EP2392013A1 - Method for preparing an electrically conductive article - Google Patents

Abstract

The invention relates to a method for manufacturing an electrically conductive article including the following steps: - a dry process step of mixing a powder including at least one thermosetting resin including at least two epoxide groups, a powder compound for hardening said resin, and an electrically conductive filler powder; - a step thermocompressing the mixture of powders obtained during the preceding step in a mould with a shape adapted to the article and at a temperature that is effective for obtaining a recirculation of the resin, at the end of which the electrically conductive article is obtained.

Description

 PROCESS FOR PREPARING A CONDUCTIVE ARTICLE OF

ELECTRICITY

DESCRIPTION

TECHNICAL AREA

The present invention relates to a method for preparing a conductive article of electricity.

This method is particularly suitable for the preparation of current collectors, in particular bipolar plates used in proton exchange membrane fuel cells (known as PEMFC cells, meaning PEMFC).

"Proton Exchange Membrane Fuel CeIl"), proton exchange membrane electrolysers and supercapacitors.

One of the main fields of the invention can therefore be considered as that of fuel cells.

STATE OF THE PRIOR ART A fuel cell comprises a stack of elementary cells in which an electrochemical reaction takes place between two reactants which are introduced continuously. Fuel, such as hydrogen, for batteries operating with H2 / O2 mixtures, or methanol for batteries operating with methanol / oxygen mixtures, is brought into contact with the anode, whereas the oxidant, generally oxygen is brought into contact with the cathode. The anode and the cathode are separated by an electrolyte, proton exchange membrane type. The electrochemical reaction, the energy of which is converted into electrical energy, splits into two half-reactions: fuel oxidation, taking place at the anode / electrolyte interface producing, in the case of H ₂ cells, protons H + , which will cross the electrolyte in the direction of the cathode, and electrons, which join the external circuit, in order to contribute to the production of electrical energy;

a reduction of the oxidant, taking place at the electrolyte / cathode interface, with production of water, in the case of batteries operating with H ₂ .

The electrochemical reaction takes place, strictly speaking, at an electrode-membrane-electrode assembly.

To ensure the operation of electrical devices, it is necessary to obtain an electrical power much greater than the power delivered by a single electrode-membrane-electrode assembly. In this context, the electrode-membrane-electrode assemblies are arranged, most often in the form of a stack, the electrical continuity between the different assemblies being ensured by means of conductive plates, called bipolar plates.

In addition to the current collector function, the bipolar plates must also perform the following functions: the distribution of the reagents and the evacuation of the products at the level of the anode and the cathode, the reactants being hydrogen and oxygen and the products the water for the batteries operating with H ₂ / O ₂ ;

the evacuation of the heat produced during the electrochemical reaction; the mechanical structuring of the stack of elementary cells constituting the stack.

Thus, the constituent materials of the bipolar plates must meet the following criteria: sufficient electrical conductivity in order to efficiently collect the electric current produced by the elementary cells; good thermal conductivity to evacuate the heat produced during the electrochemical reaction to the elementary cells; good mechanical properties so as to be able to withstand the stresses associated with the assembly of elementary cells constituting the cell and also to withstand handling during assembly of the fuel cell; a thermal stability to guarantee the integrity of the assembly in the ranges of temperature of use of the battery; chemical stability against fluids in the core of the cell (eg, water, acid) so that the performance of the material can be maintained and decomposition can be avoided of it and thus a pollution of the anode and the cathode with which it is in contact; a reagent (eg, hydrogen and oxygen) impermeability greater than that of the proton exchange membrane; surface hydrophobicity to facilitate evacuation of water formed during the electrochemical reaction; an ability to be shaped, so as to allow the formation of distribution channels on the surface of the plates and without requiring, preferably, machining phase for the formation of said channels.

The bipolar plates currently used can be subdivided into three categories:

- bipolar plates in graphite;

- the metal bipolar plates; and bipolar plates of organic composite material.

As regards the graphite plates, it is difficult to envisage using them on an industrial scale because of their very high cost, essentially due to the machining phase necessary for the production of the distribution channels.

Regarding the metallic bipolar plates, they present a set of properties

(mechanical strength, tightness, electrical conductivity, fitness), which makes them prime candidates for the design and construction of inexpensive bipolar plates. However, their density greater than that of graphite requires them to be used in the form of thin sheets shaped by stamping. Under such conditions, the channel designs are limited (the ultimate elongation limit of the material conditioning the geometry of the channels). Moreover, the stamping technique allows the formation of channels on only one face of a plate, which requires the combination of two half-plates stamped so as to obtain a resultant plate having channels on both sides .

For bipolar plates made from metal sheets shaped by stamping, the arrival of the fluids and the evacuation of the formed products are carried out in areas of the bipolar plate locally flat, which requires the use of a frame having a shape adapted and able to ensure the peripheral sealing, as described in WO 2007/03743. This technique has the disadvantage of requiring for the same bipolar plate a complementary piece intended to ensure the junction between two associated plates to form the bipolar plate and ensure the supply and evacuation of fluid and product. This results in final bipolar plates less compact than expected.

