Pd 3d_5/2 orbital of Pd₁/BNC was at a binding energy of 338.1 eV, slightly lower than that of Pd/NC (338.4 eV), implying a higher metallicity of the Pd species in Pd₁/BNC than Pd₁/NC. The main reason for this phenomenon might be that the long-term interaction between B and N could decrease the electronic density of Pd active sites and downshift the d-band center, thus increasing the polarity of the metal-N sites.³¹

Soft X-ray absorption near-edge structure (XANES) analysis was performed to examine the coordination micro-environment of Pd₁/BNC. As shown in Fig. 2e, the N K-edge spectrum of Pd₁/BNC displays broadened peaks, indicating multiple forms of N species. The peaks at 398.6 and 400 eV mainly originated from pyridinic π* and graphitic π* transitions. The fingerprints comprising π* resonance at 401.7 eV and two σ* features at 408.3 and 415.3 eV suggest the successful formation of h-BN nanostructures in the BNC supports.^31,33 The presence of h-BN domains in the Pd₁/BCN sample could be further confirmed by the B K-edge spectrum (Fig. 2f), in which a spectral fingerprint of B(sp²)–N(sp²) bonds with a sharp B 1s → π* could be found at approximately 191.6 eV and three 1s → σ* resonances could be observed at 196.8, 198.6, and 203.7 eV. The N and B K-edge XANES spectra indicate the presence of B–N polar sites on the surface of the Pd₁/BNC catalyst.³⁴ These B–N species could be separated by an appropriate distance from each other on the catalyst surface, avoiding their collapse by mutual neutralization. They may act as molecular frustrated Lewis acid-base pairs (FLPs) to activate the liquid hydrogen donors for active H species production, thus promoting the hydrogenation of the amination products.^35,36 The N species were also monitored by the N 1s XPS spectrum (Supplementary Fig. 7a), which was fitted into four peaks attributed to the N-B bond (397.8 eV), pyridinic N (398.8 eV), pyrrolic N (400.0 eV), graphitic N (401.4 eV), pyridinic N (398.4 eV), pyrrolic N (399.9 eV), and graphitic N (401.2 eV).³⁷ The above results clearly indicate that the Pd₁/BCN catalyst not only involves the h-BN nanostructures but also multiple forms of active N sites.

Amine synthesis with the catalysis of the Pd₁/BNC SACs

Optimal reaction conditions screening. The reaction between 5-HMF and NH₃ was first investigated to find out the widely applicable conditions for the SAC-mediated reductive amination reactions. Liquid hydrogen donors, rather than high pressure H₂ gas, were selected as the reductants for the reductive-amination of 5-HMF due to their high safety and environmentally friendly feature. Screening of reductants (Entries 1–3, Table 1) indicates that all the liquid hydrogen donors, i.e., formic acid, isopropanol, and NaBH₄, could efficiently convert 5-HMF, but only isopropanol could lead to the formation of the corresponding primary amine (5-hydroxymethyl-2-furan methylamine (HMFMA) with a high yield (Entry 1). Both formic acid and NaBH₄ could directly reduce 5-HMF into furan-2, 5-diyldimethanol. It should be noted that 5-HMF could also be efficiently converted into imine (5-hydroxymethyl-2-furanimide (5-HMFMI) via a direct amination process in the absence of reductants (Entry 4), suggesting that the reduction and amination reactions could occur independently.

Under the optimum reaction conditions, the turnover frequency (TOF) of the Pd1/BNC catalyst for reductive amination of 5-HMF was calculated as 1368 h^− 1, much higher than that with the carbon-supported Pd nanoparticle (5% Pd/C) catalyst (Entry 5), confirming the superior atomic efficiency of the SACs. Such a high TOF value also placed the Pd₁/BNC catalyst among the best reductive amination catalysts reported so far (Supplementary Table 1). A control experiment was conducted to examine the reaction efficiency under catalyst-free conditions. Unlike the reductant-free conditions (Entry 4), both conversion of HMF and yield of primary amine were very low (Entry 8) in the absence of catalyst, highlighting the important role of the catalyst in the reaction.

