CARBON SUPPORTED CATALYST COMPRISING A MODIFIER AND PROCESS FOR PREPARING THE CARBON SUPPORTED CATALYST

Abstract

The invention is related to a carbon supported catalyst comprising a carbon-comprising support with a BET surface area in a range from 400 m.sup.2/g to 2000 m.sup.2/g, a modifier comprising at least one mixed metal oxide, comprising niobium and titanium, and/or a mixture, comprising niobium oxide and titanium oxide, a catalytically active metal compound, wherein the catalytically active metal compound is platinum or an alloy comprising platinum and a second metal or an intermetallic compound comprising platinum and a second metal, the second metal being selected from the group consisting of cobalt, nickel, chromium, copper, palladium, gold, ruthenium, scandium, yttrium, lanthanum, niobium, iron, vanadium and titanium.

The invention is further related to a process for preparing the carbon supported catalyst.

Claims

1.-20. (canceled)

21. A carbon supported catalyst comprising a carbon-comprising support with a BET surface area in a range from 400 m.sup.2/g to 2000 m.sup.2/g, a modifier comprising: at least one mixed metal oxide, comprising niobium and titanium; a mixture comprising niobium oxide and titanium oxide; or at least one mixed metal oxide comprising niobium and titanium and a mixture comprising niobium oxide and titanium oxide; a catalytically active metal compound, wherein the catalytically active metal compound is platinum or an alloy comprising platinum and a second metal or an intermetallic compound comprising platinum and a second metal, the second metal being selected from the group consisting of cobalt, nickel, chromium, copper, palladium, gold, ruthenium, scandium, yttrium, lanthanum, niobium, iron, vanadium and titanium, wherein the carbon supported catalyst comprises 0.5% to 20% by weight of niobium and 0.5% to 10% by weight of titanium.

22. The carbon supported catalyst according to claim 21, wherein the ratio of the molar amount of niobium comprised in the carbon supported catalyst to the sum of the molar amount of niobium and the molar amount of titanium comprised in the carbon supported catalyst is in a range from 0.01 to 0.5.

23. The carbon supported catalyst according to claim 21, wherein the carbon supported catalyst comprises 10% to 50% by weight of platinum.

24. The carbon supported catalyst according to claim 21, wherein the catalytically active metal compound is present in form of nanoparticles.

25. The carbon supported catalyst according to claim 21, wherein the modifier consists of niobium, titanium and oxygen.

26. The carbon supported catalyst according to claim 21, wherein all metal comprised in the carbon supported catalyst is comprised in the modifier and in the catalytically active metal compound.

27. The carbon supported catalyst according to claim 21, wherein the carbon-comprising support comprises carbon black, graphene, graphite, activated carbon or carbon nanotubes.

28. An electrode comprising the carbon supported catalyst according to claim 21.

29. A fuel cell comprising the electrode according to claim 28.

30. A process for preparing the carbon supported catalyst according to claim 21, comprising the following steps: (a) precipitating the modifier onto the surface of the carbon-comprising support by preparing an initial mixture, comprising the carbon-comprising support with a BET surface area in a range from 400 m.sup.2/g to 2000 m.sup.2/g, at least two metal oxide precursors, a first precursor comprising niobium and a second precursor comprising titanium, and a solvent, and drying of the initial mixture to obtain an intermediate product, or heating the initial mixture to a temperature at which the initial mixture is boiling, followed by filtration, (b) loading of the catalytically active metal compound in form of particles onto the surface of the intermediate product in a liquid medium by deposition, precipitation and/or reduction of a catalytically active-metal-comprising precursor with a reducing agent, (c) heat treatment of the catalyst precursor resulting from step (b) at a temperature of at least 200° C. in a reducing atmosphere.

31. The process according to claim 30, wherein the initial mixture comprises an acid.

32. The process according to claim 31, wherein the acid is a carboxylic acid.

33. The process according to claim 30, wherein the drying in step (a) is carried out as spray-drying.

34. The process according to claim 30, wherein the drying is carried out with an inert drying gas.

35. The process according to claim 30, wherein at least one of the metal oxide precursors is an alcoholate selected from the group consisting of ethanolate, n-propanolate, iso-propanolate, n-butanolate, iso-butanolate and tert-butanolate, or at least one of the metal oxide precursors is a chloride.

36. The process according to claim 30, wherein the solvent is an alcohol, a carboxylate ester, acetone or tetrahydrofuran.

