METHOD FOR FORMING A FUEL CELL STACK

Abstract

A method for forming a fuel cell stack (1), the method comprising: Providing a pair of plates (65A, 65B) of a bipolar plate assembly (65), each of the plates comprising: a reactant channel (70), a reactant collection area (85), a coolant channel and a coolant collection area on a surface thereof; and a bead (105) formed integrally with the respective plate (65A, 65B) and projecting from the surface thereof, defining an engagement portion (107) thereon; Curing a microseal (110) deposited on the engagement portion (107) of at least a first plate (65A) of the plate pair (65A, 65B); mutually aligning the pair of plates (65A, 65B) in a facing and adjacent relationship along a stacking direction such that at least one cured microseal (110) contacts a cured microseal (110) deposited on the engagement portion (107) of the second plate (65B) of the pair of plates (65A, 65B); compressively forming a stack of such assemblies (65) aligned along the stack dimension such that upon cooperative engagement between the pair of plates (65A, 65B) of each of the assemblies (65), the engagement portions (107) and any cured microseals (110) disposed therebetween come into contact with the surface to achieve liquid isolation of a reactant or coolant which, during operation of said stack, is directed through a respective one of said channels or collection areas defined by the bead (105); Activating the adhesion between the two micro-seals (110) only after the aligned assemblies (65) are held compressed in said housing.

Description

BACKGROUND OF THE INVENTION

[0001] The present invention relates generally to a method of forming a fuel cell stack to provide improved reactant and coolant sealing in bonded or fluidly cooperating fluid transport plates.

[0002] Fuel cells convert fuel into usable electricity through electrochemical reaction. A significant benefit of this energy production is that it is achieved without relying on ignition as an intermediate step. Fuel cells have several environmentally friendly advantages over internal combustion engines (ICEs) for propulsion and similar propulsive applications. In a typical fuel cell—such as a proton exchange membrane or polymer electrolyte membrane (PEM in both cases) fuel cell—a pair of catalyzed electrodes is separated by an ion-permeable medium (such as Nafion™) in what is commonly referred to as a membrane-electrode assembly (MEA). The electrochemical reaction occurs when a first reactant, in the form of a gaseous reductant (such as hydrogen, H 2) is introduced and ionized at the anode and then passes through the ion-permeable medium, which reacts with a second reactant in the form of a gaseous oxidizing substance (such as oxygen O 2) introduced through the other electrode (the cathode); this combination of reactants forms water as a byproduct. The electrons released during the ionization of the first reactant are passed in the form of direct current (DC) through the external circuit, which typically includes a charge (such as an electric motor or various pumps, valves, compressors, or other components that carry liquid), to the cathode, where useful work can be done. The current generated by this passage of DC current can be amplified by combining numerous such cells into a larger power-generating arrangement. In this design, the fuel cells are connected along a common stacked dimension—almost like a deck of playing cards—to form a fuel cell stack.

[0003] In such a stack, adjacent MEAs are separated from each other by a series of reactant flow channels, usually in the form of gas-impermeable, bipolar plates (herein referred to as flow field plates), which—in addition to promoting the transport of reactants, coolant, and byproducts—provide structural support to the MEA and also collect or transport electrical current. In a common variant, the channels typically follow a serpentine pattern covering most of the opposing, generally flat surfaces of each plate. The juxtaposed arrangement of plate and MEA promotes the transport of reactants to or from the fuel cell, while additional channels (fluctually decoupled from the reactant channels) can also be used for coolant transport.In one embodiment, the bipolar plate itself is an assembly formed by attaching a pair of thin metal sheets (called half-plates) with channels embossed or otherwise integrally formed into their surfaces. The various reactant and coolant flow paths formed by the channels on each side usually meet at a manifold (also referred to herein as a collection region or collection area) defined by one or more opposite edges of the plate. Examples of all of these functions—as well as a common construction of such a bipolar plate assembly that may be used in PEM fuel cells—are shown and explained in commonly owned U.S. patents US 5,776,624 A and US 8,679,697 B1.

