Palladium Oxide Nanoparticles: Preparation, Characterization and Catalytic Activity Evaluation

Stable palladium oxide nanoparticles were prepared in aqueous suspension with a very simple procedure, by dissolving palladium nitrate in water at a concentration around 10−4 M. UV-visible absorption spectroscopy was adopted to follow the formation of these nanoparticles, which were characterized by TEM microscopy, along with XRD, XPS and Raman measurements. DFT calculations allowed to interpret the Raman data and to clarify the species present at the surface of the nanoparticles. The catalytic activity of the latter was evaluated by monitoring the reduction of p-nitrophenol to p-aminophenol. This investigation paves the way to the use of these colloidal nanoparticles in processes of heterogeneous catalysis, in particular those concerning the catalytic degradation of aromatic derivatives that represent a serious danger for the environment as pollutants, as in the case of p-nitrophenol.


Introduction
Due to the optimal combination of activity and selectivity, palladium is one of the most used transition metals for both homogeneous and heterogeneous catalysis of organic reactions [1,2]. In general, heterogeneous catalytic processes are preferred with respect to homogeneous ones due to the higher yields and easier reusability of the catalytic material. Since heterogeneous catalysis is based on processes taking place at the surface, the use of finely dispersed catalysts is often adopted in order to reduce the amount of material and maximize its catalytic effect [3,4]. For this purpose, palladium nanoparticles are widely used to catalyze organic reactions [5][6][7][8][9][10][11][12][13][14]. Similar to palladium, palladium(II) oxide (PdO) has also found applications in many catalytic processes [15][16][17][18], but few studies were carried out on the production of palladium oxide nanoparticles compared to the wide range of works concerning metallic palladium particles [19][20][21][22]. A problem in obtaining colloidal suspension of PdO nanoparticles derives from the agglomeration of them, so it was often necessary to introduce surfactants or supporting materials that favored their stability. In this work, we propose a simple method to obtain stable palladium oxide nanoparticles dispersed in water, without the need of additives. The characteristics of the nanoparticles have been investigated by different techniques and their catalytic activity has been evaluated by monitoring the reduction reaction of p-nitrophenol to p-aminophenol. slits and detected by an air-cooled CCD (2.5 cm −1 per pixel), filtered by a double holographic Notch filters system. Spectra were calibrated with respect to a silicon wafer at 520 cm −1 .

Density Functional Theory (DFT) Calculations
Density Functional Theory (DFT) calculations were carried out using the GAUSSIAN 09 package [23]. Optimized geometries were obtained at the DFT level of theory, employing the widely adopted Becke 3-parameter hybrid exchange functional (B3) combined with the Lee-Yang-Parr correlation functional (LYP) [24,25], along with the Lanl2DZ basis set and pseudopotential [26][27][28]. The latter was successfully used in the simulation of the Raman spectra of complexes containing heavy metals [29][30][31]. All parameters were let free to relax and all calculations converged toward optimized geometries corresponding to energy minima, as revealed by the lack of negative values in the frequency calculation. A scaling factor of 0.98 for the calculated harmonic wavenumbers was employed, as usually performed in calculations at this level of theory [32][33][34][35][36].

Catalysis Tests
NaBH 4 was added to the aqueous solution of PNP (PNP concentration: 7 µM; NaBH 4 concentration: 3.6 mM). The NaBH 4 /PNP concentration ratio was about 500. The resulting solution was stirred for 30 s and observed by UV-vis absorption measurements for 30 min. Then, after adding a small amount of the PdO colloid (Pd concentration: 5 µM), the same absorption measurements were performed.

Results
We have noted that, after dissolving Pd(NO 3 ) 2 in small amount (around 2 × 10 −4 M concentration) in water at room temperature, the aqueous solution progressively turned toward a marked yellowish color. This evidence can be monitored by the UV-visible absorption spectrum, where a band grows around 265 nm, with a pronounced tail in the visible region ( Figure 1). However, after five hours from the dissolution, no further changes in the absorption spectrum were detectable. The final pH of the solution was~4. The UV-visible absorption band around 265 nm can be confidently attributed to highly dispersed palladium oxide [37,38]. Actually, the existence of nanoparticles in aqueous suspension has also been demonstrated by TEM microscopy, after depositing drops of solution on a carbon-Cu grid (see Figure 2).
Coatings 2020, 10, 207 3 of 10 by a double holographic Notch filters system. Spectra were calibrated with respect to a silicon wafer at 520 cm −1 .