37. The process according to claim 30, wherein after filtration the intermediate product is washed with a washing liquid comprising a solvent.

38. The process according to claim 37, wherein the solvent used for washing is the same solvent as in the initial mixture.

39. The process according to claim 37, wherein the washing liquid additionally comprises an acid, preferably a carboxylic acid.

40. The process according to any of claim 30, wherein a washing step using water as washing liquid is carried out before carrying out step (b).

Description

EXAMPLES AND COMPARATIVE EXAMPLES

I. Preparation of carbon supported catalysts

EXAMPLES

[0074] Inventive catalysts with three different degrees of niobium doping in titanium oxide were prepared. The ratio of the molar amount of niobium comprised in the carbon supported catalyst to the sum of the molar amount of niobium and the molar amount of titanium comprised in the carbon supported catalyst (n.sub.Nb/(n.sub.Nb+n.sub.Ti)) was namely 0.08, 0.05 and 0.46, respectively for examples 1 to 3.

Example 1

1a) Precipitation of Mixed Niobium Titanium Oxide onto Carbon

[0075] A mixture was prepared from 60 g carbon (Black Pearls® 2000, Cabot), 455 g acetic acid with a purity of 100%, 676 g 2-propanol with a purity of 99.7%, 10.4 g niobium(V)ethoxide with a purity of 99.95%, based on the metal content, and 100 g titanium(IV)n-butoxide with a purity of 99%. In order to homogenize the components, ultra-sonication was for applied for 10 minutes. The mixture was dried in a spray-dryer. In order to prevent sedimentation, the mixture was agitated while being conveyed into the spray-tower. The flow rate of the mixture to be spray-dried was 636 g/h, the diameter of the nozzle of the spray-dryer was 1.4 mm, the nozzle pressure was 3,5 bar absolute, the nozzle gas was nitrogen, the volume flow of the nozzle gas was 3.5 Nm.sup.3/h, the temperature of the nozzle gas was room temperature, the drying gas was nitrogen, the volume flow of the drying gas was 25 Nm.sup.3/h, the temperature of the drying gas was 190° C. and the residence time in the spray-dryer was 15 seconds. For particle separation, a cyclone was applied, which is able to separate particles with a diameter of at least 10 μm. The temperature in the cyclone, corresponding to the exhaust gas temperature of the spray-dryer, was 102° C. to 104° C. All above-described production steps were carried out with exclusion of humidity. No extra water was added in any of the above-described production steps and the mixture to be spray-dried was prepared under nitrogen atmosphere.

[0076] An elementary analysis showed a niobium content of 1.3% by weight and a titanium content of 6.5% by weight, referring to the spray-dried particles. During drying in an air stream at 180° C. for analytic purposes, a mass loss of 28.7% by weight was determined.

1b) Washing

[0077] Residue organic compounds were removed by washing. 71 g of the solid obtained in step 1a) was put on a filter and water was added. A total volume of 7 L of water was used for the washing. Subsequently, the washed solid was dried in a vacuum oven at 80° C. for 10 hours.

[0078] An elementary analysis showed a niobium content of 1.7% by weight and a titanium content of 7.7% by weight, referring to the washed and dried solid. During drying in an air stream at 180° C. for analytic purposes, a mass loss of 12.2% by weight was determined.

1c) Deposition of Platinum

[0079] For the deposition of platinum, 15 g of the solid obtained in step 1 b) were suspended in 412 mL water by means of an ULTRA-TURRAX®. Then, a solution of 10.95 g platinum(II)nitrate in 161 mL water was added. Under stirring, a mixture of 354 mL ethanol and 487 mL water was added and the suspension was heated to 82° C. After six hours at 82° C., the suspension was cooled to room temperature, filtered and the solid residue was washed with 6 L water. The resulting solid was dried in a vacuum oven at 80° C.

1d) Heat Treatment at 800° C.

[0080] 12 g of the solid resulting from step 1c) were heat treated in a rotary tube furnace. In a stream comprising 95% by volume of nitrogen and 5% by volume of hydrogen, the temperature was raised by 10 Kelvin per minute to 800° C. When the temperature of 800° C. was reached, the temperature was kept constant for one hour. Subsequently, the interior of the furnace was cooled to room temperature and at a temperature below 50° C., the gas stream was switched to a stream comprising 100% by volume nitrogen. Then, the heat treated solid was passivated for 12 hours with a gas stream comprising 9% by volume air and 91% by volume nitrogen to form the carbon supported catalyst. Air typically comprises approximately 78% by volume nitrogen and 21% by volume oxygen.

