Improving on quality control methods used in commercial MEA manufacturing

Gani, Raeesah

Improving on quality control methods used in commercial MEA manufacturing

Thesis / Dissertation

2023

Abstract

With the current rate of global warming, temperatures are continuously increasing and could have severe impacts on natural and human systems. The transport sector has the largest potential for short term reduction of emissions and the fuel cell industry could be a viable replacement for fossil fuel consumption. The hydrogen economy, alongside the fuel cell industry, can target decarbonisation of long-haul and heavy-duty transport by pursuing a green hydrogen mobility ecosystem. One of the key barriers hindering fuel cell commercialisation is the lack of standard methods for quality control in membrane electrode assembly (MEA) manufacturing, specifically the lack of in-line, non-destructive and roll-to-roll quality control methods. There are currently 5 typical methods for assessing the quality of an MEA, which are gravimetric and thickness analysis, visual and optical microscopy inspection and single cell characterisation. The development of efficient, contactless and non-destructive techniques for in-line quality control testing is a critical enabler in alleviating manufacturing costs by detecting and flagging manufacturing defects thus ensuring a desired quality standard is met at each step in the process. Nonconforming products are detected early and removed at different stages in the process, reducing manufacturing costs that would have otherwise been wasted on defective products. (Yuan, et al., 2021; Phillips, Ulsh , Neyerlin, Porter, & Bender, 2018) In this work, three techniques have been proposed and developed for improving quality control. They are automated visual inspection, infrared thermography (IRT) and x-ray fluorescence (XRF). A rig was successfully designed and built for the automated visual inspection and IRT and methods were successfully developed for the use of the 3 techniques in MEA fabrication. The results from the thickness and gravimetric analysis yielded values that were on par with the experimental targets and commercial standards. A linear relationship was also demonstrated, showing the link between thickness and gravimetric loading. If the relative humidity is well controlled, one of the two quality control methods could be omitted during manufacturing to decrease manufacturing time. To overcome the downfall associated with relative humidity for both these test methods, a newer technology, XRF spectrometry, was investigated which determines the PGM loading and is independent of changes in the relative humidity. The PGM loading can then be used to calculate the approximate thickness of the electrode layers if needed and is less labour intensive and time consuming. It was discovered that the XRF needs to be specifically programmed for each matrix that it is to be used on – where the matrix includes the substrate thickness, type of substrate, type of catalyst in the ink, ionomer content and sample orientation (cathode or anode facing XRF eye). It is assumed that the matrix expands to other parameter changes as well, such as type of ionomer, ink solids content and presence of backers or covers film. However, these parameters were not tested within this study due to time and resource constraints. To assess the aesthetic and physical quality of the CLs produced, visual and optical microscopy inspection were done. Visual inspection involves human observation of the layer under light, therefore, methods were developed in this study to improve the test procedure and omit human error by digitising the test procedure. The results yielded from the visual and optical microscopy inspection were compared and were found to be almost identical. For the purpose of commercial manufacturing to identify light translucency above 10%, it would be less labour intensive and time consuming to use the automated visual inspection method developed. The downfall of either method arises when coating the anode layer (in this study) or when coating the second electrode (in general) due to themasking effect of first layer on the second, which prevents inconsistencies in the second layer being identified. The other downfall of either method is the inability to determine if a defect is a light spot or if it is a pinhole through the MEA. This is where IRT comes in as it was successful in identify pinholes. SEM and single cell performance characterisation were used to test whether the methods developed for automated visual inspection, XRF analysis and/or IR thermography had a negative or destructive effect on the MEAs – the results proved that there was no negative effect from any of the 3 methods. All three techniques were successfully developed for local use in laboratory and small-medium scale manufacturing. The automated visual inspection technique proved to be a key quality control technique and further work into its next generation and incorporation into a high-volume manufacturing line is needed as this could not be investigated due to time and budget constraints. The study was successful in developing a method for the use of XRF to determine PGM loadings and in investigating key parameters associated with its calibration procedures. The value of XRF over gravimetric analysis was shown and the XRF was successfully incorporated into a simulated high- volume line, however, further work is needed for its next generation in a high-volume line. A method for the use of IRT for defect detection was also successfully developed and an attempt to create a defect catalogue was started, however, cataloguing the appearance of defects in IRT still requires more work.

Keywords

Engineering

Reference:

Collections

Masters

Full item page