WO2025157763A1 - Gas diffusion layer for polymer-electrolyte membrane fuel cells with reduced contact resistance - Google Patents

Abstract

The invention relates to a gas diffusion layer for a fuel cell, comprising A) a flat electrically conductive material which contains at least one fiber material, said fiber material being selected from carbon fiber nonwoven materials, carbon fiber papers, carbon fiber woven fabrics, and mixtures thereof, and B) a microporous layer containing conductive carbon particles in a matrix consisting of a polymeric binder, wherein the conductive carbon particles comprise soot and graphite, and the proportion of graphite, based on the total weight of soot and graphite contained in the microporous layer, equals at least 50 wt.%, said microporous layer being applied onto at least one face of the flat electrically conductive material A). The invention additionally relates to a method for producing the gas diffusion layer, to a fuel cell containing the gas diffusion layer, and to the use of the gas diffusion layer in a polymer-electrolyte membrane fuel cell in order to reduce the contact resistance at the boundary surface between the microporous layer of the gas diffusion layer and the catalyst layer applied onto the polymer-electrolyte membrane.

Description

Gas diffusion layer for polymer electrolyte membrane fuel cells with reduced contact resistance

The present invention relates to a gas diffusion layer (GDL) for a polymer electrolyte membrane fuel cell that reduces the contact resistance at the interface between the microporous layer (MPL) of the gas diffusion layer and the catalyst layer applied to the polymer electrolyte membrane. It also relates to a method for producing the gas diffusion layer and a polymer electrolyte membrane fuel cell (PEM fuel cell) containing the gas diffusion layer.

Fuel cells use the chemical conversion of a fuel, particularly hydrogen, with oxygen to form water to generate electrical energy. In hydrogen-oxygen fuel cells, hydrogen or a hydrogen-containing gas mixture is fed to the anode, where electrochemical oxidation takes place with the release of electrons ( _H2A2H ⁺ + 2e-). The protons are transported from the anode compartment to the cathode compartment via a membrane that separates the reaction compartments from each other in a gas-tight manner and electrically insulates them. The electrons provided at the anode are conducted to the cathode via an external conductor circuit. Oxygen or an oxygen-containing gas mixture is fed to the cathode, where the oxygen is reduced and the electrons are absorbed. The oxygen anions formed react with the protons transported across the membrane to form water (1/ _2O2 + 2H ⁺ + 2e'AH2O).

For many applications, especially in automotive powertrains, low-temperature proton exchange membrane fuel cells (PEMFCs, also known as polymer electrolyte membrane fuel cells) are used. At their core is a polymer electrolyte membrane (PEM) that is only permeable to protons (or oxonium ions H ₃ O ⁺ ) and water and spatially separates the oxidizing agent, generally atmospheric oxygen, from the reducing agent. The gas-tight, electrically insulating, A catalyst layer is applied to the anode and cathode sides of a proton-conducting membrane. This layer forms the electrodes and usually contains platinum as the catalytically active metal. The actual redox reactions and charge separation take place in the catalyst layers. The membrane and catalyst layers form a unit, also known as a CCM (catalyst coated membrane). On both sides of the CCM is a gas diffusion layer (GDL), which stabilizes the cell structure and performs transport and distribution functions for reaction gases, water, heat, and electricity. The membrane, electrodes, and gas diffusion layer form the membrane electrode assembly (MEA). Flow distribution plates (so-called bipolar plates) are arranged between the membrane electrode units. These have channels for supplying the adjacent cathode and anode with process gases, and usually also have internal cooling channels. This means that in a PEM fuel cell, the GDL has direct contact with the bipolar plate.

These bipolar plates are usually metallically embossed metal plates or embossed graphitic carbon fiber mats. The webs of the bipolar plates are placed directly on the fiber side (=substrate side) of the GDL in the cell, consisting of a fiber-based substrate and a microporous layer, and form the direct contact point with the gas diffusion layer, through which electrical conduction takes place. The material flow (reaction gases oxygen and hydrogen as well as water) from the channels to the membrane and from the membrane back into the channels takes place via the channels and the pore structure of the gas diffusion layer. The gas diffusion layers located between the bipolar plates and the catalyst layers are therefore of key importance for the function and performance of the fuel cell. The process components consumed and generated in the electrode reactions must be transported through the gas diffusion layer and homogeneously distributed from the macroscopic structure of the bipolar plates to the microscopic structure of the catalyst layers. The electrons generated and consumed in the half-cell reactions must be conducted to the bipolar plates with the lowest possible voltage loss. The heat generated during the reaction must be dissipated to the coolant in the bipolar plates, so the GDL materials must also have sufficient thermal conductivity. Furthermore, the GDL must also act as a mechanical balance between the macrostructured bipolar plate and the catalyst layers. For this, component tolerances must be considered. The GDL also serves as mechanical protection for the very thin membranes, which are subject to high loads in fuel cells. Therefore, the mechanical properties of the GDL are subject to particularly stringent requirements.

Gas diffusion layers for fuel cells typically consist of a carbon fiber substrate, which is usually hydrophobically treated with fluoropolymers (e.g., PTFE) and then coated with a microporous layer (MPL). The MPL typically consists of a fluorine-containing polymer as a binder (e.g., PTFE) and a porous and electrically conductive carbon material (e.g., carbon black, graphite, graphene, carbon nanotubes (CNTs), carbon nanofibers, or a mixture thereof). The following three materials are currently used as carbon fiber substrates for GDLs:

Carbon fiber papers (wet-laid and chemically bonded carbon fiber nonwovens, with chemical binders that are carbonized),

Carbon fiber fabrics (e.g. made from yarns of oxidized but not yet carbonized polyacrylonitrile fibers, which are carbonized or graphitized after weaving),

Carbon fiber nonwovens (e.g. dry-laid, carded and hydroentangled nonwovens made of oxidized polyacrylonitrile, which are subsequently thickness calibrated and carbonized).

DE 10 2021 215 036 A1 describes a polymer electrolyte membrane fuel cell comprising a GDL and an MPL on top, wherein the MPL contains soot particles in a polymeric binder, such as PTFE. To prevent migration of the non-conductive binder into the GDL substrate, the MPL has a gradient of soot content along its thickness (z-axis).

JP200859917A describes a method for producing a gas diffusion layer for a fuel cell that is capable of regulating the water-repellent properties on the side of the flow distribution plate to prevent flooding or drying out. To avoid this, an MPL coating containing flake graphite is used.

DE 102020 121 892 A1 describes a gas diffusion layer for a fuel cell, which comprises a sheet-like electrically conductive material selected from carbon fiber nonwovens, carbon fiber wovens and mixtures thereof, wherein the sheet-like material contains at least one fluorine-containing polymer and at least one different polymer selected from polyether ketones, polyphenylene sulfides, polysulfones, polyether sulfones, partially aromatic (co)polyamides, polyimides, polyamide-imides, polyetherimides and mixtures thereof, applied thereto and/or incorporated therein, and optionally a microporous layer is applied to one of the surfaces of the electrically conductive material.

Since all components of a cell are usually joined together under high pressures, the webs of the bipolar plate can penetrate the fiber layer of the gas diffusion layer to varying degrees depending on the type of carbon fiber substrate or can mechanically change the fiber layer in its shape to varying degrees, e.g. by:

Penetration and subsequent compression of the gas diffusion layer at the web-substrate contact point (of the gas diffusion layer)

Bending or corrugation of the gas diffusion layer into the channel of the bipolar plate

Both phenomena result directly from the mechanical properties of the gas diffusion layer.

Experience has shown that the mechanical properties of a gas diffusion layer depend heavily on the type of carbon fiber substrate, which distinguishes between woven fabrics, papers, and nonwovens. In woven fabrics and dry-laid carbonized nonwovens, the fibers are mechanically bonded or intertwined by a water jet. In paper or wet-laid nonwovens, they are chemically bonded by an additional resin and then carbonized. While a woven fabric typically has an ordered structure of continuous fibers, Short-cut fibers that are chaotically and randomly oriented are used in nonwovens and papers.

The mechanical properties of the gas diffusion layer therefore have a direct influence on the effectiveness of water drainage and the pressure distribution across the cell stack. When the gas diffusion layers are installed under pressure in a fuel cell stack, they are compressed, particularly in the web area. Depending on the mechanical properties of the gas diffusion layer, this affects the contact between the MPL side of the gas diffusion layer and the carbon-coated membrane (CCM). Compression of the gas diffusion layer in the web area often results in corrugation and bending of the gas diffusion layer into the channel structure of the bipolar plate, which can lead to complete or partial delamination of the CCM to the MPL side and thus an increase in the contact resistance to the CCM. This consequently influences the current density distribution across the surface and can significantly reduce the fuel cell performance and its service life (if delamination continues, also in the web areas).

By reducing the contact area between the gas diffusion layer and the CCM, in the worst case, “bubbles” are formed between these layers, which are preferentially filled with water, which can lead to so-called “flooding”.

The contact resistance between the CCM and the gas diffusion layer significantly determines the performance of the fuel cell. A gas diffusion layer with ideal mechanical properties in terms of flexural rigidity, compressibility, and maximum contact area with the CCM can significantly reduce the contact resistance between the CCM and the gas diffusion layer and significantly increase the performance of the fuel cell.

Since the properties of the bipolar plate webs are individually designed in terms of spacing, web width, angle, and curvature depending on the application, a gas diffusion layer with specific (mechanical) properties can lead to different performance with different web profiles. The substrate of the gas diffusion layer is exposed to high stress, particularly at the web-to-channel transition. This point is particularly susceptible to damage to the microporous layer due to fiber breakage or other damage, depending on the radius and angle of the webs. Fibers have particularly sharp edges at the break point and can cause short circuits by penetrating the microporous layer and the membrane.

In the prior art, approaches to solving the problem of excessive contact resistance between a CCM and a gas diffusion layer can be found primarily in certain bipolar plate designs. However, the effects of such bipolar plate designs on the contact with the CCM are largely ignored, i.e., not discussed, in the aforementioned prior art.

For example, JP 2021 125356 A relates to a fuel cell separator with improved corrosion resistance and conductivity, comprising a substrate with a coating that is in contact with an electrode of a fuel cell. The coating contains FeCl. The separator has an electrode-side surface roughness (Sa) of 5-100 μm, preferably 10-50 μm. The coating is, for example, an electroplated layer or a sintered layer. The substrate is, for example, stainless steel, a Ti base material, or an Al base material. The electrode surface with which the coating layer is in contact is composed, for example, of a carbon base material. In this case, a contact surface pressure acting between the coating layer and the electrode surface can be 5 MPa or less, preferably 3 MPa or less. The fuel cell is, for example, a solid polymer fuel cell.

JP 2021 026909 A relates to a fuel cell in which pressure loss on the cathode side is suppressed. The fuel cell comprises a cathode separator, a cathode gas diffusion layer, a cathode catalyst electrode layer, an electrolyte layer, an anode catalyst electrode layer, an anode gas diffusion layer, and an anode separator, which are laminated in this order. The cathode separator has a flow path that allows a cathode gas to flow in the cathode gas diffusion layer. The anode separator has a flow path that allows an anode gas to flow in the anode gas diffusion layer. The cathode gas diffusion layer has higher flexural strength and in-plane air permeability per unit thickness than the anode gas diffusion layer. DE 10 2018 202 561 A1 relates to a fuel cell with an ion-selective separator, a gas diffusion layer, and a separator plate. The separator plate, together with the gas diffusion layer, forms at least one gas-conducting flow field. At least one channel web of the separator plate has an end with an upper side and an end face. The end face is configured to divide a flow impinging on the end face of the channel web in a first direction into two partial flows. The end face is configured to deflect a liquid F of the flow impinging on the end face adjacent to the upper side such that the liquid F is further away from the upper side after the deflection than before the deflection.

DE 10 2020 216 101 A1 relates to an arrangement of electrochemical cells comprising at least one gas diffusion layer, preferably with a microporous layer, a catalyst-coated membrane with a frame, and a bipolar plate. The gas diffusion layer is bonded to the catalyst-coated membrane and/or to the bipolar plate by means of a preferably electrically conductive adhesive. The adhesive is arranged on a surface of the frame of the catalyst-coated membrane, the bipolar plate, and/or optionally the microporous layer of the gas diffusion layer, and the surface is plasma-functionalized. Furthermore, the invention relates to a vehicle comprising the arrangement of electrochemical cells and a method for producing the arrangement of electrochemical cells. Through the plasma functionality of the surface or through the pretreatment of the surface with plasma, a covalent fixation of the adhesive to the frame of the membrane, the bipolar plate, or the microporous layer of the gas diffusion layer is achieved. This creates a stronger material bond and thus serves to more firmly fix the gas diffusion layer to the bipolar plate or the catalyst-coated membrane. Furthermore, better wetting of the surface with the adhesive is possible. By using the adhesive, a pressing force usually required in the arrangement of electrochemical cells, which can particularly impair the gas diffusion layer, can be significantly reduced, thus improving gas distribution in the electrochemical cell. The electrically conductive adhesive also improves the electrical contact between the bipolar plate and the gas diffusion layer. Contact resistance is reduced. Furthermore, the adhesive prevents slipping of the gas diffusion layer or the catalyst-coated membrane on the bipolar plate when stacking the assembly.

