An overview of bipolar plates in proton exchange membrane fuel cells

The need for clean energy is becoming more prevalent as the world's fossil fuel resources continue to diminish and the global demand for energy increases. Proton exchange membrane fuel cells (PEMFCs) have the potential to meet this demand. Their ability to produce electrical power efficiently is an attractive alternative to the internal combustion engine.

PEMFCs electrochemically react hydrogen and oxygen, producing electrical power without harmful emissions. Environmentally friendly operation combined with the generation of high-power densities makes PEMFCs a strong candidate to replace internal combustion engines for mobile powerplant applications.^1,2 In transportation applications, many cells are organized into a fuel cell stack. Each cell consists of a potential-producing membrane electrode assembly (MEA) separated by bipolar plates (BPPs), shown in Fig. 1. The BPPs both separate the individual cells and provide a conductive bridge from the cathode of one to the anode of the next.³ They also transport reactant gases (oxygen and hydrogen) to the electrodes (cathode and anode, respectively), provide efficient and uniform electron transport between the MEAs and the outer electrical system, and are responsible for heat and water management.^4,5 The plates are designed to facilitate an even distribution of hydrogen across one side of the membrane, and an even distribution of oxygen on the reverse side.

BPPs are the foundation of fuel cells, making up 80% of a fuel cell's weight, 50%–65% of its volume, and 40% of its total cost.^6,7 This makes it crucial to find a material that is lightweight, inexpensive, and easy to manufacture. Some other requirements for the plates include high mechanical strength, high electrical and thermal conductivity, ability to carry charge and heat away from the MEA, and ability to distribute the force applied by clamps that hold the fuel cell stack together.⁸ These requirements make the material selection an extremely important decision. The corrosive environment inside a fuel cell compounds the difficulty of finding a material that will satisfy both the previously stated demands of the plate and its operating environment.⁹ 

Materials currently in use fall into two main categories, coated metallic materials, and graphitic materials. Depending on material selection, various manufacturing methods can be utilized. Conventional manufacturing methods for metallic BPPs include powder metal forging, stamping, and hydroforming. Traditional manufacturing methods used for graphitic BPPs include compression and injection molding. However, none of these processes are truly idea for the mass-production of BPPs due to high tooling costs, inability to form complex shapes, lack of precision manufacturing, and impractical mass-production process in factories. It is especially important to use a process that can form complex shapes in the fabrication of intricate BPPs such as those with an integrated cooling jacket.⁴ Additive manufacturing (AM) techniques, such as selective laser sintering (SLS), are a viable solution to these problems. AM allows for the fabrication of far more complex parts due to the virtual absence of geometric constraints and the ability to save time and money by eliminating the waiting period and manual labor costs associated with tooling. This manufacturing technique allows for quick design changes and a rapid transition from the design phase to the manufacturing phase, unlike conventional manufacturing methods discussed for both metallic and graphitic plates.¹⁰ Novel contributions in this paper include a detailed explanation and cost analysis of traditional manufacturing and AM processes for metallic and graphitic-polymer BPPs.

Extensive efforts have been made to find innovative materials to be used in the fabrication of BPPs. These range from novel coatings such as the YZU001 like-diamond coating applied to 5052 aluminum explored by Lee et al.¹¹ to novel processing applied to a previously utilized material like the low-cost and high-power density-yielding expanded graphite developed by Yan et al.¹² The innovations also include completely novel materials fabricated specifically for application in BPPs, like a nylon-matrix stainless steel fiber composite explored by Kuo et al., the rapidly manufacturable and cost-effective thermoplastic–graphite composite produced by Cunningham et al., and the mass-producible high graphite-filled polypropylene (PP) composite tested by Derieth et al.^5,13,14 These represent some of the most novel solutions to the problem of selecting a material for use in fuel cell BPPs.

In 2017 the U.S. Department of Energy (DOE) set forth an updated list of guidelines for materials to be used for BPPs. They are as shown in Table I.

Ultimately, the requirements for BPPs will differ depending on the application (mobile/transportation or stationary) and operating conditions that the fuel cell is utilized with. Plates used for mobile/transportation applications need to be light, thin, and able to resist the demanding, corrosive environment inside the fuel cell. Stationary BPP applications do not have strict requirements for plate size or volume but need to be extremely durable. This paper will focus on BPPs for automobiles.

For BPPs with mobile/transportation applications, component volume is essential to consider. Fuel cells must compete with existing electricity-generating technology in order to be considered as a feasible alternative and commercially accepted.⁸ In order to optimize system performance and decrease stack length, plates should be made as slim as possible (minimize volume/kW).¹⁶ This requirement is placed to mediate any differences in stack performance or mass for different plate designs. Smaller fuel cell stacks allow a greater chance of it fitting in a variety of automobiles. Low-volume fuel cell stacks mean that the fuel cell active area composes most of the total stack area which in turn can increase power density and decrease cost and weight.¹⁵ 

Electrical conductivity is an important property to consider, as less parasitic resistance will lead to a more efficient cell. This effect is then multiplied as multiple cells are put in series. Thermal conductivity of a plate is important because the chemical reaction that happens in a cell produces almost as much heat as it does electricity. Allowing this heat to dissipate is crucial to the operation of the cell, as the membrane generally has a small tolerance for temperature deviation from its design point (60 to 80 °C).¹⁷ Metallic materials seem like an obvious choice given the need for high thermal and electrical conductivities and mechanical strength. When corrosion resistance is considered, metallic materials lose their edge (unless modified by a coating). Corrosion, specifically of metallic BPPs, causes ion contamination which is a detriment to the cell in a multitude of ways. The five processes by which ion contamination negatively affects the cell are explored by Ijaodola et al.,¹⁸ and consist of: reduction in proton conductivity, chemical degradation of the membrane, formation of higher resistance oxide layers, poisoning of the catalyst, and decreasing the oxygen transport rate due to decreased hydrophobicity. Preventing the poisoning of the polymer electrolyte membrane (PEM) resulting from the degradation of the BPP is a primary challenge in extending fuel cell life.¹⁹ Graphitic plates have the advantage of not suffering from corrosion, but their difficulty in machining leads to higher costs. Another potential solution to the corrosion problem is coating the metallic plates. These coatings range from metals such as gold, titanium, and zirconium, explored by Yoon et al.;²⁰ to novel coatings including carbon, investigated by Show,²¹ and a like-diamond film tested by Lee et al.¹¹ For applications such as fuel cell-powered vehicles, lower weight, volume, and cost are major factors. In the search for a material that fits these standards many material candidates have been proposed and explored, both in literature and in practical applications. These candidates can be seen with some of their properties in Table II, and they can be categorized into metallic and graphitic materials as shown in Fig. 2.

