Time-of-Flight Neutron Imaging for the Distinction of Super-Cooled Water and Ice During PEFC Cold-Starts
Muriel Siegwart,Fusen Huang,Magali Cochet,Victoria Manzi-Orezzoli,Thomas J. Schmidt,Zhang Jianbo,Pierre Boillat
DOI: https://doi.org/10.1149/ma2018-02/41/1378
2018-01-01
ECS Meeting Abstracts
Abstract:The start-up of polymer electrolyte fuel cells (PEFCs) at sub-zero temperatures (cold-start) is a challenge for fuel cell designers and operators. The freezing of water in the cell can cause shutdown of the cell and severe damages to fuel cell components. However, water produced in the electrochemical reaction is pure enough to be present in the super-cooled state (liquid) at sub-zero temperatures (1). Methods that allow visualizing the location of freezing events during cold-starts help to understand which parameters influence the phase transitions. For this purpose our group has been using energy selective neutron imaging methods since 2013 (2). We exploit the fact that the attenuation for ice and super-cooled water differs at low neutron energies (wavelength > 4 Å) while it is identical for high energies (3). When water freezes in the pores of the catalyst and/or gas diffusion layer, its volume expands in all directions and the density decreases. To account for the density change, we record images in two different energy windows - one corresponding to high and the other to low energy neutrons – and obtain a relative attenuation image, which allows us to determine the state of aggregation. Energy selective neutron imaging with a neutron filter was successfully applied while operating a fuel cell with an active area of 50 cm2 for the detection of local freezing events (4). However, the obtained contrast between liquid water and ice was very low (1.6%), making the method particularly sensitive to biases. To increase the contrast, we use time-of-flight neutron imaging (TOF-NI), a method based on neutron pulses (either at a pulsed neutron source or with the help of a rotating chopper disk) where the neutron energy is determined by the time lag between the pulse and the moment the neutron reaches the detector. Compared to conventional TOF-NI, which uses short neutron pulses (duty cycles smaller than 1%), we use a chopper disk with broad slits making a duty cycle of approximately 30% and obtain a much higher neutron flux at the cost of lower energy resolution. Here, we will present experimental results obtained during isothermal cold-starts with a cell having an active area of 4.4 cm2. Simultaneously to neutron imaging, we measured the heat fluxes in and out of the cell with a setup described in other references (5). Figure 1 contains an example of the results obtained during a cold-start performed at -7.5°C. Figure 1 (a) depicts the evolution of voltage and heat flux over time during the cold-start while Fig. 1 (b) shows the relative attenuation (ratio between low and high energy attenuation) averaged over the whole active area (A), and over the channels B and C (the different areas are represented on Fig. 1 (e)). During phase I, larger accumulations of super-cooled water were present in the gas flow channels. The phase transition of super-cooled water to ice occurred 62 minutes after starting to draw current. This is visible from the voltage drop and the peak in heat flux due to the latent heat that is released during freezing of water [Fig 1 (a)]. Phase II corresponds to the time after the freezing event where the cell was not operated and its temperature was kept constant at -7.5°C. Figure 1 (c) and (d) depict the high energy attenuation images, from which the water thickness is retrieved by using the Beer-Lambert law. Those images are representative for classical transmission images without energy selectivity. The images representing the relative attenuation (attenuation at low energy divided by attenuation at high energy) are shown in Figure 1 (e) and (f). For clarity, only channel water is shown in these images. By comparing the classical attenuation based images [Fig 1 (c-d)] with the energy selective approach [Fig 1 (e-f)] it becomes evident that the high energy attenuation does not clearly differ for phase I and II. In contrast, the relative attenuation for ice measured in phase II [Fig 1 (b and f)] is – as expected – lower compared to the one measured for super-cooled water in phase I [Fig 1 (b and e)]. By using TOF-NI we could improve the contrast between super-cooled water and ice to reach approximately 5%. P. Oberholzer et. al, J. Electrochem. Soc., 159, B235 (2011). J. Biesdorf et. al, Phys. Rev. Lett., 112 (2014). L. Torres et. al, Nuc.l Instrum. Meth. B, 251, 304 (2006). P. Stahl et. al, J. Electrochem. Soc., 163, F1535 (2016). M. Cochet et. al, Submitted for publication (2018). Figure 1