As the rate-limiting steps with sluggish kinetics, the ORR (at the cathode of PEMFCs) and OER (at the anode of PEMWEs) are the key factors determining the performance of PEMFCs and PEMWEs, respectively^[47,48]. The ORR and OER occur at the three-phase boundary^[49], which is defined as a catalyst-electrolyte interface where electrons, protons, and gases come into contact with each other. This interface involves a proton-conducting ionomer, an electron-conducting catalyst, and a gas-permeable pore structure. Figure 2 illustrates the mass transport processes in the cathode of PEMFCs and the anode of PEMWEs. Taking PEMFCs as an example, oxygen is first transported to the catalytically active sites, where it undergoes adsorption and activation. Anode-generated protons are transported to the active sites via the PEM, whereas electrons are conducted via the external circuit^[50,51]. The construction of efficient triple-phase interface regions to improve the utilization of electrocatalysts has emerged as a viable strategy to reduce noble metal catalyst loading while maintaining cell performance^[52,53].

In a typical catalyst layer, the ionomer completely covers the catalytic surface of the TPB^[54]. During the reaction, protons are conducted via the vehicle mechanism and the Grotthuss mechanism^[55]. To enhance proton conductivity, numerous researchers have attempted to modify the molecular structure^[56] and composition^[57] of the ionomer. However, due to the irregular distribution of catalyst particles in conventional CLs, proton transport pathways are inevitably disrupted, leading to the formation of inactive sites. Therefore, the optimization of the ionomer distribution and the construction of well-defined ionomer/catalyst particle interfaces are crucial. Ordered proton conductors can improve catalyst layer proton conductivity through fast, oriented transport channels. Ning et al.^[58]. fabricated nanoscale Nafion arrays and subsequently deposited Pt nanoparticle catalysts by magnetron sputtering, resulting in a proton-ordered MEA [Figure 3A]. Its ultra-high specific surface area provides a large number of active sites for catalyst loading, while its well-ordered nanostructure forms abundant proton transport channels, effectively reducing the proton transport resistance of the catalyst. In the catalyst layer, the catalyst skeleton can form continuous electron transport channels^[59]. However, some catalyst particles may be completely electronically isolated by the ionomer, resulting in the formation of “dead zones”^[60]. During the disordered electron transport process, electrons encounter increased interfacial resistance and scattering effects, which leads to a rise in ohmic losses. In contrast, ordered electron channels can provide continuous and efficient electron conduction pathways, thereby significantly reducing electron transport losses. Tian et al.^[59] used physical sputtering to deposit Pt particles on vertically aligned carbon nanotubes (VACNTs), creating a CL with an ordered electronic conductor array. The ACNTs form oriented, continuous channels within the electrode, extending from the proton exchange membrane to the gas diffusion layer, thereby precisely controlling the transport pathways of electrons, protons, and reactants at the mesoscale. This ordered structure eliminates the tortuous, random transport networks commonly found in traditional disordered catalyst layers, achieving synergistic optimization of the three-phase interface, charge conduction, and mass supply at the mesoscale. [Figure 3B and C]. Consequently, this electrode exhibits performance comparable to that of commercial electrodes with a platinum loading one order of magnitude higher even at a loading of only 35 μg cm^-2 [Figure 3D].

Improving catalyst utilization and high ORR or OER activity is highly desirable for high-performance CLs^[61]. Catalyst utilization and ORR or OER activity are directly related to the interfacial structure of the support and mass transport properties^[62,63]: higher dispersion^[64], greater ionomer coverage^[51] and improved mass transport properties will lead to better CL performance. From the mesoscale perspective, these key performance-influencing factors are essentially determined by the spatial arrangement of CL components at the mesoscale. Therefore, to design and fabricate high-performance cathode CLs with low precious metals loading and high catalyst utilization, it is essential not only to select superior catalysts but also to optimize the mesoscale structure of the CL by regulating the mesoscale distribution of components to bridge the gap between microscale active sites and macroscale transport efficiency^[65,66]. Wang et al.^[67] precisely regulated the mesoscale spatial configuration of Pt catalysts by employing a self-assembly strategy with PS-b-P4VP block copolymers to prepare metal-containing nanocomposites with ordered honeycomb-like (OHC) mesopores. This mesoscale OHC structure serves as a "nanoscale template" that constrains the growth and distribution of Pt nanoparticles: the metal can be precisely positioned on the surfaces and inner walls of the mesopores, avoiding random aggregation of Pt nanoparticles at the microscale. Xia et al.^[68] fabricated a hierarchical nanoarray-ordered CL. When used as both cathode and anode in a PEMFC, this electrode achieves a high mass-specific power density of 5.23 W mg^-1 Pt and exhibits excellent durability [Figure 3E and F]. As shown in Figure 3G, its inner layer consists of an ordered electronic conductor, polypyrrole (PPy), and the outer layer is an ordered proton conductor, Nafion ionomer. Platinum nanoparticles are distributed at the dual-functional interface, anchoring and stabilizing the active sites within an environment constructed by proton/electron transport pathways. This spatial arrangement at the mesoscale greatly enhances Pt utilization, resulting in impressive ORR activity and stability.