WO 2022/094717 A1 relates to devices and methods for providing a desired contact pressure distribution between fuel cell components in a fuel cell stack. In some embodiments, the technology relates to fuel cell bipolar plate designs and fuel cell stack compression systems, which can be used individually or in combination to provide a more uniform contact pressure distribution across the active area of fuel cells in a fuel cell stack.

DE 10 2020 209 811 A1 relates to a method for reducing the contact resistance between a metallic separator plate, for example a monopolar plate or a bipolar plate, and a gas diffusion layer of a fuel cell, in which the separator plate is provided at least partially with a carbon-based coating. According to the invention, at least one elastomer is added to the carbon as a binder to form the coating.

DE 10 2011 006 651 A1 relates to a fuel cell stack with improved freeze-thaw durability. The fuel cell stack comprises, in particular, a gas diffusion layer between a membrane electrode assembly and a bipolar plate. The gas diffusion layer has a structure that reduces contact resistance in a fuel cell and is cut at a specific angle such that the machine direction (high stiffness direction) of the GDL roll is not parallel to the main flow field direction of the bipolar plate, resulting in increased GDL stiffness in a width direction perpendicular to a main flow field direction of a bipolar plate.

DE 10 2010 002 392 A relates to a gas diffusion layer (GDL) for fuel cell applications that can prevent penetration into the channels of a bipolar plate. The gas diffusion layer is produced by cutting a GDL material at a specific angle so that a machine direction of the inherent high stiffness of the GDL material is not parallel to a main flow field direction of a bipolar plate in order to prevent GDL intrusion into the channels of the bipolar plate. Prevent this without changing an existing process for manufacturing the gas diffusion layer. The gas diffusion layer can improve the electrochemical performance of the fuel cell and can improve the manufacturing process even in cases where the width of the rolled GDL material is small.

US 2018/0006314 A1 relates to a bipolar plate for a battery that can improve battery efficiency by reducing contact resistance in contact with an electrode, and to a redox flow battery containing such a bipolar plate. According to at least one embodiment, a bipolar plate is provided that comprises a conductive thermoplastic portion formed on at least a portion of the plate for contact with an electrode, wherein the conductive thermoplastic portion is morphologically adapted to the electrode.

DE 10 2010 020 168 A1 relates to a bipolar plate for reducing electrical contact resistance between the plate and a diffusion layer used in a fuel cell. The opposing surfaces of the plate define flow channels with upstanding webs interspersed therewith. The webs of the plate form an electrically conductive contact with a diffusion layer in the fuel cell. At least a portion of the electrically conductive contact consists of a nickel-based alloy, which reduces the contact resistance between the plate and the diffusion layer as a way to achieve improved electrical current density. In one form, the alloy can be used as the primary material in the plate, while in another it can be used as a coating deposited onto a conventional stainless steel plate.

KR 101320786 B1 relates to a device for measuring the contact resistance and a method for measuring the contact resistance of a bipolar plate of fuel cells, wherein the contact resistance of each region of a bipolar plate can be measured. This improves the accuracy of the contact resistance measurement of the bipolar plate.

In addition, US 2023/0163314 A1 relates to a gas diffusion layer for an electrochemical device, comprising (a) a first side in contact with a catalyst layer, and (b) a second side, wherein the first side in contact with the catalyst layer has an increased surface area. Gas diffusion layers with engineered surface roughness and thus increased surface area increase the effective diffusivity of gas-phase reactants in electrochemical devices such as polymer electrolyte membrane fuel cells.

DE 10 2020 202 433 A1 relates to a method for removing a surface of a gas diffusion layer for a fuel cell, wherein the gas diffusion layer comprises:

- an electrically conductive network,

- an electrically insulating, hydrophobic material surrounding the electrically conductive network to hydrophobize the electrically conductive network,

- an electrode side facing a membrane of the fuel cell, and a bipolar plate side opposite the electrode side, which bipolar plate faces a bipolar plate of the fuel cell, wherein the method comprises at least the following steps: a) providing the gas diffusion layer, b) at least partially removing the electrically insulating, hydrophobic material surrounding the electrically conductive network at least in some regions on the bipolar plate side.

DE 10 2016 200 802 A1 relates to a flow body-gas diffusion layer unit for a fuel cell, in which a flow body and a gas diffusion layer are formed by a porous body and arranged between a membrane electrode assembly and the bipolar plates. The porosity of the body in the region in which it is formed as a flow body is greater than the porosity of the region of the body formed as a gas diffusion layer. This provides a layered structure for a fuel cell that optimizes the contact resistance between the components of the layered structure compared to the prior art.

EP 2722917 A1 relates to a gas diffusion layer for a fuel cell, which has a high electrical conductivity and at the same time good gas permeability and thus a improved performance. The gas diffusion layer has a microporous layer on one of its surfaces, which comprises a carbon material containing scale-like graphite. To further improve the electrical properties and gas permeability, the MPL can contain another carbon material that is embedded as a spacer between the graphite layers. The additional carbon material used in the microporous layer can have high specific surface areas of 1000 m ² /g or more. In the exemplary embodiments, coating agents containing scale-like graphite and acetylene black are used to produce the microporous layer. To produce the gas diffusion layer, the coating agent for forming the microporous layer (the MPL ink) is not applied directly to the GDL substrate, but rather to a carrier and cured thereon. The resulting layered MPL substrate is detached from the carrier and laminated to the GDL substrate. This procedure is considered critical to prevent the MPL layer from penetrating the GDL substrate and thus causing clumping. The problem of contact resistance between COM and GDL and ways to resolve it are not described in this document.

WO 2023/190153 A1 relates to a gas diffusion layer with a microporous layer on at least one side of a conductive porous base material, wherein the microporous layer contains carbon black and graphite particles with an aspect ratio of 10 or more, and the conductive porous base material is characterized in that the thickness of that part of the conductive porous base material in which the microporous layer is immersed is 5% or more and 20% or less of the thickness of that part of the conductive porous base material in which the microporous layer is not immersed. The electrical resistance within the gas diffusion layer is thereby reduced, while the gas diffusion capacity is not impaired. The use of such a gas diffusion layer in a fuel cell enables an improvement in its performance with regard to power generation.

The cited state of the art typically solves the problem of high contact resistances by adapting and optimizing the bipolar plate. This is roughened, for example, by mechanical processes or coatings in order to improve its surface (= To increase the contact area to the substrate side, they are bonded to the GDL using special concepts or the bipolar plates of the anode and cathode are milled complementarily to each other.

Only DE 10 2020 202 433 A1 , US 2023/0163314 A1 , DE 102016200 802 A1 and WO 2023/190153 A1 attempt to solve this problem by optimizing the GDL.

DE 10 2020 202 433 A1 and US 2023/0163314 A1 increase the surface area of the GDL by roughening it. This does not prevent bending into the channel and delamination.

DE 10 2016 200 802 A1 attempts to influence the porosity, pore shape, pore size and thermal conductivity by using an additional raw material in the form of another fiber.

WO 2023/190153 A1 describes the reduction of the electrical resistance of the GDL, but not the prevention of bending of the GDL into the channel and delamination of the MPL.

Thus, no solution approach attempts to solve the problem of increased contact resistance between the GDL and the COM in a fuel cell, especially a PEMFC, which is based on corrugation and bending of the GDL into the channel structure of the bipolar plate and the associated risk of delamination of the GDL from the COM, by a specifically aligned design of the microporous layer (MPL) of the GDL.

This problem is solved according to the invention by a gas diffusion layer for a fuel cell, which

A) a sheet-like electrically conductive material containing at least one fiber material, wherein the fiber material is selected from carbon fiber nonwovens, carbon fiber papers, carbon fiber fabrics and mixtures thereof, and B) a microporous layer containing conductive carbon particles in a matrix of a polymeric binder, wherein the conductive carbon particles comprise carbon black and graphite, wherein the graphite content, based on the total weight of carbon black and graphite contained in the microporous layer, is at least 50 wt.%, which is applied at least on one side to the sheet-like electrically conductive material A).

A preferred embodiment of the invention is a gas diffusion layer for a fuel cell, which

A) a sheet-like electrically conductive material containing at least one fiber material, wherein the fiber material comprises carbon fiber nonwoven fabric, and

B) a microporous layer containing conductive carbon particles in a matrix of a polymeric binder, wherein the conductive carbon particles comprise carbon black and graphite, wherein the graphite content, based on the total weight of carbon black and graphite contained in the microporous layer, is at least 50 wt.%, which is applied at least on one side to the sheet-like electrically conductive material A).

The use of a gas diffusion layer according to the invention in a fuel cell leads to a reduction in the contact resistance at the interface of the microporous layer of the gas diffusion layer, in particular to the catalyst layer applied to the polymer electrolyte membrane of a polymer electrolyte membrane fuel cell, especially in the region of the channels of the bipolar plate of the fuel cell, such as in particular in the region of the center of the channels, compared to gas diffusion layers from the prior art. The use of a combination of carbon black and graphite in the microporous layer allows for surprisingly improved contact resistances. The GDL according to the invention optimizes the mechanical

Properties and surface properties of the GDL point to the described problem.

The invention thus solves the described problem by optimizing the proportions of carbon black and graphite in the microporous layer (MPL). This results in special mechanical properties of the MPL layer. For example, by adding graphite particles, the MPL becomes particularly compressively soft, and thus the contact resistance between the microporous layer and the COM is particularly low due to the adhesion to the COM. This "compressively soft" property can be particularly well described by the compression set at 6.0 MPa. Without being bound to theory, it is assumed that when using a combination of carbon black and graphite in the MPL of a GDL, as with conventional MPLs, the substrate of the GDL will bulge into the flow channels of the bipolar plate after being pressed together with the other components to form a fuel cell, and the MPL will bulge with it. However, the large graphite flakes are able to spatially compensate for and cushion this bulge while maintaining contact with the COM.

In addition to the mechanical properties of the GDL substrate, the properties of the microporous layer can also significantly contribute to the interaction between COM and MPL. The surface properties of a microporous layer, such as roughness, deformability and elasticity under pressure, potential contact area, etc., are significantly determined by the nature of the materials used, such as carbon particles or binders such as PTFE.

Depending on the proportion of carbon black and graphite, different properties of the microporous layer can be achieved, e.g. different porosities, pore size distributions, permeability to gases and water, diffusion lengths, electrical and thermal conductivities, mechanical stability, surface roughness and behavior under pressure (for example, an MPL can be compressively soft).

To describe the relationship between channel properties and the aforementioned mechanical properties of the GDL, a measurement method was developed, the so-called strip measurement. This is explained in more detail below. Compressively stiff materials bend into the channel of the bipolar plate while retaining their thickness in the Z direction. This can lead to the MPL side losing contact with the CCM or delaminating. An increase in contact resistance at this point (channel) is observed with these materials. A further consequence could be a reduction in current density. Furthermore, delamination at the channel locations can continue across the entire surface, resulting in a further accelerated reduction in performance and a shorter service life of the fuel cell. Furthermore, corrugation of the gas diffusion layer into the channel of the bipolar plate can reduce the contact area between the bipolar plate and the gas diffusion layer, which can increase the electrical contact resistance between the two components.

Pressure-sensitive, compressible, and highly flexible gas diffusion layers, on the other hand, are capable of responding flexibly to different pressure conditions across the surface. Although they are also compressed at the contact points with the web and also bend into the channel on the substrate side, they are flexible enough to maintain contact with the CCM on the MPL side and prevent delamination at this point. This ensures a sufficiently low contact resistance.

The invention further relates to a method for producing such a gas diffusion layer, in which i) the fiber material is coated and/or impregnated with a preferably aqueous composition containing the at least one polymer and optionally further components, and the coated and/or impregnated fiber material is subsequently dried and/or sintered, wherein the sintering is carried out by thermal treatment at a temperature of 250 °C to 500 °C, preferably 300 °C to 450 °C, and in particular 350 °C to 450 °C, so that the sheet-like electrically conductive material A) is obtained, and subsequently ii) the sheet-like electrically conductive material A) is coated with a preferably aqueous composition containing the conductive carbon particles and the polymeric binder, and the coated material A) is subsequently dried and/or sintered, wherein the sintering is carried out by thermal Treatment is carried out at a temperature of 250 °C to 500 °C, preferably 300 °C to 450 °C, and in particular 350 °C to 450 °C, so that the microporous layer is produced.