Density Functional Theory (DFT) Calculations
Density Functional Theory (DFT) calculations were carried out using the GAUSSIAN 09 package [23]. Optimized geometries were obtained at the DFT level of theory, employing the widely adopted Becke 3-parameter hybrid exchange functional (B3) combined with the Lee-Yang-Parr correlation functional (LYP) [24,25], along with the Lanl2DZ basis set and pseudopotential [26][27][28]. The latter was successfully used in the simulation of the Raman spectra of complexes containing heavy metals [29][30][31]. All parameters were let free to relax and all calculations converged toward optimized geometries corresponding to energy minima, as revealed by the lack of negative values in the frequency calculation. A scaling factor of 0.98 for the calculated harmonic wavenumbers was employed, as usually performed in calculations at this level of theory [32][33][34][35][36].

Catalysis Tests
NaBH4 was added to the aqueous solution of PNP (PNP concentration: 7 μM; NaBH4 concentration: 3.6 mM). The NaBH4/PNP concentration ratio was about 500. The resulting solution was stirred for 30 s and observed by UV-vis absorption measurements for 30 min. Then, after adding a small amount of the PdO colloid (Pd concentration: 5 μM), the same absorption measurements were performed.

Results
We have noted that, after dissolving Pd(NO3)2 in small amount (around 2 × 10 −4 M concentration) in water at room temperature, the aqueous solution progressively turned toward a marked yellowish color. This evidence can be monitored by the UV-visible absorption spectrum, where a band grows around 265 nm, with a pronounced tail in the visible region ( Figure 1). However, after five hours from the dissolution, no further changes in the absorption spectrum were detectable. The final pH of the solution was ~4. The UV-visible absorption band around 265 nm can be confidently attributed to highly dispersed palladium oxide [37,38]. Actually, the existence of nanoparticles in aqueous suspension has also been demonstrated by TEM microscopy, after depositing drops of solution on a carbon-Cu grid (see Figure 2).  The TEM micrograph shows small nanoparticles (average diameter: 2.2 nm), which are mostly monodispersed and quite similar in their shapes and dimensions. The zeta potential measurements validate the stability of the colloidal dispersion, showing a peak at +35 mV. The TEM micrograph shows small nanoparticles (average diameter: 2.2 nm), which are mostly monodispersed and quite similar in their shapes and dimensions. The zeta potential measurements validate the stability of the colloidal dispersion, showing a peak at +35 mV.
A mechanism of formation of the palladium oxide nanoparticles can be proposed (Scheme 1). When Pd(NO3)2 is dissolved in aqueous solution, insoluble PdO is spontaneously produced in the water environment as small nanoparticles. The first step is the hydration of the palladium cation with the formation of Pd(II) tetrahydrate [39][40][41]. This aquacomplex, through the formation of palladium hydroxide, can give rise to PdO, which is the stoichiometry corresponding to the most stable form of palladium oxide. Scheme 1. Proposed mechanism of formation of palladium oxide nanoparticles.
In order to find a confirmation to this hypothesis and to understand the nature of the metal nanoparticles, we performed XRD, XPS and Raman measurements. By ultra-centrifuging a portion of the aqueous suspension, a dark deposit was obtained, rinsed with pure water and deposited on a silicon substrate for the XRD experiment ( Figure 3) and on a glass substrate for both XPS ( Figure 4) and micro-Raman measurements ( Figure 5). A mechanism of formation of the palladium oxide nanoparticles can be proposed (Scheme 1). When Pd(NO 3 ) 2 is dissolved in aqueous solution, insoluble PdO is spontaneously produced in the water environment as small nanoparticles. The first step is the hydration of the palladium cation with the formation of Pd(II) tetrahydrate [39][40][41]. This aquacomplex, through the formation of palladium hydroxide, can give rise to PdO, which is the stoichiometry corresponding to the most stable form of palladium oxide. The TEM micrograph shows small nanoparticles (average diameter: 2.2 nm), which are mostly monodispersed and quite similar in their shapes and dimensions. The zeta potential measurements validate the stability of the colloidal dispersion, showing a peak at +35 mV.
A mechanism of formation of the palladium oxide nanoparticles can be proposed (Scheme 1). When Pd(NO3)2 is dissolved in aqueous solution, insoluble PdO is spontaneously produced in the water environment as small nanoparticles. The first step is the hydration of the palladium cation with the formation of Pd(II) tetrahydrate [39][40][41]. This aquacomplex, through the formation of palladium hydroxide, can give rise to PdO, which is the stoichiometry corresponding to the most stable form of palladium oxide. Scheme 1. Proposed mechanism of formation of palladium oxide nanoparticles.
In order to find a confirmation to this hypothesis and to understand the nature of the metal nanoparticles, we performed XRD, XPS and Raman measurements. By ultra-centrifuging a portion of the aqueous suspension, a dark deposit was obtained, rinsed with pure water and deposited on a silicon substrate for the XRD experiment ( Figure 3) and on a glass substrate for both XPS ( Figure 4) and micro-Raman measurements ( Figure 5).