Furthermore, catalyst film electrodes, composed of smooth or nanostructured catalyst layers at the nanoscale, represent a promising candidate material for bridging microscale catalyst activity and macroscale MEA performance^[69]. Unlike traditional CLs, these film electrodes achieve mesoscale structural integration. Their ultrathin, continuous structures allow the catalyst to come into direct contact with the electrolyte membrane, thereby eliminating the carbon support and ionomer binder that would otherwise obstruct transport pathways at the mesoscale^[70,71]. This breakthrough design overcomes a key limitation of traditional catalyst layers. By replacing supported nanoparticles with a continuous catalyst film, it eliminates carrier corrosion and particle degradation. At the same time, the absence of polymer agglomeration and the shortened electron transport pathways directly enhance the kinetics of the ORR^[46,72,73]. For instance, Zeng et al.^[74] designed a 3D-ordered MEA using Pt-skin@PdCo nanotube arrays (NTAs) - the open-walled, aligned NTA structure constructs continuous mesoscale pathways for mass transport, boosting Pt utilization via dual-sided accessible sites and eliminating ionomer reliance, ultimately achieving high power densities (222.5 kW/gPt for cathode, 330.4 kW/gPt for anode). The proton conduction mechanisms on ionomer-free Pt electrode surfaces are as follows: (i) adsorbed H surface diffusion and (ii) water film-facilitated surface migration on the catalyst surface^[75-77]. It is particularly critical to provide effective proton conduction in ionomer-free catalyst film electrodes due to limited surface roughness and insufficient proton transport. Strategies aimed at enhancing proton conduction may need to balance the transport requirements of protons and oxygen, which could ultimately constrain electrode performance^[78-80]. In an effort to address these limitations, Yang et al.^[69] fabricated coaxial nanowire electrodes (CANEs) by encapsulating ionomers within Pt nanofilms [Figure 3H and I]. In this core-shell architecture, the traditional triple-phase boundary gives way to dual-phase boundaries, in which the ionomer core is responsible for proton conduction, while the Pt shell ensures rapid O₂ diffusion. This functional separation at the mesoscale decouples the transport pathways for protons and O₂, thereby achieving higher platinum utilization and durability even at low platinum loading. During the durability testing process, the CANE demonstrated remarkable resilience as depicted in Figure 3J and K. Nevertheless, it is noteworthy that the water production rate per unit thickness in catalyst film cathodes is 20-30 times higher than that of thinner Pt/C electrodes^[81]. This poses a significant challenge for water management strategies, indicating that further efforts are required to bring this promising catalyst film electrode technology into practical applications.

