Abstract
The present invention refers to a method of treating a platinum-alloy catalyst, comprising the steps of: A) providing a platinum-alloy catalyst, which comprises platinum (Pt) and at least one metal (M), which is less noble than platinum; B) exposing the catalyst to an acidic or basic medium, and exposing the catalyst to an adsorptive gas, wherein during step B) the catalyst is not subjected to an external electrical current or voltage. Further, the invention refers to a treated platinum-alloy catalyst. Moreover, and a device for carrying out the method of treating the platinum-alloy catalyst is provided.
Claims
1. A method of treating a platinum-alloy catalyst, comprising the steps of: A) providing the platinum-alloy catalyst, which comprises platinum (Pt) and at least one metal (M), which is less noble than platinum, B) exposing the catalyst to an acidic or basic medium, and exposing the catalyst to an adsorptive gas, wherein during step B) the catalyst is not subjected to an external electrical current or voltage.
2. Method according to claim 1, wherein the adsorptive gas is selected from carbon monoxide (CO), hydrogen (H.sub.2), ethylene (C.sub.2H.sub.4), SO.sub.2 or H.sub.2S.
3. Method according to claim 1, wherein the platinum-alloy of the platinum-alloy catalyst has the general formula
PtM.sub.x, wherein M is at least one metal selected from the group consisting of Cu, Ni, Co, Fe, Ag, Pd, Zn, Cr, Ti, Pb, Sn, Mo, Y and Gd, and wherein x fulfils the requirement 0<x≤5.
4. Method according to claim 1, wherein the method of treating the platinum-alloy catalyst is a method of regeneration of a used platinum-alloy catalyst.
5. Method according to claim 1, wherein the acidic medium comprises an acid having a pK.sub.a-value ranging from −1 to 10.
6. Method according to claim 5, wherein the acidic medium comprises an acid which is selected from the group consisting of carboxylic acids, preferably the carboxylic acid is selected from the group consisting of acetic acid, formic acid, propionic acid, butyric acid, valeric acid and capronic acid.
7. Method according to claim 1, wherein the basic medium comprises a base having a pK.sub.b-value ranging from 0 to 8.
8. Method according to claim 7, wherein the basic medium comprises a base having the general formula NR.sup.1R.sup.2R.sup.3, wherein R.sup.1, R.sup.2 and R.sup.3 are each independently from one another selected from the group consisting of H, Methyl, Ethyl, Propyl, Isopropyl, Butyl, Pentyl and Hexyl.
9. Method according to claim 1, wherein step B) is performed at a temperature of more than 0° C. and less than 75° C.
10. Method according to claim 1, wherein exposing the catalyst to an acidic or basic medium, and exposing the catalyst to an adsorptive gas of step B) is performed by dispersing the catalyst in the acidic or basic medium, wherein the adsorptive gas is at least partially dissolved in the acidic or basic medium and/or wherein bubbles of the adsorptive gas are present in the acidic or basic medium.
11. Method according to claim 1, further comprising: C) separating the catalyst from the acidic or basic medium.
12. Method according to claim 1, wherein step B) and step C) are repeated at least once.
13. Method according to claim 11, further comprising: D) washing the catalyst with a solvent.
14. Method according to claim 1, wherein step B) is performed in a batch reactor or a continuous reactor.
15. Method according to claim 1, wherein step B) is performed in a reactor selected from the group consisting of a fixed bed reactor (FBR), a mechanically stirred reactor (MSR) or a draft tube reactor (DTR).
16. A treated platinum-alloy catalyst, the treated platinum-alloy catalyst comprises platinum (Pt) and at least one metal (M), which is less noble than platinum, wherein the treated platinum-alloy catalyst has a core-shell-structure with a platinum-alloy core and a platinum shell.
17. The treated platinum-alloy catalyst according to claim 16, wherein the less noble metal (M) is at least one metal selected from the group consisting of Cu, Ni, Co, Fe, Ag, Pd, Zn, Cr, Ti, Pb, Sn, Mo, Y and Gd.
18. The treated platinum-alloy catalyst according to claim 16, wherein the platinum-alloy catalyst is a nanoparticulate platinum alloy catalyst, the platinum alloy nanoparticles having an average particle diameter ranging from 1 nm to 1000 nm.
19. The treated platinum-alloy catalyst according to claim 18, wherein the platinum alloy nanoparticles are supported on a support.
20. Device for carrying out the method according to claim 1, comprising: a source of an adsorptive gas, a source of an acidic or basic medium, a reactor for exposing a platinum-alloy catalyst with the acidic or basic medium and exposing the platinum-alloy catalyst with the adsorptive gas, wherein the reactor is selected from the group consisting of a fixed bed reactor (FBR), a mechanically stirred reactor (MSR) or a draft tube reactor (DTR), and is comprising: at least one inlet and at least one outlet wherein the source of the adsorptive gas is in fluid connection with the at least one inlet, and the source of the acidic or basic medium is in fluid connection with the at least one inlet, at least one separator for separating the platinum-alloy catalyst from the acidic or basic medium.
21. Device according to claim 20, further comprising means for regeneration of the acidic or basic medium, said means being in fluid connection with both the at least one outlet of the reactor and the source of the acidic or basic medium.
Description
BRIEF DESCRIPTION OF THE DRAWINGS
[0247] The method according to the first aspect of the invention and the device for carrying out the method according to the second aspect of the invention are explained in greater detail below with reference to exemplary embodiments and the associated figures, in which:
[0248] FIG. 1 shows STEM images of an as prepared PtCu.sub.3/C catalyst.
[0249] FIGS. 2a to 2d show a typical method of electrochemical activation of an as prepared PtCu.sub.3/C catalyst via cyclic voltammetry and measurements of its ORR-activity and ESA after complete activation.
[0250] FIGS. 3a and 3b show measurements of an as prepared PtCu.sub.3/C catalyst in an electrochemical flow cell, monitoring Cu- and Pt-dissolution via ICP-MS during electrochemical activation.
[0251] FIGS. 4a to 4h show TEM-EDX images and compare the chemical composition of a PtCu.sub.3/C catalyst as prepared and after electrochemical activation.
[0252] FIG. 5 compares the chemical composition obtained from EDX of a PtCu.sub.3/C catalyst, as prepared (a), and after exposure to acetic acid for varying amounts of time (b)-(d), after acetic acid exchange (e) and after additional electrochemical activation (f), wherein TEM-EDX was performed on 4 catalyst areas as large as shown in (g).
[0253] FIG. 6 shows X-ray diffraction spectra of a PtCu.sub.3/C catalyst, as prepared (a), and after exposure to acetic acid for varying amounts of time (b)-(d), and after additional acetic acid exchange (e).
