MA_Platteau/Content/Theoretical Background.tex

\chapter{Theoretical Background}
\label{cap: Theorie}


In this chapter, the fundamental concepts and components of fuel cells are explored to provide a comprehensive understanding of Proton Exchange Membrane Fuel Cells (PEMFCs). Section 2.1 covers the basic principles of fuel cells, followed by Section 2.2, which explores electrochemical fundamentals, including the thermodynamics of the cell (2.2.1). Section 2.3 focuses specifically on PEMFCs, discussing their operation (2.3.1), overpotential (2.3.2), and methods of characterisation (2.3.4). Finally, Section 2.4 examines degradation mechanisms in PEMFCs, detailing the degradation of the platinum catalyst (2.4.1), membrane degradation (2.4.2), electrochemical carbon corrosion (2.4.3), and general corrosion processes (2.4.4). This structure lays the groundwork for understanding the challenges and performance factors of PEMFCs in practical applications.


\section{Fundamentals of the Fuel Cell}
\label{sec: Revox}

As mentioned in the introduction, the clear impact of GHG and especially CO$_2$ emissions on climate change and the environment is undeniable. Therefore, new technologies such as fuel cells with practically zero emissions as well as no noise pollution are becoming a promising alternative to conventional internal combustion engines (ICEs). These engines continue to depend on fossil fuels to function, whereas fuel cells run on hydrogen and air undergoing electrochemical reactions within the fuel cell to generate electrical power. This reaction results in water as a byproduct \citep{01_wilberforce_advances_2016,02_baroutaji2015materials}.
Battery electric vehicles (BEVs) are another alternative to ICEs that could lower GHG and CO$_2$ emissions, but only if the electric energy is also produced via renewable sources. However, when comparing fuel cells with BEVs there are some advantages worth mentioning. Fuel cells, like ICEs, can be
recharged almost instantly; a feature that BEVs do not share. Besides this, fuel cells may also run on other fuels and not just on pure hydrogen. This does however depend on fuel cell type. Fuel cells do not need to be disposed of like batteries and have a much longer operation time. Finally, fuel cells have a wider range of temperatures in which they may be operated \citep{01_wilberforce_advances_2016, 02_lucia2014overview}.

Before continuing into the mode of operation and electrochemistry behind the fuel cell, it is essential to briefly explain the various types of fuel cells. These may be categorised based on the type of electrolyte membrane they use: solid oxide fuel cells (SOFCs), molten carbonate fuel cells (MFCSs), alkaline fuel cells (AFCs) and the most important for this thesis: polymer electrolyte membrane fuel cells (PEMFCs)\citep{02_wang2020fundamentals}. Also depending on the fuel cell type, operating temperatures may vary between -40 to almost 1000 °C. This phenomenon will be explained in the following\citep{02_Abderezzak2018}.

\subsubsection{Solid Oxide Fuel Cells (SOFCs)}

The operating temperatures of SOFCs is higher than in the other types of fuel cells. They range from 500 to 1000 °C \citep{SOFC_hauser2021effects}. This high operating temperature allows the fuel cell to run not only on pure hydrogen but also with gases which also contain methane (CH$_4$), carbon dioxide (CO$_2$) as well as carbon monoxide (CO), water vapour (H$_2$O) and hydrogen H$_2$ \citep{SOFC_lin_analysis_2024}. If the anode fuel has CH$_4$ and water vapour in it, the methane (CH$_4$) may be reformed in the fuel cell by the process of steam reforming as shown in the equation \ref{eq:Steam reforming}, producing hydrogen (H$_2$)and carbon monoxide (CO) \citep{SOFC_Haberman2004}.
\begin{equation}
	\mathrm{CH}_4+\mathrm{H}_2 \mathrm{O} \leftrightarrow \mathrm{CO}+3 \mathrm{H}_2, \Delta H_{298}=2.06 \times 10^5 \mathrm{KJ} / \mathrm{Kmol}
	\label{eq:Steam reforming}
\end{equation}

Another internal reaction is the water-gas-shift reaction, which can turn carbon monoxide and water vapour into carbon dioxide and hydrogen in the SOFC as shown in the following equation \ref{eq:WGS} \citep{SOFC_lin_analysis_2024,SOFC_WGS_Buttler2016}.

\begin{equation}
	\mathrm{CO}+\mathrm{H}_2 \mathrm{O} \leftrightarrow \mathrm{CO}_2+\mathrm{H}_2, \Delta H_{298}=-4.1 \times 10^4 \mathrm{KJ} / \mathrm{Kmol}
	\label{eq:WGS}
\end{equation}


\subsubsection{Molten Carbonate Fuel Cells (MCFCs)}

MCFCs are also high temperature cells with an operating temperature of about 600 to 700 °C and a electrolyte made out of molten carbonate (CO$_3$$^{2-}$). Like SOFCs, they can also be operated with both pure hydrogen (H$_2$) and biogas, which could also contain CH$_4$ as well as CO$_2$ and CO \citep{02_lucia2014overview, 02_wang2020fundamentals}.
As such, SOFCs and MCFCs do not require any external reformers to convert other fuel types into H$_2$ as they use the same reforming reactions from the equations \ref{eq:Steam reforming} and \ref{eq:WGS}
\citep{MCFScontreras2021molten}.
It is also possible to combine SOFC and MCFC with working temperatures of 550-700 °C which are slighlty higher than a normal MCFC at 650 °C
\citep{MCFS_cui2021review}. Furthermore, the electrical efficiency of the MCFCs can reach up to 60\% \citep{wang_preparation_2018}.


\subsubsection{Alkaline Fuel Cells (AFCs)}

For AFCs, the operating temperature is about 50-200 °C, much lower than that of the SOFC and MCFC, but much closer to the temperature in which PEMFCs are operated \citep{02_wang2020fundamentals}. Another important difference between AFCs and SOFCs is that the electrolyte in a AFC is liquid, unlike the solid state of the ceramics used in SOFCs. AFCs use a potassium hydroxide solution (KOH) as an electrolyte embedded in the matrix \citep{02_Abderezzak2018,02_wang2020fundamentals}. Unlike SOFCs and MCFC and due to the lower temperature, AFCs can be poisoned by carbon dioxide (CO$_2$). In this case, the alkaline electrolyte could react directly with the CO$_2$ leading to the following reaction equations \ref{eq:AFC_Poisoning} \citep{AFC_mclean2002assessment}:

\begin{equation}
	\begin{aligned}
		&\mathrm{CO}_2+2 \mathrm{OH}^{-} \rightarrow \mathrm{CO}_3^{2-}+\mathrm{H}_2 \mathrm{O} \text { and/or }\\
		&\mathrm{CO}_2+2 \mathrm{KOH} \rightarrow \mathrm{K}_2 \mathrm{CO}_3+\mathrm{H}_2 \mathrm{O}
	\end{aligned}
\label{eq:AFC_Poisoning}
\end{equation}

This reaction could reduce the ionic conductivity of the electrolyte and block the pores in the electrode. The carbonate shown in the equation \ref{eq:AFC_Poisoning} could also block the pores of the catalyst, resulting in the aforementioned reduction of ionic conductivity in the electrolyte
\citep{AFC_mclean2002assessment, AFC_AlSaleh1994_CO2, AFC_AlSaleh1994_Ni}.