The mass transport overpotential in the CL accounts for 20%-50% of the total mass transport overpotential^[82]. The core cause of the mass transfer overpotential in the catalyst layer is the hindrance to the transport of reactants to the active sites and the discharge of the product from the active sites, which results in the reaction rate being limited by mass transfer rather than the electrochemical reaction itself; such mass transport relies on the pores within the CL. Typically, catalyst particles are dispersed into Nafion solution and then sprayed onto the PEM, followed by heat treatment to produce a porous CL^[83,84]. The carbon particles contain primary pores measuring 5-20 nm in size, while secondary pores measuring 20-100 nm exist between the particles. The pore structure not only enhances electrochemical activity at the microscale but also promotes effective mass transport within the CL at the macroscale^[85,86]. O₂ transport within the CL plays a crucial role in the superior performance of hydrogen-electric energy conversion systems. Figure 4A illustrates the O₂ transport pathways in the CCL of PEMFCs. After O₂ enters the CL from the flow channel through the GDL, molecular diffusion dominates in the porous medium, while Knudsen diffusion dominates in ionomer films^[87]. The pore-scale method is an efficient approach for studying oxygen reactive transport processes in CLs^[88,89]. Micropores (< 2 nm) are unfavorable for ionomer penetration, preventing protons from reaching internal active sites, while excessive macropores (> 20 nm) significantly hinder proton conduction through the ionomer network. Mesopores (2-20 nm) enable an ideal distribution of catalyst and ionomer^[90]. Therefore, the porous structure of the CL should be designed with a considerable proportion of mesopores. Yarlagadda et al.^[91] demonstrated that particles forming accessible pores in the range of 4-7 nm can minimize oxygen transport resistance. Wang et al.^[92] systematically compared the effect of carbon support pore size on Pt/ionomer interactions using three model catalysts with well-defined pore structures. The HOMC support with ordered mesopores achieved the optimal configuration, wherein Pt nanoparticles were positioned at the mesopore mouths while the ionomer resided on the external surface. This design avoids ionomer poisoning of Pt, ensures rapid O₂ diffusion through the open mesopores, and allows effective proton conduction from the external ionomer to the Pt sites at the pore mouths. In contrast, the Vulcan support exposed Pt on the external surface, leading to direct ionomer adsorption on Pt^[92]. Although this configuration maintains proton transport, it suppresses catalytic activity and adds oxygen diffusion resistance. The high-surface-area Ketjenblack support confined Pt deep within its pore channels, where the ionomer cannot reach, making the Pt sites largely inaccessible to protons [Figure 4B]. Song et al.^[93] reported that ordered mesoporous carbon outperformed disordered wormhole-like mesoporous carbon in fuel cell reactivity, even though both carbon materials have similar pore size distributions. This is because the ordered structure has better three-dimensional connectivity, which improves mass transport and makes it easier for platinum particles to reach the electrolyte [Figure 4C].

In the practical operation of PEMFCs, in addition to ensuring the continuous transport of reactant gases and the rapid transfer of electrons and protons, it is also essential to facilitate the rapid removal of water. When flooding occurs, the accumulation of water in the porous channels directly compresses the O₂ mass transport pathways, preventing O₂ molecules from reaching the active sites in a timely manner, thereby slowing down or even halting the ORR^[94,95]. The design and synthesis of hierarchically porous materials have become a research hotspot in the field of electrocatalysis. Hierarchically porous structures are three-dimensional multiscale architectures formed by the interconnection of pores with different pore sizes (typically including at least two or more types of pores)^[96]. Such structures are beneficial for improving the mass transport efficiency and anti-flooding capability of gas porous electrodes, thereby accelerating the ORR^[97]. By constructing MEAs with hierarchically porous structures, carbon corrosion is significantly mitigated, mass transport is enhanced, and material stability is greatly improved^[98-100]. Wang et al.^[101] fabricated an anti-flooding gas porous electrode with an ordered macroporous-mesoporous interpenetrating network using a soft-hard template method. Both macropores and mesopores exhibited a regular and ordered arrangement, while Pt nanoparticles were uniformly distributed on the porous carbon support [Figure 4D]. The mass transport efficiency and antiflooding efficiency of the ordered porous electrode were four times that of commercial Pt/C [Figure 4E]. MEA testing demonstrated that the output power of the porous electrode consistently surpassed that of commercial Pt/C catalysts, with a 45% improvement in power output even under highly humidified conditions compared to traditional electrodes.