[0004] It is important to avoid leakage and associated fluid interference in a PEM fuel cell stack. To overcome such leakage, the assignee of the present invention has applied a relatively thick elastomeric seal to discrete portions of the relatively flat surface of the bipolar plate. While the thick nature of the seals is useful for achieving the required degree of sealing, such an approach is impractical in actual fuel cell stacks consisting of more than one hundred bipolar plates and MEA assemblies, as size concerns—particularly in the limited spaces of an automotive engine compartment—become paramount. Furthermore, the difficulty of ensuring consistent and repeatable seal placement makes this approach costly.

[0005] As an alternative to using thick elastomeric seals, the assignee of the present invention has developed integrally molded bipolar plate seals where embossments are formed in the plate surfaces in a manner generally similar to those forming the reactant and coolant channels, thus creating gasket-like outwardly projecting metal beads to form discrete contact points between adjacent plate surfaces. These beads (which may, for example, define a cross-section with a rectangular, trapezoidal, hemispherical, or other similar shape) are more compatible with mass production requirements than the aforementioned deposition of a thick elastomeric sealant.In particular, the applicant has applied a thin, relatively soft, compliant gasket layer where, ideally, there is no change in thickness or structural stiffness along the length of the gasket, so that the nominal gasket pressure (based on the applied stacking force per gasket length divided by gasket width) should be substantially uniform. Nevertheless, proper sealing and avoidance of pressure variations along the length of the bead is difficult to achieve, particularly given the inherent pitfalls of fuel cell stack manufacturing, which involve both dimensional tolerances of the formed beads and the misalignment of one hundred or more individual cells within the stack, making variation in the effective gasket pressure and concurrent leakage along the length of the bead in one or more regions of the plate unavoidable.

[0006] An additional difficulty arises from how the sealant is adhered within the bipolar plate assembly. In the approach investigated so far and discussed above by the applicant, the sealant first forms an adhesive bond between itself and the bead substrate. As mentioned above, while conventional thick gaskets tend to be relatively insensitive to such adhesion, the inventors have found that any attempt to reduce the sealant thickness results in significant sensitivity of the gasket pressure to how the sealant is constrained at the interface between it and the underlying substrate. For example, for a 1.1 mm wide gasket that is relatively thin (i.e., about 0.15 mm high), the locations that lose adhesion or have no adhesion to begin with may have substantially lower pressures than those of the same gasket with proper adhesion.A further difficulty arises from the fact that the long lifetime of a fuel cell stack in a harsh automotive environment often results in some delamination along the length of the cured sealant. Such previous research by the inventors has shown that if a point or section loses adhesion during the lifetime of the fuel cell stack, that area can lose 75% of the seal pressure, potentially resulting in unacceptably high reactant or coolant leakage.

[0007] WO 2011 / 026544 A1 describes a method for forming a fuel cell stack, wherein bipolar plates are provided, each comprising a pair of plates having so-called hinterland stoppers and channels. A gasket is applied to the bipolar plate, which can plastically adapt to the geometry and any tolerances of the bipolar plate. The gasket can be cured before the plates are assembled. The individual fuel cell components are then stacked, with the gasket positioned between the bipolar plate and the MEA or subgasket. The stack is then compressed, after which the gasket is cured.

[0008] Further prior art can also be found in the documents DE 101 58 772 C1, DE 103 24 157 B3, DE 10 2014 104 016 A1 and US 2010 / 0 221 638 A1. SUMMARY OF THE INVENTION

[0009] In view of the above difficulties, the inventors have developed a robust method for protecting against loss or uneven adhesion of a sealant used within a bipolar plate assembly or a fuel cell stack having such bipolar plates, as well as a method for using such a sealant in conjunction with molded metal beads.

[0010] According to the invention, a method for forming a fuel cell stack is presented, which is characterized by the features of claim 1.

[0011] The method involves providing a pair of plates used in a bipolar plate assembly, each plate including one or more reactant channels, reactant collection areas, coolant channels, and coolant collection areas on its surface, and a molded metal bead protruding from the surface. The uppermost portion of the metal bead defines a generally planar gasket-like engagement portion configured to engage one of the compatible surfaces of a facing adjacent plate, MEA, or similar surface having a microseal disposed thereon. In situations where the microseal is molded and bonded to the metal bead, this combination is referred to as a metal bead gasket (MBS).In one form, the assembly is stacked by placing a first microseal onto the engaging portion of the bead of a first plate such that the microseal cures before any stacking or similar engagement occurs between adjacent facing surfaces. This curing temporarily bonds or tacks the microseal—such as through relatively weak van der Waals forces or the like—to the surface in question (i.e., sub-gasket, engaging portion of the metal bead, MEA, or a second cured microseal). After this curing, these assemblies are aligned with corresponding MEAs along a stacking dimension and then housed in a compressed state in a housing to bring the stack to its final and proper height dimension.Activation of the adhesion between the microseal and the plate within each assembly (and optionally between adjacent assemblies) occurs only after the stack is substantially assembled, properly aligned, and held compressed within the package. This subsequent activation of the adhesion between the microseal and its adjacent substrate has the effect of more carefully and evenly distributing the microseals within the final stack. Additionally, it helps to make the seal pressure insensitive to subsequent loss of adhesion that may occur over the stack's service life.