[0254] FIG. 7 shows ORR polarization curves for ORR-activity determination of a a PtCu.sub.3/C catalyst recorded after exposure to acetic acid for 2 h (a), 4 h (b), 6 h (c) and after 6 h and additional acid exchange (d).
[0255] FIG. 8 shows CO electrooxidation measurements for ESA determination of a PtCu.sub.3/C catalyst recorded after exposure to acetic acid for 2 h (a), 4 h (b), 6 h (c) and after 6 h and additional acid exchange (d).
[0256] FIG. 9 shows ORR polarization curves (a), Tafel plots (b), CO electrooxidation experiments with follow-up cycles (c) and obtained ESA.sub.CO (d), specific activity (e), mass activity (f) of PtCu.sub.3/C catalysts that were subjected to ex-situ chemical activation (i)-(iii) and after ORR and CO measurement of ex-situ chemical activation (iii), the film was further subjected to 50 cycles of in-situ EA and ORR and CO were measured again (iv) (i=Reference Example 2; ii=Reference Example 1; iii=Example 1).
[0257] FIG. 10 shows a comparison of CO-electrooxidation experiments and ORR polarization curves between treatment in 1 M HAc saturated with CO (FIGS. 10a and 10b; Example 1) and treatment in 1 M HClO.sub.4 (FIGS. 10c and 10d; Example 2).
[0258] FIG. 11 shows Table 1, comparing the effect of using HClO.sub.4 instead of HAc as acidic medium for catalyst treatment.
[0259] FIG. 12 shows a device for treating a platinum-alloy catalyst according to the present invention, wherein for a batch reactor is used, namely a fixed bed reactor (FBR) is employed.
[0260] FIG. 13 shows a device for treating a platinum-alloy catalyst according to the present invention, wherein a mechanically stirred reactor (MSR) is employed either in batch mode or in continuous mode.
[0261] FIG. 14 shows a device for treating a platinum-alloy catalyst according to the present invention, wherein a draft tube reactor (DTR) is employed either in batch mode or in continuous mode.
DETAILED DESCRIPTION OF THE INVENTION
[0262] FIG. 1 shows scanning transmission electron microscopy images of a PtCu.sub.3/C catalyst after synthesis. The catalyst comprises PtCu.sub.3 nanoparticles supported on a carbon support. Details about the synthesis of this specific catalyst have been published previously (e.g. U.S. Pat. No. 9,147,885 B2). The catalyst is characterized by a homogenous particle size, the average particle diameter being approximately 4 nm. The PtCu.sub.3/C catalyst is a catalyst, which is particularly hard to activate without employing electrochemical activation. The reason for this is the high content and comparatively high nobility of copper, compared to other platinum-alloy-catalysts. Therefore, the PtCu.sub.3/C catalyst is a perfect model system to study efficient methods of treatment for catalyst activation. Thus for the experiments explained in the following the PtCu.sub.3/C catalyst was employed as a model system illustrating the effects of the present invention.
[0263] FIG. 2a shows a standard method of electrochemical activation (in-situ EA) of the as synthesized PtCu.sub.3/C catalyst via cyclic voltammetry. The measurements are carried out in a laboratory cell setup, namely in a three electrode half-cell configuration as explained above. A thin film of the PtCu.sub.3/C electrocatalyst was deposited on a working electrode. As a working electrode a RDE was used. The electrochemical activation was performed via subjecting the electrocatalyst to cyclic voltammetry, i.e. an external potential was subjected to the catalyst and varied within a potential window ranging between 0.05 V.sub.RHE and 1.2 V.sub.RHE. The measurements were performed in 0.1 M HClO.sub.4 (Merck, Suprapur), which was saturated with Argon. In total 200 potential cycles between 0.05 V.sub.RHE and 1.2 V.sub.RHE were performed (scan rate 300 mV.Math.s.sup.−1). Selected cyclic voltammograms (CV) are shown after the 1.sup.st, 2.sup.nd, 10.sup.th, 50.sup.th, 100.sup.th and 200.sup.th potential cycle. From the CVs it can be seen how the surface properties of the PtCu.sub.3/C catalyst change in the course of the electrochemical activation procedure. Namely, in the 1st CV the typical features of Cu can be observed (Cu.sub.OPD and Cu.sub.UPD-peaks), while none of the typical features for Pt can be seen. This shows that initially the surface of the platinum-alloy catalyst is almost completely covered with Cu. However, significant amounts of Cu are leached from the catalyst surface already during the first potential cycles. Namely, the vast majority of unstable Cu is removed from the surface and near surface area of electrocatalyst nanoparticles already after the 1st cycle of in-situ EA. In addition to the removal of elemental Cu multilayers that correspond to Cu.sub.OPD, one can also notice two additional peaks that correspond to Cu.sub.UPD removal (see labels denoting both types of deposited copper in FIG. 2a). From the 1.sup.st to about 50.sup.th potential cycle of in-situ EA, the typical Pt CV features (hydrogen under potential deposition—H.sub.UPD region between 0.05 V.sub.RHE and about 0.35 V.sub.RHE, as well as the reduction of Pt hydroxide/oxide between 0.85 to 0.55 V.sub.RHE) are poorly resolved due to trace leftover amounts of Cu, as well as constant leaking of remaining unstable Cu from the subsurface layers of the platinum-alloy catalyst with each consecutive potential cycle. Those species constantly block the Pt surface as Cu.sub.OPD and Cu.sub.UPD. Upon removal of those trace amounts, a gradual evolution of the expected Pt peaks takes place so that after about 50 potential cycles, relatively stable CVs with fully developed Pt features are recorded. Nevertheless, upon tracking the Cu dissolution during the entire 200 cycles of electrochemical activation (FIG. 3a), it can be noticed that while indeed Cu dissolution already starts to limit towards 0 after 50 cycles (between 0-383.33 seconds), a substantial amount of Cu gets removed between 50.sup.th and 200.sup.th potential cycle (between 383.33-1533.33 s). Thus, activation of a platinum-alloy catalyst is an interplay between the removal of unstable subsurface less noble metal (M) and removal of the subsequential OPD and UPD species of M on Pt. Only once the leaking of the less noble metal goes below a critical level and the subsequent OPD and UPD species of the less noble metal M on Pt block an insignificant portion of Pt surface, high and reproducible activities can be achieved (FIGS. 2b and 2c).