Finally, one of the problems inherent in the use of metal to form bipolar plates lies in the corrosion phenomena induced by the contact of the plates with aqueous media. To remedy this, some authors (such as the J. Power publication. Sources, 131 (2004), p.162-168) propose protective coatings to be carried on said plates in order to limit the phenomenon of corrosion. However, this generates a manufacturing process of these more complex to implement.

As regards the plates made of organic composite materials, these consist of plates comprising an organic polymer matrix in which electrically conductive particles are dispersed. The particles give the bipolar plates the electrical conductivity necessary for the collection of the current and the polymer matrix the mechanical strength necessary for the assembly of the different constituent elements of the cell in which the bipolar plates are arranged.

The conductive particles may be metallic, which has the advantage of good electrical conductivity. However, they have the disadvantage of having a high density and being sensitive to the chemical environment.

The conductive particles may also be carbon-based products, in the form of powders, such as powders of carbon black, graphite or carbon fibers.

Conventionally, the plates are made by incorporating the conductive particles into a liquid resin followed by curing of the resin as described in US 6,248,467. However, the use of a raw material in liquid form, in this case a liquid resin, has the following disadvantages:

instability of the system before molding due to the difficulty of controlling the crosslinking reaction of the liquid resin; a heterogeneity of the system before molding due to the combination of a liquid resin with solid particles of electrically conductive filler;

a phenomenon of exudation of the resin during the shaping of the article, which generates an article comprising a non-homogeneous distribution of the electrically conductive particles and an insulating surface of the article due to the concentration resin exuded at this level.

To overcome these difficulties, some authors have proposed to work from solid reagents (solid resin, conductive particles). However, the mixtures described in these documents must be in the melt phase before shaping, for example by hot calendering. This melting phase melting however generates a very high energy cost, which is incompatible with large-scale industrialization of the manufacture of bipolar plates. Moreover, the plates obtained do not meet the requirements of the specifications of the application including those concerning the planar electrical conductivity. Thus, there is a real need for electrically conductive articles, in particular bipolar plates, and methods of manufacturing them, which can be carried out simply, in a limited number of steps and at the least cost, by limiting the energy balance of these steps, compatible with an industrial implementation and which preferably eliminate the long and expensive machining operations, these methods must also make it possible to obtain conductive articles having effective performance in terms of electrical conductivity, mechanical strength and evacuation of water produced over time in the case of bipolar plates.

STATEMENT OF THE INVENTION

Thus, the invention relates to a method of manufacturing an electrically conductive article comprising the following steps:

a step of dry blending of a powder comprising at least one thermosetting resin comprising at least two epoxide groups, a hardener compound powder of said resin and a powder of an electrically conductive filler;

a step of thermocompression of the mixture of powders obtained in the preceding step in a mold of shape adapted to the article and at a temperature effective to obtain the crosslinking of the resin, at the end of which is obtained; conductive article of electricity. The implementation of this process results in the following advantages: an energy-saving process, in particular because the mixing step is carried out by the dry method not involving heating, as is the case with the mixing steps intended to obtain products in the form of pastes; a method that consumes less energy because the resin is a thermosetting resin, which implies that the article at the end of the process does not need to be cooled when it needs to be demolded, which is not the case generally when the article is made from thermoplastic resin;

a quick and simple process of implementation, because the article can be, after the powder mixing step, carried out in a single thermocompression step, without the need for prior transformation and without the need for a subsequent step of machining;

- A method initially implementing a stable powder mixture, because of working by the dry route, which can allow storage of the powder mixture from the first step before the shaping step by thermocompression; a method for obtaining an electrically conductive article having properties effective in terms of electrical conduction and mechanical properties; a method that can be implemented industrially on a large scale, because of the few steps necessary to obtain said articles; - An inexpensive process to implement, because of the low cost of the epoxy resins used and the ease of shaping the powder mixture without going through subsequent machining steps to form the desired article.

Thus, the method comprises a first step of dry mixing a powder comprising at least one thermosetting resin comprising at least two epoxide groups, a hardener compound powder of said resin and a powder of an electrically conductive filler.

It is specified that by dry means, conventionally, a mixing step not involving liquid compounds, such as solvents, and melted phases, the various constituents of the mixture remaining in the form of solid powders and not requiring heater. It follows a simple and energy-saving step and a mixture of powders that have not undergone a crosslinking step of the thermosetting resin before transformation thereof during the thermocompression step.

It is specified that, by thermocompression, is meant, conventionally, a step of shaping the powder mixture into the desired article comprising pressurizing said mixture to a temperature effective to ensure the cohesion of the powder mixture so as to form the article, this step can be carried out, for example, in an injection press or in a simple compression press and does not require a prefusion step of the powder mixture before shaping. The powder comprising at least one thermosetting resin comprising at least two epoxide groups may have an average particle size ranging from 10 to 500 μm, as can the hardening compound powder of said resin and the conductive filler powder of the resin. 'electricity.

Advantageously, the powder comprising at least one thermosetting resin comprising at least two epoxide groups, the hardener compound powder of said resin and the electrically conductive filler powder has an average particle size ranging from 75 to 150 μm.

The thermosetting resin is an epoxy resin comprising at least two epoxide groups, namely groups having the following pattern:

Particularly advantageous resins that can be used for the process of the invention have the following formulas:

nor representing the number of repetitions of the motive taken in square brackets;

n ₂ representing the repetition number of the motif taken in square brackets, and mixtures thereof.