[0012] In the present context, the curing of the microseal material and the activation of the adhesion at the contact points should be understood as two separate steps, with curing corresponding to the use of heat or a similar means to facilitate cross-linking within the polymer network of the microseal to create the desired structure, while the activation of the adhesion is intended to create a substantially strong chemical bond between the microseal material and the substrate (namely either the metal bead and / or the sub-seal) to which it is attached.In a preferred form, the engagement region of the bead of the second plate of the pair is in contact with the second microseal, such that upon cooperative engagement between the plate pair, the two MBSs are in contact to provide substantial liquid isolation of a reactant or coolant flowing through each of the channels or headers during stack operation. In another preferred form, the microseal may be applied to a sub-gasket used to reduce leakage at the periphery of the MEA.

[0013] In the present context, the term "microseal" serves to distinguish the thin, low aspect ratio (i.e., less than one) seal of the present invention from those with thick (i.e., high aspect ratio equal to or greater than one) designs. As mentioned above, relatively thick seals are not economically viable for high-volume production of fuel cell stacks with a large number of bipolar plates, MEAs, and associated components. Furthermore, in the present context, an effective seal pressure (or effective sealing pressure or effective pressure in all cases P eff) from a conventional sealing pressure in such a way that the former accounts for variations in the stiffness or compliance of a deposited microseal that accompany the use of such seals in very thin shapes. The inventors have determined that conventional properties of the base material do not apply to these very thin microseal structures due to geometric constraints. For example, the elastomeric material normally used for the compliant seal begins to behave in a stiffening manner when the seal is very thin relative to its width; thus, in situations where the seal is relatively wide relative to its height, spatial constraints are imposed on the seal's ability to compress in response to applied loads. These spatial constraints are more pronounced when the microseal is bonded to one or more substrates.These effects in turn tend to result in an effective elastic modulus (E. eff ) is significantly higher than the property of the base material of the elastomeric material from which the micro-seal is made. Details related to this increase in E eff and consequently the sealing pressure P effcan be found in an article entitled "The effect of compressibility on the stress distributions in thin elastomeric blocks and annular bushings" by Yeh-Hung Lai, D.A. Dillard, and J.S. Thornton in The Journal of Applied Mechanics (1992). This larger effective elastic modulus is evident because it requires a correspondingly higher compressive load to cause the same compression displacement of the microseal. Since fuel cell stacks are typically assembled under a sealing pressure between 1 and 6 MPa, limiting microseal displacement (for example, by prematurely activating an attached (or otherwise formed) adhesive on compression-contacted components) during the stack assembly process is tantamount to inhibiting the ability of a microseal to conform to the irregularities in the substrate surfaces (i.e.Metal bead, sub-gasket, or MEA), which in turn undesirably leads to an increased incidence of leakage. It is this type of prevented movement that the present invention avoids by removing the spatial limitations associated with bonding the gasket to the substrate as described herein.

[0014] According to another aspect of the present invention, a bipolar plate assembly for a fuel cell system includes a pair of plates, each having one or more reactant channels, reactant collection areas, coolant channels, and coolant collection areas on a surface thereof, as disclosed. At least one of the plates defines a molded metal bead seal extending from the surface in a manner generally similar to the plate protrusions defining the reactant or coolant channels.