\subsubsection{Polymer Electrolyte Membrane Fuel Cells (PEMFCs)}

For polymer electrolyte membrane fuel cells, the operating temperature may be lower than that of the other fuel cell types including the AFCs. PEMFCs can be operated at 40 to 80 °C \citep{02_lucia2014overview}. For high temperature PEMFCs, the operating temperature may even reach up top 150-180 °C \citep{02_wang2020fundamentals}. Because of its low operating temperature and high output power density, it is highly suited to mobile applications like the automotive industry. PEMFCs run on pure hydrogen as fuel and cannot use biogas containing methane (CH$_4$), carbon dioxide (CO$_2$) or carbon monoxide (CO) as they cannot reform it internally to pure hydrogen H$_2$ \citep{PEM_Atuomotive_arrigoni2022greenhouse}. 

Its mobile applications have turned the PEMFCs into one of the biggest research fields in the search for greener alternatives to conventional ICEs. They typically features a solid polymer electrolyte membrane and porous carbon electrodes with platinum functioning as catalyst \citep{01_wilberforce_developments_2017}.

Since this thesis focuses its attention on automotive applications, the following sections will provide a detailed explanation of the relevant fuel cell. However, before looking into their mode of operation, the subsequent section will first cover the relevant electrochemical fundamentals. 



\section{Electrochemical Fundamentals}

The function of a fuel cell is to transform the chemical energy stored in the fuel (hydrogen H$_2$) into electrical energy. During this electrochemical reaction, the fuel is transformed, but the fuel cell is not consumed by the energy production as in a battery. In a fuel cell, the electrochemical redox reaction is split in two between the cathode and the anode separated by an electrolyte \citep{Fundamentals_o2016fuel}.

\begin{align}
	\text{Anode:} \quad & \text{H}_2 \rightarrow 2\text{H}^+ + 2e^- \\
	\text{Cathode:} \quad & \frac{1}{2}\text{O}_2 + 2\text{H}^+ + 2e^- \rightarrow \text{H}_2\text{O} \\
	\text{Overall:} \quad & \text{H}_2 + \frac{1}{2}\text{O}_2 \rightarrow \text{H}_2\text{O}
	\label{eq:PEM}
\end{align}

In the anode, the oxidation part of the reaction takes place. Electrons are removed from the hydrogen H$_2$ as shown in the equation (2.4). In the cathode, these liberated electrons are consumed by the reduction reaction, and oxygen (O$_2$) and 2H$^+$ form water (H$_2$O) as a product. This is summarised in the overall reaction from the equation (2.6) \citep*{Fundamentals_o2016fuel}. The anodic reaction may be called a hydrogen oxidation reaction (HOR) and the cathodic reaction a oxygen reduction reaction (ORR) \citep{Fundamentals_scherer2012fuel}.

\subsection{Thermodynamics of the Cell}

In addition, the reaction produces heat (or enthalpy H), since there is a difference $\Delta$H between the enthalpy of the products and the enthalpy of the reactants \citep{Fund_barbir2008fuel}.
\begin{equation}
	\Delta H=\Delta H_{product}- \Delta H_{reactant}
	\label{eq:Enthalpy}
\end{equation}

The Gibbs free energy corresponds to the enthalpy in the reaction that can be converted to electricity. It is defined by the following equation (\ref{eq:Gibbs}). $\Delta$H is the difference of the enthalpy and T$\Delta$S expresses the losses in entropy ($\Delta$S) which are dependent on temperature
\citep{Fund_barbir2008fuel}.

\begin{equation}
	\Delta G=\Delta H-T \Delta S
	\label{eq:Gibbs}
\end{equation}

Furthermore, using the Faraday constant F (96,487 C/mol) and n for the number of electrons transferred in the reaction as well as a value for $\Delta$G for the Gibbs free energy, the reversible theoretical potential E$_{rev}$ of a cell and in standard conditions may be calculated via following equation
\citep{Fundamentals_scherer2012fuel}:

\begin{equation}
	E_{rev}=\frac{-\Delta G}{n F}
	\label{eq:E}
\end{equation}

E$_{rev}$ can also be referred to as E$^0$, the open circuit voltage (OCV) in standard conditions (1 atm and 25°C or 298K)\citep{Fundamentals_scherer2012fuel, F_omran2021mathematical}. Since this equation (\ref{eq:E}) can be used only in standard conditions, it cannot be used to calculate the reversible potential of a cell in non-standard conditions. In a non-standard case where the temperature or the pressure is different, the Nernst equation can be used
\citep*{Fundamentals_o2016fuel,Nernst_sahu2014performance}.

\begin{equation}
	E=E^0-\frac{R T}{n F} \ln \frac{\prod a_{\text {products }}^{v_i}}{\prod a_{\text {reactants }}^{v_i}}
	\label{eq:nernst}
\end{equation}

To be able to understand the Nernst equation, the concept of chemical potential $\mu$ must first be explained \citep{Nernst_mardle2021examination}. The chemical potential describes how the number of molecules or atoms $n_i$ of a species $i$ affects thermodynamic potentials. In this case, $a$ is the activity of the species, which for an ideal gas is $a_i$ = $p_i$/$p^0$. If the gas is non-ideal, it must be multiplied with $\gamma$, which describes how far away the gas is from an ideal one $(0<\gamma<1)$ with $\gamma=1$ as an ideal gas \citep{Fundamentals_o2016fuel}. In the Nernst equation, $v_i$ refers to the stoichiometric coefficient of the products or of the reactants. R is the gas constant (R = 8,314 J/molK).

\begin{equation}
	\mu=\mu^0+R \operatorname{Tln}(a)
	\label{eq:mu}
\end{equation}

The following equation may also be rewritten to describe how the chemical potential relates to the Gibbs free energy.

\begin{equation}
	\mu_i^\alpha=\left(\frac{\partial G}{\partial n_i}\right)_{T, p, n_{j \neq i}}
	\label{eq:chem1}
\end{equation}

Using the equations (\ref{eq:chem1}) and (\ref{eq:mu}), it is possible to calculate how the Gibbs free energy changes with the $i$ different chemical species, resulting in the following equation (\ref{eq:gibbs_2}):

\begin{equation}
	d G=\sum_i \mu_i d n_i=\sum_i\left(\mu_i^0+R T \ln a_i\right) d n_i
	\label{eq:gibbs_2}
\end{equation}

Finally, this equation (\ref{eq:gibbs_2}) can be inserted into the equation (\ref{eq:E}) to form the Nernst equation (\ref{eq:nernst}) in its general form\citep{Fundamentals_o2016fuel}.

