Pt Monolayers on Electrodeposited Nanoparticles of Different Compositions for Ammonia Electro-Oxidation

Pt monolayers (PtML) supported on nanoparticles with different compositions (i.e., Ru, Rh, Pd, Ir, and Au) were synthesized by the surface–limited redox replacement of underpotentially deposited Cu monolayers on nanoparticle supports. Nanoparticle supports with different compositions were directly deposited on the conducting substrate by a clean and one-step electrodeposition method with controlled deposition potential and time. The whole synthesis process of the electrode was free of surfactants, binders, capping agents and reductants, and without an additional coating process of electrocatalysts. The results show that the specific activity (SA) of PtML electrocatalysts depended strongly on the composition of the nanoparticle support. For example, the PtML supported on the Au nanoparticle exhibited 8.3 times higher SA than that supported on the Ru and Pd nanoparticles. The change in the SA of the PtML supported on different nanoparticles was related to the substrate–induced strain in the PtML resulting from the lattice mismatch between the PtML and the nanoparticle support. As the strain in the PtML changed from the tensile strain to the compressive strain, the SA of the PtML electrocatalysts decreased remarkably.


Introduction
Growing concern in energy and environmental issues has stimulated considerable research on the electro-oxidation of ammonia, since it addresses both clean energy supply and removal of pollutants [1][2][3][4][5][6].Ammonia is a carbon-free chemical energy carrier, which has a high capacity of hydrogen storage (17.7 wt.%) and high energy density (3000 Wh kg −1 ) [1,7].On the other hand, ammonia is an environmental pollutant and a toxic gas.Therefore, the electro-oxidation of ammonia has important applications in direct fuel cells [8][9][10], hydrogen production [1,11,12], ammonia decomposition in wastewaters [8,13] and electrochemical sensors for detecting ammonia [14].Since ammonia electro-oxidation is a sluggish reaction, electrocatalysts are required to catalyze the oxidation of ammonia to N 2 .Among various investigated electrocatalysts, Pt has been acknowledged as the most effective single-component electrocatalyst for ammonia electro-oxidation [15,16].However, it is a limited resource and has high cost which been the main obstacles for wide usage of Pt [17][18][19][20], strongly limiting the large-scale applications of ammonia electro-oxidation technologies.Therefore, great efforts have been paid to the development of Pt-based electrocatalysts with a high activity for ammonia electro-oxidation and simultaneously low Pt loading.
Developing Pt ML catalysts is one of the most promising strategies because of its unique advantages, such as an extremely high utilization degree of Pt atoms and tunable activity through the electronic and structural effects between the Pt ML and the substrate [15][16][17][18].Adzic's group successfully developed a class of electrocatalysts consisting of a Pt ML on different metals (e.g., Au, Pd, Rh, Ir, and Ru) [15,19,20], alloys (e.g., PdCo [21], PdFe [22], PdAu [23], IrNi [24], and AuNiFe [25]).Pt ML electrocatalysts can be synthesized by deposition of a Cu monolayer on the substrate through underpotential deposition (UPD), followed by a surface-limited replacement reaction between Pt ions and Cu monolayer [26,27].It has been well demonstrated that a Cu monolayer can be replaced by a Pt ML [26].Due to the high Pt utilization, the mass activity of these Pt ML electrocatalysts could be more than an order of magnitude higher than traditional all-Pt electrocatalysts [15,18,27].Since the catalytic activity of the Pt ML electrocatalysts is strongly dependent on the interaction between the Pt ML and the supporting substrate [18,19,28,29], it is of great interest to investigate the effect of the supporting substrate on the catalytic activity of electrocatalysts.Such studies are important for further enhancing the electrocatalytic activity of Pt ML electrocatalysts by rationally selecting or designing substrate materials.To date, a number of characteristics of substrate materials, such as composition [19,30,31], crystalline orientation [20,21,28,31] and particle size [17,21], have been studied as Pt ML electrocatalysts for oxygen reduction reactions and methanol electro-oxidation.However, there has so far been very limited work on Pt ML electrocatalysts for ammonia electro-oxidation [18,32].Previous work from our group synthesized the Pt ML on an Au substrate (bulk Au electrode [32] or Au particles [18]) for ammonia electro-oxidation.These Pt ML electrocatalysts exhibited several times, to a more than 10-fold increase, in the mass activity compared with the all-Pt electrocatalysts for ammonia electro-oxidation.It has also been found that the surface morphology of substrate material greatly influences the SA of Pt ML electrocatalysts [18].Nevertheless, the effect of the composition of the substrate on the catalytic activity of Pt ML nanoparticles for ammonia electro-oxidation has remained unclear.
The purpose of the present work is to illustrate whether and how the composition of the substrate material would have an influence on the activity of the surface Pt ML nanoparticles.For practical electrocatalytic applications, it is required to support Pt ML on high surface area nanoparticles.Therefore, we investigated the activity of the Pt ML on nanoparticles with various compositions including Ru, Rh, Pd, Ir and Au.A clean, facile and one-step electrodeposition method was proposed to synthesize nanoparticles with different compositions as the supporting core, synthesized directly on the surface of conducting substrate.Pt ML was synthesized on the surface of these nanoparticles by a well-established method, consisting of Cu UPD on nanoparticles and subsequently the galvanic replacement of Pt [15,33].The entire synthesis process was free of any surfactants, binders, capping agents and reductants, ensuring an extremely clean surface of electrocatalysts.In addition, the electrocatalysts were synthesized directly on the conducting substrate surface, avoiding the transfer process of the electrocatalysts.This greatly decreases the uncertainty resulting from the synthesis procedures, allowing the truly comparison of the impact of the composition of the supporting core on the electrocatalytic activity of the supported Pt ML .Furthermore, the relationship between the composition of the nanoparticle core of Pt ML electrocatalysts and the electrocatalytic activity for the ammonia oxidation was investigated by cyclic voltammetry (CV).To the best of our knowledge, this is the first time to report the synthesis of Pt ML electrocatalysts with nanoparticle core of different compositions by the electrochemical method as well as the compositional effect of the core materials on the activity of Pt ML electrocatalysts for electro-oxidation of ammonia.Interestingly, this study shows that the composition of the nanoparticle core had a great influence on the intrinsic activity of Pt ML electrocatalysts.

