Abstract
Palladium nanoparticles stabilized by poly (N-vinyl-2-pyrrolidone) (PVP) can be synthesized by corresponding Pd(acac)2 (acac = acetylacetonate) as precursor in methanol at 80°C for 2 h followed by reduction with NaBH4 and immobilized onto SiO2 prepared by sol-gel process under acidic conditions (HF or HCl). The PVP/Pd molar ratio is set to 6. The effect of the sol-gel catalyst on the silica morphology and texture and on Pd(0) content was investigated. The catalysts prepared (ca. 2% Pd(0)/SiO2/HF and ca. 0,3% Pd(0)/SiO2/HCl) were characterized by TEM, FAAS, and SEM-EDS. Palladium nanoparticles supported in silica with a size 6.6 ± 1.4 nm were obtained. The catalytic activity was tested in hydrogenation of alkenes.
1. Introduction
Poly (N-vinyl-2-pyrrolidone) stabilized metal nanoparticles have attracted considerable interest when it comes to preventing coagulation and precipitation of the metal nanoparticles. Polymer protecting agents allow preparation of metal colloids that can be stable for months with reasonable control over size as well as shape [1]. The nanoparticles are kinetically unstable with respect to aggregation or the bulk metal and should be stabilized by electrostatic or steric protection [2, 3]. Some of the protecting agents provide steric stabilization such as surfactants [4], ionic liquids [5–8], and polyoxoanions [9]. Their main disadvantage, however, is the problematic separation of the catalytic particles from the product and unused reactants at the end of the reaction. Immobilization of the particles on a solid support can facilitate the separation process, but may simultaneously lead to a decrease in activity. The nanoparticle synthesis involves addition of a polymer to the metal salt followed by chemical or thermal reduction to produce a stable black suspension of Pd(0) particles. The types of stabilizers and concentration, the solvent polarity [10], and the aging time of colloidal suspensions [11] can have an effect on the size, shape and catalytic activity of palladium nanoparticles. Platinum nanoparticles protected by PVP have been synthesized by alcohol reduction methods and incorporated into mesoporous SBA-15 silica during hydrothermal synthesis [12]. There have been several very recent reports in the literature of the catalytic properties of nanoparticles of Ag [13, 14], Rh [15], Pt [16, 17] and Pd [18] in PVP. Palladium colloid solutions stabilized by poly(N-vinyl-2-pyrrolidone) can be used as catalysts with high reactivity in C–C bond formation microwave-assisted reactions as shown by Heck [19] and Suzuki [20]. Catalytic hydrogenation reactions have been extensively evaluated with metal nanoparticles [21, 22]. We were successful in applying the new method for the synthesis of PVP-stabilized palladium (0) nanoparticles and immobilized in SiO2 employing the sol-gel process as catalysts for hydrogenation reactions. The mean diameters of the palladium nanoparticles are determined by TEM. The aim of this study is to examine the catalytic activity in hydrogenation of alkenes.
2. Experimental Section
2.1. General
Palladium(II) acetylacetonate (Pd(acac)2), poly(N-vinyl-2-pyrrolidone) (PVP-55, average molecular weight 55.000), and sodium borohydride (98%) were purchased from Sigma-Aldrich (Brazil). Pd/C (5%) was provided by Degussa (Brazil). Methanol was purchased from Synth and used as received. All other chemicals were purchased from commercial sources and used without further purification. All solutions were prepared with double deionized water. Gas chromatography analysis was performed with an Rtx-5 (30 m × 0,25 mm × 0,25 μm) gas chromatograph with an FID detector and a 30 m capillary column with 5% phenyl and 95% dimethylpolysiloxane as the carrier gas and the N2 (2,2 mL/min). The nanoparticle formation and hydrogenation reactions were carried out in a modified Fischer-Porter bottle immersed in a silicone oil bath and connected to a hydrogen tank. The temperature was maintained at 75°C by a hot-stirring plate connected to a digital controller (ETS-D4 IKA) with stirring at 1200 rpm. The fall in the hydrogen pressure in the tank was monitored with a pressure transducer interfaced through a Novus converter to a PC and the data workup via Microcal Origin 5.0.