\newpage

\section{PEMFC}
\label{sec: PEMFC}

Since this thesis focuses on the automotive applications of fuel cells, the focus of the following sections will shift to the Polymer Electrolyte Membrane Fuel Cell (PEMFC), which is the most widely used in automotive contexts due to its low operating temperature as well as its high output power density \citep{PEM_Atuomotive_arrigoni2022greenhouse}.
In the following section \ref{subsec:2_wayoffunct}, the PEMFC will be described in more detail, starting with its main
components and its mode of operation.

\subsection{Mode of Operation PEMFCs}
\label{subsec:2_wayoffunct}

To be able to produce more energy, PEMFCs not only use a single cell but a stack formed by hundreds of cells stacked on top of each other between two monopolar plates at each end, as shown in the left side of the following figure \ref{fig:PEMFC} \citep{PEMSchem_xu2020towards}.

%vielleicht eine selber machen? Was ist hier los mit der quelle? Bug
\begin{figure}[htbp]
	\centering
	\includegraphics[width=0.8\textwidth]{Figures/Theorie/PEMFC.pdf}
	\caption{Components of a PEMFC cell and its position in a fuel cell stack. Retrieved from Xu et al. page 816 [33].}
	\label{fig:PEMFC}
\end{figure}

Every cell is composed by two bipolar plates (BPs), each with its anode and cathode side. Between each BP is the membrane electrode assembly layer (MEA). The MEA consists of a proton exchange membrane in the middle of two catalyst layers (CL) and two gas diffusion layers (GDL), one on the cathode side and one on the anode side. Since the redox reaction in the fuel cell (equation \ref{eq:PEM}) is an exothermic reaction generating heat, the cell needs cooling. As such, the coolant can flow inside specific flow channels of the BP, as illustrated in the figure \ref{fig:PEMFC}, in order to prevent the system from overheating \citep{PEMSchem_xu2020towards}. The key components of the PEMFC and its functions will be explained in the following, starting with the BPs. Since the focus of this thesis is BP corrosion, this key component will be presented in more detail.

\newpage
\subsubsection{Bipolar Plate (BP)}

BPs have various functions. One important function is to distribute fuel on the anode side and oxidant on the cathode side to the reactive sites in the catalyst layer (CL). It also collects the current generated and removes byproducts from the reaction. Heat management is also a very important; special channels (flow channels) transport the coolant and remove the heat from the cell \citep{PEM_baroutaji2015materials}. 

Since BPs are responsible for 60-80 \% of the weight and 20-30 \% of the total cost of the fuel stack, the materials used have been under investigation for some time, and the Department of Energy (DOE) has set up targets for the components of the fuel cell \citep{doe_pemfc_targets}. These targets evaluate not only the cost of the materials but also durability and performance \citep{doe_pemfc_targets}. Due to their durability, excellent mechanical strength, high power density and electric conductivity, investigations into new BP materials have been primarily focused on stainless steels, titanium alloys and aluminium alloys \citep{antunes2010}. In the past, BPs have been made out of graphite. Graphite is highly corrosion resistant, but it unfortunately has drawbacks like high gas permeability and high production costs \citep{PEM_baroutaji2015materials}. 

Stainless steels, on the other hand, are more cost effective and versatile; their high mechanical strength and malleability makes possible the production of thinner BPs, leading to a weight reduction of up to 40 \% of the fuel cell stack 
\citep{SSweight_li2005review}. These optimal characteristics have attracted the automotive sector and companies such as Hyundai, GM and Honda, which have all produced fuel cell vehicles (FCV) with these stainless steels \citep{Automotive_leng2020}. Toyota, however,
uses a titanium bipolar plate nano composite (NC) as a surface treatment for the BPs used in the Mirai stack. This has, for example, ensured a reduction in the thickness of the Titanium (Ti) plates so that the Platinum (Pt) used in the stack could be reduced by 58 \% from its 2008 model to the second-generation Mirai \citep{toyota_technical_review_2021}. Since the collaboration between BMW and Toyota was announced in September 2024, BMW will also be using BPs manufactured by the Toyota Motor Corporation for its BMW iX5 Hydrogen, planned for series production in 2028 \citep{bmw_hydrogen_2024}.

Stainless steel BPs have a lower cost than Ti plates, but their durability has been questioned, since their corrosion resistance is lower than Ti and aluminium plates. When stainless steel plates corrode, they release metal ions like Fe$^{2+}$ which lead to an accelerated chemical degradation of the membrane by contaminating the MEA. \citep{eom2012}. Even though aluminium has a higher corrosion restistance than stainless steel, it would release Al$^{3+}$ ions during the corrosion process which have an even larger effect than Fe$^{2+}$ on the fuel cell catalyst \citep{sulek2011}. This form of degradation will be explained in Section \ref{sec:Degradation}. Although SS316L has been proven to have a higher corrosion resistance and is therefore used as reference material for bipolar plates, other, more cost effective options have also been studied such as stainless steels 310L, 304 and 904L\citep{papadias2015degradation,feng2011}.

\subsubsection{Membrane Electrode Assembly (MEA)} 

As previously mentioned, the Membrane Electrode Assembly (MEA) is conformed by the Gas Diffusion layer (GDL) on the cathode side, as well as the GDL on the anode side, followed by the two catalyst layers (CL) which contain platinum (Pt) and Nafion. In the middle is the proton exchange membrane (PEM) \citep{PEMSchem_xu2020towards}.
The MEA could be considered as the most important component of a PEMFC, as it is responsible for the chemical reactions, and consequently for the fuel cell's performance \citep{MEA_lim2021comparison}. 

Three methods of fabricatoin of MEA stand out due to their perfomance. They are catalyst-coated membranes (CCMs), catalyst-coated substrates (CCSs) and catalyst-coated electrodes (CCE) \citep{MEA_lapicque2012,MEA_bhosale2020}. 
 A comparative study by Bhosale et al. \citep{MEA_bhosale2020} showed that the most effective method is CCE. Nevertheless, CCM shows high performance and many studies suggest that MEA produced with a CCM method can have many advantages over CCS and CCE. Consequently it is the method most frequently used \citep{PEM_MEA_parekh2022recent}. 
 
 CCM can lead to an increase in total reactions in the MEA, as well as reduce the amount of Pt in the catalyst \citep{MEA_lim2021comparison}. Since Pt can be very expensive, methods to reduce the weight-loading percentage of the catalyst have also been under investigation, and even Pt-free catalysts have been investigated \citep{Pt_liew2014}. Moreover, Pt catalysts can be lost during dynamic operation of the cell (voltage cycling) either by Pt agglomeration or Pt dissolution, also leading to further degradation of the cell and higher mass transport losses and activation losses \citep{thiele2024realistic}. In order to better understand the MEA, its components will now also be explained.