Results and Discussion
Figure 1a-e shows the SEM images of electrodeposited Ru, Rh, Pd, Ir, and Au nanoparticles, respectively.It can be seen that all these deposited nanoparticles were randomly distributed over the surface of the glassy carbon electrode (GCE).Due to the different nature of these five metals, it is very difficult to obtain different kinds of nanoparticles with exactly the same particle size.As mentioned in the experimental part, preliminary work has tried various electrodeposition parameters including potentials and times, and finally found the optimal conditions to synthesize these five kinds of nanoparticles with a similar particle size.The particle size distribution histograms in insets of Figure 1 show that the average particle sizes for electrodeposited Ru, Rh, Pd, Ir, and Au nanoparticles were 34.2 ± 7.9 nm, 15.2 ± 3.2 nm, 18.8 ± 1.7 nm, 24.2 ± 2.9 nm, 21.4 ± 5.1 nm, respectively.It was also noticed that much larger overpotential (i.e., more negative deposition potential) was required for the electrodeposition of Ru and Ir compared to other metals.Similar results have also been reported by previous studies for the electrodeposition of Ru [34] and Ir [35], due to the slow kinetics for their deposition.For example, Le Vot et al. [35] reported that the deposition of Ir on GCE required a large overpotential for the reduction of Ir 3+ ions in 1.0 mM IrCl 3 + 0.5 M H 2 SO 4 solution.In their work, the lowest deposition potential was −0.6 V (vs.Ag/AgCl).The Ir deposits electrodeposited at this potential were characterized by large aggregates at several micrometers in size, indicating that the deposition process was very slow, and Ir tended to be deposited on previously formed Ir deposits than on the carbon substrate [35].However, in the present work, Ir nanoparticles with good dispersion and small particle size of 24.2 ± 2.9 nm could be obtained at a further decreasing deposition potential of −0.7 V (vs.reversible hydrogen electrode (RHE)) in 5 mM IrCl 3 + 0.05 M H 2 SO 4 solution (Figure 1d).Such well-dispersed Ir nanoparticles are beneficial for increasing the effective surface area.
composition of the nanoparticle core had a great influence on the intrinsic activity of PtML electrocatalysts.