It is particularly advantageous to work with epoxy resins, because these resins, after crosslinking due to the action of the hardening compound, do not generate a third species, and thus generate an article having a low shrinkage after crosslinking of the resin and excellent dimensional stability.

The use of other types of resins often gives rise, during the crosslinking, to the formation of third species such as water (which is particularly the case for phenolic resins) and therefore a residual porosity in the materials after crosslinking, which would be detrimental to electrically conductive articles to be used as bipolar plates, because the porosity could lead to an increase in gas permeability and also a decrease in mechanical properties.

Moreover, the use of resins as mentioned above gives the resulting material after crosslinking excellent properties in terms of rigidity, thermal and chemical stability, adhesion with the fillers, which makes these resins Apart from their intrinsic properties and the way in which they are implemented, they are much more attractive than the thermoplastic polymers conventionally encountered in the prior art. The hardener compound is conventionally, according to the invention, a compound capable of generating the crosslinking reaction (it could therefore also be described as a crosslinking agent) of the abovementioned resin, for example by opening the epoxide rings, these hardening compounds thus comprising at least one function capable of reacting with the epoxide rings of said resins. Suitable hardening compounds may be compounds comprising at least one amine function. These may include aliphatic amine compounds, amidoamines, polyamides, polyetheramines, cycloaliphatic amines, anhydrides, aromatic amines, imidazole compounds.

By way of example, the crosslinking reaction of an epoxy resin with an amino compound can be schematized as follows:

Other compounds can generate the crosslinking of epoxy resins. In particular, the imidazole compounds used as curative compounds are particularly preferred. Indeed, they generate a homopolymerization of an epoxy resin, which means that the patterns resulting from the polymerization of the resin are directly connected to each other, without there being any cross-linking nodes from the hardener compound molecules. . It follows, after polymerization, a denser material, more rigid and having a higher glass transition temperature than those obtained with a hardener including during the polymerization between two patterns from the resin.

By way of example of a hardening compound, mention may be made of dicyandiamide of the following formula:

By way of example of a hardening compound, mention may be made of 2-phenylimidazole of the following formula:

The electrically conductive filler powder may be a powder of any electrically conductive material. Thus, it may be metal powders, metal oxide powders, powders of carbonaceous materials.

Preferably, the electrically conductive filler powder used in the context of the process of the invention is a powder of carbon material which is particularly advantageous because of its chemical inertness and its low density.

As carbonaceous material powder, mention may in particular be made of graphite powders, carbon black powders, in particular synthetic graphite powders, which have the advantage of not including impurities harmful to the conductivity of the material, unlike the grades natural resources from mining. As carbon black powders, mention may be made of commercial Ensaco 350G powders with a specific surface area of 770 m ² / g, Ensaco 250G with a specific surface area of 65 m ² / g. As synthetic graphite powders, mention may be made of Timrex KS 150 commercial powders with a specific surface area of 3 m ² / g, Timrex KS 75 with a specific surface area of 6.5 m ² / g.

As other types of electrically conductive fillers, it is possible to envisage, in addition to those mentioned above, lamellar graphite, carbon fibers, carbon nanotubes and, in particular, lamellar graphite.

When lamellar graphite is added optionally in addition to a graphite powder, the lamellar graphite goes, in addition to conferring an electronic conductivity to the material, also allow, after the thermocompression step, when the graphite lamellae are disposed of. perpendicular to the flow of gas arriving on the formed article, to increase the tortuosity of the gas, and consequently to improve the impermeability of the material to these gases.

In addition, the mixing step may incorporate other additives, such as catalysts, plasticizers, flexibilizers, nucleating agents, hyperdispersing agents, anti-caking agents, reactive diluents.

It is understood that these additives will also be in the form of solid powders having, advantageously, an average particle size ranging from 10 to 500 μm.

In particular, it may be advisable to add a catalyst powder to the above-mentioned powder mixture so that, during thermocompression, the rate of crosslinking of the resin is improved. Suitable catalysts may be compounds capable of readily giving off hydrogen, such as phenolic, alcoholic, acidic, amino compounds. More specifically, they may be aliphatic amines, imidazole compounds or substituted ureas, such as a compound of the following formula:

In order to obtain powders having the average particle sizes mentioned above, it may be necessary to carry out a grinding and sieving step, in order to keep only the particles having the desired size.

For example, the various powders constituting the mixture have an average particle size identical.

From the point of view of the proportions of powders in the aforementioned mixture, the charge powder electrically conductive may represent from 50 to 95% by weight relative to the total mass of the mixture, while the other powders (resin powder, hardener compound powder and optionally other additive powders) may represent from 5 to 50% by weight relative to the total mass of the mixture. Advantageously, the electrically conductive filler powder represents from 80 to 92% by weight relative to the total mass of the mixture. Such a proportion of electrically conductive charging powder can make it possible, after transformation, to obtain an electrically conductive article having a very high conductivity (greater than 150 S / cm). The various powders mentioned above are mixed dry, so as to avoid any preliminary melting and a beginning of crosslinking of the resin, the mixture can be produced, for example, at room temperature. The mixing can be carried out by moving the particles relative to each other. This displacement can be generated by a stirring system (for example, the shares or ribbons of a convective mixer, the blades of a high-shear mixer), by a flow of air (for example, with a mixer with impaction) or by the rotation of a tank comprising said mixture. In this case, the rotation of the tank must cause mixing until the dynamic angle of the natural slope is exceeded, so that the particles then move in relation to each other. Preferably, the mixing step is carried out in a convective or rotating-tank mixer, which does not bring too much energy to the particles, which could lead to an increase in the temperature and possibly a crosslinking of the resin during the mixing. mixing step, which is prohibited in the method of the invention.