Primary amine synthesis. After optimization of the reaction conditions, we expanded the substrate scope from 5-HMF to various biomass-derived aldehydes. As shown in Table 2, the biomass-derived furonic aldehydes (e.g., furfural, 5-methylfurfural, derived mainly from cellulose or hemicellulose) could be efficiently converted into the corresponding primary amines with favorable yields (> 90%, Products S1 and S2). Even for 2,5-diformylfurfural with two aldehyde groups, its reductive amination to the corresponding primary amine could smoothly occur with a yield up to 87% in a 12-h reaction (Product S3). Similar to the furanic aldehydes, the reductive amination of the aromatic aldehydes (e.g., benzaldehyde, 4-hydroxybenzaldehyde, vanilline, and cinnamaldehyde, derived mainly from lignin) could also efficiently take place with favorable yields (83%-90%, Products S4-S8) to their corresponding aromatic primary amines. Notably, several sensitive functional groups abundant in biomass, including C-OH, C-O-C, and C = C, were well tolerated in this system. Unlike the aromatic and furanic aldehydes, the reductive amination of aliphatic aldehydes was hampered by an unproductive aldol condensation reaction,³⁸ offering a relatively lower yield toward its corresponding primary amines (75%, Product S9). Unlike aldehydes, the reduction of ketones is usually more demanding and requires more drastic conditions and active catalysts.¹⁸ Herein, we know that several more challenging biomass-derived ketones with various functional groups could also be effectively converted into the corresponding primary amines with high yields (75%-93%, Products S10-S15).

Secondary and tertiary amine synthesis. Inspired by the favorable performance of Pd₁/BNC for the reductive amination of various biomass-derived aldehydes and ketones for primary amine production, we applied this catalyst to the synthesis of secondary and tertiary amines. The reductive amination of 5-HMF with primary aromatic amines was tested first. The reductive-amination with electron-donating group (e.g., -H, -OH, and -CH₃) substituted aromatic amines gave high yields toward their corresponding secondary amines (Products S16-S20, yields 89%-93% within 6 h). The reactivity of the hydroxyl-substituted anilines followed an order of 4-hydroxyl > 3-hydroxyl > 2-hydroxyl (Products S18-S20) because of the different steric hindrance effects.¹⁴ In contrast, the yields for reductive-amination of the electron-drawing groups (e.g., -Cl, -COOH, and -SO₃H) substituted aromatic amines afforded relatively lower yields even after 12-h reaction (Products S21-S23, yields 78%-83%) due to the possible electronic passivation effects. ¹⁸ In addition to aromatic amines, SAC Pd₁/BNC was also efficient in the reductive amination of 5-HMF with aliphatic amines, providing yields up to 92% (Products S24 and S25).

To demonstrate the general feasibility of the single-atom Pd₁/BNC-catalyzed reaction system, the reductive N-alkylation of 5-HMF with secondary amines for tertiary amine production was also investigated. The reductive N-alkylation with three different types of secondary amines (e.g., aliphatic diethanolamine, aromatic diphenylamine, and heterocyclic 2-methylimidazole) all occurred efficiently, offering yields up to 88% toward their corresponding tertiary amines (Products S26-S28). However, a relatively lower yield (Product S28, 79%) was obtained for the reaction of 5-HMF and diphenylamine. Such a low yield might be attributed to the high steric hindrance of the two aromatic rings around the NH group.³⁹

In many practical organic synthesis processes, primary amines with different substituent groups are usually obtained from the reduction of their corresponding nitro compounds.^40,41 Therefore, compared with the traditional reductive amination of amines, one-pot reductive amination of aldehydes with nitro compounds could be more straightforward, ensuring a better step economy.^38,42,43 Similar to those of amines, the reductive-amination with electron-donating group (e.g., -H, -CH₃, -OCH₃, and -OH) substituted aromatic nitro compounds also gave high yields toward their corresponding secondary amines (Products S29-S35, yields 89%-93% within 12 h). The reactivity of the hydroxyl-substituted aromatic nitro compounds also followed the order 4-hydroxyl > 3-hydroxyl > 2-hydroxyl (Products S32-S34) due to the different steric hindrance effects. In contrast, the yields with electron-drawing group (e.g., -CHO, C = O, -Cl, -COOH, and -SO₃H)-substituted aromatic nitro compounds also gave relatively lower yields even after a 24-h reaction (Products S35-S39, yields 68%-86%) owing to the possible electronic passivation effects.