Results and Discussion
Figure 1a-e shows the SEM images of electrodeposited Ru, Rh, Pd, Ir, and Au nanoparticles, respectively.It can be seen that all these deposited nanoparticles were randomly distributed over the surface of the glassy carbon electrode (GCE).Due to the different nature of these five metals, it is very difficult to obtain different kinds of nanoparticles with exactly the same particle size.As mentioned in the experimental part, preliminary work has tried various electrodeposition parameters including potentials and times, and finally found the optimal conditions to synthesize these five kinds of nanoparticles with a similar particle size.The particle size distribution histograms in insets of Figure 1 show that the average particle sizes for electrodeposited Ru, Rh, Pd, Ir, and Au nanoparticles were 34.2 ± 7.9 nm, 15.2 ± 3.2 nm, 18.8 ± 1.7 nm, 24.2 ± 2.9 nm, 21.4 ± 5.1 nm, respectively.It was also noticed that much larger overpotential (i.e., more negative deposition potential) was required for the electrodeposition of Ru and Ir compared to other metals.Similar results have also been reported by previous studies for the electrodeposition of Ru [34] and Ir [35], due to the slow kinetics for their deposition.For example, Le Vot et al. [35] reported that the deposition of Ir on GCE required a large overpotential for the reduction of Ir 3+ ions in 1.0 mM IrCl3 + 0.5 M H2SO4 solution.In their work, the lowest deposition potential was −0.6 V (vs.Ag/AgCl).The Ir deposits electrodeposited at this potential were characterized by large aggregates at several micrometers in size, indicating that the deposition process was very slow, and Ir tended to be deposited on previously formed Ir deposits than on the carbon substrate [35].However, in the present work, Ir nanoparticles with good dispersion and small particle size of 24.2 ± 2.9 nm could be obtained at a further decreasing deposition potential of −0.7 V (vs.reversible hydrogen electrode (RHE)) in 5 mM IrCl3 + 0.05 M H2SO4 solution (Figure 1d).Such well-dispersed Ir nanoparticles are beneficial for increasing the effective surface area.Figure 2a-e shows the cyclic voltammetry (CV) curves of pure Ru, Rh, Pd, Ir, and Au nanoparticles electrodeposited on the GCE in 0.5 M H 2 SO 4 solution at a scan rate of 0.05 V s −1 .Rh, Ir and Pd nanoparticles showed characteristic hydrogen adsorption and desorption (0.05 to 0.2 V (vs.RHE)) peaks while no such hydrogen adsorption/desorption peaks could be observed for Au and Ru nanoparticles.All these CV curves exhibited very similar features compared to previously reported nanoparticles synthesized by a water-in-oil microemulsion method [36].
Catalysts 2018, 8, x FOR PEER REVIEW 4 of 15 Figure 2a-e shows the cyclic voltammetry (CV) curves of pure Ru, Rh, Pd, Ir, and Au nanoparticles electrodeposited on the GCE in 0.5 M H2SO4 solution at a scan rate of 0.05 V s −1 .Rh, Ir and Pd nanoparticles showed characteristic hydrogen adsorption and desorption (0.05 to 0.2 V (vs.RHE)) peaks while no such hydrogen adsorption/desorption peaks could be observed for Au and Ru nanoparticles.All these CV curves exhibited very similar features compared to previously reported nanoparticles synthesized by a water-in-oil microemulsion method [36].3d).Such an oxidation current peak has been extensively reported on pure Ir and also Pt electrodes and is ascribed to the electro-oxidation of ammonia to N2 [12,13,37].Previous work performed CV measurements in the absence of ammonia and found that a such characteristic current peak did not appear [13,37,38].For 4d transition metals (i.e., Ru, Rh and Pd), all CV plots showed no anodic oxidation current peak related with the ammonia electro-oxidation from 0.4 to 0.7 V (vs.RHE).This is in good agreement with previous results reported by de Vooys et al. [13] and Vidal-Iglesias et al. [36].Based on differential electrochemical mass spectroscopy (DEMS) measurements, de Vooys et al. [13] found the dehydrogenation of ammonia occurred at significantly lower potentials on Ru, Rh and Pd than that on Pt and Ir, resulting in the surface poisoning by adsorbed nitrogen atoms (Nads) at much lower potentials on Ru, Rh and Pd.As a result, Ru, Rh and Pd were inactive towards the ammonia electro-oxidation to N2 [13].It was also observed that the peak height at 0.2 V (vs.RHE) for Rh nanoparticles (Figure 3b) is remarkably higher compared to other metals.This agrees well with previous studies reported by Cooper et al. [39], which attributed such peak to the OH − adsorption by Rh. Figure 3a-e shows CVs measured on pure Ru, Rh, Pd, Ir, and Au nanoparticles electrodeposited on the GCE in 0.1 M ammonia + 1 M KOH aqueous solution at 0.05 V s −1 .Only for Ir nanoparticles did the CV curve exhibited an anodic current peak at 0.59 V (vs.RHE) (Figure 3d).Such an oxidation current peak has been extensively reported on pure Ir and also Pt electrodes and is ascribed to the electro-oxidation of ammonia to N 2 [12,13,37].Previous work performed CV measurements in the absence of ammonia and found that a such characteristic current peak did not appear [13,37,38].For 4d transition metals (i.e., Ru, Rh and Pd), all CV plots showed no anodic oxidation current peak related with the ammonia electro-oxidation from 0.4 to 0.7 V (vs.RHE).This is in good agreement with previous results reported by de Vooys et al. [13] and Vidal-Iglesias et al. [36].Based on differential electrochemical mass spectroscopy (DEMS) measurements, de Vooys et al. [13] found the dehydrogenation of ammonia occurred at significantly lower potentials on Ru, Rh and Pd than that on Pt and Ir, resulting in the surface poisoning by adsorbed nitrogen atoms (N ads ) at much lower potentials on Ru, Rh and Pd.As a result, Ru, Rh and Pd were inactive towards the ammonia electro-oxidation to N 2 [13].It was also observed that the peak height at 0.2 V (vs.RHE) for Rh nanoparticles (Figure 3b) is remarkably higher compared to other metals.This agrees well with previous studies reported by Cooper et al. [39], which attributed such peak to the OH − adsorption by Rh. Figure 4a-e shows anodic stripping voltammograms of Cu UPD adlayer on synthesized Ru, Rh, Pd, Ir and Au nanoparticles on the GCE, respectively.All these curves showed typical Cu UPD stripping peaks.The amount and the corresponding surface area of the UPD Cu layer could be estimated from the stripping charge of the UPD Cu, as calculated by integrating the area covered by the Cu stripping current peak after subtracting the background [33,40,41].To achieve the formation of PtML on those metal nanoparticles, the GCE was immersed into 5 mM K2PtCl4 and 0.05 M H2SO4 solution for the galvanic replacement of the surface Cu monolayer by Pt ions.Assuming that the Faraday efficiency of the replacement reaction between the UPD Cu and Pt ions is 100%, the amount and the corresponding surface area of the formed Pt can be calculated based on the amount of UPD Cu monolayer.The corresponding surface area of Pt can also be measured by the stripping charge of Cu UPD monolayer, assuming 480 μC cm −2 for PtML [19,42].The calculated data of the Cu stripping charge, Cu amount, Pt amount and Pt surface area are summarized in Table 1. Figure 4a-e shows anodic stripping voltammograms of Cu UPD adlayer on synthesized Ru, Rh, Pd, Ir and Au nanoparticles on the GCE, respectively.All these curves showed typical Cu UPD stripping peaks.The amount and the corresponding surface area of the UPD Cu layer could be estimated from the stripping charge of the UPD Cu, as calculated by integrating the area covered by the Cu stripping current peak after subtracting the background [33,40,41].To achieve the formation of Pt ML on those metal nanoparticles, the GCE was immersed into 5 mM K 2 PtCl 4 and 0.05 M H 2 SO 4 solution for the galvanic replacement of the surface Cu monolayer by Pt ions.Assuming that the Faraday efficiency of the replacement reaction between the UPD Cu and Pt ions is 100%, the amount and the corresponding surface area of the formed Pt can be calculated based on the amount of UPD Cu monolayer.The corresponding surface area of Pt can also be measured by the stripping charge of Cu UPD monolayer, assuming 480 µC cm −2 for Pt ML [19,42].The calculated data of the Cu stripping charge, Cu amount, Pt amount and Pt surface area are summarized in Table 1.  