Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2020 Jul 28;5(31):19497–19505. doi: 10.1021/acsomega.0c01757

Nature and Distribution of Cu and Pd Species in CuPd/TiO2–Na Bimetallic Catalysts for Glycerol Hydrodeoxygenation

Alba N Ardila A †,‡,*, Erasmo Arriola-Villaseñor †,‡, Gustavo A Fuentes †,*
PMCID: PMC7424577  PMID: 32803043

Abstract

graphic file with name ao0c01757_0010.jpg

We demonstrated recently that CuPd/TiO2–Na bimetallic catalysts synthesized by sequential wet impregnation are active, selective, and stable for the hydrodeoxygenation (HDO) of glycerol into propylene glycol at low H2 pressure. The present study reports on the nature and distribution of Cu and Pd surface species in CuPd/TiO2–Na bimetallic catalysts using different scanning transmission electron microscopy techniques that supply cluster-specific alloying details. In particular, we used atomic-resolution Z-contrast imaging, X-ray energy-dispersive spectroscopy, and electron energy-loss spectroscopy. We also include X-ray photoelectron spectroscopy results. Our analysis shows that the metallic nanoparticles adopt mainly five different structures according to how the Cu and Pd atoms coordinate among themselves: a homogeneous CuPd alloy structure (45–61%), a Cu shell/CuPd core (15–23%), a smaller number of particles formed by Cu on the surface and Pd in the nucleus (10–17%), and there are also nanoparticles formed only by Pd (4–7%) or by Cu (8–13%). We determined that there is a inhomogeneous distribution of Cu and Pd in the bimetallic nanoparticles, with Cu being predominant on the surface (between 76 and 90% of the total area analyzed for each particle). Most bimetallic nanoparticles have sizes below 6 nm, the Pd monometallic nanoparticles are in the 2–4 nm range, whereas the monometallic Cu nanoparticles are larger than 8 nm. Bimetallic nanoparticles with sizes smaller than 6–7 nm are fundamentally made up of Cu0–Pd0 and Cu1+–Pd0. The nanoparticles with sizes greater than 7 nm consist of Cu2+ and Cu2+–Pd2+. Our obtained results also help describe reports about the activation of HDO by Pd–Cu in the absence of H2, an effect apparently not observed with other bimetallic systems.

Introduction

Copper catalysts with palladium acting as the promoter have a good performance for the glycerol hydrogenolysis in aqueous solution into propylene glycol (1,2-PDO),1−4 but they have poor stability. Recently, we reported the transformation of glycerol by liquid phase hydrogenolysis into 1,2-PDO with high selectivity over CuPd/TiO2–Na bimetallic catalysts that have good stability at low hydrogen pressure in a batch system.1

The high yield toward 1,2-PDO of the CuPd bimetallic catalysts was related to the presence of stable CuPd alloy nanoparticles formed by low-temperature reduction. Their formation is evidenced by temperature-programmed reduction with H2 (H2-TPR), CO and H2 chemisorption, N2O oxidation, and H2 titration. We also used high-resolution transmission electron microscopy (HRTEM), X-ray diffraction (XRD), diffuse reflectance ultraviolet–visible spectroscopy (DRS-UV–vis), and electron diffraction of selected area (EDSA). The bimetallic particles had an average diameter below the diameter of Cu monometallic particles. Formation of this alloy stabilized Cu0 against leaching and oxidation and resulted in a recycled catalyst. Although some advances were made in determining the distribution of the surface copper and palladium species deposited on TiO2 or in other support, that question was still unresolved.5−9

To better deduce the catalytic properties of supported bimetallic catalysts, it is necessary to study both the oxidation states and the distribution of the two metals within the individual nanoparticles. As demonstrated in our previous work, some information can be obtained by the various techniques used there, but to deduce the alloying process and surface segregation behavior of the individual bimetallic nanoparticles, we present here results using different scanning transmission electron microscopy (STEM) techniques that provide cluster-specific alloying information. In particular, we used X-ray energy-dispersive spectroscopy (EDS), electron energy-loss spectroscopy (EELS), and also X-ray photoelectron spectroscopy (XPS). Thus, in the present work, a joint bulk and surface characterization study was performed in detail to identify the structure/composition/oxidation state profile of individual nanoparticles from CuPd/TiO2–Na bimetallic catalysts with sub-nanometer resolution.

Experimental Section

Catalyst Characterization and Preparation

CuPd/TiO2–Na catalysts (nominal atomic ratio Cu/Pd = 1.67) were prepared by sequential impregnation on TiO2 (Degussa, P-25 powder). Pd and Cu loading was set at 5 wt % each. The catalysts were prepared by first treating TiO2 with NaOH dissolved in methanol (5 mL methanol/g TiO2 to obtain an initial loading of 5 wt % Na), followed by drying and calcination. Cu and Pd were then impregnated on Na–TiO2 as done on TiO2, with intermediate drying and calcination steps.

