What Happens Inside a PEM Fuel Cell

Introduction

This article is for those of you that are asking yourselves: what are the physical processes taking place inside a fuel cell? While there’s a lot of literature on novel materials and advanced cell design, it’s important to understand what exactly happens inside a fuel cell so we can decide the best materials and electrode architecture for a given application, as well as understand the advantages and disadvantages of each design selection. To answer this question, this blog will focus on outlining the different processes occurring in a fuel cell and analyzing the three most critical ones: the anodic and cathodic reactions, and charge transport in the proton exchange membrane.

 

Overall physical processes

The figure below shows the different physical processes that occur inside a proton exchange membrane fuel cell. In the anode, where fuel oxidation occurs, hydrogen is introduced to the cell via the gas channels in the flow field and then it is transported, via convection and diffusion, to the reaction sites in the anode catalyst layer. In this layer, it undergoes the hydrogen oxidation reaction (HOR) at the surface of the catalyst and produces electrons and protons. The electrons are transported via the metal catalyst, usually platinum, and the carbon support in the catalyst layer, and then via the carbon fibers in the gas diffusion layer to the carbon flow field and metal current collector. The protons are released from the metal catalyst surface to the electrolyte in the catalyst layer and then transported in the water-filled nanopores in the polymer electrolyte membrane, usually Nafion (as will be discussed below), to the cathode electrolyte and catalyst surface.

To establish an electrical potential and a subsequent flow of current, the electrons and protons from the anode must recombine in the cathode following an electrochemical reaction with a higher half-cell potential than the anode. The anode has a half-cell potential of 0 V at standard temperature and pressure (the electrode is known as the standard hydrogen electrode). In most fuel cells, the oxygen reduction reaction (ORR) is the cathode half-cell reaction of choice as it has an equilibrium potential of 1.23 V at STP with respect to the standard hydrogen electrode, and requires only air, which is readily available in the atmosphere. Therefore, air is fed to the cathode electrode via gas channels from where it travels, via diffusion and convection, to the metal catalyst surface, again usually platinum. The ORR produces water, which depending on the operating temperature and pressure, can either be effectively removed as water vapour or accumulate as liquid water and removed via capillary pressure. The latter mechanism however can severely limit fuel cell performance as the liquid water can occupy a large portion of the empty pores in gas diffusion and catalyst layers, thereby hampering the transport of oxygen (see blog 2 for an explanation of voltage losses due to improper water management). Furthermore, water can accumulate in the channels leading to flow maldistribution and severe voltage/current oscillations [1].

As discussed above, there are many physical processes occurring inside a fuel cell but of paramount importance are the electrochemical reactions, and proton transport. These are discussed in more detail below.

 

Hydrogen oxidation reaction

The main reaction occurring on the surface of the anode catalyst is the hydrogen oxidation reaction, which is given by

\[H_2 \underset{k_b}{\stackrel{k_f}{\rightleftharpoons}} 2 H^+ + 2 e^-\]

This reaction involves more than one electron transfer and, as a result, can be further divided into several elementary reactions that need to proceed concurrently on the surface of the catalyst, usually platinum (Pt). The following three elementary step, outlined in the works of Tafel, Heyrovsky and Volmer, are used today to develop most hydrogen oxidation reduction models, e.g., [2,3],

\[H_2 + 2 Pt \underset{k_{-1}}{\stackrel{k_1}{\rightleftharpoons}} 2 PtH_{(ads)} \quad \text{Tafel}\] \[H_2 + Pt \underset{k_{-2}}{\stackrel{k_2}{\rightleftharpoons}} PtH_{(ads)} + H^+ + e^- \quad \text{Heyrovsky}\] \[PtH_{(ads)} \underset{k_{-3}}{\stackrel{k_3}{\rightleftharpoons}} Pt + H^+ + e^- \quad \text{Volmer}\]

Based on the Tafel-Heyrovsky-Volmer mechanism above, two reaction pathways are possible to complete the overall hydrogen oxidation reaction, i.e., 1) a Tafel step followed by two Volmer steps (TV), and 2) a Heyrovsky step followed by Volmer step (HV). In both cases, there is an interplay between the number of available Pt sites, and adsorbed species, i.e., PtH_(ads). Too many adsorbed species might block further adsorption of hydrogen, while without intermediates, the reaction cannot be completed. Therefore, it is essential to find a catalyst that can strongly adsorb the hydrogen molecule onto the surface to form the intermediate species (PtH_(ads)), yet the adsorption energy is not high enough to prevent the subsequent desorption of protons in the Volmer step, as outlined by the Sabatier principle.