Figure 5a-e shows the transmission electron microscopy (TEM) images of PtML nanoparticles with Ru, Rh, Pd, Ir, and Au cores, respectively.As a result of TEM sample synthesis, which is required to strip nanoparticles from GCE and then to transfer them to the Cu grid, it is unavoidable to result in the agglomeration of nanoparticles.Therefore, it could be seen that PtML nanoparticles with different composition cores were not dispersed uniformly, showing some extent of agglomeration (Figure 5) compared to freshly electrodeposited core nanoparticles (Figure 1).Nevertheless, it is clear that the overall structure of a single PtML nanoparticle with different composition cores consisted of the smaller, interconnected nanoparticles, suggesting a large surface area.Obviously, the PtML nanoparticles with different composition cores observed in SEM images (Figure 1) were composed of interconnected smaller nanoparticles.The insets of Figure 5 show the corresponding small nanoparticle size distribution histograms.It could be seen that the average nanoparticle sizes of the small PtML nanoparticles with Ru, Rh, Pd, Ir, and Au cores were approximately 4.1 nm, 5.0 nm, 2.6 nm, 4.9 nm and 4.4 nm respectively (the insets of Figure 5), and all of which had narrow nanoparticle size distribution.Figure 5a-e shows the transmission electron microscopy (TEM) images of Pt ML nanoparticles with Ru, Rh, Pd, Ir, and Au cores, respectively.As a result of TEM sample synthesis, which is required to strip nanoparticles from GCE and then to transfer them to the Cu grid, it is unavoidable to result in the agglomeration of nanoparticles.Therefore, it could be seen that Pt ML nanoparticles with different composition cores were not dispersed uniformly, showing some extent of agglomeration (Figure 5) compared to freshly electrodeposited core nanoparticles (Figure 1).Nevertheless, it is clear that the overall structure of a single Pt ML nanoparticle with different composition cores consisted of the smaller, interconnected nanoparticles, suggesting a large surface area.Obviously, the Pt ML nanoparticles with different composition cores observed in SEM images (Figure 1) were composed of interconnected smaller nanoparticles.The insets of Figure 5 show the corresponding small nanoparticle size distribution histograms.It could be seen that the average nanoparticle sizes of the small Pt ML nanoparticles with Ru, Rh, Pd, Ir, and Au cores were approximately 4.1 nm, 5.0 nm, 2.6 nm, 4.9 nm and 4.4 nm respectively (the insets of Figure 5), and all of which had narrow nanoparticle size distribution.Figure 6a-e shows CVs measured on PtML nanoparticles with different core compositions in 0.5 M H2SO4 solution at a scan rate of 0.05 V s −1 .Compared to Figure 2, a significant change in the electrochemical behavior could be observed after applying a PtML on nanoparticles.All CV profiles had typical voltammetric characteristics of Pt in H2SO4 aqueous solution, showing characteristic potential regions including the hydrogen adsorption/desorption region from 0.05 to 0.3 V(vs.RHE) and the double-layer region from 0.3 to 0.85 V(vs.RHE) [43].It has been extensively reported that the hydrogen desorption profile on pure polycrystalline Pt is characterized by two well-separated anodic peaks [5,36,44,45].However, PtML nanoparticles showed only one relatively broad anodic peak in the hydrogen desorption region.Similar results have also been widely reported by previous studies, which is attributed to the interaction between the PtML and the substrate [32,[46][47][48][49].The above results provide clear evidence of the formation of the PtML on these nanoparticles.Generally, the electrochemically active surface area of pure Pt can be estimated from the CV based on the charge related with the hydrogen desorption [50][51][52][53].However, in the case of PtML nanoparticles, the hydrogen desorption profile would be affected due to the interaction between underlying core particles and the PtML, therefore it is possibly not accurate to calculate the effective surface area of Pt.Therefore, previous studies, including the present work, used the amount of Cu monolayer to estimate the Pt surface area assuming the complete replacement of Cu by Pt (Table 1).Figure 6a-e shows CVs measured on Pt ML nanoparticles with different core compositions in 0.5 M H 2 SO 4 solution at a scan rate of 0.05 V s −1 .Compared to Figure 2, a significant change in the electrochemical behavior could be observed after applying a Pt ML on nanoparticles.All CV profiles had typical voltammetric characteristics of Pt in H 2 SO 4 aqueous solution, showing characteristic potential regions including the hydrogen adsorption/desorption region from 0.05 to 0.3 V(vs.RHE) and the double-layer region from 0.3 to 0.85 V(vs.RHE) [43].It has been extensively reported that the hydrogen desorption profile on pure polycrystalline Pt is characterized by two well-separated anodic peaks [5,36,44,45].However, Pt ML nanoparticles showed only one relatively broad anodic peak in the hydrogen desorption region.Similar results have also been widely reported by previous studies, which is attributed to the interaction between the Pt ML and the substrate [32,[46][47][48][49].The above results provide clear evidence of the formation of the Pt ML on these nanoparticles.Generally, the electrochemically active surface area of pure Pt can be estimated from the CV based on the charge related with the hydrogen desorption [50][51][52][53].However, in the case of Pt ML nanoparticles, the hydrogen desorption profile would be affected due to the interaction between underlying core particles and the Pt ML , therefore it is possibly not accurate to calculate the effective surface area of Pt.Therefore, previous studies, including the present work, used the amount of Cu monolayer to estimate the Pt surface area assuming the complete replacement of Cu by Pt (Table 1).To investigate the electrocatalytic activity for the electro-oxidation of ammonia, Figure 7a shows the CVs measured on five different PtML covered nanoparticles in 0.1 M ammonia + 1 M KOH.In order to reveal the SA, the current was normalized by the effective electrochemical surface area.Compared to observed CVs on bare nanoparticles (Figure 3), applying a PtML on these nanoparticles resulted in an obvious change of CV curves.All five CV curves showed a current peak around 0.7 V (vs.RHE) on the positive sweep, which is a typical feature of ammonia electro-oxidation on Pt as previously discussed.This again confirms the successful deposition of Pt on various nanoparticles by the replacement of UPD Cu.It is worth noticing that the SA of PtML covered nanoparticles strongly depends on the composition of core nanoparticles, which can be clearly seen in Figure 7b.The SA for electro-oxidation of ammonia decreases in the following order: PtML/Au nanoparticles > PtML/Ir nanoparticles ≈ PtML/Rh nanoparticles > PtML/Ru Nanoparticles ≈ PtML/Pd nanoparticles.PtML supported on Au nanoparticles showed the highest SA, which was about 8.3 times higher than that on Ru and Pd nanoparticles.In addition, compared to other Pt-based catalysts reported in previous work, the obtained PtML/Au nanoparticles exhibited relatively high SA for electrooxidation ammonia and Pt surface area among the various catalysts (Table 2).As a typical representative, as-prepared PtML/Au nanoparticles are provided in Figure 7c since it exhibited highest SA among the obtained catalysts.The high-resolution transmission electron microscopy (HRTEM) image showed the clear lattice fringes indicating a high crystallinity of the PtML/Au nanoparticles, and it also revealed that the PtML/Au nanoparticles were composed of multiple crystalline domains, suggesting a polycrystalline structure.The widely accepted mechanism of ammonia electro-oxidation on Pt was proposed by Gerischer and Mauerer [54], involving dehydrogenation of NH3,ads to NHx,ads (x = 1 or 2) intermediates and Nads.The partially dehydrogenated NH2,ads species are key precursors to promote the formation of N2H4 that is then dehydrogenated quickly