Browse Source

update

method
Daniel Ledda 2 months ago
parent
commit
5fe08a8ea6
1 changed files with 75 additions and 71 deletions
  1. +75
    -71
      Content/Method.tex

+ 75
- 71
Content/Method.tex View File

@@ -1,58 +1,58 @@
\chapter{Method}
\label{chap:Methode}

This chapter will provide a comprehensive overview of the methods used in this Thesis. Starting with the In-Situ methods in chapter \ref{sec: M_Setup} where the hardware will be explained in a more detail way as well as the characterization method used to Analyze the cells. Section \ref{sec: M_Preliminary} will dive into the methods used to determine the optimal conditions for an endurance run which accelerates corrosion. In section \ref{sec: M_Endurance run_d} the planned endurance run will be explained. An overview as well as an detailed explanation of the analytical Ex-Situ methods used to evaluate corrosion in the cell will be presented in the second chapter \ref{sec: Ex-Situ}.
This chapter will provide a comprehensive overview of the methods used in this thesis. Starting with the in-situ methods in chapter \ref{sec: M_Setup}, where the hardware will be explained in more detail, as well as the characterisation method used to analyse the cells. Section \ref{sec: M_Preliminary} will dive into the methods used to determine the optimal conditions for an endurance run that accelerates corrosion. In section \ref{sec: M_Endurance run_d}, the planned endurance run will be explained. An overview as well as an detailed explanation of the analytical ex-situ methods used to evaluate corrosion in the cell will be presented in the second chapter \ref{sec: Ex-Situ}.

\section{Experimental Setup}
\label{sec: M_Setup}


In order to have a better understanding of the In-Situ methods used to develop and analyse the endurance this section will focus on the hardware such as the PEMFC cells and Stacks as well as the different test bench setups.
In order to gain a better understanding of the in-situ methods used to develop and analyse the endurance, this section will focus on the hardware, such as the PEMFC cells and stacks as well as the different test bench setups.

\subsection{Stack}

First of all the components of the used test specimen of a PEMFC stack will be explained. Figure \ref{fig:Stack} shows a schematic representation of the different layers which compose the test specimen \citep{99_sabawa2021temperature}.
First of all, the components of the test specimen of a PEMFC stack that was used will be explained. Figure \ref{fig:Stack} shows a schematic representation of the different layers which compose the test specimen \citep{99_sabawa2021temperature}.

%Skizze von der Zelle?
\begin{figure}[htbp]
\centering
\includegraphics[width=0.7\textwidth]{Figures/Method/Stack.pdf}
\caption{Schematic of the structure in the used test specimen with its different components and position. Retrieved from Sabawa. page 43 [99].}
\caption{Schematic of the structure in the test specimen used with its different components and position. Retrieved from Sabawa. page 43 [99].}
\label{fig:Stack}
\end{figure}

In the following each numbered component will be named and explained shortly:
In the following, each numbered component will be named and briefly explained:
\begin{enumerate}
\item \textbf{Screw-in units}: This units provide a point of connection between the test specimen and the test bench for the gases and cooling water.
\item \textbf{Compression plate}: This plate is made out of aluminium and ensures the stability of the structure and that the PEMFC stays sealed.
\item \textbf{O-ring}: This ring sealing ensures, that the gas cannot escape through other parts of the plate and that it follows the predetermined channels.
\item \textbf{Clamping rods}: Ensure that the components of the MEA stay together and under pressure even when the cell is being transported.
\item \textbf{Insulation plate}: The insulation plate is made out of Ultem$^{TM}$ Resin 1000 and secures the thermal and electrical insulation from the end plate on both anode and cathode sides.
\item \textbf{Current collector}: It enables the connection of the load cable to the cell in the test bench. The current collector is made out of gold-plated copper.
\item \textbf{Stacked BPs and MEA}: The "sandwich structure" of the cells will be explained in the next part in a more detailed way.
\item \textbf{Screw-in units}: These units provide a point of connection between the test specimen and the test bench for the gases and cooling water.
\item \textbf{Compression plate}: This plate is made of aluminium and ensures the stability of the structure and that the PEMFC stays sealed.
\item \textbf{O-ring}: This ring sealing ensures that the gas cannot escape through other parts of the plate and that it follows the predetermined channels.
\item \textbf{Clamping rods}: Ensure that the components of the MEA stay together and under pressure, even when the cell is being transported.
\item \textbf{Insulation plate}: The insulation plate is made of Ultem$^{TM}$ Resin 1000 and secures the thermal and electrical insulation from the end plate on both anode and cathode sides.
\item \textbf{Current collector}: Enables the connection of the load cable to the cell in the test bench. The current collector is made of gold-plated copper.
\item \textbf{Stacked BPs and MEA}: The "sandwich structure" of the cells will be explained in the next part in more detail.
\item \textbf{Current collector}: Current collector at the cathode side.
\item \textbf{Insulation plate}: This is the insulation plate at the cathode side.
\item \textbf{Compression plate}: The compression plate on the other side with integraded pneumatic pressure pads.
\item \textbf{Pressure gauge and gas connection}: Before the cell can be used it will be brought up to the right pressure to ensure it is air tight and gas will not leak during the experiments.
\item \textbf{Feet}: Since the test specimen is placed lateral in the test bench it will be supported by the feet.
\item \textbf{Compression plate}: The compression plate on the other side with integrated pneumatic pressure pads.
\item \textbf{Pressure gauge and gas connection}: Before the cell can be used, it is brought up to the correct pressure to ensure that it is airtight and that gas will not leak during the experiments.
\item \textbf{Feet}: Since the test specimen is placed laterally in the test bench, it is supported by the feet.
\end{enumerate}