According to a particular embodiment of the invention, the mixing step can be divided into two sub-steps:

the mixture of the reagent system comprising the resin powder, the hardening compound powder and optionally the catalyst powder;

- The mixture of the reactive system with the (or) powder (s) of conductive charge electricity.

Advantageously, the powder mixtures may be the following:

a mixture of powders comprising:

an epoxy resin powder of following formula:

nor representing the repetition number of the pattern taken in square brackets, for example ni is greater than 1, 5; a dicyandiamide hardener compound powder of the following formula:

a catalyst powder of the following formula

and; a synthetic graphite powder and possibly lamellar graphite;

a mixture of powders comprising:

an epoxy resin powder of following formula:

n ₂ representing the number of repetitions of the pattern taken in square brackets;

an epoxy resin powder of following formula: nor representing the repetition number of the pattern taken in square brackets, for example ni is greater than 1.5; a dicyandiamide hardener compound powder of the following formula:

H ₉ N- -C:: N: NH

NH ₂

a catalyst powder of the following formula

and;

a synthetic graphite powder; a mixture of powders comprising:

an epoxy resin powder of following formula nor representing the number of repetitions of the motive taken in square brackets; an imidazole hardening compound powder of the following formula:

and; - a synthetic graphite powder.

These specific powder mixtures are obtained from commercial products that comply with the European RoHS directive (abbreviation corresponding to the "Restriction of the use of certain Hazardous Substances in electrical and electronic equipment") in force. The transformation of these mixtures by crosslinking the resin does not generate any toxic or harmful species.

The powder mixtures resulting from the mixing step of the process of the invention, and in particular the specific mixtures mentioned above, are particularly stable and can be stored at ambient temperature, ready for use, for several months, without degradation of the properties of the article formed by transformation of said powder mixtures. Once the mixing step has been completed, the resulting powder mixture is subjected to a step of thermocompression of the mixture of powders obtained in the preceding step in a mold of suitable shape for obtaining the desired article and at a temperature of effective to obtain the crosslinking of the resin, at the end of which is obtained the conductive article of electricity.

During this step, the mixture of powders is converted by densification of the resulting mixture generated by the crosslinking of the resin. It is advantageously exerted a pressure less than or equal to 1 t / cm ² , ideally between 250 and 750 kg / cm ² . The exercise of this pressure on the mixture of powders makes it possible, in addition to compacting the mixture, to reduce the distance between the electrically conductive filler particles.

The pressure is preferably maintained for a time longer than the curing time of the epoxy resin. In parallel and / or prior to the application of the pressure to the mixture, the mixture is heated to a temperature necessary for the crosslinking of the resin.

The temperature and duration of the heating cycle are a function of the epoxy resin / hardener couple and the catalyst, when it is present. The heating cycle can vary from 2 to 30 minutes for a temperature ranging from 50 to 250 ^° C.

As mentioned above, the step is performed in a mold whose shape is adapted to obtain the desired article, which means, in other words, that the mold must have a design and shape such as after the thermocompression step, the article resulting from this step has the desired final shape, without having to machine said article to give it its final shape.

According to a particular embodiment, the thermocompression step can be implemented as follows:

- The introduction of the mixture of powders in a hot mold (temperature above the polymerization temperature of the resin), the mold having dimensions corresponding to the dimensions of the desired article, the mold including fingerprints so as to generate the design of distribution channels on the surface of articles, if they are intended to constitute bipolar plates;

the closure of the mold followed by placing under load. The pressure can be applied gradually, in order to evacuate the air included in the material in powder form. Under the effect of heat, the compounds of the mixture begin to melt. The polymerization reaction is initiated when the compounds have reached the activation temperature of the hardener. The pressure is maintained until the end of the polymerization reaction of the resin; once the polymerization is complete, demolding the resulting article without prior cooling.

The mold can be covered with a layer of a release agent or a specific surface treatment, in order to optimize the extraction of the molded articles.

It will be specified that the use of a thermosetting resin makes it possible to reduce the energy cost of manufacturing the aforementioned articles. Indeed, if a thermoplastic resin was used, it would be necessary, after the thermocompression step, to cool the mold so that the article becomes solid and can be extracted from the mold.

The process of the invention is particularly suitable for the preparation of bipolar plates for fuel cells, especially bipolar plates advantageously having the following characteristics: an electrical conductivity of at least 150 S / cm; a thermal conductivity of at least 15 W / mK; a flexural modulus of at least 10 GPa;

- a breaking stress of at least 40 MPa;

a contact angle of at least 100 °;

a permeability to hydrogen of less than 2 * 10 ^-6 cm ³ / s. cm ² (measured at 80 ^° C. under 3 atm). The invention will now be described with reference to the following examples given for illustrative and not limiting.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a graph illustrating planar electrical conductivity σ (in S / cm) as a function of the average size of electrically conductive filler particles t (in μm) for plates manufactured in accordance with FIG. 1. FIG. 2 is a diagram illustrating the directions in which the electrical conductivity and the thermal conductivity, respectively planar (arrow 1) and transverse (arrow 2), are measured, the arrow 3 indicating the charging direction.