In this application, the terms “sintering” and “sintering” are used synonymously.

The invention also relates to a fuel cell comprising at least one such gas diffusion layer, which is typically a polymer electrolyte membrane fuel cell.

Accordingly, the invention also relates to the use of such a gas diffusion layer in a fuel cell, preferably a polymer electrolyte membrane fuel cell.

Furthermore, the invention relates to the use of such a gas diffusion layer in a polymer electrolyte membrane fuel cell for reducing the contact resistance at the interface between the microporous layer of the gas diffusion layer and the catalyst layer applied to the polymer electrolyte membrane, especially in the region of the channels of the bipolar plate of the fuel cell, such as in particular in the region of the center of the channels.

Furthermore, the invention relates to the use of conductive carbon particles comprising carbon black and graphite in the microporous layer of a gas diffusion layer, which is preferably as defined above, for reducing the contact resistance at the interface between the microporous layer of the gas diffusion layer and the catalyst layer applied to the polymer electrolyte membrane of a polymer electrolyte membrane fuel cell, especially in the region of the channels of the bipolar plate of the fuel cell, such as in particular in the region of the center of the channels.

Furthermore, the invention relates to the use of a polymer selected from polyaryletherketones, polyphenylene sulfides, polysulfones, polyethersulfones, partially aromatic (co)polyamides, polyimides, polyamideimides, polyetherimides and mixtures thereof, preferably a polyaryletherketone such as polyetheretherketone, in and/or on the fiber material of a gas diffusion layer to reduce the contact resistance at the interface between the gas diffusion layer, its MPL if present, and the catalyst layer applied to the polymer electrolyte membrane of a polymer electrolyte membrane fuel cell, especially in the region of the channels of the bipolar plate of the fuel cell, such as in particular in the region of the center of the channels. The fiber material preferably comprises carbon fiber nonwoven fabric. The gas diffusion layer is particularly preferably as defined above.

It was surprisingly found that when such a polymer is used in and/or on the fiber material of a gas diffusion layer (GDL), the GDL bulges less or not at all into the channel of the bipolar plate. This can improve the adhesion of the GDL to the COM and, in particular, prevent or at least mitigate delamination of the GDL, i.e., its MPL, if present, from the COM. This results in lower contact resistance at the interface between the GDL and the COM. Thus, by impregnating and/or coating the fiber material with such a polymer, the conformal behavior of the GDL can be suitably adapted to a specific channel geometry.

Preferred embodiments of the gas diffusion layer and fuel cell according to the invention, the method according to the invention for producing the gas diffusion layer, and the uses according to the invention are set out in the dependent claims. Preferred embodiments of one of the aforementioned inventive subjects are also preferred embodiments of the other inventive subjects, unless otherwise stated.

Gas diffusion layer

The sheet-like electrically conductive material used according to the invention and the gas diffusion layer are sheet-like structures that have a substantially two-dimensional, planar extension and a comparatively smaller thickness. The gas diffusion layer has a base area that generally corresponds substantially to the base area of the adjacent membrane with the catalyst layers and the base area of the adjacent bipolar plate of the fuel cell. The shape of the base area of the gas diffusion layer can, for example, be polygonal (n-cornered with n > 3, e.g., triangular, square, pentagonal, hexagonal, etc.), circular, circular-segment-shaped (e.g., semicircular), elliptical, or elliptical-segment-shaped. The base is preferably rectangular or circular.

The gas diffusion layer according to the invention for a fuel cell comprises A) a sheet-like, electrically conductive material containing at least one fiber material, wherein the fiber material comprises carbon fiber nonwoven fabric, and B) a microporous layer containing conductive carbon particles in a matrix of a polymeric binder, wherein the conductive carbon particles comprise carbon black and graphite, wherein the graphite content, based on the total weight of carbon black and graphite contained in the microporous layer, is at least 50 wt.%. Carbon fiber nonwoven fabrics are advantageous, among other things, because they are compression-elastic and can be easily produced on an industrial scale, e.g., in a roll-to-roll process.

The conductive carbon particles used vary considerably. Carbon particles in the form of carbon black possess a largely disordered structure. They often consist of agglomerates of very small and round primary particles with a diameter that typically, depending on production conditions, is in the lower nanometer range of 5 to several hundred nm, for example, from above 10 nm to approximately 100 nm. Such agglomerates, with a diameter of typically at most several pm, can be at least partially broken up by shear forces during processing, such as the dispersion process in water, resulting in smaller particles. Carbon black powders typically have particles (agglomerates) with a D90 value of at most a few pm, in particular at most 500 nm, determined by static light scattering at 25 °C using a Mastersizer 2000 instrument from Malvern Instruments (where a wet measurement can be performed in distilled water or, preferably, a dry measurement). The D90 value of the integral volume distribution is defined in this application as the particle diameter at which 90 volume percent of the particles have a diameter smaller than the diameter corresponding to the D90 value. In a dispersion, the measured particle sizes are often significantly smaller due to the breakup of agglomerates. Graphite is a natural form of carbon with a unique hexagonal crystal structure arranged in multiple parallel planes, the graphene layers. It is accordingly platelet-like. This anisotropic structure gives graphite special properties, such as electrical conductivity and exceptional strength along the individual layers, as well as easy cleavage and good sliding and lubricating properties. It is therefore relatively stable during processing. A distinction is also made between natural and synthetic graphites. In graphite powder, the graphite is typically already present essentially as primary particles. The particles present in the powder typically have a particle diameter of more than 500 nm, preferably in the pm range. The D90 value is usually well above 1 pm, in each case determined by static light scattering at 25 °C using a Mastersizer 2000 device from Malvern Instruments (wet measurement in distilled water or, preferably, dry measurement). The D90 value of the integral volume distribution is defined in the present application as the particle diameter at which 90 volume percent of the particles have a diameter smaller than the diameter corresponding to the D90 value.

According to the invention, the fuel cell preferably comprises a polymer electrolyte membrane to which a catalyst layer is applied, wherein the catalyst layer can be brought into contact with the surface of the microporous layer of the gas diffusion layer.

Flat electrically conductive material A)

According to the invention, the fiber material preferably has a basis weight of 15 to 400 g/m ² , preferably 20 to 300 g/m ² , more preferably 30 to 150 g/m ² and in particular 40 to 120 g/m ² such as 50 to 100 g/m ² . In the present invention, the mass or total weight of the fiber material refers to the untreated fiber material, ie without any components such as polymer or other additives.

The fiber material preferably has a thickness in the range of 50 to 500 pm, particularly preferably 100 to 400 pm. This thickness refers to the non- equipped, uncompressed state of the fiber material, ie before the GDL is installed in a fuel cell.

The fibers contained in the fiber material comprise carbon fibers (carbon fibers, carbon fibers), which comprise carbon fiber nonwoven fabric, and optionally other fibers, preferably selected from glass fibers, fibers of organic polymers such as polypropylene, polyester, polyphenylene sulfide, polyether ketones, and mixtures thereof. Specifically, the fibers contained in the fiber material consist solely of carbon fibers.

Carbon fibers can be produced in a conventional manner, preferably using polyacrylonitrile fibers (PAN fibers) as the starting material. PAN fibers are produced by radical polymerization of a monomer composition that preferably contains at least 90% by weight, based on the total weight of the monomers used for polymerization, of acrylonitrile. The resulting polymer solution is spun into filaments, e.g., by wet spinning and coagulation, and gathered into tows. Before this PAN precursor is converted into carbon fibers at high temperatures, it is generally subjected to oxidative cyclization (also referred to as oxidation for short) in an oxygen-containing atmosphere at elevated temperatures of approximately 180 to 300 °C. The resulting chemical crosslinking improves the dimensional stability of the fibers. The actual pyrolysis to carbon fibers then takes place at temperatures of at least 1200 °C. Depending on the desired fiber shape, either the original fibers or a flat fiber material can be used for this pyrolysis. Depending on the temperature during pyrolysis, a distinction is made between carbonization and graphitization. Carbonization refers to a treatment at approximately 1200 to 1500 °C under an inert gas atmosphere, which leads to the release of volatile products. Graphitization, i.e., heating to approximately 2000 to 3000 °C under an inert gas, produces so-called high-modulus or graphite fibers. These fibers are highly pure, lightweight, high-strength, and highly conductive to electricity and heat. According to the invention, the fiber material is selected from carbon fiber nonwovens, carbon fiber papers, carbon fiber fabrics, and mixtures thereof. Preferably, the fiber material is, i.e., the fiber material consists of, carbon fiber nonwovens.

Furthermore, the fiber material preferably consists of carbon fiber paper. In an alternative embodiment, the fiber material is carbon fiber nonwoven fabric in combination with carbon fiber fabric or carbon fiber paper.

In carbon fiber fabrics, the flat fiber material is produced by interlacing two thread systems: warp and weft. As with textiles, fiber bundles are flexibly but inextricably linked. Oxidized, but not yet carbonized or graphitized PAN fibers are preferably used to produce carbon fiber fabrics. Carbonization or graphitization, to impart electrical conductivity to the flat fiber material, occurs after weaving.

Oxidized PAN fibers are generally used to produce carbon fiber paper. These are shredded into fiber fragments in a conventional manner, slurried, and, analogous to papermaking, a fiber matrix is produced by sieving (laths), and dried. In a preferred embodiment, at least one binder is also incorporated into the paper. Suitable binders include, for example, phenolic, furan, polyimide resins, etc. To incorporate the binder, the paper can be impregnated with it, and the binder can then be cured if necessary. After impregnation and curing, the carbon fiber paper is subjected to another carbonization/graphitization process to convert the binder into compounds with improved electrical conductivity. In another suitable embodiment, a filled carbon fiber paper is used to provide the fiber material. The initial production process is as described above, but instead of introducing a binder and carbonization/graphitization, a filler consisting of a carbon material in a polymer binder is introduced into the still-moist paper. A carbon-PTFE filler is specifically used for this purpose. This filling increases the thermal and electrical conductivity to such an extent that carbonization/graphitization is no longer necessary. Non-oxidized or oxidized PAN fibers can be used to produce carbon fiber nonwovens. In a first preferred embodiment, the fibers are first dry-laid (carded) to form a nap and then consolidated into a nonwoven. This can be done, for example, by hydroentangling, whereby the carbon fibers are oriented, entangled, and thus mechanically stabilized. If necessary, the thickness of the consolidated nonwoven can be calibrated to a desired value. Nonwovens based on non-oxidized PAN fibers are first subjected to oxidation at elevated temperature and in an oxygen atmosphere after nonwoven laying and consolidation, and then to carbonization/graphitization in an inert gas atmosphere. Nonwovens based on oxidized PAN fibers are only subjected to carbonization/graphitization after nonwoven laying and consolidation. Carbon fiber nonwovens, for the production of which the fibers are laid dry to form a pile in a first step, are a preferred embodiment of the invention.

In a preferred embodiment, the fiber material contains at least one polymer applied to and/or incorporated into it. For this purpose, the fiber material can be treated with polymer components and, if appropriate, other additives using conventional application and impregnation processes.

Preferably, polymer a1) comprises at least one fluorine-containing polymer, preferably selected from polytetrafluoroethylene (PTFE), tetrafluoroethylene-hexafluoropropylene copolymers (FEP), perfluoroalkoxy polymers (PFA), and mixtures thereof. Perfluoroalkoxy polymers include, for example, copolymers of tetrafluoroethylene (TFE) and perfluoroalkoxy vinyl ethers, such as perfluorovinyl propyl ether. Particularly preferably, the fluorine-containing polymer a1) comprises, and in particular is, polytetrafluoroethylene. Component a1) can serve to increase the hydrophobicity of the fiber material.

The mass fraction of the fluorine-containing polymer a1) is preferably 0.5 to 40%, preferably 1 to 20%, in particular 1 to 10%, based on the mass of the fiber material. In a specific embodiment, the fluorine-containing polymer a1) is PTFE, and the mass fraction of PTFE is 0.5 to 40%, preferably 1 to 20%, in particular 1 to 10%, based on the mass of the fiber material. Likewise preferably, the polymer of a1) comprises various polymers a2) selected from polyaryletherketones (PAEK), polyphenylene sulfides (PPS), polysulfones (PSU), polyethersulfones (PES), partially aromatic (co)polyamides (high-temperature polyamides, HTPA), polyimides (PI), polyamideimides (PAI), polyetherimides (PEI), and mixtures (blends) thereof. The use of such a component a2) in and/or on the fiber material of a gas diffusion layer can lead to a reduction in the contact resistance at the interface between the gas diffusion layer and the catalyst layer applied to the polymer electrolyte membrane of a polymer electrolyte membrane fuel cell, especially in the region of the channels of the bipolar plate of the fuel cell, such as in particular in the region of the center of the channels.