Scheme 1.
Proposed mechanism of formation of palladium oxide nanoparticles.
In order to find a confirmation to this hypothesis and to understand the nature of the metal nanoparticles, we performed XRD, XPS and Raman measurements. By ultra-centrifuging a portion of the aqueous suspension, a dark deposit was obtained, rinsed with pure water and deposited on a silicon substrate for the XRD experiment ( Figure 3) and on a glass substrate for both XPS ( Figure 4) and micro-Raman measurements ( Figure 5).           Figure 3 shows the experimental diffraction pattern (black line), along with the reference diffraction patterns of PdO (blue segments) and Pd(NO 3 ) 2 (red segments), corresponding to the Power Diffraction Files (PDF)s of the XRD database. By comparing the experimental pattern with the two reference patterns, it is possible to note that the experimental data are almost completely attributable to palladium oxide and just a small amount of palladium nitrate can be recognized. It is probably due to trace of the reactant solution. On the other side, a more thorough water cleaning is not recommended in order to avoid the alteration or aggregation of the nanoparticles. Moreover, the pronounced width of the peaks in the experimental diffraction pattern can be ascribable to the nature of the analyzed solid, as a secondary proof of the nanoparticle presence.
The XPS spectrum of the Pd 3d transition region is reported in Figure 4. Due to spin-orbit coupling, this transition results in two peaks: 3d 5/2 and 3d 3/2 at about 335 eV and 340 eV, respectively. Referring to the 3d 5/2 peak, it can be fit with two components: the first one predominant at 337.4 eV and the second one minority at 339.0 eV. The peak at 337.4 eV can be confidently attributed to PdO, according to our previous investigations on palladium surfaces [42], even if an upshift is observed with respect to the value of 336.9 eV. This shift, generally shown by nanoparticles with respect to the bulk [43], can be explained on the basis of a quantic confinement. The other peak at 339.0 eV could be attributed to Pd(OH) 2 , in analogy with that observed at 338.4 eV [42], or to a species of hydrate palladium oxide. Raman measurements, along with DFT calculations, can provide useful information in this regard. Figure 5 shows two Raman spectra of the deposited colloidal nanoparticles, representative of different points of the sample, obtained with the micro-Raman apparatus. The intense band at 650 cm −1 , observed in both spectra, is attributable to PdO, according to the literature [44,45], whereas the shoulder around 625 cm −1 , with variable intensity depending on the irradiated points of the sample, could be attributed to the minority species observed in the XPS spectrum. This secondary band can be correctly assigned with the help of DFT calculations, which allow for collecting information on the species observed in the Raman spectra.
As shown in Figure 6, for the PdO molecule, a band at 649 cm −1 is predicted by the DFT calculations, fitting quite well with the Raman band observed at 650 cm −1 , whereas for Pd(OH) 2 , the Raman band occurring at 625 cm −1 is not reproduced, because the symmetric and antisymmetric stretching modes of the Pd-O bonds are predicted at 567 and 517 cm −1 , respectively. Instead, considering palladium oxide coordinated to one water molecule, we calculated two bands at 625 and 647 cm −1 , which fit well with the bands observed in the Raman spectrum, which can be attributed to the PdO stretching mode mixed with the H bending mode of the water molecule. pronounced width of the peaks in the experimental diffraction pattern can be ascribable to the nature of the analyzed solid, as a secondary proof of the nanoparticle presence.