FIG. 3 is a graph illustrating the planar electrical conductivity σ (in S / cm) as a function of the charge ratio% (in mass%) of electrically conductive particles for plates manufactured according to example 2.

FIG. 4 is a graph illustrating the planar electrical conductivity σ (in S / cm) as a function of the operating pressure P (in t / cm ² ) for plates manufactured according to example 3. FIG. a graph illustrating the evolution of the contact angle θ (in °) as a function of the pressure implemented P (in t / cm ² ) for plates manufactured according to example 3.

FIG. 6 is a graph illustrating the polarization curve, that is to say the T cell voltage (in V) as a function of the current density D (in A / cm ² ) for an electrode-membrane assembly made according to Example 4.

DETAILED PRESENTATION OF PARTICULAR EMBODIMENTS

Examples 1 to 3 above are intended to highlight the reasoned choice of the characteristics of the raw materials and the operating conditions within the scope of the invention to produce bipolar plates.

In these examples, different magnitudes (electrical conductivity, thermal conductivity, thermal properties, permeability, wettability) are measured, the measurement protocols of which are given below.

a) Electrical conductivity

The conduction of electricity by the plate reflects a possible passage of electrons within the composite material constituting the bipolar plates. This electronic transfer is done via electrically conductive filler particles.

There is an overall resistance of the composite material due to the discontinuity of the material. This overall resistance is due to both the constriction resistance (due to the low contact area between the filler particles) and the tunnel resistance (due to the thickness of the insulating film on the surface of these same particles).

The constriction resistance is a function of the geometry of the electrically conductive particles and the compaction of the composite material (notably the settling and the deformation of the particles on each other).

The tunnel resistance is, in turn, a function of the proportion of resin in the composite material. Electronic conduction is then done by electronic jump between isolated particles, generating a local electric field.

The electrical conductivity σ (in S / cm) is determined by measuring the electrical resistivity p (in Ω.cm) of the material by the Werner method, known as the 4-point method.

An alternating current I is sent into the material to be characterized from two extreme points. The potential difference V is measured between the two internal points. The points are equidistant from a distance d.

The resistivity p (in Ω.cm) corresponds to the following formula:

p = (2πd) * (V / I) the conductivity corresponding to the inverse of the resistivity.

b) Thermal conductivity From an atomic point of view, thermal conductivity is related to two types of behavior: the movement of charge carriers, electrons or holes;

the oscillation of the atoms around their position of equilibrium. The thermal conductivity is therefore linked, on the one hand, to the electrical conductivity (movement of the charge carriers) and, on the other hand, to the very structure of the material (vibration of the atoms). The thermal conductivity Λ (W / mK) is measured directly on the article of the invention (hot compression molded plate).

To do this, the sensor is positioned on the main surface of the article (to determine the planar conductivity) or on its edge (to determine the transverse conductivity).

A resistive sensor is used. The electrical current applied to the sensor generates a known amount of heat transmitted to the article. The dissipation of this amount of heat is analyzed by the sensor (which manifests as a decrease in electric current).

This measurement method is not destructive.

c) Mechanical properties

It is proceeded to the implementation of a bending test protocol (according to the NF EN ISO 178 standard) of the composite materials obtained, this stress being comparable to the forces experienced by the bipolar plates during their assembly in stacking .

The three-point bending tests are carried out according to standard NF EN ISO 178. From an experimental point of view, the force curve F-arrow Y is recorded from which it is possible to calculate the stress at break σ _R and Young modulus in bending Ef. The stresses and deformations are determined by the following equations: σ _R = (3F _R D / 2bh ³ )

E _f = (D ³ / 4bh ³ ) (F / Y) in which:

- F _R represents the breaking force;

Y represents the arrow at break; - D represents the distance between supports;

b represents the width of the test piece;

h represents the thickness of the test piece;

- F / Y represents the slope at the origin of the experimental curve force F-arrow Y. The specimens are taken from plates molded by thermocompression and tested on a dynamometer.

d) Permeability The flow of gas passing through a known thickness of the material is measured over time. The permeation coefficient of the material is deduced from the following formula:

Pe = (Φ * e) / (A * ΔP)

with Φ: gas flow (in mbar.l / s) e: thickness of the sample

A: sample surface ΔP: upstream / downstream pressure difference e) Wettability-Hydrophobicity This property is evaluated by measuring the contact angle of a drop of water on the surface of the plate. This method consists in depositing a drop of liquid on the surface of the substrate. The wettability of the surface is then characterized by the contact angle θ formed by the solid surface and the tangent to the liquid surface at the connection point.

EXAMPLE 1

This example is intended to illustrate the influence of the average particle size of the electrically conductive charge on the electrical conductivity of the material and, consequently, on the thermal conductivity.