[0264] FIG. 2b shows ORR polarization curves measured of three different batches of PtCu.sub.3/C catalyst. Three 2 g batches of PtCu.sub.3/C catalyst were prepared (For the electrochemical measurements however, only a few micrograms of the platinum-catalyst are utilized). The ORR polarization curves are measured after electrochemical activation via 200 potential cycles (performed at 600 rpm; rpm=rotations per minute). The electrolyte was exchanged prior to recording the ORR polarization curves. The ORR polarization curves were recorded at 1600 rpm. The so called kinetic-region (i.e. the potential range between 0.85 V.sub.RHE to 0.95 V.sub.RHE of the ORR polarization curves) can be used to determine the catalytic activity of the platinum-alloy catalyst for the ORR.
[0265] FIG. 2c shows the corresponding Tafel plots obtained from the ORR polarization curves of FIG. 2b. From the Tafel plot the specific activity can be directly derived for a given potential. In most cases ORR-activities are compared at a potential of 0.9 V.sub.RHE.
[0266] FIG. 2d shows CO-electrooxidation experiments for each of the three investigated batches of PtCu.sub.3/C catalyst. CO-electrooxidation experiments are used to determine the catalytically active platinum surface area of a platinum-alloy catalyst. For this purpose the platinum-alloy catalyst is subjected to a constant potential, such as 0.05 V.sub.RHE. Under these potential hold conditions subjected to the catalyst via the potentiostat a monolayer of CO is adsorbed to the catalyst surface. Thereafter, a potential cycle is performed in order to electrochemically oxidise CO to CO.sub.2. This results in a CO-electrooxidation peak at about 0.8 V.sub.RHE. Based on the charge required for CO-electrooxidation the ESA can be calculated.
[0267] As can be seen from FIGS. 2b-2d the ORR-curves, Tafel plots and CO-electrooxidation experiments indicate excellent reproducibility and thus homogeneity of the PtCu.sub.3/C catalyst.
[0268] FIG. 3a shows measurements of a PtCu.sub.3/C catalyst in an electrochemical flow cell (EFC) with online electrolyte analysis via ICP-MS (Inductively-coupled mass spectrometry). In the electrochemical flow cell the PtCu.sub.3/C catalyst is subjected to 200 potential cycles between 0.05 V.sub.RHE and 1.2 V.sub.RHE (scan rate 300 mV.Math.s.sup.−1). That is the platinum-alloy catalyst is subjected to electrochemical activation under conditions as also shown in FIG. 2a. The electrochemical flow cell allows to investigate how much Cu is dissolved into the electrolyte in the course of potential cycling. It can be noticed that while indeed Cu dissolution already starts to significantly decrease after 50 cycles (1.sup.st to 50.sup.th cycle=time between 0-383.33 seconds), still a substantial amount of Cu gets removed between the 50.sup.th and 200.sup.th potential cycle (50.sup.th to 200.sup.th cycle=time between 383.33-1533.33 s). Therefore, FIG. 3a clearly proves that the entire 200 potential cycles are required (for a catalyst, which is not subjected to any other type of activation) until a stable PtCu.sub.3/C catalyst is obtained and thus electrochemical activation is completed.
[0269] FIG. 3b shows a measurement of the PtCu.sub.3/C catalyst in an electrochemical flow cell (EFC) at the scan rate of 20 mV.Math.s.sup.−1. The measurement allows to identify Cu and Pt dissolution during a single potential cyclic voltammogram. Both metals exhibit an anodic (peak 2) and a cathodic (peak 3) transient dissolution peak. Moreover, Cu has an additional transient dissolution peak 1. This dissolution peak corresponds to OPD and UPD species of Cu on Pt that in too high quantity have a highly negative effect on ORR activity.
[0270] FIG. 4 reveals selected morphological and compositional changes occurring during electrochemical activation (in-situ EA) under potential cycling conditions as shown in FIG. 2a carried out on the as synthesized PtCu.sub.3/C electrocatalyst. FIGS. 4c to 4e show a single PtCu.sub.3 nanoparticle as prepared, i.e. after synthesis. FIGS. 4f to 4h show the same particle after being subjected to electrochemical activation. Elemental mapping for Cu (FIGS. 4c and 4f) and for Pt (FIGS. 4d and 4g) was performed via TEM-EDX (EDX=energy dispersive X-ray spectroscopy). FIG. 4e shows FIGS. 4c and 4d overlapping, FIG. 4h shows FIGS. 4f and 4g overlapping. The elemental mapping of the as-prepared PtCu.sub.3 nanoparticle shows that both metals (Pt and Cu) are fully mixed (FIG. 4c-e). After 200 potential cycles of electrochemical activation several important differences are observed (FIG. 4f-h). The Cu content in the nanoparticles drops significantly (e.g. from 62 to 37 wt. % for the particular particle displayed in FIGS. 2a and 2b). More importantly, as visible from FIG. 4h, apart from Cu.sub.UPD and Cu.sub.OPD removal, Cu is particularly leached from the near surface region, forming a Pt-rich overlayer over the PtCu.sub.3 core.
[0271] In the experiments illustrated in FIGS. 2 to 4 the as synthesized PtCu.sub.3/C catalyst was subjected to electrochemical activation in an electrochemical cell in order to remove Cu from the catalyst surface and near surface region. From the performed experiments it can be summarized that that an as synthesized PtCu.sub.3/C catalyst requires extensive in-situ electrochemical activation in order to obtain a stable and active platinum-alloy catalyst.
[0272] Based on the findings of electrochemical activation reported above, the inventors of the present invention have studied ways of activating platinum-alloy catalysts without the need of an electrochemical cell. That is, the inventors have explored conditions for an effective ex-situ chemical activation.
[0273] FIG. 5 compares the chemical composition of a PtCu.sub.3/C catalyst obtained from TEM-EDX for at least 4 catalyst regions as large as the catalyst region shown in FIG. 5g. The chemical composition is given of the PtCu.sub.3/C catalyst, as prepared (a), and after 2 h (b), 4 h (c), 6 h (d) in 1 M acetic acid and after exchange of acetic acid with fresh acetic acid with additional acetic acid washing for another 30 minutes (e) (See Reference Example 3). In addition, the composition of the as prepared PtCu.sub.3/C catalyst after being subjected to 200 potential cycles of electrochemical activation is shown in (f). Thus as part of the initial experiment, the duration of one-time acid washing (Reference Example 3) was varied (t=2, 4 or 6 h). Additionally, the experiment with t=6 h was repeated with additional exchange of the acid with a fresh one (t=30 min). Statistically representative TEM-EDX data revealed that the chemical compositions determined after a single acid wash matched well with those obtained during other types of treatments including a 200-cycle in-situ EA (FIG. 5b-f).