\subsubsection{Gas Diffusion Layer (GDL)}

Starting with the first layer of the MEA, right between the BPs and the catalyst layers (CL) on each side are the gas diffusion layer (GDL), as well as the microporous layer (MPL) \citep{PEMSchem_xu2020towards}. Their primary functions are to offer mechanical support to the MEAs, ensure the flow of the reactants and also remove products. Furthermore, they must enable electron conduction between the CLs and the BPs on each side. 

Since its main functions are related to diffusion, the GDL is made out of porous materials. Typically, it is made out of carbon paper \citep{02_wang2020fundamentals}.
To enhance water management and prevent flooding in the electrode, the carbon paper GDL has to be hydrophobic. For this reason, Polytetrafluorethylene (PTFE) is often added as treatment to achieve this hydrophobicity
\citep{02_wang2020fundamentals,GDL_zamel2011}. Since a high PTFE load can cause an obstruction in the pores of the GDL, and consequently cause mass transport limitations, it is crucial to add the right amount. Studies have shown that 20 wt. \% is an optimal load percentage for PTFE to turn the carbon paper hydrophobic without blocking the pores \citep{GDL_zamel2011}. Also important are the capillary effects of the MPL. Since the MPL is also hydrophobic, it provides great drainage as well as stable gas and electron channels \citep{ijaodola2019}. This helps the overall performance of the cell by reducing flooding, since this layer sits between the GDL and CL \citep{majlan2018}.

\subsubsection{Catalyst Layer (CT)}

The catalyst layers (CTs) are positioned between the PEM and the MPL on both sides. Electrochemical reactions take place here in the CL and, therefore, it must provide continuous pathways for the different reactants. More specifically, it must provide a route for proton transport; its porous structure must supply the gaseous reactants to the site whilst removing water, while also being able to form a conductive pathway for electrons between the CL and the current collector \citep{02_wang2020fundamentals}.

The oxygen reduction reaction (ORR) takes place at the CL and is the most critical process for a PEMFC at the anode \citep{PEM_baroutaji2015materials}. This reaction heavily relies on the platinum (Pt) catalyst. By increasing the platinum load, the ORR rate can be enhanced leading to a higher power output in the cell \citep{PEM_MEA_parekh2022recent}. In contrast to the GDL and MPL, platinum particles are not hydrophobic, since they present hydrophilic capabilities \citep{CT_malek2011}. 

As previously before, the CCM method is currently the most frequently used in the production of the MEA. This is because Pt is the most expensive part of the production, and CCM production already has a better electrochemical performance than the CCS method and exhibits a lower Pt load\citep{hnat2019}. In this method, the catalyst layer is produced by applying catalyst ink onto a PEM \citep{MEA_lim2021comparison}. However, the search for a more cost effect alternative to Pt catalyst with the same electrochemical performance continues to be a challenge \citep{PEM_MEA_parekh2022recent}.

The catalyst layer consists of a catalyst (Pt), carbon hsupport, ionomer and a void space. PTFE was substituted with recast Nafion ionomer as a binder, allowing for a significant reduction in Pt loadings. Additionally, Pt supported on carbon (Pt/C) also helps to decrease the metal content \citep{ink_zamel2016catalyst}. The ionomer serves a dual purpose as a binder for Pt/C particles and as a proton conductor. An imbalance in ionomer loading can lead to transport or ohmic losses, which will be discussed in subchapter \ref{subsec:losses}. Insufficient ionomer diminishes proton conductivity, and an excessive amount can increase resistance of gaseous reactant transport \citep{02_wang2020fundamentals}.




\subsubsection{Proton Exchange Membrane (PEM)}

The proton exchange membrane is the heart of the PEMFC. It is right in the middle of the cathode and anode, followed by CL, MLP and then GDL, in that order from inside to outside. The PEM has two primary functions. The first is to server as a barrier. It prevents the mixing of reactant gases and electrons between anode and cathode. The second is to facilitate proton conduction from the CL on the anode to the CL on the cathode side. Furthermore, the PEM is impermeable to gas. It stops oxygen and hydrogen crossover and must be electrically insulating. Another requirement for the membrane is exceptional chemical and mechanical stability in order to endure the harsh operating conditions of the PEM fuel cells \citep{ghassemzadeh2010chemical}.

The most widely used material for the membrane in a PEMFC is perfluorosulfonic acid (PFSA), also referred to as \textit{Nafion}, developed by DuPont \citep{PEM_MEA_parekh2022recent}. Figure \ref{fig:Nafion} shows the chemical structure of Nafion \citep{okonkwo2021nafion}.


\begin{figure}[htbp]
	\centering
	\includegraphics[width=0.5\textwidth]{Figures/Theorie/Nafion.pdf}
	\caption{Chemical structure of PFSA, also called Nafion. Retrieved from Chen et al., page 1436 (1) [59]}
	\label{fig:Nafion}
\end{figure}

 Since the main chain is Teflon-like, one side is hydrophobic and the sulfonic acid groups on the side chains are hydrophilic. This is a great advantage because it facilitates water adsorption and consequently proton conduction. To maintain an effective proton transport, proper hydration of the membrane is vital, while simultaneously avoiding excessive moisture that could lead to flooding in the CL and GDL \citep{zaidi2009polymer}.
 
  However the membrane can degrade when it is exposed to low humidity and high temperatures, while degrading Nafion can release F$^-$, CO$_2$, SO$_4^{2-}$, SO$_2$ as well as fluorocarbons \citep{teranishi2006}. Besides this form of degradation, the electrochemical reaction in a PEMFC can also produce hydrogen peroxide (H$_2$O$_2$) when the entry of oxygen in the PEM reacts with the hydrogen in the anode, as shown in the following equation (\ref{eq:h2o2}) \citep{ren2020degradation}:
  
\begin{equation}
 	\mathrm{O}_2+2 \mathrm{H}^{+}+2 e^{-} \rightarrow \mathrm{H}_2 \mathrm{O}_2, \mathrm{E}_{\mathrm{O}}=0.695 \mathrm{~V} \text { vs. } \mathrm{SHE}\left(75^{\circ} \mathrm{C}\right)
 	\label{eq:h2o2}
 \end{equation}
 
 
 Furthermore, H$_2$O$_2$ in the presence of ferrous ions like Fe$^{2+}$, which are released by the BP when corroding, can trigger the formation of hydroxyl radicals that also attack the membrane. It is thought that an incomplete reduction of the oxygen by the Pt catalyst can also trigger the production of H$_2$O$_2$ \citep{elferjani_coupling_2021}.