[0015] According to another aspect of the present invention, a fuel cell stack is disclosed. The stack includes numerous individual fuel cells aligned along a stack axis and compressibly contained within a housing. Each of the cells includes a pair of plates in an opposed arrangement of their surfaces, each surface defining one or more reactant channels, reactant collection regions, coolant channels, and coolant collection regions. The surfaces also include a molded metal bead projecting therefrom to define an engagement portion thereon. MEAs are disposed between at least some of the plate pairs; in such a case, each of the reactant channels of that plate pair is disposed in fluid communication with a respective anode or cathode within the MEA.A microseal is disposed at the mating portion of at least one plate pair such that the microseal is substantially cured prior to alignment, but is not substantially bonded until the numerous cells are compressed and contained within the package. Depending on the type of cell assembly, such contact may be from the microseal to an adjacent sub-seal, MEA, an mating portion of an adjacent bipolar plate assembly, and the cured microseal deposited on the mating portion of the adjacent bipolar plate assembly.

[0016] These and other aspects or embodiments will become apparent to those of ordinary skill in the art upon reading the following detailed description and the appended claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The following detailed description of the preferred embodiments of the present invention will be best understood when read in conjunction with the following drawings, in which like structures are designated by like reference numerals and the various components of which are not necessarily drawn to scale: Fig. Figure 1 shows a schematic exploded view of a fuel cell stack that can be assembled according to one aspect of the present invention; Fig. Figure 2 is a simplified illustration of a partially exploded sectional view of a portion of a fuel cell with surrounding bipolar plates; Fig. 3 is a detailed top view of a bipolar plate made of Fig. 2, which includes a metal bead capable of accommodating a microseal according to one aspect of the present invention; Fig. Figure 4 is a graphical illustration of how preventing the formation of an adhesive bond between the microseal and a bipolar plate assembly prior to individual cell stacking, alignment, and compression according to one aspect of the present invention improves leak prevention; and Fig. Figure 5 shows a simplified side view of the relative arrangement of the metal beads, microseals, and sub-seals within an adjacently disposed bipolar plate assembly according to one aspect of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] Referring first to the Fig. 1 to 3 show a simplified view of a fuel cell stack 1 in exploded view ( Fig. 1), a PEM fuel cell 30 ( Fig. 2) and a bipolar plate arrangement 65 ( Fig. 3). The stack 1 includes a housing 5 consisting of a dry end unit plate 10 and a wet end unit plate 15; this (as well as others not shown) can help to perform the compressive clamping of the compression retention system of the housing 5; such a compression retention system includes a plurality of bolts (not shown) extending through the thickness of the stack 1, as well as through various side panels 20 and rigid clamp members 25 extending vertically along the stacking direction (the Y-axis) to secure the wet end unit plate 15 to the dry end unit plate 10. Stacks of various fuel cells 30 are securely held in a compressed relationship along the stacking direction by the action of the bolts, clamp members 25, and other components in the housing 5. While the stacking axis of the fuel cell 1 in the present context is along a substantially vertical (i.e.Y) Cartesian axis, therefore, the majority of the generally flat surfaces of each fuel cell 30 lie in the XZ plane. Regardless, those skilled in the art will recognize that the precise orientation of the cells 30 within the stack 1 is not critical.

[0019] With special reference to the Fig. 2 and Fig. 3, the fuel cell 30 includes a substantially planar proton exchange membrane 35, an anode catalyst layer 40 in facing contact with one side of the proton exchange membrane 35, and a cathode catalyst layer 45 in facing contact with the other side. Collectively, the proton exchange membrane 35 and the catalyst layers 40 and 45 are referred to as an MEA 50. An anode diffusion layer 55 is in facing contact with the anode catalyst layer 40, while a cathode diffusion layer 60 is in facing contact with the cathode catalyst layer 45. Each of the diffusion layers 55 and 60 is of a generally porous construction to allow the passage of gaseous reactants toward the catalyst layers 40 and 45.Together, anode catalyst layer 40 and cathode catalyst layer 45 are referred to as electrodes and may be formed as separate layers of their own as shown or, alternatively (as mentioned above), as at least partially embedded in diffusion layers 55 and 60, respectively, and partially embedded in the opposite sides of proton exchange membrane 35.

