Xiangdi
Zeng
a,
Zi
Wang
a,
Meng
He
a,
Wanpeng
Lu
a,
Wenyuan
Huang
b,
Bing
An
a,
Jiangnan
Li
b,
Mufan
Li
b,
Ben F.
Spencer
c,
Sarah J.
Day
d,
Floriana
Tuna
ae,
Eric J. L.
McInnes
a,
Martin
Schröder
*a and
Sihai
Yang
*ab
aDepartment of Chemistry, University of Manchester, Manchester, M13 9PL, UK. E-mail: M.Schroder@manchester.ac.uk
bCollege of Chemistry and Molecular Engineering, Beijing National Laboratory for Molecular Sciences, Peking University, Beijing, 100871, China. E-mail: Sihai.Yang@pku.edu.cn
cDepartment of Materials, The University of Manchester, Manchester, M13 9PL, UK
dDiamond Light Source, Harwell Science Campus, Didcot, Oxfordshire OX11 0DE, UK
ePhoton Science Institute, University of Manchester, Manchester, M13 9PL, UK
First published on 31st October 2024
Phenylacetylene is a detrimental impurity in the polymerisation of styrene, capable of poisoning catalysts even at ppm levels and significantly degrading the quality of polystyrene. The semi-hydrogenation of phenylacetylene to styrene instead of ethylbenzene is, therefore, an important industrial process. We report a novel cerium(IV)-based metal–organic framework (denoted as Ce-bptc), which comprises {Ce6} clusters bridged by biphenyl-3,3′,5,5′-tetracarboxylate linkers. Ce-bptc serves as an ideal support for palladium nanoparticles and the Pd@Ce-bptc catalyst demonstrates excellent catalytic performance for semi-hydrogenation of phenylacetylene, achieving a selectivity of 93% towards styrene on full conversion under ambient conditions with excellent reusability. In situ synchrotron X-ray powder diffraction and electron paramagnetic resonance spectroscopy reveal the binding domains of phenylacetylene within Ce-bptc, and details of the reaction mechanism are discussed.
Herein, we report a novel Ce(IV)-based MOF (Ce-bptc), consisting of {Ce6} clusters bridged by biphenyl-3,3′,5,5′-tetracarboxylate linkers. Incorporation of Pd nanoparticles onto Ce-bptc affords an efficient catalyst Pd@Ce-bptc for the semi-hydrogenation of phenylacetylene, yielding a selectivity of 93% towards styrene on full conversion under ambient conditions. The catalytic mechanism has been elucidated by in situ synchrotron X-ray powder diffraction (SXPD), X-ray photoelectron spectroscopy (XPS), and electron paramagnetic resonance (EPR) spectroscopy.
In Ce-bptc six Ce(IV) centers are assembled into an octahedron, where the eight faces are occupied by μ3-O2− or μ3-OH− groups in each {Ce6O4(OH)4(CO2)12} cluster. The Ce(IV) atom is coordinated to eight O atoms, four of which belong to four different bptc4− ligands and the remaining four are from μ3-O2−/OH− groups. Each bptc4− ligand is connected to four different {Ce6} clusters with each carboxylate coordinated by two adjacent Ce(IV) centers in the same cluster. Ce-bptc shows cubic cages with {Ce6} clusters on the vertices and a planar bptc4− linker on the faces (Fig. 1b). The cubic cages have a dimension of ∼9.4 Å and are interconnected through smaller tetrahedral cages (dimension of ∼3.5 Å) located at the twelve edges of the cubic cages (Fig. 1c). Thermogravimetric analysis (TGA) shows that Ce-bptc is stable up to ca. 250 °C before decomposition (Fig. S10†).