The tortuous and random structure of traditional electrodes limits the transport of O₂ and water^[72]. In addition to the design of ordered pores, the ordered shape design of catalyst layers, including ridges, pillars, and channels, is also a method to reduce mass transport resistance and enhance the transport capabilities of reactant gases and water. Lee et al.^[102] proposed a multifunctional electrode with a highly ordered ridge and groove structure. The grooved electrode consists of catalyst ridges with high ionomer content, separated by empty grooves, providing effective pathways for H⁺ and O₂ transport. This design simultaneously enhances H⁺ and O₂ transport, achieving significant improvements in performance and durability. They used multiphysics simulation to analyze the local transport behavior of reactants and products within the electrode and calculated the spatial distribution of the oxygen reduction reaction rate (i_ORR). At an I/C (ionomer-to-carbon ratio) of 0.9, the introduction of grooves had a limited impact on performance due to the high H⁺ transport resistance in the electrode; however, as the I/C ratio increased, the grooves significantly influenced i_ORR, leading to notable performance improvements [Figure 4F]. The i_ORR distribution in the grooved electrode was more uniform, directly demonstrating improved oxygen and proton transport. Additionally, the oxygen concentration distribution within the grooved electrode was also more uniform [Figure 4G], further corroborating the above conclusion. Xuan et al.^[103] synthesized a novel double-sided ordered membrane electrode for PEMFC using a TiN nanorod array [Figure 4H]. Due to its low mass transfer resistance, this ordered structure significantly enhances gas diffusion efficiency. Additionally, the uniformly arranged channel structure in the double-sided ordered membrane electrode generates radial capillary forces. Combined with its low surface tension, this significantly increases the water drainage rate, thereby preventing water accumulation and channel blockage. The prepared ordered membrane electrode exhibited an almost linear change in output voltage as the current density increased, with no noticeable sudden voltage drop observed. After 200 h of accelerated stress test (AST) at 90 °C and 30/30% relative humidity, the peak performance only dropped by 4.8%. In contrast to the introduction of TiO₂ arrays, Wang et al.^[104] directly grew TiO₂ nanoarrays on the surface of the gas diffusion layer, thereby developing a novel self-humidifying membrane electrode. This catalytic layer structure effectively maintains humidity in the anode region and promptly replenishes moisture in the proton exchange membrane, thereby ensuring smooth proton conduction under low-humidity conditions. Lee et al.^[105] invented an in-plane channel-structured CL [Figure 4I]. A comparison between patterned CLs with structures parallel and perpendicular to the rib/channel configuration demonstrated that in-plane channels traversing both the channel and rib regions improved overall gas and water transport. The enhanced performance in the direction perpendicular to the channels was attributed to the higher utilization of the CL segments under the rib regions due to gas transport from the channel regions [Figure 4J]. Oxygen gain measurements provided additional evidence for the improved oxygen transport [Figure 4K]. Measurements of the limiting current oxygen transport resistance revealed that the observed performance increase stemmed from shorter effective diffusion distances, which promoted both molecular and Knudsen diffusion.

Ordered CLs provide short and directional transport paths at the beginning of life, but their advantage can diminish if the ordered architecture reconstructs under realistic operation. Understanding these end-of-life (EOL) degradation pathways is essential for assessing the practical durability of ordered MEA technologies^[106,107]. In PEMFC cathodes, Pt dissolution/redeposition, Ostwald ripening, carbon corrosion at high potentials, ionomer chemical attack, and repeated hydration-dehydration cycles can redistribute catalyst particles and ionomer domains, increasing local oxygen resistance and reducing electrochemically active surface area^[108-111]. Ordered arrays and ultrathin films may be especially sensitive because their high surface energy, small feature size, and limited mechanical redundancy can accelerate coarsening, detachment, or collapse under compression. In PEMWE anodes, Ir dissolution, Ti oxidation/passivation, ionomer degradation, oxygen-bubble blockage, and porous transport layer (PTL)/CL interfacial fatigue can change the true contact area and convert initially continuous pathways into isolated reaction zones^[112-114].

Ordered CLs, particularly those with high surface-area-to-volume ratios such as nanowires and nanoporous architectures, are susceptible to significant electrochemical active surface area (ECSA) losses during AST^[115]. Recent modeling studies have established that Pt degradation in the cathode catalyst layer follows a complex sequence involving Ostwald ripening, mass loss, and precipitation within the PEM, ultimately leading to Pt band formation^[116]. For ordered nanowire arrays, a unique dissolution-redeposition mechanism has been identified: during potential cycling, material dissolves from regions of high positive curvature and is redeposited along the surface during cathodic sweeps, leading to coarsening of the porous structure and eventual transformation into dense nanoparticle-like morphologies. This mechanism is accentuated by the high tortuosity of nanostructured architectures, which increases the probability of dissolved cations contacting the surface prior to elution into the bulk electrolyte^[117,118]. The stability of the ionomer phase in ordered CLs is equally critical. In ordered CLs where the ionomer/catalyst interface is precisely engineered, any degradation of the ionomer distribution directly disrupts the designed proton transport pathways, potentially causing more severe performance loss than in random CLs where alternative pathways may exist^[119]. Ordered carbon-based structures in CLs and GDLs face severe corrosion under long-term operation, particularly at high potentials and under start-stop conditions^[54]. The corrosion of carbon supports leads to catalyst detachment, collapse of ordered pore structures, and loss of electrical connectivity. For vertically aligned carbon nanotube arrays, while the ordered structure provides excellent initial transport properties^[120], the high surface energy and mechanical stress concentration sites may accelerate carbon oxidation compared to randomly distributed carbon powders^[54,115]. The most critical EOL challenge for ordered structures is the loss of orderedness itself. As degradation proceeds, the initially ordered transport channels may become blocked by Pt precipitation, ionomer redistribution, or carbon corrosion products^[121]. Unlike random porous networks that possess inherent fault tolerance through tortuous alternative pathways, ordered channel architectures may suffer catastrophic transport blockage if critical pathways are compromised. This ‘all-or-nothing’ degradation mode represents a fundamental durability risk that must be addressed through the design of hierarchical ordered structures with built-in redundancy^[118]. For instance, Huang et al.^[122] designed an orderly interlocking membrane electrode assembly to prevent damage to the array structure during assembly. In the 300-h stability test of this electrode, the attenuation rate was only 18.9 μV/h @ 2 A/cm².