Now that all the components of the PEMFC test specimen have been explained the actual structure of the cell and its sandwich structure will be explained. Figure \ref{fig:Stack_1} shows a detailed schematic of a cell.
Now that all the components of the PEMFC test specimen have been explained, the actual structure of the cell and its sandwich structure will be explained. Figure \ref{fig:Stack_1} shows a detailed schematic of a cell.

\begin{figure}[htbp]
\centering
\includegraphics[width=0.7\textwidth]{Figures/Method/Stack_1.pdf}
\caption{Schematic of the sandwich structure of the BPs and the MEA within the cell stack. In (a) and (c) the BPs plates and its active area in a darker grey are shown. (b) shows the MEA placed between the BPs. Retrieved from Sabawa page 46 [99].}
\caption{Schematic of the sandwich structure of the BPs and the MEA within the cell stack. In (a) and (c) the BPs plates and their active area in a darker grey colour are shown. (b) shows the MEA placed between the BPs. Retrieved from Sabawa page 46 [99].}
\label{fig:Stack_1}
\end{figure}

As seen in the schematic the MEA (b) is placed between two BPs (a, c). Each BP has an anodic and a cathodic side. The cells can be stacked by following this sandwich structure (BP, MEA, BP, MEA, BP) to generate more power. The darker grey part of the BPs represent the active area. Channels within the BPs ensure the distribution of the coolant flow while hydrogen flows through the flow fields in the active area of the cell, as can be seen in the figure \ref{fig:PEMFC}. On both sides manifolds ensure that the Hydrogen, air, water and coolant can be distributed from the first cell to the las cell and afterwards the used coolant as well as produced water can be transported away from the cell again through the manifolds and into the outlets. The GDL is sealed by adjusting the pressure on the pneumatic pads which can be found in the compression plate at the end of the stack.
As seen in the schematic, the MEA (b) is placed between two BPs (a, c). Each BP has an anodic and a cathodic side. The cells can be stacked by following this sandwich structure (BP, MEA, BP, MEA, BP) to generate more power. The darker grey part of the BPs represent the active area. Channels within the BPs ensure the distribution of the coolant flow while hydrogen flows through the flow fields in the active area of the cell, as can be seen in the figure \ref{fig:PEMFC}. On both sides, manifolds ensure that the hydrogen, air, water and coolant can be distributed from the first cell to the last cell, and afterwards the coolant used and produced water can be transported away from the cell again through the manifolds and into the outlets. The GDL is sealed by adjusting the pressure on the pneumatic pads which can be found in the compression plate at the end of the stack.

Two types of cells where used in the experiment. Type one is made out of titanium with a carbon coating (Ti-C) to enhance the electrical conductivity of the cell and has an active area of 273 cm$^2$. The second type is made out of stainless steel 316L like refereed to in the chapter \ref{subsec: BP Corrosion} and an active area of 285 cm$^2$.
Two types of cells were used in the experiment. Type one is made of titanium with a carbon coating (Ti-C) to enhance the electrical conductivity of the cell and has an active area of 273 cm$^2$. The second type is made out of stainless steel 316L, as referred to in chapter \ref{subsec: BP Corrosion}, and an active area of 285 cm$^2$.

\subsection{Testbench}

Two different test bench setups where used in this Thesis to be able to develop an endurance run with reinforced corrosion conditions and then perform the endurance run. The development of the corrosion endurance run was performed on a \textit{Horiba Fuel Con Shortstack S005} test bench as shown in the figure \ref{fig:Setup_Preliminary} \citep{horiba_fuelcon_2024}. The PEMFC 4-cell stack was placed in the middel of the test chamber and connected to the different supply lines as well as safety and devices.
Two different test bench setups were used in this thesis in order to develop an endurance run with reinforced corrosion conditions and then perform the endurance run. The development of the corrosion endurance run was performed on a \textit{Horiba Fuel Con Shortstack S005} test bench as shown in the figure \ref{fig:Setup_Preliminary} \citep{horiba_fuelcon_2024}. The PEMFC 4-cell stack was placed in the middle of the test chamber and connected to the different supply lines as well as safety devices.

\begin{figure}[htbp]
\centering
@@ -61,9 +61,9 @@ Two types of cells where used in the experiment. Type one is made out of titaniu
\label{fig:Setup_Preliminary}
\end{figure}

In this particular case the test bench was modified, so that the produced water from the cathode and anode could be collected separately. The two condenser bottles where placed under the test bench to ensure, that the produced water flows away from the cell and does not accumulate in the cell after a shutdown. Therefore minimizing the corrosion at the outlet caused by standing product water and also enabling the separate analysis of the product water on the cathode and anode after a test. Since the Test Bench was designed for a 4-cell specimen it allows a maximum gas supply of 200 Nl/min and a minimum of 21,4 Nl/min. Consequently the gas stoichiometry for tests performed with a 10-cell stack at high current densities where lower than the planed $\lambda_{air}$ = 2 at the cathode and $\lambda_{H_2}$ = 1,5 at the anode.
In this particular case, the test bench was modified so that the product water from the cathode and anode could be collected separately. The two condenser bottles were placed under the test bench to ensure that the product water would flow away from the cell and would not accumulate in the cell after a shutdown. This minimised the corrosion at the outlet caused by standing product water and also enabled separate analysis of the product water on the cathode and anode after a test. Since the test bench was designed for a 4-cell specimen, it allowed a maximum gas supply of 200 Nl/min and a minimum of 21,4 Nl/min. Consequently, the gas stoichiometry for tests performed with a 10-cell stack at high current densities were lower than the planned $\lambda_{air}$ = 2 at the cathode and $\lambda_{H_2}$ = 1,5 at the anode.

