The selective hydrogenation of carbon-carbon triple bonds to double bonds is an important process in the fine chemicals industry, and is relevant to commodities and other specialty chemicals production [1, 2]. The partial hydrogenation of 1, 4-butynediol (BYD) to 1, 4-butenediol (BED) is especially important, because BED is an important intermediate for producing endosulfan [3], vitamins A and B6, and is used in the synthesis of n-methyl pyrrolidone [4]. BED is also used as an additive in resin manufacturing [5]. Lindlar catalysts [6] (Pd supported on CaCO3 with a secondary doping of lead) are typical selective supported metal catalysts for the liquid phase hydrogenation of BYD. In these catalysts, lead is used as a dopant to promote the selective reduction of BYD to BED. A basic compound such as quinoline is also added during the reaction, to reduce the rate of the subsequent hydrogenation of BED to 1, 4-butanediol (BDO) [1]. The outstanding selectivity of Lindlar catalysts results from the synergistic effect of the lead compounds and other soluble additives. Various Pd and Ni-based catalysts can also be used in this reaction, but require combining with one or more mixed compounds of Cu, Zn, Ca, Cd, and Ga, and/or an organic base [10]. The conversion of BYD is significantly lower using these catalyst combinations, and toxic compounds are generally required to obtain high purity products [18, 19]. Developing alternative reusable catalysts for a more sustainable transformation is therefore of interest.
Metal-organic frameworks (MOFs) have attracted much recent attention in the field of heterogeneous catalysis [20]. The high surface area, narrow pore diameter, and specific composition of MOFs can yield nanoparticles (NPs) with a uniform size distribution and interesting catalytic activity and selectivity [21, 22]. MOF catalysts with nitrogen-containing groups have been investigated for tailoring the acidity/basicity, solubility/dispersibility, surface area, and selectivity toward the target products in hydrogenation reactions [26]. In our previous work, a polyvinyl-pyrrolidone (PVP)-protected Pd nano-sol supported on ZIF-8 was shown to be a highly active and selective catalyst for the hydrogenation of BYD [27].
Herein, an acetate-protected Pt nano-sol with a uniform particle size distribution of 1-2 nm was added during the synthesis of ZIF-8 at room temperature. The obtained Pt@ZIF-8 catalyst was used in the hydrogenation of BYD. We investigated the activity, selectivity, and recyclability of Pt@ZIF-8, and the specific role of ZIF-8 in this reaction.
The Pt NP colloid was prepared through the chemical reduction of H2PtCl6·6H2O by ethylene glycol in the presence of CH3COONa as a stabilizer, at 160 ℃ for 3 h [28]. Then, 11.35 g of 2-methy imidazole and 12 mL of the as-prepared Pt colloid ethanol solution were dissolved in 40 mL of H2O. An aqueous solution of Zn(NO3)2·6H2O (4 mL) was then added under stirring. The mixture was stirred for a further 1.5 h at 25 ℃, after which the black powder was collected by centrifugation. The product was washed twice with H2O and then twice with methanol. Drying under vacuum at 120 ℃ for 12 h yielded the 1.0 wt% Pt@ZIF-8 catalyst. The Pt loading could be controlled by adjusting the amount of Pt colloid added during the synthesis. The precise Pt content as determined by inductively coupled plasma atomic emission spectroscopy (ICP-AES) was 0.98 wt%. Fig. 1 shows the procedure for preparing Pt@ZIF-8.
Powder X-ray diffraction (XRD) patterns were recorded using a D/MAX 2400 diffractometer with monochromatic Cu Kα radiation (λ = 0.15418 nm), operated at 40 kV and 100 mA. The surface area, pore volume, and pore size distribution of the ZIF-8 support and Pt@ZIF-8 catalyst were determined from N2 adsorption-desorption isotherms at -196 ℃, using a Quantachrome Autosorb-IQ apparatus. The average particle size and size distribution of the samples were investigated by transmission electron microscopy (TEM), using a Philips CM200 apparatus operated at 120 kV. Powder samples were sonicated in ethanol, and dispersed on copper grids. Elemental analysis was performed by ICP-AES, using a Perkin-Elmer Optima 2000 DV apparatus.
The catalyst was reduced at 200 ℃ for 1.5 h under an atmosphere consisting of Ar:H2 = 2:1. The freshly reduced catalyst (0.05 g) was mixed with BYD (0.568 g) and 1, 2-propylene glycol (an internal standard required for gas chromatography analysis) in 10 mL of H2O. The mixture was transferred into a 50 mL batch reactor. The reactor was flushed with H2 three times, and catalytic hydrogenation was then carried out at a given H2 pressure and temperature. A blank test was performed following a conventional catalytic test procedure at 120 ℃ and p(H2) = 2 MPa, using pure ZIF-8 as the catalyst. Less than 2% conversion was obtained, indicating that ZIF-8 was inactive for the hydrogenation of BYD.
