Moving Beyond “Black Paper”

Dr Jason Morgan, AvCarb, USA, outlines the steps for how a properly designed gas diffusion layer can remove bottlenecks and help keep up with a changing industry.

Polymer electrolyte membrane fuel cells (PEMFCs) have emerged as a cornerstone technology in the global transition towards decarbonisation, offering high power density and efficiency for a variety of clean energy applications. As the PEMFC industry accelerates its transition from stationary and material handling applications to high-power, heavy-duty transport, and distributed power systems, the engineering focus has shifted towards a well-documented challenge. Consequential system failures frequently associated with the catalyst layer are now starting to appear in the porous transport layers and interfacial regions that surround it. At the centre of this challenge is the Gas Diffusion Layer (GDL), a component once treated as a commodity, is now recognised as a critical determinant of stack performance and lifetime.

As OEMs push the boundaries of technology via thinner membrane electrode assemblies (MEAs), higher current densities, and extended target lifetimes, the limitations of conventional GDL designs have become well-documented performance bottlenecks, specifically in compressibility, electrical resistance, mechanical durability, and surface topology. This realisation has driven a shift toward greater material engineering rigour, culminating in the development of the FLEX-GDLTM. By examining the specific mechanical and electrical failure modes of conventional GDLs in modern fuel cells, this article will outline the engineering approach to overcoming these failures, ultimately illustrating why this release represents the next evolution in fuel cell material science.

The failure cascade: Why the GDL matters

Within a fuel cell the GDL must simultaneously manage reactant gas transport, water removal, electrical conduction, mechanical compliance, and heat dissipation. The failure of any single function creates a cascade effect that reduces the stack performance. For instance, a piece of GDL with high compressibility will easily conform around the plate during the stack build to help create the seal necessary for the fuel cell to work. However, if the GDL thickness is not uniform, both down-web and cross-web, then when compressed to a hard stop the pressure applied to individual GDL pieces may be different. This difference will impact the porosity of the GDL and thus the diffusion rate, perhaps creating additional mass transport loss or ‘low cells’ (individual cells in a stack that underperform due to localised transport restrictions). Even worse, if the GDL compressibility is inelastic then, as the membrane in the MEA swells and contracts against these thicker pieces during operation, the GDL thickness can creep lower, eventually becoming too thin to maintain the seal. This will cause a leak that compromises the entire stack for the thickness variation of a single piece of GDL. While the impact of GDL variations can be localised initially, these failures propagate stress to adjacent areas, reducing efficiency and ultimately compromising the performance of the entire stack. To mitigate these types of risks, the FLEX-GDL was engineered to address four critical failure points inherent in traditional designs.

Step one: Resolving the compressibility paradox

Conventional GDLs were designed with high compressibility to accommodate manufacturing tolerances in bipolar plates and other cell components. While this aided sealing, it introduced ‘silent efficiency killers’ such as channel intrusion (increasing pressure drop), permanent compression set (creep), and inconsistent contact pressure (increased resistance). Attempts to stiffen the material by increasing fibre density work, but often resulted in higher costs, reduced permeability, and increased risk of membrane damage. These solutions also limit the reduction in GDL thickness, preventing improvement in power density by reduction of cell pitch.

The engineering response

Technologies like FLEX-GDL reimagine the binder distribution strategy. Rather than relying on dense inter-fibre webbing, the binder reinforces individual carbon fibre junctions, thereby preserving open pore pathways while building load-bearing capacity. The result is a non-linear compression profile as seen in Figure 1. The material compresses readily below 1 MPa to facilitate sealing, but exhibits near-elastic behaviour above 1 MPa, with less than 15 μm of thickness change between 1 and 2 MPa and a permanent compression set of only 10 μm after cycling to 3 MPa. This dual-phase behaviour prevents channel intrusion and creep, ensuring consistent interfacial contact throughout the stack’s operational life.

Figure 1.

Through-plane thickness as a function of compressive stress on FLEX-GDLTM. The material exhibits a dual-phase compression response with easy compression below 1 MPa for easy sealing and less compression at higher forces to prevent channel intrusion.

Step two: Minimising Ohmic losses via advanced hydrophobicity

Hydrophobic agents, typically polyterafluoroethylene (PTFE), are essential for water management and flood prevention. However, these agents act as electrical insulators. In standard GDLs, achieving the necessary hydrophobicity (contact angle - 140°) often requires PTFE loadings that significantly increase Ohmic resistance, forcing operators to choose between water management efficiency and electrical conductivity.