In order to determine whether there were changes in the CuPd catalyst during its use, we analyzed the structure of a fresh bimetallic catalyst (5Cu5Pd5Na-F), that of a catalyst reused after simple filtering, water rinsing, and lixiviation with water prior to reaction,1 which suffered a 10-fold decrease in Na content during synthesis (5Cu5Pd5Na-W), and that of a catalyst reused three times (5Cu5Pd5Na-R3). In the last case, the catalyst was filtered and washed with deionized water after each use and placed again in the reactor for a new run. Atomic absorption spectroscopy analysis of the reaction solution evidenced that close to 80% of the Na initially present in the fresh bimetallic catalyst (5Cu5Pd5Na-F) was found in the solution after the first use. Additionally, the leaching of Na decreased during further uses or washes until Na in the catalysts reached a stable residual level of 0.5 wt % for the 5Cu5Pd5Na-W and 5Cu5Pd5Na-R3. Data for these samples are given in Table 1. The reused catalysts did not lose activity when compared with a fresh catalyst, which is an indication of the stability of the active structure.

Table 1. Nomenclature Used for the Catalysts.

catalyst code
5%Cu-5%Pd/TiO2-5%Na-fresh 5Cu5Pd5Na-F
5%Cu-5%Pd/TiO2-5%Na-washed 5Cu5Pd5Na-W
5%Cu-5%Pd/TiO2-5%Na-third reused 5Cu5Pd5Na-R3
Open in a new tab

The details about the preparation and pretreatment of the materials are described in our previous report,1 as well as H2 (H2-TPR), CO and H2 chemisorption, N2O oxidation and H2 titration, HRTEM, XRD, DRS-UV–vis, and EDSA results.

XPS analyses were carried out using a Thermo VG Scientific ESCALAB 250 spectrometer equipped with a hemispherical electron analyzer and an Al Kα radiation source (1486.6 eV) powered at 20 kV and 30 mA, respectively. The binding energy (BE) was determined by using the carbon C (1s) line as the reference with a BE of 284.6 eV. The spectrometer was operated at a pass energy of 23.5 eV, and the base pressure in the analyzing chamber was maintained at 3 × 10–8 mbar. Peak fitting was done by using XPSPEAK 41 with Shirley background. High-resolution core-level scans were acquired from the surface of the freshly fractured catalysts for the Ti 2p, O 1s, Na 1s, Cu 2p, and, Pd 3d photoelectron peaks. Relative concentrations of species were calculated from the integral intensities of the XPS peaks using the cross-sections according to Scofield.

The catalysts also were analyzed by scale Z-contrast imaging, EELS, and EDS techniques using a scanning transmission electron microscope (JEOL 2100F electron microscope at 200 kV and a 0.19 nm resolution). TEM images from bimetallic particles using HRTEM were also recorded. Catalyst samples were deposited on Au grids. EELS and EDS analysis of elements was performed by using a Si (Li) model INCA, Oxford.

Catalytic Tests

Catalytic tests were performed in a batch reactor (Parr Instruments), which was loaded with 50 mL of a 20 wt % aqueous solution of glycerol and 0.3 g of the previously reduced catalyst. The catalysts were transferred from the reduction tube to a reactor while trying to keep them in an inert atmosphere, but a brief exposure to air cannot be excluded. After the reactor was sealed, it was flushed with flowing N2 at 0.1 MPa for 5 min to remove air from the headspace. N2 was subsequently flushed with flowing H2 at 0.1 MPa for 5 min. The reactor was then pressurized to 0.7 MPa with H2 and heated under moderate agitation (100 rpm) to a final reaction temperature of 220 °C. Once this temperature was reached, an initial liquid sample was drawn to mark the start of the reaction, and the rate of agitation was increased to 480 rpm. The reaction was allowed to proceed for 6 h, while the liquid phase was sampled every hour.

Results and Discussion

Activity and Selectivity during Hydrodeoxygenation

In our earlier study, we found that Na-promoted bimetallic catalysts showed higher activity than monometallic catalysts while maintaining high 1,2-PDO selectivity (76–93%) for a types of Cu catalysts. We obtained similar results for the initial glycerol turnover frequency (TOF). The best results were obtained with 5Cu5Pd5Na-F at 220 °C and 0.7 MPa H2. The initial TOF (based on Cu + Pd sites) was 0.14 s–1; the selectivity to 1,2-PDO reached 85%.1 The good activity and selectivity of these catalysts involves a synergistic effect between the two metallic components, that is, the activity of the bimetallic catalyst is larger than the sum of the effects for the two separate monometallic catalysts. The active sites were stable, as evidenced by the 5Cu5Pd5Na-F catalyst that maintained high stable activity and high 1,2-PDO selectivity after being reused three times. Our previous characterization results1 showed that our bimetallic catalysts consisted primarily of CuPd alloy nanoparticles with an average diameter of about 4 nm, which is smaller than the diameters of Cu or Pd nanoparticles in the monometallic samples. Cu is stabilized in the alloy, and it is mainly present on the surface of the nanoparticles.

Therefore, the presence of Na in the bimetallic catalysts was necessary to maintain high activity and selectivity during operation. An analysis of the actual content after use showed that Na was leached from the catalysts, but a stable concentration was reached that helped maintain their activity. The limiting content is 0.5 wt %, and it is very stable. In the previous article,1 we showed that the decrease in activity during the first reuse appears to be caused by the initial lixiviation of Na from the fresh catalyst to the reaction solution, which led to the formation of Na2CO3 on the catalyst. Also, a detailed analysis of the micrographs showed the presence of NaOx and Na2CO3; the latter in bimetallic reused catalysts. Multilayers of NaOx on both TiO2 and CuPd are readily observed on these materials.