The endurance run was performed on a \textit{Horiba Fuel Con C1000} test bench. This specific test Bench allows a maximum gas supply of 100 Nl/min and up to 1000A \citep{horiba_fuelcon_2024}. Despite having a different maximum gas supply and the modifications made to the first \textit{Horiba Fuel Con Shortstack S005} test bench to be able to gather and extract the product water the two test benches have a very similar setup. Figure \ref{fig:P&ID} presents a simplified piping \& instrumentation Diagram in order to give a better overview of the components used to change the operating parameters.
The endurance run was performed on a \textit{Horiba Fuel Con C1000} test bench. This specific test bench allowed for a maximum gas supply of 100 Nl/min and up to 1000A \citep{horiba_fuelcon_2024}. Despite having a different maximum gas supply and despite the modifications made to the first \textit{Horiba Fuel Con Shortstack S005} test bench to be able to gather and extract the product water, the two test benches had a very similar setup. Figure \ref{fig:P&ID} presents a simplified piping \& instrumentation diagram to give a better overview of the components used to modify operating parameters.

\begin{figure}[htbp]
\centering
@@ -72,35 +72,35 @@ The endurance run was performed on a \textit{Horiba Fuel Con C1000} test bench.
\label{fig:P&ID}
\end{figure}

The gases have a purity of 99,999\% and can be dosed by the mass flow controllers (MFC) as seen in the figure \ref{fig:P&ID} with the number 2. After the MFC the dew point of the gases can be adjusted with the help of the bubbler-humidifiers (4) at the anode and the cathode. The water in the bubbler can only be heated and not actively cooled which limits the process of ramping down the dew point temperature. Number 5 is the dry gas bypass, this allows for a dew point temperature range from 10-90°C. Furthermore the gases can be heated up to a temperature ranging from 40-180°C in step 6 with the heating hoses on the inlet of the PEMFC. Part 9 of the chart is the PEMFC which for the endurance run consisted of a 4-cell stack of the type two cells made out of stainless steel 316L. Number 11 shows the temperature control of the cooling circuit. The cell can be cooled or heated from a temperature of 35°C to 105°C. It is also possible to change the pressure to be able to reach water temperatures above the boiling point. Number 13 shows the safety sensors which can shut down the cell and the test bench in case of a failure. 15 allows the control of the membrane pressure. The current collectors are connected with a potentiostat to set a defined voltage or current in the FC (18). To monitor the cell voltage the cell is connected to the test bench with a cell voltage monitoring connector (CVM), so that each individual cell voltage can be measured (19).
The gases have a purity of 99,999\% and can be dosed by the mass flow controllers (MFC) as seen in the figure \ref{fig:P&ID} with the number 2. After the MFC the dew point of the gases can be adjusted with the help of the bubbler-humidifiers (4) at the anode and the cathode. The water in the bubbler can only be heated and not actively cooled, which limits the process of ramping down the dew point temperature. Number 5 is the dry gas bypass; this allows for a dew point temperature range from 10-90 °C. Furthermore, the gases can be heated up to a temperature ranging from 40 to 180 °C in step 6, with the heating hoses on the inlet of the PEMFC. Part 9 of the chart is the PEMFC, which for the endurance run consisted of a 4-cell stack of the type two cells made out of stainless steel 316L. Number 11 shows the temperature control of the cooling circuit. The cell can be cooled or heated from a temperature of 35 °C to 105 °C. It is also possible to change the pressure to be able to reach water temperatures above the boiling point. Number 13 shows the safety sensors which can shut down the cell and the test bench in case of failure. 15 allows the control of the membrane pressure. The current collectors are connected with a potentiostat to set a predefined voltage or current in the FC (18). To monitor the cell voltage, the cell is connected to the test bench with a cell voltage monitoring connector (CVM) so that each individual cell voltage can be measured (19).

\subsection{Dew Point Calculation}

As explained before the dew point can be varied by changing the temperature of the bubbler-humidifier to adjust the humidity of the cathode and anode separately. Since the experiment aims for a specific level of humidity on the anode and the cathode of about 30 \% and 50\% the right dew point temperatures have to be calculated. The Sonntag formula for saturation pressure of water was used to calculate the relative humidity at both anode and cathode based on the pressure, temperature of the cell, molecular weight as stated in the British Standard 1339:1 \citep{sonntag1990important}.
As previously explained, the dew point can be varied by changing the temperature of the bubbler-humidifier to adjust the humidity of the cathode and anode separately. Since the experiment aims for a specific level of humidity on the anode and the cathode of about 30 \% and 50 \%, the correct dew point temperatures have to be calculated. The Sonntag formula for saturation pressure of water was used to calculate the relative humidity at both anode and cathode based on the pressure, temperature of the cell and molecular weight as stated in the British Standard 1339:1 \citep{sonntag1990important}.

\subsection{Calculation of the Stochiometrie}
\subsection{Stochiometry Calculation}

Since the stoichiometry of the reactants at the anode and cathode and therefore of the gas dosed plays a very important role to avoid phenomena like hydrogen starvation (which could be fatal for the cell) it is important to understand how the gas flow was calculated. To avoid hydrogen starvation the test where conducted with a stoichiometry of $\lambda_{air}$ = 2 at the cathode and $\lambda_{H_2}$ = 1,5 at the anode.
Since the stoichiometry of the reactants at the anode and cathode and therefore of the dosed gas both play a very important role to avoid phenomena like hydrogen starvation (which could be fatal for the cell), it is important to understand how the gas flow was calculated. To avoid hydrogen starvation, the tests were conducted with a stoichiometry of $\lambda_{air}$ = 2 at the cathode and $\lambda_{H_2}$ = 1,5 at the anode.