to N2 [54].To investigate the electrocatalytic activity for the electro-oxidation of ammonia, Figure 7a shows the CVs measured on five different Pt ML covered nanoparticles in 0.1 M ammonia + 1 M KOH.In order to reveal the SA, the current was normalized by the effective electrochemical surface area.Compared to observed CVs on bare nanoparticles (Figure 3), applying a Pt ML on these nanoparticles resulted in an obvious change of CV curves.All five CV curves showed a current peak around 0.7 V (vs.RHE) on the positive sweep, which is a typical feature of ammonia electro-oxidation on Pt as previously discussed.This again confirms the successful deposition of Pt on various nanoparticles by the replacement of UPD Cu.It is worth noticing that the SA of Pt ML covered nanoparticles strongly depends on the composition of core nanoparticles, which can be clearly seen in Figure 7b.The SA for electro-oxidation of ammonia decreases in the following order: Pt ML /Au nanoparticles > Pt ML /Ir nanoparticles ≈ Pt ML /Rh nanoparticles > Pt ML /Ru Nanoparticles ≈ Pt ML /Pd nanoparticles.Pt ML supported on Au nanoparticles showed the highest SA, which was about 8.3 times higher than that on Ru and Pd nanoparticles.In addition, compared to other Pt-based catalysts reported in previous work, the obtained Pt ML /Au nanoparticles exhibited relatively high SA for electrooxidation ammonia and Pt surface area among the various catalysts (Table 2).As a typical representative, as-prepared Pt ML /Au nanoparticles are provided in Figure 7c since it exhibited highest SA among the obtained catalysts.The high-resolution transmission electron microscopy (HRTEM) image showed the clear lattice fringes indicating a high crystallinity of the Pt ML /Au nanoparticles, and it also revealed that the Pt ML /Au nanoparticles were composed of multiple crystalline domains, suggesting a polycrystalline structure.The widely accepted mechanism of ammonia electro-oxidation on Pt was proposed by Gerischer and Mauerer [54], involving dehydrogenation of NH 3,ads to NH x,ads (x = 1 or 2) intermediates and N ads .The partially dehydrogenated NH 2,ads species are key precursors to promote the formation of N 2 H 4 that is then dehydrogenated quickly to N 2 [54].This mechanism has been later confirmed by Catalysts 2019, 9, 4 9 of 15 experimental methods [55,56] and density functional theory calculations [57], supporting that NH 2 species are active intermediates, and therefore a Pt surface with a higher binding energy of NH 2 intermediates has higher activity for ammonia electro-oxidation.Previous studies found that due to the lattice mismatch between the Pt ML and the substrate, different level of tensile or compressive strain is exerted on the Pt ML depending on the substrate material [19].For example, Au exerts a tensile strain on Pt ML (over 4%) while other metals such as Pd, Ru and Rh exert on it a comprehensive strain (about −2 ~−4%) [19].This affects the binding abilities of the Pt ML and thus changes the surface activity.In consistence with theoretical calculations [58,59], Li et al. [19] demonstrated that the tensile surface strain in the Pt ML (e.g., Pt ML supported on Au) tends to upshift the weighted center of the d-band (ε d ) in energy and thus increases the binding of CO and OH.This contributes to a significant enhancement in the activity of the Pt ML on Au towards electrocatalytic oxidation of methanol and ethanol [19].On the contrary, Pt ML on other substrates (i.e., Ru, Rh, Ir and Pd) is under compressive strain, resulting in decreased activity for the electro-oxidation of methanol and ethanol [19].It is interesting to note that the substrate-induced surface strain effect on activity was also been found in the present work.For example, the Pt ML supported on electrodeposited Au nanoparticles showed significantly higher SA compared to that supported on electrodeposited Pd and Ru nanoparticles.Similarly, the tensile surface strain in the Pt ML induced by the substrate Au is expected to increase the binding of the Pt ML for active intermediates such as NH 2 to activate ammonia electro-oxidation and thus improves the SA.On the contrary, Pd and Ru nanoparticles exert a compressive strain on the Pt ML [19] and therefore they have an opposite effect compared to Pt ML on Au, exhibiting low specific activities among all investigated Pt ML nanoparticles.However, it was noticed that Pt ML on electrodeposited Rh and Ir nanoparticles had remarkably higher activity than those on Pd and Ru nanoparticles despite of the compressive strain exerted by Rh or Ir on Pt, which is different with that reported by previous studies [19].This was due to the positive synergistic effect between Pt and Rh or Ir.Cooper et al. [39] pointed out that the presence of Rh along with Pt will increase active sites on the Pt surface for ammonia electro-oxidation.The synergic effect of Ir to Pt has also been reported by previous studies, which was attributed to the ability of Ir to dehydrogenate ammonia molecules at lower potentials [36,60,61].Our results agree well with these studies [36,39,60].
Catalysts 2018, 8, x FOR PEER REVIEW 9 of 15 This mechanism has been later confirmed by experimental methods [55,56] and density functional theory calculations [57], supporting that NH2 species are active intermediates, and therefore a Pt surface with a higher binding energy of NH2 intermediates has higher activity for ammonia electro-oxidation.Previous studies found that due to the lattice mismatch between the PtML and the substrate, different level of tensile or compressive strain is exerted on the PtML depending on the substrate material [19].For example, Au exerts a tensile strain on PtML (over 4%) while other metals such as Pd, Ru and Rh exert on it a comprehensive strain (about −2 ~ −4%) [19].This affects the binding abilities of the PtML and thus changes the surface activity.In consistence with theoretical calculations [58,59], Li et al. [19] demonstrated that the tensile surface strain in the PtML (e.g., PtML supported on Au) tends to upshift the weighted center of the d-band (εd) in energy and thus increases the binding of CO and OH.This contributes to a significant enhancement in the activity of the PtML on Au towards electrocatalytic oxidation of methanol and ethanol [19].On the contrary, PtML on other substrates (i.e., Ru, Rh, Ir and Pd) is under compressive strain, resulting in decreased activity for the electro-oxidation of methanol and ethanol [19].It is interesting to note that the substrate-induced surface strain effect on activity was also been found in the present work.For example, the PtML supported on electrodeposited Au nanoparticles showed significantly higher SA compared to that supported on electrodeposited Pd and Ru nanoparticles.Similarly, the tensile surface strain in the PtML induced by the substrate Au is expected to increase the binding of the PtML for active intermediates such as NH2 to activate ammonia electro-oxidation and thus improves the SA.On the contrary, Pd and Ru nanoparticles exert a compressive strain on the PtML [19] and therefore they have an opposite effect compared to PtML on Au, exhibiting low specific activities among all investigated PtML nanoparticles.However, it was noticed that PtML on electrodeposited Rh and Ir nanoparticles had remarkably higher activity than those on Pd and Ru nanoparticles despite of the compressive strain exerted by Rh or Ir on Pt, which is different with that reported by previous studies [19].This was due to the positive synergistic effect between Pt and Rh or Ir.Cooper et al. [39] pointed out that the presence of Rh along with Pt will increase active sites on the Pt surface for ammonia electro-oxidation.The synergic effect of Ir to Pt has also been reported by previous studies, which was attributed to the ability of Ir to dehydrogenate ammonia molecules at lower potentials [36,60,61].Our results agree well with these studies [36,39,60].  a] CV was performed at room temperature. [b] Pt surface area is the effective electrochemical surface area of Pt normalized by the Pt amount. [c] SA was calculated by the peak oxidation current density normalized by the effective electrochemical surface area of Pt.