Bulk Chemical Composition of Individual Bimetallic Nanoparticles

Composition, distribution, and oxidation states of Cu and Pd within the individual nanoparticles were determined using EDS. About 50 particles from different sections and samples were analyzed for each catalyst (all prepared using the same method). The average nanoparticle size in the analyzed bimetallic samples is in the 5.0–5.7 nm range, regardless of the material. Details of the size and the structure of the nanoparticles in the catalysts are given in the earlier manuscript. The EDS elemental general map of 5Cu5Pd5Na-F and its corresponding Z-contrast image, shown in Figure 1, showed the presence of Cu, Pd, Na, and Ti. The elemental maps consistently indicated that most individual nanoparticles were composed of both Cu and Pd, see Figure 2a,b, that is, bimetallic particles were formed. We also identified Na K signals originating from the support as a result of the initial impregnation with NaOH during synthesis of the catalyst.

Figure 1.

Figure 1

Open in a new tab

EDS elemental maps showing the distribution of Cu, Pd, and Ti in 5Cu5Pd5Na-F.

Figure 2.

Figure 2

Open in a new tab

EDS spectra from bimetallic particles on the Cu5Pd5Na-F catalyst (a,b) from different areas.

In Table 2 we report the approximate percentage of Cu, Pd, or mixed CuPd particles for each catalyst. To estimate these percentages, we based our analysis on the presence of either the Cu K, Pd L, or Cu K + Pd L signals. As can be seen, bimetallic nanoparticles containing both Cu and Pd predominate in all materials, with number percentages larger than 80% in all cases. Consequently, the percentage of Cu- or Pd-only particles was relatively small (≤13%).

Table 2. Percentage Estimate of Cu, Pd, or CuPd Particles in the Samples Studied.

    particle percentage
catalyst number of nanoparticles analyzed Pd-L Cu-K Cu-K and Pd-L
5Cu5Pd5Na-F 50 10.0 4.0 86.0
5Cu5Pd5Na-W 52 12.0 4.0 84.0
5Cu5Pd5Na-R3 48 13.0 5.0 82.0
Open in a new tab

Based on the EDS results, we also estimated the at. % composition of the bimetallic nanoparticles, as reported in Table 3. These values evidence the nonuniform distribution of Cu and Pd in the bimetallic particles, with predominance of Cu for the three catalysts surpassing the global at. % of our samples, 62.6%, that corresponds to the 5 wt % of each of Cu and Pd impregnated during synthesis. The variation in the atomic composition of the bimetallic nanoparticles suggests the existence of different nanostructures on the surface of the catalysts.

Table 3. Estimated Chemical Composition of Individual Nanoparticles in the Catalyst Samples.

  percentage of bimetallic particles (%)
catalyst 0–25 at. % (Cu K) 25–50 at. % (Cu K) 50–75 at. % (Cu K) 75–100 at. % (Cu K)
5Cu5Pd5Na-F 28.0 15.0 14.0 43.0
5Cu5Pd5Na-W 21.0 17.0 35.0 27.0
5Cu5Pd5Na-R3 19.0 31.0 20.0 30.0
Open in a new tab

Structure of the Bimetallic Nanoparticles

To study the structure of the bimetallic nanoparticles in our catalysts, we used representative high-angle annular dark-field (HAADF)/STEM images coupled with an EDS scan along the diameter of the nanoparticles studied in each catalyst (Figure 3a–j). In Figure 3a,f, the Cu and Pd signals show similar intensities along the diameter of the particles, suggesting a nearly homogeneous distribution of both elements (CuPd alloy). In Figure 3b,g, the Cu-K signals are strong at the periphery and weak at the core of the nanoparticles, suggesting Cu surface enrichment, as confirmed by the strong Pd-L signal in the central part of the nanoparticle. This corresponds to a Cu shell/Pd core structure. Another type of structure can be observed in Figure 3c,h, in which there is a core formed by a mixture of Cu and Pd covered mainly by Cu (Cu shell/CuPd core). Other images, shown in Figure 3d,i,e,j, reveal only scanning lines for Pd-L or Cu-K across the entire diameter of the particles, which confirms the presence of some particles consisting only of Cu or Pd.

Figure 3.

Figure 3

Open in a new tab

(a–e) HAADF/STEM representative images and (f–j) STEM/EDS Cu and Pd elemental scan lines.

In summary, our results indicate that the nanoparticles in our CuPd/TiO2 catalysts can adopt five different structures according to how the Cu and Pd atoms coordinate among themselves and with each other: Cu only, Pd only, Pd encapsulated in the core with a Cu shell, a CuPd core covered by Cu, and a homogeneous CuPd alloy (see models in Scheme 1). These structural characteristics have not been reported previously in the literature. We include, as the Supporting Information (Figures S1–S3), additional HAADF-STEM images and STEM/EDS scan lines of the different particle structures. Additionally, Table 4 shows the total number of particles analyzed and the percentage of each configuration for each sample. The results suggest that for the three catalysts, there are a significant number of particles composed of a CuPd alloy (45–61%) and Cu shell/CuPd core (15–23%), and there are also a smaller number of particles formed by Cu on the surface and Pd in the core (10–15%). In addition, the percentage of particles formed only by Pd (4–7%) or Cu (8–13%) is small. The obtained results are consistent with literature reports about the formation of CuPd alloy nanoparticles with their surface enriched by Cu.10−12

Scheme 1. Models Proposed for the Different Configuration of the Particles in the Bimetallic Catalysts.