Although graphite has been established as a benchmark material for BPP fabrication,^23–26 its high brittleness, cost, weight, volume, and permeability are major deterring components that prevent the commercialization of graphite plates.²⁷ Metals have been identified as a promising alternative to traditional graphite due to its lower relative cost, high mechanical strength, ease of manufacturing, better durability to shocks and vibration, and little to no gas permeability.^3,10 Metallic materials also possess the ability to create plates with low volume.²⁸ However, metals have the tendency to corrode and therefore suffer from an excessive interfacial contact resistance (ICR).²⁹ These are both serious issues as BPPs are typically exposed to an operating environment with a pH of 2–3 at temperatures of around 80 °C, prime corrosion conditions for metal plates.³⁰ The self-passivating nature of metals also results in a high contact resistance leading to a high Ohmic overpotential.³¹ Corrosion-resistant metals such as stainless steel, Ni-based alloys, titanium, zirconium, niobium, tantalum, and hafnium present a possible option for BPP fabrication, but their high material cost discourages use for commercial fuel cells.^32–34 Coatings on metallic BPPs are a much more feasible option to protect metal plates from corrosion and have subsequently been heavily researched.¹⁹ 

Show²¹ investigated the effects of electrically conductive amorphous carbon (a-C) coatings on metal BPPs for PEMFCs. This study used titanium (Ti) BPPs with deposited a-C film coatings and compared those with coatings to those without. Ti has a highly resistive oxidized layer on their surface which increases the BPP's ICR and therefore lowers electric power generation efficiency. The oxidized layer was removed from the metal and replaced with an a-C coating. Show's research finds that the temperature at which the coating was applied directly impacted its electrical conductivity and subsequently the output power of the fuel cell. When deposited at high temperatures (above 550 °C), the a-C film became more electrically conductive (10⁻³ Ω cm). This reduced the ICR between the film surface and the MEAcausing low series resistance and high output power. Show's study demonstrates that the a-C coating technique can improve fuel cell efficiency.

Cho et al.³⁵ analyzed the performance of a 1 kW-class fuel cell stack that utilized BPPs made of 316 stainless steel and coated with tin. Cho recognized that one of the main issues in the development of metallic BPPs is preventing corrosion on the plates which can ultimately lead to the degradation of PEMFCs. In this experiment, the performance of graphite, noncoated, and Ti–N-coated AISI 316 stainless steel plates were compared in a single-cell and 12-cell 1 kW stack. High Ohmic resistances were observed in the metallic BPPs, especially the noncoated ones.

Metallic plate corrosion was linked to the contamination of the electrolyte membrane by corrosion of the alloys. The charge transfer resistance in the order of decreasing values: noncoated (35 mΩ cm⁻²), coated (31.2 mΩ cm⁻²), and pure graphite (30.7 mΩ cm⁻²). This order was a common theme throughout the experiment and was seen again in the single-cell and 12-cell 1 kW stack tests. So, although both the metallic plates demonstrated higher resistance values and thus lower performance, the coating improved the plate's performance.

Another example of using carbon-based films in BPPs is the study done by Fu et al.³⁶ In their research, they studied carbon-based films coated on 316L stainless steel (SS) BPPs. SS is an excellent candidate for a BPP material, but its corrosion resistance and ICR are still inadequate for BPP requirements. This means that any coating created must be corrosion resistant and provide high interfacial conductivity in order to even be considered. Three kinds of carbon films were tested in this experiment: C, C–Cr, and C–Cr–N. The pure C film had the highest ICR with the lowest conductivity while the C–Cr–N film had slightly better values but not by a significant difference. The C–Cr film demonstrated the lowest ICR with the highest conductivity, so this specific coating was selected for further testing. Potentiodynamic and potentiostatic tests were used to characterize the corrosion resistance of the plate. These tests found that the coated plate exhibited better corrosion resistance than the pure base metal in all simulated PEMFC environments. One important observation noted during testing was that the coated BPP has a much higher contact angle (91°) with water than that of an uncoated sample (73°). A higher contact angle means that water droplets will not adhere easily to the plate, and therefore decelerate the corrosion of the plate.

Carbon-based films are not the only type of coatings investigated. Zhang et al.³⁷ evaluated the corrosion behavior of TiN-coated stainless steel BPPs. Magnetron sputtering (MS) and pulsed bias arc ion plating (PBAIP) were used to prepare the TiN or multilayer Ti₂N/TiN coating. ICR and voltage drop measurements as well as an electrochemical test were performed to determine the effects of coating. The polarization curves obtained show that the coated samples have higher corrosion potentials and lower corrosion currents, demonstrating that the coatings have a significant effect on the corrosion resistance of the samples. Calculated results from the polarization curves also show that the multilayered samples that were prepared by the MS process have a higher corrosion potential. The ICR measurements were taken at different compaction forces and a clear relationship between contact resistance and compaction force was determined. It was found that the contact resistance decreases with increasing compaction forces; when the pressure increases, the effective contact area is increased. Both coated specimens had similar electrical resistances, and both helped to reduce the ICR of the SS samples. Another important observation made in this study is the relationship between the corrosion behavior of the coatings and their microstructures. Coatings deposited using the MS technique have tightly bonded grains (∼100 nm). After undergoing corrosion testing, one can see that corrosion occurred on grain boundaries and therefore increased the grain boundary width significantly. These corroded grain boundaries led to a higher contact resistance due to the decrease in the effective surface. The PBAIP deposited coatings had larger bonded particles on the surface of the coated film with no trace of grain boundary, and this lack of grain boundary caused the PBAIP sample to be more difficult to corrode compared to the MS coated sample. While the MS technique samples experienced corrosion mainly on grain boundaries, the PBAIP samples experienced corrosion from the large particles on the coating. Ultimately both coatings provided low contact resistance, but the PBAIP Ti₂N/TiN multilayer coating provided the better corrosion protection for stainless steel.

Apart from coatings, investigating the ability of metallic BPPs to produce low-volume stacks and their durability is of interest to research. Lee et al.³⁸ investigated the durability of a 1 kW stack of 64 stainless steel plates. The plates were composed from 304 stainless steel and were coated with a corrosion protective material using physical vapor deposition (PVD). The 1 kW stack volume of 10 L is a 20% volume reduction compared to a graphite stack with the same conditions. Water contact angle was studied as surface energy is an important influence in cell performance. A water contact angle of the coated plate was found to be 62.58° which is a relatively low contact angle. This low contact angle gives the plate a lower surface energy and can lead to a higher chance of flooding of the cathode site. In order to lessen the effects of the water contact angle, the stack was operated at a lower current density. After 500 h of daily startup and shut-down (DSS) under a constant output power of 1.2 kW, the stack degradation rate was interpolated as 27μV h⁻¹ and the durability of the stack to be estimated at 17 000 h. Longer degradation times need to be tested in order to determine the long-term characteristics.