Electrode Synthesis and Characterization
All electrochemical syntheses and characterizations were performed on a three-electrode setup.A glassy carbon electrode (GCE, 5 mm in diameter) served as the working electrode, a Pt plate (1 × 1 cm 2 ) electrode and a mercury sulfate electrode (MSE) served as counter and reference electrodes respectively.For the convenience of comparison, the potential measured by MSE was referred to reversible hydrogen electrode (vs.RHE).Prior to the electrodeposition, the GCE was first polished and then ultrasonically cleaned in ultrapure water for 3 min.Through controlling the deposition potential under potentiostatic conditions, nanoparticles with different compositions were synthesized on the surface of GCE.Some preliminary experiments had been carried out to find out the optimal electrodeposition potentials and times, and the following conditions were employed to synthesize nanoparticles.Ru nanoparticles were synthesized on the GCE at −0.8 V (vs.RHE) for 8 s in 5 mM RuCl 3 + 0.05 M H 2 SO 4 solution.Rh nanoparticles were synthesized on the GCE at −0.05 V (vs.RHE) for 0.2 s in 5 mM RhCl 3 + 0.05 M H 2 SO 4 solution.Pd nanoparticles were synthesized on the GCE at 0 V (vs.RHE) for 0.2 s in 5 mM PdCl 2 + 0.05 M H 2 SO 4 solution.Ir nanoparticles were synthesized on the GCE at −0.7 V (vs.RHE) for 800 s in 5 mM IrCl 3 + 0.05 M H 2 SO 4 solution.Au nanoparticles were synthesized on the GCE at 0 V (vs.RHE) for 0.2 s in 5 mM HAuCl 4 + 0.05 M H 2 SO 4 solution.The synthesis of Pt ML on electrodeposited nanoparticles was achieved by a well-established method, as described as follows.After the synthesis of core nanoparticles on the GCE, the modified GCE, rinsed thoroughly with ultrapure water, was transferred into 0.05 M H 2 SO 4 + 0.05 M CuSO 4 aqueous solution for the subsequent Cu UPD.The modified GCE was initially kept at 0.8 V (vs.RHE) for 20 s to make sure there is no Cu on the surface [18,33].The electrode potential was subsequently swept at 0.4 V s −1 to the Cu UPD potential of 0.36 V (vs.RHE) and held at this potential for 60 s [18,33].When the Cu UPD was finished, the corresponding GCE, rinsed thoroughly with ultrapure water, was immediately transferred into solution containing 5 mM K 2 PtCl 4 and 0.05 M H 2 SO 4 under the protection of high-purity Ar (99.999%) atmosphere to achieve the replacement of Cu with Pt 2+ ions for 20 min.After the replacement reaction, the Pt ML covered nanoparticles on the GCE were rinsed thoroughly with ultrapure water and dried in a nitrogen stream.The amount of the deposited Cu monolayer by UPD could be measured by the anodic stripping technique.For anodic stripping tests, the electrode potential was kept at 0.36 V (vs.RHE) for 60 s to complete the Cu UPD, and then swept back to 0.8 V (vs.RHE) at 0.4 V s −1 to strip off the Cu layer [33].
The particle size and dispersion degree of the Pt ML decorated nanoparticles were determined by scanning electron microscopy (SEM, S-4800, Hitachi, Tokyo, Japan) and transmission scanning electron microscopy (TEM, JEM-2100F, JEOL, Tokyo, Japan).For the TEM measurements, the electrodeposited nanoparticles were gently scraped and dispersed in ethanol on surface of GCE, and then the nanoparticle dispersions were transferred onto a Cu grid, which was directly conducted on surface of GCE.

