MA_Platteau/Content/SummaryAndOutlook.tex

\chapter{Summary and Outlook}
\label{cap:SummaryAndOutlook}

The final chapter of this thesis will present a brief summary of the work as well as its motivation in the first section \ref{sec:5_Summary}. The second and last part of this chapter \ref{sec:5_Outlook} will offer an outlook on potential directions for future research.


\section{Summary}
\label{sec:5_Summary}

Since global temperature is on the rise with projections of an increase of 1,5 °C by 2030, it is of utmost importance to implement effective measures to mitigate climate change. GHG and especially CO$_2$ emissions have been identified as the main contributor to global climate, accounting for 80,6\% of emissions. The transportation sector in Germany is responsible for 19,8\% of emissions in the country. It is therefore crucial to transition from ICEs to more sustainable alternatives such as BEVs and FCEVs. While FC technology is on the rise, manufacturing costs continue to be a limiting factor as the BPs constitute 45\% of manufacturing costs. The search for more cost effective alternatives has suggested a change from Ti BPs to stainless steel plates. Although the plates are more cost effective, the corrosion of BP plates raises questions about their durability and how it may affect PEMFC performance. Therefore, this thesis aims to deepen the understanding of the corrosion in stainless steel plates by developing a corrosion reinforcing endurance run and studying the effect of corrosion on the performance of PEMFCs, as well as analysing corrosion damage in cells.

Chapter \ref{cap: Theorie} lays the theoretical background of this thesis. The fundamentals of the fuel cell are explained as well as the components of a PEMFC (PEM, GDL, CL and BP). The membrane is composed of Nafion (PFSA) and the CL has Pt as catalyst for the electrochemical reactions. Before explaining the degradation mechanisms, the overpotentials are first explained using the polarisation curve. At low current densities, the activation losses dominate the polarisation curve. At medium current densities, the ohmic losses dominate the form of the curve and at high current densities the form of the polarisation curve is determined by the concentration polarisation or mass transport losses related to the diffusion limitations of the cell. Afterwards, the main degradation mechanisms are explained, starting with Pt catalyst dissolution and agglomeration, moving on to electrochemical carbon corrosion, membrane degradation and finally corrosion. In the membrane degradation, the fenton reaction is of utmost importance, and this reaction is catalysed by metal ions such as Fe$^{2+}$, which react with H$_2$O$_2$ to form hydroxyl radicals that attack the membrane. In this process, F$^-$ from the membrane is released. F$^-$ decreases the pH of the product water which can then attack the BPs or reinforce corrosion. Pitting corrosion and crevice corrosion as well as galvanic corrosion can lead to structural damages to the cell, cause pinholes or even gas crossover, which will all dramatically shorten the lifespan of the cell.

Chapter \ref{chap:Methode} explains the methods used to analyse the operating conditions that reinforce corrosion and afterwards design a corrosion reinforcing endurance run. In the preliminary investigations, the pH and the electrical conductivity of the product water at three different operating points is tested. The effect of the operating temperature is then measured by searching for the lowest pH due to membrane degradation and release of F$^-$, and subsequently the highest electrical conductivity as a sign from the metal ions released by the BP during a corrosion mechanism. Afterwards, two endurance runs are performed: one with corrosion reinforcing operating parameters, and a second high temperature endurance run for compariso. The cell components of the 4 cell stack are then analysed in the ex-situ investigations using microscopy, LIBS as well as REM and EDX.