In particular, polymer component a2) comprises at least one polyaryl ether ketone. Specifically, polymer component a2) consists of at least one polyaryl ether ketone. Polyaryl ether ketones are semi-crystalline thermoplastics that have an alternating structure in which each aryl group is followed by a keto group (carbonyl group) or ether group. The proportions of keto and ether groups are variable and can differ in the substitution pattern on the aryl rings. Suitable polyaryl ether ketones a2) are polyether ketones (PEK), polyether ether ketones (PEEK), polyether ketone ketones (PEKK), etc. Preferably, polymer component a1) comprises at least one polyether ether ketone; in particular, polymer component a1) consists of at least one polyether ether ketone.

Suitable partially aromatic (co)polyamides a2) are, in particular, the polymers known as high-temperature polyamides (HTPAs). These are semi-crystalline or amorphous, thermoplastic, partially aromatic polyamides. These preferably contain at least one aromatic dicarboxylic acid as polymerized units, in particular selected from terephthalic acid, isophthalic acid, and mixtures of terephthalic acid and isophthalic acid. Preferred partially aromatic (co)polyamides a2) are selected from PA 6.T, PA 10.T, PA 12.T, PA 6.I, PA 10.1, PA 12.1, PA 6.T/6.I, PA 6.T/6, PA 6.T/10T, PA 10.T/6.T, PA 6.T/12.T, PA12.T/6.T, and mixtures thereof. Another specific embodiment of the polyamides a2) is polyphthalamide (PPA). Suitable polyimides a2) are polysuccinimide (PSI), polybismaleimide (PBMI), polyimidesulfone (PISO) and polymethacrylimide (PMI).

Suitable polymers a2) are also (semi-)aromatic polyesters, such as polyethylene terephthalate (PET) or polybutylene terephthalate (PBT), polycarbonates (PC) and temperature-resistant melamines, such as melamine foams filled with nanoporous SiO2 aerogels.

Preferably, the polymer a2) as defined above is selected from so-called high-performance plastics, which are characterized by properties such as a high glass transition temperature, a high melting temperature, good temperature resistance, good chemical resistance, and good mechanical properties. The polymers a2) preferably have a continuous operating temperature (continuous use temperature) of at least 150°C. Semi-aromatic and aromatic polymers are preferred as polymers a2). The polymers a2) are preferably thermoplastics.

The mass fraction of the polymer a2) is preferably 0.5 to 40%, preferably 1 to 20%, based on the mass of the fiber material a). In a specific embodiment, the polymer a2) is PEEK, and the mass fraction of the PEEK a2) is 0.5 to 40%, preferably 1 to 20%, based on the mass of the fiber material.

In a particularly preferred embodiment, the polymer a1) comprises at least one fluorine-containing polymer and/or a2) at least one polymer different from a1), as defined above, in order to simultaneously achieve a suitable hydrophobicity of the fiber material and a reduced contact resistance as explained above.

Preferably, the proportion of the fluorine-containing polymer a1), based on the total weight of fluorine-containing polymer a1) and polymer a2) different from a1), is 10 to 100 wt.%, preferably 20 to 90 wt.%, more preferably 30 to 80 wt.%, even more preferably 40 to 75 wt.%, such as in particular 40 to 60 wt.% or 60 to 75 wt.%. If a bonded fiber material is used as the fiber material, it is selected in particular from mechanically bonded fiber materials. Chemical bonding, especially with carbonizable polymeric binders, can adversely affect the flexural properties of the gas diffusion layer. Specifically, the fiber materials used according to the invention do not contain any polymers other than polymers a1) and a2) added as binders.

In one embodiment of the invention, the fibers are wet-laid in a first step. In this embodiment, the term "carbon fiber nonwoven fabric" also encompasses, for example, a wet-laid material consisting of short-cut carbon fibers, carbon black, at least one polymer a1), especially PTFE, and at least one polymer a2), especially PEEK. In contrast to carbon fiber papers known from the prior art, the wet-laid fiber materials used according to the invention contain no or only a very small proportion of phenolic resins as binders. The mass fraction of phenolic resins is preferably 0 to 10%, more preferably 0 to 5%, in particular 0 to 1%, based on the mass of the fiber material.

In many cases, the fiber material already has good electrical and thermal conductivity due to the carbon fibers used, even without conductivity-enhancing additives. To improve the electrical and thermal conductivity, however, the fiber material can additionally be treated with at least one conductivity-enhancing additive a3). Preferably, the conductivity-enhancing additive a3) is selected from metal particles, carbon black, graphite, graphene, carbon nanotubes (CNTs), carbon nanofibers, and mixtures thereof. Preferably, the conductivity-enhancing additive a3) comprises carbon black or consists of carbon black. The treatment of the fiber material with at least one conductivity-enhancing additive a3) can, for example, be carried out together with a polymer, such as, for example, polymer a1) and/or a2), and/or other additives. An aqueous dispersion is preferably used to treat the fiber material.

Preferably, the mass fraction of the conductivity-improving additive a3) is 0.5 to

45%, preferably 1 to 25%, based on the mass of the fiber material. In a specific embodiment, the conductivity-improving additive a3) comprises carbon black or consists of carbon black and the mass fraction is 0.5 to 45%, preferably 1 to 25%, based on the mass of the fiber material.

In a preferred embodiment, the fiber material contains, based on the total weight of components a1), a2) and a3), 10 to 50 wt.%, preferably 20 to 40 wt.%, of at least one fluorine-containing polymer a1), 0 to 40 wt.% of at least one polymer a2 different from a1), and 20 to 90 wt.%, preferably 50 to 80 wt.%, of at least one conductivity-improving additive a3), applied thereto and/or incorporated therein. The total weight of components a1), a2), if present, and a3) is 3 to 50 wt.%, preferably 5 to 35 wt.%, in particular 10 to 25 wt.%, based on the total weight of the fiber material.

In one embodiment, the fiber material further comprises at least one additive a4) applied thereto and/or incorporated therein, which additive is selected, for example, from polymer binders a41) different from polymer a1) and polymer a2) as defined above, such as furan resins, polyimide resins, surface-active substances a42) and further additives and auxiliaries a43).

The fiber material can be treated with at least one further additive a4), for example, together with a polymer, such as polymer a1) and/or a2), and/or other additives. The binders a41) can optionally be subsequently cured. This can be done, for example, together with drying and/or sintering following treatment with a polymer, such as polymer a1) and/or a2), or separately.

The total mass fraction of further additives a4) is preferably 0 to 80%, preferably 0 to 50%, based on the mass of the fiber material. If the fiber materials additionally contain at least one further additive a4), the total mass fraction of further additives a4) is 0.1 to 80%, preferably 0.5 to 50%, based on the mass of the fiber material.

Specifically, the fiber material contains various other polymers a41) in addition to the fluorine-containing polymers a1) and the polymers a2) in a weight proportion of at most 5%, preferably at most 1%, particularly preferably at most 0.5%, in particular at most 0.1%, based on the total weight of the fiber material, applied thereto and/or incorporated therein. More specifically, the fiber material contains no additives of further polymers a41) that are different from the fluorine-containing polymers a1) and the polymers a2). This applies especially to polymeric binders a41) that would carbonize under the manufacturing conditions of the sheet-like electrically conductive material A).

The fiber material can be provided with the polymer components such as polymers a1) and/or a2) and optionally further components such as a3) and/or a4) by conventional methods. Suitable coating and impregnation methods are described in more detail below. The subsequent drying and/or sintering, preferably drying and sintering, of the coated and/or impregnated fiber material to produce the sheet-like electrically conductive material A) according to the invention is also described in more detail below. The binding effect of the polymer is achieved through sintering. The sintering step can also be carried out without a prior drying step. However, a prior drying step is preferably carried out, which promotes a more uniform binder distribution in the sintered product of the sheet-like electrically conductive material A).

Accordingly, the present application also relates to a gas diffusion layer according to the invention as defined above, wherein the sheet-like electrically conductive material A) is obtainable by coating and/or impregnating the fiber material with a preferably aqueous composition containing at least one polymer and optionally further components, and subsequently drying and/or sintering, preferably drying and sintering, the coated and/or impregnated fiber material, wherein the sintering is carried out by thermal treatment at a temperature of 250 °C to 500 °C, preferably 300 °C to 450 °C, and in particular 350 °C to 450 °C. Accordingly, the gas diffusion layer preferably comprises a sheet-like electrically conductive material A) resulting from drying the coated and/or impregnated fiber material. Likewise preferably, the gas diffusion layer comprises a material obtained by thermal treatment at a temperature of 250 °C to 500 °C, preferably 300 °C to 450 °C, and in particular 350 °C to 450 °C. Sintered product of the sheet-like electrically conductive material A). The finishing weight of the preferably aqueous composition, ie its mass of solids used relative to the mass of the unfinished fiber material, is typically 5 to 25%, preferably 10 to 20%, such as about 15%.

Process for producing the flat electrically conductive material A)

In the first step of the method according to the invention, at least one fiber material is provided, which comprises a carbon fiber nonwoven fabric. Preferably, the fiber material consists of a carbon fiber nonwoven fabric. In an alternative embodiment, the fiber material is a carbon fiber nonwoven fabric in combination with a carbon fiber fabric or carbon fiber paper. Regarding suitable and preferred fiber materials, reference is made in full to the above statements.

The fiber material provided in the first step is coated and/or impregnated in the second step with a preferably aqueous composition containing a polymer such as at least one fluorine-containing polymer a1) and/or at least one polymer a2) different from a1), as defined above, and optionally further additives. The preferably aqueous composition can be in the form of a dispersion or a solution. Typically, it is in the form of an aqueous dispersion.

The finishing of the fiber materials by coating and/or impregnation is carried out using conventional application methods known to those skilled in the art. Preferably, a method selected from among padding, doctor blade coating, spraying, splattering, and combinations thereof is used for coating and/or impregnating the fiber materials.

In the padding process, the fiber material is passed through a padding tank containing the additive-containing solution or dispersion and then squeezed out to the desired application quantity of additive using a pair of rollers with adjustable pressure and, if necessary, gap. In the squeegee process, a distinction is made between gravure and screen printing. In gravure printing, the squeegee used is a knife-like steel belt, with or without a support blade. It serves to wipe off excess additive-containing solution or dispersion from the webs of the printing cylinder (scraping). In screen printing, however, the squeegee is usually made of rubber or plastic with a sharp or rounded edge.

During spray application, the additive-containing solution or dispersion is applied to the fiber material to be finished using a slotted nozzle.

The kiss-roll coating process is used to coat the underside of horizontally running webs. The coating medium can be applied to the web in either a counter- or co-rotating direction. Transfer rollers allow for indirect coating with small application quantities.

In a particular embodiment, the padding process is used to finish the cleaning articles according to the invention.

In the third step of the process according to the invention, the coated and/or impregnated fiber material is subjected to drying and/or sintering, preferably drying and sintering, to produce the flat, electrically conductive material A). Suitable processes for drying fiber materials coated and/or impregnated with additive-containing solutions or dispersions, such as nonwovens or woven fabrics, are known in principle. After application, for example, at least a portion of the solvent, especially the water, can be suctioned out of the fiber material, e.g., by passing it through a suction opening from which the liquid is removed by means of applied negative pressure. Alternatively or additionally, the fiber material can be dried at elevated temperature. In addition, drying can take place at reduced pressure. The fiber material is preferably dried at a temperature in the range of 20 to 250°C, particularly preferably 40 to 200°C.

In addition, i.e. after, or alternatively to drying, the coated and/or impregnated fiber material can be subjected to sintering in the third step. Sintering is preferably carried out by thermal treatment at a temperature of 250 °C to 500 °C, preferably 300 °C to 450 °C, and in particular 350 °C to 450 °C.

Microporous layer B):

The microporous layer (MPL) according to the invention contains conductive carbon particles in a matrix of a polymeric binder, wherein the conductive carbon particles comprise carbon black and graphite to reduce contact resistance as described above, wherein the graphite content, based on the total weight of carbon black and graphite contained in the microporous layer, is at least 50 wt. %. It is applied to at least one of the surfaces of the sheet-like electrically conductive material A), i.e., applied thereto on at least one side. In a preferred embodiment, it is applied to only one side of the sheet-like electrically conductive material A). In an alternative embodiment, it is applied to both sides of the sheet-like electrically conductive material A). An MPL according to the invention can also be applied to one side of the sheet-like electrically conductive material A) and an MPL not according to the invention can be applied to the other side of the sheet-like electrically conductive material A). The MPL not according to the invention also contains conductive particles, typically conductive carbon particles, preferably carbon black and/or graphite, in a matrix of a polymeric binder, such as preferably the aforementioned fluorine-containing polymers, especially polytetrafluoroethylene (PTFE). Typically, an MPL is applied only to one side of the sheet-like electrically conductive material A), as this creates a porosity gradient within the resulting GDL. This facilitates both the finely divided supply of the combustion gases to the catalyst layer and the removal of the combustion gas (in the form of water).