Electrochemical Tests
Electrochemical tests were also carried out on the electrochemical workstation (IviumStat, Eindhoven, Netherland).Electrodeposited pure nanoparticles, or Pt ML covered nanoparticles, on the GCE served as the working electrode.A Pt plate (1 × 1 cm 2 ) served as the counter electrode.A CV technique was used to investigate the electrochemical performance of nanoparticles with different compositions before and after the coverage of the Pt ML .For the CVs tested in 0.5 M H 2 SO 4 aqueous solution, an MSE was used as the reference electrode with a scan rate of 0.05 V s −1 .The electro-oxidation of ammonia was conducted in aqueous solution containing 1 M KOH + 0.1 M ammonia by CV at 0.01 V s −1 , and a Hg/HgO (filled with 1 M KOH) served as the reference electrode in this case.All the electrochemical measurements were performed in testing solutions saturated with argon gas (99.999%).All the measured potentials by the reference electrode were referred to as the RHE for the convenience of comparison.All experiments were carried out at controlled temperature of 25 ± 1 • C.

Conclusions
The effect of the composition of nanoparticle supports on the electrocatalytic activity of Pt ML electrocatalysts for ammonia oxidation was investigated.A clean electrochemical approach free of binders, surfactants, capping agents and reductants was proposed to synthesis nanoparticle supports with various compositions including Ru, Rh, Pd, Ir and Au.Pt ML supported on different nanoparticles were obtained by the Cu UPD and subsequent replacement of Pt ions.The SA of the Pt ML electrocatalysts was significantly dependent on the composition of nanoparticle supports, which decreased in the following order: Pt ML /Au nanoparticles > Pt ML /Ir nanoparticles ≈ Pt ML /Rh