Since the tests where carried out at different stoichiometries and by setting the current density the flow rate had to be adjusted and calculated. The first step of this calculation is to multiply the current density with the active area of one BP to determine the electrical current $I$ in A. In the second step the with the help of Faradays law and Faraday's constant $F$ 96485,3 As/mol the flow rate can be calculated using the following equation \citep{Fundamentals_scherer2012fuel}.
Since the tests where carried out at both different stoichiometries and via setting the current density, the flow rate had to be adjusted and calculated. The first step of this calculation is to multiply the current density with the active area of one BP to determine the electrical current $I$ in A. In the second step, with the help of Faradays law and Faraday's constant $F$ 96485,3 As/mol, the flow rate can be calculated using the following equation \citep{Fundamentals_scherer2012fuel}.

\begin{equation}
\varphi=\frac{1}{z \cdot F} \cdot V_{\mathrm{mol}} \cdot \frac{60 \mathrm{~s}}{\mathrm{~min}}
\end{equation}

In this equation z is equal to 2 for H$_2$ (anode) and equal to 4 for O$_2$ (cathode). Furthermore, $V_{mol}$ = 22,414 l/mol and indicates the molar volume of an ideal gas. Then $\varphi$ can be multiplied with the electrical current and the result can be multiplied with the stochiometry factor. The last step is to multiply this with the number of cells to get the final gas flow in Nl/min for each new current density at the selected stoichiometry.
In this equation z is equal to 2 for H$_2$ (anode) and equal to 4 for O$_2$ (cathode). Furthermore, $V_{mol}$ = 22,414 l/mol, and indicates the molar volume of an ideal gas. Then, $\varphi$ can be multiplied with the electrical current and the result can be multiplied with the stochiometry factor. The last step is to multiply this with the number of cells to get the final gas flow in Nl/min for each new current density at the selected stoichiometry.

\subsection{Measurement of pH and Electrical Conductivity}

In order to determine which set of operating parameters of the PEMFC have a reinforcing impact on the corrosion of the BPs the pH and the conductivity of the product water was measured. The measurement was made by using the seven excellence pH meter from the company \textit{Mettler Toledo} \citep{mettler_toledo_ph_meters_2024}. To avoid sudden changes of pressure in the cell which could cause an alteration of the results as well as hard shut down caused by the security sensors, the product water was extracted after the normal shut down and then brought back to the laboratory to be measured.
In order to determine which set of operating parameters of the PEMFC have a reinforcing impact on the corrosion of the BPs the pH and the conductivity of the product water was measured. The measurement was made by using the seven excellence pH meter from the company \textit{Mettler Toledo} \citep{mettler_toledo_ph_meters_2024}. To avoid sudden changes of pressure in the cell, which could cause an alteration of the results as well as hard shutdown by the security sensors, the product water was extracted after the normal shutdown and then brought back to the laboratory to be measured.

\subsection{Characterization of Cells}
\subsection{Characterisation of Cells}

The method used for the In-Situ characterization of the cell in this thesis is the polarization curve as mentioned in the second chapter in the section \ref{subsec: Polarizaiton}. To monitor the state of health of the cell and therefore also the degradation of the cell the polarization curves will be performed at the start of life of the cell and then periodically repeated after the test and in the end of life characterization.
The method used for the in-situ characterisation of the cell in this thesis is the polarisation curve. as mentioned in the second chapter in section \ref{subsec: Polarizaiton}. To monitor the state of health of the cell and therefore also the degradation of the cell, the polarisation curves will be performed at the start of life of the cell and then periodically repeated after the test and in the end of life characterisation.

\subsubsection{Polarization Curves}
\subsubsection{Polarisation Curves}

As a standard procedure three different polarization curves will be tested. Between them the temperature of the cell which should be very similar to the coolant temperature ($T_{coolant,in}$), the gas temperature at the anode and cathode ($T_{gas,A}$, $T_{gas,C}$) and the dew point temperature will be increased while the cell is at a safe point before performing the next curve. The exact parameters for the three polarization curves can be found in the following table \ref{tab:PolKurve}.
As a standard procedure, three different polarisation curves will be tested. Between them the temperature of the cell, which should be very similar to the coolant temperature ($T_{coolant,in}$), the gas temperature at the anode and cathode ($T_{gas,A}$, $T_{gas,C}$) and the dew point temperature will be increased while the cell is at a safe point before performing the next curve. The exact parameters for the three polarisation curves can be found in the following table \ref{tab:PolKurve}.

\begin{table}[h]
\centering
@@ -113,23 +113,23 @@ As a standard procedure three different polarization curves will be tested. Betw
90& 105 & 62& 2&105& 62& 2\\
\end{tabular}
\caption{Operating parameters of the polarization curves at 60, 80 and 90 °C and 2 bar pressure to characterize the cell performance.Temperatures in [°C] and pressure in [bar].}
\caption{Operating parameters of the polarisation curves at 60, 80 and 90 °C and 2 bar pressure to characterise the cell performance. Temperatures in [°C] and pressure in [bar].}
\label{tab:PolKurve}
\end{table}

Since the bubblers can not be cooled actively the characterization starts with the 60°C polarization curve to avoid long waiting periods. In each polarization curve the current density is increased gradually from 0 A/cm$^2$ to 2,2 A/cm$^2$ and then back down to 0 A/cm$^2$. Each step is maintained for 120s. With each change the Volume flow of the gases on cathode and anode is adjusted as well to avoid H$_2$ starvation.
Since the bubblers can not be cooled actively, the characterisation starts with the 60 °C polarisation curve to avoid long waiting periods. In each polarisation curve, the current density is increased gradually from 0 A/cm$^2$ to 2,2 A/cm$^2$ and then back down to 0 A/cm$^2$. Each step is maintained for 120s. With each change, the volume flow of the gases on cathode and anode is also adjusted to avoid H$_2$ starvation.