Scheme 1

Open in a new tab

Table 4. Distribution of Nanoparticle Configurations Determined by HAADF-STEM Images and STEM/EDS Scan Lines for the Different CuPd/TiO2 Catalysts.

    particle configuration
catalyst number of nanoparticles studied Pd Cu CuPd alloy Cu shell/CuPd core Cu shell/Pd core
5Cu5Pd5Na-F 24 6.0 8.0 61.0 15.0 10.0
5Cu5Pd5Na-W 26 7.0 13.0 45.0 18.0 17.0
5Cu5Pd5Na-R3 23 4.0 9.0 52.0 23.0 12.0
Open in a new tab

Surface Chemical Composition of the Bimetallic Nanoparticles

Cu and Pd can originate from ordered Cu3Pd, CuPd, or CuPd3 phases below 900 K or disordered alloys over a wide range of composition at high temperatures.13−22 To determine the surface composition of the alloyed nanoparticles formed in our study, we followed the procedure reported by Heinrichs et al.23 and Venezia et al.24,25 It is relevant to mention that we considered only Pd and Cu atoms present in alloyed particles in the calculation. In the mathematical developments, Pd atoms present in pure Pd particles were not considered. Moreover, in agreement with results presented earlier, it is assumed that Cu is present only in the form of a CuPd alloy. The results, shown in Table 5, reveal strong surface enrichment by Cu of bimetallic CuPd particles in all bimetallic catalysts (76–90%) and therefore, the existence of only a small fraction of Pd on the surface of the nanoparticles. These observations confirm the data obtained by other techniques such as XPS, chemisorption, and CO-Fourier transform infrared spectroscopy.

Table 5. Surface Composition of the Alloyed CuPd Particles in CuPd/TiO2 Catalysts.

catalyst XPd (at. %) ds (nm) DCuPd (%) DPd (%) XPds (at. %) XCus (at. %)
5Cu5Pd5Na-F 0.77 4.5 0.24 5.5 18.0 82.0
5Cu5Pd5Na-W 0.69 5.2 0.20 8.7 29.4 70.6
5Cu5Pd5Na-R3 0.82 4.5 0.24 9.4 33.0 67.0
Open in a new tab

The superficial enrichment of the CuPd bimetallic particles with Cu has been reported by Molenbroek et al.,13 Pintar et al.,14,26 and Edelmann et al.15 They affirm that most of the atoms of Cu in their bimetallic catalysts must be near the Pd atoms. Pintar et al.27 also state that the changes in the Cu extended X-ray absorption fine structure (EXAFS) spectra are more pronounced compared to those for Pd.

Most investigations performed affirm that during the reduction process, Pd–Cu alloys are formed with Cu segregated onto the surface of the nanoparticles,2,13,23,24,27,28 which is related to the lower surface free energy of Cu, the exothermicity of the Pd–Cu alloy formation process, and the strain resulting from the difference in the atomic radii of Cu and Pd. It has also been reported that the degree of surface segregation of Cu depends on the Cu content of the alloy nanoparticles.13,14,25 It has also been found by XPS and XRD that no alloying occurs in alumina-supported Cu–Pd catalysts, suggesting that the metal–support interaction also has a role in the final structure of the bimetallic nanoparticles.2,13,19 These studies indicate the findings of this investigation, however, to be relevant and indicate that the alloying behavior of bimetallic Pd–Cu catalysts is a complex process, and the final material structure may strongly depend on the initial metal loadings and the specific reduction process used to make the alloy.13,22

In Table 5, XPd is the fraction of Pd atoms in the bulk of CuPd alloy particles estimated from XRD; ds is the mean surface diameter of CuPd alloy particles, and Σnidi3/Σnidi2, is estimated from TEM; DCuPd is the overall dispersion of CuPd alloy particles estimated from TEM; DPd (%) is the Pd dispersion estimated from CO chemisorption; XPds is the fraction of Pd atoms present on the surface of CuPd alloy particles estimated from the combination of CO chemisorption, XRD, and TEM results; XCus = 100% – XPds.

Cu and Pd Local Oxidation States in the Bimetallic Nanoparticles

The XPS spectra in the Cu 2p and Pd 3d regions for bimetallic catalysts and their XPS parameters are summarized in the Supporting Information (Figures S4–S5 and Tables S1–S3). The surface chemical compositions calculated from XPS peaks are also described. The three samples exhibit XPS spectra typical of copper with Cu 2p3/2 and Cu 2p1/2 main peaks.29,30 In addition, all the materials exhibit an intense doublet attributed to Pd 3d5/2 and Pd 3d3/2 centering at 334.5 and 339.7 eV.

Interestingly, a comparison between the XPS parameters of the bimetallic catalysts shows that there are no appreciable differences between the materials studied. As the spectra in Figure S4 and Tables S1 and S3 show, in the three catalysts, only zero-valent or partially reduced metal species (Cu0, Cu1+) were identified. In addition, the atomic percentage of Cu0 or Cu1+ surface species is similar to that in the three materials. Therefore, the percentage of surface species of Pd (Figure S5 and Tables S2 and S3) is also similar. Furthermore, a comparison of the XPS spectra and parameters of Cu and Pd of the materials shows minor differences. In addition, the results indicate that the total surface concentration of Pd is around 1.32, 1.28, and 1.26 at. % in the bimetallic catalysts 5Cu5Pd5Na-F, 5Cu5Pd5Na-W, and 5Cu5Pd5Na-R3, respectively, whereas the Cu total concentration is close to 3.82, 3.54,and 3.37 at. % in 5Cu5Pd5Na-F, 5Cu5Pd5Na-W, and 5Cu5Pd5Na-R3, respectively.