Mesoscale regulation of the CL focuses on optimizing reaction interfaces and mass transport pathways to improve catalyst utilization and reduce transport resistance. Through the design of ordered pore/shape structures and the construction of ordered proton/electron transport channels, the accessibility of active sites, mass transport efficiency, and flooding resistance are enhanced. These mesoscale design strategies enable the catalyst layer to achieve high performance at a low platinum loading. To facilitate the rational selection of fabrication strategies for ordered catalyst layers, Table 1 systematically compares the representative methods discussed above, providing a practical reference for researchers navigating the diverse landscape of ordered CL fabrication.

Constructing ordered proton transport channels within the membrane can improve the entangled structure of the PEM and create uninterrupted pathways for water molecules and protons^[133-136]. Ordered PEMs stand in stark contrast to the random and less controlled environment of traditional PEMs as shown in Figure 5A, where proton transfer channels in PEMs suffer from inefficient proton transport due to the amorphous nature of polymer segments^[136]. However, achieving sufficient proton conductivity requires maintaining a high ion exchange capacity, which inevitably leads to excessive swelling under high humidity, thereby compromising mechanical properties^[137]. In other words, the development of polymeric proton exchange membranes is constrained by the trade-off between proton conductivity and mechanical performance^[138]. These challenges have prompted researchers to explore novel proton exchange membranes with long-range ordered proton transfer channels and stable membrane structures^[139]. Introducing suitable fillers into Nafion or other sulfonated polymers can phase-separate the matrix, raise the density of acid-base pairs, and improve hydrogen-bond network connectivity, thereby optimizing proton transport channels^[133,140].

Li et al.^[141] prepared zirconium phosphate (ZrP) poly (diallyldimethylammonium chloride)-modified sulfonated halloysite nanotube (PSHNT) composites via electrostatic self-assembly, and incorporated them into a sulfonated polyimide (SPI) matrix to obtain a novel composite proton exchange membrane. The synergy between the high acidity of ZrP and the excellent water absorption of HNTs provided more proton transfer sites within the membrane and created long-range ordered conduction pathways as shown in Figure 5B. The SPI-ZrP/PSHNT-6 composite membrane exhibited a water uptake of 45.5% [Figure 5C], and a proton conductivity of 288.2 mS·cm^-1 at 100% relative humidity and 80 °C [Figure 5D]. The peak power density of a single cell based on this membrane was 230.2 mW·cm^-2. Shao et al.^[142] employed a sulfonation strategy to synthesize a two-dimensional hydrophilic covalent organic framework (ZUT-COF-SO₃H) rich in sulfonic groups. They incorporated this framework as a nanofiller into a Nafion matrix to prepare a composite proton exchange membrane [Figure 5E]. The well-ordered one-dimensional channels of ZUT-COF-SO₃H provide space for the formation of long-range ordered hydration pathways, while its high density of sulfonic acid sites significantly promotes proton conduction. Following the formation of such ordered pathways, proton migration becomes more direct and unimpeded, adhering to the Grotthuss hopping mechanism, resulting in a marked reduction in the activation energy for proton transport [Figure 5F]. At 80 °C and 80% relative humidity, the COF-Nafion membrane doped with 10% ZUT-COF-SO₃H exhibited proton conductivity 2.4 times that of the pure Nafion membrane. When assembled into a fuel cell as a proton exchange membrane, the composite membrane demonstrated excellent performance, achieving a maximum power density of 304.056 mW·cm^-2 at a current density of 640.179 mA·cm^-2.