[0020] In addition to providing a substantially porous flow path for reactant gases to reach the appropriate side of the proton exchange membrane 35, the diffusion layers 55 and 60 establish electrical contact between the electrode-catalyst layers 40, 45 and a bipolar plate assembly 65, which in turn acts as a current collector. Furthermore, the generally porous nature of the diffusion layers 55 and 60 provides a conductor for the removal of product gases generated at the catalyst layers 40, 45. In addition, the cathode diffusion layer 60 generates significant amounts of water vapor within the cathode diffusion layer. This property is important for keeping the proton exchange membrane 35 hydrated. The permeation of water through the diffusion layers can be controlled by adding small amounts of polytetrafluoroethylene (PTFE) or similar materials.

[0021] Although theoretically with a thick-walled structure in Fig. 2, the individual plates 65A and 65B (also referred to herein as half plates) that make up the assembly 65 preferably utilize the thin or foil-like structure (as in connection with Fig. 3 is shown and described in more detail); therefore, Fig. 2 cannot be used to derive the relative thickness of the assembly 65. Simplified opposing surfaces defined by the oppositely adjacent half-plates 65A and 65B are provided to delineate each MEA 50 and associated diffusion layers 55, 60 from the adjacent MEAs and layers (both not shown) in stack 1. One half-plate 65A engages the anode diffusion layer 55, while a second half-plate 65B engages the cathode diffusion layer 60. The two thin, opposing metal sheets that comprise half-plates 65A, 65B define, with appropriate compression and associated joining techniques, the plate assembly 65. Each half-plate 65A and 65B (which, when assembled as a unified whole, would constitute the bipolar plate 65) defines various reactant gas flow channels 70 along a respective plate side.Although the bipolar plate 65 is shown (for stylized purposes) as defining exclusively rectangular reactant gas flow channels 70 and surrounding structures, those skilled in the art will recognize that a more accurate (and preferred) embodiment generally utilizes channels 70 following a serpentine pattern. The tops of the channels define lands 72 that act as engagement surfaces with complementarily shaped lands 72 of opposing plates.

[0022] During operation, a first gaseous reactant such as H 2 from the half-plate 65A to the anode side of the MEA 50 through the channels 70, while a second gaseous reactant, such as O 2(usually in the form of air) is conducted from half-plate 65B to the cathode side of the MEA 50 through the channels 70. Catalytic reactions occur at the anode 40 and cathode 45, respectively, generating protons that migrate through the proton exchange membrane 35 and electrons that result in an electric current that can be transmitted through the diffusion layers 55 and 60 and the bipolar plate 65 due to contact between the latter and layers 55 and 60. Associated channels (not shown) can be used to transport coolant for regulating the temperature generated by the fuel cell 1. In situations where the half-plates 65A, 65B are configured to conduct coolant, their functions are comparable to those of the plates conducting reactants; they are similarly constructed and will not be discussed further herein.

[0023] Sub-gaskets 75 (a portion of which is shown in cross-section) may be disposed at many locations within the stack 1 for improved sealing. In a preferred form, they are made of a non-conductive and gas-impermeable material (such as plastic) affixed to the perimeter of the MEA 50 to separate the various electrically conductive layers (such as the electrode 40 and gas diffusion layer 55 on the anode side, and the electrode 45 and gas diffusion layer 60 on the cathode side). Another important function of the sub-gasket 75 is to prevent junction leakage and associated mixing of the reactants at the edge of the MEA 50. In one form, the sub-gasket 75 defines a generally planar frame-like member circumferentially disposed to protect the edge of the MEA 50.As such, the sub-gasket 75 is preferably located where the elastomer seal (discussed below) contacts either the MEA 50 or the facing surface of one or more metal beads (also discussed below). This helps reduce leakage of reactant gases and coolant to the outside, as well as their mixing, in the collection area 85. Additionally, the sub-gasket 75—which is preferably between about 50 µm and 250 µm thick—is often used to extend the separation of gases and electrons between the catalyst layers 40 and 45 to the edge of the MEA 50, thereby increasing the active surface area of ​​the membrane 35.

