Hollow platinum nanoparticles for fuel cells

Abstract

The present invention relates to hollow platinum nanoparticles with a diameter comprised between 3 and 20 nm which comprise a first central cavity and optionally at least one second cavity at the periphery of the first cavity, the shell of which is dense and single-crystal with a thickness comprised between 0.2 and 5 nm. The invention also relates to a method for manufacturing such nanoparticles, as well as to their use as an electrocatalyst in fuel cells.

Claims

1. A method for manufacturing hollow platinum nanoparticles comprising a dense and single-crystal shell having a thickness of from 0.2 nm to 5 nm and a first central cavity and optionally at least one second cavity at a periphery of the first central cavity, wherein the nanoparticle has a diameter in a range of from 3 nm to 20 nm, is quasi-spherical in shape, and does not comprise a non-noble metal alloy element, the method comprising the following steps: a) placing Pt.sub.xM/C nanoparticles into a polarization device, where: Pt is platinum; x is in a range of from 0.2 to 10; M is a metal less noble than platinum; C is carbon; and Pt and M are supported on the carbon; b) subjecting the nanoparticles for a plurality of times to a sequence of polarizations between the two following potentials: a reducing potential of from 0 to 0.5 volts relative to a reversible hydrogen electrode (V/RHE), and an oxidizing potential of from 0.6 to 1.4 V/RHE, until single-crystal hollow platinum nanoparticles are obtained having a diameter of from 3 nm to 20 nm and a dense shell having a thickness of from 0.2 nm to 5 nm, wherein the polarization device is a rotary conducting drum, and the step b) is carried out in an electrolytic solution.

2. The method according to claim 1, wherein step b) is carried out until single-crystal hollow platinum nanoparticles are obtained with a diameter of from 3 nm to 10 nm and a dense shell having a thickness of from 0.277 nm to 0.831 nm.

3. The method according to claim 1, wherein the sequence of polarizations of step b) is carried out with: a succession of linear ramps of rising and falling potentials at a potential scanning rate in a range of 1 to 1,000 mV s.sup.1; a succession of plateaus of potentials with a duration of from 1 to 36,000seconds; or a mixture of ramps and plateaus for rising and falling potential phases.

4. The method according to claim 1, wherein the sequence of polarizations has a rectangular profile, for which: the reducing potential is in a range of 0.05 V/RHE to 0.5 V/RHE; and the oxidizing potential is in a range of 0.8 to 1 V/RHE.

5. The method according to claim 1, wherein the Pt.sub.xM/C nanoparticles are obtained by the method comprising: a) forming colloidal nanoparticles of the metal M on a carbon support from a M.sup.z+metal salt to obtain M/C nanoparticles, and b) depositing platinum on the M/C nanoparticles obtained at the end of step a).

6. The method according to claim 1, further comprising after step b): c) cleaning the nanoparticles in an acidic solution.

7. The method according to claim 6, further comprising after step c): c) heating the nanoparticles in an oven at a controlled temperature in a range of 100 C. to 600 C. for 10 minutes to 10 hours.

8. The method according to claim 6, further comprising after step c): c) subjecting the nanoparticles to a sequence of polarizations in an electrolytic solution that is different from the electrolytic solution of step b).

9. The method according to claim 7, wherein step c) is carried out under a controlled atmosphere which is neutral, oxidative, or reductive.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 is a graph indicating the interatomic distance (expressed in nanometers) of the nanoparticles of a cathodic electrocatalyzer versus the duration of use of a PEMFC during actual operation.

(2) FIG. 2 is a graph expressing the electrocatalytic activity of an electrocatalyst comprising either hollow platinum particles obtained after more than 1,000 hours of use of a PEMFC during actual operation, or nanoparticles of commercial electrocatalysts of the Pt/C type and of the Pt.sub.3Co/C type.

(3) FIG. 3 illustrates an X-ray diffraction spectrum of hollow platinum nanoparticles according to the invention.

