Patent Application
20200208283
The invention relates to a gas diffusion electrode (ODE) for reducing carbon dioxide (CO2), which is based on porous silver powder as electrode catalyst, and the use thereof for the electrochemical reduction of carbon dioxide to CO.
The invention proceeds from gas diffusion electrodes known per se, which usually comprise an electrically conductive support, a gas diffusion layer and a catalytically active component and are used in chloralkali electrolysis. The known electrodes are used for cathodic oxygen reduction.
The gas diffusion electrodes are electrodes in which three states of matter solid, liquid and gaseous—are in contact with one another and the solid, electron-conducting catalyst catalyzes an electrochemical reaction between the liquid phase and the gaseous phase.
Various proposals for reducing carbon dioxide in electrolysis cells are in principle known on the laborious scale from the prior art. The basic idea here is to replace the hydrogen-involving cathode of the electrolysis (for example in chloralkali electrolysis) by the carbon dioxide ODE.
The carbon dioxide ODE has to meet a number of fundamental requirements in order to be usable in industrial electrolyzers. Thus, the catalyst and all other materials used have to be chemically stable. Likewise, a high degree of mechanical stability is required since the electrodes are installed and operated in electrolyzers having a large size of usually more than 2 m2 in area (industrial size). Further properties are: a high electrical conductivity, a low layer thickness, a high internal surface area and a high electrochemical activity of the electrocatalyst. Suitable hydrophobic and hydrophilic pores and an appropriate pore structure for conducting gas and electrolyte are likewise necessary, as is impermeability so that the gas and liquid space remain separated from one another. Long-term stability and low production costs are further particular requirements which an industrially usable oxygen-depolarized electrode has to meet.
A further important property is a low potential at a high current density of preferably greater than 4 kA/m2 and a high selectivity to carbon monoxide. Apart from standard silver particles as catalyst, different silver morphologies and also gold- and carbon-based catalysts are known for the electrochemical reduction of carbon dioxide to carbon monoxide.
Hori et al. state that a polycrystalline gold catalyst achieves a selectivity of 87% to carbon monoxide at a current of 5 mA/cm2.
Y. Chen et al. (JAGS (2012) 19969) have developed gold nanoparticles which display a selectivity of 98% at a current density of 6 mA/cm2 and a potential of −0.4 volt. Apart from gold, silver is also known as catalyst for the reduction of carbon dioxide to carbon monoxide.
Lu et al. were able to show that nanoporous silver as electrode catalyst for producing carbon monoxide from carbon dioxide displays a selectivity of 90% at a current density of 20 mA/cm2 and −0.6 volt. (Nat. Corn. 5 (2014))
The article by Lu and Jiao (Nanoenergy 2016) describes still further nanoporous silver systems. All have a similar performance (selectivity of 90% as a current density of 20 mA/cm2). The production of these nanoporous systems is very complicated, difficult to implement on an industrial scale and it is difficult to increase the porosity.
In addition, significantly higher current density of from at least 200 to 400 mA/cm2 have to be achieved for industrial use. This has hitherto not been demonstrated in the literature for any catalyst.
The production of metallic silver and very finely divided silver nanoparticles is well known and has been described in many publications and patents. Porous silver materials can be produced via a colloidal route by, for example, crystallizing monodisperse polystyrene particles, filling the interstices between the particles with silver and subsequently leaching out the polystyrene particles. This process is very complicated and unsuitable for industrial use (Chem. Mater. 2002, 14, 2199-2208). In another process, a polymer gel is utilized as template instead of the colloidal particles (Chem. Mater. 2001, 13, 1114-1123), but this is similarly complicated. In addition, all these processes are multistage processes and also require high sintering temperatures of up to 800° C.
In other processes, an AlAg or CuAg alloy is firstly produced in a complicated manner in order then to leach out the copper or aluminum; hereto, high temperatures are also necessary for producing the alloy (Nanoenergy 2016). In addition, monoliths, i.e. very large particles, which are unsuitable for further processing to give a GDE are usually obtained.