Graphite is the traditional material of choice for BPPs due to its excellent chemical resistance, good thermal stability, and established application in phosphoric acid fuel cells.^39,40 However, its brittleness and porosity make machining flow fields into pure graphite plates an extremely time-consuming and expensive process while also leaving these plates vulnerable to gas permeation.⁴¹ Graphite's brittleness can be attributed to the weak Van der Waals forces between the graphite layers, giving it weak flexural strength and thus increasing the chance of fracture during fabrication.^42–44 Its brittleness requires each plate to be several millimeters thick, causing the fuel cell to also become heavy and large.⁴⁵ These qualities—brittleness, porosity, inadequate mechanical strength—are major issues that prevent the mass production of pure graphite plates. The high cost (especially machined graphite), high mass, and high volume also pushes researchers to seek other materials.⁴⁶ Scientists have investigated different solutions to these issues such as using corrosion-resistant coatings to alleviate its porosity as well as experimentation with graphite composite materials, both of which have been heavily investigated.^47,48

Graphite–polymer composites have been a subject of interest for researchers looking to resolve the disadvantages that graphite presents.⁴⁹ These graphite–polymer composites are advantageous compared to metallic materials due to improved corrosion resistance, lighter weight, and lower cost.^50,51 However, a major drawback of these composites is their low electrical conductivity. The addition of carbon fillers does help improve electrical conductivity, but too high of a filler content leads to difficulties in processing and ultimately decreases the overall mechanical strength.⁵² The polymeric binders used for graphite-composite plates can be divided into two categories: thermoplastics and thermosets.

Several types of commercially available graphite–polymer composites consist of thermoplastics like PP, polyphenylene sulfide (PPS), and polyvinylidene fluoride (PVDF).^53,54

BPP composites using PP, a thermoplastic resin, along with filler reinforcements offer good mechanical properties and easy, cost-effective manufacturing processing.⁵⁵ However, it is important for fillers to be easily and widely dispersed in composites in order to obtain optimal properties, but PP's lack of polar functional groups in its backbone causes the dispersion of graphite in this resin to be more difficult.⁵⁶ This difficulty of dispersion leads to lower electrical conductivity, which is not ideal. However, modification of PP through grafting with polar molecules can be used to solve this issue.⁵⁷ 

PPS is a semi-crystalline thermoplastic polymer resin and an excellent candidate for composite BPPs because of its exceptional chemical resistance, mechanical properties, dimensional stability, and high-temperature resistance.⁵⁸ In addition, PPS is also able to be recycled, a highly attractive quality for BPPs.⁵⁹ Despite its excellent properties, there has been limited research on its corrosion resistance and this still needs to be explored.^60,61

PVDF is a thermoplastic fluoropolymer that has been frequently used with graphite particles and carbon additions to produce BPPs.⁶² These PVDF blends typically have good bulk conductivity but have weak flexural strength.⁶³ 

Examples of thermosetting binders include furan resin, phenolic resin, epoxy resin, melamine resin, and unsaturated polyester.⁶⁴ 

Out of all these thermosetting binders, phenolic resins have been most heavily used for BPP applications. The thermosetting phenolic resin, which exists in a liquid resole form and a powder novolac form, is most suitable for PEMFC BPP applications because of their corrosion resistance and ability to operate at temperatures between 80 and 100 °C.⁶⁵ 

Current investigations show that graphitic materials are still a viable option for use in low-cost BPPs. Scholta et al.⁶² tested a novel graphite composite (SGL 001) BPP created by SGL Technik GmbH and found it comparable to the metal-based reference cell used. Their investigation was concerned with achieving DOE requirements as mentioned previously with their novel type of low-cost BPP. They performed a 120-h test to examine current–voltage curves, gas tightness, and total resistivity. Their experimentation showed acceptable current–voltage curves, sufficient gas tightness, and a reduced total resistivity. Although these results are promising, extended life tests should be performed, and efforts should be made to further reduce the materials resistance.

Hui et al.⁶⁶ studied the electrical and mechanical properties of a graphite–polymer BPP produced using bulk-molding compound process. The graphite–polymer composite plate was created using natural flake graphite and carbon black compounded with low-cost novolac epoxy. They found the following relationships. First, different resin content showed varied conductivity and flexural strength. For a resin content over 15 wt. %, the conductivity significantly decreased while the flexural strength increased over 20 wt. %. The ideal resin content range was found to be between 15 and 20 wt. %. Second, the effect of graphite particle size was investigated. It was found that as the particle size increased and the number of particles decreased, the plate conductivity increased but flexural strength decreased. After that, they examined the effects of molding pressure and curing temperature. An increase in molding pressure led to an increase in the conductivity and flexural strength. The curing temperature only produced an increase in conductivity and flexural strength between 145 and 175 °C. Finally, they studied aging behavior and thermal properties. The graphite–polymer plate experienced minimal aging mass loss after 25 days, and although the flexural strength decreased rapidly at first (decrease in about 4 MPa), there was no additional change after ten days. In terms of thermal properties, it was confirmed through thermo gravimetric analysis (TG) and differential scanning calorimetry (DSC) curves that the composite BPP was suitable for applications with a working temperature of 80–100 °C. Overall, Hui et al.'s research found their combination of natural graphite, carbon black, and novolac epoxy to produce a composite plate suitable for PEMFC. However, it seems that more long-term testing and multi-cell stacking needs to be performed to determine the long-term stability of this composite plate.

Yan et al.¹² evaluated the performance of a PEMFC stack using expanded graphite BPPs. Their expanded graphite BPP was created by first impregnating raw expanded graphite plates with a low viscosity thermosetting resin and then stamping out the final product. A pure graphite, composite graphite plate, and a metal-composite plate were also created along with the expanded graphite plate in order to run three different single-cell tests. 1 and 10 kw PEMFC stacks using expanded graphite were also assembled. Through the single-cell tests, they found that when the pressure between the two BPPs increases to a lower level (F <50 N cm⁻²), the contact resistance between the BPPs and gas diffusion layers (GDLs) decreases rapidly. The expanded graphite BPP exhibited the highest contact resistance, but this did not affect the performance of the fuel cell. The single-cell tests proved that the expanded graphite BPP did not perform as well as the pure graphite plate and that the metal-composite plate performed just as well as the expanded graphite. These single-cell tests are useful for examining the BPPs and its properties, but 1 kW and 10 kW PEMFC stacks are preferred to validate the feasibility of using this kind of BPP for fuel cells. The 1 kW fuel cell stack showed that the output power reached 1.6 kW and the power density exceeded 0.37 W cm⁻². When the PEMFC ran under higher working pressure the power density reached 2.4 kW and the power density exceeded 0.54 W cm⁻². From these data, Yan et al. were able to conclude that expanded graphite works well for this kind of PEMFC cell stack. For the 10 kw cell stack, its output power reached 14.3 kW and has a more uniform cell voltage indicating the stack's ability to run steadily and therefore the suitability of using expanded graphite as a BPP material.