[0274] FIG. 6 shows X-ray diffraction spectra of as prepared PtCu.sub.3/C (a) and after (b) 2 h, (c) 4 h, (d) 6 h in 1 M acetic acid as well as (e) after exchange of acetic acid with a fresh acetic acid and extra acetic acid washing for 30 minutes as described in Reference Example 3. The peak at approximately 25° 2θ corresponds to graphitic structure of the carbon support while peaks at approximately 42.4° and 49.3° 2θ correspond to the most intense peaks of PtCu.sub.3 pm3m crystal phase (111 and 200). Comparison of XRD spectra of samples that had been treated with one-time acid washing in 1 M HAc (t=2, 4 or 6 h) and two-time acid washing (t=6 h+acid exchange) revealed that already the first and shortest washing (t=2 h) produced a stable structure that did not further develop during additional washing treatment (FIG. 6b-e).
[0275] In summary, FIGS. 5 and 6 show that after all of these treatments the crystal structure of the nanoparticles is very similar.
[0276] However, despite the compositional and structural similarity of these samples, their electrocatalytic performance measured by TF-RDE varies significantly (FIG. 7-8). This strongly suggests that the main difference is in the surface properties of the samples, possibly in the degree of Cu.sub.OPD and Cu.sub.UPD surface coverage.
[0277] FIG. 7 shows ORR polarization curves (0.05-1.0 V.sub.RHE, 20 mV.Math.s.sup.−1, 1600 rpm, O.sub.2 saturated electrolyte) measured in 0.1 M HClO.sub.4 directly after (a) 2 hours, (b) 4 hours, (c) 6 hours and (d) 6 hours+additional acid exchange ex-situ acid wash (full) as described in Reference Example 3. Furthermore, in all cases additional 100 potential cycles (dashed) of in-situ electrochemical activation were performed (0.05-1.2 V.sub.RHE, 300 mV, 600 rpm, Ar saturated electrolyte). As can be seen from FIG. 7a the PtCu.sub.3/C catalyst after being exposed to acetic acid for 2 h shows very poor ORR activity. This can be seen from the ORR-curve: First, the so called diffusion limiting current (i.e. about −1.1 mA for 1600 rpm) is not reached for the potential region between 0.05 and 0.8 V.sub.RHE. Moreover, the magnitude of the currents is very low in the so called kinetic region of the ORR-curve between about 0.85 and 0.95 V.sub.RHE. That is, the PtCu.sub.3/C catalyst after being exposed to acetic acid for 2 h shows only slightly negative currents in the kinetic region, indicating poor ORR activity. This means that the overpotential for the ORR is high and the ORR activity low. The diffusion limiting currents in the diffusion limiting region are only reached after the catalyst is subjected to additional electrochemical activation via 100 potential cycles (dashed line). Moreover, the current becomes more negative in the kinetic region after the catalyst is subjected to said additional 100 potential cycles (dashed line). That is, in case of an exposure of the catalyst to acetic acid for only 2 h the chemical activation is very poor and a significant amount of further electrochemical potential cycling is required.
[0278] The results of the one-time washing experiment reveal that increasing the duration of washing resulted in a higher limiting current, lower overpotential and more stable ORR polarization curve CV (FIG. 7a-c). The longer the washing time, the more the results of ex-situ chemical activation (black full, FIG. 7a-c) approach those obtained when using additional 100 potential cycles subsequent to the chemical activation (dashed line). All the parameters are further improved if acid washing (6 h) is upgraded with acid electrolyte exchange (FIG. 7d). However, even in this case no complete activation is achieved.
[0279] FIG. 8 shows carbon monoxide electrooxidation cyclovoltammograms as well as the follow-up cycle (0.05-1.0 V.sub.RHE, 20 mV s.sup.−1, Ar saturated electrolyte) measured in 0.1 M HClO.sub.4 directly after (a) 2 hours, (b) 4 hours, (c) 6 hours and (d) 6 hours exposure to acetic acid plus additional acid exchange ex-situ chemical activation (full) as described in Reference Example 3 and after additional 100 cycles (dashed) of in-situ electrochemical activation (0.05-1.2 V.sub.RHE, 300 mV, 600 rpm, Ar saturated electrolyte). While the differences in measured CO electrooxidation peaks and the follow-up cycles are more subtle than in the case of ORR polarization curves, they can still offer a few important conclusions. As mentioned before, it is assumed that the degree of Cu adatoms coverage drops in the order of a>>b>c>d. If so, then the results show that a lower degree of Cu adatoms coverage leads not only to a higher ESA.sub.CO (=electrochemical active surface area determined via CO electrooxidation) but also to a more symmetric CO electrooxidation peak that is shifted towards higher potentials. Finally, it also results in more pronounced Pt CV features in the follow-up cycle (black full) in comparison to Pt CV in the follow-up cycle after additional in-situ EA cycles (black dash). By performing this experiment, the inventors of the present invention have concluded: [0280] Time of a single acid wash is important—there exists an equilibrium concentration between Cu surface species and the concentration of Cu in the electrolyte. But more importantly, [0281] exchanging the acid has shown positive effects that are more important than the prolonged time of a single acid wash.
[0282] FIG. 9 compares the electrochemical performance of PtCu.sub.3/C catalysts for the ORR after being subjected to different methods of activation treatment (i)-(iv): [0283] (i) PtCu.sub.3/C catalyst is activated by one-time acid washing with 1 M HAc (See Reference Example 2). [0284] (ii) PtCu.sub.3/C catalyst is activated by four-times acid washing with 1 M HAc (See Reference Example 1). [0285] (iii) PtCu.sub.3/C catalyst is activated by four-times acid washing with 1 M HAc, while at the same time being saturated with carbon monoxide (See Example 1). [0286] (iv) Experiment (iii), wherein the catalyst was subjected to additional 50 cycles of in-situ EA (0.05 to 1.2 V.sub.RHE, 300 mV.Math.s.sup.1, 600 rpm) and subsequent exchange of electrolyte (0.1 M HClO.sub.4) prior to recording the ORR polarization curves and CO electrooxidation experiment.
[0287] FIG. 9a shows ORR polarization curves of the PtCu.sub.3/C catalyst treated according to (i)-(iv). FIG. 9b shows the corresponding Tafel plots, which allow to determine the specific activity (SA) of the catalyst. FIG. 9c provides CO electrooxidation experiments in order to determine the electrochemically active surface area (ESA.sub.CO). FIG. 9d compares the ESA.sub.CO, FIG. 9e the SA and FIG. 9f the MA of the catalysts treated according to (i)-(iv).