These results show that several reduction-sensitive groups, such as C-Cl, C-OH, O-CH₃, COOH, and SO₃H, were well tolerated in this reaction system, providing the corresponding secondary amines in 78 to 91% yields. An exception is that this reaction system showed a relatively poor tolerance to aldehyde and ketone groups. The reactions with aldehyde- or ketone-substituted aromatic nitro compounds offered yields of 68% and 76% to their corresponding secondary amines, respectively, after a 24-h reaction (Products S35 and S36). Significant amounts of byproducts of the reductive amination of the aromatic aldehyde or ketone groups were detected in the two cases. These results further confirm that SAC Pd₁/BNC was highly efficient for the reductive amination of aldehydes.

Catalyst recycling and reaction upscaling

Reusability is crucial for heterogeneous catalysts, as it not only obviously reduces production costs but also considerably facilitates product purification. ³⁸ Before the reusability evaluation, we first examined the heterogeneous catalytic nature of this reaction via a hot filtration operation (Supplementary Fig. 8a). In the initial 180 min, the reaction occurred smoothly, providing yields of the corresponding amines and imines of 58% and 17%, respectively. When the SAC Pd₁/BNC was filtered, the reaction almost stopped in the subsequent 180 min, and the yields of amines/imines remained almost unchanged (Supplementary Fig. 8b). These results demonstrate that the reductive amination with the Pd₁/BNC catalyst was indeed a heterogeneous catalytic process, in which the irreversibly leached Pd species had almost no influence on the reaction.

Seven successive cycles of 5-HMF reductive amination were conducted to evaluate the durability of the Pd₁/BNC catalyst. As shown in Supplementary Fig. 8c, the HMF conversion slightly decreased from 99–94% after seven cycles. Correspondingly, the conversion rate slightly decreased from 1.375×10^− 2 to 1.306×10^− 2 mmol HMF/min after seven successive cycles. Meanwhile, the HMFMA yield only slightly decreased from 96–89% in the seven cycles with a slight decrease in the conversion rate from 1.333 × 10^− 2 to 1.236 × 10^− 2 mmol HMF/min. The release of Pd to the reaction system in each cycle was analyzed to further evaluate the stability of the Pd₁/BNC catalyst. As shown in Supplementary Fig. 9, after the first cycle, the Pd concentration in the reaction mixture was only 3.31 µg/L and continuously decreased to 1.72 µg/L in the subsequent cycles. In all seven cycles, less than 0.2% of the Pd in the catalyst was released into the reaction mixture. These results confirm the favorable reusability and stability of the Pd₁/BNC catalyst. The HR-TEM image of the Pd₁/BNC catalyst after cycle reuse shows that after seven reuse cycles, the Pd maintained its atomic dispersion state without significant aggregation (Supplementary Fig. 10), further confirming the favorable stability of the catalyst.

An upscaling reaction with a 100-fold higher substrate (50 mmol 5-HMF, 6.30 g) was conducted to evaluate the applicability of the single-atom Pd₁/BNC-catalyzed reductive amination for multigram-scale synthesis. After a 12-h reaction, the reductive amination of 5-HMF with NH₃ could yield 7.4 g of the corresponding amine (light yellow liquid, Supplementary Fig. 11) with a yield of 91% (GC yield) and a purity of 94% (Supplementary Figs. 12 and 13). Such a yield was comparable to that in the 0.5 mmol scale reaction, confirming the application potential of our reaction system.

Reaction pathway and catalytic mechanism

Figure 3a displays the HMF conversion and product yields as a function of the reaction time, showing that a large proportion of imine (19%) was produced at the initial stage (within 120 mi) of the reaction and then continuously decreased to 1% in the following 240 min. Meanwhile, the yield of 5-HMFMA continuously increased as the reaction proceeded, reaching a final value of 96%. These results suggest that the imine should be a key intermediate in the reductive amination of aldehydes/ketones. As the imine is mainly formed from the amination reaction between NH₃ (or other amino source) and aldehydes/ketones, its quick accumulation in the initial reaction stage implies that the amination is not a rate-determining step. Instead, the kinetics profile shows that after the yield of imine reached the maximum value, its subsequent reduction to amine proceeded slowly, implying that imine reduction should be the rate-determining step in the reductive amination reaction.