Ion channels in conventional PEMs typically adopt a tortuous morphology, and their proton conduction behavior is often isotropic, or even exhibits undesirable anisotropy in which in-plane conduction exceeds out-of-plane conduction. In order to address this challenge, researchers have explored electric-field alignment and magnetic-field alignment techniques to orient conductivity in the through-plane direction^[143,144]. In PEMs, protons are bound within the negatively charged -SO₃H groups^[145]. When subjected to a uniform external electric field, these charged components can form dipoles and undergo rotation^[146,147]. As described below^[148], the energy generated by this rotation can force the molecules to align in a specific direction:

(1) $$ U = \text{n}Eqd(1 − cos\theta) $$

where U denotes dipole rotation energy; n stands for the total amount of involved molecules; E refers to the intensity of the externally applied uniform electric field; q represents the charge magnitude of polarized units; d is the spacing separating paired opposite-charge species; and θ characterizes the included angle between the dipole axis and the direction of the applied electric field.

Dielectric polarization is the primary driving force behind the morphological reorganization of block copolymer nanostructures under the influence of an external electric field. Studies have shown that even after the removal of an external magnetic field, the aligned ionic clusters can remain stable, indicating that the formation of ion-conducting channels does not rely on the external magnetic force^[149]. Based on this characteristic, the introduction of an external electric field can effectively regulate the positioning and distribution of ion-conducting nanochannels within the membrane^[150,151]. By applying an electric field during the membrane casting process, Ryu et al.^[131] obtained recast Nafion with more ordered microstructures and reduced ion cluster sizes. Wang et al.^[152] incorporated proton-exchange montmorillonite (MMT) nanosheets into a polybenzimidazole (PBI-O) matrix containing ether linkages. Throughout the solvent evaporation process, an alternating electric field was used to align the MMT fillers in situ within the initial cast film solution [Figure 5G]. SEM images verified the generation of continuous proton-conduction pathways inside hybrid films, among which the composite filled with 5 wt.% MMT delivered the most regular and complete conductive channel network. These morphological observations solidified the positive contribution of the external alternating electric field to constructing ordered ion migration pathways and thus boosting the through-plane proton conduction of PBI-O-based high-temperature membranes [Figure 5H-K]. The 5% aligned MMT nanosheets increased the conductivity of the pure PBI-O membrane by 400% without sacrificing other properties [Figure 5L].

Magnetic field-induced alignment offers unique advantages in regulating material structures, as it only requires magnetic materials to possess sufficient saturation magnetization to respond to the field. By introducing magnetizable materials and aligning them under a magnetic field, the orientation of channels can be conveniently controlled to optimize material performance. Although most metals exhibit some degree of magnetism, only a few, such as Fe, Co, and Ni, display ferromagnetism. Among these, Fe ranks first in magnetic strength, with Co coming second, while other magnetic metals are typically paramagnetic with relatively weak magnetism. Therefore, Fe₃O₄ is often chosen as a prevalent doping candidate to realize directional filler arrangement under external magnetic fields due to its excellent magnetic properties^[153,154]. Han et al.^[132] prepared poly-dopamine-coated, phosphotungstic acid-loaded magnetic ferrite nanoparticles (DMNPs@HPW) with excellent magnetic responsiveness [Figure 5M]. By incorporating this functional filler into a sulfonated poly(ether ether ketone) matrix and utilizing a magnetic field to induce the oriented alignment of the filler, an ordered proton conduction pathway spanning the entire membrane thickness was established within the membrane. Hyun et al.^[155] synthesized one-dimensional magnetic WS₂ nanotubes by dispersing a surface-modified polystyrene sulfonic acid layer within a Nafion matrix and loading them with Fe₃O₄ nanoparticles [Figure 5N-P]. By applying a weak magnetic field of 0.035 T during the casting process, they successfully achieved the perpendicular alignment of proton conduction channels along the thickness of the membrane [Figure 5Q]. Molecular dynamics simulations indicate that the high density of sulfonic acid groups on the nanotube surface induces the formation of a continuous hydration layer at the Nafion/nanotube interface, providing a high-speed pathway for protons. Compared to recast Nafion membranes of the same thickness, the vertically aligned composite membrane exhibited a 69% increase in out-of-plane proton conductivity at 30% relative humidity, and fuel cells based on this membrane demonstrated a 51% increase in peak power density under the same humidity conditions.