A mild room temperature procedure was used to synthesize ZIF-8, which ensured that the structure of the Pt nano-sol was preserved. XRD patterns of ZIF-8 and Pt@ZIF-8 prepared by the one-step synthesis are shown in Fig. 2(a). The XRD patterns show that introducing the Pt nano-sol had negligible effect on the formation of ZIF-8. Both ZIF-8 and Pt@ZIF-8 exhibited very high crystallinity. The characteristic diffraction peaks of Pt could not be detected, because of the low Pt loading and high crystallinity of ZIF-8. N2 adsorption measurements of ZIF-8 and Pt@ZIF-8 are shown in Fig. 2(b). Type I isotherms were observed for both ZIF-8 and Pt@ZIF-8, according to IUPAC classifications. High surface areas of 1786 m2/g for ZIF-8 and 1747 m2/g for Pt@ZIF-8 were observed, along with narrow pore size distributions of 1-2 nm. This indicated that the pore structure remained intact after loading with Pt.
TEM images of the Pt NPs and Pt@ZIF-8 are shown in Fig. 3. The ZIF-8 framework was well crystallized. It had a smooth surface, and consisted of particles of about 100-200 nm in size. No additional impurities were observed in the TEM images, in agreement with the XRD results. The TEM images show that the Pt nano-sol exhibited very little change after being encapsulated in the ZIF-8 crystals. Pt NPs of size of 1-2 nm were randomly dispersed throughout the ZIF-8 support.
The performance of Pt@ZIF-8 as a catalyst was tested in a batch reactor, using H2O as a solvent. The general hydrogenation pathways of BYD are shown in Scheme 1. BYD can be easily and completely hydrogenated to BDO by monometallic Pd or Pt catalysts. The selectivity to mono-ene products can be suppressed by poisoning with bismuth, sulfur adatoms, lead, and Ag [1]. However, these catalytic systems usually provide high selectivity to partially hydrogenated products, as well posing a hazard because of their toxicity. It would advantageous if a high partial hydrogenation selectivity could be obtained without using these additives.
When the reaction temperature was 50 ℃, a low BYD conversion of 10% was obtained when using Pt@ZIF-8 as a catalyst under a H2 pressure of 3 MPa for 4 h. The BYD conversion significantly increased with increasing temperature, and complete transformation was achieved at 120 ℃. Higher temperatures often lead to the further hydrogenation of BED to BDO. However, a high BED selectivity (>92%) was observed in the current study, regardless of the BYD conversion. The influence of temperature is shown in Fig. 4(a).
The pressure also affects the activity (Fig. 4(b)), and a decrease in conversion was anticipated at lower pressure. However, an excessively high pressure was also found to decrease the activity of the catalyst. This was attributed to the high internal surface area and the presence of metal cation sites in ZIF-8, which facilitated hydrogen adsorption. These properties mean that ZIF-8 supported Pt materials are promising candidates for hydrogen storage [2]. H2 pressure of over 3 MPa resulted in increased adsorption of H2 molecules into the micropores of ZIF-8, which limited the ability of the BYD substrate to reach Pt sites encapsulated within the ZIF-8 framework. The selectivity to BED was consistently high at approximately 95%, under various H2 pressure.
Kinetic data for the hydrogenation of BYD over 1 wt% Pt@ZIF-8 at 120 ℃ and 3 MPa of H2 pressure is shown in Fig. 5(a). Analysis of kinetic data can yield information about the interaction between structural properties and catalytic activity. Within the first 4 h, the BYD concentration decreased linearly with increasing reaction time, confirming the zero order kinetics with respect to BYD. This observation was consistent with previous results of the hydrogenation of BYD over noble metals [3]. At low reaction time (i.e. lower conversions), the hydrogenation of BYD primarily produced BED. The selectivity to BED was approximately 94% after 4 h of reaction time. BYD was almost completely converted within 4 h, and high selectivity was maintained during the entire reaction process. A further 2 h of reaction time after the complete consumption of BYD yielded no obvious hydrogenation to cis-BED or isomerization. The selectivity to BED remained high (93%) after 6 h of reaction, indicating that the reaction stopped after the first hydrogenation. The Pt@ZIF-8 catalyst in the hydrogenation of BYD in aqueous solution gave selectivity towards BED of up to 94% at conversions of up to 100%. This was further confirmed by kinetic data for BED hydrogenation performed under the same conditions, as shown in Fig. 5(b). Less than 20% of BED was converted to BDO after 5 h, and no side products were observed. In comparison, a conventional Pt/C catalyst exhibited poor selectivity for BED (<60%), under the same condition. The BED target product was also readily further hydrogenated to BDO, after the BYD substrate had been completely consumed [4]. These results suggested that the selectivity of the Pt@ZIF-8 catalyst resulted from its structural properties, rather than the reaction kinetics.