Several control experiments were conducted to obtain further information on which pathway was the rate-determining step in the reductive amination process. As shown in Fig. 3b, the direct reduction of 5-HMF in the absence of NH₃ hardly occurred, in which less than 10% 5-HMF was converted within 3 h, yielding a trace amount of 2,5-dihydroxymethyl furan. When NH₃ was introduced into the reaction system, the conversion of 5-HMF became significant, yielding the corresponding imines, which were further reduced to amines. By comparison, the direct amination of 5-HMF could occur smoothly in the absence of the reductant isopropanol, yielding imine as a main product, which could be reduced into the corresponding amine when the reductant was introduced into the reaction (Fig. 3c). With these results, we could further infer that reduction rather than amination should be the rate-determining step in the reductive amination reaction.

Apart from the rate-determining steps, exploring how the H-donor (reductant) was activated by metal–N_x sites on the SAC is also essential for mechanism elucidation. ¹⁸ With the previously reported results on Pt-O_x,⁴⁴ Ni-N_x,⁴⁵ and Ru-N_x SACs,¹⁸ we propose that the frustrated Lewis pairs (FLPs) formed between the central Pd and its neighboring N and B atoms might be responsible for activating the H-donor to form active H species in a heterolysis manner.^46,47 To demonstrate the FLP mechanism, the Lewis acid and basic sites were selectively poisoned by potassium thiocyanate (KSCN) and pyrrole, respectively. As shown in Fig. 3d and e, after the Lewis basic or acid sites were poisoned by KSCN or pyrrole, the final amine yield significantly decreased from 96–17% or 4%, respectively, while the imine yield increased sharply from 1–55% or 34%. This result clearly demonstrates that the FLP was responsible for activating the H-donor to form active H species for the reduction of imine intermediates.

To understand the role of Pd single atoms in governing the reaction route, we performed density functional theory (DFT) calculations for the reductive amination of HMF with NH₃ in the presence of Pd₁/BNC. The adsorption and dissociation of isopropanol to generate the surface H-atom (H*) at the Pd site is the first step for HMF reduction. To show the electron redistribution at the interface, electron density difference analysis was conducted. For isopropanol and the Pd₁/BCN complex, accumulated electron density was found in the region between H and Pd, giving rise to the formation of Pd-H bonds (Fig. 4a and Supplementary Fig. 16). Partial density of states (PDOS) analysis also confirms the activity of Pd₁/BCN with the highest occupied molecular orbital (HOMO) of the Pd atom near the Fermi level (Fig. 4b), which is beneficial for H* formation. Pd₁/NC also showed enriched electron density between the H and Pd atoms but was less active in generating H* compared with Pd₁/BCN. To analyze the reductive activity of H*, PDOS analysis of H-1s orbitals was conducted. The H-1s at Pd₁/BNC was lower than the Fermi level, while that of H* at Pd₁/NC was near the Fermi level (Fig. 4c), indicating that the H* at Pd₁/BNC was less active. Moreover, to evaluate the selectivity of HMF reduction at the Pd single atoms, an energy diagram of the reaction routes was calculated, including the formation of H*, the reduction of the –CH = NH group to the –CH₂NH₂ group (target product) and the competing reduction of the –CH = O group to the –CH₂OH group (byproduct, Fig. 4d). When isopropanol was adsorbed at Pd₁/BNC, the surface H* was found to form more readily (-0.25 eV) than that at Pd₁/NC (2.20 eV, Fig. 4e), which is consistent with the electronic analysis. The resulting H* of Pd₁/BNC had a higher activity for the reduction of the –CH = NH group with a transition barrier of 0.94 eV and a limiting reaction energy change of 1.08 eV (Fig. 4f). In contrast, Pd₁/BNC was less active to the competing reduction of the –CH = O group with a larger transition barrier of 1.62 eV and a limiting step energy change of 2.11 eV (Fig. 4e). Therefore, a high selectivity of the reductive amination product CH-NH₂ over the byproduct CH-OH was achieved by Pd₁/BNC.

