2.1. Characterization of catalysts
Mono- and bimetallic palladium and palladium-silver catalysts modified with polysaccharides such as chitosan (Chit), 2-hydroxyethyl cellulose (HEC) and pectin (Pec) have been prepared by adsorption method in aqueous medium under ambient conditions and constant stirring. The solution of polymer and then metal salts were sequentially added into a zinc oxide suspension. The resulting composites were washed with water and dried in air. As result the following catalysts were obtained: Pd-HEC/ZnO, PdAg-HEC/ZnO, PdAg-Chit/ZnO, PdAg-Pec/ZnO.
The palladium and silver content in the catalysts was evaluated using spectrophotometry and potentiometry methods, respectively. Analysis of supernatant solution before and after sorption process showed that 91-99% of the introduced Pd and 99-100% of Ag were adsorbed on the polymer-modified ZnO. The calculated total metal content (Pd and Ag) in the obtained polymer-modified mono- and bimetallic catalysts was 0.46-0.49 wt % that close to expected value of 0.5 wt% (
Table 1).
Table 2 shows the results of elemental analysis of the catalysts. The palladium content in Pd-HEC/ZnO, PdAg-HEC/ZnO, PdAg-Chit/ZnO and PdAg-Pec/ZnO catalysts was found to be 0.49%, 0.36%, 0.44% and 0.48%, respectively. The silver content in the bimetallic PdAg-HEC/ZnO, PdAg-Chit/ZnO and PdAg-Pec/ZnO catalysts was found to be 0.18%, 0.15%, and 0.16 %, respectively. Thus, the total metal content in all catalysts was not less than 0.5 wt%, suggesting the almost complete adsorption of metals (Pd and Ag) on the polymer modified support materials (
Table 2). This is consistent with data obtained using spectrophotometry and potentiometry methods.
The results of X-ray powder diffraction analysis (XRD) of the ZnO, HEC/ZnO and PdAg-HEC/ZnO is shown in
Figure 1. All XRD patterns showed the characteristic peaks at 37.0⁰, 40.2⁰, 42.3⁰,55.8⁰, 66.7⁰, 74.5⁰, 78.9⁰, 80.9⁰, 82.3⁰, 86.8⁰, 92.6⁰, which correspond to the [100], [002], [101], [102], [110], [103], [200],[112], [201], [004], [202] planes of ZnO wurtzite structure (JCPDS card no. 79-0206) [
50]. A broad peak at 23º observed in polymer modified materials can be attributed to the amorphous phase of the HEC [
51]. No peaks related to palladium (Pd or PdO) and silver (Ag, AgCl) species were observed on the XRD pattern of PdAg-HEC/ZnO catalyst. This can be explained by low metal (Pd, Ag) content in the catalyst and small particles sizes [
52].
The modification of ZnO with HEC is also confirmed by standardized Brunauer–Emmett–Teller (BET) method (
Table 3). HEC/ZnO and PdAg-HEC/ZnO composites were characterized by decreased surface area in comparison with initial zinc oxide. It should be noted that PdAg-HEC/ZnO catalyst demonstrated the higher surface area to compare with that of HEC/ZnO composite. This is consistent with data obtained for similar systems modified with polyvinylpirollidone (PVP) [
53]. In [
53] decreasing the surface area from ZnO to PVP/ZnO explained by blockage of micropores in inorganic material after modification with the polymer, while increasing the surface area from PVP/ZnO to Pd-PVP/ZnO explained by a decreasing the surface coverage of the ZnO with PVP shell through changing in an orientation of the polymer functional groups from ZnO to Pd. Another explanation for such changing the surface area from unmodified zinc oxide to polymer-modified catalyst can be in changing the degree of agglomeration of ZnO particles after modification with a polymer and following adsorption of metal ions on the polymer/ZnO composite. That is, adding of HEC solution to zinc oxide suspension can lead to agglomeration of ZnO particle probably due to ZnO-HEC-ZnO bonding. Adsorption of metal ions on HEC/ZnO composite, on the contrary, can lead to decrease the agglomeration of ZnO particles due to formation ZnO-HEC-Pd(Ag) bonds. It should be noted that according to these assumptions, HEC can interact with both zinc oxide and metal (Pd and Ag) ions.