The XPS spectra allowed the evaluation of the surface composition of the alloyed particles, but this technique did not provide information about the distribution of the oxidation states of Cu and Pd in the particles.

EELS mapping was performed in different positions of each bimetallic nanoparticle to determine the oxidation states of the metals on the individual bimetallic nanoparticles. About 15 particles from different areas, specimens, and sizes were analyzed. Figure 4 shows Z-contrast image (a) and EEL Cu (b) and Pd (c) core-loss spectra acquired from the sites marked from 1 to 6 in bimetallic particles with sizes in the range of 1–5 nm. EEL spectra 1–6 in Figure 4b consist of Cu-L2,3 (931 eV) edges characteristic of metallic Cu. The same numbers in spectra of Figure 4c, except for the one marked 4, correspond to Pd-M4,5 (459 eV) edges, an indication that most of the Pd is present in the metallic state. This suggests that in the regions marked as 1, 2, 3, 5, and 6, the nanoparticles consist of a CuPd bimetallic alloy. The O–K (533 eV) peak in spectrum 4 in Figure 4c indicates that in some particles, oxidized Pd is combined with metallic Cu.

Figure 4.

Figure 4

Open in a new tab

(a) TEM Z-contrast image of nanoparticles in the range 1–5 nm; (b,c) EEL Cu and Pd core-loss spectra, respectively, acquired from the sites marked from 1 to 6 in (a).

Figure 5 shows the Z-contrast image (a) and EEL Cu (b) and Pd (c) core-loss spectra acquired from the sites marked from 1 to 6 in a bimetallic particle with a diameter close to 3 nm. In this case, the EEL spectra 1–6 in Figure 5b,c correspond to edges characteristic of metallic Cu-L2,3 (931 eV) and Pd-M4,5 (459 eV), respectively.

Figure 5.

Figure 5

Open in a new tab

(a) TEM Z-contrast image of one particle with size around 3 nm; (b) and (c) EEL Cu and Pd core-loss spectra, respectively, acquired from the sites marked from 1 to 6 in (a).

The Z-contrast TEM image and EEL spectra of particles with diameters around 2 nm is reported in Figure 6a–c. In this case, EELS is related to the presence of metallic and partially oxidized Cu, as well as oxidized and metallic Pd.

Figure 6.

Figure 6

Open in a new tab

(a) TEM Z-contrast image of one particle with size around 2 nm; (b,c) EEL Cu and Pd core-loss spectra, respectively, acquired from the specific sites marked from 1 to 4 in (a).

In the case of large-sized nanoparticles (diameters ∼ 8 nm), the results in Figure 7a–c show a representative Z-contrast image of one of the particles analyzed; Figure 7b,c correspond to EEL Cu and Pd core-loss spectra acquired from the specific sites marked from 1 to 4 in the bimetallic particle. A large qualitative difference is observed for the EEL spectra of Cu1+ and Cu2+, meaning that spectra can be used to determine the Cu-oxidation states.17,31 Thus, the shape of O–K edge in the EEL spectra in the region of low loss of CuO and Cu2O are different, for example, the CuO EEL spectrum presents an asymmetric peak and at least three structures marked by arrows in Figure 7b (spectra 2 and 3). Otherwise, the Cu2O EEL spectrum presents a more symmetric peak for the O–K with only a broad less peak (spectra 1 and 4). In addition, EEL spectra 1 and 2 shown in Figure 7c, did not reveal peaks related to Pd, while spectra 3 and 4 revealed the presence of oxidized Pd. Thus, this particle is constituted basically by oxidized and partially oxidized Cu together with oxidized Pd.

Figure 7.

Figure 7

Open in a new tab

(a) TEM Z-contrast image of one particle with sizes around 8 nm; (b,c) EEL Cu and Pd core-loss spectra respectively, acquired from the specific sites marked from 1 to 4 in (a).

Additionally, Figure 8 shows a TEM image representative of another particle with a size around 4 nm. Figure 8a and the EEL spectra for Cu Figure 8b and Pd Figure 8c acquired from specific sites marked with numbers from 1 to 4 in the particle. Similarly, the EEL spectra of the four positions marked in Figure 8c did not reveal peaks related to Pd, whereas the Cu spectra evidenced the presence only of oxidized Cu.

Figure 8.

Figure 8

Open in a new tab

(a) TEM Z-contrast image of one particle with size around 4 nm; (b,c) EEL Cu and Pd core-loss spectra respectively, acquired from the specific sites marked from 1 to 5 in (a).

Interestingly, the results for particles having different diameters showed that the bimetallic nanoparticles have sizes ≤6 nm; the size of the monometallic particles of Pd is in the range of 2–4 nm, while the monometallic Cu particles have sizes greater than 8 nm. This confirms that probably the largest particles are composed only of Cu and the smallest are composed either of metallic Pd or CuPd; however, the percentage of particles formed only by Pd is small compared to that corresponding to bimetallic particles.32

According to the data obtained from the analysis of several particles (Table 6), it can be concluded that nanoparticles with sizes smaller than 7 nm are composed of Cu0–Pd0 and Cu1+–Pd0, whereas nanoparticles with sizes greater than 7 nm consist mostly of Cu2+ and Cu2+–Pd2+. These results indicate that the distribution of oxidation states of Cu and Pd is not uniform and depends on the size of the particles.