[0024] With special reference to Fig. 3 shows an exploded view of two adjacently stacked half-plates 65A, 65B forming the bipolar plate assembly in more detail. In particular, the individual half-plates 65A, 65B each include both an active region 80 and a collection region 85, the former forming a planar opposing relationship with the electrochemically active region associated with the MEA 50 and the diffusion layers 55 and 60, and the latter associated with an edge (as shown) or peripheral region (not shown) where openings formed through the plates 65A, 65B serve as conductors for the transport of reactants, coolants, or by-products to and from the stacked fuel cells 30. As can be seen from the exploded view in Fig. 3, these two half-plates 65A, 65B can be used to form a sandwich-like structure with MEA 50 and anode and cathode diffusion layers 55, 60, and then repeated as many times as necessary to form a fuel cell stack 1. In one form, the anode half-plate 65A and / or the cathode half-plate 65B are made of a corrosion-resistant material (such as 304L SS or the like). The generally serpentine-shaped flow channels 70 form a tortuous path from near the edge 90 adjacent to the collection region 85 to near the opposite edge 95 adjacent to the opposite collection region 85.As can be seen, the reactant (in the case where one plate 65A, 65B is in a facing relationship with MEA 50) or the coolant (in the case where one plate 65A is in a facing relationship with the back of another plate 65B, thereby forming coolant channels) is directed from a series of repeating exits or recesses to the channels 70 that form a header region 100 located between the active region 80 and the collection region 85 of one (e.g., supply) edge 90; a similar implementation is located at the opposite (e.g., waste) edge 95. In an alternative embodiment (not shown), the supply and waste collection regions may be located near the same edge (e.g., either 90 or 95).In situations where the individual half-plates 65A, 65B are made of a formable material (such as the aforementioned stainless steels), the various surface features (including recesses, channels, lands, or the like) are preferably embossed using techniques already known, thereby ensuring that both the channels 70, the lands 72, and their corresponding structures, in addition to the metal beads (which are explained in more detail below), are integrally formed from a sheet of material.

[0025] Next, with reference to Fig. 5, in one embodiment, a generally planar portion of a metal bead 105 of the respective plates 65A, 65B directly engages the sub-gasket 75; this planar engagement portion acts as a gasket to which the sub-gasket 75 can be connected by a micro-seal 110. In another embodiment (not shown), a generally planar portion of a metal bead 105 of the respective plates 65A, 65B directly engages the metal bead 105 of the opposing bipolar plates 65B, 65A. In either configuration, the gasket-like structure of the metal bead 105 and the microseal 110 together define the MBS 115. The gasket-like nature of the metal bead 105 arises from the fact that it is formed as an upstanding rectangular, trapezoidal (as shown), or slightly curved protrusion by stamping from the thin metal material from which the respective plates 65A and 65B are made.The metal beads 105 preferably define a height of approximately 300 µm to 600 µm and a width between approximately 1 mm and 4 mm. The upper side defines an engagement portion 107 that is substantially similar in structure and function to the lands 72, which may also be formed within one or both plates 65A, 65B. As such, the engagement region 107 corresponds to the portion of the metal bead 105 configured for facing contact with the microseal 110, sub-seal 75, MEA 50, or adjacent metal bead 105.Essentially, the microseal 110 functions (a) to fill the surface imperfections of the metal bead 105 or sub-seal 75 in the mating portion 107, (b) to create a more uniform sealing force per length along the metal bead 105 by providing a compliant cushion to compensate for the unevenly compressed height of the metal bead 105, (c) to prevent liquid permeation (such as a reactant) through its bulk alloy, and (d) to prevent leakage flow through the interface between either (i) the sub-seal 75 and microseal 110 or (ii) the metal bead 105 and microseal 110, depending on the precise mating during formation of the stack 1.The elastomeric microseal 110 is shown on the engagement portion 107, although it will be appreciated by those skilled in the art that the microseal 110 may also be formed on the surfaces of the sub-seals 75 as well as (or instead) directly on the metal bead 105.

[0026] In a preferred embodiment, the microseal 110 is between about 30 µm and 300 µm thick and between about 1 mm and 3 mm wide. With such dimensions, the microseal 110 can be deformed under the high compressive loads encountered during the formation of the stack 1. Spatial confinement and the inherent incompressibility of the material of the microseal 110 can cause stresses in the microseal, particularly at its contact point with the corresponding substrates of the engaging portion 107 of the metal bead 105 or the sub-seal 75, where the microseal 110 is bonded. The inventors have discovered that by substantially delaying the formation of the adhesive bond until after assembly and compression of the stack 1, a significant reduction in stress can be realized by softening the edge boundaries and similar effects.This, in turn, reduces the likelihood of premature stress-induced failure of the microseal 110. In an idealized sense, where latent adhesion may be used in accordance with the present invention, no bonding occurs between the microseal 110 and an adjacent substrate prior to assembly and compression of the stack 1. Therefore, in the present context, descriptions that include the formation of an adhesive bond (such as the phrase "substantial activation of the bond" between the microseal 75 and an adjacent substrate) are understood to encompass those situations where a small amount of bonding may occur at the interface between the mating surfaces prior to assembly and compression of the stack 1, as long as the substantial bulk of this bonding is avoided until such assembly and compression is complete.

