WO2016124625A1

WO2016124625A1 - Method for the operation of an electrolyser or a pemfc fuel cell, for increasing the service life thereof

Info

Publication number: WO2016124625A1
Application number: PCT/EP2016/052243
Authority: WO
Inventors: Frédéric FOUDA-ONANA; Marion CHANDESRIS; Samira CHELGHOUM; Victor MEDEAU; Dominique THOBY
Original assignee: Commissariat A L'energie Atomique Et Aux Energies Alternatives
Priority date: 2015-02-04
Filing date: 2016-02-03
Publication date: 2016-08-11
Also published as: FR3032206A1; FR3032206B1

Abstract

The invention relates to a novel method for the operation of a PEM electrolyser or a PEMFC fuel cell, for increasing the service life thereof. According to the invention, the management of the service life of a PEM electrolyser or a PEMFC fuel cell consists essentially in reducing the operating temperature and the applied or supplied current to an operating point which is acceptable in terms of degradation of the membrane of the at least one MEA assembly and in terms of performance.

Description

METHOD OF OPERATING ELECTROLYSET OR PEMFC FUEL CELL TO INCREASE LIFETIME

Technical area

The present invention relates to the field of electrolysers and proton exchange membrane fuel cells as electrolyte (acronym PEM for "Protons Exchange Membrane" or "Polymer Electrolyte Membrane").

The present invention is more particularly intended to propose a strategy for operating a fuel cell or a PEM electrolyser in order to increase its lifetime.

Although described with reference to a PEM electrolyser, the invention is applicable to all proton exchange membrane (PEMFC) fuel cells.

Prior art

The operating principle of a PEM electrolyser consists in separating hydrogen and oxygen from the water by supplying electrical energy at relatively low temperatures, typically below 100 ° C.

A membrane-electrode assembly 1 or AME constitutes the main element of a PEM electrolyser. FIG. 1 shows schematically such an assembly 1. It comprises a polymer membrane 3, which constitutes the electrolyte allowing the exchange of protons, inserted between a cathode 2 and an anode 4.

The cathode 2 consists of a layer 20 of porous catalytic material, typically based on platinum (Pt) directly in contact with the membrane 3 and a current collector 21, typically based on carbon, superimposed on the catalytic layer 20.

The anode 4 consists of a layer 40 of porous catalytic material, typically based on iridium oxide (IrO ₂ ) directly in contact with the membrane 3 and a current collector 41, typically based on titanium (Ti) superimposed on the catalytic layer 40.

The assembly 1 is generally obtained by hot pressing of the two electrodes (cathode 2 and anode 4) against the membrane 3.

As shown diagrammatically in FIG. 1, the operation of the assembly 1 AME is as follows: the water is oxidized with oxygen at the anode 4 and the electrons released by the reaction anodic are used for the electrochemical reduction of hydrogen protons at cathode 2.

The realization of AME assemblies requires the use of noble materials both as catalysts, typically Pt and Ir, for the anode oxygen evolution reaction and the cathode hydrogen evolution reaction. This involves a high cost for an AME assembly.

In addition, the cost of the current collector materials, which are generally made of porous titanium at the anode, and that of the electrical or fluidic interconnection devices, usually called bipolar plates, also made of titanium, is added to the production cost of an AME assembly.

These bipolar plates allow the flow of current and the flow of gas in a PEM electrolyser stack of several AMES. In addition to their high cost constitutive material, these bipolar plates are machined in the form of teeth and channels to homogenize the separation of water over the entire surface of the electrodes. This machining on surfaces that can go up to 1000 cm ² strongly impacts the final cost of hydrogen production (H ₂ ) in € / Nm ³ .

Thus, it is estimated that for a PEM electrolyser, approximately half of the realization cost comes from the stack of AME assemblies necessary to obtain the desired power. It is further estimated that in a given stack, about 75% of the cost of implementation is due to that of current collectors and bipolar plates.

In the end, the cost of producing a PEM electrolyser is of the order of 2,000 € / kW compared to less than 1,000 € / kW for an electrolyser of the alkaline type, i.e. operating with a basic liquid electrolyte.

In addition to their relatively low production costs, an alkaline electrolyser has a lifetime approximately twice as long as a PEM electrolyser in an acid medium.