One possible reason forthe high partial hydrogenation selectivity may have been the inhibiting effect from the protective agent around the Pt NPs. This hypothesis can be verified by testing the catalytic performance of the Pt nano-sol, and the result is shown in Fig. 6(a). The acetate-protected Pt nano-sol exhibited a higher activity, requiring only 2 h to fully convert BYD. This may have resulted from the better contact between the Pt nano-sol and reactants, in the absence of the ZIF-8 support. The highest selectivity to BED (68%) was obtained at 1.5 h, and further hydrogenation and isomerization occurred as the reaction continued. Similar catalytic performance was observed when the ZIF-8 support was replaced with SBA-15, as shown in Fig. 6(b). Good selectivity was not observed in the absence of the ZIF-8 support, indicating that the favorable hydrogenation properties of Pt@ZIF-8 resulted from the role of ZIF-8. Zn- and N-containing organic ligands contained within the ZIF-8 structure are a common poison. Metal and organic inhibitors were also present to adjust the selectivity of the catalyst for the hydrogenation of BYD. The narrow pore diameter of ZIF-8 may also have suppressed isomerization.
The Pt@ZIF-8 catalyst could be reused without reactivation treatment, except requiring washing twice with H2O. The recyclability of 1.0 wt% Pt@ZIF-8 at 120 ℃ and p(H2) = 3 MPa is shown in Fig. 7(a). The kinetic data for BYD hydrogenation in Fig. 5(a) showed that BYD was fully converted in approximately 4 h. Thus, the reaction data in Fig. 7(a) was collected after 4 h. The sample exhibited reproducible performance over five runs, sustaining very good conversion of BYD (>91%) and excellent selectivity for BED (>92%).
The XRD patterns of Pt@ZIF-8 before and after reaction are shown in Fig. 7(b). No obvious differences were observed between the patterns of fresh Pt@ZIF-8 and that after two reaction cycles, indicating that the ZIF-8 structure remained intact. The characteristic XRD peaks of ZnO began to emerge after four reaction cycles, indicating the decomposition of ZIF-8. The characteristic XRD peaks of ZIF-8 completely disappeared after five reaction cycles, indicating complete decomposition of the support at the high reaction temperature. The residual XRD peaks were consistent with ZnO. BET results showed that the specific surface area decreased from 1747 to 40 m2/g, over the five reaction cycles. The metal composition remained largely unchanged, at 21.3% and 19.8% for fresh Pt@ZIF-8 and that subjected to five reaction cycles. This indicated that the high surface area and narrow pore dimeter of ZIF-8 were not the main reasons for the high selectivity of the Pt@ZIF-8 catalyst for BED. ZnO was obtained by calcining ZIF-8 in air at 500 ℃ for 3 h. The Pt/ZnO catalyst was prepared by adding ZnO to the Pt colloid. Dilute hydrochloric acid was then added to lower the pH to <3. This in turn lowered the concentration of the glycolate colloid stabilizer. The kinetics for the hydrogenation of BYD by the Pt/ZnO catalyst are shown for comparison. The Pt/ZnO catalyst exhibited poor selectivity for BED (<65%). After four reaction cycles, the crystal structure of ZIF-8 began to degrade, and the 2-methylimidazole ligand on ZIF-8 began to partially decompose and adsorb to the Pt surface, which affected the catalysts performance. These results showed that nitrogen-containing organic ligands can play an important role in the partial hydrogenation. A possible reason for the good reusability of the catalyst even after ZIF-8 decomposition was that most of the Pt nano- sol was located on the outer surface of ZIF-8.
We prepared a catalyst containing Pt supported on ZIF-8, via a rapid dynamic crystallization method at room temperature. The Pt@ZIF-8 catalyst exhibited enhanced selectivity for 1, 4-butenediol in the hydrogenation of 1, 4-butynediol, because of the role of Zn2+ and N-containing organic ligands in ZIF-8. The high activity of Pt@ZIF-8 resulted from the catalysts with high specific surface area and well-dispersed Pt NPs. Pt@ZIF-8 exhibited excellent performance and reusability, although the ZIF-8 structure began to degrade after more than two reaction recycles.
2016,
Vol. 37 Issue (9): 1555-1561 
PDF
2016,
Vol. 37 Issue (9): 1555-1561 PDF