[0288] Upon inspecting the ORR polarization curves and obtained Tafel plots (FIGS. 9a and 9b), the following ORR specific activity trend is found: (i)<<(ii)<<(iii)≈(iv) (FIG. 9c). By careful inspection of CO electrooxidation peaks of (i)-(iv), as well as the follow-up cycle from FIG. 9c, it is observed that the total charge corresponding to CO electrooxidation is similar for all the cases. The comparison of ESA.sub.CO of FIG. 9d shows that the differences in ESA between treatments (ii)-(iv) are comparatively small. In the case of treatment (i), it can be noted that the obtained charge is most likely a mixture of oxidation of Cu.sub.OPD and Cu.sub.UPD species, as well as CO electrooxidation. In this case, the shape of ORR polarization curve and substantially lower limiting current clearly indicate that ORR partially takes place via the much slower two-electron ORR pathway (FIG. 9a, i=dots), while the poorly resolved H.sub.UPD region in the follow-up cycle after CO electrooxidation (FIG. 9c, dots) additionally confirms the assumption of substantial Cu adatoms coverage. Subsequently, both specific activity (FIG. 9e) and mass activity (FIG. 9f) are very poor for treatment (i).
[0289] In the case of experiment (ii), acid washing was performed 4 times. Each time a certain equilibrium concentration of Cu adatoms species was removed so, progressively, more and more Cu.sub.OPD and Cu.sub.UPD was removed from the catalyst surfaces. An immense improvement can be observed in the case of ORR polarization curve after treatment (ii) compared to treatment (i) (FIG. 9a). From the shape of the ORR polarization curve and improvement in the limiting current value it can be noted that ORR takes place on a much higher surface fraction—via the faster ORR four-electron pathway. However, due to remaining blockage of active surface by Cu adatoms that are most likely still present in significant quantity, both the overpotential and the limiting current are still negatively affected. Thus, the measured ORR activity is still rather moderate, approximately on the order of pure Pt/C catalysts (i.e. nanoparticulate platinum only catalysts supported on carbon). The improved removal of Cu.sub.OPD and Cu.sub.UPD is further confirmed by CO electrooxidation, where a peak is observed at about 0.75 V.sub.RHE, and by the much better resolved H.sub.UPP and Pt hydroxides/oxides regions (FIG. 9c).
[0290] In the case of experiment (iii), which is an embodiment of the present invention, apart from using a 4-time acid washing with 1 M HAc, the suspension was additionally purged by CO. Surprisingly it can be observed that the additional exposure to CO results in a significant further improvement in ORR activity. Without being bound by theory this may be attributed to a very strong binding of CO to platinum atoms at the surface of the platinum-alloy catalyst, thus protecting the platinum from re-deposition of already dissolved copper species. It is believed by the inventors that exposure with CO in addition to the acid washing results in disruption of Cu.sub.UPD formation. This may explain the dramatic further improvement, which is observed for treatment (iii) (treatment iii=Example 1 of the present invention). As seen in FIG. 9a for treatment (iii), the maximum (geometrical) limiting current was achieved already in the 1.sup.st cycle after immersion into electrolyte without any prior cyclic voltammetry. This points towards a complete removal of Cu.sub.OPD and making the four-electron ORR pathway the dominant mechanism.
[0291] After additional 50 cycles of in-situ electrochemical activation via cyclic voltammetry and an exchange of electrolyte (=treatment iv) only a minor improvement in the specific ORR activity measured at 0.9 V.sub.RHE is observed (FIG. 9e), whereas the limiting current remained the same (FIG. 9a, dash).
[0292] This indicates that treatment (iii) according to Example 1 of the present invention results in a nearly complete activation of the as synthesized PtCu.sub.3/C catalyst.
[0293] The fact that still a minor improvement can be achieved when objected to another 50 potential cycles is presumably because a small quantity of O.sub.2 stabilized Cu—(OH) adatoms from leftover Cu.sub.UPD may still block some of the most active sites for ORR in the case of (iii)—slightly affecting the slope of ORR polarization curve, but not the mechanism of the reaction.
[0294] Nevertheless, treatment (iii) provides specific activities and mass activities, which are in the range of electrochemically activated catalysts (iv) (FIGS. 9e and 9f). This activation is sufficient for the catalysts treated with treatment (iii) to be employed for instance in fuel cell, without the need of further electrochemical activation. In contrast to that other chemical activation treatments, such as treatments (i) or (ii) do result in specific activities or mass activities, which are approximately a factor of 4 or 5 below the activities obtained with the treatment (iii) according to Example 1 of the present invention.
[0295] The fact that still a very minor improvement is achieved upon further 50 potential cycles in treatment (iv), furthermore provides evidence, that treatment (iii) does not degrade the catalyst, as otherwise further potential cycling would result in a decrease of catalyst activity. The inventive method therefore, results in an approximately complete activation, however without causing catalyst degradation.
[0296] Moreover, the remaining difference between treatment (iv) and treatment (iii) is attributed to the above mentioned minor leftover from Cu.sub.UPD. This is attributed to the fact that treatment (iii) was performed in a batch reactor and not a continuous/flow reactor.
[0297] The inventors of the present invention have understood that by using a continuous/flow reactor instead of a batch reactor an even further improvement of the excellent chemical activation according to treatment (iii) can be achieved. In a continuous/flow reactor the catalyst is continuously exposed to fresh acidic or basic medium. In this way the less noble metal can be removed even more efficiently from the catalyst surface, thus enabling complete activation.
[0298] FIG. 10 shows a comparison between ex-situ chemical activation performed in 1 M HAc saturated with CO (FIGS. 10a and 10b-Example 1) or 1 M HClO.sub.4 (FIGS. 10c and 10d-Example 2). FIGS. 10a and 10c illustrate CO electrooxidation measurements, while FIGS. 10b and 10d show ORR polarization curves. The CO electrooxidation curves and ORR polarization curves of the PtCu.sub.3/C catalysts after chemical activation correspond to the curves with full-lines, while the curves with dashed-lines belong to the same catalysts after subjected to another 50 potential cycles of in-situ electrochemical activation (0.05 to 1.2 V.sub.RHE, 300 mV.Math.s.sup.−1, 600 rpm).
