WO2009081183A1

WO2009081183A1 - Improvements in catalysts

Info

Publication number: WO2009081183A1
Application number: PCT/GB2008/051171
Authority: WO
Inventors: Sonia Garcia Lopez; Robert John Potter
Original assignee: Johnson Matthey Public Limited Company
Priority date: 2007-12-21
Filing date: 2008-12-10
Publication date: 2009-07-02
Also published as: GB0724909D0

Abstract

An electrocatalyst comprises sub-nanometre-sized cobalt moieties deposited on an active carbon. It is believed that the moieties are cobalt co-ordination complexes. The electrocatalysts are stable and productive catalysts for the two-electron reaction producing hydrogen peroxide, which can be obtained at high concentration.

Description

IMPROVEMENTS IN CATALYSTS

The present invention concerns improvements in catalysts, and more especially concerns a novel metal catalyst and its potential use in industry, particularly in the production of hydrogen peroxide.

Hydrogen peroxide is used widely as an environmentally benign oxidising agent. It is used in the pulp and paper, textile and chemical industries and production is approximately 1 ,000,000 tonnes per annum in Europe. At present hydrogen peroxide is manufactured mainly by a chemical route based on the reduction of an anthraquinone with hydrogen followed by its oxidation with air. The hydrogen peroxide is then extracted in an aqueous stream and must be further purified to eliminate organic contaminants. The method requires a large-scale installation to be economical and the hydrogen peroxide must be transported to the point of use. The anthraquinone process carries certain environmental risks, and is energy-intensive.

An electrochemical cell for hydrogen peroxide production is disclosed in US Patents 5,565,073 and 5,647,968. Oxygen is reduced and water is oxidised in a liquid alkaline electrolyte. The process requires significant input of electrical energy, and the hydrogen peroxide is produced in an alkaline solution (generally a sodium hydroxide solution). Separation of the peroxide from the base is difficult, making the process expensive and inefficient for the production of free hydrogen peroxide. An advantage of this electrochemical process over the anthroquinone process, however, is that it does offer an opportunity to build small point-of-use hydrogen peroxide generators, if other technological problems can be overcome.

Monsanto's WO2007/098432 describes transition metal catalysts , processes for their preparation and their use as fuel cell catalysts, and is particularly directed to carbon- supported cobalt-nitrogen compositions. The skilled person would understand this lengthy and wide-ranging disclosure to be directed to the use of nitrides. The skilled person notes that the intended purpose of the catalyst is to generate power in a fuel cell and that the generation of hydrogen peroxide is undesirable since this reaction pathway is less energy efficient than the production of water.

A fuel cell is fundamentally an electrochemical cell formed from two electrodes and an intervening electrolyte. In a fuel cell, the stored chemical energy of a fuel and an oxidant is converted into electrical energy. The fuel is usually hydrogen or methanol, and this is combined with oxygen to produce power. The hydrogen or methanol is oxidised at the anode and oxygen is reduced at the cathode. The electrodes are porous to gas diffusion and are both in contact with the electrolyte. The electrolyte may be liquid or solid, acid or alkaline in nature.

There has recently been a high level of research and development in fuel cell technology. It is envisaged that fuel cells will provide power sources for a wide range of applications, particularly as a source of power for small items such as laptops or mobile phones, potentially as a replacement for internal combustion engines in vehicles, and in a combined heat and power generator for domestic or other small scale use.

When hydrogen and oxygen are fed into a fuel cell, the reaction at the anode is the oxidation of hydrogen:

2H₂ → 4H^t + 4e- the reaction at the cathode is the four electron reduction of oxygen:

O₂ + 4H⁺ + 4e → 2H₂O and the product of the overall reaction is water:

2H₂ + O₂ → 2H₂O The Gibbs energy of this reaction is converted to electrical energy.

Hydrogen peroxide can be produced instead of water if the four electron reduction of oxygen is inhibited and the two electron reduction of oxygen at the cathode is promoted:

O₂ + 2H⁺ + 2e^' → H₂O₂ The overall cell reaction becomes:

O₂ + H₂ → H₂O₂

Again, the Gibbs energy of the reaction is converted into electrical energy, so that the synthesis of hydrogen peroxide is accompanied by energy production.