Different formulations have been prepared with various particle size charges by dry mixing for identical charge rates. The formulations developed and implemented consist of the same matrix (reagent system). Only the particle size of the electrically conductive filler varies, which is synthetic graphite of the type defined in the table of Example 4 below. The mass content of charge for each formulation is 85% by weight relative to the total mass of the mixture.

The reagent system, for its part, comprises the following ingredients: a DGEBA resin of formula as defined in the example 4 below of functionality 2, epoxide equivalent 475-550 g / eq (DER 671 from Dow)

(100 parts) having particle size characteristics such that 90% by volume of the particles have an average particle size of less than or equal to 480 μm and of the total particle population, 50% by volume of the particles have a lower average particle size or equal to 150 μm; a dicyandiamide hardener of formula as defined in Example 4 below (Air Products Amylide CG 1200) (2.5 percent parts resin) having an average particle size of less than 50 μm;

a substituted urea catalyst of formula as defined in Example 4 below (Amicure UR2T from Air Products) (0.5 percent resin parts) having an average particle size of less than 50 μm.

The aforementioned mixtures are shaped by thermocompression according to an identical process for each of the mixtures: thermocompression at 180 ^° C. under 1 t / cm ² for 30 minutes.

The resistivity of the parts obtained is measured and the values are shown in FIG. 1 representing the planar electrical conductivity σ (in S / cm) as a function of the average size of electrically conductive filler particles t (in μm).

It is specified that the planar electrical conductivity corresponds to the conductivity measured in a material plane perpendicular to the load direction.

These results show that the larger the average particle size of the conductive fillers, the higher the planar electrical conductivity of the material.

However, on the other hand, the larger the average particle size, the poorer the interface between the matrix and the filler particles, resulting in a weakening of the mechanical properties of the composite material.

What's more, the size of the filler particles should not be too large, so as not to affect the design of the article. Thus, taking into account all these requirements, the authors have been able to determine, in a reasoned manner, that, in this case, an electrically conductive composite material comprising graphite particles must have an average particle size. (Namely in this case an average particle diameter due to the overall shape of the particles) of the latter advantageously greater than 10 microns and less than 500 microns, preferably ranging from 75 to 150 microns. It may be envisaged to add to the composite material, in addition to the above-mentioned graphite particles, lamellar filler particles, so as to limit the formation in the material of preferential paths for the gas molecules and thus to favor the average free path ( or tortuosity) of these, the lamellar filler particles may be foamed graphite sheets.

During the process of shaping the material by thermocompression, it was proceeded to a shaping promoting an orientation of expanded graphite sheets parallel to the main plane of the article and perpendicular to the direction of underloading.

From a concrete point of view, it was proceeded to the design of an article according to the invention from a mixture M ₄ as defined in the table of Example 4 below. The synthetic graphite used (Timrex KS150 TIMCAL Ltd) has an average particle size such that 90% by volume of the particles have an average particle size of less than 150 μm and the lamellar graphite is in the form of sheets of several hundred ² μm surface area and a few μm thick.

The thermocompression parameters are as follows: 30 minutes at 180 ^° C. under lt / cm ² .

From the plate obtained, the thermal conductivity and the electrical conductivity in the plane perpendicular to the load direction (referred to in the table below as "planar") and in the plane were measured. parallel to the load direction (referred to in the table below "transverse") (indicated respectively by the arrows 1 and 2 in Figure 2, the direction of loading being indicated by the arrow 3). The results are shown in the table below.

In the case of the incorporation of a lamellar charge in addition to the load of non-lamellar graphite, it is interesting, in this case, that this lamellar charge is arranged in a plane perpendicular to the direction of movement of the gas molecules ( this provision being induced by the implementation of the method), which is also the charging direction. The lamellar filler arranged in the form of parallel planes interferes with the progression through the plate of the reaction gas molecules.

EXAMPLE 2

This example is intended to illustrate the influence of the rate of electrically conductive filler particles on the properties of the constituent material of the bipolar plate produced according to this example.

To do this, it was proceeded to the preparation of different solid composite mixtures by varying the content of electrically conductive particles (values shown in Figure 3). The different mixtures such as those explained in the first part of Example 1 are used again, the load being a synthetic graphite (Timrex KS150 TIMCAL Ltd) with a particle size such that 90% by volume of the particles have an average size of particles less than or equal to 150 μm.

The aforementioned mixtures are then shaped according to a method of thermocompression at 180 ^° C. under 1 t / cm ² for 30 minutes.

The planar electrical conductivity of the parts obtained is then measured. The results are reported in FIG. 3, which represents the planar electrical conductivity σ (in S / cm) as a function of the% charge rate (in% by weight) of the electrically conductive particles.

From this graph, one can, for this case, put forward the following observations: below a certain proportion of electrically conductive particles (more precisely below 80% by mass), the resistivity composite material is high. Indeed, the organic matrix of the material in large amounts forms a thick insulating film on the surface of the electrically conductive particles, thus generating a high tunnel resistance of the material;

the electrical conductivity is optimal for the particle size used here (D ₉₀ = 150 μm, D ₉₀ meaning that 90% of the volume of the particles has an average particle size of less than or equal to 150 μm) and a proportion of conductive particles of between 85 and 92% by weight; beyond 92% by mass, the organic matrix is not in sufficient proportion, which generates a porous and fragile composite material.