2.2. Palladium Nanoparticles Solution
The suspensions of palladium nanoparticles were prepared by the alcohol reduction method [23]. The Palladium nanoparticles solution was prepared from a solution containing Pd(acac)2 (30 mg, 0.1 mmol) and it was dissolved in methanol (10 mL) and poly(N-vinyl-2-pyrrolidone (33 mg, 0.6 mmol of monomeric units, Mw~55.000) as a stabilizer. Sodium borohydride (20 mg, 0.6 mmol) was then added after reflux. The solution was refluxed for 2 h, resulting in a dark brown solution. The color change from yellow to dark indicates that the formation of palladium nanoparticles was completed.
2.3. Synthesis of Palladium Nanoparticles Immobilized in Silica
Silica immobilized Pd(0) nanoparticles were prepared by the sol-gel process under acidic conditions. Typical procedure: the Pd(0) nanoparticles solution (prepared in Section 2.2) was introduced in a Becker under vigorous stirring for 10 minutes and 2 mL of tetraethoxy orthosilicate (2 g, 9 mmol) was then added. The acid solution (HF or HCl) was introduced in a Becker under vigorous stirring at 50°C. The temperature was kept at 50°C and left to stand for a further 24 h. The resulting material was isolated by centrifugation (3000 rpm for 5 min) and washed several times with water and methanol and dried under vacuum.
2.4. Adsorption and Desorption Isotherms
The specific surface area analysis and pore size distribution of the samples was performed in a Gemini analyzer (Micromeritics Tristar 3020). The adsorption data was obtained at liquid-N2 temperature. The 100 mg of samples were preheated at 150°C for 4 h under vacuum. Specific area was assessed by the BET method. Average pore diameter and its distribution were obtained by BJH method.
2.5. Hydrogenations Reactions
The catalysts (50 mg) were placed in a Fischer-Porter bottle and the alkene (1 g) was added. The reactor was placed in an oil bath at 75°C and hydrogen was admitted to the system at constant pressure (4 atm) under stirring until the consumption of hydrogen stopped. The organic products were recovered by decantation and analyzed by GC. The reuse of the catalysts was performed by simple extraction of the organic phase (upper phase) followed by the addition of the alkenes.
2.6. Scanning Electron Microscopy (SEM) and Electron Dispersive Spectroscopy (EDS) Elemental Analysis
The morphology of the materials was analyzed by SEM using a JEOL model JSM 6060 with 20 kV and 5000 magnification and the electron dispersive spectroscopy (EDS) analysis was performed using a JEOL model JSM 5800 with 20 kV and 5000 magnification. The same instrument was used for the EDS with a Noran detector (20 kV with an acquisition time of 100 s and 5000 magnification).
2.7. Sample Preparation and TEM Analysis
The morphologies of the obtained particles were determined on a JEOL JEM-2010 equipped with an EDS system and a JEOL JEM-120 EXII electron microscope, operating at accelerating voltages of 120 kV. The TEM samples were prepared by deposition of the Pd(0) nanoparticles or Pd(0)/SiO2/HF isopropanol dispersions on a carbon-coated copper grid at room temperature. The histograms of the nanoparticles size distributions were obtained from the measurement of around 300 diameters (palladium nanoparticles) and 600 diameters (Pd(0)/SiO2/HF) and reproduced in different regions of the Cu grid assuming spherical shapes.
2.8. Flame Atomic Absorption (FAAS)
The Palladium present in the Pd(0)/SiO2/HF was measured using an Analytik Jena (Jena, Germany) flame atomic absorption spectrometer, equipped with a deuterium background correction system and a transversely heated graphite atomizer, using an air-acetylene (10 : 2.5 L min−1) flame under optimized conditions. Pyrolytically coated graphite tubes, hollow cathode lamp (operated at 3 mA), and deuterium lamp were supplied by Analytik Jena. Hollow cathode lamps of Pd (λ) 247.6 nm) from the same manufacturer were used as radiation sources.