The XPS spectrum of the Pd 3d transition region is reported in Figure 4. Due to spin-orbit coupling, this transition results in two peaks: 3d5/2 and 3d3/2 at about 335 eV and 340 eV, respectively. Referring to the 3d5/2 peak, it can be fit with two components: the first one predominant at 337.4 eV and the second one minority at 339.0 eV. The peak at 337.4 eV can be confidently attributed to PdO, according to our previous investigations on palladium surfaces [42], even if an upshift is observed with respect to the value of 336.9 eV. This shift, generally shown by nanoparticles with respect to the bulk [43], can be explained on the basis of a quantic confinement. The other peak at 339.0 eV could be attributed to Pd(OH)2, in analogy with that observed at 338.4 eV [42], or to a species of hydrate palladium oxide. Raman measurements, along with DFT calculations, can provide useful information in this regard. Figure 5 shows two Raman spectra of the deposited colloidal nanoparticles, representative of different points of the sample, obtained with the micro-Raman apparatus. The intense band at 650 cm −1 , observed in both spectra, is attributable to PdO, according to the literature [44,45], whereas the shoulder around 625 cm −1 , with variable intensity depending on the irradiated points of the sample, could be attributed to the minority species observed in the XPS spectrum. This secondary band can be correctly assigned with the help of DFT calculations, which allow for collecting information on the species observed in the Raman spectra.
As shown in Figure 6, for the PdO molecule, a band at 649 cm −1 is predicted by the DFT calculations, fitting quite well with the Raman band observed at 650 cm −1 , whereas for Pd(OH)2, the Raman band occurring at 625 cm −1 is not reproduced, because the symmetric and antisymmetric stretching modes of the Pd-O bonds are predicted at 567 and 517 cm −1 , respectively. Instead, considering palladium oxide coordinated to one water molecule, we calculated two bands at 625 and 647 cm −1 , which fit well with the bands observed in the Raman spectrum, which can be attributed to the PdO stretching mode mixed with the H bending mode of the water molecule.   Finally, we have evaluated the catalytic activity of our PdO nanoparticles by monitoring the reduction of p-nitrophenol (PNP) in borohydride solution, which is only possible in the presence of an appropriate catalyst [46] and represents a classical test for its efficiency [47][48][49][50][51]. By adding NaBH 4 , the aqueous solution of PNP becomes alkaline and the absorption band of p-nitrophenolate occurs at 400 nm in the absorption spectrum, which does not appreciably decrease in time. When, instead, a small quantity of the colloidal suspension of PdO nanoparticle is added to the aqueous solution, the peak at 400 nm decreases exponentially, while the absorption peak of p-aminophenol grows around 300 nm, as shown in Figure 7. This evidence points to a catalytic reduction of PNP by the effect of the palladium oxide nanoparticles added to the aqueous solution.
the aqueous solution of PNP becomes alkaline and the absorption band of p-nitrophenolate occurs at 400 nm in the absorption spectrum, which does not appreciably decrease in time. When, instead, a small quantity of the colloidal suspension of PdO nanoparticle is added to the aqueous solution, the peak at 400 nm decreases exponentially, while the absorption peak of p-aminophenol grows around 300 nm, as shown in Figure 7. This evidence points to a catalytic reduction of PNP by the effect of the palladium oxide nanoparticles added to the aqueous solution.

Conclusions
We proposed a very simple method to obtain stable palladium oxide nanoparticles dispersed in water, without the aid of additional material, by dissolving palladium nitrate in water at a concentration around 10 -4 M. TEM microscopy showed the presence of very small nanoparticles, mostly monodisperse, whose colloidal stability was also proven by zeta-potential measurements. The particle characterization, performed with different techniques, such as UV absorption, XRD diffraction, XPS and Raman spectroscopy, along with DFT calculations, allowed both to validate the mechanism of formation of the nanoparticles and to identify their chemical composition. The colloidal nanoparticles are mostly constituted by Pd(II) oxide, along with a minority species formed by PdO strongly coordinated to water. These nanoparticles showed a marked catalytic activity, evaluated by monitoring the reduction of p-nitrophenol to p-aminophenol. Hence, they can be usefully employed in processes of heterogeneous catalysis, in particular those concerning the catalytic degradation of aromatic derivatives that represent a serious danger for the environment as pollutants, as in the case of p-nitrophenol.