Another example of using expanded graphite in a BPP for PEMFC is in Du et al.'s⁶⁷ research on the preparation and properties of thin epoxy/compressed expanded graphite (CEG) composite BPPs for PEMFCs. They prepared plates by vacuum resin impregnation in compressed expanded graphite (CEG) sheets with a thickness of only 1 mm. From their results, they found that the electrical properties stayed constant, but the mechanical properties and gas impermeability increased with increasing resin content between the range of 4%–30%. Through their single-cell test, it was observed that the power density of the cell continuously increases up to the current density of 1400 mA cm⁻² with a maximum power density of 674 mW cm⁻². This single cell's excellent performance demonstrates that using resin vacuum impregnation with CEG sheets are suitable for PEMFC BPPs.

Just as different materials must be considered for various fuel cell applications, different manufacturing methods are more suitable for different materials. Different manufacturing processes for metallic, graphitic, and composite plates can be seen in Fig. 3.

As shown in Fig. 3, there are several different manufacturing techniques capable of forming metallic BPPs. Traditionally, the three main types of manufacturing that are considered when fabricating metallic BPPs are powder forging metallurgy, stamping, and hydroforming.

Powder forging metallurgy is defined as a process in which a nonsintered, presintered, or sintered powder metal preform is forged in a confined or trapped die.⁶⁸ Typically, a metal powder is fed into a powder hopper, then pushed into a preform press, and compacted into a green body. It is very important that the density of this green body is relatively high as this will help prevent shrinkage and distortion during sintering. This results in better component properties, as well as allowing the desired sintered density to be achieved at a lower temperature.⁶⁹ The green body is extremely weak and will fall apart with minimal applied force. After the part is sent through a furnace the powder particles are bound together and create a mechanically stable component. Sintering does not always guarantee a fully dense part as micropores tend to always be present and they often to act as initiation sites for cracks during service.⁷⁰ In order to help eliminate some of these micropores, post-processing treatment such as sanding, forging, extrusion, shot peening, and nitriding can be applied. Powder forging metallurgy is an attractive BPP manufacturing process for several reasons. First, it allows the user to be able to use a variety of metal and nonmetal powders. Materials that are difficult to process such as W, Mo, and Be can also be used with this manufacturing process to create components.⁷¹ This option greatly increases the user's ability to experiment with different BPP materials due to its convenience and wide range of available powders. It also allows for the conservation of material. Unlike traditional subtractive methods in which a block of metal is removed in order to create the final component, powder forging metallurgy only uses the amount of powder necessary to form the component, therefore reducing waste and costs. Although powder forging has been heavily researched and has many benefits, using this technique requires a large investment and is most suitable for mass production (upwards of millions) of components.⁷² 

Stamping is a manufacturing technique that has been used in a wide variety of industries such as aerospace, electronics, machine tools, automobiles, and refrigeration.⁷³ A specific form of stamping known as progressive die stamping has been acknowledged as one of the most appropriate methods to mass produce BPPs.⁷⁴ Progressive die stamping involves moving a continuous flat sheet of metal down an assembly line in which the metal sheet is stamped at multiple stations. Each station involves a unique die to form a different element of the product as seen in Fig. 4.⁷⁵ Not only is this method a more affordable practice for mass production, but the ability to stamp plates with thicknesses as low as 0.051 mm makes this process an attractive choice compared to its heavier graphitic and machined metallic plate counterparts.⁶ Its cost-effectiveness and high controllability of sheet metal dimensions also make this technique highly attractive for BPP mass production.⁷⁶ Stamping is a relatively fast process, but forming defects like cracks, thinning, warping, and wrinkles can happen.⁷⁷ Mass-produced BPPs fabricated through the stamping process can also experience spring-back which can result in channels that are deformed and irregular.^78–80 Spring-back is a phenomenon that occurs when sheet metal is stamped and then elastically relaxes back toward its original shape, often resulting in an unacceptable loss of dimensional control.⁸¹ Not being able to closely meet channel dimensions will lead to nonuniform gas distribution and ultimately cause low efficiency, low power output, and poor gas utilization.⁸² 

Hydroforming is a nontraditional sheet-metal forming process that uses pressurized liquid to produce components rather than a conventional punch as seen in stamping.^83,84 Hydroforming can be divided into, tube hydroforming and sheet hydroforming, with BPPs falling into the latter type of manufacturing.⁸³ Sheets of metals are placed between an upper female die and a lower male die, then a pressurized forming chamber is pushed into a cavity between the metal sheets, and finally the pressure is released to reveal the final product.⁸⁵ This process, shown in Fig. 5, is highly advantageous over conventional sheet forming processes (deep drawing, stretch forming, spinning, roll forming, etc.) as it demonstrates flexibility, better surface quality, less spring-back, ability to form complex sheet metal parts, shorter production cycle, and lower cost.^86,87 Sheet hydroforming is also capable of turning a process that may require multiple stages to produce a component, as seen in stamping, reducing it to a single-stage process.⁸⁶ Research done by Mahabunphachai et al.,^88,89 Peng et al.,⁸⁵ and Belali-Owsia et al.⁹⁰ have also shown the repeatability and controllability of the hydroforming process for producing channels on stainless steel plates. Although significant progress has been made in advancing sheet hydroforming technology, it is still slower than that of tube hydroforming due to numerous process difficulties and industry problems. Problems that need to be solved include increasing process efficiency, improving the capacity of local deformation, reducing press costs, and improving the automation of the equipment.^86,91,92

There are several different methods of producing metallic BPPs but stamping and hydroforming are considered the most suitable manufacturing methods to mass produce metallic BPPs out of all previously mentioned methods because of their low-cost, high-production rate capabilities.⁹³ However, the question of which manufacturing technique is superior to the other is often asked. In a 2017 BPP cost analysis done by Strategic Analysis,^93,94 they evaluated an 80 KW_net PEMFC stack using the Design for Manufacture and Assembly (DFMA) costing technique to determine that hydroforming has cost advantages over progressive die stamping. Progressive die stamping requires multiple different stamping stations in order to finish a single BPP. So, in order to meet high-volume demands, there must be several parallel production lines all working at the same time which increases equipment and tooling costs and therefore increases the BPP cost.⁹⁵ In fact, Strategic Analysis reports that for progressive die stamping to meet a rate of 500 000 vehicles per year, 111 parallel production lines running all day, night, and year are required. Hydroforming on the other hand does experience an increase in cycle time when asked to form four plates simultaneously rather than only two plates, but the cost of forming four plates does not exceed the cost of forming two plates.⁹³ Strategic analysis reports that hydroforming would only require 72 parallel production lines running all day, night, and year. The main disadvantage of this technique is that extra stamping steps are needed to cut the manifold holes and separation of parts.^93,94

As mentioned previously, graphite possesses excellent chemical resistance properties, great electrical conductivity, high strength, thermal stability, and corrosion resistance.^39,45 However, pure graphitic plates must be thicker to compensate for graphite's innate brittleness, leading the plate to become heavy and voluminous as well as expensive due to the material's high cost.^58,96,97 These pure graphitic plates also experience difficulties when using simple machining to produce flow fields which is an expensive, time-consuming process not suitable for mass production.^98–100 Pure graphite is not a suitable material for BPP production, but the desired properties of graphite can still be utilized in conjunction with other additives such as polymer resins or carbon fillers to fabricate graphite-composite plates.^11,101 With these alternative BPP materials, different manufacturing methods must be used.