[0027] In fact, a small residual amount of non-latent adhesion prior to assembly of stack 1 may even be beneficial to promote improved handling of the individual parts prior to assembly and compression. As a result, the residual adhesion resulting from the bonding of microseal 110 is not so great as to cause permanent facing alignment between adjacent assemblies 65 (or individual components within an assembly 65) prior to formation of stack 1, but sufficient to avoid relative sliding between adjacent surfaces in a plane and thus facilitate handling. To this end, the inventors have determined that it may be desirable to have a weak form of adhesion between microseals 110 and metal beads 105 after microseal 110 has cured, but before the permanent bonding step is activated.Therefore, during assembly of the stack 1, the mechanical stresses in the intermediate region due to the compression force are expected to break these relatively weak bonds, allowing the microseal 110 to spread along the intermediate region. In one form, relatively weak bonds (such as through van der Waals forces or similar interactions) between the microseal 110 and the metal bead 105 may be supported after the microseal 110 is applied and cured to effect this temporary degree of adhesion. In the present context, these weak (or temporary) forms of adhesion are to be distinguished from more permanent variants, such as those due to covalent bonds, which create a strong chemical bond.Therefore, essentially all bond activation occurs only after all fuel cells within stack 1 are aligned and pressed together, with the possible exception of the residual adhesion.

[0028] The material used to form the microseal 110 is comprised of a compliant plastic or elastomer (including polyacrylate, hydrated chlorosulfonated polyethylene, ethylene acrylic, chloroprene, chlorosulfonated polyethylene, ethylene propylene, ethylene vinyl acetate, perfluoroelastomer, fluorocarbon, fluorosilicone, hydrogenated nitrile, polyisoprene, micronuclear polyurethane, nitrile rubber, natural rubber, polyurethane, styrene-butadiene rubber, TFE / propylene, silicone, carboxylated nitrile, or the like) and is preferably applied by a screen printing process, although other approaches such as ink pad printing, injection molding, or other deposition methods may be used. As mentioned above, in a preferred form, the layer formed by the microseal 110 has a thickness between about 30 and 300 µm, while the preferred width across the engagement portion 107 is between about 1 mm and 3 mm.In a particular form, the material used in the microseal 110 includes at least silicone (for example in the form of vinyl polydimethylsiloxane, PDMS), a structural reinforcement (such as silicon dioxide SiO. 2), a bonding catalyst (such as a platinum-containing vinyl-SiH bonding catalyst), and a coupling agent (such as 1,2 bis(triethoxysilyl)ethanes). By using one of these preferred formulations, the microseal 110 exhibits a two-part property, the first of which promotes rapid curing and structural assembly, while the second of which delays the formation of the interlayer bond until after assembly and compression of the fuel cell stack 1. Details related to these materials—as well as the use of screen printing to deposit them on a suitable metal bead 105 or sub-seal substrate 75—can be found in concurrently filed U.S. patent application Ser. No. 2017 / 0226392A1. Additional screen printing features unique to seal formation are disclosed by way of example in U.S. patent No. 4,919,969A.