To encapsulate Pd nanoparticles, Pd precursors (at 0.3, 2.0, 3.7 and 5.4 wt%) were dissolved in water and then embedded within Ce-bptc using a double solvent method17 This was followed by treatment under a flow of H2 at 150 °C for 2 h to yield Pd@Ce-bptc, which shows retention of the framework structure of Ce-bptc as confirmed by powder X-ray diffraction (PXRD, Fig. 2a and S1†). No obvious Bragg peaks for Pd nanoparticles or PdOx were observed, indicating the excellent dispersion of Pd species in Ce-bptc. The diffuse reflectance infrared Fourier transform spectra (DRIFTs) of Ce-bptc and Pd@Ce-bptc confirm the retention of the structure and stretching mode of μ3-OH− groups upon Pd doping (Fig. 2b). The encapsulation of Pd nanoparticles in Ce-bptc is also accompanied by a reduction of the Brunauer–Emmett–Teller (BET) surface area from 532 to 390 m2 g−1 for Ce-bptc and Pd@Ce-bptc, respectively (Fig. S3†). Scanning electron microscopy (SEM) and energy dispersive X-ray (EDX) analysis show that Pd nanoparticles are distributed homogeneously in the MOF (Fig. S7 and S8). For comparison, we also prepared another benchmark MOF, Ce-UiO-66,12 which shows reduced crystallinity and almost no porosity upon the encapsulation of Pd (Fig. S2 and S9†).
XPS experiments were conducted to study the electronic state of Ce and Pd sites (Fig. 2c and d). The Ce 3d spectra were analysed according to well-established models;18 it was found that both Ce(IV) and Ce(III) species (73.4% and 26.6%, respectively) exist in Ce-bptc, which is consistent with other reported Ce(IV)-based MOFs.19 Upon reduction, the Ce(IV) satellite peak at 916.8 eV decreased significantly, and the concentration of Ce(III) increased from 26.6% to 62.1%; the increased amount of Ce(III) may facilitate subsequent electron transfer. To investigate the interactions between the Pd nanoparticles and MOF host, the Pd 3d XPS spectra of Pd@Ce-bptc and Pd/C were measured (Fig. 2d). The Pd 3d5/2 peak for Pd@Ce-bptc was located at 335.7 eV (calibrated against the C peak 1s = 284.8 eV). Compared with the Pd 3d5/2 peak of Pd/C (335.1 eV),20 that for Pd@Ce-bptc was found at a higher binding energy, indicating electron transfer from Pd to Ce in Pd@Ce-bptc.21 This shift in binding energy for Pd 3d5/2 is not due to issues with charge calibration as other peak positions are at the expected values after calibration.
The catalytic performances of a range of Pd-based catalysts (Pd@Ce-bptc, Pd@Zr-bptc, Pd@Ce-UiO-66, Pd@Zr-UiO-66, Pd@CeO2, and Pd@ZrO2) were tested for the semi-hydrogenation of phenylacetylene (Fig. S12†). The impact of different loadings of Pd nanoparticles (0.3, 2.0, 3.7 and 5.4%) in Pd@Ce-bptc was studied. At 25 °C and under 1 bar of H2 for 1 h, the conversion of phenylacetylene increased from 33% to 91% upon increasing loading of Pd nanoparticles, and the TOF decreased from 2385 to 362 h−1 (Fig. 3a). To balance the TOF and conversion, Pd@Ce-bptc with 2.0 wt% of Pd was used for further tests and comparison with other catalysts.
For the reaction at 25 °C and 1 bar of H2 for 2.5 h, Pd@Ce-bptc-2.0% shows a selectivity of 93% towards styrene on full conversion of phenylacetylene (Fig. 3b). Control experiments with Pd/C (Pd supported on carbon black) demonstrate >99% conversion with only 10% selectivity towards styrene (the major product is ethylbenzene), while bare Ce-bptc shows no activity (Table 1). A powdered mixture of 2% Pd/C and Ce-bptc shows a selectivity of 45% towards styrene on full conversion, demonstrating the important role of confinement of Pd nanoparticles in the pores to promote semi-hydrogenation rather than full hydrogenation. The selectivity of Pd@Ce-bptc-2.0% is higher than that of the commercial Lindlar catalyst (which shows a selectivity of 87% on full conversion), other supported Pd@MOF catalysts (Pd@Ce-UiO-66, Pd@Zr-bptc, and Pd@Zr-UiO-66) and metal oxides (Pd@CeO2 and Pd@ZrO2) (Table 1). To explore its reusability, used Pd@Ce-bptc was tested for 5 consecutive runs, and the conversion of phenylacetylene remained at 97%. After 5 consecutive runs, the PXRD pattern of the used catalyst did not show any obvious changes, demonstrating its high stability (Fig. 3c and Fig. S14†). In comparison, the conversion over Pd@Zr-bptc decreased from 98% to 88% after 5 runs (Fig. S15†). The improved stability of Pd@Ce-bptc compared with Pd@Zr-bptc is consistent with the XPS result, which shows stronger interaction between Pd nanoparticles and Ce-bptc (Fig. 2d and S12†).