Hydrogen peroxide production using fuel cell technology would provide many advantages over the current methods of manufacture. The process is essentially a zero-emission process and no electricity is required to power the cell; indeed energy is produced by the electrochemical cell and can be used to reduce the energy costs elsewhere, for example in a chemical process. The generation of hydrogen peroxide can be achieved in a small modular unit that can be integrated into a larger system. Accordingly hydrogen peroxide can be generated at the point of use and there are no transportation costs or difficulties.

Two distinct approaches have so far been adopted in the development of fuel cells for hydrogen peroxide production. One approach is similar to that adopted in the O₂/H₂O electrochemical cells. A wide variety of electrode surfaces such as carbon, gold and compounds of cobalt are known to promote two electron oxygen reduction in alkaline media, so a liquid alkaline electrolyte is used. An oxygen diffusion electrode is combined with a hydrogen diffusion electrode, and the hydrogen peroxide is produced in the alkaline (usually hydroxide) solution. The generation of hydrogen peroxide in an alkaline fuel cell is described in J. Electrochem. Soc. Vol. 145, No. 10, 3444-3449.

Another approach has followed current fuel cell technology and uses a solid membrane electrolyte. In proton exchange membrane fuel cells (PEM FC), the electrolyte is a solid, proton-conducting polymer electrolyte membrane. The combined structure of a membrane and two gas diffusion electrodes is known as the membrane electrode assembly (MEA). Electro catalysts are incorporated on either side of the membrane in the MEA to increase the rates of the desired electrode reactions. Membrane electrolytes are clearly more convenient than liquid electrolytes because they are light and compact, and there are no problems with separating the product from the electrolyte. The most common membranes are based on polymeric perfluorosulfonic acid, eg the commercially available Nafion®. Patent applications WO 97/13006 and WO 95/30474 (both The Dow Chemical Company) disclose an electrochemical cell for hydrogen peroxide synthesis containing a Nafion® based membrane. The choice of cathode electrocatalyst is limited because most known oxygen reduction catalysts promote the four electron reduction of oxygen in acidic media. Electrocatalysts based on metals such as zinc, gadolinium and lanthanum have been proposed because they appear to favour the two electron reduction. Despite careful catalyst choice, however, the selectivities of the disclosed processes are quite poor (generally below 70mol%) and only very weak solutions of hydrogen peroxide in water are produced (up to about 4wt%). In addition, it is likely that the integrity of catalyst materials such as zinc would be damaged by the highly acidic electrolyte environment (acid concentrations of over 3mol/l prevail in the polymer). There has been some study of cobalt-based electrocatalysts, primarily in the context of power-generating fuel cells: Estimation of specific interaction between several Co porphyrins and carbon black: its influence on the electrocatalytic O₂ reduction by the porphyrins, J. Electroanal. Cem 576 (2005) 253, Oxygen reduction electrocatalysis: ageing of pyrolyzed Co macrocycles dispersed on an active carbon, Electrochim. Acta, 44 (1999) 2653, and the aforesaid WO2007/098432.

There remains a need for a stable electrocatalyst that will be effective for the production of hydrogen peroxide using a PEM fuel cell structure. Additionally, any practical electrocatalyst must be stable in the sense that after several months of stop- start operation, there is no significant movement of catalytic material into the membrane or onto the other electrode.

The present invention provides a stable electrocatalyst comprising sub-nanometre- sized particles of cobalt (including metallic cobalt, cobalt alloys, oxide and/or hydroxide) moieties deposited on a high surface area active carbon carrier, the moieties comprising particles having one or more partial or complete coatings of a nitrogen-containing material or having bondings incorporating nitrogen. It is desirable that the electrocatalysts of the invention are substantially non-ferromagnetic. It is also desirable that the electrocatalysts of the invention comprise a partial or complete coating of an organic nitrogen-containing material. At the present time, the precise form of the cobalt moieties is not clear. The particles are not detectable by transmission electron microscopy ("TEM"). It is believed that cobalt moeties of size greater than 2 nni would be detectable by the particular model TEM.

The catalyst of the invention was assessed by a number of analytical techniques. X- ray diffraction ("XRD") was able to identify only crystalline NaBr, believed to be formed as a by-product. XPS indicated that the cobalt moieties included Co(II) with nitrogen bonding. EXAFS supported the XPS results but indicated that the moieties were different from those formed between cobalt and a macrocycle such Co phthalocyanine. From the EXAFS results, the moieties do not appeal' to have metallic or metal oxide character.

When nominally metallic particles are sub-nanometre-sized, they do not exhibit all the characteristics of bulk metal, but it is not believed that this effect governs the cobalt moieties of the present invention, although it can be expected that this effect does contribute to the characteristics of the cobalt moieties of the present invention.

Without wishing to be bound by any theory, it is presently believed that the cobalt moieties are in the form of a co-ordination complex. The complex may be in a square planar co-ordination. There are indications that the cobalt is Co¹¹, and it is hypothesised that this is in a low spin state.

The catalysts of the invention are distinct from those prepared according to WO2007/098432. We have prepared for comparison purposes a cobalt-containing catalyst according to this prior art, and have assessed both its magnetic properties and its hydrogen peroxide-generating properties, as well as using a variety of analytic techniques to study the prior art catalysts and the catalysts according to the invention. The preparation and tests are described in greater detail below.

In the present invention, the loading of cobalt, calculated as metal, on the active carbon is desirably less than 10 wt%, and is preferably approximately 2 wt%. The invention also provides a method of preparing the electrocatalysts of the invention, comprising preparing an aqueous solution of a cobalt salt and preparing from the solution a suspension of sub-nanometre-sized cobalt (this term including metallic cobalt, cobalt alloys, oxide and hydroxide) particles, depositing the particles onto a high surface area active carbon carrier, and depositing one or a plurality of partial or complete coatings or bonding precursors incorporating nitrogen, and firing the product, preferably at a temperature from 600 to 700° C, preferably to yield a partial or complete organic nitrogen-containing coating on the cobalt moieties..

The method includes the variant of forming the cobalt particles in a medium in which the active carbon carrier is present, so that the formation and deposition steps are largely simultaneous.

Preferably, the method comprises the steps of forming a solution of cobalt chloride, and in the presence of a surfactant stabiliser, and using a base hydrolysis reaction, for example using sodium carbonate to form the particles. More preferably, the base hydrolysis is carried out at elevated temperature, for example at approximately 80° C, with stirring or other agitation, for an extended period, for example overnight or 15 hours. The stabiliser is desirably a tetraalkyl ammonium bromide product (CTAB), although we have also successfully used succinic anhydride, and such products are commercially available.

To the suspension of the particles, is added a high surface area active carbon, for example 200 mVg surface area or greater and the mixture stirred. The resulting product is dried, optionally with preliminary filtering. Preferably, there is an intermediate firing, for example at 300° C, to remove the stabiliser. The dried product is subsequently impregnated with a precursor of a nitrogen-containing coating or bonding, preferably an organic nitrogen donor, more preferably an aromatic nitrogen donor, most especially a chelating agent such as o-phenylene diamine.

The resulting mixture is dried and fired under an inert gas atmosphere, conveniently under argon. It has been found that firing at temperatures below approximately 700° C, say at 675° C, yield an electrocatalyst that favours the 2-electron reaction useful for hydrogen peroxide manufacture, whereas firing at temperatures above approximately 800° C, say at 900° C, tends to yield an electrocatalyst that favours the 4-electron reaction more useful in fuel cells.

The electrocatalyst of the invention may be used as such or in association with one or more other catalysts or promoters.

The CoPt3 electrocatalyst has been shown to be very effective in fuel cells but to lose performance due to dissolution of the Co under operating conditions; it is considered that the present invention may offer increased stability. It is not necessary that the catalytic reaction involved in any use of the electrocatalyst of the invention is one that would normally be recognised as an electrocatalytic reaction.

The present invention additionally provides components and a method for manufacturing hydrogen peroxide. The electrocatalysts may be incorporated into a MEA using conventional technology such as forming an electrocatalytic ink for printing onto a PEM membrane or onto a gas diffusion electrode to form a cathode. The MEA may be incorporated into a PEM fuel cell or fuel cell stack. In use, the cathode is supplied with humidified oxygen, and a standard catalysed anode supplied with humidified hydrogen. No external power is required. Operating at temperatures of approximately 10° C (much lower than for power generating fuel cells), hydrogen peroxide solution may be withdrawn from the cell at industrially-useful concentrations. Under preferred conditions, and using optimised conditions to reduce oxygen stoichiometry and water, hydrogen peroxide solutions of 30% by wt have been seen. Furthermore, the cells have operated in the laboratory for several months without any noticeable loss of performance, indicating excellent stability.

Indicative cell operating conditions are 40-100 mV and a current of 300 mA/sq cm. There does not appear to be active hydrogen evolution, suggesting that the risk of explosion during operation is not significant. Operation relatively close to the hydrogen potential reduces any tendency for corrosion of the cathodic electrocatalyst.

The hydrogen peroxide-generating fuel cells of the invention permit generation of hydrogen peroxide locally,, at a point of use. Potentially, this may even be generated in small industry or domestic environments, such as for swimming pools, laundry and cleaning etc It is conceivable that the hydrogen peroxide may be combined with other materials, for example acetic acid, or with other biocides, to be an effective biocide. This may control biofilm or algal growth. The hydrogen peroxide may also be used as a reagent in chemical processes.

The invention will now be described by way of example.

Example 1 Electrocatalyst preparation

Into a 1 1 round flask, were charged an aqueous solution Of NaCO₃ (8.30 g in 160 ml of water) and CTAB (29.08 g in 160 ml of water) at room temperature. To the flask was then added dropwise over a period of 30 minutes an aqueous solution of CoCl₂-OH₂O (4.75 g in 80 ml). The mixture was heated to 80° C for ca. 16 hours. The reaction mixture was then cooled to room temperature and weighed (418.83 g).

208 g of the reaction mixture was added to 26 g of commercial activated carbon XC72R to form a slurry. 100 ml of deionised water was added to allow better stirring. Stirring was continued with heating until most of the water had been evaporated. The remaining slurry was dried completely by keeping in an oven at 76° C over the weekend. The dried material was fired at 300° C under nitrogen or argon.

After cooling, 33.48 of the resulting catalyst precursor was charged into a beaker and 16.5g (150 mrnol) of o-phenylenediamine was added with the minimum amount of hot water (ca. 300 ml) and the mixture stirred. The beaker was heated to evaporate most of the water, then the product was dried overnight at 76° C. 55.94 g of dried product was collected.

15.03 g of the dried product was fired in a furnace oven under an Arc atmosphere as follows:

Ramp temperature 10° C/min to 300° C, then 60 min hold at this temperature;

Ramp temperature 10° C/min to 675° C, hold for 60 min at this temperature.

After cooling, the fired electrocatalyst (10.36 g) was collected. Analysis showed a loading of 3.34 wt% of Co.

Example 2 Electrocatalyst preparation 205 g of the reaction mixture prepared as in Example 1 was added to 26 g carbon Black Pearl 2000 to form a slurry. 100 ml of deionised water was added to facilitate stirring. The mixture was stirred and heated until most of the water had evaporated. The slurry was transferred to an oven at 76 C to completely dry the product.

37.94 g of the product was transferred to a beaker and 16.5 g of o-phenylenediamine was added in the minimum of hot water (ca. 300 ml) and the mixture stirred. Most of the water was evaporated then the beaker transferred to an oven at 76° C where the remainder of the water was evaporated overnight. 15 g of the dried product was fired according to the same firing schedule as in Example 1, yielding 10.11 g of product electro catalyst.

Example 3 Hydrogen Peroxide

A sample of electrocatalyst, having a nominal 2 wt% Co, is converted into an ink and deposited on Toray paper in conventional manner to form an elctrode. The catalyst layer thickness is approximately 40-50 microns. The electrode is converted into a membrane electrode assembly using conventional hot pressing, in combination with a Nafion 115 membrane and a conventional cathode having 40 wt% Pt on carbon catalyst. The resulting MEA, having a 49 cm 2 active area, was used in a Ballard Mk rv fuel cell (optimised for power production), fed with hydrogen and oxygen.

The cell is operated at 10° C under differing conditions. In one run, the fluid in the cell is sampled and tested for hydrogen peroxide levels, over a 22 day period. The Faradaic efficiency was calculated at the same time, and the results are shown below.

Under preferred conditions, and having allowed the cell to stabilise, a sample of the water in the cell was found to contain approximately 30% by weight of peroxide.

In an alternative run, using an electrocatalyst which had been fired at 900° C₅ a current/voltage plot was generated, which is shown in the attached Figure. This illustrates that significant currents can be drawn, whilst still being able to make several % of peroxide. For some uses, this balance of power vs peroxide make may be desirable.

Comparative testing was carried out by preparing a catalyst according to WO2007/062396 as follows:

Example 4 Electrocatalyst preparation "Monsanto" (Comparative)

Ketjen EC300J (8 g) carbon was dispersed in water (400ml) using a Silverson mixer (όOOOrpm) connected to a pH sensor. The slurry was warmed to ca. 40° C. A solution of CoCl₂OH₂O (8.1 g) in deionised water (100 ml) was fed into the stirred carbon slurry at ca. 10ml/min together with NaOH (0.5M) at a variable rate to maintain the pH at 7.0. When the addition was complete the slurry was stirred at 40-50 ⁰C for Ih. The reaction was stirred and boiled for Ih and then allowed to cool overnight. Catalyst precursor was then recovered by filtration and washed with ca. 2 1 of water. It was then dried in the oven at 105° C for ca. 16 hours.

The catalyst precursor (10 g) was then charged into a tube furnace. The tube furnace was purged with Ar for ca. 1 hour and then the catalyst was fired under an atmosphere of NH₃/N₂ (5% NH₃) using the following programme:

Ramp Temperature: 10 °C/min

Final Temperature: 300 ⁰C

Firing time at final temperature: 60 min

Ramp Temperature: 10 °C/min

Final Temperature: 675 ⁰C

Firing time at final temperature: 60 min

The final catalyst (8.45 g) was then collected. The resulting catalyst "Monsanto" and two catalysts prepared according to the present invention, one fired at 675° C and one fired at 900° C, were tested in an in-house rotating-ring-disk experiments ("RRDE")- The instrumentation consisted of an Autolab PGStat 30 electrochemical work station (Eco Chemie B. V., The Netherlands) connected to a Pine Instruments (Raleigh, NC, USA) AFMSRX rotator with MSRX speed control and E6 series RRDE electrodes. The RRDE electrode consisted of a Pt ring (inside diameter. 6.5mm, o.d. 7.5mm) and a glassy carbon disc (o.d. 5mm), pre- cleaned with gentle wiping with alcohol and ultrasonication in deionised water. The electrochemical cell was a jacketed glass vessel of approximately 100ml capacity utilising a dimensionally stable counter electrode consisting of a Ti mesh coated with a proprietary dimensionally stable RuO2/Ta2O5/TiO2 formulation (KaiDa Technology Limited 145-157 St John Street

London United Kingdom EClV 4PY), and a mercury/mercury sulphate reference electrode (Radiometer Analytical S. A., France). The electrolyte was one molar sulphuric acid (Fischer Scientific) and the cell was temperature-controlled via an integral water jacket set at 20⁰C.

The working electrode glassy carbon disc was coated with the catalyst to be studied by evaporating an aqueous Nafion (Aldrich Chemical Company) suspension of the carbon-supported catalyst at approximately 50wt% catalyst introduced onto the glassy carbon surface by micro-pipetting approximately 50 microlitres of the catalyst suspension and then rapidly drying the droplet by moving the RRDE under a commercial infra-red lamp for a few tens of seconds. Care is needed to ensure that the droplet does not spread onto the ring section of the RRDE. After drying the electrode is then ready for testing in the electrochemical cell.

The electrochemical measurement protocol is adapted from well-known literature procedures whereby the electrolyte is purged with nitrogen for 30 minutes to provide an oxygen-free 'base-line' condition. The electrode is swept at a fixed rotation rate from approximately 1.0V vs the reversible hydrogen potential (RHE) to OV v RHE and back again at 25mV/s. Concurrently, the ring electrode is fixed at 1.2V v RHE and the current recorded. The potential-sweep is repeated for a range of rotation rates e.g. from 5 to 30Hz. Oxygen or air is then bubbled into solution for 30 minutes and the potential-sweep/rotation procedure repeated. Currents from the oxygen or air runs are then subtracted from the nitrogen runs to provide a direct measure of the reduction current due to dissolved oxygen reaction.

The ring currents are used to determine the amount of peroxide generated. The nitrogen-purged background current is subtracted from the oxygen/air runs to determine the magnitude of the oxidation current at 1.2V on the ring. In the absence of peroxide generation, no oxidation currents above background are observed. To derive the absolute level of peroxide generation in terms of the Faradaic yield, the collection efficiency of the RUDE assembly needs to be measured using a standard protocol described in many electrochemical texts (see e.g. Instrumental Methods in Electrochemistry - Southampton Electrochemistry Group, Horwood Publishing Chemical Science Series , 2001). The collection efficiency for this system was approximately 22%.

The results are shown in the table below:

It can readily be seen that both of the catalysts according to the invention convert more current into the generation of hydrogen peroxide than the comparative Monsanto catalyst. For catalysts fired at the same temperature of 675° C₅ the catalyst of the invention is approximately 50% more effective.

It has also been discovered, in a simple test for magnetic properties, the Monsanto catalyst exhibited a significant ferromagnetic property, whereas the catalysts according to the invention exhibited a very small ferromagnetic property. The test involved contacting a glass phial containing a sample of the respective catalyst to a magnet and inverting the phial while still in contact with the magnet. In the case of the Monsanto catalyst, the vast majority of the catalyst (at least about 90%) remained clumped together at the top of the phial. In contrast, with catalysts according to the invention, the vast majority of the catalyst (at least about 90%) fell away under the effect of gravity. This indicates the presence in the Monsanto catalyst of metallic cobalt particles of such a size that feiTomagnetism is observed.

Additional analytical tests indicate that, in the Monsanto catalyst, rather than the nitride expected, the cobalt is more likely to be bonded to oxygen, while in the catalysts of the invention, the cobalt is associated with organic nitrogen. Studies of the dominant nitrogen signal indicate that there is essentially no nitride present.

Claims

1. A stable electrocatalyst comprising sub-nanometre-sized particles of cobalt (including metallic cobalt, cobalt alloys, oxide and/or hydroxide) moieties deposited on a high surface area active carbon carrier, the moieties comprising particles having one or more partial or complete coatings of a nitrogen-containing material or having bondings incorporating nitrogen,

2. An electrocatalyst according to claim 1, wherein the cobalt is present in an amount by weight of the electrocatalyst less than 10%, preferably approximately 2%.

3. An electrocatalyst according to claim 1 or 2, wherein the active carbon has a surface area of 200 ni^z/g or more.

4. An electrocatalyst comprising sub-nanometre sized particles of cobalt which are co-ordination complexes, deposited on high surface area active carbon.

5. An electrocatalyst according to any one of the preceding claims, which is substantially nonferromagnetic.

6. An electrocatalyst according to any one of the preceding claims, in which the cobalt moieties comprise a partial or complete coating of an organic nitrogen- containing material.

7. A method of preparing electrocatalysts, comprising preparing an aqueous solution of a cobalt salt and preparing from, the solution a suspension of sub- nanometre-sized cobalt (this term including metallic cobalt, cobalt alloys, oxide and hydroxide) particles, depositing the particles onto a high surface area active carbon carrier, and depositing one or a plurality of partial or complete coatings or bonding precursors incorporating nitrogen, and firing the product, preferably at a temperature from 600 to 700° C, preferably to yield an organic nitrogen-containing coating or partial coating on the cobalt moieties.

8. A method according to claim 7, comprising using base hydrolysis to form the suspension from the solution.

9. A method according to claim 8, wherein the base hydrolysis is carried out at an elevated temperature, desirably at approximately 80° C.

10. A method according to any one of claims 7 to 9, wherein the solution of the cobalt salt also comprises a surfactant stabiliser.

11. A method according to any one of claims 7 to 10, wherein a bonding precursor which is an aromatic nitrogen donor is used.

12. A method according to claim 9, wherein the bonding precursor is o-phenylene diamine.

13. A method of making hydrogen peroxide, in a fuel cell, comprising using an electrocatalyst according to any of claims 1 to 6, or as prepared according to any of claims 7 to 12.

14. A method according to claim 13, wherein the fuel cell is operated at a temperature below ambient, preferably approximately 10° C.

15. A cleaning or laundry device, comprising a fuel cell using an electrocatalyst according to any of claims 1 to 6, or as prepared according to claims 7 to 12.