For hydrogen‑electric energy interconversion systems, the interfacial configurations between components are as decisive as the intrinsic material properties of each part in determining overall performance and operational stability. Therefore, in addition to the intrinsic design of the PEM, the interface between the PEM and the CL must also be considered. The PEM/CL interface structure significantly influences the mass transport and charge transfer processes in the MEA^[154,156]. With technological advancements, the initial two-dimensional interface structure achieved by the CCM method has evolved into three-dimensional engineered interfaces between the PEM and CL, including engineered surface-patterned PEMs and porous membrane designs [Figure 6A]^[19,60]. Among these, surface-patterned PEMs enable precise control over the PEM surface structure, allowing the fabrication of the PEM with ordered patterns. This effectively optimizes the contact interface between the PEM and the CL, shortens the proton transport path, and reduces the proton transport resistance^[157]. Lee et al.^[158] constructed an ion-pair membrane with an inverted pyramid microstructure pattern using a solvent-assisted patterning technique. This design demonstrated higher transfer efficiency, enhanced capillary penetration of phosphoric acid (PA) into the catalyst layer, and reduced PA loss due to compression [Figure 6B]. Furthermore, compared to traditional CCS and flat CCM, the three-dimensional interlocking membrane/catalyst layer structure exhibited significant improvements in interfacial bonding strength and electrochemical active surface area [Figure 6C]. This indicates a uniform distribution of PA along the proton transport pathways, increasing the number of active sites. The patterned CCM also exhibited the lowest oxygen transport resistance [Figure 6D], attributed to changes in the morphology and porosity of the catalyst layer induced by the patterned membrane. Benefiting from these improvements, the patterned CCM, composed of an ion-pair membrane and protonated PA ionomer, achieved a high power output of 680 mW cm^-2 under dry H₂/ air conditions at 160 °C with a loading of 0.5 mgPt cm^-2 [Figure 6E].

The employment of patterned membranes has been demonstrated to enhance the reaction rate by means of two principal mechanisms. Firstly, it leads to an expansion of the membrane-electrode interface, and secondly, it results in a reduction of the thickness of the CL. In addition to fabricating single-scale patterns, multiscale patterns with various micro- and nanoscale morphological features can also be designed on the membrane surface^[159]. Li et al.^[160] fabricated a multi-scale structured PEM by replicating the hierarchical architecture of natural lotus leaves. The assembled MEA equipped with this biomimetic multi-layered PEM achieves an ohmic resistance of only 84 mΩ cm², outperforming single-scale patterned counterparts [Figure 6F]. It features microscale columnar structures and nanoscale needle-like protrusions. This multi-scale structured PEM combines dual advantages: enhanced surface area from nanoscale features and reduced thickness through optimized microscale architecture [Figure 6G]. The distinctive hierarchical surface structure facilitates enhanced water management. The loose contact region full of voids is constructed between the textured membrane and catalytic layer, enabling unobstructed oxygen transport and fast drainage of cathodic liquid water [Figure 6H]. Single-cell testing demonstrates that the MEA equipped with this multiscale membrane enhanced discharge performance by a factor of 1.96 in comparison with a MEA with a flat PEM. Mesoscale regulation of PEM focuses on constructing ordered proton transport channels and optimizing the PEM/CL interface. Through filler modification, external field-induced alignment, and surface patterning (including multiscale patterns), the proton conductivity of PEMs is improved while maintaining mechanical stability, and the interfacial mass transport and charge transfer efficiency are enhanced.

A range of advanced manufacturing methods, including 3D additive manufacturing, laser etching, and mask lithography, enable researchers to precisely control pore size, geometric profile, spatial distribution, and total porosity, thereby fabricating ordered GDLs with uniform microstructures and customizable properties^[123]. Table 3 presents a comparative overview of ordered gas diffusion layer fabrication methods. Huang et al.^[173] utilized 3D printing technology to fabricate a nickel-based gas diffusion layer with straight-through pores and a three-dimensional periodic structure [Figure 7A]. They further characterized the bubble escape behavior, wettability, ohmic resistance, and mass transport resistance of the two GDLs using an in-situ observation system, contact angle measurements, and electrochemical impedance spectroscopy. As shown in Figure 7B-D, in the anion exchange membrane water electrolyzer (AEMWE), the GDL_{3D print} exhibits excellent electrolytic performance. This advantage stems from three structural benefits: a periodic grid arrangement, surface properties that balance gas-affinity and gas-repulsion, and lower internal ohmic resistance. This customized porous structure suppresses excessive bubble growth and shortens gas residence time, while simultaneously enhancing electron conduction between the layer and the flow plate, thereby synergistically reducing ohmic polarization losses and concentration polarization losses. In a 1 mol/L KOH solution, the NiFe/ GDL_{3D print}. only requires 240 mV overpotential to sustain 10 mA cm^-2 OER activity. Furthermore, the AEMWE with the GDL_{3D print}. under pure water supply achieved an industrial current density of 1A cm^-2 at 1.8 V. Lu et al.^[174] modified the GDL using laser perforation technology to control pore size and pattern. The perforation pattern consisted of a square array of 100 μm diameter holes spaced 1 mm apart, applied only to the microporous layer (MPL) [Figure 7E]. The single cell with the perforated GDL demonstrated high performance under conditions of 90 ℃/30% relative humidity and 40 ℃/100% relative humidity. To elucidate the impact of perforation, ex-situ water migration was analyzed using neutron imaging [Figure 7F]. In the reference cell, the water thickness fluctuated within a range of 0.2-1.3 mm, whereas in the cell with the perforated GDL, the water thickness remained relatively uniform. Additionally, the breakthrough pressures for the reference GDL and the perforated GDL were 25.0 kPa and 9.5 kPa, respectively. These findings were corroborated by an ex-situ water transport experiment, which demonstrated that localized perforation of the porous layer effectively achieved uniform water distribution and improved water transport in the GDL [Figure 7G]. Wang et al.^[175] fabricated an MPL with basket-weave patterned grooves approximately 80 µm wide [Figure 7H and I]. Research has found that the grooves formed between the CL and the ordered GDL do not increase contact resistance and actually promote water retention. Molecular dynamics simulation results [Figure 7J] validated this strengthened water-retaining characteristic of structured MPLs. Under low-humidity operating environments, the extra retained moisture lowers internal ohmic loss and boosts proton conduction within electrodes, so MEAs equipped with grooved microporous layers deliver much better electrochemical output compared to flat counterparts. Zhu et al.^[176] fabricated ordered PTLs by directionally stacking titanium fibers, which effectively eliminated dead zones and improved permeability.

A good interfacial contact at the transport layer and CL interface enables high catalyst utilization^[177]. To maximize this contact and eliminate discrete interfaces, researchers have explored using the porous transport layer directly as the supporting skeleton for catalysts. In such architectures, the porous material not only acts as a structural framework but also serves as an intrinsic electrocatalytic medium. Meng et al.^[123] proposed a method for fabricating ordered membrane electrodes for PEMFCs using a highly ordered Pt-Co/VACNTs bifunctional layer. The highly ordered hydrophobic VACNTs served as the microporous layer, and Pt-Co alloy catalysts were deposited on the VACNTs via dual-target magnetron sputtering. Finally, a pure carbon paper without a MPL was hot-pressed to form the complete membrane electrode structure. In traditional membrane electrodes [Figure 7K], the random distribution of particles in the MPL and CL makes it challenging to precisely control the microstructure^[123]. In addition, the highly tortuous pore pathways within the MPL, combined with the insulating and hydrophobic nature of polytetrafluoroethylene (PTFE), hinder the transport of oxygen, electrons, and liquid water around the oxygen reduction active sites^[178]; moreover, these mass transfer limitations are further exacerbated under high-current conditions^[179]. As shown in Figure 7L, vertically aligned carbon nanotube arrays feature a straight, continuous pore structure; this ordered framework can significantly accelerate multiphase mass transfer processes within the electrode assembly, thereby enhancing reaction rates and stability. Both the ohmic polarization and charge transfer polarization measured in single cells were reduced. This result confirms that the Janus bifunctional catalytic film optimizes the interface between the proton exchange membrane and the diffusion anode, ensuring uniform dispersion of the Nafion-doped catalyst particles and accelerating the diffusion rate of oxygen into the TPB [Figure 7M]. This membrane electrode exhibits high electrochemical stability [Figure 7N] and Pt utilization, achieving a maximum power density of 1.41 W cm^-2, outperforming membrane electrodes prepared using the CCM method and traditional carbon paper with an MPL.

Mesoscale regulation of GDLs mainly involves precise control of pore structure and interface properties through advanced manufacturing technologies and catalyst integration strategies. Standalone ordered GDLs with controlled pore size, shape, and distribution improve gas diffusion and water management efficiency. Catalyst-integrated ordered GDLs achieve synergistic optimization of catalyst utilization and mass transport, enhancing the overall performance and stability of the MEA.