[0029] Next, with reference to Fig. Figure 4 is a graph showing predicted seal pressure versus displacement as an example to demonstrate the improvement of the present invention over the prior art. As mentioned above, the nominal seal pressure in a fuel cell stack ranges from approximately 1 to 6 MPa. In both the prior art and the present invention cases shown in the graph, the values ​​for the width and thickness of the microseal were chosen to be 1.1 mm and 0.15 mm, respectively.In the prior art, the adhesion between the microseal and the metal bead is activated before the cell is assembled and reaches a nominal sealing pressure of 3 MPa at a displacement of 0.046 mm (corresponding to location A), whereas in the present invention, the adhesion between the microseal 110 and the metal bead 105 is activated after the microseal 110 has cured; this was also subjected to a nominal sealing pressure of 3 MPa, and at this time a displacement of 0.092 mm had occurred (corresponding to location C). Therefore, the present invention requires more displacement for the same sealing pressure of 3 MPa; this, in turn, allows the microseal 110 to move during alignment, stacking, and (at least) parts of the compression process, so that it fills any gaps and asperities between the adjacent metal beads 105 of the bonded bipolar plates (such as plates 65A and 65B in FIG. Fig.2). Another advantage of the present invention is its ability to maintain seal pressure during loss of adhesion (corresponding to the arrow from location A to location B) throughout the lifetime of the stack 1. In a fuel cell stack, it is known that the distance between adjacent bipolar plates is typically kept constant, which in the present invention results in constant displacement of the metal bead 105 and microseal 110. When a conventional sealant loses adhesion, it spreads along the interface, which in turn results in a decrease in seal pressure from 3 MPa to less than 1 MPa, as demonstrated by the movement from location A to location B.On the other hand, the sealing pressure of the sample of the present invention does not decrease upon loss of adhesion because the relatively weakly adhesive intermediate region, even during alignment and compression, allows the microseal 110 to expand laterally along the intermediate region during stack assembly, thus only reaching its final shape and dimensions at the time of stack formation and subsequent adhesive bonding. Furthermore, excessive sealing pressure (which can occur, for example, when attempting to achieve the same nominal pressure by compressing to a large displacement in the prior art) has detrimental effects not only on the microseal 110 but also on the other components in the stack 1, particularly through the development of large material stresses, creep, or the like.

[0030] Although not depicted, a particular application for a system based on a stack of PEM fuel cells 1 could be an automobile or similar vehicle. In the present context, it will be appreciated that the term "vehicle" may refer to cars, trucks, vans, sport utility vehicles (SUVs), or other such automotive forms such as buses, aircraft, watercraft, spacecraft, and motorcycles; all of which have the potential to be made compatible with the present invention to generate motive or locomotion energy.

[0031] As explained above, with the specificity of ensuring that no more than weak, temporary forms of adhesion are used between adjacent microseals 110 and their associated substrates, the term "substantially" in modifying the assembly of the fuel cell stack 1 is used herein to represent that some of these temporary or residual adhesives may be used herein without changing the basic function of the matter being addressed; therefore, the inclusion of some of these weaker, more temporary ways of holding the various components of the stack 1 during assembly does not detract from the fact that the substantial (i.e., more permanent) form of adhesion is not used until the various stacked cells are aligned, crimped, and secured within the stack 1.

Claims

[1] A method for forming a fuel cell stack (1), the method comprising: Providing a pair of plates (65A, 65B) of a bipolar plate assembly (65), each of the plates comprising: a reactant channel (70), a reactant collection area (85), a coolant channel and a coolant collection area on a surface thereof; and a bead (105) formed integrally with the respective plate (65A, 65B) and projecting from the surface thereof, defining an engagement portion (107) thereon; Curing a microseal (110) deposited on the engagement portion (107) of at least a first plate (65A) of the plate pair (65A, 65B); mutually aligning the pair of plates (65A, 65B) in a facing and adjacent relationship along a stacking direction such that at least one cured microseal (110) contacts a cured microseal (110) deposited on the engagement portion (107) of the second plate (65B) of the pair of plates (65A, 65B); compressively forming a stack of such assemblies (65) aligned along the stack dimension such that upon cooperative engagement between the pair of plates (65A, 65B) of each of the assemblies (65), the engagement portions (107) and any cured microseals (110) disposed therebetween come into contact with the surface to achieve liquid isolation of a reactant or coolant which, during operation of said stack, is directed through a respective one of said channels or collection areas defined by the bead (105); Activating the adhesion between the two micro-seals (110) only after the aligned assemblies (65) are held compressed in said housing. [2] The method of claim 1, wherein curing of the microseal (110) occurs on the engagement portion (107) such that substantially no adhesion occurs therebetween until the stack has been formed by compression as said. [3] The method of claim 1, wherein said microseal (110) defines an applied thickness between 30 µm and 500 µm. [4] The method of claim 1, wherein said microseal (110) defines an applied width between 0.5 mm and about 3.0 mm. [5] The method of claim 1, wherein said microseal (110) consists of at least silicone, a structural reinforcement, a bonding catalyst, and an adhesion promoter.

Images