Table 6. Total Percentage of Bimetallic Nanoparticles with Different Cu and Pd Oxidation States.

  particle percentage
sizes (nm) Cu1+–Pd0 Cu0–Pd0 Cu2+–Pd0 Cu1+–Pd2+ Cu0–Pd2+ Cu2+–Pd2+ Cu2+ Pd0
<7 21 41 5 6 10 8 0 9
≥7 0 0 15 7 0 19 59 0
Open in a new tab

A word of caution is needed at this point because the oxidation states reported here were determined after the samples were briefly handled in air as they were loaded into the different equipment used for analysis. Full or partial oxidation may occur in some cases. It is true, however, that the observed oxidation state differences point to different activities of the nanoparticles as a function of their size.

Conclusions

The three bimetallic catalysts did not exhibit significant differences in the percentage of Cu, Pd, or CuPd particles, the chemical composition of individual nanoparticles, the distribution of nanoparticle configurations, and the surface composition of the alloyed CuPd particles. On the other hand, all the analyzed catalysts are qualitatively and quantitatively similar in nature and distribution of Cu and Pd species. Additionally, the results showed that the leaching of some of the Na (5Cu5Pd5Na-F and 5Cu5Pd5Na-W) and reusing of the catalysts (5Cu5Pd5Na-R3) did not have any effect on the different configuration of the particles in the bimetallic catalysts.

However, detailed EELS and EDS analyses of the individual nanoparticles showed the formation of CuPd bimetallic nanoparticles with nonuniform surface distribution of both elements, showing regions of oxidized Pd, metallic Cu, and areas of a bimetallic CuPd alloy. The smaller particles are formed mainly of the CuPd alloy, whereas the larger particles are constituted by oxidized Cu and Pd. The nanoparticles adopt five different configurations: mixed CuPd alloy, Cu shell/Pd core, Cu shell/CuPd core, Cu only, and Pd only, with the largest percentage corresponding to bimetallic mixed CuPd alloy nanoparticles.

The nanoparticles with sizes smaller than 7 nm are constituted mainly of Cu0–Pd0 and Cu1+–Pd0, while particles with sizes greater than 7 nm are constituted mostly of Cu2+ and Cu2+–Pd2+. These results indicate that the distribution of oxidation states of Cu and Pd is not uniform and depends on the size of the particles. In summary, characterization of the samples was consistent with the formation of CuPd alloy nanoparticles with their surface enriched with Cu, thus explaining the high selectivity to 1,2-PDO of the bimetallics. The increased activity of the CuPd bimetallic catalyst stems from the significantly smaller average diameter of the CuPd alloy nanoparticles and their good stability. Besides the effect upon the nanoparticle structure, Pd promoted hydrodeoxygenation (HDO) in part by the activation of glycerol aqueous phase reforming to produce H2 at the surface level. Cu was stabilized in the bimetallic samples by the formation of an alloy with Pd in the catalyst.

A combined bulk and surface characterization study was successfully performed to determine the structure/composition/oxidation state profile of individual nanoparticles from CuPd/TiO2–Na bimetallic catalysts with sub-nanometer resolution, an aspect of surface chemistry that had not been elucidated and tested for the active bimetallic nanoparticles on the glycerol HDO to 1,2-propanediol. Thus, our findings will serve as a model system to improve the understanding of the selective conversion of glycerol to 1,2-PDO over specific bimetallic catalysts in Cu.

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Acknowledgments

This research was made possible by the financial support from the Universidad A. Metropolitana,Iztapalapa, México. A.N.A.A. acknowledges the Politécnico Colombiano Jaime Isaza Cadavid, Enlaza Mundos (Colombia) and CONACYT (Mexico) for a graduate fellowship. The STEM and XPS results in this work were obtained at the Nanoscience and Nanotechnology Center, National University, Mexico, and the authors greatly appreciate the support provided

Glossary

Abbreviations

H2-TPR

temperature-programmed reduction with H2

HRTEM

high-resolution electron microscopy

XRD

X-ray diffraction

DRS-UV–vis

diffuse reflectance ultraviolet–visible spectroscopy

EDSA

electron diffraction of selected area

STEM

scanning transmission electron microscopy

EDS

X-ray energy-dispersive spectroscopy

EELS

electron energy-loss spectroscopy

XPS

X-ray photoelectron spectroscopy

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.0c01757.

  • Additional HAADF-STEM images and STEM/EDS scan lines of the different particle structures and XP spectra in the Cu 2p and Pd 3d regions for bimetallic catalysts and their XPS parameters (PDF)

The authors declare no competing financial interest.