The engineering response

The new GDL design utilises a conformal coating process in which PTFE is deposited directly onto individual fibre surfaces rather than pooling at fibre junctions or bridging pore throats. This approach achieves equivalent hydrophobicity (contact angle ≥ 140°) with an 80% reduction in total PTFE content compared to conventional GDLs. The practical consequence is measurable: through-plane electrical resistivity is reduced from 7 mΩ·cm2 to 5 mΩ·cm2 at 1 MPa (a 30% reduction vs the prior-generation baseline) while water management performance is preserved. The result is a simultaneous improvement in electrical efficiency and material sustainability, without the traditional trade-off between hydrophobicity and conductivity.

Figure 2.

Mechanical strength comparison between MB-39 and FLEX-GDL.

Table 1.

Physical property comparison between AvCarb MB-39 and AvCarb FLEX-GDL

Figure 3.

Confocal laser microscope images of the FLEX-GDL MPL surface at nominal pressure (left) and at 2 MPa compressive load (right). Surface morphology is visually unchanged and measured average surface roughness (Ra) are identical at 2.4 and 2.35 μm respectively.

Figure 4.

Three-dimensional CT reconstruction showing the layered architecture of FLEX-GDL. The smooth, continuous MPL surface (upper face) contrasts with the open fibrous substrate below, illustrating the structural separation between the two functional layers.

Step Three: Enhancing mechanical robustness for manufacturability

As GDLs have become thinner (≤ 200 μm) to meet increased power density targets, mechanical fragility has emerged as a significant manufacturing challenge. Thin carbon papers are susceptible to web tears and may crack under winding tension during MEA manufacturing. They can also shed debris (e.g., broken fibres or loosened carbon particles, etc.) particularly when compressed at or above 3 MPa in the system. This particulate contamination is a known risk factor for membrane pinhole formation and cell shorting.

The advantage

At a nominal thickness of 150 μm, the new GDL design achieves a machine direction (MD) break strength of over 3000 N/m shown in Figure 2 (roughly double that of standard commercial papers like MB-39) and can be wound onto 3 in. cores for roll-to-roll processing. In bending trials, the material sustains contact over 270° of arc around a 2 in. roller without web fracture, a geometry that would catastrophically fail most conventional carbon papers. Critically, the binder architecture that provides this flexibility also helps prevent fibre debonding under compression, eliminating the debris generation mechanism that puts ultra-thin membranes at risk during stack assembly.

Step four: Protecting the interface with a resilient MPL

With the advent of ultra-thin membranes (≤ 10 μm), the surface topology of the microporous layer (MPL) is critical. Traditional MPL designs lack inherent structural strength; under compression, they deform, allowing substrate fibres to penetrate the surface and increase roughness. This creates a hostile environment for delicate ultra-thin membranes.

The engineering response

FLEX-GDL incorporates a high-modulus binder in the MPL formulation that maintains structural integrity under compressive load. Unlike conventional carbon black MPLs that utilise PTFE as an additional binder, this system resists lateral deformation, preventing substrate fibre protrusion through the MPL surface. Surface profilometry data confirms the practical effect with average surface roughness (Ra) measured under compression utilising a soft lithography method and finding an average of 2.4 μm at nominal conditions (no compression) and 2.35 μm when compressed at 2 MPa. When compared to an average of 5 μm roughness found for the prior-generation baseline (MB-39), it is over a 50% improvement in surface smoothness maintained under cell assembly loads (Figures 3 and 4). This stable, smooth interface is a prerequisite for reliable deployment of ultra-thin membranes below 10 μm, where even modest fibre protrusion constitutes a shorting risk.

Conclusion: The next step in GDL evolution

The transition from simple ‘black paper’ substrates to an engineered, high-performance graphitic substrate marks a new era in fuel cell technology. The release of a new generation of GDL technologies like the FLEX-GDL represents the next step in GDL evolution, moving beyond simple substrate functionality to become an active, tailored solution for the rigorous demands of heavy-duty transport.

By systematically addressing the interrelated challenges of compression set, electrical resistance, mechanical fragility, and surface topology, as described in Table 1, a new material with improved performance has been introduced. The highly graphitic substrate chemistry also enables compatibility with both acidic and alkaline electrochemical environments, extending applicability beyond PEMFC into electrolyser and CO2 reduction platforms, an increasingly relevant requirement as the hydrogen economy matures. As the industry moves toward higher power densities and longer target lifecycles, materials like FLEX-GDL illustrate how systematic GDL engineering can remove bottlenecks that have historically constrained the entire stack.