\newpage
\section{Preliminary Investigation}
\label{sec: M_Preliminary}

In order to design an endurance run in which the conditions will reinforce corrosion mechanism within the cell without inducing a total failure of the cell before corrosion becomes visible, It is essential to look into how different parameters affect the cell. Therefore, before starting an endurance run various parameters like the temperature of the coolant at the inlet ($T_{coolant,in}$), gas temperature at the anode and cathode ($T_{gas,A}$ , $T_{gas,C}$ ) as well as the dew point temperature at the anode and cathode ($T_{dp,A}$) where tested to measure the effects they have on the cell.
It is essential to look into how different parameters affect the cell when designing an endurance run, as it takes place under conditions that reinforce the corrosion mechanism within the cell and must avoid inducing total failure of the cell before corrosion becomes visible. Therefore, before starting an endurance run, various parameters such as the temperature of the coolant at the inlet ($T_{coolant,in}$), gas temperature at the anode and cathode ($T_{gas,A}$, $T_{gas,C}$ ) and the dew point temperature at the anode and cathode ($T_{dp,A}$) were tested to measure their effects on the cell.

High humidity as well as a low pH in the product water where identified as conditions which could intensify the corrosion of the BPs. Therefore the following experiment was designed to be able to measure the pH and the electrical conductivity of the product water and trace metal ions dissolving from the BP which could potentially lead to an increased electrical conductivity. The pH could also be theoretically lowered by the Nafion degradation and a potential release of F$^-$.
High humidity along with a low pH in the product water were identified as conditions that could intensify the corrosion of the BPs. Therefore, the following experiment was designed to be able to measure the pH and the electrical conductivity of the product water and trace metal ions dissolving from the BP which could potentially lead to an increased electrical conductivity. The pH could also be theoretically lowered by the Nafion degradation and a potential release of F$^-$.

\subsection{Experimental Setup}

The test specimen used for the preliminary investigations was a 4-cell stack of the type one cell made out of titanium and a carbon coating (Ti-C) and an active area of 273 cm$^2$. As stated before this test was performed on the \textit{Horiba Fuel Con Shortstack S005} test bench which was modified as seen in the figure \ref{fig:Setup_Preliminary} to be able to extract the product water of the anode and cathode separately.
The test specimen used for the preliminary investigations was a 4-cell stack of the type one cell made of titanium and a carbon coating (Ti-C) and an active area of 273 cm$^2$. As previously stated, this test was performed on the \textit{Horiba Fuel Con Shortstack S005} test bench, modified as seen in the figure \ref{fig:Setup_Preliminary} in order to be able to extract the product water of the anode and cathode separately.

\subsection{Experimental Execution}

@@ -137,25 +137,25 @@ The preliminary investigation can be divided into the following 4 steps:

\begin{enumerate}
\item Startup and activation of the cell.
\item Begin of life characterization with polarization curves.
\item Parameter variation tests with voltage cycling.
\item End of life characterization of the cells.
\item Begin of life characterisation with polarisation curves.
\item Parameter variation tests with voltage cycling.
\item End of life characterisation of the cells.
\end{enumerate}

This Steps will be explained in a more detailed way in the following. Starting with the activation of the cell after the startup.
These steps will be explained in a more detail in the following. Starting with the activation of the cell after the startup.

\subsubsection{Activation of the Cell}

After the startup the cell it is not completely ready to perform. Even though the materials where carefully stored in a clean room to avoid the drying out of the membrane which could cause mechanical stress and damage it. To avoid any damage to the MEA the cell has to be carefully activated to ensure that the membrane has the right humidity in order to perform.
After the start up the cell is at a temperature of 60°C, the gas temperature at 85°C and the dew points at 46°C. Therefore, to activate the cell the 60°C polarization curve is performed 3 times in a row and afterwards the 80°C polarization curve is performed. After this process the cell should have reached an optimal performance and the polarization curves should look stable.
After startup, the cell is not completely ready to perform. Even though the materials were carefully stored in a clean room to avoid having the membrane dry out, which could cause mechanical stress and damage to it. To avoid any damage to the MEA, the cell must be carefully activated to ensure that the membrane has the right humidity in order to perform.
After startup, the cell is at a temperature of 60 °C, the gas temperature at 85 °C and the dew points at 46 °C. Therefore, to activate the cell the 60 °C polarisation curve is performed three times in a row and then the 80 °C polarisation curve is performed. After this process, the cell should have reached an optimal performance and the polarisation curves should look stable.

\subsubsection{Begin of Life Characterization}
\subsubsection{Begin of Life Characterisation}

For the begin of life characterization the three polarization curves at 60,80 and 90 °C are performed to set a benchmark which can then be analyzed and compared to the following characterizations to be able to see the drop in the performance due to the degradation.
For the begin of life characterisation, the three polarisation curves at 60, 80 and 90 °C are performed to set a benchmark which can then be analysed and compared with the following characterisations so that the drop in performance due to the degradation can be seen.

\subsubsection{Parameter Variation}

Three different operating parameters where tested to see how the cell and gas temperature could affect the pH and electrical conductivity of the product water. The cell was tested at 60°C, 75°C and 90°C. The exact parameters for the gas temperatures and pressures can be found in table \ref{tab:3_pH_T}.
Three different operating parameters were tested to see how the cell and gas temperature could affect the pH and electrical conductivity of the product water. The cell was tested at 60 °C, 75 °C and 90 °C. The exact parameters for the gas temperatures and pressures can be found in table \ref{tab:3_pH_T}.

\begin{table}[h]
\centering
@@ -172,21 +172,25 @@ Three different operating parameters where tested to see how the cell and gas te
\label{tab:3_pH_T}
\end{table}

The dew point temperatures where adjusted so that the cell could have a relative humidity of about 30\% at the anode and 50\% at the cathode. Furthermore, since the first test with a low temperature and high humidity at 60°C should have a dew point temperature of 37°C at the anode and the normal start up rises the dew point temperature to 46°C the startup script had to be adjusted. Consequently the temperature of the dew point at the anode would just be elevated to 37°C.
The dew point temperatures were adjusted so that the cell could have a relative humidity of about 30 \% at the anode and 50 \% at the cathode. Furthermore, since the first test with a low temperature and high humidity at 60 °C should have a dew point temperature of 37 °C at the anode, and normal startup raises the dew point temperature to 46°C, the startup script had to be adjusted. Consequently, the temperature of the dew point at the anode were simply elevated to 37°C.