Figure 1 .
Figure 1.Scanning electron microscopy (SEM) images of (a) Ru, (b) Rh, (c) Pd, (d) Ir, and (e) Au nanoparticles electrodeposited on the GCE, respectively.Insets of SEM images show the corresponding particle size distribution histogram.

Figure 1 .
Figure 1.Scanning electron microscopy (SEM) images of (a) Ru, (b) Rh, (c) Pd, (d) Ir, and (e) Au nanoparticles electrodeposited on the GCE, respectively.Insets of SEM images show the corresponding particle size distribution histogram.

Figure
Figure3a-e shows CVs measured on pure Ru, Rh, Pd, Ir, and Au nanoparticles electrodeposited on the GCE in 0.1 M ammonia + 1 M KOH aqueous solution at 0.05 V s −1 .Only for Ir nanoparticles did the CV curve exhibited an anodic current peak at 0.59 V (vs.RHE) (Figure3d).Such an oxidation current peak has been extensively reported on pure Ir and also Pt electrodes and is ascribed to the electro-oxidation of ammonia to N2[12,13,37].Previous work performed CV measurements in the absence of ammonia and found that a such characteristic current peak did not appear[13,37,38].For 4d transition metals (i.e., Ru, Rh and Pd), all CV plots showed no anodic oxidation current peak related with the ammonia electro-oxidation from 0.4 to 0.7 V (vs.RHE).This is in good agreement with previous results reported by de Vooys et al.[13] and Vidal-Iglesias et al.[36].Based on differential electrochemical mass spectroscopy (DEMS) measurements, de Vooys et al.[13] found the dehydrogenation of ammonia occurred at significantly lower potentials on Ru, Rh and Pd than that on Pt and Ir, resulting in the surface poisoning by adsorbed nitrogen atoms (Nads) at much lower potentials on Ru, Rh and Pd.As a result, Ru, Rh and Pd were inactive towards the ammonia electro-oxidation to N2[13].It was also observed that the peak height at 0.2 V (vs.RHE) for Rh nanoparticles (Figure3b) is remarkably higher compared to other metals.This agrees well with previous studies reported byCooper et al. [39], which attributed such peak to the OH − adsorption by Rh.

Figure 3 .
Figure 3. CVs of the ammonia electro-oxidation on pure (a) Ru, (b) Rh, (c) Pd, (d) Ir, and (e) Au nanoparticles electrodeposited on the GCE in aqueous solution containing 1 M KOH + 0.1 M ammonia at 0.05 V s −1 , respectively.

Figure 3 .
Figure 3. CVs of the ammonia electro-oxidation on pure (a) Ru, (b) Rh, (c) Pd, (d) Ir, and (e) Au nanoparticles electrodeposited on the GCE in aqueous solution containing 1 M KOH + 0.1 M ammonia at 0.05 V s −1 , respectively.

Figure 5 .
Figure 5. TEM images of PtML particles with (a) Ru, (b) Rh, (c) Pd, (d) Ir, and (e) Au cores electrodeposited on the GCE, respectively.Insets of TEM images show the corresponding particle size distribution histogram.

Figure 5 .
Figure 5. TEM images of Pt ML particles with (a) Ru, (b) Rh, (c) Pd, (d) Ir, and (e) Au cores electrodeposited on the GCE, respectively.Insets of TEM images show the corresponding particle size distribution histogram.

Figure 7 .
Figure 7. (a) CVs of the ammonia electro-oxidation on PtML supported on Ru nanoparticles, Rh nanoparticles, Pd nanoparticles, Ir nanoparticles, and Au nanoparticles on the GCE in aqueous solution containing 1 M KOH + 0.1 M ammonia at 0.05 V s −1 , respectively.(b) Comparison of the SA of PtML supported on Ru nanoparticles, Rh nanoparticles, Pd nanoparticles, Ir nanoparticles, and Au nanoparticles on the GCE.(c) HRTEM image of as-prepared PtML/Au nanoparticles.

Figure 7 .
Figure 7. (a) CVs of the ammonia electro-oxidation on Pt ML supported on Ru nanoparticles, Rh nanoparticles, Pd nanoparticles, Ir nanoparticles, and Au nanoparticles on the GCE in aqueous solution containing 1 M KOH + 0.1 M ammonia at 0.05 V s −1 , respectively.(b) Comparison of the SA of Pt ML supported on Ru nanoparticles, Rh nanoparticles, Pd nanoparticles, Ir nanoparticles, and Au nanoparticles on the GCE.(c) HRTEM image of as-prepared Pt ML /Au nanoparticles.

Table 1 .
Calculated values of the Cu stripping charge, Cu amount, Pt amount and Pt surface area.

Table 1 .
Calculated values of the Cu stripping charge, Cu amount, Pt amount and Pt surface area.

Table 2 .
Comparison of SA for ammonia electro-oxidation and Pt surface area of as-prepared electrocatalysts with previous work.