A PEM electrolyser still has a number of advantages, including the following three:

it can operate at much higher current densities than an alkaline electrolyzer. Typically, the current density of an alkali electrolyzer is of the order of 250 mA / cm ² to 1.8V against 2000 mA / cm ² at 1.8V for a PEM electrolyser. For the same amount of electrical energy, the production of hydrogen is 10 times more important in acidic medium, ie for a PEM electrolyser compared to an alkaline electrolyzer.

- Given the presence of a solid membrane as an electrolyte, gas permeation is less important in a PEM electrolyser than for an alkaline electrolyzer for which the electrolyte is liquid. Therefore, the purity of the gases from a PEM electrolyser is much better than that of an alkaline electrolyzer.

the low ionic strength of the polymer membrane, generally composed of a perfluorinated sulphonic acid polymer, such as Nafion®, with a thickness generally of between 125 and 180 μιη, enables a PEM electrolyser to have cooling times; responses of the order of one hundred milliseconds, which is very advantageous for coupling with renewable energy (s), such as wind and solar energy.

However, to improve the competitiveness of a PEM electrolyser compared to an alkaline electrolyzer, it is important to demonstrate its reliability over a long period of time.

To do this, it is necessary to develop innovative operating strategies to significantly increase the life of PEM electrolysers.

To date, there are very few publications on operating strategies to improve the life of PEM electrolysers. Among the reasons that may explain this, we could mention the complexity of degradation phenomena that are often coupled: chemical degradation, electrochemical, corrosion etc .... Moreover, to better understand these phenomena, long and expensive tests are often necessary and even if they are made, the conclusions are difficult to extrapolate to other test conditions.

Currently, only two strategies have been considered.

The first is to monitor the degradation of the membrane by measuring the rate of fluorine emitted at the outlet of the electrolyser PEM: see publication [1]. By making the approximation that the measured fluorine comes for the most part from the degradation of the membrane, most industrial electrolyser manufacturers estimate the life of an electrolyser by defining as a limit point, a critical thickness of the membrane at beyond which the mixture of hydrogen H ₂ in oxygen 0 ₂ is less than a value threshold. This threshold value is chosen with respect to the lower explosive limit (LEL), and it is equal to 4% of the latter. One could just as well choose this threshold value as being equal to 2% of the LEL.

The second strategy envisaged to limit the chemical degradation of the proton exchange membrane consists in using an ion exchange resin at the electrolyzer inlet: see publication [2]. This resin is intended to capture the metal ions that are dissolved in water and that are fixed in the membrane. In the absence of this resin, the membrane loses its proton conductivity in favor of the other ions, and consequently, the performance of a PEM electrolyser is deteriorated because the operating point must be at higher potentials. which causes premature corrosion of titanium materials.

In addition, the chemical degradation of the membrane is the consequence of the so-called Fenton reaction, which uses metal ions as catalysts and generates highly oxidizing radicals, according to the following equation:

Fe ²⁺ (aq) + H ₂ 0 ₂ → Fe ³⁺ (aq) + OH (aq) + ΌΗ.

The degradation comes from rad radicals that attack the polymeric chains of the membrane.

In order to control the presence of the metal ions in the water entering the electrolyzer, the electrical conductivity of the water is measured, which indicates the amount of metal ions present and thus makes it possible to determine the purity of the water. A threshold value may for example be 1 mS / cm. If the conductivity of the water, after passing over the resin, exceeds this value, it means that it would be necessary to replace the resin.

These two strategies each have disadvantages that can be detailed as follows.

With regard to the first strategy based on the measurement of fluorine output electrolyser, the chemical degradation of the membrane is assumed to follow the following mechanism:

permeation of the oxygen produced at the anode to the cathode;

electrochemical reduction of oxygen to hydrogen peroxide at the cathode;

formation of free radicals from hydrogen peroxide by the Fenton reaction. - attack of free radicals resulting from the Fenton reaction on the polymer chains of the membrane.

However, given that the thickness of the membrane thins throughout the life of the electrolyser and this thinning amplifies the phenomena of permeation and therefore degradation, it is not accurate to consider from the measurement of the rate of fluorine released at a given moment, the degradation of the membrane over time is linear. Thus, the thinner the membrane, the more it is permeable and therefore the more it degrades. The phenomenon of degradation does not follow a linear law but rather an exponential law. Therefore, considering that the rate of fluoride emitted at a given instant t would be constant and extrapolable over several thousand hours could lead to overestimated life estimates.

The second strategy is based on the monitoring of the ionic conductivity of the water entering the electrolyzer, which requires maintenance every 500 to 800 hours of operation, which can generate a significant associated cost.