\subsection{Department of Energy Targets}
\label{subsec:2_DOE}

The Deparment of Energy (DOE) of the United States(U.S.) with the help from the U.S. DRIVE partnership has set targets for the components of PEMFC to help FC developers develop them without needing to test the full system
\citep{doe_pemfc_targets}. The U.S DRIVE FC team aims to develop a PEMFC system for transportation that is able to resist 8000 hours and with a mass production cost of 35\$ per Kilowatt (kW) by 2025 \citep{trabia2016}.

Targets of the DOE include specifications for the MEA, PEM, electrocatalysis and bipolar plates. The goal for BPs is to reduce the plate cost from 5,4\$ to 2\$ per kW by 2025. Some other goals include the weight reduction of the BPs and increased corrosion resistance, as well as higher electric conductivity \citep{PEM_MEA_parekh2022recent}. Overall, the DOE intends to increase cell performance and simultaneously reduce production costs to allow PEMFCs in FCV and fuel cell electrical vehicles (FCEV) to become a cost effective and green alternative to ICE in the series production.


\subsection{Overpotentials of the PEMFC}
\label{subsec:losses}


\begin{figure}[htbp]
	\centering
	\includegraphics[width=0.8\textwidth]{Figures/Theorie/Polarization.pdf}
	\caption{Polarisation curve of a fuel cell including the different losses. Retrieved from Jung et al., page 741 (4) [64].}
	\label{fig:losses}
\end{figure}

The Nernst equation \ref{eq:E} calculates the reversible cell potential, which is the current that should be drawn by the PEMFC. However the actual open-circuit voltage (OCV) measured is lower than the theoretical one calculated by the equation \citep{Loss_mardle2021examination}. 
This deviation can be observed in the figure \ref{fig:losses} \citep{Loss_jung2010dynamic}. This first deviation is caused by the hydrogen (H$_2$) crossover. This occurs when H$_2$ diffuses through the membrane, leading to a mixed potential that lowers the overall open-current potential (OCP). Internal short circuits also lead to OCV losses in this stage \citep{Loss_mazzeo2024assessing}.

In addition to the first loss, the polarisation curve experiences other deviations as the current density begins growing, starting with the activation losses, then the Ohmic losses and at high current densities the mass transport losses. These will all be explained in the following \citep{02_lucia2014overview}.


\subsubsection{Activation Polarisation}

The activation loss, also called activation polarisation loss, is driven by the voltage loss caused by the activation energy required for the electrochemical reaction to start as the protons move through the reaction interface. Therefore, it is a loss related to the kinetic of the cathode and anode electrodes \citep{Loss_li2022new}. As shown in the image \ref{fig:losses}, it takes place at a region with low current densities. It$\eta_{act}$ can be calculated using the following equation (\ref{eq:Loss_N}) \citep{ren2020degradation}.

\begin{equation}
	\eta_{\text {act }}=\frac{R T}{\alpha n F} \ln \left(\frac{i_{\text {loss }}}{i_0}\right)
	\label{eq:Loss_N}
\end{equation}

In this equation the following parameters are taken into account: F for the Faraday constant, $i_0$ as the exchange current density for the active area of the FC, $\alpha$ as charge transfer coefficient, R is the gas constant and n, the number of molecules or atoms. Furthermore, $i_{loss}$ is formed as the addition of $i_{short}$ for the current density of short-circuits and $i_{crossover}$, the gas-crossover current density \citep{ren2020degradation,jouin2016}.

\begin{equation}
	i_{\text {loss }}=i_{\text {crossover }}+i_{\text {short }}
\end{equation}

It is worth mentioning that the exchange current density of the oxygen reduction reaction (ORR) can be perceived as a limiting factor in low temperature PEMFC such as the ones used for the automotive sector \citep{Loss_mazzeo2024assessing}. 


\subsubsection{Ohmic Polarisation}

As the current density increases, ohmic polarisation loss becomes the dominant factor in the polarisation curve. Voltage decreases in an almost linear way with increasing current density \citep{Loss_li2022new}. Ohmic losses are associated with the resistance encountered by the flow of the electrons through various components of the FC \citep{Loss_mazzeo2024assessing}. The resistance of the hydrogen ion flow into the electrolyte is a significant factor. This resistance is heavily influenced by membrane's hydration level as well as operating temperatures and current density \citep{springer1991}. Mathematically it can be described with the following equation (\ref{eq:Loss_ohm})\citep{ren2020degradation}.

\begin{equation}
	\eta_{\mathrm{ohm}}=\left(R_{\mathrm{ion}}+R_{\mathrm{ele}}+R_{\mathrm{con}}\right) \cdot i
	\label{eq:Loss_ohm}
\end{equation}

In this equation R$_{ion}$ represents the ionic resistance, R$_{con}$ the contact resistance and R$_{ele}$ the electronic resistance. For this section, the polarisation behaves linearly since it is multiplied with the current density ($i$) \citep{ren2020degradation}.

\subsubsection{Concentration Polarisation}

At high current densities, concentration polarisation or concentration loss occurs. The reactants are consumed very quickly during the electrochemical reactions at a high current density. Because of transport and diffusion resistance, the availability of the reactants at the reaction sites decreases, limiting the reactions and thereby the efficiency of the PEMFC \citep{Loss_li2022new}.
The ohmic polarisation can be calculated using the following equation (\ref{eq:Loss_con}) 
\citep{ren2020degradation}:

\begin{equation}
	\eta_{\text {con }}=\left(1+\frac{1}{\alpha}\right) \frac{R T}{n F} \ln \left(\frac{i_{\mathrm{L}}}{i_{\mathrm{L}}-i}\right)
	\label{eq:Loss_con}
\end{equation}

The parameters of the equation (\ref{eq:Loss_con}) are the same as in those previously discussed, the only new one being $i_L$, which stands for the limiting current density.

At such high current densities it is important to avoid undersupply of the anode, which could damage the PEMFC. Therefore, changing the hydrogen-oxygen stoichiometric ratio from 1:1 to 1.5:2.2 can improve the performance of the PEMFC, as well as reduce damage caused by the operation on high current densities \citep{liu2024study}.


\subsection{Characterisation of PEMFC}
\label{subsec: Polarizaiton}

As this thesis includes an endurance run and preliminary investigations to better understand the operating conditions of the PEMFC and identify the optimal point for triggering cell corrosion, this section will detail the in-situ methods employed to characterise the cells. Parameters such as cell potential and current density can provide insight into the state of the cell's health. By using predefined characterisation curves between a specific number of voltage cycles, the cell degradation may be tracked. Figure \ref{fig:PolCurve} shows an example of the polarisation curves after a specific number of voltage cycles \citep{mohsin2020electrochemical}.

\begin{figure}[htbp]
	\centering
	\includegraphics[width=0.7\textwidth]{Figures/Theorie/PolCurve.pdf}
	\caption{Example of a polarisation curve of a PEMFC after different numbers of voltage cycles (VC)  . Retrieved from Mohsin et al., page 24096 (4) [69].}
	\label{fig:PolCurve}
\end{figure}


In these polarisation curves, the potential of the cell is plotted over the current densities. Degradation of the membrane, corrosion, carbon corrosion or, as a consequence thereof, platinum catalyst dissolution, all cause the polarisation curve to have a higher drop in the potential at much lower current densities after more cycles, as shown in the aforementioned figure \ref{fig:PolCurve} \citep{Pol_thiele2024realistic}. These mechanisms will be explained in the following section \ref{sec:Degradation}. After a larger number of voltage cycles the higher current densities can no longer be reached as a consequence of the degradation as well as larger activation, ohmic and concentration losses in the cell \citep{mohsin2020electrochemical}.