In a further alternative embodiment, an MPL according to the invention can also be combined with an MPL not according to the invention on at least one side of the sheet-like electrically conductive material A). The microporous layer according to the invention may contain further conductive particles, preferably further carbon particles such as in particular graphene, carbon nanotubes (CNT), carbon nanofibers or mixtures thereof.

In a preferred embodiment of the MPL according to the invention, the total weight of carbon black and graphite contained in the MPL, based on the total weight of the conductive carbon particles contained in the microporous layer, is at least 50%, preferably at least 90%, more preferably at least 95% and in particular 100%.

The polymeric binder of the MPL preferably comprises at least one fluorine-containing polymer. The fluorine-containing polymer is preferably selected from polytetrafluoroethylenes, tetrafluoroethylene-hexafluoropropylene copolymers, perfluoroalkoxy polymers, and mixtures thereof. Polytetrafluoroethylene (PTFE) is preferably used.

In a preferred embodiment, the graphite content of the microporous layer, based on the total weight of carbon black and graphite contained in the microporous layer, is 50 to 90 wt. %, more preferably 60 to 85 wt. %, and in particular 65 to 80 wt. %, such as 70 to 75 wt. With increasing graphite content of the MPL, the contact resistance at the interface between the MPL of the gas diffusion layer and the catalyst layer applied to the polymer electrolyte membrane can be further reduced, as described above. However, at very high graphite contents in the MPL, its thermal and electrical conductivity as well as the air permeability of the GDL according to Gurley typically deteriorate.

The microporous layer preferably contains carbon black in a proportion of 10 to 45 wt.%, preferably 15 to 40 wt.%, based on the total weight of the microporous layer. The microporous layer preferably contains graphite in a proportion of 30 to 80 wt.%, preferably 40 to 70 wt.%, based on the total weight of the microporous layer. The microporous layer preferably contains carbon black in a proportion of 10 to 45 wt.%, preferably 15 to 40 wt.%, and graphite in a proportion of 30 to 80 wt.%, preferably 40 to 70 wt.%, based on the total weight of the microporous layer. In a further preferred embodiment, the polymeric binder is present in the microporous layer in a proportion of 5 to 50 wt.%, preferably 15 to 30 wt.%, in particular 20 to 25 wt.%, based on the total weight of the microporous layer. In a further preferred embodiment, the microporous layer contains carbon black and graphite together in a proportion of 40 to 95 wt.%, preferably 60 to 90 wt.%, in particular 70 to 80 wt.%, based on the total weight of the microporous layer. In a further preferred embodiment, the microporous layer contains the polymeric binder in a proportion of 5 to 50 wt.%, preferably 15 to 30 wt.% and in particular 20 to 25 wt.%, and carbon black and graphite together in a proportion of 40 to 95 wt.%, preferably 60 to 90 wt.% and in particular 70 to 80 wt.%, in each case based on the total weight of the microporous layer.

In a special version, the MPL contains at least one pore-forming agent. Suitable pore-forming agents are commercially available plastic particles, e.g., made of polymethyl methacrylate (PMMA). A suitable particle size ranges from 10 to 100 pm.

The loading of the microporous layer is preferably 10 to 50 g/m ² , more preferably 10 to 30 g/m ² and especially 15 to 25 g/m ² .

Accordingly, the microporous layer B) preferably has a thickness in the range of 5 to 150 pm, particularly preferably 10 to 100 pm. This thickness refers to the uncompressed state of the microporous layer B), i.e., before the GDL is incorporated into a fuel cell.

In contrast to the macroporous sheet-like electrically conductive material A), the MPL B) is microporous with pore diameters that are generally well below five micrometers, preferably of at most 900 nm, particularly preferably of at most 500 nm, in particular of at most 300 nm. The pore size distribution can be determined using mercury porosimetry, as described in DIN ISO 15901-1:2019-03: Mercury porosimetry. The pore diameter can exhibit a bimodal or polymodal distribution curve. For example, when using a mixture of carbon black and graphite, a pore diameter distribution with multiple pore peaks can be obtained.

The presence of MPL has a significant impact on the water balance of the fuel cell. Due to the high proportion of polymeric binders such as PTFE and the smaller pores of MPL, flooding of the GDL and the electrode is made more difficult by the MPL acting as a liquid water barrier, thus favoring the mass transport of gaseous reactants to the catalyst.

The microporous layer preferably has a mean roughness value R _a of at most 10 pm, more preferably at most 5 pm, determined according to DIN 4768-1:1974-08.

In addition, the microporous layer preferably has an average roughness depth R _z of at most 60 pm, more preferably at most 30 pm, determined according to DIN 4768-1:1974-08.

Procedure for the

According to the invention, to produce the microporous layer B), the sheet-like electrically conductive material A) is typically coated with a preferably aqueous composition containing the conductive carbon particles and the polymeric binder, and the coated material A) is then dried and/or sintered, preferably dried and sintered.

The application of the MPL B) to the flat electrically conductive material A) can be accomplished in various ways. While spray, screen printing, or Meyer rod processes are often used in discontinuous production, doctor blade, slot die, and roller processes are preferred for continuous coating.

Suitable drying methods are known in principle. For example, drying can be carried out at elevated temperatures. Drying can also be carried out at reduced pressure. Drying is preferably carried out at a temperature in the range of 20 to 250 °C, particularly preferably 40 to 200 °C, such as 100 to 200 °C. Sintering is carried out by thermal treatment at a temperature of 250°C to 500°C, preferably 300°C to 450°C, and especially 350°C to 450°C. The binding effect of the polymer is achieved through sintering. The sintering step can also be performed without a prior drying step. However, a prior drying step is preferably performed, which promotes a more uniform binder distribution in the sintered product of the MPL B).

Accordingly, the present application also relates to a gas diffusion layer according to the invention as defined above, wherein the microporous layer B) is obtainable by coating the sheet-like electrically conductive material A) with a preferably aqueous composition containing the conductive carbon particles and the polymeric binder, and subsequently drying and/or sintering the coated material A), wherein the sintering is carried out by thermal treatment at a temperature of 250 °C to 500 °C, preferably 300 °C to 450 °C, and in particular 350 °C to 450 °C.

The particle diameter values D10, D50, and D90 for graphite and carbon black can be determined according to ISO 13320:2020-01 (Particle size analysis - Laser diffraction method). In a preferred embodiment, the graphite has a particle diameter value D10 of typically 0.5 to 15 pm, preferably 1 to 10 pm, in particular 3 to 7 pm, determined by static light scattering at 25°C using a Mastersizer 2000 instrument from Malvern Instruments. In a further preferred embodiment, the graphite has a particle diameter value D50 of typically 1 to 50 pm, preferably 2 to 30 pm, in particular 10 to 25 pm, determined by static light scattering at 25°C using a Mastersizer 2000 instrument from Malvern Instruments. In a further preferred embodiment, the graphite has a particle diameter value D90 of typically 2 to 100 pm, preferably 5 to 70 pm, in particular 30 to 60 pm, determined by static light scattering at 25 °C using a Mastersizer 2000 device from Malvern Instruments. The determination of the particle diameter value by static light scattering at 25 °C using a Mastersizer 2000 device from Malvern Instruments is carried out as a wet measurement, typically with the addition of distilled water. The D10, D50 or D90 value of the The integral volume distribution is defined in the present application as the particle diameter at which 10, 50 or 90 volume percent of the particles have a smaller diameter than the diameter corresponding to the D10, D50 or D90 value, respectively.

The graphite preferably has a BET surface area, determined according to ISO 9277:2010, in the range from 0.5 to 50 m ² /g, preferably from 1.0 to 25 m ² /g.

In a further preferred embodiment, the carbon black has a particle diameter value D10 of typically 5 to 50 nm, preferably 10 to 30 nm, determined by static light scattering at 25°C using a Mastersizer 2000 from Malvern Instruments. In a further preferred embodiment, the carbon black has a particle diameter value D50 of typically 20 to 200 nm, preferably 50 to 150 nm, determined by static light scattering at 25°C using a Mastersizer 2000 from Malvern Instruments. In a further preferred embodiment, the carbon black has a particle diameter value D90 of typically 40 to 500 nm, preferably 80 to 300 nm, determined by static light scattering at 25°C using a Mastersizer 2000 from Malvern Instruments. The determination of the particle diameter value by static light scattering at 25 °C using a Mastersizer 2000 instrument from Malvern Instruments is carried out as a wet measurement, typically with the addition of distilled water.

The carbon black preferably has a BET surface area, determined according to ISO 9277:2010, in the range from 10 to 600 m ² /g, preferably from 20 to 400 m ² /g.

According to the invention, a single type of carbon black or a combination with at least one other carbon black can be used as carbon black. Accordingly, the particle size distribution of the carbon black in the preferably aqueous composition can be monomodal or multimodal, such as bimodal. Likewise, a single type of graphite or a combination with at least one other graphite can be used as graphite. Accordingly, the particle size distribution of the graphite in the preferably aqueous composition can be monomodal or multimodal, such as bimodal. Accordingly, the microporous layer B) according to the invention is preferably obtainable from a preferably aqueous composition whose graphite and/or carbon black has such particle diameters.

Properties of the gas diffusion layer

The gas diffusion layer according to the invention preferably has a thickness at a compressive force of 0.025 MPa in the range of 100 to 300 pm, more preferably 130 to 250 pm, and in particular 150 to 200 pm.

As already mentioned above, according to the invention, high compression set values are advantageous with regard to reducing the contact resistance at the interface of the GDL.

Plastic deformation occurs when a material, such as a gas diffusion layer, does not fully return to its original shape after being subjected to stress, but rather undergoes a permanent deformation. Part of the deformation is elastic and thus reversible, while only a certain part is plastic and remains permanently. The property of a material to permanently change its shape when stress is applied, i.e., its deformability, is also referred to as "settling." Materials with low plastic deformability exhibit low settling behavior. The compression set value is a measure of how a material, in this case a GDL, behaves during compression deformation and subsequent relaxation. GDLs are typically heavily compressed when used in fuel cells. The proportion of elastic and plastic deformation can be used to characterize the properties of a GDL due to compression. The compression set is the permanent deformation that remains after the applied force is removed. Gas diffusion layers with low settling behavior are characterized by low compression set values. The compression set value can be determined as described in detail below under the test methods. It is possible to simultaneously determine the values of other physical parameters, such as thickness, gas permeability, and electrical resistance, each at a specific compressive force and after single or multiple force applications. Preferably, the gas diffusion layer has a compression set value at 1.0 MPa of at least 4.0 pm, preferably at least 5.0 pm, more preferably at least 7.0 pm and in particular at least 10.0 pm, such as at least 20 pm, determined on an annular sample with an inner diameter of 45 mm and an outer diameter of 56 mm, wherein the sample is subjected to three loading cycles from 0.025 MPa to 1.0 MPa and the compression set value results from the difference between the thickness measured at 1.0 MPa in the first loading cycle and in the third loading cycle.

Likewise preferably, the gas diffusion layer has a compression set value at 2.0 MPa of at least 4.0 pm, preferably at least 5.0 pm, more preferably at least 7.0 pm and in particular at least 10.0 pm, such as at least 20 pm, determined on an annular sample with an inner diameter of 45 mm and an outer diameter of 56 mm, wherein the sample is subjected to three loading cycles from 0.025 MPa to 2.0 MPa and the compression set value results from the difference between the thickness measured at 2.0 MPa in the first loading cycle and in the third loading cycle.

Likewise preferably, the gas diffusion layer has a compression set value at 6.0 MPa of at least 5.0 pm, preferably at least 10.0 pm, more preferably at least 15.0 pm and in particular at least 20.0 pm, determined on an annular sample with an inner diameter of 45 mm and an outer diameter of 56 mm, wherein the sample is subjected to three loading cycles from 0.025 MPa to 6.0 MPa and the compression set value results from the difference between the thickness measured at 6.0 MPa in the first loading cycle and in the third loading cycle.

In a preferred embodiment, the gas diffusion layer also has an air permeability, also called gas permeability or gas permeability, according to Gurley, determined according to ISO 5636-5:2013-11, of 1 to 50 seconds, preferably 5 to 25 seconds, more preferably 10 to 20 seconds.

In a further preferred embodiment, the gas diffusion layer has a

Dry diffusion length of 250 pm to 900 pm, preferably 300 to 800 pm.

Dry diffusion length refers to the actual length of the path that a The distance traveled by a gas molecule through the flat, electrically conductive material A) and the microporous layer B) in pm. It is determined using a stationary Wicke-Kallenbach cell.

fuel cell

The present invention also relates to a fuel cell which contains at least one gas diffusion layer according to the invention as described above.