DETAILED DESCRIPTION OF EMBODIMENTS

(4) Hollow platinum nanoparticles were obtained in the following way:

(5) A PEMFC was used for 3,500 hours under the following conditions: cathodic electrocatalyst of the Pt.sub.3Co/C type, humidified air (65% relative humidity, 1.1 bars absolute, stoichiometry 2.5); anodic electrocatalyst: pure dihydrogen (0% relative humidity, 1.3 bars absolute, blocked mode (dead-end mode) stoichiometry 2.5); temperature of 70 C.; 50 ampere intensity current.

(6) The cathodic electrocatalyst was observed with a transmission electron microscope in a scanning mode with a wide angle annular dark field detector (STEM-HAADF).

(7) STEM-HAADF is a technique allowing characterization of bimetal nanoparticles, since it provides images including chemical information, called a Z contrast image (Z being the atomic number of the element).

(8) In this imaging mode, the intensity of the image (I) of an atomic column is in a first approximation, proportional to the product of the number (N) of the atoms present in the atom column and to the square of the average atomic number Z of the atoms making it up:
I.sub.HAADF=kNZ.sup.Equation 2
wherein k is a constant and 2.

(9) Analysis of the cathodic electrocatalyst after about 1,000 hours of operation of the PEMFC has shown the presence of two types of nanoparticles: Some of the nanoparticles had a bright shell covering a darker core, generally with a large intensity difference between the core and the shell. These are hollow particles. The remainder of the nanoparticles had a structure similar to that of the cathodic electrocatalyst of the Pt.sub.3Co/C type (homogeneous HAADF contrast). These are solid particles.

(10) After 3,500 hours of operating of the PEMFC, the fraction of hollow platinum nanoparticles had been evaluated to about 35% of the total number of nanoparticles. This gives the possibility of inferring from a structural point of view, that these nanoparticles are stable on the long term under the real operating conditions of a PEMFC.

(11) Even after 3,500 hours of operation of the PEMFC, there was always coexistence of hollow nanoparticles and of solid nanoparticles. This is explained by the fact that a restricted fraction of electrocatalyst (of the order of 20% to 50%) is actually active for the cathodic reaction. Only the particles meeting the triple contact condition: i.e. the nanoparticles in contact with the carbon and the ionomer are capable of being active. To this limitation are added other limitations related to the dynamic operating conditions of the PEMFC.

(12) This is why, within the scope of the present invention, it is particularly advantageous to manufacture hollow platinum nanoparticles, for example as this was described above in the embodiment of the invention via a chemical route, before incorporating them into a membrane-electrodes assembly rather than making in situ nanoparticles from an electrocatalyst of the Pt.sub.xM/C type, a so-called commercial electrocatalyst. In this way, the electrocatalyst only comprises hollow platinum nanoparticles, which will probably be never the case if an electrocatalyst of the Pt.sub.xM/C type, a so-called commercial type, is used, already applied in an electrode or in a membrane-electrodes assembly.

(13) Further, the contraction of the lattice parameter of the shell of the hollow platinum nanoparticles was determined.

(14) No relaxation of the lattice parameter of the platinum was observed up to durations close to 3,500 hours.

(15) The graph of FIG. 1 shows that the lattice parameter of the shell of the hollow platinum nanoparticles is slightly contracted relatively to that of platinum nanoparticles of equivalent size. This contraction is maintained on the long term during the use of the PEMFC during actual operation.

(16) Thus, the hollow platinum nanoparticles, once they are formed, are stable over a period of 2,500 hours, while solid platinum nanoparticles of the Pt/C type or of the Pt.sub.xM/C type would be greatly modified under identical conditions.

(17) Measurement of the Electrocatalytic Activity:

(18) Measurement of the specific activity (referred to 1 cm.sup.2 of platinum) for the oxygen reduction reaction was conducted by using a conventional electrocatalysis methodology.

(19) To do this, a film of a thickness of a few micrometers consisting of the electrocatalyst, as detailed above, to be tested, was immobilized at the surface of a glassy carbon electrode.

(20) This film was formed by drying a mixture containing in identical mass proportions (50-50 wt %) the electrolyzer and the Nafion ionomer in ultrapure water. Thus, the electrocatalyst and ionomer were homogeneously distributed and made the test comparable with an electrocatalytic layer of a PEMFC electrode.