Compression molding is one of the main processes that is used to manufacture a high volume of composite components (Fig. 6).¹⁰² This technique can be separated into four stages: precharge preparation and placement, mold closure, curing, and part release.¹⁰³ 

The first stage of precharge preparation and placement involves feeding the preheated molding material, known as a precharge or charge, into a hot mold.¹⁰⁴ Two important factors to consider in this stage are the precharge dimension and position in the mold. The precharge should initially cover about 50% of the mold surface area and the positioning of the precharge must be carefully considered as an inappropriate positioning will correlate with poor part quality.

The second stage involves the mold closure. As soon as the charge is placed into the mold, the movable upper half of the mold will quickly move down to touch the top of the charge. Once it touches the top of the charge, it will slowly (5–10 mm s⁻¹) compress the charge and the molding pressure will begin to increase. This increase in mold pressure will cause the material to fill the mold and cause the air to escape through vents or edges in the mold. Mold closing speed and mold temperature are both important factors that contribute to product quality and process performance.

The third stage is the curing stage. In this stage, the charge has filled the mold cavity and the mold is kept closed until the resin/binder is cured. Curing time depends on the resin/binder used, part thickness, and mold temperature. If a thermo-set binder is used, several minutes are required for it to be sufficiently cross-linked whereas thermoplastic binders need only to be cooled below its melting temperature before the plate can be removed.⁸² 

The final stage is part release. Once the part is solidified, it is removed from the mold using the ejector pins. When the part is removed, it is cooled completely, the mold cavity is cleaned, and a mold release agent is applied for the next molding cycle.

Compression molding is an excellent manufacturing method because it creates components that do not require extra machining steps, thus reducing any scrap material, costs, and stresses seen in machined parts.¹⁰⁵ This method is a relatively slow process because there is a waiting time for both the resin/binder to cure and the mold cavity to cool before the part can be removed. Despite being a slower manufacturing method, compression molding remains the best technique for manufacturing plates that are loaded with a high volume of carbon fillers (>50 vol. %).¹⁰⁶ Other manufacturing methods like injection molding have a more difficult time processing plates with such high carbon concentrations.

Injection molding is commonly used to produce graphite-composite plates because of its potential for low-cost mass production.¹⁰⁷ This manufacturing method is composed of three main stages: melting plastic, plastic solidification, and ejecting the solidified component.¹⁰⁸ 

In the first stage, as shown in Fig. 7, powdered materials are fed into a hopper and moved into the barrel where they are melted by heaters in the press. In graphite–polymer composites, graphite powder, polymer binder, and other additives are continuously compounded and granulated in a twin-screw extruder.¹⁰⁷ 

During the second stage, the ram and screw on the press then work together in order to compress and inject the resin through a nozzle to the mold cavity where it will solidify to the shape of the cavity.¹⁰⁹ 

Finally, once the mold is filled and the part has solidified, pins will eject the part, and the mold is closed to begin the next cycle.¹¹⁰ 

The injection molding process is extremely useful as it permits simple, complex, small, and large shapes to be quickly and automatically manufactured.¹¹¹ Products can be molded to very specific tolerances, thicknesses, and weights, giving the user very few constraints.¹¹² Additionally, it has been observed that as the number of plates produced increase, the cost of material and processing drop significantly, making this manufacturing method especially attractive.¹¹³ However, injection molding can suffer from excessive mold wear, limited size to thickness ratio and poor thermal and electrical conductivity.^114,115 Injection molding also requires a sizable upfront investment, making it expensive if mass-production is not required.

Both compression and injection molding are useful and promising candidates to produce graphite-polymer BPPs. However, little research has been published on the effects of these manufacturing methods on the plates themselves. BPP flow fields have many options for channel patterns and dimension,¹⁰⁵ so a different mold for compression and injection molding must be fabricated for different designs which is expensive and time-consuming.¹¹⁶ Therefore, research and development on BPPs for both compression and injection molding is limited.

One study by Derieth et al. focused on examining the effects of graphite morphology as well as compression or injection molding on electrical conductivity and on the internal structure of the BPP.¹⁴ The plates were created using a combination of graphite flakes, a PP binder, and carbon black through either the compression or injection molding process. First, they determined the graphite material that is most suitable for creating BPPs. An investigation between spherical and different sized graphite flakes determined that an increase in size of particles leads to an increase in density and therefore improvement in rate of production. On the other hand, a decrease in size of particles improved the electrical conductivity immensely while increasing the difficulty of the process. Keeping mass manufacturing requirements in mind, a screening process with different combinations of graphite materials and carbon blacks was initiated. From that screening process, they determined that a flake-like graphite (5 μm) was the most suitable.

In terms of manufacturing processes, compression molding typically uses a higher processable filling load of graphite than injection molding which in turn leads to a higher bulk conductivity of the basis compound. This process leads to graphite particles being oriented in a plane perpendicular to the direction of compaction force during molding and therefore perpendicular to the current flow. As a result, the through-plane conductivity is much worse than the in-plane conductivity. Injection-molded plates, on the other hand, offer a different structure compared to compression molded plates. This manufacturing technique requires a degree of lower viscosity which leads to a lower bulk conductivity but can also allow a less filled compound with a better conductivity. At the surface of the plate, the graphite flakes were perpendicular to the current flow which is similar to compression molded plates. However, the flakes in the core of the injection-molded BPPs are aligned in a direction that is similar to the direction of current flow which will improve both the in plane and through plane conductivity. Derieth et al. concluded that both compression and injection molding produce plates with good electrical conductivity, and that the main influence in production rate of the plates depended on the morphology of the graphite material.

The molding processes revealed that plates produced by injection molding have a similar or better conductivity compared to those produced by compression molding.