The interaction of the polysaccharide with other components of PdAg-HEC/ZnO catalyst and formation of polymer-metal complex on ZnO surface was confirmed by infrared spectroscopy (IRS).
Figure 2 shows the IR spectra of HEC, HEC/ZnO and PdAg-HEC/ZnO. HEC shows characteristic bands at 1061 and 1122 cm
-1, corresponding to the C-O-C stretching vibrations in the glucopyranose structure and C-O anti-symmetric vibrations, respectively.
Other characteristic bands at 1460 and 1385 cm
–1, have been attributed to O-H “plane deformation” and for C-H “symmetric bending vibrations” respectively in -CH
2O- [
54,
55]. The shifting of absorption bands of O-H, C-O-C, C-O and -CH
2O- groups in the IR spectrum of the catalyst, compared to the same bands in HEC and HEC/ZnO (
Figure 2, spectra 1 to 3), confirms the interaction of HEC with both ZnO and Pd (or Ag) ions.
In our prior studies [
53,
56] have been shown that interaction of polymers with metal ions on a support surface led to formation smaller active phase nanoparticles (~2 nm) to compare with those in similar unmodified supported catalysts. The transmission electron microscopy (TEM) image of Pd-HEC/ZnO catalyst showed the formation of finely dispersed palladium nanoparticles of ~2 nm evenly distributed on the surface of zinc oxide modified with HEC (
Figure 3). This confirms the stabilization role of the polysaccharide immobilized on ZnO.
The study of the Pd-HEC/ZnO catalyst using X-ray photoelectron spectroscopy (XPS) also confirms the presence of both polysaccharide and palladium on the surface of zinc oxide (
Figure 4).
Analysis of XPS spectrum of the Pd 3d(5/2) region (
Figure 4a) showed that palladium in the catalyst occurred in both oxidized (Pd
2+) and reduced (Pd
0) states with the binding energies of ~337.3 and ~336.0 eV, respectively [
57]. The presence of the small amount of zero-valent palladium is probably caused by the photocatalytic reduction of Pd
2+ in the presence of zinc oxide, which is known as an efficient photocatalyst [
58]. This suggests that a small part of the introduced palladium interacts with zinc oxide. This is confirmed by the fact that binding energy of Pd
0 has a positive shift (ca. 0.6 eV) due to strong metal-support interaction and formation of PdZn species [
59]. In the case of palladium in oxidized state such shifting was not observed, confirming stabilization of PdO particles with the polymer. The peak of the Zn 2p(3/2) at 1020.4 eV corresponds to Zn
2+ ions in a zinc oxide crystal lattice (
Figure 4b) [
60]. The XPS spectrum of the O 1s region (
Figure 4c) was de-convoluted into two peaks with binding energies at 530.6 (peak 1) and 532.5 eV (peak 2). According to the literature data [
61], peak 1 is related to O
2− ions in the Zn-O bonding of the wurtzite structure, and peak 2 is attributed to oxygen of OH groups on the zinc oxide surface. Peak 2 can also be assigned to the C-O oxygen in the HEC repeated unit [
62]. The de-convolution of the C 1s line in the XPS spectrum of the catalyst indicated that carbon is represented by two components (
Figure 4d). The main component centered at 285.0 eV is attributed to C-C bond in HEC macromolecule and no shifting of binding energy for this peak is observed. In case of the minor peak at 288.0 eV related to C-O bond in HEC some positive shift of binding energy (ca. 1.3 eV) was observed probably due to the interaction of C-O-C, C-O and -CH
2O- groups of HEC with both Pd
2+ and Zn
2+ of zinc oxide [
62].