Scalability And Sustainability Evaluation

Encouraged by the abovementioned excellent catalytic performance, we further conceived a conceptual integrated process in which lignocellulosic biomass was directly used as a feedstock to produce low carbon-footprint amines. In this process (Fig. 5a and b), the lignocellulosic biomass was first separated into lignin, cellulose and hemicellulose using a biomass fractionation method described by Dumesic et al. (Supplementary Text 4).⁴⁸ The as-separated lignin was depolymerized under electrochemical oxidation conditions to produce various aromatic aldehydes and ketones (Supplementary Text 5 and Supplementary Fig. 14), while the cellulose and hemicellulose were first hydrolyzed into glucose and xylose with the catalysis of HCl and then further dehydrated to form 5-HMF and furfural with the catalysis of Lewis acid (AlCl₃) (Supplementary Fig. 15). The formed aldehydes and ketones were then converted into the corresponding amines via reductive amination. The final amine products were separated and purified in the same way as the upscaling reaction. Figure 5c presents the carbon flow for the production of amines from the lignocellulosic biomass. With the proposed process, 1000 kg of fir wood biomass could be finally converted into 538 kg aromatic and furonic amines, as well as 96 kg of aromatic monomers (guaiacols and other phenols) and 207 kg of carbohydrate pulps (monosaccharides and oligosaccharides).

The amine production process from lignocellulosic biomass makes sense only if a lower CO₂ footprint can be achieved. Therefore, based on the experimental data, we performed a simple life-cycle assessment (LCA) to evaluate the global warming potentials (GWPs) of biomass-based amines compared with their fossil-based counterparts (Supplementary Text 7, Supplementary Table 2, and Supplementary Fig. 17). Figure 5c and d displays the GWPs for the production of biomass-based aromatic amines (originating from lignin) and furonic amines (originating from cellulose/hemicellulose), respectively. The GWPs of aromatic and furonic amines were calculated as 0.14 and 0.47 kg CO₂ per kg products when the electricity used in the system was produced from fossil fuels. However, when the electricity used in the system was produced from renewable sources (e.g., hydraulic power, wind, and solar power), the corresponding GWPs all turned to negative values (-1.57 to -2.85 kg CO₂ per kg products for aromatic amines and − 1.76 to -2.93 kg CO₂ per kg products for furonic amines). These negative GWP values indicate net CO₂ consumption, namely, a carbon-negative effect on their production.

Preparation of single-atom Pd₁/BNC catalysts

First, 3.0 g of urea and 100 mg of boric acid were dispersed in 10 mL of deionized water under ultrasonication for 15 min. A 2,2′-bipyridinepalladium dichloride ethanol solution (5 mL) of 2 mM and 200 mg of F127 in ethanol (5 mL) were injected into the above solution, and the resulting mixture was stirred for 30 min. Then, the precursor was obtained by freeze-drying. Subsequently, the obtained yellow precursor was ground in a quartz mortar and carbonized at 800°C for 2 h at a heating rate of 5°C/min under flowing Ar gas. Moreover, the synthetic procedure of BNC samples is similar to that of the Pd₁/BNC product except for the addition of 2,2′-bipyridinepalladium dichloride into the precursor solution.

Reductive Amination Procedures

Five milligrams of Pd₁/BNC single-atom catalyst, 63 mg of 5-HMF (0.5 mmol), 5 mL of isopropanol (acting as the reductant and solvent), 0.5 mL of NH₃ŸH₂O solution, and 155 mg of biphenyl (1 mmol, acting as an internal standard) were mixed in a round-bottom flask (25 mL capacity) fitted with a reflux condenser. The resulting mixture was vigorously stirred (700 rpm) at 80°C for a given reaction time. The reaction mixture was then centrifuged to separate the catalyst, and anhydrous MgSO₄ was added to the liquid phase to remove possible moisture. Thereafter, a certain amount of isopropanol was added to the dried liquid to bring its volume to 10 mL. Finally, the conversion of 5-HMF and the yield of amines were determined by a 7890A gas chromatograph (GC, Agilent Co., USA) equipped with an HP-5MS capillary column (30 m × 0.25 mm × 0.25 µm) using biphenyl as an internal standard, and the structure of the products was confirmed by GC-mass spectrometry (MS). The procedures of the reductive amination of other biomass-derived aldehydes are almost the same except 5-HMF is replaced with other aldehydes. For the reductive amination of 5-HMF with other amines, the procedures were almost the same except for replacing 0.5 mL of NH₃ŸH₂O solution with 1.2 mmol of the corresponding amines or nitro compounds.