AMis a nontraditional manufacturing process that fabricates geometrically complex components, which are not feasible with machining procedures, in a layer-by-layer fashion using a laser or electron beam and powdered material, as shown in Fig. 5.¹¹⁷ Although AM techniques have existed for decades, it was initially restricted to prototyping and rapid manufacturing of porous parts.¹¹⁸ However, the advancement of AM technology provided AM techniques the ability to freely create complex shapes, to make customizable parts, to reduce material waste, and to have shorter lead times.¹¹⁹ AM technologies also provide the flexibility to change fuel cell BPP designs without needing to change the entire setup as would be required with conventional manufacturing methods.¹¹⁷ 

SLS is a novel AM method that has been determined to be one of the most suitable fabrication techniques for graphite–polymer composite plates, and it can also be used to produce metallic plates (Fig. 8).¹²⁰ During the SLS process shown in Fig. 5, a roller spreads an even layer of powder particles across the powder bed. Then a high-power laser selectively raster scans the layer of particles to sinter that single layer of the component's geometry. Once this layer has been successfully sintered, the powder bed will lower itself for the roller to push another layer of powder across the bed and start the process over again until the component is finished. Some of the biggest advantages of this technique are its ability to accurately build complex flow channel patterns as well as to conserve time and materials.¹²¹ This solid freeform fabrication technique can be further classified into two groups: direct SLS and indirect SLS.

Direct SLS produces parts directly from the SLS process without the use of a polymer binder and without the addition of any post-processing steps.¹²² 

Chen et al.¹²³ investigated use of direct SLS for fabrication of BPPs. They concluded that this AM technique is not practical for the fabrication of metallic and pure graphitic BPPs. The melting point of graphite is above 3000 °C, which prompts the need for a high laser power for the heating source. However, this is not possible with current commercial SLS machines as their highest power output is 50 W.¹²⁴ An even greater challenge is the use of metallic materials in direct SLS for fuel cell BPPs. Balling phenomenon is a common issue in the direct SLS of metal,^125,126 and Chen found that it imparted a rough surface finish on the BPP, leading to geometric inaccuracy and an inability to meet the DOE's technical requirements.

Other studies about direct SLS involved improving the use of metal or combination of different metal powders with this technique. A study done by Murali et al.¹²⁷ researched the effects of direct SLS on an iron–graphite powder mixture. When comparing the direct SLS material to the conventional powder metal sintered material, they found that direct SLS still created parts with porosity, but its constant re-melting of the powdered material helped to reduce the porosity. The direct SLS material also showed higher hardness values and the potential for good wear characteristics. Agarwala et al.¹²⁶ extensively reviewed the use of direct SLS to fabricate metal parts. Previous experiments using direct SLS have been made using Cu–Sn, Cu–solder, and Ni–Sn but this review found that the best results came from metal powder blends with constituents of different melting behavior. Having partial liquid and solid phases during direct SLS led to smooth and strongly sintered layers with minimum balling. Although there are certainly ways to improve this technique, current technologies do not support the use of this method for the fabrication of PEMFC BPPs.

Indirect SLS produces a porous green part held by a polymer binder that is later removed through various post-processing steps; unlike direct SLS it utilizes both polymer binders and post-processing techniques. Another study done by Chen et al. used this method in their research to fabricate a graphite–polymer composite BPP because of its ability to use a carbon-based composite material.¹²⁴ A combination of graphite particulates (GrafTech GS150E) and phenolic resin (GP5546) was used to fabricate the plates and post-processing steps including binder carbonization, epoxy infiltration, and heat treatment were performed. They measured physical properties such as flexural strength, electrical conductivity, corrosion rate, and gas permeability. The average flexural strength was 1730 psi and the load/strain curves showed that the SLS-fabricated plate is ductile and has a yield strength that matches its ultimate strength. The plate's electrical conductivity was also determined to be 80 S cm⁻¹, which is competitive in comparison to other carbon–polymer composite BPPs.¹²⁸ When looking at corrosion resistance, the Tafel extrapolation method was selected because it is a quick and simple technique, and a corrosion current density of 6 μA cm⁻² was found. This corrosion current density is lower in comparison to metallic BPPs but makes sense because of the graphite used to fabricate the plate. Finally, the gas permeability was tested, and an average leak rate measured was 5 × 10⁻⁶ cm³ cm⁻² s⁻¹. This is a promising result and can be improved using multiple epoxy infiltrations. Overall results from the testing of an SLS-fabricated graphite–polymer BPP were satisfactory and met DOE requirements.

Another example of using indirect laser sintering is in Bourell et al.'s research.¹²⁹ Their group's goal was to research how the addition of carbon fibers in a laser sintered (LS) graphite BPP affects the parts final electrical conductivity and strength. The plate was created by laser sintering a mixture of graphite powder, fine phenolic powder transient binder, and 0–26 vol. % carbon fibers to create a porous green part. The green part was then heated so the phenolic binder would be converted to carbon. This resulted in the final component known as the brown part (a part subjected to thermal treatment during post-processing steps), which is then infiltrated with epoxy or cyanoacrylate to create a gas-impervious plate. Bourell et al. found that increasing the carbon fiber percentage gave both the green and brown parts a higher flexural strength (increase in 2 MPa for straight carbon–phenolic mixtures and 4 MPa for 26% carbon fiber addition). They also found that at a dissociation temperature of 1200 °C, a carbon fiber addition of 26% allowed the brown part strength to go from 3 MPa to 10 MPa. The carbon fibers increased the strength of the final parts from 35 MPa to about 50 MPa.

In terms of electrical conductivity, the phenolic dissociation temperature played a significant role in electrical conductivity but too much of the carbon fiber negatively affected the electrical conductivity. Using a 600 °C-dissociation temperature led to low electrical conductivity, but at dissociation temperatures at 800 °C, a sharp rise in electrical conductivity was observed (rose to 70 S cm⁻¹), and was independent of carbon fiber content. For dissociation temperatures above 800 °C, the carbon fiber additions had a negative effect on electrical conductivity. It was further observed that any increase in carbon fiber volume percentage over 5% would lead to a significant decrease in electrical conductivity (400 S cm⁻¹ to 50 S cm⁻¹).

Ultimately, an increase in volume percentage up to 26% of chopped carbon fibers also led to an increase in flexural strength for both green and brown parts but decreased electrical conductivity of finished parts. Additionally, lower carbon fiber additions are better for achieving an electrical conductivity that satisfies DOE requirements.

Another AM technique that has been researched for the fabrication of metallic BPPs is direct metal laser sintering (DMLS).