Platinum is an excellent catalyst for the HOR. In fact, due to the extremely fast reactions, quantification of kinetic parameters using the standard rotating disk electrode technique is difficult and hydrogen pumps and ultra-low loading floating electrodes must be used for accurate determination [4]. It is for this reason that, despite the high cost of platinum, it is commonly used as the anode catalyst since very low catalyst loading (below 0.1 mg/cm²) can be used without significantly affecting cell performance. Many other precious catalysts have been studied in the literature, e.g., Ir/C, Pd/C and Rh/C [4], and many articles in the literature have also tried to identify non-precious catalysts; however, despite its high cost, platinum remains thus far the most used catalyst.

Each electrochemical reaction above requires a voltage difference between the solid and electrolyte phases to proceed. This difference is known as the overpotential of an electrochemical reaction, and it is given by

\[\eta = \phi_s - \phi_{el} – E^{eq}\]

where φ_s is the solid/metal catalyst potential, φ_el is the electrolyte potential, and E^eq is the half-cell equilibrium potential. The overpotential is the necessary voltage difference required to activate the electrochemical reaction and results in an irreversible energy loss that needs to be minimized. The current density of each individual reaction is given by the Butler-Volmer equation, which assuming negligible mass transport losses, is given by

\[i = i_0 \left( \exp \left( \frac{\alpha_a F \eta}{RT} \right) - \exp \left(-\frac{\alpha_c F \eta}{RT} \right) \right)\]

where i₀ is the exchange current density and is dictated by the catalyst and α_i is the transfer coefficient given by the preferential reaction pathway in the given catalyst. Using this equation for each elementary step, complex kinetic models can be developed as shown in references [1,2].

Proton transport in the membrane

High power density fuel cells are a result of the discovery of Nafion, a perfluorinated sulfonic-acid (PFSA) ionomer, in the 1970s by DuPont. PFSAs are ion-conductive polymers that have remarkable mechanical and chemical stability enabling the manufacture of ultra-thin proton conducting membranes. These membranes allow for the anode and cathode electrode separation to be reduced to several micrometers, thereby reducing ohmic losses due to ion transport considerably.

Nafion is a co-polymer obtained by modifying polytetrafluoro-ethylene (PTFE) by the inclusion of pendant side-chains terminated in a hydrophilic ionic group, SO₃⁻ (poly sulfonyl fluoride vinyl ether). In the presence of water, the dissimilar nature of the PTFE backbone and the side-chains results in a phase-separated morphology that can readily transport protons by means of a water-filled nanochannel network within the material. Transport properties, such as proton conductivity, water diffusivity and electro-osmotic drag coefficient (i.e., average number of molecules dragged per ion transported in the membrane) depend on the amount of water in the membrane, which is usually quantified in terms of moles of sorbed water per moles of ionic groups and represented with the variable lambda (λ). The amount of water (λ) depends on the environment the membrane is exposed to, i.e., either liquid water or gas at a given relative humidity, with the relation between water content and water activity known as the sorption isotherm. Transport properties are obtained as a function of water content and reported in the literature, e.g., [5]. For a given hydration, the properties of PFSA can be modified by changing ion-exchange capacity (IEC), given in terms of moles of ionic acid group per gram of dry polymer, and side-chain chemistry and length.