For the cycle performance evaluation, 50 mg of the Pd₁/BNC single-atom catalyst, 630 mg of 5-hydromethylfurfural (5 mmol), 50 mL of isopropanol (acting as the reductant and solvent), 5 mL of NH₃ŸH₂O solution, and 1.55 g of biphenyl (10 mmol, acting as an internal standard) were mixed in a round-bottom flask (250 mL capacity) fitted with a reflux condenser. The experimental procedures for the cycle performance evaluation are similar to the conventional HMF reductive-amination process discussed above. After each cycle reaction, the spent Pd₁/BNC single-atom catalyst was recovered via centrifugation, washed several times with isopropanol, and then used directly for another cycle. For hot filtration, the reaction lasted for the initial 3 h, then the catalyst was hot filtrated, and the remaining liquid reacted for an additional 3 h.

Procedures For The Gram-scale Reaction

A total of 270 mg of Pd₁/BNC SCA, 6.30 g of 5-hydroxyfurfural (50 mmol), 150 mL of isopropanol (acting as the reductant and solvent), and 30 mL of NH₃ŸH₂O solution were mixed in a round-bottom flask (500 mL capacity) fitted with a reflux condenser. The resulting mixture was vigorously stirred (700 rpm) at 90°C for 12 h. After the reaction, the reaction mixture was centrifuged to separate the catalyst, and the remaining liquid was evaporated at 50°C under reduced pressure to remove the isopropanol solvent. Thereafter, the remaining residue was dissolved with ethyl acetate, and a certain amount of HCl (35%, v/v) was added to the solution. Amine was converted into its quaternary ammonium form and crystallized out from the solution. The crystallized quaternary ammonium salt was separated via filtration, washed with CH₂Cl₂ several times, and dried under vacuum for 24 h. The dried solid was then weighed to determine the yield, and its structure and purity were determined with GC‒MS and ¹³C/¹H-NMR (AV 400, Bruker Inc., Switzerland) after reconversion into its amine form by heating at 80°C under reduced pressure.

Acknowledgements

The authors thank the National Natural Science Foundation of China (22122608, 21976170, U20A20325, 51821006, 52100195 and 52192684) for supporting this work.

Author contributions

W.J.L., H.Q.Y. and Y.W. devised the concept. H.Q.Y. and Y.W. supervised the project. W.J.L. conducted the reduction amination experiments, product analysis, and mechanism illustration. X.Z. performed the catalyst synthesis and characterizations. Y.M. conducted the DFT calculations. J.W.H. helped with the product analysis. J.J.C. helped with the DFT calculations. W.J.L. X.Z. and Y.M. wrote the manuscript. H.Q.Y., Y.W. and W.J.L. revised the manuscript.

Competing interests

The authors declare no competing interests.