DMLS is a novel AM process that is used to create a component by depositing metallic powders layer by layer.¹³⁰ Just like SLS, a laser beam is used to sinter the powders to the geometry for that specific layer. The platform is then lowered, another layer of metallic powder is deposited, and the process repeats again. This manufacturing method has been applied to a wide range of applications in the automobile, food technology, aerospace, and medical industry due to its ability to create functional complex parts.¹³¹ However, its main disadvantages are its long build time (6–12 h) for even small pieces and to maintain its part accuracy; reducing the build time is important for commercialization.¹³² For DMLS to be commercially viable and accepted, it is essential for this process to meet all the required part parameters at economical costs and practical fabrication rates.¹³³ 

Lyons and Gould¹³⁴ investigated the Naval Research Laboratory's (NRL's) use of DMLS in printing titanium alloy (Ti-6Al-4V) BPPs. The NRL designed a 4 cm × 8 cm × 3 mm-thick BPP with a 22 cm² active area and a serpentine flow-field. To prevent corrosion resistance, a thin layer of Au nanoparticles was deposited onto a TiO₂ layer which was then deposited onto a BPP through the thermal spray process.¹³⁵ An important observation made was the rough surface which contributed to a high interfacial resistance between the DMLS-printed plates.¹³⁶ Surface features up to 20 μm with voids and an overall surface roughness of 7.6 μm were clear contributors to the high interfacial resistance. In order to mediate this issue, the plates were polished, and the interfacial resistance decreased significantly.

During assembly of stacks, another issue that was found was warping. Warping is a result of residual strains caused by the constant re-melting of powdered layers in DMLS. This warping not only made assembly difficult, but also resulted in leaks, poor electrical contact, and poor stack performance.¹³⁷ 

AM technology is a hot topic because of its promise for rapid prototyping, but there is still very limited research on using AM techniques such as SLS and DMLS to fabricate PEMFC BPPs. To develop these processes on a commercial basis, more research needs to be conducted on the repeatability and consistency of manufactured parts, improvement on AM part surface finish, and advances in novel processes and technology.¹³⁸ 

For PEMFCs to become commercially accepted, cost is among the most important things to consider and reduce. According to the Fuel Cell Technical Team Roadmap report, current BPP costs are $5.4/kW which is much higher than the $2/kW expected by the 2025 DOE technical targets.¹³⁹ Aside from material choice, choosing the proper manufacturing process is an important method to reduce plate cost. This section will provide a cost analysis of both traditional and AM processes.

Traditional manufacturing processes like stamping, hydroforming, injection molding, and compression molding are all well-established techniques that have been used for decades. However, as our society progresses and new technologies become available, it is important to continually re-evaluate traditional techniques to determine if they are competitive options, especially in terms of cost.

Lovell et al.¹⁴⁰ estimated that in the U.S. over $100 × 10⁹ is spent each year on the design, fabrication, and assembly of stamped parts. Stamping is a highly energy-intensive process due to the employment of casting and machining to create the die parts used for stamping.¹⁴¹ Steel and aluminum are common materials used to create the die parts, but this type of metal production accounts for over 10% of global carbon dioxide emissions.¹⁴² 

For production of any size, custom die molds are needed. It can take up to 10 weeks to manufacture a matched die-set¹⁴² and requires a large up-front investment. Costs of die-sets for automobile body parts typically range between $39 000 and $53 500.¹⁴¹ In addition to die mold and material costs, labor costs are also important to consider. Cooper et al.¹⁴¹ interviewed a range of car and die-makers to find that the average labor cost for stamping is $65/h including overhead costs while equipment depreciation is equal to $42/h assuming a utilization period of 4000 h a year.

Huya-Kouadio et al.⁹⁵ recently analyzed the use of large-scale manufacturing of metallic BPPs using the stamping method and identified three major challenges.

First, it is very difficult to manufacture small and detailed BPP flow channels. This kind of flow channel requires a greater stamping tonnage and ultimately increases the machine capital cost and the stamping cycle time by a considerable amount, $1000/ton and 2.5 s, respectively.⁹⁵ To achieve a high accuracy and aspect ratio, multiple stamping cycles are also needed, leading to high equipment cost and a long manufacturing time.

Second, the die sets for BPPs are extremely expensive and require a considerable amount of time to fabricate. Huya-Kouadio et al.⁹⁵ report that it can take several thousand hours and $600 000 per die set.

Finally, the mass production process of BPPs experiences a bottleneck issue. Each stamping line is only able to produce 5100 fuel cell vehicles with 3360 h of operation time and costs $6.50/kW which is well over twice of the DOE target.

Although the stamping process is considered one of the more promising mass production techniques for metallic BPPs, there are still major cost improvements that need to be made for mass application of PEMFCs to be possible.

Whereas it is more difficult for stamping technology to produce the small, complex flow channels that are typically required of BPPs, the hydroforming process can remedy the limitations stamping presents.

Sheet hydroforming is commonly used among automotive manufacturers in producing car panels such as fenders, liftgates, front bumpers, etc., due to its high degree of formability and lower tooling cost.¹⁴³ Unlike machining and stamping processes, hydroforming can reduce material scraps and tooling costs due to its ability to create the finished product in one or fewer forming operations. Cost modeling also shows that sheet hydroforming has a smaller initial investment cost especially when multiple low-volume stamping dies are needed.¹⁴⁴ 

Although hydroforming may have lower tooling costs, this process requires presses of a higher capacity,¹⁴⁵ causing equipment to cost up to 30% more than stamping tools.¹⁴⁶ According to Golle et al.¹⁴⁷ this higher press cost might not be offset by the lower tool cost depending on part geometry, production volume, and part mix; if the part is large, more complex, or has a high material cost the product cost will be significantly larger. Another important thing to note when using sheet hydroforming for BPP production is that parts with sharp corners require a press with a very high capacity and results in an increase in the investment cost.¹⁴⁸ BPPs traditionally are in a planar form and typically have relatively sharp corners.

Sheet hydroforming is often compared to the stamping process and especially in the realm of BPPs. While hydroforming offers solutions for some of the prohibitions that stamping has, hydroforming is not definitively more cost-effective than stamping.

A study by Minke et al.¹⁴⁹ shows that compression molding has the potential to produce 36 000 BPPs annually and requires a $180 000 investment for the machine and a $96 000 investment for tooling costs, about 300% less and 20% less, respectively, than that required for injection molding.

Compression molding presents a much smaller investment cost compared to injection molding but suffers from high cycle-times and limited parts per batch.⁹⁵ Cooling times depend heavily on the wall thickness of the parts and can take up to several hours for large, thick parts which is undesirable.¹⁵⁰ These longer cycle/cooling times are in exchange for the significantly cheaper machine and tooling costs. A modest labor cost is considered with compression molding, but it will add very little to the variable cost of each part.

A consensus is that compression molding is typically a good choice for very simple plastic parts and can be a much more affordable option vs injection molding depending on the type of part one is intending to produce. However, due to its high cycle times and difficulty in producing more complex patterns it may not be the most cost-effective choice for BPP production.

Injection molding is admired for its high production rates, long mold life, as well as its excellent quality and accuracy of parts.¹⁵¹ According to Franchetti et al.,¹⁵² a mold for injection molding may take 2–6 days, which is a much shorter time compared to the previously mentioned manufacturing methods, and machine setup may take well over 10 h with a $20 an hour labor cost.