After the start up in all three test the desired temperatures found in table \ref{tab:3_pH_T} where set in a specific order:
After the start up in all three tests the desired temperatures found in table \ref{tab:3_pH_T} were set in a specific order:
\begin{enumerate}
\item Increase anode pressure to 2,4 bar.
\item Decrease cathode pressure to 1,5 bar.
\item Increase cell temperature by increasing the coolant temperature $T_{coolant,in}$.
\item Increase the gas temperatures $T_{gas,A}$ and $T_{gas,C}$ to the desired temperature.
\item Increase the dew point temperatures $T_{dp,A}$ and $T_{dp,C}$ to reach the desired relative humidity.
\item Ramp up current density and Volume flow of the gases until 2 A/cm$^2$.
\item Start Voltage cycling for 2h.
\item Increase the dew point temperatures $T_{dp,A}$ and $T_{dp,C}$ to reach the desired relative humidity.
\item Ramp up current density and volume flow of the gases until 2 A/cm$^2$.
\item Start voltage cycling for 2h.
\item Shut down.
\end{enumerate}

After reaching the set parameters the cell Voltage cycling of the cell was manually started. In this process the cell switched between 15s at 0,85V and 10s at 0,6V. This can be clarified by the following figure. After the two hours the cell could be shut down and after the shot down the product water was collected at the cathode and anode to be measured at the laboratory.
After reaching the set parameters the voltage cycling of the cell was
manually started. In this process, the cell switched between 15s at 0,85V and
10s at 0,6V. This can be clarified by the following figure. After the two hours
the cell could be shut down and after shutdown the product water was
collected at the cathode and anode to be measured at the laboratory.

\begin{figure}[htbp]
\centering
@@ -198,19 +202,19 @@ After reaching the set parameters the cell Voltage cycling of the cell was manua
\section{Developement of Endurance Run}
\label{sec: M_Endurance run_d}

The endurance run was developed after the preliminary investigations with the three different settings and voltage cycling. Some changes where made to the methods which will be explained in this section. In order to compare the temperature effects on the corrosion and other degradation mechanisms like Pt dissolution and agglomeration two different endurance runs where performed. The low temperature high humidity endurance run to trigger the corrosion mechanisms and a high temperature endurance run to trigger other degradation mechanisms like Pt dissolution, agglomeration and membrane degradation. The changes to the mehtod will be explained in the next sections.
The endurance run was developed after the preliminary investigations with the three different settings and voltage cycling. Some changes were made to the methods which will be explained in this section. In order to compare the temperature effects on the corrosion and other degradation mechanisms like Pt dissolution and agglomeration, two different endurance runs were performed. First, the low temperature high humidity endurance run, in order to trigger the corrosion mechanisms, and second, a high temperature endurance run, in order to trigger other degradation mechanisms such as Pt dissolution, agglomeration and membrane degradation. The changes to the mehtod will be explained in the following sections.

\subsection{Experimental Setup}

Both endurance runs where performed at the same time on two identical \textit{Horiba Fuel Con C1000} test benches \citep{horiba_fuelcon_2024}. The cells used for both 4-cell stacks where the type two cells made out of stainless steel 316L and an active area of 285 cm$^2$. This time no product water was extracted from the cell because of the space limitations of the two test benches.
Both endurance runs were performed at simultaneously on two identical \textit{Horiba Fuel Con C1000} test benches \citep{horiba_fuelcon_2024}. The cells used for both 4-cell stacks were the type two cells made of stainless steel 316L and an active area of 285 cm$^2$. This time, no product water was extracted from the cell because of the space limitations of the two test benches.

\subsection{Experimental Execution}

First the experimental execution of the low temperature corrosion endurance run will be explained and afterwards the high temperature endurance run. It is worth mentioning that the two endurance runs had the same activation procedure and characterization procedure between each cycle. Since both differed from the one used in the preliminary investigations they will be explained prior to outlining the plan and parameters for both endurance runs.
First, the experimental execution of the low temperature corrosion endurance run will be explained, and next the high temperature endurance run. It is worth mentioning that the two endurance runs had the same activation procedure and characterisation procedure between each cycle. Since both differed from that used in the preliminary investigations, they will be explained prior to outlining the plan and parameters for both endurance runs.

\subsubsection{Activation of the Cell}

Since the 60°C polarization curve presents a higher humidity than the 80°C polarization curve the last is better equipped for the activation of the type two cells. Therefore, for the activation of the endurance run cells the 80°C polarization curve was repeated four times as an activation. After the third the difference between the curves was minimal so that the hydration of the membrane was optimal as well as the performance of the cell. The results can be seen in chapter \ref{chap:Ergebnisse und Diskussion}.
Since the 60 °C polarisation curve presents a higher humidity than the 80 °C polarisation curve, the latter is better equipped for the activation of the type two cells. Therefore, for the activation of the endurance run cells the 80 °C polarisation curve was repeated four times as an activation. After the third, the difference between the curves was minimal so that the hydration of the membrane as well as the performance of the cell could be optimal. The results can be seen in chapter \ref{chap:Ergebnisse und Diskussion}.