[0081] By elementary analysis, a niobium content of 1.0% by weight, a titanium content of 5.8% by weight and a platinum content of 33% by weight, referring to the carbon supported catalyst, was determined.

[0082] The carbon supported catalyst was further analyzed by powder X-ray diffractometry. The average crystallite size of the platinum comprised in the carbon supported catalyst was calculated from the powder X-ray diffractometry results applying the Scherrer formula. A bimodal distribution of 3.2 nm and 32 nm was determined for the platinum crystallite. This integral method, combined with TEM results, indicated that most of the platinum particles were of a small size of approximately 3 nm and, in addition, a group of larger platinum particles with an average crystallite size of approximately 32 nm was present. Further, a crystallographic phase of TiO.sub.2 (anatase) was observed in the carbon supported catalyst by powder X-ray diffractometry.

[0083] FIG. 1 shows pictures obtained by transmission electron microscopy (TEM) of the carbon supported catalyst produced in example 1. The transmission electron microscopy was coupled with energy-dispersive X-ray spectroscopy (EDX) analysis. A first image (high angle annular dark field, HAADF), provides an overview of the distribution of material density, referring to the electron density, in the sample. Platinum shows the highest contrast, grey areas are assigned to carbon and oxides of niobium and titanium. In three further images, the distribution of the single elements niobium, platinum and titanium are represented separately. For all images the given comparative scale is 90 nm. The elements platinum, niobium and titanium were homogeneously dispersed on the surface of the carbon supported catalyst.

Example 2

2a) Precipitation of Mixed Niobium Titanium Oxide onto Carbon

[0084] A mixture was prepared from 60 g carbon (Black Pearls® 2000, Cabot), 455 g acetic acid with a purity of 100%, 676 g 2-propanol with a purity of 99.7%, 4.92 g niobium(V)ethoxide with a purity of 99.95%, based on the metal content, and 100 g titanium(IV)n-butoxide with a purity of 99%. In order to homogenize the components, ultra-sonication was for applied for 10 minutes. The mixture was dried in a spray-dryer. In order to prevent sedimentation, the mixture was agitated while being conveyed into the spray-tower. The flow rate of the mixture to be spray-dried was 743 g/h, the diameter of the nozzle of the spray-dryer was 1.4 mm, the nozzle pressure was 3.5 bar absolute, the nozzle gas was nitrogen, the volume flow of the nozzle gas was 3.5 Nm.sup.3/h, the temperature of the nozzle gas was room temperature, the drying gas was nitrogen, the volume flow of the drying gas was 25 Nm.sup.3/h, the temperature of the drying gas was 190° C. and the residence time in the spray-dryer was 15 seconds. For particle separation, a cyclone was applied, which is able to separate particles with a diameter of at least 10 μm. The temperature in the cyclone, corresponding to the exhaust gas temperature of the spray-dryer, was 101° C. to 104° C. All above-described production steps were carried out with exclusion of humidity. No extra water was added in any of the above-described production steps and the mixture to be spray-dried was prepared under nitrogen atmosphere.

[0085] An elementary analysis showed a niobium content of 0.6% by weight and a titanium content of 5.6% by weight, referring to the spray-dried particles. During drying in an air stream at 180° C. for analytic purposes, a mass loss of 31% by weight was determined.

2b) Washing

[0086] Residue organic compounds were removed by washing. 71 g of the solid obtained in step 2a) was put on a filter and water was added. A total volume of 7 L of water was used for the washing. Subsequently, the washed solid was dried in a vacuum oven at 80° C. for 10 hours.

[0087] An elementary analysis showed a niobium content of 0.9% by weight and a titanium content of 8.6% by weight, referring to the washed and dried solid. During drying in an air stream at 180° C. for analytic purposes, a mass loss of 4.1% by weight was determined.

2c) Deposition of Platinum

[0088] For the deposition of platinum, 15 g of the solid obtained in step 2b) were suspended in 414 mL water by means of an ULTRA-TURRAX®. Then, a solution of 10.95 g platinum(II)nitrate in 159 mL water was added. Under stirring, a mixture of 354 mL ethanol and 487 mL water was added and the suspension was heated to 82° C. After six hours at 82° C., the suspension was cooled to room temperature, filtered and the solid residue was washed with 6 L water. The resulting solid was dried in a vacuum oven at 80° C.