(21) The catalytic activities were measured under identical conditions allowing direct comparison of the results for different electrocatalysts: E=0.85 V/RHE Aqueous electrolyte: 0.1 mol.L.sup.1 of sulfuric acid T=25 C. v=1 mV s.sup.1.

(22) FIG. 2 is a graph comparatively showing the catalytic activity of: the electrocatalyst containing hollow platinum nanoparticles obtained after 1,000 hours of use of a PEMFC during actual operation, the cathodic electrocatalyst being of the Pt.sub.3Co/C type (referenced on the graph as: Hollow Pt nanoparticles) as described above for FIG. 1, a commercial electrocatalyst of the Pt.sub.3Co/C type (referenced on the graph as: Pt.sub.3Co/C), an electrocatalyst of the Pt/C type (referenced on the graph as: Pt/C),
and this depending on the interatomic distance of said electrocatalysts (as determined from X diffractograms).

(23) According to FIG. 2, the electrocatalyst comprising hollow platinum nanoparticles has better catalytic activity than commercial electrocatalysts of the Pt/C or Pt.sub.3Co/C type.

(24) It should be noted that this experimental result would be exacerbated upon considering the mass activity (as referred to the platinum mass).

(25) Hollow platinum nanoparticles according to the invention were made in the following way:

(26) Nanoparticles on a carbon support of the Pt.sub.3Co/C type were incorporated into a membrane-electrodes assembly and were used as an electrocatalyst at the cathode of a PEMFC.

(27) The initial average size of these nanoparticles was comprised between 3 and 5 nm and the distribution of the platinum and cobalt atoms within these nanoparticles was homogeneous. These nanoparticles were initially dense.

(28) A sequence of polarizations was carried out on said nanoparticles between the two following potentials: the selected reducing potential applied for 30 minutes was 0.2 V (this corresponds to the cell voltage or potential difference between the cathode and the anode, during operation with hydrogen (at the anode)/nitrogen (at the cathode). At this voltage, the potential of the cathode approximately had a value of 0.2 V/RHE (i.e. measured relatively to a reversible hydrogen electrode). the oxidizing potential was reached by carrying out 2 potential scanning cycles between E=0.05 V and E=1.23 V at a rate v=0.020 V/s.

(29) This alternation of reducing and oxidizing potentials was repeated 130 times for an approximate period of 60 hours.

(30) Hollow platinum nanoparticles with an approximate diameter of 5 nm, close to the initial size of the nanoparticles, were thereby obtained.

(31) The shell was dense with an approximate thickness of 1.5 nm and was confirmed by STEM-HAADF microscopy observations.

(32) The X-ray diffraction measurements of FIG. 3 also confirm the formation of these hollow nanoparticles with an average interatomic distance of 0.2758 nm, smaller (contraction of the lattice parameter) than the one measured on a platinum nanoparticle of identical size (0.2775 nm).

(33) Table 1 below indicates that the specific activity of hollow platinum nanoparticles according to the invention for reducing oxygen (SA.sub.0.85V/Acm.sup.2.sub.Pt) is better than that of Pt.sub.3Co/C nanoparticles.

(34) TABLE-US-00001 TABLE 1 Hollow platinum Pt.sub.3Co/C nanoparticles according nanoparticles to the invention SA.sub.0.85 V (A cm.sup.2.sub.Pt) 120 130 d.sub.V (nm) 5.3 5.7 Specific activity for the oxygen reduction reaction measured in a liquid electrolyte of 0.1 mol. L.sup.1 of sulfuric acid at E = 0.85 V/RHE (SA.sub.0.85 V) and with a nanoparticle size (d.sub.V)

(35) Hollow platinum nanoparticles 1 to 3 were made according to the manufacturing method of the invention.

(36) More precisely, nanoparticles of PtNi/C were used as starting materials and were obtained via the chemical route that was detailed above as an embodiment for obtaining Pt.sub.xM/C nanoparticles.

(37) The PtNi/C nanoparticles were made by co-reduction of the salts Pt(NH.sub.3).sub.4Cl.sub.2 and NiCl.sub.2 by NaBH.sub.4 in aqueous medium.