[0299] FIG. 11 shows Table 1, which summarizes the results of the experiments of FIG. 10. Table 1a shows the values for ESA.sub.CO, SA at 0.9 V.sub.RHE (=jk.sub.CO0.9) and MA at 0.9 V.sub.RHE (=jm.sub.0.9). When using strong acids, such as 1 M HClO.sub.4 (Example 2, Table 1b) instead of less aggressive acids, such as 1 M HAc (Example 1, Table 1a), exposure to CO in combination with four times acid washing can damage the platinum-alloy catalyst. This results in a loss of ESA.sub.CO due to severe particle degradation (FIG. 11b compared to FIG. 11a). This is most likely due to the fact that Pt slowly dissolves in a strong acid such as 1 M HClO.sub.4 in the presence of CO. Thus, using a milder acid, such as 1 M HAc, is beneficial in order to reduce the risk of catalyst degradation. Nevertheless also strong acids or strong bases may be used. However, in this case care must be taken with respect to time of exposure, temperature, repetitions of exposure and other experimental conditions in order to reduce the risk of catalyst degradation. By using mild acids or bases for chemical activation, a less aggressive way for chemical activation is provided, which results in high ESA.sub.CO, SA at 0.9 V.sub.RHE (=jk.sub.CO0.9) and MA at 0.9 V.sub.RHE (=jm.sub.0.9), while catalyst degradation can be easily avoided.
[0300] FIGS. 12 to 14 show preferred embodiments of devices for carrying out the method of the present invention. The devices illustrate how chemical activation and re-activation of platinum-alloy catalysts according to the method of the present invention can be achieved on an industrial scale.
[0301] It is to be understood, that the Figures and description of the Figures are shown for illustration purpose only and as such are not intended to restrict the scope of the invention.
[0302] FIG. 12 shows a device, which is preferably operated batch-wise, while devices shown on FIGS. 13 and 14 can be operated in either batch or continuous way.
[0303] FIG. 12 shows a device (10) for carrying out the method of the present invention, which comprises a batch reactor (3), which is preferably a fixed bed reactor (FBR). The reactor has at least one inlet (3a) and at least one outlet (3b). The platinum-alloy catalyst can be filled into the fixed bed reactor (FBR) or a retaining device therein, prior to the start of the treatment. Alternatively, platinum-alloy particles may be transported to the reactor e.g. through piping (not shown in FIG. 12), e.g. with the aid of air or any other suitable gas, known as pneumatic conveying. Due to small size of the particles of most common platinum-alloy catalysts instead of pneumatic conveying, catalyst particles may be transported to the reactor in form of a liquid suspension. The reactor can comprise means for separating the catalyst from the acidic or basic medium (4). For instance, the reactor comprises a porous member (4) at its bottom, which holds particles to be treated, while at same time allows gas and liquid to pass. It can also serve as filter (4) separating particles from carrier medium during reactor loading operation. Frit or mesh, made of ceramic, metal or polymer or any other suitable material can serve as porous member (4). Reactor enclosure has connections for introduction of gas, liquid and outlet connection for waste fluids. It also has means (not shown on FIG. 12) for loading reactor with particles and removing those particles as part of normal operation, or for inspection and repair as result of reactor blockage or other process interruptions. More than one liquid (e.g. pure water and acid/base) and more than one gas (e.g. carbon monoxide and nitrogen) can be applied in the method of the present invention for treating and thus chemically activating the platinum-alloy catalyst. For this purpose the device comprises a source of an adsorptive gas (1). In FIGS. 12-14 the adsorptive gas is shown to be CO. However, said embodiments are not restricted to CO, but also other adsorptive gases may be used. Moreover the device comprises a source of an acidic or basic medium (2). In FIGS. 12-14 the acidic or basic medium is an acid. However, again said embodiments are not restricted thereto. Each fluid, either gas or liquid, can have either its own connection on reactor enclosure or the connection can be the same for all fluids. Only the latter option is shown on FIG. 12. Separate connections can also be provided for means of particle transport e.g. by mentioned liquid suspension conveying. Chemical activation of platinum particles can be achieved either by operation of reactor in three phase mode, known to those, skilled in the art as trickle bed regime or in two phase mode. It will also be apparent to those, skilled in the art, that operation in trickle bed mode requires proper set of operating conditions, most importantly, but not limited to, proper choice of liquid and gas flow rates, to achieve proper liquid and gas distribution over surface of particles, e.g. that surface of particles is entirely covered by liquid. Also a properly designed liquid distributor (not shown on FIG. 12) is beneficial above the fixed catalyst bed to enable uniform liquid distribution across the complete cross section of bed of particles. Operation in two phase mode is simpler in respect to flow distribution inside particle bed but it may require higher pressures to dissolve more adsorptive gas (such as carbon monoxide) in the liquid and proper means of contacting of two phases prior to introduction to reactor (not shown in FIG. 12). Contacting between the adsorptive gas (e.g. CO) and the acidic or basic medium (e.g. an acid) can be achieved, but not limited to, by aid of static mixers, packed column or simply by using a long enough pipe. It will be apparent to those skilled in the art, that operation at higher pressures requires suitable means of pressure control in the reactor, for example, but not limited to, mechanical back pressure control valves. The device for carrying out the invention is equipped with one or more pumps for introduction of acid and/or water into the reactor. Any kind of pump can be used. Metering pumps such as diaphragm or high pressure piston pumps are however preferred. In case of high pressure operation, high adsorptive gas compressor (not shown on FIG. 12) can be included, but preferably carbon monoxide with high pressure at its source, e.g. in gas cylinder is used. Optionally the device can be equipped with an acid/base regeneration module (5) and/or gas recycle compressor. An acid regeneration module (operating e.g. on electrolytic principle) drastically reduces consumption of fresh acid and amount of generated waste. Both factors can considerably reduce operating cost. Gas recycle compressor recycles any unconverted adsorptive gas back to reactor, reducing both amount of fresh carbon monoxide and amount of generated waste gases. If a recycle compressor is used, need for fresh gas compressor is eliminated. Fresh carbon monoxide is introduced into the device at inlet of recycle compressor (not shown on FIG. 12). If one or both additional options are used, gas-liquid separator is placed at reactor outlet. Any type of gas-liquid separator can be used but one, operating on gravity separation principle is preferred. A preferred method of treatment of a platinum-alloy catalyst, i.e. chemical activation, in the device is as follows: The reactor is loaded with platinum-alloy particles to be treated. As described, loading can be done by pumping a liquid suspension of particles into the reactor and filtering them on a frit, located at the bottom of the reactor enclosure. Simply pouring a liquid suspension into the reactor can also be