3. Results and Discussion
The sol-gel process allows us to obtain solid products by creating an oxide network via progressive polycondensation reactions of molecular precursors in a liquid medium. The process essentially consists of two steps: hydrolysis and condensation. Both reactions are affected by the nature of the catalyst [24]. Therefore, in the present study, two main acids were evaluated: HF or HCl. The textural properties were further characterized by nitrogen adsorption. Specific area, pore diameter, and pore volume were calculated by the BET method (Table 1). The pore volume was shown to be dependent of the acidic conditions (HF or HCl). Nevertheless, the pore diameter was shown to be smaller for the materials prepared in the presence of HCl as catalyst (entry 2). The investigation of the palladium elemental concentrations in the catalytic samples is shown in Table 1. The concentrations of incorporated Pd(0) was determined using FAAS. The concentrations are expressed as % (m/m). It is evident that the Pd(0) metal concentration increased for the materials prepared in the presence of HF as catalysts (entry 1).
The metal distribution was determined by SEM-EDS analyses. Mapping showed a homogeneous Pd(0) distribution in the silica grains, for the materials prepared in the presence of HF as catalysts. Figure 1 illustrates the micrography of samples prepared by acids.
(a)
(b)
According to Figure 1, particle morphologies are in accordance to that usually observed for pure silica synthesized by these acid-catalyzed conditions. In this case a less organized, plate-like structure was observed for both cases.
Figure 2(a) shows the micrograph of the isolated Pd(0) particles, the mean size was shown to be ca. nm with irregularly shaped.
(a)
(b)
In the case of Pd(0)/SiO2/HF prepared by acid catalysis (HF), both the morphology and size (ca. nm) were maintained within the silica framework (Figure 3). It is clear that the morphological structure of the nanoparticles did not change with the presence of silica.
(a)
(b)
The supported catalysts were evaluated in hydrogenation reactions 1-hexene, 1-decene and cyclohexene hydrogenation reactions (Figure 4). For comparative purposes, the data concerning the catalytic activity of commercial Pd/C (5%) (Table 2) was also included.
As shown in Table 2, all the supported systems were more active exhibiting higher TOF in comparison to those of commercial Pd/C (5%). The structure generated in Pd(0)/SiO2/HCl might have afforded more active systems because the immobilized Pd content is less than Pd(0)/SiO2/HF. Besides, according to porosimetric measurements, the pore diameter was much smaller for the Pd(0)/SiO2/HCl system.
Finally, the catalytic material Pd(0)/SiO2/HF can be recovered by simple decantation and reused for at least nine times without any significant loss in catalytic activity (Figure 5).
(a)
(b)
4. Conclusion
The palladium nanoparticles protected with PVP were successfully supported in silica prepared by sol gel process (acid catalysis). The Pd(0) content in the resulting xerogels was shown to be dependent of the preparative route. In particular the silica-based systems prepared under acidic conditions (HF) were shown to be the most active and stable. The use of Pd(0)/SiO2/HF as catalysts in hydrogenation of alkenes gives better yields and TOF values under moderate conditions and shorter reaction times. Recycling experiments show that Pd(0)/SiO2/HF could be used nine times with essentially no loss in activity for the hydrogenation of 1-hexene. The palladium nanoparticles/silica combination exhibits an excellent synergistic effect that enhances the activity and durability of the catalyst for the hydrogenation of alkenes.
Conflict of Interests
The authors declare that there is no conflict of interests regarding the publication of this paper.
Acknowledgments
The authors would like to thank CAPES for partial financial support and PPGQTA for scholarships. They would also like to thank Prof. Dr. Jairton Dupont (UFRGS) and CME-UFRGS for the TEM and SEM microscopy analyses. They would also like to thank Dr. Fábio Andrei Duarte (UFSM) for performing the FAAS analysis.