Supplementary Material

ao0c01757_si_001.pdf (1MB, pdf)

References

  1. Ardila A. N.; Sánchez-Castillo M. A.; Zepeda T. A.; Villa A. L.; Fuentes G. A. Glycerol hydrodeoxygenation to 1,2-propanediol catalyzed by CuPd/TiO2-Na. Appl. Catal., B 2017, 219, 658–671. 10.1016/j.apcatb.2017.08.006. [DOI] [Google Scholar]
  2. Gopinath R.; Lingaiah N.; Sreedhar B.; Suryanarayana I.; Sai Prasad P. S.; Obuchi A. Highly stable Pd/CeO2 catalyst for hydrodechlorination of chlorobenzene. Appl. Catal., B 2003, 46, 587–594. 10.1016/s0926-3373(03)00321-7. [DOI] [Google Scholar]
  3. Jin X.; Dang L.; Lohrman J.; Subramaniam B.; Ren S.; Chaudhari R. V. Lattice-matched bimetallic CuPd-graphene nanocatalysts for facile conversion of biomass-derived polyols to chemicals. ACS Nano 2013, 7, 1309–1316. 10.1021/nn304820v. [DOI] [PubMed] [Google Scholar]
  4. Tao W.; Wang A.; Yin H. Interaction between Pd and Cu nanoparticles in bimetallic CuPd nanoparticles and its impact on oxidation of 1,2-propanediol to aliphatic acids. Chin. J. Chem. Eng. 2020, 28, 1085–1094. 10.1016/j.cjche.2020.02.005. [DOI] [Google Scholar]
  5. Yamauchi M.; Abe R.; Tsukuda T.; Kato K.; Takata M. Highly selective ammonia synthesis from nitrate with photocatalytically generated hydrogen on CuPd/TiO2. J. Am. Chem. Soc. 2011, 133, 1150–1152. 10.1021/ja106285p. [DOI] [PubMed] [Google Scholar]
  6. Le S. D.; Nishimura S. Highly Selective Synthesis of 1,4-Butanediol via Hydrogenation of Succinic Acid with Supported Cu-Pd Alloy Nanoparticles. ACS Sustainable Chem. Eng. 2019, 7, 18483–18492. 10.1021/acssuschemeng.9b04447. [DOI] [Google Scholar]
  7. Xi Z.; Li J.; Su D.; Muzzio M.; Yu C.; Li Q.; Sun S. Stabilizing CuPd Nanoparticles via CuPd Coupling to WO2.72 Nanorods in Electrochemical Oxidation of Formic Acid. J. Am. Chem. Soc. 2017, 139, 15191–15196. 10.1021/jacs.7b08643. [DOI] [PubMed] [Google Scholar]
  8. Diyarbakir S.; Can H.; Metin Ö. Reduced graphene oxide-supported cupd alloy nanoparticles as efficient catalysts for the sonogashira cross-coupling reactions. ACS Appl. Mater. Interfaces 2015, 7, 3199–3206. 10.1021/am507764u. [DOI] [PubMed] [Google Scholar]
  9. Mun Y.; Lee S.; Cho A.; Kim S.; Han J. W.; Lee J. Cu-Pd alloy nanoparticles as highly selective catalysts for efficient electrochemical reduction of CO2 to CO. Appl. Catal., B 2019, 246, 82–88. 10.1016/j.apcatb.2019.01.021. [DOI] [Google Scholar]
  10. Ma S.; Sadakiyo M.; Heima M.; Luo R.; Haasch R. T.; Gold J. I.; Yamauchi M.; Kenis P. J. A. Electroreduction of carbon dioxide to hydrocarbons using bimetallic Cu-Pd catalysts with different mixing patterns. J. Am. Chem. Soc. 2017, 139, 47–50. 10.1021/jacs.6b10740. [DOI] [PubMed] [Google Scholar]
  11. Chen Z.; He Y.-C.; Chen J.-H.; Fu X.-Z.; Sun R.; Chen Y.-X.; Wong C.-P. PdCu Alloy Flower-like Nanocages with High Electrocatalytic Performance for Methanol Oxidation. J. Phys. Chem. C 2018, 122, 8976–8983. 10.1021/acs.jpcc.8b01095. [DOI] [Google Scholar]
  12. Goulas K. A.; Sreekumar S.; Song Y.; Kharidehal P.; Gunbas G.; Dietrich P. J.; Johnson G. R.; Wang Y. C.; Grippo A. M.; Grabow L. C.; Gokhale A. A.; Toste F. D.; et al. Synergistic effects in bimetallic palladium-copper catalysts improve selectivity in oxygenate coupling reactions. J. Am. Chem. Soc. 2016, 138, 6805–6812. 10.1021/jacs.6b02247. [DOI] [PubMed] [Google Scholar]
  13. Molenbroek A. M.; Haukka S.; Clausen B. S. Alloying in Cu/Pd nanoparticle catalysts. J. Phys. Chem. B 1998, 102, 10680–10689. 10.1021/jp9822081. [DOI] [Google Scholar]
  14. Batista J.; Pintar A.; Gomilšek J. P.; Kodre A.; Bornette F. On the structural characteristics of γ-alumina-supported Pd-Cu bimetallic catalysts. Appl. Catal., A 2001, 217, 55–68. 10.1016/s0926-860x(01)00580-4. [DOI] [Google Scholar]
  15. Edelmann A.; Schießer W.; Vinek H.; Jentys A. Oxidation state of bimetallic PdCu catalysts during liquid phase nitrate reduction. Catal. Lett. 2000, 69, 11–16. 10.1023/a:1019032931079. [DOI] [Google Scholar]
  16. Strukul G.; Pinna F.; Marella M.; Meregalli L.; Tomaselli M. Sol-gel palladium catalysts for nitrate and nitrite removal from drinking water. Catal. Today 1996, 27, 209–214. 10.1016/0920-5861(95)00189-1. [DOI] [Google Scholar]