Unless harsh conditions are described in Chem. Commun., 2009, 301-303. Here, the porous particles are produced in an ionic liquid under pressure. However, ionic liquids are expensive, so that this approach also will not lead to inexpensive catalysts.
It was therefore an object of the invention to provide a gas diffusion electrode and a process for the production thereof, by means of which the reduction of carbon dioxide occurs at a high current density (>2 kA/m2) and with high selectivity (>50%).
It has surprisingly been found that a selective electrocatalyst which is based on a porous powder and can be used successfully in a GDE for carbon dioxide reduction is obtained when the syntheses of silver nanoparticles are modified by increasing the concentration of silver nitrate and a stabilizer is omitted. Micron-sized silver particles should be formed by increasing the concentration of the starting materials such as silver nitrate, sodium citrate and sodium borohydride and by growing silver nuclei. Surprisingly, it is not bulk particles but instead porous particles which sometimes have, depending on the process salts used, remarkably high BET surface areas, for example up to 8 g/m2, which are formed. The porous particles consist of agglomerated nanoparticles. The size of the nanoparticles and thus also the porosity can be controlled by the manner of addition, of mixing and of concentration of the starting materials. The primary particles preferably have a diameter of less than 100 nm. To produce the porous catalyst, silver nitrate and trisodium citrate are dissolved in water. A solution consisting of a reducing agent such as NaBH4, KBH4 or formaldehyde dissolved in water is added thereto while stirring. The porous particles are formed with a particle size of greater than 1 μm, are then filtered off, washed and dried.
Selective GDEs are obtained by means of these porous particles when the porous particles are mixed with a fluoropolymer by the present process according to the invention and the powder mixture obtained is subsequently pressed onto a support element.
The invention provides a gas diffusion electrode for reducing carbon dioxide, where the gas diffusion electrode comprises at least one sheet-like, electrically conductive support and a gas diffusion layer and electrocatalyst applied to the support, where the gas diffusion layer consists at least of a mixture of electrocatalyst and a hydrophobic polymer and silver acts as electrocatalyst, characterized in that the electrocatalyst contains silver in the form of highly porous agglomerated nanoparticles and said nanoparticles have a surface area measured by the BET method of at least 2 m2/g.
The thickness of the catalytically active coating consists of PTFE and silver of the gas diffusion electrode is preferably from 20 to 1000 urn, particularly preferably from 100 to 800 μm, very particularly preferably from 200 to 600 μm.
The proportion of electrocatalyst is preferably from 80 to 97% by weight, particularly preferably from 90 to 95% by weight, based on the total weight of electrocatalyst and hydrophobic polymer.
The proportion of hydrophobic polymer is preferably from 20 to 3% by weight, particularly preferably from 10 to 5% by weight, based on the total weight of electrocatalyst and hydrophobic polymer.
Further preference is given to an embodiment of the novel gas diffusion electrode, in which the hydrophobic polymer is a fluorine-substituted polymer, particularly preferably polytetrafluoroethylene (PTFE).
A further preferred variant of the gas diffusion electrode is characterized in that the electrode has a total loading of catalytically active component in a range from 5 mg/cm2 to 300 mg/cm2, preferably from 10 mg/cm2 to 250 mg/cm2.
Preference is given to an embodiment of the gas diffusion electrode which is characterized in that the silver particles are present as agglomerates of silver nanoparticles having an average agglomerate diameter (d50 measured by means of laser light scattering) in the range from 1 to 100 μm, preferably in the range from 2 to 90 μm.
Preference is also given to a gas diffusion electrode in which the silver nanoparticles have an average diameter in the range from 50 to 150 nm, determined by means of scanning electron microscopy with image analysis.
The novel gas diffusion electrode preferably has a support consisting of a material selected from the group consisting of silver, nickel, coated nickel, e.g. silver-coated nickel, polymer, nickel-copper alloys or coated nickel-copper alloys, e.g. silver-plated nickel-copper alloys, from which sheet-like textile structures have been produced.
The electrically conductive support can in principle be a gauze, nonwoven, foam, woven mesh, braid or expanded metal. The support preferably consists of metal, particularly preferably nickel, silver or silver-plated nickel. The support can have one or more layers. A multilayer support can be made up of two or more superposed gauzes, nonwovens, foams, woven meshes, braids or expanded metals. The gauzes, nonwovens, foams, woven meshes, braids, expanded metals can be different here. They can, for example, have different thicknesses or different porosities or have a different mesh opening. Two or more gauzes, nonwovens, foams, woven meshes, braids or expanded metals can, for example, be joined to one another by sintering or welding. Preference is given to using a gauze composed of nickel or silver and having a wire diameter of from 0.04 to 0.4 mm and a mesh opening of from 0.2 to 1.2 mm.
The support of the gas diffusion electrode is preferably based on nickel, silver or a combination of nickel and silver.
Preference is also given to a form of the gas diffusion electrode in which the support is present in the form of a gauze, woven mesh, formed-loop unit, drawn-loop unit, nonwoven, expanded metal or foam, preferably a woven mesh.
The various forms of the carbon dioxide electrolysis can in principle be distinguished by how the GDE is installed and how the distance between the ion exchange membrane and the ODE is established thereby. Many cell designs allow a gap between the ion exchange membrane and the ODE, known as the finite-gap arrangement. There the gap can be from 1 to 3 mm, and KHCO3, for example, flows through the gap. Flow can, in an upright arrangement of the electrode, occur from the top downward (for the principle of the falling-film cell, see, for example, WO 2001/057290A2) or from the bottom upward (gas pocket principle, see, for example, DE 4 444 114 A2).
A particular embodiment of the invention provides polymer-bonded electrodes, with the gas diffusion electrode being equipped with both hydrophilic and hydrophobic regions. These gas diffusion electrodes have high chemical stability, in particular when PTFE (polytetrafluoroethylene) is used.
Regions having a high proportion of PTFE are hydrophobic and electrolyte cannot penetrate here but can at places having a low proportion of PTFE or no PTFE. The catalyst itself has to be hydrophilic.
The production of such PTFE-catalyst mixtures is in principle carried out by, for example, use of dispersions of water, PTFE and catalyst. To stabilize PTFE particles in the aqueous solution, emulsifiers, in particular, are added, and thickeners are preferably used for processing the dispersion. An alternative to this wet production process is production by dry mixing of PTFE powder and catalyst powder.
The gas diffusion electrodes of the invention can, as described above, be produced by wet, dispersion and dry processes. Particular preference is given to the dry production process.
Dispersion processes are chosen mainly for electrodes with a polymeric electrolyte, for example as introduced successfully in the PEM (polymer-electrolyte membrane) fuel cell or HCl GDE-membrane electrolysis (WO 2002/18675).
When the GDE is used in liquid electrolytes, the dry process gives more suitable GDEs. In wet or dispersion processes, heavy mechanical pressing can be dispensed by evaporation of the water and sintering of the PTFE at 340° C. These electrodes are usually very open-pored. On the other hand, cracks through which the liquid electrolyte can penetrate can quickly be formed in the electrode under incorrect drying conditions. For this reason, the dry process has become established for applications using a liquid electrolyte, for example the zinc-air battery or the alkaline fuel cell.
In dry processes, the catalyst is intensively mixed with a polymer component (preferably PTFE). The powder mixture can be formed by pressing, preferably by pressing by means of a rolling process, to give a sheet-like structure which is subsequently applied to the support (see, for example, DE 3 710 168 A2; EP 144 002 A2). A preferred alternative which can likewise be used is described in DE 102005023615 A2; here, the powder mixture is sprinkled on a support and pressed together with the latter.
In the dry process, the electrode is, in a particularly preferred embodiment, produced from a powder mixture consisting of silver and/or oxides thereof and PTFE. It is likewise possible to use doped silver and/or oxides thereof or mixtures of silver and/or oxides thereof with silver and PTFE.
The catalyst and PTFE are, for example, treated in a dry mixing process as described in U.S. Pat. No. 6,838,408 and the powder is compacted to give a sheet.
The sheet is subsequently pressed together with a mechanical support. Both the sheet formation process and the pressing of sheet and support can, for example, be carried out by means of a rolling process. The pressing force has, inter alia, an influence on the pore diameter and the porosity of the GDE. The pore diameter and the porosity have an influence on the performance of the GDE.
As an alternative, the production of the GDE according to the invention can be carried out by applying the catalyst powder mixture directly to a support, in a manner analogous to that described in DE 10 148 599 A1.
In a particularly preferred embodiment, the powder mixture is produced by mixing the catalyst powder and the binder and optionally further components. Mixing preferably occurs in a mixing apparatus which has fast-rotating mixing elements, e.g. beater blades. To mix the components of the powder mixture, the mixing elements preferably rotate at a speed of from 10 to 30 m/s or at a rate of rotation of from 4000 to 8000 rpm. After mixing, the powder mixture is preferably sieved. Sieving is preferably carried out by means of a sieving apparatus which is equipped with meshes or the like having mesh openings of from 0.04 to 2 mm.
Mixing in the mixing apparatus having rotating mixing elements produces energy into the powder mixture, as a result of which the powder mixture undergoes strong heating. If the powder is heated too strongly, an impairment of the GDE performance is observed, so that the temperature during the mixing process is preferably from 35 to 80° C. This can be achieved by cooling during mixing, e.g. by addition of a coolant, e.g. liquid nitrogen or other inert heat-absorbing substances. A further possible way of controlling the temperature is to interrupt mixing in order to allow the powder mixture to cool or to select suitable mixing apparatuses or to change the amount of filling quantity in the mixer.
The application of the powder mixture to the electrically conductive support is, for example, carried out by sprinkling. Sprinkling of the powder mixture on the support can, for example, be effected by means of a sieve. It is particularly advantageous to place a frame-like template on the support, with the template preferably being selected so that it just encompasses the support. As an alternative, the template can also be made smaller than the area of the support. In this case, an uncoated margin of the support remains free of electrochemically active coating after sprinkling-on of the powder mixture and pressing together with the support. The thickness of the template can be selected as a function of the amount of powder mixture to be applied to the support. The template is filled with the powder mixture. Excess powder can be removed by means of a scraper. The template is then removed.
In the next step, the powder mixture is, in a particularly preferred embodiment, pressed together with the support. Pressing can, in particular, be effected by means of rollers. Preference is given to using a pair of rollers. However, it is also possible to use one roller on a substantially flat plate, with either the roller or the plate being moved. Furthermore, pressing can be carried out by means of a pressing punch. The forces during pressing are, in particular, from 0.01 to 7 kN/cm.
A GDE according to the invention can in principle be made up of one or more layers. To produce multilayer GDEs, powder mixtures having different compositions and different properties are applied in layers to the support. The layers of different powder mixtures are preferably not pressed individually with the support, but instead are firstly applied in succession and subsequently pressed together with the support in one step. For example, a layer of a powder mixture which has a higher binder content, in particular a higher content of PTFE, than the electrochemically active layer can be applied. Such a layer having a high PTFE content of from 6 to 100% can act as gas diffusion layer.
As an alternative or in addition, a gas diffusion layer composed of PTFE can also be applied. A layer having a high PTFE content can, for example, be applied as lowermost layer directly onto the support. Further layers having different compositions can be applied to produce the gas diffusion electrode. In the case of multilayer GDEs, the desired physical and/or chemical properties can be set in a targeted manner. These include, inter cilia, the hydrophobicity or hydrophilicity of the layer, the electrical conductivity, the gas permeability. Thus, for example, a gradient of a property can be built up by the magnitude of the property increasing or decreasing from layer to layer.
The thickness of the individual layers of the GDE can be set by the amount of powder mixture which is applied to the support and also by the pressing forces during pressing. The amount of powder mixture applied can, for example, be set via the thickness of the template which is placed on the support in order to sprinkle the powder mixture onto the support. In the process of DE 10 148 599 A1, a sheet is produced from the powder mixture. Here, the thickness or density of the sheet cannot be set independently of one another since the parameters of rolling, e.g. roller diameter, roller spacing, roller material, clamping force and circumferential velocity, have a critical influence on these properties.
The pressing force during pressing of the powder mixture or layers composed of different powder mixtures together with the support is carried out, for example, by roller pressing with a linear pressing force in the range from 0.01 to 7 kN/cm.
The carbon dioxide GDE is preferably connected as cathode, in particular in an electrolysis cell for the electrolysis of alkali metal chlorides, preferably of sodium chloride or potassium chloride, particularly preferably of sodium chloride, or of hydrochloric acid.
The carbon dioxide GDE is particularly preferably used as cathode in chlorine electrolysis or O2 electrolysis.
The invention therefore further provides for the use of the novel gas diffusion electrode for the electrolysis of carbon dioxide to give carbon monoxide, in particular in chloralkali electrolysis.
The invention also provides a process for the electrochemical conversion of carbon dioxide into carbon monoxide, characterized in that the carbon dioxide is reacted cathodically at a novel gas diffusion electrode as described above to form carbon monoxide, and chlorine or oxygen is simultaneously produced on the anode side.
In a preferred process, the current density in the reaction is at least 2 kA/m2, preferably at least 4 kA/m2.
The invention also provides an electrolysis apparatus comprising a novel gas diffusion electrode as carbon dioxide-depolarized cathode.
The invention additionally provides a gas diffusion electrode, characterized in that the gas diffusion electrode comprises at least one sheet-like electrically conductive support element and a gas diffusion layer and an electrocatalyst applied to the support element, characterized in that in that the gas diffusion layer consists of a mixture of silver particles and PTFE, with the silver particles and the fluoropolymer having been applied in powder form to the support element and compacted and with the silver particles forming the electrocatalyst.
Preference is given to a gas diffusion electrode which has been obtained from a production process according to the invention as described above.
The GDEs produced according to the following examples were used in oxygen electrolysis. A laboratory cell which consisted of an anode space and, separated off by an ion exchange membrane, a cathode space was used for this purpose. A KHCO3 solution having a concentration of 300 g/l was used in the anode space in which oxygen was produced at a commercial DSA with iridium-coated titanium electrode. The cathode space was separated from the anode space by a commercial cation exchange membrane from Asahi Glass, Type F133. Between GDE and the cation exchange membrane, there was an electrolyte gap in which an NaHCO3 solution having a concentration of 300 g/l was circulated by pumping. The GDE was supplied via a gas space with carbon dioxide whose concentration was greater than 99.5% by volume. Areas of anodes, membrane and gas diffusion electrodes were each 3 cm2. The temperature of the electrolytes was 25° C. The current density in the electrolysis was 4 kA/m2 in all experiments.
The GDEs were produced as follows: 3.5 kg of a powder mixture consisting of 7% by weight of PTFE powder, 93% by weight of silver powder (e.g. Type 331 from Ferro) were mixed in an Ika model A11 basic mill in such a way that the temperature of the powder mixture did not exceed 55° C. This was achieved by the mixing operation being interrupted and the powder mixture being cooled down. In total, mixing was carried out three times at a mixing time of 10 seconds. After mixing, the powder mixture was sieved through a sieve having a mesh opening of 1.0 mm. The sieved powder mixture was subsequently applied to an electrically conductive support element. The support element was a gauze composed of nickel having a wire thickness of 0.14 mm and a mesh opening of 0.5 mm. Application was carried out with the aid of a 1 mm thick template, with the powder being applied using a sieve having a mesh opening of 1.0 mm. Excess powder which projected above the thickness of the template was removed by means of a scraper. After removal of the template, the support with the applied powder mixture was pressed by means of a roller press using a pressing force of from 0.4 to 1.7 kN/cm. The gas diffusion electrode was taken from the roller press.
400 ml of a 0.1 molar AgNO3 solution (6.796 g of AgNO3) were admixed with 0.8 g of trisodium citrate. 400 ml of a 0.2 molar sodium borohydride (3.024 g of NaBH4) solution were added quickly while stirring to the first solution (about 15 s, Re>10000) and stirred for 1 hour. The precipitate was filtered off, washed with water and dried overnight at 50° C.
The powder was characterized by means of BET, laser light scattering and scanning electron microscopy.
Particle size is about 145 nm in diameter and the BET surface area is 2.23 m2/g (N2 adsorption).
400 ml of a 0.1 molar AgNO3 solution (6.796 g of AgNO3) are admixed with 0.8 g of trisodium citrate, 400 ml of a 0.2 molar sodium borohydride (3.024 g of NaBH4) solution are slowly added dropwise to the first solution (about 1 hour) while stirring and stirred for 1 hour. The precipitate was filtered off, washed with water and dried overnight at 50° C. The powder is characterized by means of BET, laser light scattering and scanning electron microscopy.
Particle size is about 290 nm in diameter and the BET surface area is 0.99 m2/g (N2 adsorption).
Production of GDE with Porous Silver
The GDE was produced by the dry process, with 93% by weight of silver powder as per example 1 and 2 and LCP-1 silver from Ames Goldsmith, and 7% by weight of PTFE from DYNEON TF2053 being mixed in an Ika model A11 basic mill and subsequently pressed by means of a roller press at a force of 0.5 kN/cm. The electrode was used in the above electrolysis cell and operated at 2 and 4 kA/m2. The Faraday efficiency for CO is shown in the table below.
The examples show that both carbon dioxide GDEs produce carbon monoxide even at high current densities. However, it can be seen very clearly that the electrode containing more porous silver has a significantly higher selectivity to carbon monoxide than conventional silver. The selectivities at 2 kA/m2 are in an order of magnitude which is of great interest for industrial use. If LCP-1 silver particles, whose porosity is normally lower, are used, no CO rather only hydrogen is produced.
The BET measurements were carried out under the following conditions.
The physisorption of gases under cryogenic temperature conditions is used to determine the specific surface area (SSA) of compact finely divided or porous solids. Nitrogen is used as gas at 77K in the pressure range from 0.05 to 0.30 p/p0 (p0=saturation pressure of nitrogen at the measurement temperature) in order to determine the SSA of a sample. The amount of nitrogen which is physisorbed on the accessible surface area of the sample is measured in a static volumetric analyzer by introduction of a well-defined amount of nitrogen gas into the measurement cell containing the sample. At the same time, the pressure rise due to the introduced gas is recorded after the equilibrium state has been reached. The pressure rise (at equilibrium) is all the smaller, the larger the total area in the measurement cell, since the amount of nitrogen adsorbed on the surface cannot contribute to the pressure rise. The molar amount of nitrogen adsorbed on a sample enables the total area of the sample to be calculated by multiplication of the molar amount by the known adsorption cross section of the gas being adsorbed.
Before the adsorption measurement at 77 K, all desorbable molecules have to be vaporized from the sample surface. Thus, the sample was maintained under vacuum conditions for a number of hours at 200° C.
The measurement is then carried out in a manner analogous to the DIN ISO Standard 9277 using nitrogen of the purity class 5.0
Preparation instrument: SmartVacPrep (from Micrometrics) and gas adsorption analyzer: Gemini 2360.
The particle sizes were obtained by means of laser light scattering on a Malvern Mastersizer MS2000 Hydro MU instrument.
| Number | Date | Country | Kind |
|---|---|---|---|
| 17177031.6 | Jun 2017 | EP | regional |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2018/066293 | 6/19/2018 | WO | 00 |