The semi-hydrogenation of phenylacetylene-based substrates with various para-substitutions was tested over Pd@Ce-bptc to investigate the impact of substitution groups on the reaction (Table 1). The reaction time for all substrates was set at 1 h, and conversions for 4-methylphenylacetylene, 4-tert-butylphenylacetylene, 4-chlorophenylacetylene and 4-bromophenylacetylene of 55%, 85%, 42% and 20%, respectively, were observed (Fig. S16–20†). These results suggest that substrates with an electron-donating group show higher conversions compared with phenylacetylene, while substrates with an electron-withdrawing group show lower conversions. Overall, these results suggest that Pd@Ce-bptc is a highly active and stable catalyst for semi-hydrogenation at room temperature.
To investigate the host–guest interaction between the adsorbed substrate and Ce-bptc, SXPD data were collected for phenylacetylene-loaded Ce-bptc, and highly satisfactory Rietveld refinement revealed the binding domains for the substrate (Fig. 4a, S20 and Table S1†). The ethynyl group of adsorbed phenylacetylene interacts with the bridging oxygen and carboxylate oxygen centres of the {Ce6O6} clusters to form hydrogen bonds [C–H⋯bridge = 346(5) Å, C–H⋯ligand = 3.15(8) Å]. These adsorbed molecules are further stabilised by parallel-displaced π⋯π interactions between the CC bond/benzene rings of guest molecules and ligands of Ce-bptc, with inter-planar distances of 4.02(1) Å and 4.09(1) Å, respectively. Thus, the SPXRD study confirms the presence of notable host–guest binding interactions that facilitate charge transfer and thus promote hydrogenation.22,23
It has been reported that the hydrogenation of phenylacetylene to styrene and ethylbenzene may proceed via two different mechanisms: so-called one-pathway or two-pathway.24 The one-pathway mechanism involves direct hydrogenation of the triple bond to a single bond, while the two-pathway mechanism proposes a two-step hydrogenation process involving the formation of single or multiple adsorbed species.25,26 Given the high selectivity towards styrene observed for Pd@Ce-bptc, it is suggested that the reaction predominantly follows the two-pathway mechanism involving adsorbed intermediates.
To identify any radical species in this reaction, in situ EPR spin-trapping experiments were conducted. Considering the transient nature of many free radicals, with lifetimes significantly shorter than the acquisition time of EPR spectra, N-tert-butyl-α-phenylnitrone (PBN) was employed as a spin trap reagent27 to identify radicals as long-lived PBN-radical adducts. When H2 and the catalyst were present in the reaction solution, an intense seven-line EPR signal with a g-factor of 2.0055 and hyperfine coupling constants AN = 15.0 G and AH1 = AH2 = 7.4 G was detected. This signal is unequivocally attributed to the PBN-H radical,28 indicating that H2 is activated directly by the confined Pd nanoparticles, leading to homolytic dissociation of the H–H bond and the generation of H˙ radicals (Fig. 4b, c and Table S2†). Interestingly, no PBN-H signal was observed when the substrate was added to the reaction solution, suggesting that the hydrogenation involving H˙ radicals and the substrate is extremely rapid, preventing H˙ radicals from being trapped by PBN (Fig. S22†). Thus, the EPR study confirms the reaction pathway (Fig. 4d): H2 is first activated by the Pd nanoparticles, resulting in the homolytic dissociation of the H–H bond and the release of H˙ radicals, which subsequently attack the CC bond of adsorbed phenylacetylene, forming the desired product, styrene, which can readily desorb from the adsorption site in competitive adsorption with phenylacetylene, thus avoiding over-hydrogenation to ethylbenzene.
Footnote |
† Electronic supplementary information (ESI) available. CCDC 2205374 and 2205379. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d4qi02225d |
This journal is © the Partner Organisations 2025 |