\section{Degradation Mechanisms}
\label{sec:Degradation}

Automotive conditions can be very stressful for the PEMFC. Start-stop procedures, idling conditions, operation at maximum power, as well as quick changes from full power to stopping can all speed up the degradation progress of the cell and consequently shorten its lifespan \citep{pei2008}. Accelerated stress tests (AST) are one way components of a PEMFC may be tested in a controlled environment without them needing to be in the actual vehicle. It can shorten the test duration by accelerating the degradation processes and simulating different conditions and automotive scenarios \citep{Pol_thiele2024realistic}. This section will provide information on a few of the most important degradation mechanisms that can be found in a PEMFC, such as platinum catalyst dissolution, membrane degradation, carbon corrosion and finally corrosion in general . It is also important to mention that there are many more mechanisms that may contribute to the degradation of the fuel cell, and that these mechanisms all impact one another\citep{Pol_thiele2024realistic}.

\newpage


\subsection{Platinum Catalyst Dissolution and Agglomeration}
\label{subsec: Pt}

Carbon-supported platinum nanoparticles in the CL of the PEMFC increase the oxygen reduction reaction (ORR) at the cathode making the cell more efficient. With such a crucial task it is of utmost importance to understand the degradation mechanism. Studies have shown that corrosive acidic environments in the PEMFC under a positive potential can lead to the dissolution of platinum, causing a reduction in the catalyst performance \citep{cherevko2015}.
Pt loss during PEMFC operation is a major contributor to the degradation of the CL. This is driven by processes such as platinum dissolution, Pt-detachment, Pt-migration and Pt-agglomeration. In a study by Luo et al., a 10 cell stack was operated for 200 hours at a temperature of 60 °C. When analysed, the stack showed a reduction from an initial Pt content of 20 \% to 13,5 \% \citep{luo2010}.

 While Pt can remain stable at potentials below 1,188 V, at high cell voltages and OCV direct electrochemical dissolution may occur at the cathode. During normal operation conditions or load cycling, the dissolution of Pt is more likely to occur \citep{wallnofer2024main}. Lower electrode potentials as well as voltage cycling can cause Pt oxide dissolution which can be described by the following equations \citep{takei2016}:


\begin{equation}
	\mathrm{Pt} \rightarrow \mathrm{Pt}^{2+}+2 e-\mathrm{E}_0=1.188 \mathrm{~V}
\end{equation}
\begin{equation}
\mathrm{Pt}+\mathrm{H}_2 \mathrm{O}+2 e^{-} \rightarrow \mathrm{PtO}+2 \mathrm{H}^{+} \quad \mathrm{E}_0=0.98 \mathrm{~V}
\end{equation}
\begin{equation}
	\mathrm{PtO}+2 \mathrm{H}^{+} \rightarrow \mathrm{Pt}^{2+}+\mathrm{H}_2 \mathrm{O}
\end{equation}

Since water is produced in the reaction (2.21) the higher water content in the ionomer leads to a greater mobilty of the dissolved Pt ions, which can facilitate the Ostwald ripening of the particles beneath it \citep{takei2016}.

Platinum migration is another problem that can degrade the PEMFC by loss of CL performance. The Pt particles may diffuse into the ionomer phase and then precipitate within the membrane (PEM). Furthermore, hydrogen migrating from the anode to the cathode can reduce the Pt ions, forming Pt$^{2+}$ and Pt$^{4+}$. This can again cause the oxidation of Pt to PtO as shown in the previous reactions, and subsequently decrease the cell performance due to the lingering oxygen \citep{pavlivsivc2018platinum,okonkwo2021platinum}. The agglomeration process can facilitate the formation of oxygenated functional groups on the carbon surface which then lead to an increased hydrophilicity of the carbon support. This altered hydrophilicity can influence the displacement of oxygen towards the Pt by controlling the flooding in the CL. Flooding can limit oxygen access to the active reaction sites within the CL and hence decrease the efficiency of the PEMFC \citep{okonkwo2021platinum}.

Losses of activity in the reaction sites can be categorised into two groups. The first being the unrecoverable losses and the second being those that are recoverable. Pt detachment as well as agglomeration, dissolution, carbon corrosion and Pt redeposition are all associated with the first group, the unrecoverable losses. Start and end scenarios expose the cell to very rapid changes in these parameters. Additionally, operating under extreme conditions can accelerate the degradation of the cell and favour the aforementioned mechanisms. 
 \citep{okonkwo2021platinum}. The recovery loss was linked either to the reduction of platinum oxide or the removal of carbon monoxide produced via carbon corrosion \citep{okonkwo2021platinum} .

There is also a second way of classifying the degradation mechanisms. Since one degradation mechanism can trigger or favour another, they may also be classified as primary or secondary, depending on their ability to start or intensify another mechanism. For example, carbon corrosion is a primary mechanism, since it can be responsible for Pt agglomeration and detachment, leading to an increase in degradation \citep{okonkwo2021platinum}.

\subsection{Electrochemical Carbon Corrosion}
\label{subsec: Carbon corrosion}

Since carbon corrosion is a primary degradation mechanism for PEMFCs, it is essential to examine this process in greater detail. A deeper understanding of electrochemical carbon corrosion can provide insights into its impact on other mechanisms, and especially on the performance and durability of the fuel cell.