Lawrence, S. A. Amines: synthesis, properties and applications. (Cambridge University Press, 2004).
Ricci, A. Amino group chemistry: from synthesis to the life sciences. (John Wiley & Sons, 2008).
Murugesan, K. et al. Catalytic reductive aminations using molecular hydrogen for synthesis of different kinds of amines. Chem. Soc. Rev. 49, 6273–6328 (2020).
Afanasyev, O. I., Kuchuk, E., Usanov, D. L. & Chusov, D. Reductive amination in the synthesis of pharmaceuticals. Chem. Rev. 119, 11857–11911 (2019).
Murugesan, K., Beller, M. & Jagadeesh, R. V. Reusable nickel nanoparticles-catalyzed reductive amination for selective synthesis of primary amines. Angew. Chem. 131, 5118–5122 (2019).
Trowbridge, A., Walton, S. M. & Gaunt, M. J. New strategies for the transition-metal catalyzed synthesis of aliphatic amines. Chem. Rev. 120, 2613–2692 (2020).
Nugent, T. C. Chiral amine synthesis: methods, developments and applications. (John Wiley & Sons, 2010).
Tripathi, R. P., Verma, S. S., Pandey, J. & Tiwari, V. K. Recent development on catalytic reductive amination and applications. Curr. Org. Chem. 12, 1093–1115 (2008).
Baxter, E. W. & Reitz, A. B. Reductive aminations of carbonyl compounds with borohydride and borane reducing agents. Org. React. 59, 1–714 (2004).
Chatterjee, M., Ishizaka, T. & Kawanami, H. Reductive amination of furfural to furfurylamine using aqueous ammonia solution and molecular hydrogen: an environmentally friendly approach. Green Chem. 18, 487–496 (2016).
Cukalovic, A. & Stevens, C. V. Production of biobased HMF derivatives by reductive amination. Green Chem. 12, 1201–1206 (2010).
Karve, V. V. et al. Efficient reductive amination of HMF with well dispersed Pd nanoparticles immobilized in a porous MOF/polymer composite. Green Chem. 22, 368–378 (2019).
Roylance, J. J. & Choi, K.-S. Electrochemical reductive amination of furfural-based biomass intermediates. Green Chem. 18, 5412–5417 (2016).
Xie, C. et al. Ambient reductive amination of levulinic acid to pyrrolidones over Pt nanocatalysts on porous TiO₂ nanosheets. J. Am. Chem. Soc. 141, 4002–4009 (2019).
Besson, M., Gallezot, P. & Pinel, C. Conversion of biomass into chemicals over metal catalysts. Chem. Rev. 114, 1827–1870 (2014).
Mika, L. T., Cséfalvay, E. & Németh, Á. Catalytic conversion of carbohydrates to initial platform chemicals: Chemistry and sustainability. Chem. Rev. 118, 505–613 (2017).
Wang, T. et al. Rational design of selective metal catalysts for alcohol amination with ammonia. Nat. Catal. 2, 773–779 (2019).
Senthamarai, T. et al. Simple ruthenium-catalyzed reductive amination enables the synthesis of a broad range of primary amines. Nat. Commun. 9, 4123 (2018).
Wang, Y.-G., Mei, D., Glezakou, V.-A., Li, J. & Rousseau, R. Dynamic formation of single-atom catalytic active sites on ceria-supported gold nanoparticles. Nat. Commun. 6, 6511 (2015).
Matsubu, J. C. et al. Adsorbate-mediated strong metal–support interactions in oxide-supported Rh catalysts. Nat. Chem. 9, 120–127 (2017).
Wang, A., Li, J. & Zhang, T. Heterogeneous single-atom catalysis. Nat. Rev. Chem. 2, 65–81 (2018).
Yang, X.-F. et al. Single-atom catalysts: a new frontier in heterogeneous catalysis. Acc. Chem. Res. 46, 1740–1748 (2013).
Hannagan, R. T., Giannakakis, G., Flytzani-Stephanopoulos, M. & Sykes, E. C. H. Single-atom alloy catalysis. Chem. Rev. 120, 12044–12088 (2020).
Liu, L. & Corma, A. Metal catalysts for heterogeneous catalysis: From single atoms to nanoclusters and nanoparticles. Chem. Rev. 118, 4981–5079 (2018).
Lang, R. et al. Hydroformylation of olefins by a rhodium single-atom catalyst with activity comparable to RhCl(PPh₃)₃. Angew. Chem. Int. Ed. 55, 16054–16058 (2016).
Zhang, L., Ren, Y., Liu, W., Wang, A. & Zhang, T. Single-atom catalyst: a rising star for green synthesis of fine chemicals. National Sci. Rev. 5, 653–672 (2018).
Yan, H., Su, C., He, J. & Chen, W. Single-atom catalysts and their applications in organic chemistry. J. Mater. Chem. A 6, 8793–8814 (2018).
Wan, J. et al. In situ phosphatizing of triphenylphosphine encapsulated within metal–organic frameworks to design atomic Co₁–P₁N₃ interfacial structure for promoting catalytic performance. J. Am. Chem. Soc. 142, 8431–8439 (2020).
Wang, Y. et al. High-efficiency oxygen reduction to hydrogen peroxide catalyzed by nickel Single-atom catalysts with tetradentate N₂O₂ coordination in a three-phase flow cell. Angew. Chem. Int. Ed. 59, 13057–13062 (2020).