EXAMPLE 3

This example is intended to illustrate the influence of the pressure implemented during the thermocompression step on the properties of the constituent material of the bipolar plate made according to this example.

To do this, the same mixture (mixture Mi as explained in the table of Example 4) is shaped by thermocompression at different pressures (0.25, 0.5, 0.75, 1; 25 and 1.5 t / cm ² ) for an unchanged time and temperature (30 minutes at 180 ^° C.).

The planar electrical conductivity σ (in S / cm) is then measured.

The results are reported in FIG. 4, which represents a graph illustrating the planar electrical conductivity σ (in S / cm) as a function of the pressure implemented P (in t / cm ² ). From these results, it is clear for the particular mixture used, the following findings: up to 0.5 t / cm ² , the compactness and thus the planar electrical conductivity are increased; for a certain range of pressures applied (in this case between 0.5 and 0.75 t / cm ² ), the planar electrical conductivity has maximum values. This can be explained by the fact that the particles of electricity are approaching each other thus increasing their contact surface; for too high pressures

(Especially for pressures greater than 0.75 t / cm ² ), the planar electrical conductivity begins to decrease. This can be explained by the fact that the organic matrix exudes, thus helping to form an insulating film on the surface of the composite material.

This phenomenon of exudation is also detectable with the measurement of the contact angle of a drop of water placed on the surface of the plates resulting from the thermocompression.

Thus, it was also measured the contact angle of a drop of water placed in contact with the plate as a function of the pressure applied to the mixture during the thermocompression step.

The results are shown in the graph of FIG. 5, which is a graph illustrating the evolution of the contact angle θ (in °) as a function of the pressure implemented P (in t / cm ² ).

From this graph, it can be observed that the contact angle θ decreases sharply from a pressure of 750 kg / cm ² , which reflects an increase in the hydrophilic nature of the surface of the material due to the exudation of the matrix. organic. Compared to the mixture used, it appears the need to work at pressures that are not too high to overcome the phenomenon of exudation of the organic matrix. The pressure applied for this type of mixture is preferably less than 1 t / cm ² , ideally between 200 and 750 kg / cm ² , so as to sufficiently compact the composite material to ensure optimal properties.

EXAMPLE 4

This example illustrates the preparation of various powder mixtures (respectively Mi to M ₄ ) to be shaped by thermocompression according to the method of the invention.

The characteristics of these mixtures are shown in the table below.

(1) Dow Supplier

(2) Air Products Supplier

(3) Supplier Shikoku Chemicals Corporation

(4) Supplier TIMCAL Graphite & Carbon

(5) Supplier Carbone Lorraine

The DGEBA resin has the following formula:

nor representing the repetition number of the pattern taken in square brackets, nor being greater than 1.5.

The Novolac modified DGEBA resin (EPN) is a mixture of DGEBA resin of formula above and a resin of the following formula:

n ₂ representing the repetition number of the pattern taken in parentheses, the mixture comprising at least 80% by weight of DGEBA resin.

The dicyandiamide hardener has the following formula:

H ₉ N- -C: -M-: NH

NH ₂ The "substituted urea" catalyst has the following formula:

The "2-phenylimidazole" catalyst has the following formula:

For the various mixtures, the solid epoxy resins, supplied in the form of crystals, are premixed to the desired particle size so that 90% by volume of the particles have an average particle size of less than or equal to 300 μm and from the total population of particles, 50% by volume of the particles have an average particle size less than or equal to 150 microns.

The grinding is carried out using a sieve mill provided with a suitable perforated grid. Once the components of the mixture have been weighed, they are dry blended (corresponding to the terminology "dry blend") in a powder mixer, specifically a rotating bowl mixer. In this mixer, an oscillating cradle is set in motion by two axes rotating in opposite directions. The resulting movement combines rotation, translation and inversion. This random movement of the particles guarantees a homogeneous mixture without segregation between components. The mixing step takes place under optimal hygiene conditions, without dust, thus reducing the cleaning operations.

The mixing step can be split into two sub-steps:

the mixture of the reagent system comprising the resin, the hardener and the catalyst;

- The mixture of the reactive system with the (or) load (s). The resulting mixture has the particularity of being very stable and can be preserved in the state for several months without degradation.

From the mixtures mentioned above (mixtures Mi to M ₄ ), bipolar plates are molded by thermocompression.

From a practical point of view, the molding is implemented as follows:

the mixture of powders is introduced into a hot mold (temperature higher than the polymerization temperature of the resin), the mold having dimensions corresponding to the dimensions of the plate; desired, the mold including fingerprints so as to generate the design of the distribution channels on the surface of the plates; the mold is then closed and put under load. The pressure can be applied gradually, in order to evacuate the air included in the material in powder form. Under the effect of heat, the compounds of the mixture begin to melt. The polymerization reaction is initiated when the compounds have reached the activation temperature of the hardener. The pressure is maintained until the end of the polymerization reaction of the resin;

- Once the polymerization is complete, the resulting plate is then demolded, without prior cooling.

The precise operating conditions of the thermocompression forming step of the compositions are summarized in the table below for the various mixtures.