In principle, the gas diffusion layer according to the invention is suitable for all common fuel cell types. Preferably, the invention relates to a polymer electrolyte membrane fuel cell. In this case, the fuel cell according to the invention comprises a polymer electrolyte membrane to which a catalyst layer is applied, wherein the catalyst layer is in contact with the surface of the microporous layer of the gas diffusion layer according to the invention.

As described above, the gas diffusion layer according to the invention is suitable, compared to gas diffusion layers from the prior art, in a polymer electrolyte membrane fuel cell for reducing the contact resistance at the interface between the microporous layer of the gas diffusion layer and the catalyst layer applied to the polymer electrolyte membrane, especially in the region of the channels of the bipolar plate of the fuel cell, such as in particular in the region of the center of the channels.

In a fuel cell according to the invention, the channels K of the bipolar plates can have different geometries, among other things with regard to the width B and the radius R of the curves (wherein the channels K can be arranged linearly and parallel on the bipolar plate, for example). The geometries (B1, R1), (B1, R2), (B2, R1) and (B2, R2) are mentioned here as examples, where “B” stands for the width and “R” for the radius of the curves of the channels K, which are each adjoined by webs S (for illustration, see Figure 1, which schematically shows a section of a bipolar plate from the prior art (in cross section)). “B1” stands for a channel width of 0.3 mm and “B2” for a channel width of 0.6 mm. “R1” stands for a radius of the channel curve of 0.13 mm and “R2” for a radius of the channel curve of 0.26 mm. The contact resistance of a GDL applied to a bipolar plate at its interface with a catalyst layer applied to a polymer electrolyte membrane can be simulated using Test Method F (strip measurement). This applies both to the channel regions of the fuel cell bipolar plate, particularly the channel center regions, and to the land regions of the fuel cell bipolar plate. The contact resistance in question is simulated by determining the contact resistance of the GDL applied to the bipolar plate in the corresponding regions at the interface with the measuring electrodes instead. For details of Test Method F, including the measuring apparatus, please refer to the following section on measurement methods.

Since, after the individual components of a fuel cell have been pressed together, the pressure to which the GDL is exposed is lower in the area of the channels of the bipolar plate, and particularly in the center of the channels, compared to the area of the webs, the contact resistance at the interface with the catalyst layer is comparatively high. Surprisingly, it was found that the GDL according to the invention leads to a significant reduction in the contact resistance at the interface with the catalyst layer in the pressed fuel cell compared to prior art gas diffusion layers. This applies particularly in the area of the channels, and especially their center.

In a fuel cell according to the invention, when using bipolar plates with a comparatively small channel width, as is the case, for example, with the geometry (B1, R1) or (B1, R2) (where the channels on the bipolar plate can be arranged linearly and parallel, for example), the contact resistance at the interface between the GDL and the catalyst layer in the region of the center of the channels of the bipolar plate, at a compression pressure of 1.0 MPa, is as follows: if the GDL is arranged on the bipolar plate such that the machine direction of the GDL runs parallel to the main flow field direction of the bipolar plate, typically < 15 mOhm*cm ² , preferably < 10 mOhm*cm ² . if the GDL is arranged on the bipolar plate such that the machine direction of the GDL runs at a 90° angle to the main flow field direction of the bipolar plate (ie the transverse direction of the machine of the GDL is parallel to the main flow field direction of the bipolar plate), typically also < 15 mOhm*cm ² , preferably < 10 mOhm*cm ² , whereby the contact resistances are typically lower in comparison.

Gas diffusion layers have a machine direction (MD), which runs parallel to the production direction of a web and thus along the roll winding, and a cross-machine direction (CMD), which runs transversely, i.e. at a 90° angle, to the production direction and thus transversely to the roll winding.

If the channels of a bipolar plate intended for use in a fuel cell are anisotropic (for example, if they run linearly and parallel), the contact resistance between the GDL and the CCM in the region of the channels can be minimized by a suitable arrangement of the GDL on the bipolar plate. Preferably, the GDL is not arranged on the bipolar plate such that the machine direction of the GDL runs parallel to the main flow field direction of the bipolar plate. Particularly preferably, the GDL is arranged on the bipolar plate such that the machine direction of the GDL forms an angle of 25° to 90°, preferably 45° to 90°, to the main flow field direction of the bipolar plate, which angle is in particular 90°. The machine direction of a GDL typically has a comparatively high flexural rigidity, so that if it is arranged not parallel, and particularly preferably at a 90° angle to the main flow field direction of the bipolar plate, the GDL intrusion into the channels of the bipolar plate and thus the contact resistance between the GDL and CCM in the region of the channels can be minimized.

In addition, in a fuel cell according to the invention, when using bipolar plates with a comparatively high channel width, as is the case, for example, with the geometry (B2, R1) or (B2, R2) (where the channels on the bipolar plate can be arranged linearly and parallel, for example), the contact resistance at the interface of the GDL to the catalyst layer in the region of the center of the channels of the bipolar plate, at a compression pressure of 1.0 MPa, is as follows: if the GDL is arranged on the bipolar plate such that the machine direction of the GDL runs parallel to the main flow field direction of the bipolar plate, typically < 40 mOhm*cm ² , preferably < 25 mOhm*cm ² and in particular < 15 mOhm*cm ² . if the GDL is arranged on the bipolar plate such that the machine direction of the GDL runs at a 90° angle to the main flow field direction of the bipolar plate (ie the transverse machine direction of the GDL runs parallel to the main flow field direction of the bipolar plate), typically < 30 mOhm*cm ² , preferably < 20 mOhm*cm ² and in particular < 10 mOhm*cm ² .

As already mentioned above, gas diffusion layers according to the invention exhibit relatively high compression set values due to the compressively soft MPL. The compression set at 2.0 MPa and the compression set at 6.0 MPa is particularly preferably above 20 μm. In such gas diffusion layers, the contact resistance at the interface between the MPL and the catalyst layer in the region of the channels of the bipolar plate is typically so reduced that it differs at most slightly from the contact resistance in the region of the webs. This means that the adhesion and electrical conductivity differ only slightly between the channel and web regions.

The invention is explained using the following non-limiting examples.

Measurement methods

The following describes in more detail the test methods underlying the present application and, in particular, the evaluation of the (comparative) examples:

Roughness (Test Method A)

The determination of the roughness, for example of the microporous layer, can be carried out using conventional stylus methods known to the person skilled in the art, such as those described in DIN 4768-1:1974-08 entitled "Determination of the roughness measurement quantities R _a , R _z , Rmax with electrical stylus instruments; basic principles". The mean roughness value R _a (mean distance of a measuring point on the surface from the center line) and the mean roughness depth R _z were determined. The measurements were performed using a Mahrsurf XCR20 measuring instrument with an MFW-250 free-surface probe. The values are averages of six measurements: three in the machine direction (MD) and three perpendicular to the machine direction (CD).

thickness

Thicknesses in the uncompressed state can be determined according to DIN 53855-1:1993-08 "Determination of the thickness of textile fabrics" (test method B1).

The determination of the thickness of a GDL at a specific compressive force (e.g. at 0.025 MPa) can be carried out in a compression set measuring device as described in detail below (Test Method B2).

Three samples (left, right, and center) are taken across the entire width of the GDL to be tested, and an average value is calculated. If the material has a machine direction due to manufacturing, the samples are taken perpendicular to the machine direction (CMD). The samples are ring-shaped with an inner diameter of 45 mm and an outer diameter of 56 mm. The sample area is 8.72577 ^cm2 . In a testing machine, the samples are subjected to a time-varying compressive force acting perpendicular to the surface of the sample. A sensor determines the change in the thickness of the GDL over time for each applied pressure. The sample is mounted on a device for determining elastic and plastic deformation using force sensors, with the movement being transmitted to the sample via springs. The travel distance until the maximum compressive force is reached is determined using displacement sensors. Since the deformation of the sample is non-linear, the measurement curve is adapted to the relative change. One measurement cycle, ie a single loading to maximum pressure and the subsequent unloading, lasts 1 minute. The sample undergoes three loading cycles. The initial value, at which only a small force is exerted on the sample, is 0.025 MPa. Typical pressure values for determining the compression set value (and other physical quantities, such as thickness, electrical conductivity or sheet resistance, gas permeability, etc.) are 0.6 MPa, 1.0 MPa, 2.0 MPa, 2.4 MPa and 6.0 MPa.

The compression set value for a specific pressure is the difference between the thickness measured at that pressure in the first load cycle and the thickness measured at that pressure in the third load cycle.

Air permeability (gas permeability, gas permeability) according to Gurley (test method D)

Gas permeability perpendicular to the material plane can be determined using a Gurley measurement, which can be performed using an automated Gurley densometer from Gurley Precision Instruments. The measurement determines the time in seconds until 100 ^cm³ of air has flowed vertically through the GDL sample at a constant pressure difference, with a sample area of 6.42 ^cm² . The determination of air permeability according to Gurley is described in ISO 5636-5.

Dry diffusion length (test method E)

The dry diffusion length of a GDL is determined using a stationary Wicke-Kallenbach cell.

The contact resistance of a GDL applied to a bipolar plate at its interface with a catalyst layer applied to a polymer electrolyte membrane can be simulated using strip measurement. This applies both to the channel regions of the fuel cell bipolar plate, particularly the channel center regions, and to the bridge regions of the fuel cell bipolar plate. The respective contact resistance is simulated by determining the contact resistance of a circular GDL sample applied to the channel structure of the corresponding bipolar plate at its interface with measuring electrodes, specifically in the desired region (i.e., for example, in the channel center region and/or bridge region). For this purpose, the circular GDL sample is clamped between the respective channel structure at the top and the measuring electrodes 1-20 below at a compression pressure of 1.0 MPa, in the specified Angle to the main flow field direction of the channel structure. The line of the 20 measuring electrodes is arranged in the apparatus perpendicular to the main flow field direction of the channel. The channels can, for example, have the geometry (B1, R1), (B1, R2), (B2, R1) or (B2, R2) as defined above. The circular GDL sample is large enough that its maximum width covers at least one channel and the two adjacent webs. The measuring electrodes 1 and 2 or 19 and 20 are arranged in the region of the webs on the GDL sample; the measuring electrodes 10 and 11 are also arranged in the region of the channel center on the GDL sample. Further details of the measuring method, including the measuring apparatus, can be found in the Journal of The Electrochemical Society, 159 (6) B709-B713 (2012), which is hereby incorporated by reference into the disclosure of the present application. This concerns in particular the experimental part of the said publication, including the figures, in particular Figure 1, to which reference is made therein (and in particular the statements under ‘Measurement of the contact resistance’).

The invention is explained using the following non-limiting examples.

FIGURE DESCRIPTION

Figure 1 schematically illustrates a section of a prior art bipolar plate, such as can also be used in a fuel cell according to the invention (sectional drawing). "B" represents the width and "R" the radius of the curves of the channel K shown, which is adjoined by webs S on both sides.

EXAMPLES

I) Production of gas diffusion layers

As described below, gas diffusion layers according to the invention and as a comparison were produced, each containing a sheet-like electrically conductive material A) and a microporous layer B). The impregnation composition and MPL- An aqueous dispersion was used in the GDL production of the paste. In the present application, additives such as surfactants are not considered to be part of the solids of an impregnation composition or MPL paste, particularly since they typically decompose during sintering.

Comparative example 1 (VB1)

To produce a flat, electrically conductive material, a nonwoven fabric I made of 100% carbon fibers with a basis weight of 40 g/ ^m² was used. To finish the nonwoven fabric, an impregnation composition was mixed which contained, based on the solids, 70 wt.% carbon black and 30 wt.% PTFE. The finishing was carried out by pad impregnation with an aqueous dispersion with 15% finish weight based on the mass of the GDL substrate (i.e., nonwoven fabric I, which corresponds to a finish of 6.0 g/ ^m² ). This was followed by drying for 5 minutes at 160 °C and sintering for 10 minutes at 400 °C. An MPL was then applied. For the MPL coating, an MPL paste containing, based on the solids, 20 wt.% PTFE and 80 wt.% carbon black in distilled water was applied to the fiber material. The fiber material was then dried at 130 °C and sintered at 400 °C to produce the desired gas diffusion layer. The resulting MPL loading was 15 g/m ² .

Example 2 (B2)

The production of the gas diffusion layer in Example 2 differs from that of Comparative Example 1 only in the MPL paste used. For the MPL coating, an MPL paste containing, based on the solids content, 20 wt.% PTFE, 40 wt.% carbon black, and 40 wt.% graphite in distilled water was applied to the fiber material. The fiber material was then dried at 130 °C and sintered at 400 °C to produce the desired gas diffusion layer. The resulting MPL loading was 15 g/ ^m² .

