Chemical States of Water in Anion Exchange Membrane for Fuel Cells Using Raman and Coherent Anti-Stokes Raman Spectroscopies

Abstract

Purpose Low membrane durability remains a major challenge in anion exchange membrane fuel cells (AEMFCs). This is despite the AEMFCs’ proven potential to eliminate the dependency of Fuel cells on Pt catalysts and bring about the mass commercialization of the technology. This problem can be effectively addressed through combined optimization of the AEMFCs’ operating conditions during evaluation, followed by finetuning of the membrane chemistry to address any observed weaknesses. After synthesis, water distribution is the primary operation condition determining a given AEM’s performance and lifespan [1]. This is partly because water affects the ionic conductivity and, thus, maximum current density, which is known to play a significant role in the longevity of membranes. Therefore, establishing the water distribution during power generation for any given membrane is highly desirable. For this to happen, suitable and reliable techniques must exist to measure and compare the water distribution in situ and operando. In this paper, we used in-house assembled Raman spectroscopy and coherent anti-Stokes Raman scattering spectroscopy (CARS) systems to measure the water in an operational AEMFC for the first time. CARS comes with a high time resolution of 0.1 s, which is essential in the transient response study of FCs. Experimental Method QPAF-4 membrane film (Fig. 1) of the thickness of 30 µm and IEC of 2.0 meq g-1 was solution cast as reported [2]. The cell was then assembled, as reported [3]. An airtight transparent quartz window 200 μm thick was mounted at the cathode endplate for optical access to the QPAF-4 membrane through a 500 mm pinhole in the GDE. A thin Pt disk was placed on the opposite side of the membrane to reflect the signal to the spectrometer through the ×50 objective lens used. For Raman spectroscopy, a 632 nm, 1 mW laser with 5.5 min total exposure was used. For CARS, an 11 mW 785nm pump laser and 16 mW Stokes laser with 200 ms total exposure were used. The flow rate was 100 ml min-1 for all gases. The cell temperature was maintained at 60 °C. The relative humidity of the gas varied from 30% to 100% in steps of 10% RH. The system was allowed to stabilize for 3 hours at each humidity level before measurement. Results and discussion Figures 2 and 3 show the normalized in situ spectra recorded at the center of the membrane at different relative humidity values using CARS and Raman, respectively. The peak positions are comparable. However, the OH peak in CARS is taller relative to C=C than in Raman. CARS signal intensity is generally stronger than Raman [3]. Within the CARS peak, the purely OH portion of the signal is more prominent relative to the rest of the peak than in Raman. This is attributed to the fact that the Infrared laser associated with the stokes light (817-1144nm) of CARS is known to be more sensitive to OH vibration than the lower wavelength laser (632nm) used in Raman [4]. Conclusion We successfully developed a 785nm CARS system capable of operando water measurement in AEMFCs. The results from this technique are comparable to those from micro-Raman spectroscopy but with better signal intensity and time resolution. This increases the number of practical methods available to researchers interested in measuring water content in AEMFCs allowing for faster development of the technology. The water distribution results during power generation will be discussed at the conference. With CARS, higher temporal/time resolution is achieved, which is advantageous when cell response needs to be studied.