In addition, this strategy has no effect on ions already present in the membrane when the ion exchange resin at the electrolyzer inlet is replaced. If the resin is replaced, it means that the ion content in the water upstream of the PEM electrolyser is important and that certainly the amount of ions in the membrane is also important. The fact of replacing the resin does not change the content of the metal ions in the membrane and thus the phenomena of chemical degradation of the membrane will not be slowed down with the change of resin.

There is therefore a need to improve the operating strategies of a PEM electrolyser in order to increase its lifetime, and in particular by avoiding the inaccurate estimates of degradation of the constituent membrane of an AME assembly, and by reducing maintenance costs.

An object of the invention is to respond at least in part to this need.

Presentation of the invention

To this end, the invention relates, in a first aspect, to a method of operating a proton exchange membrane type electrolyser (PEM) comprising at least one membrane-electrode assembly (AME), the method comprising the steps following:

a / determination of the amount of fluorine leaving the electrolyser; b / when the amount of fluorine determined is substantially equal to a given value corresponding to a given thickness of the membrane of the assembly (1) AME, then cleaning at least the membrane,

Once the cleaning of the assembly AME has been performed, the temperature and the operating current density of the electrolyser are reduced.

The inventors were first of all interested in the parameters that affect the lifetime of a polymer membrane of an assembly AME of a PEM electrolyser. They have observed experimentally that the two major parameters are the operating temperature and the current density. Thus, as a first approach, the conditions that make it possible to maximize the lifetime of an AME assembly would ideally be operation at ambient temperature and at the highest possible current density.

However, in practice, such an ideal operating point with these conditions can not be retained for the following reasons.

The reactions to the electrodes are governed by the Arrhenius law which explains the variation of the speed of a chemical reaction as a function of temperature. Thus, according to this law, the higher the temperature, the more the speed of the electrochemical reaction is improved. In other words, in the context of a PEM electrolysis, it is not really possible to envisage an ambient temperature. To do so would be to greatly reduce the performance of PEM electrolysis.

Moreover, the curve of the voltage to be applied as a function of the current follows an Ohm law and therefore, at high current density, the associated voltage is important and beyond a certain value, especially beyond 2V. it could oxidize materials, such as the constituent titanium of the current collector at the anode and if necessary bipolar plates.

The inventors have therefore come to the conclusion that a compromise is necessary in order to have a sufficiently high operating point to obtain both a large hydrogen production and a control of the degradation of the membrane and preferably the degradation of the materials, such as titanium constituting the current collector at the anode and optionally electrical and fluidic interconnection devices (bipolar plates). Thus, rather than applying a constant nominal temperature and / or current density condition throughout the lifetime of a PEM electrolyser as in the state of the art, the invention progressively adapts these two parameters so that the chemical stress imposed on the membrane is the least aggressive and the production of hydrogen is the largest possible. It goes without saying that it must be ensured that throughout the life of a PEM electrolyser, the production of hydrogen according to the process of the invention is greater than in conventional operation.

In other words, the invention firstly states that when the measured fluorine mass exceeds a critical value, which corresponds to the fact that the thickness of the membrane has reached a critical thickness for a given operating point. It is then necessary to perform a cleaning of the membrane in the case of a PEM electrolyser and a change of operating point.

Indeed, if the given operating point was maintained, then the degradation of the membrane would be accelerated. It is therefore necessary to lower the operating temperature and also the associated current density, to a final state in which the membrane retains a sufficient thickness and the performances of the PEM electrolyser are satisfactory.

In summary, the management of the service life of a PEM electrolyser or a PEMFC fuel cell according to the invention essentially consists of reducing the operating temperature and also the current applied or supplied, up to a point of operation. acceptable in terms of degradation of the membrane of the AME assembly (s) and in terms of performance.

According to an advantageous characteristic, the PEM electrolyser membrane is composed of a perfluorinated sulfonic acid polymer, such as Nafion®, the cleaning of the membrane being carried out by washing the AME assembly with sulfuric acid.

According to an advantageous embodiment, during step c1, the operating temperature is decreased by a predetermined value while the applied current is decreased so that the applied potential remains substantially constant. According to this mode, the operating temperature is advantageously reduced by 10 ° C while the applied current is decreased so that the applied potential is of the order of 1.9V. According to an advantageous application, when the applied current is supplied by a system for producing electricity by renewable energy, step c1 consists solely in reducing the operating temperature of the electrolyser.

The electrolyser may include a titanium based anode current collector and a carbon based cathode current collector.

The electrolyser preferably comprises an iridium oxide-based anode catalytic layer (IrO ₂ ) and a platinum-based (Pt) cathode catalytic layer, preferably Pt-Sn, Pt- Ru or Pt-Mo.

Another aspect of the invention is a method for operating a fuel cell (PEMFC) of the proton exchange membrane type comprising at least one membrane-electrode assembly (AME), the process comprising the following steps:

determining the amount of fluorine exiting the fuel cell; b when the quantity of fluorine determined is substantially equal to a given value corresponding to a given thickness of the membrane of the assembly AME, then decreasing the temperature and the operating current density of the fuel cell.

On the basis of the observation that the chemical degradation of the polymer membranes is substantially identical in operation as a PEM electrolyser and in battery operation, the inventors have also envisaged applying the foaming process according to the invention to a PEMFC cell by varying the conditions nominal throughout the life of the battery, that is to say by gradually lowering the current and temperature.

But, unlike a PEM electrolyser, the operation of a PEMFC cell does not induce the "trapping" of metal ions, such as Fe ²⁺ ions in the membrane, so it is not necessary to clean when changing the operating point.

The membrane of a PEMFC cell used in the invention may advantageously be composed of a perfluorinated sulfonic acid polymer, such as Nafion®.

According to an advantageous embodiment, steps a1 to c1 or the above steps a '/ and b are repeated from the initial state in which the membrane has a thickness initially when performing the assembly AME, to a final state in which the membrane has reached a predetermined final thickness.

According to this mode, the steps a / to cl or steps a 7 and b 7 are preferably repeated each time the thickness of the membrane is decreased by a value substantially equal to or less than 10% from its previous value.

The PEM electrolyser or the PEMFC fuel cell according to the invention comprises a membrane-electrode assembly stack (AME) and a plurality of electrical and fluidic interconnection devices arranged individually between two adjacent AME assemblies. The electrical and fluidic interconnection devices are advantageously based on titanium.

detailed description

Other advantages and characteristics of the invention will emerge more clearly from a reading of the detailed description of exemplary embodiments of the invention, given for illustrative and nonlimiting purposes, with reference to the following figures among which:

FIG. 1 is a schematic view showing the general operating principle of a PEM water electrolyser;

FIG. 2 is a schematic view showing the method of operation of a PEM electrolyser according to the invention;

FIG. 3 illustrates the operating point change step of a PEM electrolyser according to the invention;

FIG. 4 illustrates, in the form of experimental curves, the effect of the operating point on the quantity of fluorine ions F ^" measured at the outlet of a PEM electrolyser;

FIG. 5 illustrates, in the form of experimental curves, the effect of temperature on the performance of a PEM electrolyser;

FIG. 6 illustrates, in the form of modeled curves, the performances of a PEM electrolyser and the effect of the operating temperature;

FIG. 7 illustrates, in the form of modeled curves, the estimation of the lifetime of a PEM electrolyser according to the invention;

FIG. 8 is a schematic view showing an advantageous example of the method of operation of a PEM electrolyser according to the invention; FIG. 9 illustrates, in the form of a bar graph, the evolution of the service life of an electrolyser AME assembly as a function of the operating regime respectively according to the invention and according to the state of the art;

FIG. 10 illustrates, in the form of a bar graph, various parameters of an electrolyser AME assembly as a function of the operating regime respectively according to the invention and according to the state of the art.

Figure 1 relating to the general operating principle of a PEM electrolyser has already been commented in the preamble. It will not be detailed below.

It is specified here throughout the present application, the terms "upstream", "downstream", "input", "output" are to be understood by reference to the flow of fluids in a PEM electrolyser according to the invention.

2 schematically shows the method of operation ^of an electrolyser according to the invention. At the outlet ^of an assembly MEA 1 is arranged a gas-liquid separator 5 adapted to separate water from the hydrogen product. Downstream of the separator 5 is arranged a fluorimeter 6 which makes it possible to measure the fluoride ions F present and thus to determine the mass of fluorine released by the membrane of the AM1.

When the measured fluorine mass exceeds a predetermined critical value, this is a sign that the thickness of the membrane has reached a critical thickness for the operating point in question.

A control-command unit 7 can then trigger a cleaning operation of the membrane with sulfuric acid, as shown in FIG.

Once the cleaning is done, the unit 7 can then proceed to a change of operating point, that is to say a decrease in the operating temperature and the current density applied to ΓΑΜΕ.

To achieve the definition of the ^invention, the inventors have compared the experimental curves of the models ^that they have developed and used to ^estimate the performance of an MEA assembly and its life.

During operation of an MEA assembly PEM electrolyser, the inventors ^firstly found that:

- A PEM electrolyser degrades all the faster than its operating temperature is high, which results in a greater amount of fluorine at 80 ° C than 60 ° C; - there is a critical current that maximizes the degradation of the AME membrane. This critical current is characterized by a high concentration of fluoride ions F ^" in solution, with a density of 0.4 A / cm ² at 80 ° C and 0.3 A / cm ² at 60 ° C.

These findings are reported in the form of experimental curves in Figure 4.

It is clear that the conditions that limit the degradation of the membrane are a low temperature and a high current density.

Figure 5 shows experimentally that the performance of a PEM electrolyser ^are even worse than the funct ioning temperature decreases. As ^an example, a current density equal to lA / ^cm2 must be provided 1, 8V at 40 ° C against 1 65V at 80 ° C.

Knowing that the amount of ^hydrogen produced is directly proportional to the operating current according to the formula ^V H ₂ (Nl / h) "0.45 i (A) _el q _ue i _th potential U ^of an ^electrolytic cell increases proportional to the current from 1 A / cm ² and that this parameter influences both the life span of materials, such as titanium and the cost of hydrogen production (electricity cost in kWh / c €)), then the amount of electrical energy required for hydrogen production can be approximated according to the equation:

Figures 4 and 5 show what is the operating strategy to implement to maximize U · production.

More precisely, FIG. 5 shows that the operating point would be at an elevated temperature because at a given current density, the electrical energy to be supplied is lower at a high temperature than at a low temperature. For example, the voltage of 1.8V at a current density of 1 A 5 cm ^to 80 ° C against 1, 9V to, 5A / cm ² at 40 ° C. However, it appears from FIG. 4 that at 40 ° C. the membrane is less degraded than at 80 ° C., whatever the imposed current density.

The inventors then defined models to estimate:

the performance of an AME assembly as a function of the physical properties of FAME and the test conditions, as illustrated in FIG. 6; - the lifetime of an MEA assembly based on the chemical degradation ^of a membrane by coupling ^the thinning of the membrane with ^the increase of the degradation processes, as illustrated in Figure 7.

It is clear from these models that the lifetime of an MEA assembly is longer at a lower temperature: is typically observed a gain ^of a factor of 4 between 333 K and 353 K. In addition, the lifetime is largely overestimated by considering a linear degradation of the membrane rather than considering it according to the model coupling the thinning of the membrane with the increase of the processes of degradation. This is visible in FIG. 7, in which the dashed estimate corresponds to a linear degradation and the dashed estimate corresponds to the coupled model between membrane thinning and the increase of these degradation processes.

Thus, on the basis of the experimental results and the established models, the inventors propose an exemplary operating strategy as illustrated in FIG. 8.

To determine the criterion triggering the new operating point, the inventors considered that the composition of the Nation® is 82% fluorine. It is therefore possible to go back to the thickness of the lost membrane on the basis of the fluorine found in the water leaving the electrolyser PEM.

When the membrane has lost 10% of its initial thickness, a search for a new operating point is put in place.

It is specified here that the fluorine mass criterion that triggers the nominal point change is based on the finding that after 10% membrane thinning, the gap between a linear model and the coupled aging model starts to diverge, which means that the rate of thinning becomes faster and faster.

Considering a membrane of 180 μπι and a surface of 25 cm with a membrane density of 2 g / cm ³ , when the amount of fluorine would be 74 mg, it would be necessary to change operating point because the membrane would have lost 10% of its initial thickness.

It should be noted that the next criterion on the measurement of fluorine will have to take into account the new thickness of the membrane of 162 πι and consequently the mass of fluorine to be considered is 66 mg. To achieve the change of operating point, one proceeds first ^of all to the regeneration of the membrane Nafion®.

Thus, it is known phenomena of chemical degradation of the membrane that the metal ions, in particular Fe ^{2 '} which would come from steel pipes, are the catalysts of the Fnton reaction.

Before proceeding to a new nominal point it is essential to regenerate the proton exchange membrane by cleaning it with sulfuric acid so as to exchange the metal ions captured in the membrane by the protons of the acidic solution.

As the authors of the publication [3] have shown, this method seems to work well.

As illustrated in FIG. 4, the effect of temperature on the aging of the membrane is explicitly demonstrated. Therefore, even if the decrease in temperature is not favorable to obtaining good performance, it is necessary to consider decreasing the temperature to increase the life, especially as the reduction in membrane accelerates with time.

Therefore, the choice of a new nominal point is guided by the need to reduce the operating temperature and to adjust the current so that the potential is less than 2 V in order to avoid the titanium oxidation that occurs according to the equation:

Ti + 2H ₂ O → TiO, + 4 (Pf + e)

As previously stated, the operating temperature is lowered by 10 ° C from the previous operating point.

Given the new membrane thickness and the new temperature, it is possible to estimate the value of the current to be applied that meets the criterion on the potential (E = 1, 9 V). The new operating point is then imposed with the new current value and the new temperature value until the criterion on the fluorine measurement is faulty.

The process steps are then repeated until the thickness of the membrane is 120 iim approximately: at this thickness the safety criterion of the percentage of H ₂ in O .: is always respected.

With reference to FIGS. 9 and 10, an example of operation of a PEM electromagnet assembly AME according to the invention will now be described. comparison with examples of the operation of a PEM electromagnet assembly AME according to the state of the art, ie at constant operating temperature and current.

In these figures 9 and 10:

AME-1 said in self-operating operation corresponds to an operation according to the invention, with as initial operating conditions, an initial temperature of 80 ° C and an initial current applied such that the potential is equal to 1, 9 V. In accordance with the invention, when the thickness of the membrane has lost 10% of its initial thickness, the temperature is lowered by 10 ° C and the operating current is adjusted so that the potential is equal to 1, 9V. These steps are repeated until the membrane reaches a predetermined final thickness value equal to 1 18 μm.

AME-2, AME-3, AME-4 said stationary foaming each correspond to operation according to the state of the art, with constant operating conditions a constant temperature of 80 ° C, 60 ° C and 40 ° C and a stabilized current so that the potential is equal to 1, 9V. These conditions are kept constant until the membrane reaches the predetermined final thickness value equal to 1 18 πι.

Comparing the stationary operations AME-3, AME-3 and AME-4 with each other, it is found that operating a lower temperature AME assembly makes it possible to envisage a longer service life, as is apparent from FIG. 9.

On the other hand, the associated current is lower and consequently the production rate of 1 · · is less important. This results in a higher manufacturing cost for a PEM electrolyzer because with a low operating point, it is more bulky to produce as much hydrogen.

Figure 10 compares the production of hydrogen produced throughout the service life of AME assemblies. This figure 10 shows that. ^the ilisat ut ion, at elevated temperature operating point leads to a high operating current, but with an amount of hydrogen produced relatively low due account of the short life of ΓΑΜΕ. This is reflected in Figure 9 by the surface corresponding to the example AME-2 according to the state of the art.

Conversely, at a low relative temperature, the flow rate of hydrogen produced is low, but the amount of hydrogen over the lifetime of the product is relatively large. This is reflected in Figure 9 by the surface corresponding to the example AME-4 according to the state of the art.

It is thus found that with a funct ioning said self according to ^the invention, it is possible to achieve both an average rate of ^hydrogen production and a consequent relatively high life.

The inventors have quantified this advantage obtained with the method according to ^the invent ion, using merit indicators. ^The indicator should be high when the operating current and lifetime i'AME are high.

Typically, the inventors have chosen the product of the current by the shelf life as a merit indicator. The latter is shown in FIG. 10, from which it is clear that the temperature and current operating point management strategy according to the invention (AME-1) makes it possible to improve the hydrogen productivity ^. by a factor of 1.5 relative to an AME assembly at 40 ° C (AME-4) and a factor of 4 over the case at 80 ° C (AME-2).

Of course these figures are indicative and are based on a model that allowed inventors to report their experimental results. While quantitative results can not be directly exploited for another électroiyseur PEM including a different type of assembly MEA and / or a stack of MEAs, the solution according to ^the invent ion can be implemented for any électroiyseur PEM , as seen from the v ient ^been described.

Although described with reference to a PEM électroiyseur, ^the inv ention can equally well be applied to any fuel cell PEMFC. Indeed, the chemical degradation of the polymer membranes is substantially identical in operation in PEM electromagnet and battery operation. Therefore, it can be quite consider ^applying the funct ioning method according to ^the invention in a PEM fuel cell by v arier nominal condit ions throughout the duration of u a of the stack, c ^'t that is, gradually lowering the current and the temperature. The only essential difference in carrying out the method according to the invention between a PEM électroiyseur and a PEM fuel cell is that it ^is not necessary to perform a cleaning of the membrane stage for the battery.

Other improvements and variations may be envisaged without departing from the scope of ^the invent ion. An interesting application envisaged is that in which a PEM electrolyzer is coupled with a renewable energy (ENR) power generation system, such as solar, or wind power. In this application, only the operating temperature can be lowered gradually, according to the invention, because the applied current is imposed by the system.

The invention is not limited to the examples which have just been described; it is possible in particular to combine with one another characteristics of the illustrated examples within non-illustrated variants.

References cited

[1]. http://www.hydrogen.energy.gov/pdfs/reviewl3/pd030_hamdan_2013_o.pd pp 7.

[2]. "Hydrogen and Fuel cells Fundamentals, Technologies and Applications ^" "John Wiley & Sons, Deltlef Solten Edition (Wiley-VCH) 2010 page 285.

[3]. Sun et al. J. Power Sources 267 (2014) 515-520 page 517.

Claims

A method of operating a proton exchange membrane type electrolyser (PEM) comprising at least one membrane-electrode assembly (AME), the method comprising the steps of:

a / determination of the amount of fluorine output of the electrolyser; b / when the amount of fluorine determined is substantially equal to a given value corresponding to a given thickness of the membrane of the assembly (1) AME, then cleaning at least the membrane,

2. A method of operating an electrolyzer (PEM) according to claim 1, the membrane being composed of a perfluorinated sulfonic acid polymer, such as Nafion®, the cleaning of the membrane being carried out by washing the assembly (1) AME with sulfuric acid.

3. A method of operating an electrolyzer (PEM) according to claim 1 or 2, wherein in step c / the operating temperature is decreased by a predetermined value while the applied current is decreased so that the applied potential remains substantially constant.

4. Method of operating an electrolyser (PEM) according to the claim

3, the operating temperature is decreased by 10 ° C while the applied current is decreased so that the applied potential is of the order of 1.9V.

5. A method of operating an electrolyser (PEM) according to one of the preceding claims, wherein the applied current is provided by a renewable energy generation system, the step cl consisting solely in the temperature decrease operating the electrolyser.

6. A method of operating an electrolyzer (PEM) according to one of the preceding claims, the electrolyzer comprising a titanium-based anode current collector and a current collector at the carbon-based cathode.

7. A method of operating an electrolyzer (PEM) according to one of the preceding claims, the electrolyzer comprising a catalyst layer at the anode to iridium oxide base (Ir0 ₂ ) and a platinum (Pt) cathode catalytic layer, preferably Pt-Sn, Pt-Ru or Pt-Mo.

A method of operating a fuel cell (PEMFC) of the proton exchange membrane type comprising at least one membrane electrode assembly (AME), the method comprising the steps of:

determining the amount of fluorine exiting the fuel cell; b when the quantity of fluorine determined is substantially equal to a given value corresponding to a given thickness of the membrane of the assembly (1) AME, then decreasing the temperature and the operating current density of the fuel cell.

9. A fuel cell operating method (PEMFC) according to claim 8, the membrane being composed of a perfluorinated sulfonic acid polymer, such as Nafion®.

10. A method of operating an electrolyzer according to one of claims 1 to 7, or a fuel cell (PEMFC) according to one of claims

8 or 9, the steps a / to cl or the steps a '/ and b7 being reiterated from the initial state in which the membrane has an initial thickness during the realization of the assembly (1) AME, until a final state in which the membrane has reached a predetermined final thickness.

11. The operating method according to claim 10, the steps a / to cl or the steps a '/ and b7 being repeated each time the thickness of the membrane is decreased by a value substantially equal to or less than 10% relative to at its previous value.

12. A method of operating an electrolyzer according to one of claims I to 7 and 10 to 11, or a fuel cell (PEMFC) according to one of claims 8 to 11, the electrolyser or the battery to fuel cell comprising a membrane-electrode assembly stack (AME) and a plurality of electrical and fluid interconnection devices arranged individually between two adjacent AME assemblies.

13. Operating method according to claim 12, the electrical and fluid interconnection devices being based on titanium.