(38) The steps of the synthesis were the following: 0.3 g of powder of carbon support (Vulcan XC-72 from Cabot) with 140 mL of deionized ultrapure water (18.2 M cm, quantity of total organic impurities less than 5 ppb) and 10 mL of ethanol were mixed together under magnetic stirring; 0.46 mmol of Pt(NH.sub.3).sub.4Cl.sub.2 and 0.46 mmol of NiCl.sub.2 were added to the powder of carbon in suspension with the solvents; Reductive agent was added, i.e. 5 mL of a solution of NaBH.sub.4 with a concentration of 0.22 mol.L.sup.1; The mixture was stirred under magnetic stirring during one hour; It was filtered and rinsed with deionized ultrapure water; It was dried at a temperature of 110 C. during 45 minutes.

(39) The nature of the salts conducts to a sequential reduction. The salt of nickel is firstly reduced by NaBH.sub.4 and then, the reduction of the salts of platinum and the galvanic exchange of nickel atoms previously formed by the ions Pt.sup.2+ occur simultaneously that conducts to the formation of hollow PtNi/C nanoparticles (with a low amount of nickel).

(40) The nanoparticles 1 were obtained at the end of step b) of the manufacturing method according to the invention that was carried out on a thin active layer deposited on the tip of a rotating disk electrode, according to the following parameters: Under an atmosphere of argon; Electrolyte comprising a solution of 0.1 mol.L.sup.1 of sulfuric acid; 20 cycles at 20 mVs.sup.1 with a reducing potential of 0 V/RHE and an oxidizing potential of 1.23 V.RHE (total duration of 41 minutes).

(41) The nanoparticles 2 were obtained at the end of step b) of the manufacturing method according to the invention that was carried out on a thin active layer deposited on the tip of a rotating disk electrode, according to the following parameters: Under atmosphere of argon; Electrolyte comprising a solution of 0.1 mol.L.sup.1 of sulfuric acid; 60 cycles at 20 mVs.sup.1 with a reducing potential of 0 V/RHE and an oxidizing potential of 1.23 V.RHE (total duration of 123 minutes).

(42) The nanoparticles 3 were obtained at the end of step b) of the manufacturing method according to the invention that was carried out on a thin active layer deposited on the tip of a rotating disk electrode, according to the following parameters: Under an atmosphere of dioxygen; 4 cycles at 2 mVs.sup.1 with a reducing potential of 0.4 V/RHE and an oxidizing potential of 1.05 V.RHE (total duration of 43.3 minutes).

(43) In the difference of nanoparticles 1 and 2, for these nanoparticles 3, step b) was preceded with an acidic washing during 12 hours in a solution of sulfuric acid at a concentration of 1 mol.L.sup.1 that finalized the above detailed chemical route. This last step of washing was destined to eliminate the residual nickel atoms which were present on the surface of the carbon support and not covered with platinum atoms or present in the shell of the hollow nanoparticles.

(44) The catalytic activities of these nanoparticles 1 to 3 were measured under identical conditions allowing a direct comparison of the results: Aqueous electrolyte: 0.1 mol.L.sup.1 of sulfuric acid; T=25 C.; v=2 mV s.sup.1; Measurement at E=0.85 V/RHE and E=0.9 V/RH.

(45) The specific activity of these hollow platinum nanoparticles 1 to 3 for the oxygen reduction was on average: 550 A cm.sup.2.sub.pt (measured at E=0.85 V/RHE), against 180 A cm.sup.2.sub.pt for solid Pt/C nanoparticles from E-TeK (i.e. a commercial catalyst of the reference E-TeK) with an identical carbon support Vulcan XC-72 from Cabot, m.sub.pt/(m.sub.pt+m.sub.c)=40%) and 200 A cm.sup.2.sub.pt (measured at E=0.9 V/RHE), against 50 A cm.sup.2.sub.pt for solid Pt/C nanoparticles from E-TeK (i.e. a commercial catalyst of the reference E-TeK) with an identical carbon support: Vulcan XC-72 from Cabot, m.sub.pt/(m.sub.pt+m.sub.c)=40%).

(46) Thus, the specific activity of the hollow platinum nanoparticles 1 to 3 for the oxygen reduction is higher than: the specific activity of solid Pt/C nanoparticles, by a factor from about 3 to 4, and the specific activity of Pt.sub.3Co/C nanoparticles (see Table 1 above).