applied. Appropriate flow rates of acid and adsorptive gas (e.g. CO) are set. In case dissolved adsorptive gas is used, pressure in reactor is preferably also set. Process is continued until predefined time is reached. This predefined time is at least as long as necessary to achieve chemical activation across entire bed of particles. On-line process measurements may also be used to determine end point of chemical activation. After finishing the chemical activation step, the reactor is preferably flushed with demineralized water or other suitable solvent to remove residues of acid and carbon monoxide out of reactor. Inert gas purging or successive vacuuming may also be applied to remove carbon monoxide from reactor volume where liquid cannot reach. Nitrogen can be preferably applied as such inert gas. Other suitable inert and harmless gases can also be applied. The chemically activated platinum-alloy catalyst can be removed from the reactor either manually through an opening provided in reactor enclosure or with aid of the above mentioned transfer means (e.g. by turning treated material into liquid suspension). Preferably this is done by introduction of transfer liquid from bottom of reactor in order to raise particles into suspension by up-flowing liquid. FBR device is preferably used for small production capacities. For larger capacities either mechanically stirred reactor (MSR) device or draft tube reactor (DTR) device, shown on are preferred. Another embodiment for a device (10) for carrying out the inventive method is shown in FIG. 13. The device comprises a mechanically stirred reactor (3) (MSR). MSR reactors are well known and widely used in process industries for applications ranging from mixing, reactions, extractions to crystallization and many more. They are well known to those skilled in the art. Mechanically stirred reactor is in its basic form a vessel, equipped with stirrer. Electrical motor (M) is preferably used as a drive for the stirrer but other types of drive can also be used. Many different well known standard types of stirrers can be employed. Examples are pitched blade turbine, propeller, and Rushton turbine. As will be immediately clear to those, skilled in the art, stirrer types with higher capability to maintain solid particles in suspension are preferred to those with lower capability. Reactor in this embodiment of invention is also equipped with means for introduction and dispersion of gas. Different types of gas introduction and dispersion means can be used, examples being (but not limited to) perforated tubes, porous frits, perforated plates and nozzles. It will be apparent to those skilled in the art, that proper means for gas introduction and dispersion should be designed in such a way, that fine bubbles are obtained and dispersed in a reactor vessel. Such a design increases gas to liquid surface area and this way increases mass transfer of gas into liquid. During chemical activation step at least part of a gas, being introduced into reactor, is composed of adsorptive gas, such as carbon monoxide. Before and after chemical activation, gas can be switched to suitable inert gas (not shown on FIG. 13), with purpose being either to remove air, containing oxygen, prior to starting the entire process of chemical activation (or to remove gases containing the adsorptive gas) after shutting down the process of chemical activation. MSR reactor of preferred embodiment of the invention is equipped with at least two filters (4) located inside (shown in FIG. 13) or outside reactor vessel (not shown in FIG. 13). First option is preferred if reactor is custom built for the process of chemical activation while second one is preferred if existing reactor, built for other application(s) is to be converted for a process of chemical activation of platinum. In the preferred embodiment an acidic or basic medium, e.g. an acid and optionally in addition another liquid, as for example demineralized water for washing and cleaning, is introduced into reactor through one filter and removed through the other. Both filters are connected to a system of at least one stream selection valve. Two three-way valves are shown as an example on FIG. 13. Stream selection valves serve to select which filter(s) serve as reactor inlet filters and which filter(s) serve as reactor outlet filters. Reactor outlet filters are used to remove platinum particles from reactor effluent liquid. In the process platinum-alloy particles, filtered out of suspension, may accumulate on a filter surface and its pores, slowly clogging the filter. On the other hand, liquid flowing through the filter in opposite direction may serve to remove particles, accumulated on the filter. Stream selection valves are used to change direction of flow in filters and in that way regenerate filters which are brought from outlet service into inlet service. Any suitable type of filter can be used. Porous candle type sintered filters can be used. It will be apparent to those, skilled in the art, that some overpressure in reactor is preferred to force liquid through outlet filter(s). This is achieved with gas pressure which is controlled by suitable means, for example (but not limited to) mechanical back pressure valve (not shown on FIG. 13) on reactor effluent gas stream. Level of liquid suspension inside reactor is maintained by level control means. This may be composed of suitable level sensor, controller and final control element (e.g. pump as shown on FIG. 13) or by a mechanical control device, composed of float and valve. Any suitable level sensor may be used but preferably it is suited for three phase operation in a mixture of gas, liquid and solid. As shown in FIG. 13, a pump is preferred as means of acid pressure increase but other means e.g. gravity or acid tank (not shown on FIG. 13) gas pressurization may be used as well. As in case of FBR device and with same aim and benefits, acid regeneration loop (5) and gas recycle loop may be applied. Reactor can be equipped with suitable means of heating and/or cooling (not shown on FIG. 13) which enable to control the temperature inside reactor. The MSR of the device can be operated either in batch or continuous mode but continuous mode is preferred. In batch mode platinum-alloy particles to be treated may be loaded into reactor manually through suitable opening in reactor vessel (not shown in FIG. 13) or preferably introduced into reactor in form of a liquid suspension as shown in FIG. 13. The suspension can be prepared in a suspension preparation tank and pumped into reactor with a pump. Other means such as gravity or gas pressurization may also be applied. It will be apparent to those, skilled in the art, if a pump is to be used, it should be suitable for suspension, e. g. centrifugal pump. A suspension of chemically treated platinum-alloy catalyst particles is removed from reactor into one or more filters where it is collected as a product. If process is to be operated in continuous mode, preferably more than one filter is used. This enables one filter to be in service, while other filter(s) are in different stages of operation (e.g. unloading, washing and/or drying). Instead of batch type filter(s) one continuous filter like drum or try filter may be applied in continuous mode of operation. Removal of product particles may also be achieved by other mechanical means, e.g. by different types of centrifuges. MSR embodiment of the invention has advantage of being mainly built from standard equipment, readily available on the market. Existing equipment, originally made for other applications may be easily converted to chemical activation process according to MSR embodiment.