Studies have shown that start-stop cycles of fuel cells primarily initiate surface corrosion of the carbon support in the CL. Repeated cycles of start-stop conditions can modify the crystalline carbon which will be transformed into a more corrosion-prone amorphous carbon \citep{park2016effects}. Due to the carbon corrosion, the CL experiences a loss of thickness, resulting in detachment of the Pt particles, especially at the cathode. This mechanism is not only intensified by start-stop cycles, but also by square wave cycles and triangular wave cycles. The former triggering the corrosion on its surface and internally, while the latter targets surface defects \citep{zhao2021carbon}. It is also worth mentioning that in the results of an AST in which the cell was exposed to various conditions, like load change cycles and start-stop cycles, the decreasing perfomance could be attributed by one third to the shutdown and startup cycles, proving how demanding this step can be for a cell and how it can increase the degradation \citep{zhao2021carbon,lin2015investigating}. Dependening on the cell conditions, either one of the following three reactions can lead to carbon corrosion \citep{wallnofer2024main}.

\begin{equation}
	\mathrm{C}+2 \mathrm{H}_2 \mathrm{O} \rightarrow \mathrm{CO}_2+4 \mathrm{H}^{+}+4 e^{-} \mathrm{E}_0=0.207 \mathrm{~V}
\end{equation}
\begin{equation}
	\mathrm{C}+\mathrm{H}_2 \mathrm{O} \rightarrow \mathrm{CO}+2 \mathrm{H}^{+}+2 e^{-} \mathrm{E}_0=0.518 \mathrm{~V}
\end{equation}
\begin{equation}
	\mathrm{CO}+\mathrm{H}_2 \mathrm{O} \rightarrow \mathrm{CO}_2+2 \mathrm{H}^{+}+2 e^{-} \mathrm{E}_0=-0.103 \mathrm{~V}
\end{equation}

Another effect of carbon corrosion is the reduction in the hydrophobicity of the GDL, which could be attributed to the loss of PTFE and its hydrophobic properties if the cell is flooded while operating at a high current \citep{pei2008}. As mentioned previously, carbon corrosion may be increased not only by start-stop conditions but also by high potentials. This can then be observed in the polarisation curve, causing a larger activation loss at low current densities \citep{Pol_thiele2024realistic}. 

Due to more hydrophilic behaviour due to carbon corrosion and loss of PTFE, the membrane responds slowly to quick changes from high to low load, resulting in water accumulating on the anode side and therefore a reduced hydrogen supply. Furthermore, the pressure difference between inlet and outlet on the anode prevents the water from being removed. As a consequence of this, when the load is increased again quickly, the anode can suffer from partial hydrogen starvation, again intensifying the carbon corrosion due to the high electrode potential forming on the cathode side \citep{Pol_thiele2024realistic}.

Lastly, the activation polarisation is affected by the carbon corrosion and the ohmic loss is increased. This increase in the ohmic loss results from the decrease in activity of the Pt catalyst, since the loss in thickness from the CL and its consequent detachment of Pt particles at the cathode weaken its activity \citep{ren2020degradation}. To be more precise, the degradation of the porous structure in the CL causes extended pathways for electrons, increasing the contact resistance of the PEMFC \citep{wallnofer2024main}.

\subsection{Membrane Degradation}
\label{subsec:membrane degradation}

As membrane degradation plays a significant role in the performance of the cell, it is important to understand how it works and how the PEMFC can be affected by these mechanisms. Since the polyelectrolyte membrane (PEM) is formed by perfluorsulfonic acid (PFSA), also known as Nafion, it is important to look at its degradation mechanism \citep{okonkwo2021nafion}.

 The chemical degradation of the PFSA ionomer is linked to the membrane decay and can lead to pinhole formations. It is driven by hydrogen peroxide radicals, which are known to be formed at potentials below 0,682V and in acidic environments \citep{wallnofer2024main}. Since the membrane transports water (H$_2$O), protons (H$^+$) and oxygen (O$_2$) from the anode to the cathode, it is possible that the Pt in the CL catalyses the reaction of the oxygen with the protons from the hydrogen (H$^+$), forming hydrogen peroxide (H$_2$O$_2$) as an intermediate product \citep{frensch2019impact}. This can be described using the following reaction equation (\ref{eq: h2o2}) \citep{ruvinskiy2011using} .
 
 \begin{equation}
 	\mathrm{O}_2+2 \mathrm{H}^{+}+2 \mathrm{e}^{-} \rightarrow \mathrm{H}_2 \mathrm{O}_2
 	\label{eq: h2o2}
 \end{equation}

The PFSA membrane can also suffer from chemical degradation caused by an attack of free radicals. Hydroxyl radicals (OH), hydroperoxyl radicals (OOH) and hydrogen radicals (H) can be the species responsible for this attack to the membrane, again possibly leading to the formation of pinholes in the membrane, therefore causing its failure \citep{ren2020degradation}. With the presence of metal ions like Fe$^{2+}$ or Cu$^{2+}$, the formation of radicals from hydrogen peroxide can be catalysed. This mechanism is called the Fenton reaction and can be seen in the following equation
\citep{frensch2019impact, ruvinskiy2011using}.

\begin{equation}
	\mathrm{H}_2 \mathrm{O}_2+\mathrm{Fe}^{2+} \rightarrow \mathrm{Fe}^{3+}+\mathrm{HO}^{+}+\mathrm{HO}^{-}
\end{equation}

It is believed that metal components like the BPs made out of stainless steels can release metal ions which can then travel to the membrane and either stay in it or be transported out by one of the outlets \citep{elferjani_coupling_2021}.
Hydrogen peroxide radicals released by the Fenton reaction cause degradation at the weaker points in the ionomer. This could be, for example, the functional end groups or C-H bonds in the PTFE chain which sometimes arise from the manufacturing process as well as substituted C-F bonds. The attack leads to either breakdown of the ionomers main or side chain or to elimination of the end groups, subsequently accelerating the degradation process of the membrane and the PEMFC \citep{wallnofer2024main}.

Pt dissolution, as mentioned before, can also have a huge impact on the membrane degradation. Studies have shown that Pt band formation, although minimal during open circuit conditions, can also degrade the membrane. However, Pt particles which infiltrate into the membrane may also act as a catalyst for direct generation of OH free radicals bypassing the intermediate formation of H$_2$O$_2$ \citep{ohma2008}.

Membrane degradation can have a series of devastating consequences in the PEMFC, such as the formation of pinholes, which lead to a high gas crossover rate and as such to high voltage losses. or even reversal of current in specific cells \citep{Weber_2008}. The loss of functional end groups in the ionomer is also known to increase the membrane resistance and to change water management properties of the membrane \citep{wallnofer2024main}.

\subsection{Corrosion}
\label{subsec: BP Corrosion}

The use of stainless steel in metallic bipolar plates (BPs) has become increasingly common in PEMFCs due to its low cost and excellent mechanical and electrical properties. However, the implementation of stainless steels as a BP material has raised some questions as to how its corrosion may affect the durability of FCs, as automotive conditions have been known to accelerate the corrosion rate and surface destruction \citep{Corr_ren2022corrosion}. The many different types of corrosion include uniform corrosion, galvanic corrosion, interangular corrosion, crevice corrosion and pitting corrosion \citep{jones1996principles}.

Before discussing the types of corrosion that may occur on metallic BPs, it is important to first explore the various metals used for their construction. Understanding these materials can provide insight into how specific corrosion mechanisms can impact them and in turn damage the PEMFC. Because of their higher corrosion resistance and great properties, stainless steels like 304L, 316L and 904L have been under investigation. The composition of these three stainless steels is very similar since they are made of iron (Fe), chromium (Cr) and nickel (Ni) which could all contaminate the MEA \citep{novalin2023demonstrating}. Furthermore, the iron in the aforementioned stainless steels can catalyse the Fenton reaction seen in equation (\ref{eq: h2o2}) leading to chemical degradation and the formation of pinholes in the membrane \citep{ruvinskiy2011using,novalin2023demonstrating}.