The results along with their discussion are then presented in Chapter \ref{chap:Ergebnisse und Diskussion}. Two types of cells were used in this thesis: type 1 made of Ti-C with an active area of 273 cm$^2$ (used for the preliminary investigation) and type 2 cells, made of stainless steel 316L and an active area of 285 cm$^2$ (used in the endurance runs). The preliminary investigations concluded that the lowest pH as well as the highest electrical conductivity was found at a lower temperature (60 °C) at the cathode and the highest pH and lowest electrical conductivity at a temperature of 90 °C after 2 hours of voltage cycling between 10s at 0,6 V and 15s at 0,85 V. The pH found in the literature of 2 was not reached; this could have been caused by the higher corrosion resistance of Ti-C compared to 316L. As such, fewer metal ions moved from the BP to the membrane to catalyse the fenton reaction and the degradation was less than it would have been with a stainless steel plate. The corrosion endurance run was performed at a cell temperature of 66 °C with a 4 cell stack made of type 2 cells (316L). After the activation of the cells using four 80 °C polarisation curves, the BoL characterisation was performed with a 60° C polarisation curve and a 80° C polarisation curve as in-situ characterisation. Then, the cell was set up for 12500 VC of 10s at 0,6 V and 10s at 0,88 V, followed by the in-situ characterisation. Next, the 25000 VC was followed by the in-situ characterisation and finally 42500 VC followed by the last in-situ characterisation. The 60 °C polarisation curve showed a higher degradation than the 80 °C polarisation curve. A significant difference in the cell voltages at high current densities was measured in the 60°C polarisation curve between cells 1 and 4. Cell 1 presented a voltage of 0,6 V while cell 4 had a voltage of 0,53V. Due to this difference, the two cells were further analysed in the ex-situ section. The 80 °C polarisation curve showed almost no signs of degradation after 81000 VC, which can be attributed to the high stoichiometry of 1,5 at the anode and 2 at the cathode, as well as to the low HFR measured. The low HFR after 81000 VC indicates an optimal humidity level of the cell and therefore a lower ohmic polarisation. This results in improved cell performance.

The ex-situ analysis of cells 1 and 4 of the corrosion endurance run showed signs of corrosion in the welding seams of both cells as well as corrosion in the cathode outlet of the cell 4. The CCM of both BP 1 and 4 showed Pt agglomeration as well as cracks and a wave structure. The LIBS analysis performed on the welding seam and the cathode outlet showed an increased oxide layer and decreased carbon percentage in the measurement. The Fe, Ni and Cr percentages could not be analysed correctly due to their high standard deviations. The oxide layer and the Pt agglomerations could have caused increased losses in performance of the cell, although the losses could also be attributed to membrane degradation and carbon corrosion (i.e. decrease in carbon percentage of the material). A REM and EDX analysis of the cathode CCM in BP 1 and 4 showed no traces of metals from the BP (Fe, Ni, Cr), and thus no signs of the migration from the BP to the CCM. This could be attributed to the duration of the endurance run of 400h. Finally, the next section will present an outlook on the topic of this thesis.


\section{Outlook}
\label{sec:5_Outlook}

Due to time constraints and problems with the first test bench in which the preliminary investigations were conducted, some of the experiments will have to be repeated. After a promising result in the pH and electrical conductivity measurements, the next step would have been to repeat these analyses with a stack made of the type 2 stainless steel cell. This could lead to lower pH and a higher electrical conductivity as a sign of corrosion, which was not the case with the type 1 plates made of Ti-C. Furthermore, an online pH and electrical conductivity measurement of the endurance runs could lead to a better understanding of the operating conditions and the condition of the cell in the different states of the endurance run. Moreover, a more detailed analysis of the product water, including Fe, Cr, Ni, Si, Mo, magnesium (Mg) and fluoride could provide a better overview of the corrosion mechanism and consequently the dissolution of the metal ions, enabling them to be traced in the product water. It would also be helpful to create a correlation between the fluoride content and the metal ions that cause the membrane degradation (or degradation of Nafion) catalysed by the metal ions in the Fenton reaction.

The use of other analytical methods such as X-ray diffraction analysis (XRD), X-ray photoelectron spectroscopy (XPS) and inductively coupled plasma mass spectrometry (ICP-MS) could also produce a much better result when analysing components of the PEMFC. Metal traces of Fe, Ni or Cr could be detected on the CCM, GDL and MPL. XPS could provide deeper insights into the oxide layer on the BP, therefore offering a more detailed understanding of BP corrosion.

Potentiostatic and potentiodynamic measurements of the stainless steel could also be performed to evaluate the BP material and compare it to the targets set by the DOE for 2025 as stated in the theoretical background in section \ref{subsec:2_DOE}. Lastly, the research could also focus on other types of stainless steels. Potential candidates might be 304, 904L, 321 or even other coatings for the different BP materials to improve corrosion resistance.