Thus, the characterization of polysaccharide modified catalysts using physical chemical methods such as spectrophotometry, potentiometry, elemental analysis, XRD, IRS, BET, TEM and XPS indicated that polysaccharides and metal (Pd, Ag) ions quantitatively adsorbed on zinc oxide and the polymers interacted with both ZnO and active phase particles formed. The role of polymers is both to fix Pd and Ag species on zinc oxide surface and their stabilization. This is consistent with data obtained in our prior study for similar Pd-PVP/ZnO catalyst [
53]. However, in case of Pd-HEC/ZnO a small amount of Pd was also interacted with ZnO forming PdZn species. This can be explained by lesser affinity of HEC functional groups to metal ions in comparison with those of PVP.
2.2. Catalytic test
The obtained palladium and palladium-silver catalysts modified with natural polysaccharides (chitosan (Chit), 2-hydroxyethyl cellulose (HEC) or pectin (Pec)) were tested in hydrogenation process under mild conditions (0.1 MPa, 40 °C). The hydrogenation of 2-hexyn-1-ol was chosen as the model reaction. The possible pathways of the reaction are illustrated in
Scheme 1. The first step triple bond of the alkynol is reduced to C=C double one forming cis/trans-isomers which then hydrogenated to alkanol.
The test results of the developed catalysts in the hydrogenation process of 2-hexyn-1-ol are given in
Table 4. Polysaccharide-containing palladium-silver catalysts were found to be effective in this process. In descending order of influence on the hydrogenation rate, the stabilizing polymers form the following series: HEC > Pec > Chit (
Figure 5). However, the maximum reaction rate is achieved on monometallic Pd catalysts modified with HEC (
Table 4), probably due to the higher palladium content. The introduction of silver into the HEC-containing catalyst improved selectivity to 2-hexen-1-ol to compare with the selectivity on monometallic Pd catalyst (
Figure 6). The cis-alkenol selectivity of process is maximal on PdAg catalyst containing HEC and achieves 97.2%. Thus, the maximum yield of the alkenol was observed on the bimetallic PdAg catalyst stabilized with HEC and reached 89.4% at 93.0% conversion (
Table 4).
*Reaction conditions presented in
Figure 5.
The optimum temperature and catalyst loading for the PdAg catalysts modified with HEC and Chit have been determined (
Table 5). The initial reaction rate increased with increasing temperature up to 40°C.
Further temperature increase to 50°C lead to a significant decrease in the reaction rate, which could be explained by the shrinking the surface polymer layer of the catalyst and blocking the active centers [
53].
The increase in the catalyst loading from 0.01 g to 0.1 g leads to the increase in the reaction rate. The highest rate was achieved on the 0.1 g load of the both catalysts. At the same time, the maximum yield and selectivity to cis-hexene-1-ol decreased, probably due to accelerated hydrogenation of double bond of the cis-alkene into an alkane.
According to the chromatographic analysis, in the presence of HEC-modified catalyst, the conversion of 2-hexyn-1-ol to cis-2-hexen-1-ol bigan in the first half of the process (
Figure 7a). After almost complete disappearance of acetylenic alcohol in the reaction medium, a process of isomerisation of cis-alkenol to trans-form is initiated and parallel reduction of double bond of alkenols to saturated alcohol is performed. A similar change in the composition of the reaction mixture is observed when the process is carried out with the catalyst containing chitosan. A little difference is that the isomerization of cis- to trans-2-hexen-1-ol on this catalyst starts slightly before the complete conversion of the initial alkynol (
Figure 7b).
Reusability of PdAg-Chit/ZnO and PdAg-HEC/ZnO catalysts was studied by hydrogenation of successive portions of 2-hexyn-1-ol on one the same load of catalyst (
Figure 8). High stability demonstrated by reaction rate on the both catalysts remains nearly the same at least in 10 runs demonstrating high stability. This can be explained by swelling ability of polymer-metal shell of the catalysts in ethanol and negligible leaching active phase by preventing effect of the polymers [
53].
Thus, a comparison of the performance of palladium and palladium-silver catalysts stabilised with derivatives of natutal polysaccharies (
Table 5) as well as pectin confirms the prospectives of their use in synthesis of metal nanoparticles and further application in catalysis.