\subsubsection{Endurance Runs}

@@ -218,24 +222,24 @@ Since the 60°C polarization curve presents a higher humidity than the 80°C pol
\begin{figure}[htbp]
\centering
\includegraphics[width=0.8\textwidth]{Figures/Method/In-Situ.pdf}
\caption{Structure for the corrosion endurance run and high temperature endurance run with the three steps of 12500, 25000 and 43500 voltage cycles. Between each step the in-situ characterization is performed.}
\caption{Structure for the corrosion endurance run and high temperature endurance run with the three steps of 12500, 25000 and 43500 voltage cycles. The in-situ characterisation is performed between each step.}
\label{fig:In-Situ}
\end{figure}

Figure \ref{fig:In-Situ} shows the structure of both the corrosion endurance run and the high temperature endurance run since they have the same amount of voltage cycling cycles. For the in-situ characterization of the cells after the activation and in-between each set of voltage cycles two polarization curves are planned. This time only the 60°C and the 80°C polarization curves will be used to characterize the cell and its performance loss. Due to the high temperature and volume flow of the 90°C polarization curve that could intensify the degradation before the next cycle as well as trigger a hard shut down on the test bench. To avoid any damage to the cell caused because of the hard shut down the decision was made to only work with the other two curves for this tests.
Figure \ref{fig:In-Situ} shows the structure of both the corrosion endurance run and the high temperature endurance run, since they have the same number of voltage cycling cycles. For the in-situ characterisation of the cells after the activation and in between each set of voltage cycles, two polarisation curves are planned. This time, only the 60 °C and the 80 °C polarisation curves will be used to characterise the cell and its performance loss. Due to the high temperature and volume flow of the 90°C polarisation curve, this could intensify the degradation before the next cycle, as well as trigger a hard shutdown on the test bench. To avoid any damage to the cell caused by the hard shutdown, the decision was made to only work with the other two curves for these tests.

The voltage cycling in the two endurance runs is also different from the one used in the preliminary study. Since the endurance run would be a lot longer than the first tests each cycle was modified and now consists of 10s hold time at 0,88V and then another 10s at 0,6V. This modified cycle is illustrated in figure \ref{fig:ECU}.
The voltage cycling in the two endurance runs is also differs from that used in the preliminary study. Since the endurance run would be much longer than the first tests. each cycle was modified and now consists of a 10-second hold time at 0,88 V and then another 10 seconds at 0,6 V. This modified cycle is illustrated in figure \ref{fig:ECU}.

\begin{figure}[htbp]
\centering
\includegraphics[width=0.6\textwidth]{Figures/Method/ECU.pdf}
\caption{Voltage cycling of the cell between 10s at 0,88V and 10s at 0,6V.}
\caption{Voltage cycling of the cell between 10 seconds at 0,88 V and 10 seconds at 0,6 V.}
\label{fig:ECU}
\end{figure}

\subsubsection{Corrosion Reinforcing Endurance Run}

Before starting each cycle of voltage cycling as seen in figure \ref{fig:In-Situ} the cell first hast to get to the desired operating parameters. For this corrosion reinforcing endurance run the specific parameters can be found in the table \ref{tab:3_ER_S}.
Before starting each cycle of voltage cycling as seen in figure \ref{fig:In-Situ}, the cell first must reach the desired operating parameters. For this corrosion reinforcing endurance run, the specific parameters can be found in the table \ref{tab:3_ER_S}.

\begin{table}[h]
\centering
@@ -245,15 +249,15 @@ Before starting each cycle of voltage cycling as seen in figure \ref{fig:In-Situ
\hline
65& 85 & 45& 2,3&85 &53& 1,4\\
\end{tabular}
\caption{Temperature and pressure parameters of the corrosion reinforcing endurance run.Temperatures in [°C] and pressure in [bar].}
\caption{Temperature and pressure parameters of the corrosion reinforcing endurance run. Temperatures in [°C] and pressure in [bar].}
\label{tab:3_ER_S}
\end{table}

This time the relative humidity of both cathode and anode is increased compared to the preliminary study. The cathode is run at a relative humidity of about 54,66\% and the anode at 36,65\% to increase corrosion even more. The volume flow of the gases is increased to cope with an current density of 1,5 A/cm$^2$ before starting the voltage cycling. The stoichiometry stays at $\lambda_{air}$ = 2 at the cathode and $\lambda_{H_2}$ = 1,5 at the anode to increase the performance at the cell and avoid H$_2$ starvation at the anode.
This time, the relative humidity of both cathode and anode is increased compared to the preliminary study. The cathode is run at a relative humidity of about 54,66 \% and the anode at 36,65 \% to further increase corrosion. The volume flow of the gases is increased to cope with a current density of 1,5 A/cm$^2$ before starting the voltage cycling. The stoichiometry stays at $\lambda_{air}$ = 2 at the cathode and $\lambda_{H_2}$ = 1,5 at the anode to increase the performance at the cell and avoid H$_2$ starvation at the anode.

\subsubsection{High Temperature Endurance Run}

For the high temperature endurance run the parameters are different. Starting with a much higher cell temperature due to the cooling temperature of the cell which is at 103°C compared to the 65°C of the corrosion endurance run. The temperature and pressure parameters of the high temperature endurance run can be seen in table \ref{tab:3_ER_HT}.
For the high temperature endurance run, the parameters are different. To start with, a much higher cell temperature is used due to the cooling temperature of the cell, which is at 103 °C compared to the 65 °C of the corrosion endurance run. The temperature and pressure parameters of the high temperature endurance run can be seen in table \ref{tab:3_ER_HT}.

\begin{table}[h]
\centering
@@ -267,19 +271,19 @@ For the high temperature endurance run the parameters are different. Starting wi
\label{tab:3_ER_HT}
\end{table}

Since the hight temperature already accounts for a higher product water production in the cell the pressure in the anode is not lowered to 1,4 bar but only to 1,8 bar to avoid an accumulation of water at the cathode exit. The relative humidity is also lower than in the corrosion endurance run with a relative humidity of about 30,1\% at the anode and 51\% at the cathode. As explained before the two setups, characterization methods and number of cycles in the endurance run are the same as in the corrosion run, just the parameters have changed.
Since the high temperature already accounts for a higher product water production in the cell, the pressure in the anode is not lowered to 1,4 bar, but only to 1,8 bar in order to avoid an accumulation of water at the cathode exit. The relative humidity is also lower than in the corrosion endurance run with a relative humidity of about 30,1 \% at the anode and 51 \% at the cathode. As previously explained, the two setups, characterisation methods and number of cycles in the endurance run are the same as in the corrosion run, only the parameters have been changed.