  17. Sun K.; Liu J.; Nag N. K.; Browning N. D. Atomic scale characterization of supported Pd-Cu/γ-Al2O3 bimetallic catalysts. J. Phys. Chem. B 2002, 106, 12239–12246. 10.1021/jp0265889. [DOI] [Google Scholar]
  18. Fernández-García M.; Anderson J. A.; Haller G. L. Alloy formation and stability in Pd-Cu bimetallic catalysts. J. Phys. Chem. 1996, 100, 16247–16254. 10.1021/jp9608133. [DOI] [Google Scholar]
  19. Renouprez A. J.; Lebas K.; Bergeret G.; Rousset J. L.; Delichère P. Supported Pd-Cu catalysts prepared from bimetallic organo-metallic complexes: relation between surface composition measured by Ion Scattering and reactivity. Stud. Surf. Sci. Catal. 1996, 101, 1105–1114. 10.1016/s0167-2991(96)80322-3. [DOI] [Google Scholar]
  20. Bradley J. S.; Hill E. W.; Chaudret B.; Duteil A. Surface chemistry on colloidal metals. Reversible adsorbate-induced surface composition changes in colloidal palladium-copper alloys. Langmuir 1995, 11, 693–695. 10.1021/la00003a002. [DOI] [Google Scholar]
  21. Bradley J. S.; Via G. H.; Bonneviot L.; Hill E. W. Infrared and EXAFS Study of Compositional Effects in Nanoscale Colloidal Palladium–Copper Alloys. Chem. Mater. 1996, 8, 1895–1903. 10.1021/cm960093c. [DOI] [Google Scholar]
  22. Toshima N.; Wang Y. Preparation and catalysis of novel colloidal dispersions of Copper/noble metal bimetallic clusters. Langmuir 1994, 10, 4574–4580. 10.1021/la00024a031. [DOI] [Google Scholar]
  23. Heinrichs B.; Noville F.; Schoebrechts J.-P.; Pirard J.-P. Palladium-Silver Sol-Gel Catalysts for Selective Hydrodechlorination of 1,2-Dichloroethane into Ethylene. J. Catal. 2000, 192, 108–118. 10.1006/jcat.2000.2816. [DOI] [Google Scholar]
  24. Lambert S.; Heinrichs B.; Brasseur A.; Rulmont A.; Pirard J.-P. Determination of surface composition of alloy nanoparticles and relationships with catalytic activity in Pd-Cu/SiO2 cogelled xerogel catalysts. Appl. Catal., A 2004, 270, 201–208. 10.1016/j.apcata.2004.05.005. [DOI] [Google Scholar]
  25. Venezia A. M.; Liotta L. F.; Deganello G.; Schay Z.; Guczi L. Characterization of Pumice-Supported Ag-Pd and Cu-Pd Bimetallic Catalysts by X-Ray Photoelectron Spectroscopy and X-Ray Diffraction. J. Catal. 1999, 182, 449–455. 10.1006/jcat.1998.2343. [DOI] [Google Scholar]
  26. Dietrich P. J.; Akatay M. C.; Sollberger F. G.; Stach E. A.; Miller J. T.; Delgass W. N.; Ribeiro F. H. Effect of co loading on the activity and selectivity of ptCo aqueous phase reforming catalysts. ACS Catal. 2014, 4, 480–491. 10.1021/cs4008705. [DOI] [Google Scholar]
  27. Pintar A.; Batista J.; Arčon I.; Kodre A. Characterization of γ-Al2O3 supported Pd-Cu bimetallic catalysts by EXAFS, AES and kinetic measurements. Stud. Surf. Sci. Catal. 1998, 118, 127–136. 10.1016/s0167-2991(98)80175-4. [DOI] [Google Scholar]
  28. Popov A. A.; Shubin Y. V.; Plyusnin P. E.; Sharafutdinov M. R.; Korenev S. V. Experimental redetermination of the Cu-Pd phase diagram. J. Alloys Compd. 2019, 777, 204–212. 10.1016/j.jallcom.2018.10.332. [DOI] [Google Scholar]
  29. Corma A.; Palomares A.; Márquez F. Determining the Nature of the Active Sites of Cu-Beta Zeolites for the Selective Catalytic Reduction (SCR) of NOxby Using a Coupled Reaction-XAES/XPS Study. J. Catal. 1997, 170, 132–139. 10.1006/jcat.1997.1739. [DOI] [Google Scholar]
  30. Morales J.; Sánchez L.; Martín F.; Ramos-Barrado J. R.; Sánchez M. Use of low-temperature nanostructured CuO thin films deposited by spray-pyrolysis in lithium cells. Thin Solid Films 2005, 474, 133–140. 10.1016/j.tsf.2004.08.071. [DOI] [Google Scholar]
  31. Fox E. B.; Velu S.; Engelhard M. H.; Chin Y.-H.; Miller J. T.; Kropf J.; Song C. Characterization of CeO2-supported Cu-Pd bimetallic catalyst for the oxygen-assisted water-gas shift reaction. J. Catal. 2008, 260, 358–370. 10.1016/j.jcat.2008.08.018. [DOI] [Google Scholar]
  32. Freitas I. C.; Manfro R. L.; Souza M. M. V. M. Hydrogenolysis of glycerol to propylene glycol in continuous system without hydrogen addition over Cu-Ni catalysts. Appl. Catal., B 2018, 220, 31–41. 10.1016/j.apcatb.2017.08.030. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ao0c01757_si_001.pdf (1MB, pdf)

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES