CARBON-COATED PART

Description

technical field

[0001] The invention relates to the technical field of coatings for fuel cell or electrolyzer electrodes. Previous art

[0002] The invention relates to electrochemical systems such as fuel cells and electrolyzers, and in particular to proton exchange membrane fuel cells, or "PEMFCs" according to the English acronym for "proton exchange membrane fuel cells".

[0003] The operation of these electrochemical systems requires an acidic or basic environment, oxidizing at the cathode, a temperature ranging from 60 to 160 °C, and the possible presence of halides. This environment promotes corrosion of the system's components, such as the interconnecting plates, also known as electrodes, interconnectors, or bipolar or monopolar plates.

[0004] In particular, bipolar plates are one of the critical components for the durability of these systems: they are made of metal sheets approximately 100µm thick. They must be protected by a coating to maintain sufficient surface electrical conductivity to minimize electrical losses at the interfaces, and to prevent corrosion of the metal sheets in the aggressive environment of the fuel cell.

[0005] The surface conduction of a bipolar plate made of metallic material, even in corrosive environments, is generally achieved by depositing a carbon- or gold-based functional layer on the outermost surface of a substrate. In the case of carbon layers, the prior deposition of a sub-layer on the substrate can improve the adhesion of the functional layer and ensure the mechanical strength of the stack.

[0006] In general, the adhesion between the layer and the substrate, as well as the mechanical strength of the functional layer, which results in an absence of damage by cracking or delamination, are important parameters.

[0007] Indeed, the barrier function of this layer must not degrade over the duration of operation of the electrochemical system, in order to protect the metallic substrate from oxidation, in order to prevent the emission into the cell environment of metallic cations from the substrate, even in small quantities.

[0008] Local and generally temporary variations in the conditions of the electrochemical system place increased stress on the functional layer, within which defects such as gaps, cracks, holes, intercolumnar spaces, can cause rapid degradation of the substrate, notably through galvanic coupling with the functional layer.

[0009] Furthermore, when the fuel cell membrane is made of fluoropolymer, it can release fluoride ions (F-) which, among other things, promote pitting corrosion of stainless steel substrates. This can then lead to rapid and catastrophic failure of the entire fuel cell.

[0010] In order to protect a battery from such failures, while meeting the objectives of mechanical strength and electrical conductivity described above, it is known from the prior art to deposit a functional layer, notably carbon-based.

[0011] Physical vapor deposition (or "PVD" for physical vapor deposition) is characterized by the possibility of obtaining a wide range of materials with properties that can differ widely, including when it is only carbon deposition, for example by playing on the proportions of sp2 hybridized atoms compared to sp3 hybridized atoms, it is possible to obtain very different usage properties between them.

[0012] To improve the characteristics of the deposited layer, additional energy is generally supplied during deposition. It is also possible to introduce varying proportions of hydrogen within the carbon layer.

[0013] Good performance has been achieved by adding other elements to carbon. Thus, document WO2011077755 describes a bipolar fuel cell plate coated with a carbon-based functional layer, doped with nitrogen with a content between 3 and 20 at.%, and with hydrogen with a content greater than 0% and less than or equal to 20 at.%.

[0014] The document WO2007136131 describes a conductive organ with an amorphous carbon layer, doped with silicon at a content of less than 1% at., and containing hydrogen with a maximum content of 30% at.

[0015] The document WO2013114836 describes a layer of amorphous carbon containing hydrogen, nitrogen and oxygen with maximum contents of 30, 20 and 3 atomic percent.

[0016] However, controlling the levels of additional elements makes the process more complex and the treatment more expensive.

[0017] Document JP6512577 describes a coating comprising two distinct layers of amorphous carbon of different types, and applications PCT / FR2022 / 051631 and PCT / FR2022 / 051642 on behalf of the Applicant provide solutions to meet the requirements of the U.S. Department of Energy (DOE). However, the lifespan of some coatings can still be improved, particularly at high potential or high temperature.

[0018] Finally, one difficulty lies in the contradiction between having the thinnest possible layer so that its deposition is rapid, and the lower durability of a thin layer, which will degrade more quickly.

[0019] One aim of the invention is therefore to overcome the disadvantages described above. Description of the invention

[0020] To this end, the invention aims firstly to provide a carbon coating whose intrinsic characteristics are optimal to ensure high performance for electrodes, for example, of fuel cells.

[0021] Another objective of the invention is to provide a carbon coating whose intrinsic characteristics are compatible with reversible systems (“URFC” according to the anglicism “Unitised Regenerative Fuel Cell”), which operate at a higher potential than PEMFCs, or for high-temperature fuel cell components (“HT PEMFC”), which operate between 120 and 200°C.

[0022] The invention also relates to a part adapted for specific fuel cell applications, which is efficient and inexpensive.

[0023] The invention then relates to a deposition process for obtaining such a carbon coating.

[0024] To this end, a component was developed comprising a metallic substrate and a layer of amorphous carbon-based material featuring sp2 hybridized and sp3 hybridized atoms, the layer exhibiting: a first content of sp3 hybridized atoms on the substrate side, and a second content of sp3 hybridized atoms on the side of an external surface of the layer, lower than the first content.

[0025] According to the invention, the average sp3 hybridized bond content within the layer is between 5% and 65%, and preferably between 5% and 45%, and the sp3 hybridized bond content varies continuously within the layer. There is a gradient in content from the substrate to the outer surface. Conversely, the sp2 hybridized atom content shows an inverse gradient within the layer (the sum of the sp2 and sp3 contents equals 100%).

[0026] In this way, the carbon-based material layer exhibits optimal intrinsic characteristics to ensure high electrode performance. The higher content of sp3 hybridized atoms allows the substrate to benefit from the advantages of the sp3 form of carbon, namely its good resistance to oxidation and degradation, while avoiding the disadvantage of the high resistivity of the sp3 form of carbon on the external surface, thanks to a higher content of sp2 bonds.

[0027] The part according to the invention therefore exhibits an excellent lifespan thanks to the resistance of the sp3 form, while having low interfacial contact resistance and good surface conductivity thanks to the sp2 form of carbon.

[0028] Since the evolution of the content of sp3 hybridized atoms within the layer is continuous, neither the distribution of internal stresses nor the Young's modulus within the carbon layer presents any discontinuity or interface that would be preferential sources of degradation, leading for example to spalling or delamination.

[0029] The cohesion of the carbon layer is therefore improved, which contributes to the durability of the part, in particular by increasing its resistance to transient phenomena.

[0030] In the specific use case within a fuel cell, the part is part of a monopolar or bipolar type fuel cell plate. In this case, the part meets the DOE requirements.

[0031] The average sp3 hybridized atom content is measured by Raman spectroscopy at several laser excitation wavelengths (325 nm, 442 nm, 488 nm, 633 nm). The laser power is limited to prevent modification of the material under irradiation. The intensity spectra as a function of the Raman shift are fitted by two Gaussian curves for the G (stretching mode) and D (vibration mode) peaks of the carbon bonds. Plotting the position of the G peak (in Raman shift) as a function of wavelength yields a straight line whose slope represents the dispersion of the G peak (cm⁻¹ / nm). This dispersion is proportional to the sp3 content of the analyzed material. A nomogram can be found, for example, in the article "Structure of diamondlike carbon films deposited by femtosecond and nanosecond pulsed laser ablation" by Sikora et al., Journal of Applied Physics 108, 113516 (2010).

[0032] The local content of sp3 hybridized atoms can be determined from the local density of the material constituting the layer (C). This can be measured by transmission electron microscopy (TEM) on cross-sections using electron energy loss spectroscopy (EELS), and spatially resolved with electron beam sizes limited to a fraction of a nanometer. There is a proportional relationship between the layer density and the content of sp3 hybridized atoms (see, for example, the publication "Diamond-Like Amorphous Carbon" by J. Robertson, in Materials Science and Engineering R 37 (2002) 129-281, figure 32).

[0033] According to another feature independent of the features underlying the invention, or in combination with the preceding features, carbon is implanted into a material receiving the layer, whether it be the substrate material or the material of a sub-layer. This implantation is achieved during the layer deposition and advantageously contributes to the adhesion of the layer to its support. Surprisingly, the Applicant has found that the carbon implantation, during an initial phase of the deposition, provides the part with excellent properties with regard to the aforementioned requirements.

[0034] In other words, it could entirely be considered to provide a part comprising a metallic substrate and a layer of amorphous carbon-based material featuring sp2-hybridized and sp3-hybridized atoms, the layer exhibiting: a first content of sp3 hybridized atoms on the substrate side, and a second content of sp3 hybridized atoms on the side of an external surface of the layer, lower than the first content; in which carbon is implanted in a material receiving the layer, whether it is the substrate material or the material of a sub-layer, the implantation of the carbon being obtained during the deposition of the layer; without worrying about the average content, within the layer, of sp3 / sp2 hybridized bonds.

[0035] In a preferred embodiment, the part comprises a single layer of carbon-based material. The manufacturing process for the part involves only one carbon deposition step. It is therefore simplified, and the cost of the part is reduced. Even more preferably, the part comprises a single layer of material made entirely of carbon.

[0036] To promote the conductivity of the part, the average sp3 hybridized bond content is limited. Therefore, the average sp3 hybridized bond content within the layer is between 10% and 45%. Preferably, the average sp3 hybridized bond content within the layer is between 10% and 30%. In some modes, the material forms a thin film commonly referred to as "aC". Preferably, the sp3 hybridized bond content is homogeneous within the layer, and not localized by means of clusters of crystals with a high sp3 hybridized bond content, forming a carbon layer with a heterogeneous sp3 hybridized bond content.

[0037] To reduce manufacturing costs, the carbon-based material is free of metallic dopants. Furthermore, dopants form carbides within the carbon layer, and these localized carbides can become defect sites and preferential corrosion sites. When deposition occurs with a dopant, metallic droplets can also form, creating preferential corrosion sites. These drawbacks are avoided.

[0038] The best durability performance, particularly at high potential, was achieved with layers of carbon-based material free of nitrogen and hydrogen.

[0039] By "hydrogen- and nitrogen-free" and "dopant-free," we mean that these elements are not intentionally added to the deposited layer and can only be present in insufficient quantities to impart any particular properties to the deposited layer. Preferably, these elements are present only in trace amounts.

[0040] Advantageously, the layer has a thickness greater than or equal to 20 nm, preferably between 20 nm and 250 nm, more preferably between 50 nm and 150 nm, even more preferably between 60 nm and 100 nm.

[0041] To improve the adhesion of the layer to the part, and to protect the substrate from possible oxidation, the part includes an underlayer located between the substrate and the layer of carbon-based material, in contact with the layer.

[0042] Advantageously, the underlayer material is selected from chromium, titanium, zirconium, tantalum, niobium, or their alloys, as well as their nitrides and carbides, or the underlayer material is a ceramic. Preferably, the underlayer is an electrically conductive material that becomes passivated upon oxidation. Titanium is a preferred material.

[0043] In order to have a favorable compromise between the deposition time of the metallic undercoat and the improvement in adhesion that it confers, its thickness is between 5 nm and 200 nm, preferably between 20 nm and 100 nm, and is for example 50 nm.

[0044] In a particular embodiment adapted to the field of fuel cells, the substrate comprises stainless steel, titanium, a titanium alloy, or a nickel-chromium-iron alloy, which is preferably an Inconel®.

[0045] Preferably, the substrate is a plate with a thickness between 10 µm and 1000 µm.

[0046] Preferably, the part is part of a monopolar or bipolar type plate for fuel cell, as it has the technical characteristics ensuring durability and sufficient performance to meet the requirements of the field of fuel cells.

[0047] The invention also relates to a method for vacuum vapor deposition of a layer of an amorphous carbon-based material onto a substrate, notable in that a parameter influencing the sp2 or sp3 hybridized bond ratio of the carbon varies continuously during the deposition of the layer so that it exhibits: a first sp3 hybridized bond content on the substrate side, and a second sp3 hybridized bond content on an external surface of the layer, lower than the first content, and in that an average content within the layer of sp3 hybridized bonds is between 5% and 65%, and preferably between 5% and 45%, and in that the content of sp3 hybridized bonds evolves continuously within the layer.

[0048] By "varies continuously," we mean that the parameter does not exhibit discontinuities during deposition, such as during the deposition of successive distinct layers, each with a different sp3 hybridized bond content. It may, however, exhibit strictly increasing or decreasing variations, or variations that are not strictly increasing or decreasing (oscillations, for example). Brief description of the drawings

[0049] [ Fig.1 ] is a diagram of a cross-section of a part according to the invention. Fig. 2 ] is a graph illustrating the variation in the content of sp3 hybridized carbon atoms within a layer of carbon-based material in such a part. Fig.3 [ ] is a graph illustrating the variation in the content of another part according to the invention. [ Fig. 4 [ ] is a graph illustrating the variation in the content of another part according to the invention. [ Fig. 5 [ ] is a graph illustrating the variation in the content of another part according to the invention. [ Fig. 6 [ ] is a graph illustrating the variation in the content of another part according to the invention. [ Fig. 7 ] is a graph illustrating the corrosion current density during a test. Fig. 8 [ ] is a graph illustrating the corrosion current density during another test. Fig. 9 ] is a graph illustrating the corrosion current density during another test. Detailed description of the invention

[0050] In the field of vacuum deposition, several types of technologies exist, each with its own advantages and disadvantages. In the context of processing parts (P), and in particular monopolar or bipolar plates for fuel cells, the Applicant sought to optimize known deposition processes.

[0051] The Applicant carried out various series of tests and interpretations aimed at obtaining a deposit of a carbon-based material, forming a layer (C) on a substrate (S), and exhibiting good properties including mechanical strength, corrosion resistance, adhesion, and electrical conductivity.

[0052] The deposition is preferably carried out by cathodic arc. In this method, the installation used to implement an embodiment of the process comprises a secondary vacuum chamber equipped with a pumping system, a plasma source, an arc spray source, and a substrate holder on which the substrate(s) (S) to be treated are mounted.

[0053] The pumping system creates a secondary vacuum within the chamber, meaning a pressure on the order of 10⁻⁷ mbar to 5 x 10⁻⁵ mbar. This pumping system, or another independent system, is capable of introducing a gas into the vacuum chamber. The gas is intended to be ionized; argon is preferred.

[0054] The arc spray source is conventional and operates on a continuous power supply. Alternatively, it can be pulsed. Electrical discharges onto a graphite target generate carbon ions. The arc supply can be switched on intermittently (e.g., 60 seconds on and 10 seconds off).

[0055] The substrate holder is polarized, meaning that a negative voltage or potential difference is applied across its terminals to create a flow of carbon ions towards the substrate holder. This acceleration of the ions occurs in the vicinity of the substrates (S), since the electric field resulting from the polarization of the parts (P) extends over a short distance, on the order of a few millimeters.

[0056] In the case of substrate (S) polarization in a plasma, the polarization voltage is applied between the substrates (S) and the plant ground. A potential difference is established between the substrates (S) and the plasma. It is in this potential drop zone, over approximately 0.5 to 2 mm of the substrate (S) surface, that the ions are accelerated.

[0057] Within the installation, various tests are carried out by modifying the parameters of the thickness of the layer (C), the presence or absence of a sublayer (SC) between the substrate (S) and the layer (C), the bias voltage (its value, its continuous or pulsed application mode, and the duty cycle in the case of a pulsed voltage), the ignition current of the source (current value, or duty cycle of the intermittency).

[0058] The parts (P) are coated with a functional layer (C) of carbon-based material, intended to protect the substrate (S) from oxidation to ensure the service life of the part (P), while exhibiting low interfacial contact resistance to ensure the performance of a battery or electrolyzer comprising the part (P).

[0059] The lifespan of the deposited layer (C) is evaluated by subjecting it to an electrochemical corrosion test.

[0060] Under normal fuel cell operating conditions, the cathode potential is less than 0.9V compared to a standard hydrogen electrode (NHE, notated 0.9V / NHE). However, under accidental and transient conditions, the electrode could experience higher potentials, up to 1.6V / NHE.

[0061] This is particularly the case when the battery is started or stopped, when the presence of H2 / air or H2 / O2 fronts on a plate can generate reverse current phenomena: the powered part of the battery discharges into the unpowered part of the battery, generating very high local potentials (electrolysis conditions).

[0062] These transient phenomena can last longer if the fuel cell has been poorly drained after shutdown (presence of residual water), and the ambient temperature is very low: distribution channels can be blocked by ice from residual water, which delays the filling and homogenization of the compartment.

[0063] It is therefore necessary that the interconnecting plates be able to withstand very high potentials for several hours, relative to the battery's lifespan. This is all the more important for heavy-duty batteries.

[0064] The electrochemical tests performed by the Applicant are therefore more stringent than the DOE requirements and are conducted in an acidic solution with a pH of 3, at 80°C, and with 0.1 ppm of fluoride ions. This test medium is defined by the DOE in the United States of America to simulate the operating environment of a PEMFC. In the present tests, the potential is set at +1.6 V / NHE on the working electrode to which the material under test is mounted. The potential used is higher than that recommended by the DOE (0.8 V / NHE) in order to make the test more rigorous and to select only the best deposits capable of withstanding accidental conditions. The addition of air bubbling simulates the cathode compartment of a fuel cell. The test is carried out over a period of 20 hours.

[0065] The corrosion current reflects the rate of degradation of a part (P) comprising a substrate (S) coated with a layer (C) of material. Indeed, the higher the corrosion current, the more the part (P) is oxidizing, meaning that the layer (C) of material is not adequately fulfilling its protective role. In practice, a corrosion current density below 2000 nA / cm² is considered acceptable after 20 hours at a potential of 1.6 V / NHE.

[0066] The surface conductivity of the coating is evaluated by measuring its interfacial contact resistance, or "ICR". A coating with good surface conductivity has a low ICR, for example less than 10 mΩ.cm 2.

[0067] The RCI measurement is performed on a stack consisting of a Copper block - Carbon Foil - Deposit, on substrate (S) - Nickel paint - Copper block, on which a current of 100 mA is applied for an area of ​​1 cm², then the resistance of the assembly is calculated from the measured voltage.

[0068] This stack represents the contact between a coated bipolar plate and the gas diffusion layer (C). A pressure of 138 N / cm² is applied to it by a lever arm system with weights; this pressure is representative of that applied to an electrochemical cell during its assembly.

[0069] The total resistance R obtained is the sum of the equation: From the resistance of the Cu-Cu system, the interfacial contact resistance of Copper-Carbon (R Cu / C), the resistance of the Carbon felt (RC), the interfacial contact resistance between the deposit and the Carbon (RC / deposit), the linear resistance of the deposit (R deposit), the resistance of the 316L steel plate (R 316L), and the interfacial contact resistance of Nickel-Copper (R Ni / Cu). [Math. 1] Rtotal=Roffset+RCu / C+RC / deposit+Rdeposit+RNi / Cu

[0070] The RCI is determined using the following equation. [Math. 2] R C / d é pôt + R d é pôt = RCI = R total − R offset − R Cu / C − R Ni / Cu

[0071] It is relevant to consider corrosion resistance after the previously described severe corrosion test. Indeed, a substrate (S) coated with a metallic underlayer (SC) and then a carbon layer (C) may exhibit good corrosion resistance, but this can be explained in some cases by passivation of the underlayer (SC) material if the carbon layer (C) degrades. However, this passivated material is not sufficiently conductive on its surface, meaning that a bipolar plate functionalized with such a coating protects a fuel cell from accidental degradation, but the fuel cell's performance would be reduced. Therefore, the cumulative consideration of these two parameters is necessary.

[0072] The passivated underlayer ensures that the stainless steel of the substrate (S) will not emit metallic cations into the electrochemical system.

[0073] Another criterion is the visual appearance of the layer (C), which must not have been macroscopically damaged, consumed or delaminated at the end of the test.

[0074] Within the installation, several series of tests were carried out. The parts (P) tested are test specimens whose substrate (S) is made of 316L stainless steel, intended to be coated on both sides, in order to simulate the coating of bipolar plates.

[0075] The substrate (S) is positioned on a fixture, cleaned and blown to remove any contaminants and dust that may be present on its surface. It is then introduced into a vacuum deposition system.

[0076] The pumping system is activated so that the pressure in the enclosure is less than 5x10-6 mbar, and the enclosure is heated to desorb the water from the walls.

[0077] The surfaces of the substrate (S) to be coated are heated and then bombarded with argon ions in order to remove the water adsorbed on these surfaces and strip away a layer (C) of oxide present on the surface.

[0078] The pumping system then regulates the argon flow rate within the chamber to maintain an argon pressure of approximately 2 x 10⁻⁴ mbar during the deposition phase. The argon is used for ionization and is not intended to be contained within the carbon-based layer (C).

[0079] We are not seeking to perform a reactive deposition of the (C) layer: the pumping system does not introduce nitrogen or other additive elements into the enclosure, such as doping elements: the internal volume of the enclosure, and therefore the deposited (C) layer, is free of nitrogen or other additive elements.

[0080] Layer (C) does not contain hydrogen, nitrogen, or any other doping element such as tungsten. In this technical field, "does not contain" means zero, trace, or at least less than 1 at.%. In all cases, the aim is to obtain a sufficiently small amount of these elements so that they cannot impart any properties to the deposited layer (C). None of these elements are intentionally added during deposition.

[0081] However, despite the heating, some adsorbed water always remains within the chamber. In practice, at the vacuum pressures of approximately 10⁻⁵ mbar used, virtually only water vapor remains inside the chamber, provided there are no significant leaks. Therefore, it is not possible to completely eliminate the presence of oxygen in the chamber through the water, and this component can end up in the deposit during its formation. Oxygen can also be present in the carbon-based target, as the graphite target is generally porous.

[0082] Unlike hydrogen, oxygen is not easily removed from the (C) layer during deposition, and its concentration depends on the ion bombardment conditions applied, which are themselves specific to the method used. Therefore, only oxygen can be present in the (C) layer, with the obvious exception of atoms from the inert gas (argon), which may be present in minute quantities.

[0083] An underlayer (SC) of metallic material may optionally be deposited on the substrate (S), for example by sputtering a target corresponding to the desired material (metal or ceramic, possibly within a reactive atmosphere to modify the composition of the underlayer (SC) only).

[0084] Next, a layer of carbon-based material is formed from a graphite target by arc deposition.

[0085] An initial series of tests is carried out by modifying: the value of the bias voltage (whether continuous or pulsed); where applicable, the duty cycle of the bias voltage; the duty cycle of the intermittent ignition of the graphite target (arc ignition).

[0086] All the samples tested have a titanium sublayer (SC) 50nm thick, and a carbon layer (C) 70nm thick.

[0087] Characterizations of the parts (P) were then carried out to discriminate the different deposits obtained.

[0088] Following the tests carried out, the Applicant found that despite the prejudice, for example described in documents WO2022049245A1 and WO202128399A1, that a high content of sp3 hybridized form would be more protective, some of the sp3 bond-rich samples (56% sp3) exhibit a significant corrosion current, as well as a strong increase in RCI after corrosion testing: the sole criterion of average sp3 bond content does not appear to be sufficient to discriminate the parts (P) obtained.

[0089] Furthermore, parts (P) with moderate sp3 content exhibit very different characteristics: a sample containing on average 20% sp3 bonds may be non-compliant, while a sample containing 45% is.

[0090] Unexpectedly, the presence of oxygen within the (C) layer is not a discriminating criterion.

[0091] In a first embodiment, the sp3 bonding rate evolves continuously within layer (C), according to a decreasing gradient from the substrate (S) to the external surface (SE). This characteristic explains why the sample with an average sp3 bonding content of 45% is compliant, while the samples with average contents of 20% and 56% are not.

[0092] In practice: when the deposition begins, the sp3 content is high (therefore on the substrate side (S)); during the deposition, the sp3 content gradually decreases without discontinuity; at the end of the deposition, the sp3 content is lower (therefore on the external surface side (SE) of the part (P)).

[0093] In another embodiment, the gradient is strictly decreasing from the substrate (S) towards the external surface (SE). "Strictly decreasing" means that the sp3 bonding content never increases. However, a plateau in the content is possible. In this embodiment, the first content (sp3%1h) is equal to the maximum sp3 content (sp3%max) within layer (C), and the second content (sp3%2b) is the minimum sp3 content (sp3%min).

[0094] In another embodiment, the sp3 bond content exhibits oscillations, such that it increases over at least a portion of the layer, according to variations in the influential parameter. In this embodiment, the first content (sp3%1h) is not the maximum sp3 content (sp3%max) within layer (C), and the second content (sp3%2b) is not the minimum content (sp3%min).

[0095] In both cases, the sp3 bond content shows an overall downward trend within layer (C). Indeed, a higher sp3 bond content at the external surface would imply excessively high RCI, and an excessively high sp2 bond content at the substrate level would imply poor corrosion resistance.

[0096] The variation in the sp3 hybridization rate modifies the characteristics of the (C) layer. It is known that a carbon layer comprising between 10% and 35% sp3 bonds (the remainder being sp2) is of the family of amorphous carbon layers "aC", while a carbon layer comprising between 60% and 85% sp3 belongs to the family of tetrahedral bonded carbon layers "ta-C", and that within the family of tetrahedral bonded carbon layers "ta-C", a carbon layer comprising 65% sp3 bonds does not exhibit the same characteristics (mechanical, electrical) as a carbon layer comprising 85% sp3.

[0097] The sp3 bond content of a material therefore gives a particular nature to that material, and the variation of the sp3 bond content within the layer (C) makes it possible to obtain several natures of materials within the layer.

[0098] With reference to the figure 1, part (P) present within the carbon layer (C): a first zone (e1) on the side of the substrate (S), which is rich in sp3 bonds, and within which a first composition (C1) of the layer (C) is predominantly present; a second zone (e2) on the side of the external surface (SE), which is less rich in sp3 bonds, and within which a second composition (C2) of the layer (C) is predominantly present; the first zone (e1) and the second zone (e2) being separated by an intermediate zone considered negligible with regard to the invention.

[0099] The first composition corresponds to an sp3 bond content greater than a first threshold (sp3%1b), or to an sp3 bond content between a first threshold (sp3%1b) and the first content (sp3%1h).

[0100] The second composition corresponds to an sp3 bond content below a second threshold (sp3%2h), or to an sp3 bond content between the second content (sp3%2b) and a second threshold (sp3%2h).

[0101] The first composition (C1) is for example defined by a first threshold (sp3%1b) of the sp3 bond content, greater than 80%, or 75%, or 70%, or 65%, or 60%, or 55%, or 50%, or 45%, or 40%, or 35%, or 30%, or 25%, or 20%, or 15%.

[0102] The second composition (C2) is defined by a second threshold (sp3% 2h) of the sp3 bond content, lower than that of the first composition (C1), and for example less than 4%, or 5%, or 10%, or 15%, or 20%, or 25%, or 30%, or 35%, or 40%, or 45%, or 50%, or 55%, or 60%.

[0103] Depending on the values ​​retained for the first composition and for the second composition, the first zone (e1) and / or the second zone (e2) each measure between 1 nm and 30 nm, or between 5 nm and 15 nm, or between 7 nm and 10 nm.

[0104] The first zone (e1), rich in sp3 bonds, exhibits good mechanical characteristics and resistance to oxidation.

[0105] The second zone (e2), less rich in sp3 bonding, has the advantage of conducting electricity better, and therefore implies a decrease in RCI.

[0106] The combination of these two zones within the same layer (C) thus makes it possible to combine the advantages of each type of coating. It is possible to produce layers thinner than those of the prior art, with equivalent performance according to DOE requirements, or even exceeding DOE requirements since the invention provides enhanced durability with high potential.

[0107] A thinner layer, in addition to a shorter deposition time, also provides better resistance to deformation. If layer (C) is deposited on a flat part to be subsequently shaped (creation of bipolar plate channels), the part (P) is less likely to exhibit defects such as cracks or delaminations after shaping due to its thinness and the presence of an sp3-rich layer near the interface (I) between layer (C) and the receiving material.

[0108] Furthermore, the continuous variation in sp3 bond content improves the layer's durability. The absence of discontinuities: allows for a better distribution of internal stresses in the carbon (C) layer, and therefore avoids spalling phenomena; allows for better cohesion of the material within the (C) layer, and therefore avoids the presence of preferential sites for corrosion attacks.

[0109] The variation in sp3 bond content is achieved by modifying an influential parameter during deposition. For example, the influential parameter could be the value of the bias voltage, whether continuous or pulsed: It is possible to begin the deposition of layer (C) with a bias voltage adapted so that the carbon ions being deposited are deposited onto the part (P) with high energy, thus promoting sp3 hybridization. During deposition, the bias voltage is gradually decreased, so that the impact energy of the ions also gradually decreases. This progressively reduces sp3 hybridization, and consequently, their concentration decreases.

[0110] The influencing parameter is chosen according to the deposition process implemented, the criterion being the continuous variation of the amount of energy supplied during deposition. For example, in the case of high-power pulsed magnetron sputtering (HiPIMS), the influencing parameter could be the quantity of ions generated by the plasma source, which depends on the pulse parameters.

[0111] Alternatively, the influential parameter could be the arc current applied to the cathode, or the ignition ratio of a pulsed bias. In the case of a pulsed bias voltage, it is worth recalling that for a given average value, the substrate carrier is supplied with a voltage inversely proportional to the ignition duration. For example: For an average voltage of 100V and an ignition ratio of 50%, the substrate carrier (S) can be at 0V half the time, and at 200V the other half (average of 100V); for an average voltage of 100V and an ignition ratio of 25%, the substrate carrier (S) can be at 0V three-quarters of the time, and at 400V one-quarter of the time (average of 100V).

[0112] Although the average voltage is identical, it is observed that significant differences in instantaneous voltage alter the energy supply to carbon ions during the deposition of the (C) layer.

[0113] By adapting the variation of the influential parameter (linear or non-linear), it is possible to obtain different gradient profiles.

[0114] With reference to figures 2 to 4Three different sp3 content gradient profiles (sp3%) were represented as a function of position within the thickness (e) of layer (C). Depending on the content variation profile, layer (C) will comprise a first zone (e1) and a second zone (e2) of varying thicknesses. The first zone (e1) extends from the interface (I) between layer (C) and the receiving material to a first depth (z1), while the second zone (e2) extends from a second depth (z2) to the external surface (SE).

[0115] There figure 2 illustrates a first example in which the sp3 bond content decreases regularly. Then the first zone (e1) and the second zone (e2) are of equal thickness.

[0116] There figure 3This illustrates a second example in which the sp3 bond content decreases irregularly, but with a first plateau at the beginning of deposition and a second plateau at the end. These plateaus can be obtained by momentarily stopping the variation of the influential parameter. The first zone (e1) and the second zone (e2) are of equal thickness, because the two plateaus are of equivalent lengths, but their thicknesses are greater than in the example of the figure 2 However, the thickness of the intermediate zone is reduced compared to that of the first example.

[0117] There figure 4 illustrates a third example in which the sp3 bond content decreases irregularly, with only an initial plateau at the beginning of the deposition. In this case, the first zone (e1) is thicker than the second zone (e2).

[0118] There figure 5illustrates a fourth example, in which the intermediate zone constitutes the major part of the layer (C), because the first zone (e1) and the second zone (e2) are of small thicknesses, on the order of a few nanometers, for example between 1 nm and 15 nm, or between 5 nm and 10 nm.

[0119] There figure 6This illustrates a fifth example, in which the content evolves in a non-strictly decreasing manner: variations in the influential parameter lead to oscillations in the sp3 bond content. Nevertheless, the second sp3 bond content (sp3%2b) is strictly lower than the first sp3 bond content (sp3%1h), such that layer (C) exhibits a first zone (e1) and a second zone (e2). It is possible to obtain an intermediate zone constituting the major part of layer (C), depending on the downward trend of the sp3 bond content and the amplitude of the oscillations. The first zone (e1) and the second zone (e2) can be of small thicknesses, on the order of a few nanometers, for example, between 1 nm and 20 nm, or between 5 nm and 10 nm.

[0120] The intermediate zone is a transition zone between the first zone (e1) and the second zone (e2). It preferably has a limited average sp3 bond content to maintain low overall resistivity of the layer. For example, the average sp3 bond content within the intermediate layer is less than 40%, and preferably less than 25%. However, the initial sp3 bond content (sp3%1h) is preferably at least 30% to 40% to protect the substrate from oxidation.

[0121] The second sp3 bond content (sp3%2b) is preferably at most 5% to 25%, in order to guarantee the good surface conductivity of the part (P).

[0122] It is therefore possible to favour one composition over the other, by adapting the thickness of the first zone (e1) relative to that of the second zone (e2).

[0123] The first sp3 bond content (sp3%1h) is between 30% and 85%, or between 35% and 80%, or between 40% and 75%, or between 45% and 60%, or between 50% and 65%, or in any range obtained by combining these values.

[0124] The second sp3 bond content (sp3%2b) is chosen to be less than the first content (sp3%1h) and is between 3% and 70%, or between 15% and 65%, or between 20% and 60%, or between 25% and 55%, or between 30% and 50%, or between 35% and 45%, or in any range obtained by combining these values.

[0125] The ratio between the second content (sp3%2b) and the first content (sp3%1h) is between 0.05 and 1 (1 excluded), or between 0.2 and 0.9, or between 0.3 and 0.8, or between 0.4 and 0.7, or between 0.5 and 0.6, or in any range obtained by combining these values.

[0126] The average sp3 bond content of the layer, that is, the average value of the content over the entire thickness of the layer, is between 5% and 65%, or between 10% and 60%, or between 15% and 55%, or between 20% and 50%, or between 25% and 40%, or between 30% and 35%, and is, for example, 9%, or 15%, or 19%, or 22%, or 45%, or 56%, or is within any range obtained by combining these values.

[0127] Depending on the choice of the first concentration (sp3%1h) and the second concentration (sp3%2b), it is possible to obtain: a layer (C) of amorphous carbon of type aC, with a first sp3 content of less than about 40%; a layer (C) of amorphous carbon of type ta-C, with a second sp3 content of more than about 60%; a layer (C) of carbon having: * a first zone (e1) rich in sp3 bonds, with a first content (sp3%1h) of more than about 40%; and * a second zone (e2) rich in sp2 bonds with a second content (sp3%2b) of less than about 40%.

[0128] In a particular embodiment, taken in isolation from the features at the basis of the invention, or in combination, the energy of the ions is sufficiently high so that at the interface (I) between the layer (C) and the supporting material receiving it (whether it is the substrate (S) or the sublayer (SC)), carbon is implanted within this material.

[0129] The implantation occurs over a thickness of approximately 1 nm to 3 nm, for example 1.5 nm to 2 nm or 2.5 nm. These carbon implantations help to promote the adhesion of the (C) layer.

[0130] In this mode, the invention makes it easier to manufacture a part (P) with a layer (C) exhibiting both good electrical and electrochemical performance, and good adhesion to the part (P), because the adjustment of the energy level is configured both to implant carbon in the part (P) and to deposit carbon in the form of sp3 bonds.

[0131] The performance of a part (P) corresponding to the invention was compared to that of prior art devices.

[0132] In this mode, the part (P) consists of a 316L substrate (S), a titanium sublayer (SC) measuring 50nm thick, and a 70nm thick carbon layer (C) deposited by cathodic arc, within which the sp3 hybridized bond content decreases, in a direction from the substrate (S) to the external surface (SE).

[0133] The carbon (C) layer contains neither dopants nor nitrogenous elements, in order to avoid the presence of preferential corrosion sites. However, the presence of oxygen within the (C) layer is not a distinguishing criterion.

[0134] The comparative samples are: Comparative sample 1 (EC1): a part consisting of a 316L substrate (S), a titanium sublayer (SC) measuring 50nm thick, and a carbon layer not corresponding to the invention, 70nm thick; Comparative sample 2 (EC2): a 316L substrate, without coating; Comparative sample 3 (EC3): a titanium substrate, without coating.

[0135] Corrosion tests at a potential of +1.6V / NHE were carried out.

[0136] With reference to the figure 6 The current density of the part (P) according to the invention drops rapidly and is less than 200 nA / cm² at the end of the test, a value comparable to that of the bare titanium substrate (EC3). The coating remains intact, unlike that of the comparative sample 1 (EC1), whose carbon layer (C) is consumed. At the end of this more stringent test, the RCI of the part (P) is low, which is not the case for bare titanium.

[0137] The part (P) according to the invention therefore exhibits both very good conductivity (low RCI), and very good resistance to oxidation (low corrosion current, no consumption of layer (C)).

[0138] Other corrosion tests at a potential of +1.6V / NHE were carried out.

[0139] With reference to the figure 7 , we compared the corrosion resistance of a part (P) according to the invention with respect to the comparative sample 1 (EC1).

[0140] The characteristics of the sp3 content according to the invention have an impact on the high-potential performance of layer (C): the layer (C) of the part (P) is not consumed, and the corrosion current density is very low; comparative sample 1 (EC1) has a low final corrosion current, but the carbon layer (C) is completely consumed, and the metallic sub-layer (SC) is exposed on the surface (which corresponds to the finding of a higher corrosion current between 0h and 6h).

[0141] With reference to the figure 9 , another corrosion resistance test is carried out, this time at a potential of 1.2V / NHE for 10h and at a temperature of 90°C.

[0142] Following this further severe test, it is observed once again that the part (P) according to the invention exhibits better corrosion resistance (approximately 10 mΩ.cm 2< only), than the comparative sample (EC1) (more than 100 mΩ.cm 2<).

[0143] A person skilled in the art will be able to adapt the deposition parameters and in particular the influential parameter, in order to continuously change the proportion of sp3 hybridized atoms relative to sp2 hybridized atoms, and thus conform to the invention.

[0144] It is recalled that the invention has the advantage of conferring greater durability to the parts (P), particularly at higher temperature and higher potential (which are two more severe conditions compared to standard conditions as described by the DOE), compared with state-of-the-art solutions for proton exchange membrane fuel cells.

[0145] Furthermore, part (P) can be shaped differently from the examples given without departing from the scope of the invention, which is defined by the claims.

[0146] In some embodiments, the part (P) has a stack of several sublayers (SC) arranged between the substrate (S) and the outer layer (C) of carbon-based material.

[0147] The carbon layer (C) is described as external because it is the layer (C) intended to be in contact with the electrochemical environment of the cell, and must ensure both the protection of the component (P) against corrosion and promote electrical conductivity. This terminology does not preclude the presence of an additional thin carbon layer, for example of the aC or aC:H type, deposited on top of the carbon layer (C).

[0148] The sp3 bonding content within the (C) layer may exhibit profiles other than those illustrated.

[0149] Furthermore, the technical characteristics of the various embodiments and variants mentioned above can be combined, in whole or in part. Thus, the part (P) can be adapted in terms of cost, functionality, and performance.

Claims

1. A part (P) comprising a metal substrate (S) and a layer (C) of amorphous carbon-based material having sp2 hybridised bonds and sp3 hybridised bonds, the layer (C) having: - a first content (sp3%1h) of sp3 hybridised bonds on the substrate (S) side, and - a second content (sp3%2b) of sp3 hybridised bonds on the side of an external surface (SE) of the layer (C), - the first content (sp3%1h) being greater than the second content (sp3%2b), characterised in that an average content within the layer (C) of sp3 hybridised bonds is between 5% and 65%, and preferably between 5% and 45%, and in that the content of sp3 hybridised bonds evolves continuously within the layer (C).

2. The part (P) according to claim 1, characterised in that it comprises a single layer (C) of carbon-based material.

3. The part (P) according to either of the preceding claims, characterised in that the carbon-based material comprises carbon in amorphous form, "a-C".

4. The part (P) according to any of the preceding claims, characterised in that the carbon-based material is free of dopants.

5. The part (P) according to any of the preceding claims, characterised in that the carbon-based material is free of nitrogen and hydrogen, and even more preferentially free of oxygen.

6. The part (P) according to any of the preceding claims, characterised in that the carbon-based material has an average content of sp3 hybridised bonds of between 10% and 45%, and preferentially between 10% and 30%.

7. The part (P) according to any of the preceding claims, characterised in that the layer (C) has a thickness of between 20 nm and 250 nm, preferably between 50 and 150 nm.

8. The part (P) according to any of the preceding claims, characterised in that it comprises an underlayer (SC) which is located between the substrate (S) and the layer (C) of carbon-based material and is in contact with the layer (C).

9. The part (P) according to claim 8, characterised in that the material of the underlayer (SC) is selected from chromium, titanium, zirconium, tantalum, niobium, or the alloys thereof and the nitrides, nitrocarbides and carbides thereof, or the material of the underlayer (SC) is a ceramic.

10. The part (P) according to any of the preceding claims, characterised in that carbon is implanted in a material receiving the layer (C).

11. A method for the vacuum vapour deposition of a layer (C) of a material based on amorphous carbon on a substrate (S), characterised in that a parameter influencing the ratio of sp2 or sp3 hybridised bonds in the carbon varies continuously during the deposition of the layer (C) such that it has: - a first content (sp3%1h) of sp3 hybridised bonds on the substrate (S) side, and - a second content (sp3%2b) of sp3 hybridised bonds on the side of an external surface (SE) of the layer (C), which second content is lower than the first content (sp3%1h), and in that an average content within the layer (C) of sp3 hybridised bonds is between 5% and 65%, and preferably between 5% and 45%, and in that the content of sp3 hybridised bonds evolves continuously within the layer (C).