2d) Heat Treatment at 800° C.

[0089] 15 g of the solid resulting from step 2c) were heat treated in a rotary tube furnace. In a stream comprising 95% by volume of nitrogen and 5% by volume of hydrogen, the temperature was raised by 10 Kelvin per minute to 800° C. When the temperature of 800° C. was reached, the temperature was kept constant for one hour. Subsequently, the interior of the furnace was cooled to room temperature and at a temperature below 50° C., the gas stream was switched to a stream comprising 100% by volume nitrogen. Then, the heat treated solid was passivated for 12 hours with a gas stream comprising 9% by volume air and 91% by volume nitrogen to form the carbon supported catalyst.

[0090] By elementary analysis, a niobium content of 0.58% by weight, a titanium content of 6.2% by weight and a platinum content of 30% by weight, referring to the carbon supported catalyst, was determined.

[0091] The carbon supported catalyst was further analyzed by powder X-ray diffractometry. The average crystallite size of the platinum comprised in the carbon supported catalyst was calculated from the powder X-ray diffractometry results applying the Scherrer formula. A bimodal distribution of 3.2 nm and 32 nm was determined for the platinum crystallite size. Further, a crystallographic phase of TiO.sub.2 (anatase) was observed in the carbon supported catalyst by powder X-ray diffractometry.

Example 3

3a) Precipitation of Mixed Niobium Titanium Oxide onto Carbon

[0092] A mixture was prepared from 60 g carbon (Black Pearls® 2000, Cabot), 455 g acetic acid with a purity of 100%, 676 g 2-propanol with a purity of 99.7%, 43.49 g niobium(V)ethoxide with a purity of 99.95%, based on the metal content, and 46.51 g titanium(IV)n-butoxide with a purity of 99%. In order to homogenize the components, ultra-sonication was for applied for 10 minutes. The mixture was dried in a spray-dryer. In order to prevent sedimentation, the mixture was agitated while being conveyed into the spray-tower. The flow rate of the mixture to be spray-dried was 516 g/h, the diameter of the nozzle of the spray-dryer was 1.4 mm, the nozzle pressure was 3.5 bar absolute, the nozzle gas was nitrogen, the volume flow of the nozzle gas was 3.5 Nm.sup.3/h, the temperature of the nozzle gas was room temperature, the drying gas was nitrogen, the volume flow of the drying gas was 25 Nm.sup.3/h, the temperature of the drying gas was 190° C. and the residence time in the spray-dryer was 15 seconds. For particle separation, a cyclone was applied, which is able to separate particles with a diameter of at least 10 μm. The temperature in the cyclone, corresponding to the exhaust gas temperature of the spray-dryer, was 100° C. to 107° C. All above-described production steps were carried out with exclusion of humidity. No extra water was added in any of the above-described production steps and the mixture to be spray-dried was prepared under nitrogen atmosphere.

[0093] An elementary analysis showed a niobium content of 5.3% by weight and a titanium content of 2.8% by weight, referring to the spray-dried particles. During drying in an air stream at 180° C. for analytic purposes, a mass loss of 26% by weight was determined.

3b) Washing

[0094] Residue organic compounds were removed by washing. 71 g of the solid obtained in step 3a) was put on a filter and water was added. A total volume of 7 L of water was used for the washing. Subsequently, the washed solid was dried in a vacuum oven at 80° C. for 10 hours.

[0095] An elementary analysis showed a niobium content of 6.4% by weight and a titanium content of 3.9% by weight, referring to the washed and dried solid. During drying in an air stream at 180° C. for analytic purposes, a mass loss of 14.3% by weight was determined.

3c) Deposition of Platinum

[0096] For the deposition of platinum, 15 g of the solid obtained in step 3b) were suspended in 414 mL water by means of an ULTRA-TURRAX®. Then, a solution of 10.95 g platinum(II)nitrate in 159 mL water was added. Under stirring, a mixture of 354 mL ethanol and 487 mL water was added and the suspension was heated to 82° C. After six hours at 82° C., the suspension was cooled to room temperature, filtered and the solid residue was washed with 6 L water. The resulting solid was dried in a vacuum oven at 80° C.

3d) Heat Treatment at 800° C.

[0097] 15 g of the solid resulting from step 3c) were heat treated in a rotary tube furnace. In a stream comprising 95% by volume of nitrogen and 5% by volume of hydrogen, the temperature was raised by 10 Kelvin per minute to 800° C. When the temperature of 800° C. was reached, the temperature was kept constant for one hour. Subsequently, the interior of the furnace was cooled to room temperature and at a temperature below 50° C., the gas stream was switched to a stream comprising 100% by volume nitrogen. Then, the heat treated solid was passivated for 12 hours with a gas stream comprising 9% by volume air and 91% by volume nitrogen to form the carbon supported catalyst.

[0098] By elementary analysis, a niobium content of 4.7% by weight, a titanium content of 2.9% by weight and a platinum content of 34% by weight, referring to the carbon supported catalyst, was determined.

[0099] The carbon supported catalyst was further analyzed by powder X-ray diffractometry. The average crystallite size of the platinum comprised in the carbon supported catalyst was calculated from the powder X-ray diffractometry results applying the Scherrer formula. A bimodal distribution of 2.9 nm and 27 nm was determined for the platinum crystallite size. Further, a crystallographic phase of TiO.sub.2 (anatase) was observed in the carbon supported catalyst by powder X-ray diffractometry.

Example 4

4a) Reactive Deposition of Mixed Niobium Titanium Oxide onto Carbon

[0100] A mixture was prepared from 15 g carbon (Black Pearls® 2000, Cabot), 114 g acetic acid with a purity of 100%, 169 g 2-propanol with a purity of 99.7%, 2.61 g niobium(V)ethoxide with a purity of 99.95%, based on the metal content, and 24.99 g titanium(IV)n-butoxide with a purity of 99%. This mixture was transferred to a flask equipped with a magnetic stirrer, an oil bath and a water-cooled condenser. After purging with nitrogen, the mixture was heated under reflux at 94° C. for 1 h. The mixture was cooled to room temperature, filtrated and washed with a mixture of 570 g acetic acid with a purity of 100%, 845 g 2-propanol with a purity of 99.7%. Subsequently, the powder was washed with water of 60° C. until the filtrate's pH reached a value of 7. The washed solid was dried in a vacuum oven at 80° C. for 10 hours.

[0101] An elementary analysis of the dried solid showed a niobium content of 1.4% by weight and a titanium content of 6.8% by weight. During drying in an air stream at 180° C. for analytic purposes, a mass loss of 1.1% by weight was determined.

4b) Deposition of Platinum

[0102] For the deposition of platinum, 10 g of the solid obtained in step 4a) were suspended in 276 mL water by means of an ULTRA-TURRAX®. Then, a solution of 7.30 g platinum(II)nitrate in 106 mL water was added. Under stirring, a mixture of 236 mL ethanol and 326 mL water was added and the suspension was heated to 82° C. After six hours at 82° C., the suspension was cooled to room temperature, filtered and the solid residue was washed with 6 L water. The resulting solid was dried in a vacuum oven at 80° C.

4c) Heat Treatment at 800° C.

[0103] 15 g of the solid resulting from step 3c) were heat treated in a rotary tube furnace. In a stream comprising 95% by volume of nitrogen and 5% by volume of hydrogen, the temperature was raised by 10 Kelvin per minute to 800° C. When the temperature of 800° C. was reached, the temperature was kept constant for one hour. Subsequently, the interior of the furnace was cooled to room temperature and at a temperature below 50° C., the gas stream was switched to a stream comprising 100% by volume nitrogen. Then, the heat treated solid was passivated for 12 hours with a gas stream comprising 9% by volume air and 91% by volume nitrogen to form the carbon supported catalyst.

[0104] By elementary analysis, a niobium content of 0.96% by weight, a titanium content of 4.8% by weight and a platinum content of 28% by weight, referring to the carbon supported catalyst, was determined.

[0105] The carbon supported catalyst was further analyzed by powder X-ray diffractometry. The average crystallite size of the platinum comprised in the carbon supported catalyst was calculated from the powder X-ray diffractometry results applying the Scherrer formula. A bimodal distribution of 3.1 nm and 29 nm was determined for the platinum crystallite size. Further, a crystallographic phase of TiO.sub.2 (anatase) was observed in the carbon supported catalyst by powder X-ray diffractometry.

COMPARATIVE EXAMPLES

Comparative Example 1

C1a) Deposition of Platinum onto Unmodified Carbon

[0106] 20 g of Black Pearls® 2000 were suspended in 550 mL water by means of an ULTRA-TURRAX®. Then, a solution of 14.6 g platinum(II)nitrate in 215 mL water was added. Under stirring, a mixture of 471 mL ethanol and 650 mL water were added to the suspension and the suspension was heated to 82° C. After six hours at 82° C., the suspension was cooled to room temperature, filtered and the solid residue was washed with 6 L water. The resulting solid was dried in a vacuum oven at 80° C.

[0107] By elementary analysis, a platinum content of 28.1% by weight, referring to the produced catalyst from comparative example 1, was determined. The resulting catalyst was analyzed by X-ray diffractometry and applying the Scherrer formula the average platinum crystallite size was calculated. A bimodal distribution of 1.8 and 6.5 nm was obtained.

Comparative Example 2

C2a) Precipitation of Niobium Oxide onto Carbon

[0108] A mixture was prepared from 120 g carbon (Black Pearls® 2000, Cabot), 1090 g acetic acid with a purity of 100%, 1217 g 2-propanol with a purity of 99.7%, 104.9 g niobium(V)ethoxide with a purity of 99.95%, based on the metal content. In order to homogenize the components ultra-sonication was applied for 10 minutes. The mixture was dried in a spray-dryer. In order to prevent sedimentation, the mixture was agitated while being conveyed into the spray tower. The flow rate of the mixture to be spray-dried was 700 g/h. The diameter of the nozzle of the spray-dryer was 2.3 mm, the nozzle pressure was 3.5 bar absolute, the nozzle gas was nitrogen, the volume flow of the nozzle gas was 3.5 Nm.sup.3/h, the temperature of the nozzle gas was room temperature, the drying gas was nitrogen, the volume flow of the drying gas was 25 Nm.sup.3/h, the temperature of the drying gas was 190° C. and the residence time in the spray-dryer was 15 seconds. For particle separation, a cyclone was applied, which is able to separate particles with a diameter of at least 10 μm. The temperature in the cyclone, corresponding to the exhaust gas temperature of the spray-dryer, was 101° C. to 103° C. All above-described production steps were carried out with exclusion of humidity. No extra water was added in any of the above-described production steps and the mixture to be spray-dried was prepared under nitrogen atmosphere.

[0109] By elementary analysis, a niobium content of 10.6% by weight, referring to the spray-dried solid, was observed. During drying in an air stream of 180° C. for analytic purposes, a mass loss of 24.0% by weight was determined.

C2b) Deposition of Platinum

[0110] 20 g of the solid obtained in step C2a) were suspended in 444 mL water by means of an ULTRA-TURRAX®. Then, a solution of 11.98 g platinum(II)nitrate in 174 mL water was added. Under stirring, a mixture of 380 mL ethanol and 524 mL water was added and the suspension was heated to 82° C. After 6 hours at 82° C., the suspension was cooled to room temperature, filtered and the solid residue was washed with 6 L water. The resulting solid was dried in a vacuum oven at 80° C.

C2c) Heat Treatment at 800° C.

[0111] The solid resulting from step C2b) was heat treated in 3 portions, which comprised 9.1 g, 10.4 g and 10.5 g, respectively. The heat treatment was carried out in a rotary tube furnace. In a stream comprising 95% by volume of nitrogen and 5% by volume of hydrogen, the temperature was raised by 10 Kelvin per minute to 800° C. When a temperature of 800° C. was reached, the temperature was kept constant for one hour. Subsequently, the interior of the furnace was cooled to room temperature and at a temperature below 50° C., the gas stream was switched to a stream comprising 100% by volume of nitrogen. Then, the heat treated solid was passivated for 12 hours with a gas stream comprising 9% by volume of air and 91% by volume of nitrogen to form the carbon supported catalyst. The three portions of this solid were mixed with a spatula and in all subsequent steps this mixture of the portions was used.

[0112] By elementary analysis, a niobium content of 9.6% by weight and a platinum content of 33% by weight, referring to the carbon supported catalyst, were determined.

[0113] The carbon supported catalyst was analyzed by powder X-ray diffractometry and the average platinum crystallite size was calculated applying the Scherrer formula. A bimodal distribution of 3 and 22 nm was obtained.

Comparative Example 3

C3a) Precipitation of Niobium Oxide onto Carbon Having a Low Specific Surface Area

[0114] A mixture was prepared from 120 g carbon (Vulcan XC72®, Cabot), possessing a specific BET surface of approximately 250 m.sup.2/g, 1099 g acetic acid with a purity of 100%, 1217 g 2-propanol with a purity of 99.7% and 209.8 g niobium(V)ethoxide with a purity of 99.95%, based on the metal content. In order to homogenize the components ultra-sonication was applied for 10 minutes. A mixture of 178 g water and 178 g 2-propanol was added dropwise. The mixture was dried in a spray-dryer. In order to prevent sedimentation, the mixture was agitated while being conveyed into the spray-tower. The flow rate of the mixture to be spray-dryer was 521 g/h, the diameter of the nozzle of the spray-dryer was 2.3 mm, the nozzle pressure was 3.0 bar absolute, the nozzle gas was nitrogen, the volume flow of the nozzle gas was 3.5 Nm.sup.3/h, the temperature of the nozzle gas was room temperature, the drying gas was nitrogen, the volume flow of the drying gas was 25 Nm.sup.3/h, the temperature of the drying gas was 190° C. and the residence time in the spray-dryer was 15 seconds. For particle separation, a cyclone was applied, which was able to separate particles with a diameter of at least 10 μm. The temperature in the cyclone corresponding to the exhaust temperature of the spray-dryer was 104° C. to 107° C.

[0115] By elementary analysis a niobium content of 13.5% by weight was determined, referring to the spray-dried solid. During drying in an air stream at 180° C. for analytic purposes, a mass loss of 12.8% by weight was determined.

C3b) Deposition of Platinum

[0116] 10 g of the solid obtained in step C3a) were suspended in 229 mL water by means of an ULTRA-TURRAX®. Then, a solution of 6.18 g platinum(II)nitrate in 89 mL water was added. Under stirring, a mixture of 196 mL ethanol and 270 mL water was added and the suspension was heated to 82° C. After 6 hours at 82° C., the suspension was cooled to room temperature, filtered and the solid residue was washed with 4 L water. The resulting solid was dried in a vacuum oven at 80° C.

C3c) Heat Treatment at 800° C.

[0117] 12.7 g of the solid resulting from step C3b) were heat treated in a rotary tube furnace. In a stream comprising nitrogen the temperature was raised by 10 Kelvin per minute to 400° C. After the temperature of 400° C. was reached, the gas stream was switched to a stream comprising 95% by volume of nitrogen and 5% by volume of hydrogen. The temperature was raised by 10 Kelvin per minute to 800° C. When the temperature of 800° C. was reached, the temperature was kept constant for one hour. Subsequently, the interior of the furnace was cooled to room temperature and at a temperature below 50° C., the gas stream was switched to a gas stream comprising 100% by volume of nitrogen. Then, the heat treated solid was passivated for 12 hours with a gas stream comprising 9% by volume air and 91% by volume nitrogen to form the carbon supported catalyst.

[0118] By elementary analysis, a niobium content of 13.5% by weight and a platinum content of 28.5% by weight, referring to the carbon supported catalyst, were determined. Further, a crystallographic phase of Nb.sub.2O.sub.5 and NbO.sub.2, respectively, was observed in the carbon supported catalyst by powder X-ray diffractometry.

II. Electrochemical Testing of Carbon Supported Catalysts

[0119] The carbon supported catalysts resulting from example 1 and comparative examples 1, 2 and 3 were tested in the oxygen reduction reaction (ORR) on a rotating disk electrode (RDE) at room temperature. The setup comprised three electrodes. As counter electrode a platinum foil and as reference electrode an Hg/HgSO.sub.4 electrode were installed. The noted potentials refer to a reversible hydrogen electrode (RHE). An ink, comprising the carbon supported catalyst, was prepared by dispersing approximately 0.01 g carbon supported catalyst in a solution, consisting of 4.7 g demineralized ultra-pure water with a conductivity of less than 0.055 μS/cm, 0.04 g of a solution of 5% by weight of Nafion®, which is a perfluorinated resin solution, commercially available from Sigma-Aldrich Corp., comprising 80% to 85% by weight of lower aliphatic alcohols and 20% to 25% by weight of water, and 1.2 g of 2-propanol. The ink was treated by ultra-sonication for 15 minutes.

[0120] 7.5 μL of the ink were pipetted on a glassy carbon electrode with a diameter of 5 mm. The ink was dried without rotation of the electrode in a flow of nitrogen. As electrolyte a 0.1 M solution of HClO.sub.4 was applied, which was saturated with argon.

[0121] Initially, cleaning cycles and cyclovoltamograms for background subtraction (Ar-CV) were applied. These steps are further defined as steps 1 and 2 in table 1.

[0122] Subsequently, the electrolyte was saturated with oxygen and the oxygen reduction activity was determined (step 3, table 1).

[0123] Thereafter, an accelerated degradation test was applied in argon-saturated electrolyte. Therefore, the potential was changed according to square wave cycles (step 5, table 1).

[0124] Subsequently, the electrolyte was exchanged against a fresh 0.1 M HClO.sub.4 solution and the steps of cleaning and Ar-CV in argon-saturated electrolyte were repeated (steps 6 and 7 in table 1) and the oxygen reduction (ORR) activity was measured again in oxygen saturated electrolyte (step 8 in table 1).

TABLE-US-00001 TABLE 1 Examination steps Saturation Rotation No of Scan rate or Step No. Type gas rate cycles Potential range hold time 1 Cleaning Argon   0 rpm 5  50-1400 mV 1000 mV/s 2 Ar-CV Argon   0 rpm 3  10-1000 mV  20 mV/s 3 ORR-CV Oxygen 1600 rpm 3  10-1000 mV  20 mV/s 4 Ar-CV Argon   0 rpm 3  10-1000 mV  20 mV/s 5 Degradation Argon   0 rpm 20,000 100-1000 mV 0.5 s/0.5 s 6 Cleaning Argon   0 rpm 5  50-1400 mV 1000 mV/s 7 Ar-CV Argon   0 rpm 3  10-1000 mV  20 mV/s 8 ORR-CV Oxygen 1600 rpm 3  10-1000 mV  20 mV/s

[0125] The electrochemical performance of the different carbon supported catalysts is expressed by the comparison between the ORR activities before (step 3) and after (step 8) the degradation tests (step 5).

[0126] From the anodic part of the third ORR-CV the Ar-CV from the prior step was subtracted, in order to remove the background currents. The platinum-mass-related kinetic activity l.sub.kin was calculated by taking into account the current at 0.9 V (l.sub.0.9V) the limiting current at approximately 0.25 V (l.sub.lim) and the mass of platinum on the electrode (m.sub.Pt):

l.sub.kin=l.sub.0.9V.Math.l.sub.lim/(l.sub.lim−l.sub.0.9V)m.sub.Pt

[0127] The assumptions made for this calculation method and further details thereof are described in Paulus et al., in Journal of Electroanalytical Chemistry, 495 (2001), pages 134 to 145.

TABLE-US-00002 TABLE 2 Stability of the carbon supported catalysts ORR activity/mA/mg.sub.Pt fresh catalyst after degradation (step 3) (step 8) Example 1 315 287 Example 2 296 284 Example 3 277 275 Example 4 407 354 Comparative Example 1 321 219 Comparative Example 2 232 235 Comparative Example 3 121 124

[0128] The amount of platinum required for a certain performance in applications for example in fuel cells strongly depends on the stability of the carbon supported catalyst as well as on the initial activity of the fresh carbon supported catalyst. The residual activity of the used carbon supported catalyst after degradation test is a crucial parameter, mimicking the degradation of the catalytically active metal phase in a real fuel cell to a large extent.

[0129] The catalyst according to the invention being modified with the oxide comprising niobium and titanium and prepared in example 1 showed with 287 mA/mg.sub.Pt the highest residual activity after degradation over all examples and comparative examples. All inventive catalysts of examples 1, 2 and 3 showed a higher stability against electrochemical degradation with higher residual activities after degradation in comparison with catalysts without modifier or with catalysts comprising only niobium oxide as modifier instead of an oxide comprising both niobium and titanium.

[0130] The concentration of the oxidic modifier comprised in the carbon supported catalysts resulting from example 1 and comparative example 2 was similar. Hence, the higher residual activity for the inventive carbon supported catalyst resulting from example 1 can be attributed to the modification of the carbon support with an oxide comprising both niobium and titanium.

[0131] Still, the niobium-oxide-modified catalyst resulting from comparative example 2 showed a higher residual activity after the degradation test than the catalyst without any modifier resulting from comparative example 1.

[0132] The catalyst resulting from comparative example 3, which was modified with niobium oxide and which comprised a carbon-comprising support with a low surface area showed clearly the lowest residual activity over all example and comparative examples, even though the content of niobium oxide and platinum in the carbon supported catalysts were similar.