The different bipolar plates obtained with the formulations Mi to M ₄ have been characterized in terms of electrical conductivity σ and thermal Λ (respectively planar and transverse), mechanical resistance (3-point bending method)

(maximum tensile strength Max Contr (in MPa) and flexural modulus (in GPa)), surface tension (method of contact angle) and hydrogen permeability. The thermal stability was also evaluated by measuring the glass transition temperature Tg of the reactive system (ie comprising a resin, a hardener and optionally a catalyst).

These data are summarized in the table below.

From these data, the following conclusions emerge:

the plates obtained with an EPN resin have a greater thermal stability and a better resistance to bending than for the plates obtained with a DGEBA resin; since the EPN resin is more functional (more than two epoxide sites per molecule), the crosslinking density of the material formed is greater than for a difunctional resin (such as DGEBA);

the plates obtained with an imidazole type hardener have excellent rigidity and therefore increased mechanical strength over an extended temperature range; the presence of a lamellar filler makes it possible to improve the impermeability to gases.

The plate made with the mixture Mi was subjected to stack tests carried out using a single-cell test bench. These tests consist in testing an electrode-membrane assembly positioned between two composite plates made by the method of the invention allowing the distribution of gases and the collection of electrons. More specifically, the tested assembly consists of a Nafion NRE212 membrane and two Nafion-impregnated commercial "GDE" electrodes (E-tek HT250EW 0.5 mg Pt / cm ² , 0.7 mg Nafion / cm ² ). .

The results of the stack tests carried out under optimal operating conditions for the assembly (70 ^° C., under 2 bars, gas supply H2 / O2, dry gas) are shown in FIG. 6. This figure represents a graph illustrating the polarization curve, that is to say the cell voltage T (in V) as a function of the current density D ( in A / cm ² ) obtained with an assembly as mentioned above. Bipolar plates do not deteriorate the electrochemical performance of the battery core. The new plates are therefore not militant from the point of view of the electrochemical performance of the PEMFC cell core, as can be seen from the curve of FIG.

Claims

A method of manufacturing an electrically conductive article comprising the steps of: a step of dry blending of a powder comprising at least one thermosetting resin comprising at least two epoxide groups, a hardener compound powder of said resin and a powder of an electrically conductive filler; a step of thermocompression of the mixture of powders obtained in the preceding step in a mold of shape adapted to the article and at a temperature effective to obtain the crosslinking of the resin, at the end of which is obtained; conductive article of electricity. 2. Process according to claim 1, in which the powder comprising at least one thermosetting resin comprising at least two epoxide groups has an average particle size ranging from 10 to 500 μm, as well as for the hardener compound powder of said resin and charge powder electrically conductive. 3. Process according to claim 1 or 2, in which the powder comprising at least one thermosetting resin comprising at least two epoxide groups has an average particle size ranging from 75 to 150 μm, as well as for the powder of hardener compound and electrically conductive filler powder. The method of any one of the preceding claims wherein the thermosetting resin has one of the following formulas: nor representing the number of repetitions of the motive taken in square brackets; n ₂ representing the repetition number of the pattern taken in square brackets or may be a mixture thereof. 5. Method according to any one of the preceding claims, wherein the hardening compound is a compound capable of generating the crosslinking reaction of the resin. 6. Process according to any one of the preceding claims, in which the hardening compound is chosen from aliphatic amino compounds, amidoamines, polyamides, polyetheramines, cycloaliphatic amines, anhydrides, aromatic amines and imidazole compounds. 7. Process according to any one of the preceding claims, in which the hardening compound is the diacyanamide of the following formula: H ₉ N- -C: -M-: NH NH ₂ The process according to any one of the preceding claims, wherein the curative compound is 2-phenylimidazole of the following formula: The method of any of the preceding claims, wherein the electrically conductive filler powder is a powder of carbon material selected from graphite powders, carbon black powders. 10. A method according to any one of the preceding claims, wherein the mixing step further implements lamellar graphite. 11. The method of any one of the preceding claims, wherein the mixing step further comprises a catalyst powder. The process according to claim 11, wherein the catalyst is a substituted urea of the following formula: 13. The process as claimed in any one of the preceding claims, in which the mixture of powders obtained at the end of the mixing step is one of the following mixtures: a mixture of powders comprising: an epoxy resin powder of following formula: nor representing the number of repetitions of the motive taken in square brackets; a dicyandiamide hardener compound powder of the following formula: a catalyst powder of the following formula and; a synthetic graphite powder and possibly lamellar graphite; a mixture of powders comprising: an epoxy resin powder of following formula: n ₂ representing the number of repetitions of the pattern taken in square brackets; an epoxy resin powder of following formula: nor representing the number of repetitions of the motive taken in square brackets; a dicyandiamide hardener compound powder of the following formula: a catalyst powder of the following formula and a synthetic graphite powder; a mixture of powders comprising: an epoxy resin powder of following formula nor representing the number of repetitions of the motive taken in square brackets; an imidazole hardening compound powder of the following formula: and; - a synthetic graphite powder. 14. The method as claimed in any one of the preceding claims, in which the electrically conductive filler powder represents from 50 to 95% by weight relative to the total mass of the mixture, preferably from 80 to 92%. 15. Method according to any one of the preceding claims, wherein the thermocompression step is carried out at a pressure less than or equal to 1 t / cm ² , preferably between 250 and 750 kg / cm ² . The method of any of the preceding claims, wherein the electrically conductive article is a bipolar fuel cell plate.