Example 3 (B3) The production of the gas diffusion layer in Example 3 differs from that of Comparative Example 1 only in the MPL paste used. For the MPL coating, an MPL paste containing, based on the solids content, 20 wt.% PTFE, 20 wt.% carbon black, and 60 wt.% graphite in distilled water was applied to the fiber material. The fiber material was then dried at 130 °C and sintered at 400 °C to produce the desired gas diffusion layer. The resulting MPL loading was 20 g/ ^m² .

Example 4 (B4)

The production of the gas diffusion layer from Example 4 differs from the production of the gas diffusion layer from Comparative Example 1 in that (i) a nonwoven fabric II made of 100% carbon fibers with a basis weight of 63 g/m ² was used to produce the sheet-like electrically conductive material (the impregnation composition was also applied with a 15% finish weight based on the mass of the GDL substrate (i.e., nonwoven fabric II, which corresponds to a finish of 9.5 g/m ² )), and (ii) by the MPL paste used. For the MPL coating, an MPL paste was applied to the fiber material, which contained, based on the solids, 20 wt.% PTFE, 20 wt.% carbon black, and 60 wt.% graphite in distilled water. The fiber material was then dried at 130 °C and sintered at 400 °C to produce the desired gas diffusion layer. The resulting MPL loading was 13 g/m ² .

Example 5 (B5)

The production of the gas diffusion layer from Example 5 differs from the production of the gas diffusion layer from Comparative Example 1 in that (i) a nonwoven fabric II made of 100% carbon fibers with a basis weight of 63 g/ ^m² was used to produce the sheet-like electrically conductive material (the impregnation composition was also applied at a weight of 15% based on the mass of the GDL substrate (i.e., nonwoven fabric II)), and (ii) by the MPL paste used. For the MPL coating, an MPL paste containing, based on the solids, 20 wt.% PTFE, 40 wt.% carbon black, and 40 wt.% graphite in distilled water was applied to the fiber material. The fiber material was then dried at 130 °C and sintered at 400 °C to produce the desired gas diffusion layer. The resulting MPL loading was 25 g/m ² .

Comparison example 6 (VB6)

To produce a flat, electrically conductive material, a nonwoven III made of 100% carbon fibers with a basis weight of 63 g/ ^m² was used. To finish the nonwoven, an impregnation composition was mixed which contained, based on the solids, 70 wt.% carbon black and 30 wt.% PTFE. The finishing was carried out by pad impregnation with an aqueous dispersion with 15% finish weight based on the mass of the GDL substrate (i.e., nonwoven III, which corresponds to a finish of 9.5 g/ ^m² ). This was followed by drying for 5 minutes at 160 °C and sintering for 10 minutes at 400 °C. An MPL was then applied. For the MPL coating, an MPL paste containing, based on the solids, 20 wt.% PTFE and 80 wt.% carbon black in distilled water was applied to the fiber material. The fiber material was then dried at 160 °C and sintered at 400 °C to produce the desired gas diffusion layer. The resulting MPL loading was 23 g/m ² .

Comparison example 7 (VB7)

The production of the gas diffusion layer from Comparative Example 7 differs from that of Comparative Example 6 only in the impregnation composition used. For finishing nonwoven III, an impregnation composition was mixed containing, based on the solids, 60 wt.% carbon black, 22 wt.% PTFE, and 18 wt.% PEEK. The finishing was carried out by padding impregnation with an aqueous dispersion containing 15% of the finishing weight based on the mass of the GDL substrate (i.e., nonwoven III). This was followed by drying for 5 minutes at 160 °C and sintering for 10 minutes at 400 °C.

Example 8 (B8) The production of the gas diffusion layer in Example 8 differs from that of Comparative Example 6 only in the MPL paste used. For the MPL coating, an MPL paste containing, based on the solids content, 20 wt.% PTFE, 40 wt.% carbon black, and 40 wt.% graphite in distilled water was applied to the fiber material. The fiber material was then dried at 160 °C and sintered at 400 °C to produce the desired gas diffusion layer. The resulting MPL loading was 15 g/ ^m² .

Example 9 (B9)

The production of the gas diffusion layer from Example 9 differs from the production of the gas diffusion layer from Comparative Example 6 in (i) the impregnation composition used. For example, to finish nonwoven fabric III, an impregnation composition was mixed that contained, based on solids, 60% carbon black, 22% PTFE, and 18% PEEK. The finishing was carried out by padding impregnation with an aqueous dispersion containing 15% finishing weight based on the mass of the GDL substrate (i.e., nonwoven fabric III). This was followed by drying for 5 minutes at 160°C and sintering for 10 minutes at 400°C. The production of the gas diffusion layer from Example 9 further differs from the production of the gas diffusion layer from Comparative Example 6 in (ii) the MPL paste used. For the MPL coating, an MPL paste containing 20 wt% PTFE, 40 wt% carbon black, and 40 wt% graphite in distilled water, based on the solids, was applied to the fiber material. The fiber material was then dried at 160 °C and sintered at 400 °C to produce the desired gas diffusion layer. The resulting MPL loading was 15 g/ ^m² .

Example 10 (B10)

The production of the gas diffusion layer from Example 10 differs from the production of the gas diffusion layer from Comparative Example 6 only in the MPL paste used. For the MPL coating, an MPL paste containing, based on the solids, 20 wt.% PTFE, 20 wt.% carbon black, and 60 wt.% graphite in distilled water was applied to the fiber material. The fiber material was then dried at 160 °C and sintered at 400 °C to produce the desired gas diffusion layer. The resulting MPL loading was 20 g/m ² .

Example 11 (B11)

The production of the gas diffusion layer from Example 11 differs from the production of the gas diffusion layer from Comparative Example 6 in (i) the impregnation composition used. For example, to finish nonwoven fabric III, an impregnation composition was mixed that contained, based on the solids, 60 wt.% carbon black, 22 wt.% PTFE, and 18 wt.% PEEK. The finishing was carried out by padding impregnation with an aqueous dispersion containing 15% finish weight based on the mass of the GDL substrate (i.e., nonwoven fabric III). This was followed by drying for 5 minutes at 160°C and sintering for 10 minutes at 400°C. The production of the gas diffusion layer from Example 11 differs from the production of the gas diffusion layer from Comparative Example 6 further in (ii) the MPL paste used. For the MPL coating, an MPL paste containing 20 wt% PTFE, 20 wt% carbon black, and 60 wt% graphite in distilled water, based on the solids, was applied to the fiber material. The fiber material was then dried at 160 °C and sintered at 400 °C to produce the desired gas diffusion layer. The resulting MPL loading was 20 g/ ^m² .

Comparison example 12 (VB12)

The production of the gas diffusion layer from Comparative Example 12 differs from the production of the gas diffusion layer from Comparative Example 1 only in the impregnation composition used. For finishing nonwoven fabric I, an impregnation composition was mixed containing, based on the solids, 60 wt.% carbon black, 28 wt.% PTFE, and 12 wt.% PEEK. The finishing was carried out by padding impregnation with an aqueous dispersion containing 15% of the finishing weight based on the mass of the GDL substrate (i.e., nonwoven fabric I). This was followed by drying for 5 minutes at 160 °C and sintering for 10 minutes at 400 °C.

Comparative Example 13 (VB13) The production of the gas diffusion layer from Comparative Example 13 differs from the production of the gas diffusion layer from Comparative Example 1 only in the impregnation composition used. For finishing nonwoven fabric I, an impregnation composition was mixed containing, based on the solids, 50 wt.% carbon black, 26 wt.% PTFE, and 24 wt.% PEEK. The finishing was carried out by padding impregnation with an aqueous dispersion containing 15% of the finishing weight based on the mass of the GDL substrate (i.e., nonwoven fabric I). This was followed by drying for 5 minutes at 160 °C and sintering for 10 minutes at 400 °C.

Example 14 (B14)

The production of the gas diffusion layer from Example 14 differs from the production of the gas diffusion layer from Comparative Example 1 in (i) the impregnation composition used. Thus, to finish nonwoven fabric I, an impregnation composition was mixed which contained, based on the solids, 60 wt.% carbon black, 28 wt.% PTFE, and 12 wt.% PEEK. The finishing was carried out by padding impregnation with an aqueous dispersion with 15% finish weight based on the mass of the GDL substrate (i.e., nonwoven fabric I). This was followed by drying for 5 minutes at 160°C and sintering for 10 minutes at 400°C. The production of the gas diffusion layer from Example 14 also differs from the production of the gas diffusion layer from Comparative Example 1 in the MPL paste used. For the MPL coating, an MPL paste containing 20 wt% PTFE, 40 wt% carbon black, and 40 wt% graphite in distilled water, based on the solids, was applied to the fiber material. The fiber material was then dried at 130 °C and sintered at 400 °C to produce the desired gas diffusion layer. The resulting MPL loading was 15 g/ ^m² .

Example 15 (B15)

The production of the gas diffusion layer from Example 15 differs from the production of the gas diffusion layer from Comparative Example 1 in (i) the impregnation composition used. For example, an impregnation composition containing, based on the solids, 50 wt.% carbon black, 26 wt.% PTFE, and 24 wt.% PEEK was mixed to finish the nonwoven fabric. The finishing was carried out by padding. Impregnation with an aqueous dispersion with 15% finish weight based on the mass of the GDL substrate (i.e., nonwoven fabric I). This was followed by drying for 5 minutes at 160 °C and sintering for 10 minutes at 400 °C. The production of the gas diffusion layer from Example 15 differs from the production of the gas diffusion layer from Comparative Example 1 further in the MPL paste used. For the MPL coating, an MPL paste was applied to the fiber material containing, based on the solids, 20 wt.% PTFE, 40 wt.% carbon black, and 40 wt.% graphite in distilled water. The fiber material was then dried at 130 °C and sintered at 400 °C to produce the desired gas diffusion layer. The resulting MPL loading was 15 g/m ² .

Example 16 (B16)

The production of the gas diffusion layer from Example 16 differs from the production of the gas diffusion layer from Comparative Example 1 in (i) the impregnation composition used. Thus, to finish the nonwoven fabric, an impregnation composition was mixed which contained, based on the solids, 60 wt.% carbon black, 28 wt.% PTFE, and 12 wt.% PEEK. The finishing was carried out by padding impregnation with an aqueous dispersion with 15% finish weight based on the mass of the GDL substrate (i.e., nonwoven fabric I). This was followed by drying for 5 minutes at 160°C and sintering for 10 minutes at 400°C. The production of the gas diffusion layer from Example 16 also differs from the production of the gas diffusion layer from Comparative Example 1 in the MPL paste used. For the MPL coating, an MPL paste containing 20 wt% PTFE, 20 wt% carbon black, and 60 wt% graphite in distilled water, based on the solids, was applied to the fiber material. The fiber material was then dried at 130 °C and sintered at 400 °C to produce the desired gas diffusion layer. The resulting MPL loading was 20 g/ ^m² .

Example 17 (B17)

The production of the gas diffusion layer from Example 17 differs from the production of the gas diffusion layer from Comparative Example 1 in (i) the impregnation composition used. Thus, for the finishing of nonwoven fabric I, a Impregnation composition containing, based on the solids, 50 wt.% carbon black, 26 wt.% PTFE, and 24 wt.% PEEK was mixed. The finishing was carried out by padding impregnation with an aqueous dispersion with 15% finishing weight based on the mass of the GDL substrate (i.e., nonwoven fabric I). This was followed by drying for 5 minutes at 160 °C and sintering for 10 minutes at 400 °C. The production of the gas diffusion layer from Example 17 differs from the production of the gas diffusion layer from Comparative Example 1 further in the MPL paste used. For the MPL coating, an MPL paste was applied to the fiber material which, based on the solids, contained 20 wt.% PTFE, 20 wt.% carbon black, and 60 wt.% graphite in distilled water. The fiber material was then dried at 130 °C and sintered at 400 °C to produce the desired gas diffusion layer. The resulting MPL loading was 20 g/m ² .

Example 18 (B18)

The production of the gas diffusion layer in Example 18 differs from the production of the gas diffusion layer in Example 2 in the choice of carbon layer. In Example 18, a carbonized carbon fiber paper with a basis weight of 44.5 g/ ^m² and a thickness (at 0.25 bar) of 192 μm was used. This carbon fiber paper was impregnated with a PTFE dispersion containing 10% finishing weight based on the mass of the GDL substrate. This was followed by drying for 5 minutes at 160 °C and sintering for 10 minutes at 400 °C. For the MPL coating, an MPL paste containing, based on the solids, 20 wt.% PTFE, 40 wt.% carbon black, and 40 wt.% graphite in distilled water was applied to the fiber material. The fiber material was then dried at 160 °C and sintered at 400 °C to produce the desired gas diffusion layer. MPL loading was performed analogously to Example 2.