 

Oxygen reduction reaction

The oxygen oxidation reaction is given by,

\[½ O_2 + 2 H^+ + 2 e^-\underset{k_b}{\stackrel{k_f}{\rightleftharpoons}} H_2O\]

The ORR is a more sluggish reaction than the HOR and it usually dominates the kinetic losses in a fuel cell. The reason for the sluggishness of the reaction is that it requires four electron transfers to proceed, and it involves several intermediates, e.g., PtO, PtOH, PtOOH, and other oxides. Despite the large number of experimental and theoretical studies, a generally accepted reaction pathway for oxygen reduction on platinum still does not exist; however, there are several commonly accepted elementary steps that are used to build ORR models, i.e.,

\[O_2 + H^+ + e^- + Pt \underset{k_{-1}}{\stackrel{k_1}{\rightleftharpoons}} PtOOH_{(ads)} \quad \text{Associative Adsorption}\] \[PtOOH_{(ads)} + Pt \underset{k_{-2}}{\stackrel{k_2}{\rightleftharpoons}} PtOH_{(ads)} + PtO_{(ads)} \quad \text{Dissociative Transition}\] \[PtO_{(ads)} + H^+ + e^- \underset{k_{-3}}{\stackrel{k_3}{\rightleftharpoons}} PtOH_{(ads)} \quad \text{Reductive Transition}\] \[PtOH_{(ads)} + H^+ + e^- \underset{k_{-4}}{\stackrel{k_4}{\rightleftharpoons}} H_2O + Pt \quad \text{Reductive Desorption}\]

Density functional theory studies have shown that for materials that bind oxygen intermediates, i.e, PtO_(ads), PtOH_(ads) and PtOOH_(ads), too weakly, the reaction is usually limited by the formation of PtOOH_(ads) (step 1), while for strong-binding materials it is limited by the reductive desorption step [6,7]. Theoretical studies have also shown that the binding energy of PtOH_(ads) and PtOOH_(ads) are related to each other by a constant amount (≈3.2 eV), and, as a results, there is a universal descriptor of the reaction, which is usually taken to be the adsorption energy of PtOH_(ads) [7]. As in the previous case, the trade-off between binding energies leads to a Sabatier-type “volcano plot” with a range of catalyst that are more or less suitable depending on their binding energy.

Thus far, Pt-based catalysts remain the most active for the oxygen reduction reaction, these include shape-controlled Pt nanoparticles and nanowires, where the (111) more active facets are favoured over the (100) facets [7] and, alloys of Pt and Ni, Pt and Co, and Pt and rare earths, such as Pt–Y, Pt–Gd, and Pt–Tb [4]. In a fuel cell, dealloyed Pt–Ni and Pt-Co have shown some of the highest mass activities in the literature. To replace the use of expensive platinum, high-temperature treated, heterogeneous M-N-C (M = Mn, Fe, Co) catalysts have also been studied and shown to be able to achieve moderate activity and stability.

 

Conclusion

In this blog, we have reviewed the main physical processes occurring in a fuel cell. Of paramount importance are proton transport, and the hydrogen oxidation and oxygen reduction reactions, which have been discussed in some detail. Proton transport is enabled by Nafion, a co-polymer capable of transporting protons while surviving the harsh operating conditions in a fuel cell. The two electrochemical reactions in a fuel cell are multi-step reactions that require advanced catalyst to activate the reactions so that they can proceed quickly and without requiring a large overpotential loss. The development of advanced ionomer and catalysts is critical to the development of more cost effective and durable fuel cells.

 

---

References:

[1] Kosakian et al., Electrochimica Acta, 469, 143221, 2023.

[2] Kucernak and Zalitis, J. Phys. Chem. C, 120, 10721−10745, 2016

[3] Wang et al., Journal of the Electrochemical Society, 153(9):A1732-A1740, 2006

[4] Wei et al., Adv. Mater., 31, 1806296, 2019

[5] Kusoglu and Weber, Chem. Rev, 117, 3, 987–1104, 2017

[6] Sargeant et al., Electrochimica Acta, 426, 140799, 2020.

[7] Viswanathan et al. ACS Catalysis, 2 (8), 1654-1660, 2012.