\subsubsection{Stainless Steel 316L}

 Stainless steel 316 differentiates itself from 304 because of the added molybdenum (Mo), which reinforces its corrosion resistance and offers a higher protection against mechanisms like pitting and crevice corrosion \citep{novalin2023demonstrating}.

\begin{figure}[htbp]
	\centering
	\includegraphics[width=0.6\textwidth]{Figures/Theorie/SS316L.pdf}
	\caption{Comparison of polarisation curves at 70 °C and ambient temperature for stainless steel 316L. Retrieved from Wang et al. page 60 [89].}
	\label{fig:SS316L}
\end{figure}

In a study performed by Wang et al., the electrochemical behaviour of stainless steel 316L was tested in a potentiodynamic test in 0,5M H$_2$SO$_4$ with a potential reaching from -0,1V to 1,2V with a scanning rate of 1mV/s at room temperature and 70 °C, as shown in figure \ref{fig:SS316L}
\citep{Corr_Mat_wang2010electrochemical}. This polarisation curve can be divided into three different parts\citep{Corr_Mat_wang2010electrochemical}:
\begin{enumerate}
	\item Active region: OCP to -0,15V.
	\item Passive region: -0,15V to 0,9V.
	\item Transpassive region: 0,9V to 1,2V
\end{enumerate}

Since the curve at high temperatures shows a higher current density, a higher operation temperature of a PEMFC can also be associated to a higher corrosion rate. Furthermore, the formation of the passive region shows that the Cr in 316L is able to produce a passive film that inhibits further corrosion until the transpassivation is reached with a higher potential \citep{Corr_Mat_wang2010electrochemical}. Although the corrosion is enhanced at the passive region with the formation of an oxide layer, since this layer is less reactive it can also contribute to the performance degradation of the PEMFC
\citep{laedre2017materials}.

Startup and shutdown conditions in the PEMFC can lead to an increase in the cathode potential, which results in the potential being at the transpassivation region \citep{Corr_ren2022corrosion}. Cycling between the transpassivation region and the passive or passivation region causes the dissolution of Cr species as well as Fe species leading to extensive structural damage, also causing nonuniformity scratches and defects in the surface of the BP\citep{Corr_ren2022corrosion}. The cathode environment, because of the contact with oxygen as well as the water produced at the outlet, hosts an environment that is favourable to the accumulation of metallic elements \citep{Corr_kumagai2012high}.


\subsubsection{Pitting Corrosion}

Pitting corrosion is a mechanism which causes localised depassivation. This usually happens at vulnerable surface sites like defects, grain boundaries or impurities \citep{novalin2023demonstrating}. Additionally, changes in the pH can alter the composition of the passivation layer. In the presence of fluoride (F$^-$), this process is intensified, leading to more severe pitting corrosion which causes the corrosion current density to increase. This can be detected by the density of pits formed on the surface \citep{Corr_ren2022corrosion}. Once one or more pits are initiated, the material undergoes rapid dissolution, further compromising the integrity of the material \citep{elferjani_coupling_2021}.

The F$^-$ is most commonly encountered in the FC environment due to the membrane degradation (or PFSA). High concentrations of F$^-$  may come from localised evaporations of water droplets, which consequently increase the corrosiveness of the run-off water and overall lead to a more severe degradation of the PEMFC \citep{talbot2018corrosion}. Stainless steel plates are sensitive to changes in temperature, humidity and pH. Therefore, a change of any of these parameters can have a big influence on its corrosion resistance. Furthermore, 316L has a known depassivation at pH levels ranging from 1,5 to 2 \citep{elferjani_coupling_2021}. Consequently, conditions in PEMFCs create an environment that may promote pitting corrosion \citep{novalin2023demonstrating}.



\subsubsection{Crevice Corrosion}

Crevice corrosion has a similar effect to pitting corrosion but, unlike pitting corrosion, it is driven by the geometric features of the components that lead to the creation of highly corrosive micro-environments \citep{talbot2018corrosion}. This type of corrosion may be found in the flow field of the BPs depending on its design. As mentioned before, the molybdenum in stainless steel 316L provides a higher resistance against crevice corrosion \citep{novalin2023demonstrating}.


\subsubsection{Interangular Corrosion}

Intergranular corrosion typically occurs along the boundaries of the grains of the stainless steel alloys and it therefore is often associated with the welding process \citep{talbot2018corrosion}. In this mechanism, the boundary acts as the anode, while the surrounding metal serves as the cathode. Due to the big size difference from this anode to the cathode, a rapid and concentrated attack on the metal takes place \citep{pe2009fundamentals}. This process leads to significant localised degradation of the material \citep{pe2009fundamentals}.

\subsubsection{Galvanic Corrosion}

Galvanic corrosion can occur when two different metals are in electrical contact in a conductive corrosive environment \citep{al2016modeling}. In this case the driving force is the potential difference between the two metals, and the more active metal will act as the anode and corrode. Meanwhile, the more noble metal will function as the cathode and be protected from degradation \citep{al2016modeling}. Since galvanic corrosion leads to degradation from the anodic metal it will also trigger pitting corrosion \citep{saeed2013effect}.

\subsubsection{Uniform Corrosion}

In uniform corrosion, the material corrodes consistently across the entire metal surface and leads to the gradual thinning of the material over time, weakening its structure \citep{pe2009fundamentals}. The material must be in contact with the corrosive environment with equal access to the entire area for it to be degraded evenly \citep{pe2009fundamentals}.

\subsubsection{Effects of Corrosion}

In addition to the degradation of the BP, it is important to consider that the corrosion reaction releases metal ions into the cell which may lead to the contamination of other FC components in the PEMFC. For example, metal ions from Fe or Cr can migrate throughout the cell and potentially poison the MEA, potentially further contributing to loss in performance and degradation of the cell \citep{low2024understanding}. In a study performed by Mele et al., a high accumulation of Fe was found in the MEA and especially in the GDL \citep{Corr_mele2010localised}.

Structural changes in the BPs, such as variations in shape and thickness caused by corrosion, can affect the gas flow and disrupt the water management of the cell, leading to cell flooding or air starvation, dramatically shortening the lifespan of the PEMFC \citep{low2024understanding}. Pinholes or cracks in the BPs due to corrosion can also lead to gas crossover as well as liquid leakage, possibly destroying the cell \citep{low2024understanding}. 

The cathode CL is susceptible to corrosion as a consequence of Pt disintegration or dissolution, which is particularly pronounced during fatigue because of voltage cycling or even high potentials applied to the anode electrode \citep{matsutani2010}.