It is very important to consider part complexity, shape, and size when determining product and equipment operating cost for injection molding. Relatively simple shapes and designs only require a single mold set and do not require any extra special attention. However, larger, or more complex shapes may require several sets of molds to perform necessary operations and more skilled operation, resulting in higher product, tooling, and labor costs. The geometrical features of the product will also significantly affect manufacturing costs. Designs with thin walls typically lead to additional mold costs in the injection molding process.¹⁵¹ A typical price for molds can range from $3000 to $10 000.¹⁵² 

Production amount also plays a significant role in the cost of the product. When implemented in BPP production, it was found that as the number of produced plates increased, the material costs and processing costs dropped by a significant amount.¹¹³ Injection molding is much more suitable for mass production quantities rather than individual runs.

AM has created a multitude of opportunities and is a powerful prototyping tool for many industries. 80%–90% of life-cycle design costs are determined in the first 10%–20% of the design phase, making it extremely important to reduce and control the product's cost early on.¹⁵³ The traditional manufacturing methods presented in this paper require extensive research, time, and funds to even prepare a mold that will meet cost and performance expectations. AM overcomes these limitations by eliminating the need for a mold and providing its users the resources to create innovative and unique products specifically tailored to its users' needs in a reasonable amount of time.

A major contributor to AM's significant growth over the past decade is the dedication of companies such as MakerBot Inc., Ultimaker Inc., Stratasys, and 3D Systems to lower the cost barrier for consumers to produce AM parts.¹⁵⁴ In 2015, there was $4.2 × 10⁹ spent on AM in the U.S. compared to an approximate $800 × 10⁶ spent on AM in 2005 because of an increase in accessibility of AM machines.^155–157 

In 2011, it was discovered that there was $537 × 10⁹ worth of inventories in the manufacturing industry (about 10% of that year's revenue).¹³⁹ Companies often struggle with having high inventory costs and too much product sitting in warehouses, occupying space, and deteriorating. An excellent benefit of AM is its ability to rapidly produce parts on demand and on-site. This ability would not only reduce inventory amount and its related costs, but also allow companies to focus its resources elsewhere. It is also important to note that traditional manufacturing could require multiple parts to be made at multiple different locations and then shipped to a facility where they are assembled. AM eliminates the need for different manufacturing locations, transportation, and extra labor costs.

As mentioned previously, AM also gives the user more design flexibility and freedom with creating their components. Researchers are constantly searching for new ways to improve BPPs, whether it be through the material selection, channel design, or even the plate architecture. AM is an excellent manufacturing tool as it can produce custom plates for little to no expense. It is much more expensive to experiment with plate designs using traditional manufacturing methods as it is costly. Not only this, but also for traditionally manufactured methods, there is typically a long setup time associated with it. Molds for those methods can take several weeks and require specialized labor to create whereas AM machines have virtually zero tooling and minimal labor.¹⁵² 

A study done by Atzeni et al.¹⁵⁷ compared the cost of using injection molding and SLS to manufacture a lamp holder and found that SLS was adequate for medium lot productions (up to 87 000 pieces) even for mass customization products. The major cost for injection molding was made up by the mold (84.6%–97.7% of the total cost). For SLS, the machine cost per part and material cost per part made up between 58.7% and 65.9% and 29.1%–30.4%, respectively. Ultimately, it was proven that injection molding is much more appropriate for production of more than 87 000 parts (large batch production) and AM is more suitable for small batch production. While AM machines only require a small initial investment, the cost of the product increases proportionately with number of parts produced. On the other hand, injection molding requires a substantial upfront investment but sees a decrease in product cost with an increase in number of parts produced. Despite AM not being a viable mass production method at the present moment, this does not change its unmatched prototyping ability to customize parts without needing to change the entire setup/create another costly mold and to reduce the transportation costs associated with traditional manufacturing methods which is highly attractive.

AM is a manufacturing method that should be considered in the conversation of BPP production. Although it is not as effective for large-scale production and has a higher per unit cost compared to other traditional manufacturing methods such as injection molding, AM is highly useful for quickly producing parts with complex designs that would be more expensive to manufacture with traditional methods. It would be an especially useful prototyping tool as users are able to experiment with different materials and designs, but more research needs to be done comparing additively manufactured plates to traditionally manufactured ones.

Due to its high efficiency, low-temperature, high-power density, and near-zero harmful emission operation, the PEMFC offers a powerful alternative to current nonrenewable energy sources. BPPS play a significant role in PEMFCs and make up almost 80% of its weight, 50%–65% of its volume, and 40% of the total stack cost, posing a barrier for commercialization.^6,7 The ability to reduce this key component's material cost, reduce its manufacturing cost, and improve its longevity will increase the PEMFC's chance of becoming commercially available. This paper has attempted to review commonly used materials and manufacturing methods for BPPs in PEMFCs as well as recent research conducted to improve both material selection and manufacturing techniques. In summary:

1. Pure graphite BPPs have excellent chemical resistance but lack strong mechanical properties and are also an expensive material option. More affordable graphite–polymer composites offer improved mechanical properties and manufacturability over pure graphite plates, but still face difficulties in reaching adequate electrical conductivity.

2. Metallic BPPs offer high strength, electrical conductivity, easy formability, and well-known manufacturing processes. However, they are highly susceptible to corrosion, which negatively affects their performance. This can be offset by using corrosion-preventing coatings, but more research needs to be conducted on different types of coatings used, their effectiveness, and methods of coating application.

3. Stamping and hydroforming are both feasible metal BPP forming techniques for mass production. However, both exhibit forming defects and require a significant investment for mass production lines.

4. Injection and compression molding can produce sufficiently electrically conductive graphite–polymer BPPs but require a sizable investment and are prone to mold wear over time. Also, the morphology of graphite material can significantly influence the production rate of the plates.

5. AM technology can lower the costs and increase production rates. Current research demonstrates that PEMFC BPPs can be additively manufactured using both SLS and DMLS; however, a lot more research still needs to be conducted in order to better understand which AM technique is most suitable for BPP production. AM is a worthwhile choice for BPP production.

Future work includes focusing on advanced protective coatings, novel metallic or graphitic–polymer materials, and future manufacturing processes. Another research suggestion includes investigating the use of recycled materials, such as plastic or scrap metals, for BPPs with the objective of furthering its sustainability.

All authors contributed equally to this work.

Financial support was given through the Advanced Manufacturing for Energy Systems (AMES) fellowship at the University of Connecticut, funded by the U.S. Department of Energy Advanced Manufacturing Office traineeship program, Grant No. DE-EE0008302. This initiative aims to address workforce training needs in the early stage technology area of advanced materials and process technologies in energy-related manufacturing.

Data sharing is not applicable to this article as no new data were created or analyzed in this study.