Example 19 (B19)

The production of the gas diffusion layer in Example 19 differs from the production of the gas diffusion layer in Example 4 in the choice of carbon layer. For example, in Example 19, a carbonized carbon fiber paper with a basis weight of 44.5 g/ ^m² and a thickness (at 0.25 bar) of 192 μm was used. This carbon fiber paper was impregnated with a PTFE dispersion at a weight of 10% based on the mass of the GDL substrate. This was followed by drying for 5 minutes at 160 °C and sintering for 10 minutes at 400 °C. The fiber material was then dried at 130 °C and sintered at 400 °C to produce the desired gas diffusion layer. The MPL loading was carried out analogously to Example 4.

Example 20 (B20)

The production of the gas diffusion layer in Example 20 differs from that of Example 3 in the choice of carbon layer. In Example 20, a carbonized carbon fiber paper with a basis weight of 44.5 g/ ^m² and a thickness (at 0.25 bar) of 192 μm was used. This carbon fiber paper was impregnated with a PTFE dispersion at 10% finish weight based on the mass of the GDL substrate. This was followed by drying for 5 minutes at 160 °C and sintering for 10 minutes at 400 °C. The MPL loading was carried out analogously to Example 3.

II) Properties of the gas diffusion layers

Table 1 below shows the contact resistances at the interface between the microporous layer (MPL) of the gas diffusion layers from the (comparative) examples and the catalyst layer deposited on a polymer electrolyte membrane, as simulated using test method F (strip measurement). Bipolar plates with linear and parallel channels of various channel geometries were used for the strip measurement, namely (B1, R1), (B1, R2), (B2, R1), and (B2, R2). It also shows the compression set values of the gas diffusion layers from the (comparative) examples at different pressures:

bThe GDL sample is arranged on the bipolar plate so that the machine direction (MD) of the GDL is parallel to the

Main flow field direction of the bipolar plate. cThe GDL sample is arranged on the bipolar plate so that the transverse direction of the machine (CMD) of the GDL is parallel to the

5 Main flow field direction of the bipolar plate (i.e. the GDL is arranged on the bipolar plate so that the machine direction (MD) of the GDL forms an angle of 90° to the main flow field direction of the bipolar plate).

As shown, for example, by a comparison of inventive examples 2 and 3 with comparative example 1, of inventive examples 8 and 10 with comparative example 6, of inventive examples 14 and 16 with comparative example 12, of inventive examples 15 and 17 with comparative example 13, and of inventive examples 4 and 5 with one another, the additional use of graphite in a soot-containing MPL, particularly in the region of the channels, leads to a significant reduction in the contact resistance at the interface between the MPL of the corresponding gas diffusion layer and the COM, with excellent contact resistances being able to be achieved, particularly with a high graphite content of, for example, 3:1 compared to the soot content. The MPL of the gas diffusion layers from the inventive examples is comparatively “compressively soft,” as the compression set values of the gas diffusion layers in question show.

The aforementioned results also show that the arrangement of the GDL sample on the bipolar plate influences the contact resistance due to the anisotropic arrangement of the channels on the bipolar plate. In the presence of such anisotropic channel arrangements, particularly low contact resistances can be achieved if the GDL is arranged on the bipolar plate such that the machine direction (MD) of the GDL forms an angle of 90° to the main flow field direction of the bipolar plate. Reduced contact resistances in the channel center can also be achieved by choosing a comparatively narrow channel width, as is the case, for example, with geometry (B1, R1) or (B1, R2) (compare with the contact resistances for geometry (B2, R1) or (B2, R2)).

As shown, for example, by comparing Examples 6 and 7 with each other, Examples 8 and 9 with each other, Examples 10 and 11 with each other, Examples 12 and 13 with Example 1, Examples 14 and 15 with Example 2, and Examples 16 and 17 with Example 3, the use of a polyetheretherketone in and/or on the fiber material of the sheet-like electrically conductive material of a gas diffusion layer leads to a significant reduction in the contact resistance at its interface with the CCM. The contact resistance can also be influenced by the weight ratio of PEEK to PTFE in the impregnation composition used. The contact resistance of gas diffusion layers according to the invention with a carbon black and graphite containing MPL at its interface to the CCM can thus be further reduced by a suitable choice of impregnation of the fiber material.

Claims

Patent claims 1. Gas diffusion layer for a fuel cell, which A) a sheet-like electrically conductive material containing at least one fiber material, wherein the fiber material is selected from carbon fiber nonwovens, carbon fiber papers, carbon fiber fabrics and mixtures thereof, and B) a microporous layer containing conductive carbon particles in a matrix of a polymeric binder, wherein the conductive carbon particles comprise carbon black and graphite, wherein the graphite content, based on the total weight of carbon black and graphite contained in the microporous layer, is at least 50 wt.%, which is applied at least on one side to the sheet-like electrically conductive material A). 2. Gas diffusion layer for a fuel cell, which A) a sheet-like electrically conductive material containing at least one fiber material, wherein the fiber material comprises carbon fiber nonwoven fabric, and B) a microporous layer containing conductive carbon particles in a matrix of a polymeric binder, wherein the conductive carbon particles comprise carbon black and graphite, wherein the graphite content, based on the total weight of carbon black and graphite contained in the microporous layer, is at least 50 wt.%, which is applied at least on one side to the sheet-like electrically conductive material A). 3. Gas diffusion layer according to claim 1 or 2, which has one or more of the following properties: the graphite has a particle diameter value D10 of 0.5 to 15 pm, preferably 1 to 10 pm, in particular 3 to 7 pm, determined according to ISO 13320:2020-01 by means of static light scattering at 25 °C, the graphite has a particle diameter value D50 of 1 to 50 pm, preferably 2 to 30 pm, in particular 10 to 25 pm, determined according to ISO 13320:2020-01 by means of static light scattering at 25 °C, the graphite has a particle diameter value D90 of 2 to 100 pm, preferably 5 to 70 pm, in particular 30 to 60 pm, determined according to ISO 13320:2020-01 by means of static light scattering at 25 °C. 4. Gas diffusion layer according to one of claims 1 to 3, which has one or more of the following properties: the carbon black has a particle diameter value D10 of 5 to 50 nm, preferably 10 to 30 nm, determined according to ISO 13320:2020-01 by means of static light scattering at 25 °C, the carbon black has a particle diameter value D50 of 20 to 200 nm, preferably 50 to 150 nm, determined according to ISO 13320:2020-01 by means of static light scattering at 25 °C, the carbon black has a particle diameter value D90 of 40 to 500 nm, preferably 80 to 300 nm, determined according to ISO 13320:2020-01 by means of static light scattering at 25 °C. 5. Gas diffusion layer according to one of claims 1 to 4, wherein the carbon black has a BET surface area, determined according to ISO 9277:2010, in the range from 10 to 600 m ² /g, preferably from 20 to 400 m ² /g. 6. Gas diffusion layer according to one of claims 1 to 5, wherein the graphite has a BET surface area, determined according to ISO 9277:2010, in the range from 0.5 to 50 m ² /g, preferably from 1.0 to 25 m ² /g. 7. Gas diffusion layer according to one of the preceding claims, wherein the fiber material is carbon fiber nonwoven fabric. 8. Gas diffusion layer according to one of the preceding claims, wherein the fiber material contains at least one polymer applied thereto and/or incorporated therein. 9. Gas diffusion layer according to claim 8, wherein the polymer a1) comprises at least one fluorine-containing polymer, preferably selected from polytetrafluoroethylene, tetrafluoroethylene-hexafluoropropylene copolymers, perfluoroalkoxy polymers and mixtures thereof, and in particular polytetrafluoroethylene, and/or a2) at least one polymer different from a1) selected from polyaryletherketones, polyphenylene sulfides, polysulfones, polyethersulfones, partially aromatic (co)polyamides, polyimides, polyamideimides, polyetherimides and mixtures thereof, preferably selected from polyaryletherketones such as polyetheretherketones. 10. Gas diffusion layer according to claim 8 or 9, wherein the fiber material contains at least one conductivity-improving additive a3) applied thereto and/or incorporated therein, wherein the conductivity-improving additive a3) is preferably selected from metal particles, carbon black, graphite, graphene, carbon nanotubes (CNT), carbon nanofibers and mixtures thereof, and in particular carbon black. 11. Gas diffusion layer according to one of claims 8 to 10, wherein the sheet-like electrically conductive material A) is obtainable by coating and/or impregnating the fiber material with a preferably aqueous composition which contains the at least one polymer and optionally further components, and subsequently drying and/or sintering the coated and/or impregnated fiber material, wherein the sintering is carried out by thermal treatment at a temperature of 250 °C to 500 °C, preferably 300 °C to 450 °C, and in particular 350 °C to 450 °C. 12. Gas diffusion layer according to one of the preceding claims, wherein the graphite content of the microporous layer, based on the total weight of carbon black and graphite contained in the microporous layer, is 50 to 90 wt.%, more preferably 60 to 85 wt.%, and in particular 65 to 80 wt.%, such as 70 to 75 wt.%. 13. Gas diffusion layer according to one of the preceding claims, wherein the microporous layer contains carbon black in a proportion of 10 to 45 wt.%, preferably 15 to 40 wt.%, based on the total weight of the microporous layer. 14. Gas diffusion layer according to one of the preceding claims, wherein the microporous layer contains graphite in a proportion of 30 to 80 wt.%, preferably 40 to 70 wt.%, based on the total weight of the microporous layer. 15. Gas diffusion layer according to one of the preceding claims, wherein the polymeric binder is contained in the microporous layer in a proportion of 5 to 50 wt.%, preferably 15 to 30 wt.%, in particular 20 to 25 wt.%, based on the total weight of the microporous layer. 16. A gas diffusion layer according to any one of the preceding claims, wherein the polymeric binder of the microporous layer comprises, and is preferably, a fluorine-containing polymer such as polytetrafluoroethylene. 17. Gas diffusion layer according to one of the preceding claims, wherein the microporous layer B) is obtainable by coating the sheet-like electrically conductive material A) with a preferably aqueous composition containing the conductive carbon particles and the polymeric binder, and subsequently drying and/or sintering the coated material A), wherein the sintering is carried out by thermal treatment at a temperature of 250 °C to 500 °C, preferably 300 °C to 450 °C, and in particular 350 °C to 450 °C. 18. A method for producing a gas diffusion layer for a fuel cell according to one of claims 8 to 17, in which i) the fiber material is coated and/or impregnated with a preferably aqueous composition containing the at least one polymer and optionally further components, and the coated and/or impregnated fiber material is subsequently dried and/or sintered, wherein the sintering is carried out by thermal treatment at a temperature of 250°C to 500°C, preferably 300°C to 450°C, and in particular 350°C to 450°C, so that the sheet-like electrically conductive material A) is obtained, and subsequently ii) the sheet-like electrically conductive material A) is coated with a preferably aqueous composition containing the conductive carbon particles and the polymeric binder, and the coated material A) is subsequently dried and/or sintered, wherein the sintering is carried out by thermal treatment at a temperature of 250°C to 500°C °C, preferably 300 °C to 450 °C, and in particular 350 °C to 450 °C, so that the microporous layer is formed. 19. A fuel cell comprising at least one gas diffusion layer according to one of claims 1 to 17. 20. A fuel cell according to claim 19 comprising a polymer electrolyte membrane having a catalyst layer applied thereto, wherein the catalyst layer is in contact with the surface of the microporous layer of the gas diffusion layer. 21. Use of a gas diffusion layer according to one of claims 1 to 17 in a fuel cell, wherein the fuel cell is preferably a polymer electrolyte membrane fuel cell. 22. Use of a gas diffusion layer according to one of claims 1 to 17 in a polymer electrolyte membrane fuel cell for reducing the contact resistance at the interface between the microporous layer of the Gas diffusion layer and the catalyst layer applied to the polymer electrolyte membrane, especially in the area of the channels of the bipolar plate of the fuel cell, such as in particular in the area of the center of the channels. 23. Use of conductive carbon particles comprising carbon black and graphite in the microporous layer of a gas diffusion layer, preferably as defined in one of claims 1 to 17, for reducing the contact resistance at the interface between the microporous layer of the gas diffusion layer and the catalyst layer applied to the polymer electrolyte membrane of a polymer electrolyte membrane fuel cell, especially in the region of the channels of the bipolar plate of the fuel cell, such as in particular in the region of the center of the channels. 24. Use of a polymer selected from polyaryletherketones, polyphenylene sulfides, polysulfones, polyethersulfones, partially aromatic (co)polyamides, polyimides, polyamideimides, polyetherimides and mixtures thereof, preferably a polyaryletherketone such as polyetheretherketone, in and/or on the fiber material of a gas diffusion layer, which is preferably as defined in one of claims 1 to 17, for reducing the contact resistance at the interface between the gas diffusion layer and the catalyst layer applied to the polymer electrolyte membrane of a polymer electrolyte membrane fuel cell, especially in the region of the channels of the bipolar plate of the fuel cell, such as in particular in the region of the center of the channels.