[0304] The third preferred embodiment of a device (10) for carrying out the inventive method, comprises a draft tube reactor (3) (DTR) and is shown in FIG. 14. It is similar to the reactor for hydrogenation of working solution in hydrogen peroxide manufacture, as disclosed in the patent GB 718307 A. The device of FIG. 14 can be operated in batch or continuous mode. The reactor in this embodiment is preferably composed of reactor enclosure with connections, internal filter(s) (4), gas introduction and dispersion means and draft tube(s). Each draft tube is located above gas introduction and dispersion device known as gas sparger. During normal operation, the majority of the reactor volume is occupied by a suspension containing platinum-alloy catalyst particles to be treated. The adsorptive gas is introduced through the sparger at bottom of draft tube, stream of finely dispersed gas bubbles is created inside the tube. This lowers overall density of mixture inside draft tube. Heavier suspension without bubbles, located outside draft tube(s) has a tendency to expel lighter mixture inside and force it upwards the draft tube(s). In this way circular motion is created inside reactor vessels, where mixture moves in upward direction inside draft tube(s) and downward direction in space between vessel and tube(s). Fresh acidic or basic medium, such as an acid, is introduced into the system through gas sparger or other connection or nozzle, located at bottom part of draft tube. In the following we refer to an acid as shown in the Figure. The embodiment is however not restricted to the use of acids. Platinum-alloy catalyst particles to be treated move inside reactor in close similarity to circular liquid motion and are in this way repeatedly exposed to fresh acid being introduced at bottom of draft tube. It was surprisingly discovered that repeated change of fresh acid is particularly advantageous for smooth chemical activation process. This makes this embodiment advantageous for treatment of particles on large scale or where efficient treatment of particles with high final quality is desired. Liquid reactor effluent is treated in filters to remove product particles from waste stream. As in case of previous two embodiments part of the acid may be regenerated and introduced back into reactor. Also part of the adsorptive gas may be recycled to the bottom by aid of compressor. Both additions serve to lower consumption of fresh materials and amount of wastes generated and in turn reducing cost. As in case of MSR device, different means of particle introduction into reactor and removal of product particles from reactor may be applied. Pumping of particle suspension from suspension preparation tank(s) and removing particles by suitable system of filters, as shown in FIG. 14 are however preferred. During chemicals activation phase at least part of gas, introduced into reactor is composed of adsorptive gas, such as carbon monoxide. In start-up and shut-down phases the adsorptive gas is replaced by inert gas such as nitrogen (not shown on FIG. 14) which serves to remove oxygen or carbon monoxide from reactor and in this way helps to prevent formation of potentially explosive mixtures. Acid may be replaced by water or other suitable solvent in start-up phase to fill reactor. During shut down phase acid replacement with water (or other suitable solvent) helps to remove acid and other soluble components from reactor.
Examples and Reference Examples
[0305] In the following the inventive method is further explained by means of examples and reference examples:
[0306] Example 1: 5 grams of as prepared PtCu.sub.3/C electrocatalyst was dispersed in 1 L of 1 M HAc (Merck, glacial, 100%) and ultrasonicated for 2 minutes in an ultrasound bath (Iskra, sonis 4). Afterwards, a magnetic stirrer was added in order to mix the suspension during the ex-situ CA protocol. The suspension was stirred under CO saturation for 30 minutes. At the end of the acid wash, the electrocatalyst was filtered and redispersed in 1 L of fresh 1 M HAc (Merck, glacial, 100%)—repeating the same acid wash process 3 more times. After the last filtration, the electrocatalyst was redispersed in 1 L of Mili-Q water and heated until boiling while stirring. This process of Mili-Q water washing was repeated 3 more times in order to ensure a complete neutralization. After the last filtration, the electrocatalysts were left to dry at 50° C. overnight.
[0307] Example 2: 50 mg of as prepared PtCu.sub.3/C electrocatalyst was dispersed in 50 mL of 1 M HClO.sub.4 (Merck, suprapur) and ultrasonicated for 2 minutes in an ultrasound bath (Iskra, sonis 4). Afterwards, a magnetic stirrer was added in order to mix the suspension during the ex-situ CA protocol. The suspension was stirred under CO saturation for 30 minutes. At the end of the acid wash, the electrocatalyst was filtered and redispersed in 50 mL of fresh 1 M HClO.sub.4 (Merck, suprapur)—repeating the same acid wash process 3 more times. After the last filtration, the electrocatalyst was redispersed in 200 mL of Mili-Q water and heated until boiling while stirring. This process of Mili-Q water washing was repeated 3 more times in order to ensure a complete neutralization. After the last filtration, the electrocatalysts were left to dry at 50° C. overnight.
[0308] Reference Example 1: 50 mg of as prepared PtCu.sub.3/C electrocatalyst was dispersed in 50 mL of 1 M HAc (Merck, glacial, 100%) and ultrasonicated for 2 minutes in an ultrasound bath (Iskra, sonis 4). Afterwards, a magnetic stirrer was added in order to mix the suspension during the ex-situ CA protocol. The suspension was stirred for 30 minutes. At the end of the acid wash, the electrocatalyst was filtered and redispersed in 50 mL of fresh 1 M HAc (Merck, glacial, 100%)—repeating the same acid wash process 3 more times. After the last filtration, the electrocatalyst was redispersed in 200 mL of Mili-Q water and heated until boiling while stirring. This process of Mili-Q water washing was repeated 3 more times in order to ensure a complete neutralization. After the last filtration, the electrocatalysts were left to dry at 50° C. overnight.
[0309] Reference Example 2: 50 mg of as prepared PtCu.sub.3/C electrocatalyst was dispersed in 50 mL of 1 M HAc (Merck, glacial, 100%) and ultrasonicated for 2 minutes in an ultrasound bath (Iskra, sonis 4). Afterwards, a magnetic stirrer was added in order to mix the suspension during the ex-situ CA protocol. The suspension was stirred for 30 minutes. At the end of the acid wash, the electrocatalyst was filtered and redispersed in 200 mL of Mili-Q water and heated until boiling while stirring. This process of Mili-Q water washing was repeated 3 more times in order to ensure a complete neutralization. After the last filtration, the electrocatalysts were left to dry at 50° C. overnight.
[0310] Reference Example 3: 0.5 g of as prepared PtCu.sub.3/C electrocatalyst was dispersed in 0.5 L of 1 M HAc (Merck, glacial, 100%) and ultrasonicated for 2 minutes in an ultrasound bath (Iskra, sonis 4). Afterwards, a magnetic stirrer was added in order to mix the suspension during the ex-situ CA protocol. 50 mL aliquot of the suspension was pipetted out of the reaction mixture at t=2, 4 and 6 hours. Each aliquot of the suspension was filtered and redispersed in 200 mL of Mili-Q water and heated until boiling while stirring. This process of Mili-Q water washing was repeated 3 more times in order to ensure a complete neutralization. After the last filtration, the electrocatalysts were left to dry at 50° C. overnight. The remaining aliquot of the suspension was filtered and redispersed in fresh 350 mL of 1 M HAc (Merck, glacial, 100%) and ultrasonicated for 2 minutes in an ultrasound bath (Iskra, sonis 4). The suspension in fresh acid was stirred for 30 minutes. At the end of the acid wash, the electrocatalyst was filtered and redispersed in 200 mL of Mili-Q water and heated until boiling while stirring. This process of Mili-Q water washing was repeated 3 more times in order to ensure a complete neutralization. After the last filtration, the electrocatalysts were left to dry at 50° C. overnight.
[0311] The description made with reference to exemplary embodiments does not restrict the invention to these embodiments. Rather, the invention encompasses any novel feature and any combination of features, including in particular any combination of features in the claims, even if this feature or this combination is not itself explicitly indicated in the claims or exemplary embodiments.