Zhao, Y. et al. Can boron and nitrogen co-doping improve oxygen reduction reaction activity of carbon nanotubes? J. Am. Chem. Soc. 135, 1201–1204 (2013).
Chen, S. et al. Designing boron nitride islands in carbon materials for efficient electrochemical synthesis of hydrogen peroxide. J. Am. Chem. Soc. 140, 7851–7859 (2018).
Wang, B. et al. Hydrochlorination of acetylene on single-atom Pd/N-doped carbon catalysts: Importance of pyridinic-N synergism. Appl. Catal. B 272, 118944, (2020).
Schiros, T. et al. Connecting Dopant Bond Type with Electronic Structure in N-Doped Graphene. Nano Letters 12, 4025–4031, doi:10.1021/nl301409h (2012).
Stephan, D. W. & Erker, G. Frustrated lewis pair chemistry: Development and perspectives. Angew. Chem. Int. Ed. 54, 6400–6441 (2015).
Primo, A., Neatu, F., Florea, M., Parvulescu, V. & Garcia, H. Graphenes in the absence of metals as carbocatalysts for selective acetylene hydrogenation and alkene hydrogenation. Nat. Commun. 5, 5291 (2014).
Chen, X. et al. Metal-free H₂ activation for highly selective hydrogenation of nitroaromatics using phosphorus-doped carbon nanotubes. ACS Appl. Mater. Interf. 12, 654–666 (2020).
Baik, S. & Lee, J. W. Effect of boron–nitrogen bonding on oxygen reduction reaction activity of BN Co-doped activated porous carbons. RSC Adv. 5, 24661–24669 (2015).
Jagadeesh, R. V. et al. MOF-derived cobalt nanoparticles catalyze a general synthesis of amines. Science 358, 326–332 (2017).
Yagafarov, N. Z. et al. The synthesis of sterically hindered amines by a direct reductive amination of ketones. Chem. Commun. 52, 1397–1400 (2016).
Zhou, P. et al. High performance of a cobalt–nitrogen complex for the reduction and reductive coupling of nitro compounds into amines and their derivatives. Sci. Adv. 3, 1601945 (2017).
Liu, W.-J., Tian, K. & Jiang, H. One-pot synthesis of Ni–NiFe₂O₄/carbon nanofiber composites from biomass for selective hydrogenation of aromatic nitro compounds. Green Chem. 17, 821–826 (2015).
Zhou, P. et al. A versatile cobalt catalyst for the reductive amination of carbonyl compounds with nitro compounds by transfer hydrogenation. Appl. Catal. B 210, 522–532 (2017).
Zhou, P. & Zhang, Z. One-pot reductive amination of carbonyl compounds with nitro compounds by transfer hydrogenation over Co–N_x as catalyst. ChemSusChem 10, 1892–1897 (2017).
Ren, Y. et al. Unraveling the coordination structure-performance relationship in Pt1/Fe₂O₃ single-atom catalyst. Nat. Commun. 10, 4500 (2019).
Liu, W. et al. A durable nickel single-atom catalyst for hydrogenation reactions and cellulose valorization under harsh conditions. Angew. Chem. Int. Ed. 57, 7071–7075 (2018).
Stephan, D. W. Frustrated Lewis pairs. J. Am. Chem. Soc. 137, 10018–10032 (2015).
Zhang, S. et al. Solid frustrated-Lewis-pair catalysts constructed by regulations on surface defects of porous nanorods of CeO₂. Nat. Commun. 8, 15266, doi:10.1038/ncomms15266 (2017).
Alonso, D. M. et al. Increasing the revenue from lignocellulosic biomass: Maximizing feedstock utilization. Science Advances 3, 1603301 (2017).

Tables 1 and 2 are available in the Supplementary Files section.

There is NO Competing Interest.

Download PDF

Version 1

posted

You are reading this latest preprint version

Modulation of Electronic Structure of Single-Atom Catalysts for Sustainable Production of Full-Spectrum Amines from Renewable Biomass

Status:

Version 1

Abstract

Figures

Introduction

Results

Discussion

Reaction pathway and catalytic mechanism

Scalability And Sustainability Evaluation

Methods

Preparation of single-atom Pd₁/BNC catalysts

Reductive Amination Procedures

Procedures For The Gram-scale Reaction

Declarations

References

tables

Additional Declarations

Supplementary Files

Status:

Version 1

Modulation of Electronic Structure of Single-Atom Catalysts for Sustainable Production of Full-Spectrum Amines from Renewable Biomass

Status:

Version 1

Abstract

Figures

Introduction

Results

Discussion

Reaction pathway and catalytic mechanism

Scalability And Sustainability Evaluation

Methods

Preparation of single-atom Pd1/BNC catalysts

Reductive Amination Procedures

Procedures For The Gram-scale Reaction

Declarations

References

tables

Additional Declarations

Supplementary Files

Status:

Version 1

Preparation of single-atom Pd₁/BNC catalysts