\newpage

\section{Ex-Situ analysis}
\label{sec: Ex-Situ}

After the endurance runs the components of the PEMFC test specimen are analysed to detect and evaluate the corrosion damage caused by the operating conditions. Therefore, the cell is brought to the laboratories and into a clean room to avoid any contaminations of the BPs and the MEA while opening. The stacked cells are then dismounted to be optically analysed.
After the endurance runs, the components of the PEMFC test specimen are analysed to detect and evaluate the corrosion damage caused by the operating conditions. Therefore, the cell is brought to the laboratories and into a clean room to avoid any contaminations of the BPs and the MEA while opening. The stacked cells are then dismounted to be visually analysed.

\begin{figure}[htbp]
\centering
\includegraphics[width=0.8\textwidth]{Figures/Method/FlowField.pdf}
\caption{Matrix of the active area of the cell for to clarify the exact positions analysed.}
\caption{Matrix of the active area of the cell to clarify the exact positions analysed.}
\label{fig:Matrix}
\end{figure}

@@ -287,16 +291,16 @@ Figure \ref{fig:Matrix} provides a matrix of the BP to enable a clearer understa

\subsection{Microscopy}

The first step after the cells have been removed from the stack is to look for abnormalities or discolorations in the stainless steel 316L which could be caused by the operating conditions and could include corrosion damages. In this step the whole cathode side of the BP is analysed including the cathode inlet and outlet. Since the humidity is higher at the cathode the positions M1 and A6 will be analysed in greater detail as well as the cathode outlet. For this analysis a microscope from the company \textit{Keyence} model VHX 7000 is used as well as a VHX 6000 \citep{keyence_digital_microscope_2024}.
The first step after the cells have been removed from the stack is to search for abnormalities or discolorations in the stainless steel 316L that could be caused by the operating conditions and could also include corrosion damages. In this step, the whole cathode side of the BP is analysed, including the cathode inlet and outlet. Since the humidity is higher at the cathode, the positions M1 and A6 as well as the cathode outlet will be analysed in greater detail. For this analysis, a microscope from the company \textit{Keyence} model VHX 7000 is used as well as a VHX 6000 \citep{keyence_digital_microscope_2024}.

\subsection{LIBS Analysis}

After the microscopical analysis the positions M1, A6, cathode outlet and the selected points of interest will be further investigated using laser induced breakdown spectroscopy (LIBS). LIBS will be performed by an EA-300 from the company \textit{Keyence} \citep{keyence_ea_300_2024}. This atomic-emission spectroscopy method enables for a quick material analysis. A high energy laser converts the material into a plasma plume which cools down and breaks down really fast and breaks down into ionic atomic species which are excited. Thanks to the quick cool down the optical breakdown of the plasma can be analysed with the spectrometer. This could give more insights into the material composition of the selected sites and help understand the changes in the surface composition related to corrosion.
After the microscopical analysis, the positions M1, A6, cathode outlet and the selected points of interest will be further investigated using laser induced breakdown spectroscopy (LIBS). LIBS will be performed by an EA-300 from the company \textit{Keyence} \citep{keyence_ea_300_2024}. This atomic-emission spectroscopy method enables a quick material analysis. A high energy laser converts the material into a plasma plume which cools and breaks down very quickly, breaking down into excited ionic atomic species. Thanks to the quick cooldown, the visual breakdown of the plasma can be analysed with the spectrometer. This could give more insights into the material composition of the selected sites and help understand the changes in the surface composition related to corrosion.


\subsection{SEM/EDX Analysis}

The corrosion mechanism can cause severe damage of the metal structure and cause the dissolution of the metal. As stated in the theoretical part of this thesis the metal ions can move from the BPs to the MEA and catalyse the Fenton mechanism which degrades the Membrane \citep{ruvinskiy2011using, Corr_mele2010localised}. Therefore, scanning electron spectroscopy (SEM) as well as energy dispersive x-ray spectroscopy (EDX) will be used evaluate the BP and try to find cases of metal dissolution or formation of an oxide layer as a result of the corrosion. Furthermore, the MEA will be analysed to try to find metal contaminations of Ni, Fe and Cr on the GDL and CCM and prove the movement from the metal ions from the BPs into the MEA as a result from the corrosion \citep{105_novalin2022concepts}. Both SEM and EDX will be performed on a Zeiss EVO 10 with smart EDX \citep{zeiss_evo_sem_2024}.
The corrosion mechanism can cause severe damage to the metal structure and dissolution of the metal. As stated in the theoretical part of this thesis, the metal ions can move from the BPs to the MEA and catalyse the Fenton mechanism, degrading the membrane \citep{ruvinskiy2011using, Corr_mele2010localised}. Therefore, scanning electron spectroscopy (SEM), as well as energy dispersive x-ray spectroscopy (EDX), will both be used to evaluate the BP and attempt to find cases of metal dissolution or formation of an oxide layer due to corrosion. Furthermore, the MEA will be analysed to attempt to find metal contaminations of Ni, Fe and Cr on the GDL and CCM and prove the movement from the metal ions from the BPs into the MEA are a result of the corrosion \citep{105_novalin2022concepts}. Both SEM and EDX will be performed on a Zeiss EVO 10 with smart EDX \citep{zeiss_evo_sem_2024}.





Loading…
Cancel
Save