1340
(Invited) Elucidating PEMFC Voltage Loss Contributions Via Mathematical Modeling and Differential Cell Testing

Tuesday, 31 May 2016: 10:40
Indigo 204 B (Hilton San Diego Bayfront)
W. Gu, S. Arisetty, S. Kumaraguru, and M. Mathias (General Motors, Fuel Cell Activities)

Proton-exchange-membrane fuel cells (PEMFC) have been greatly advanced towards commercially viable hydrogen-powered fuel cell systems for automotive application in recent years. Lower fuel cell costs have been achieved by reducing the loading of Pt catalyst in the electrodes. Unfortunately lower catalyst loadings accompany higher losses in cell voltage. To quantify various voltage loss terms and thus identify the opportunity for improvement, mathematical models with parameters extracted from experiments are commonly employed. A differential cell operates at high stoichiometry flow with minimal pressure drop down the channel, making it amenable to treatment by a one-dimensional model across the cell. Hence the experiments are conducted on it to extract model parameters. The cell performance can be described by:

    Ecell = Erev - iRΩ - ηHOR - |ηORR| - i(RH+,a + RH+,c) - ηtx,O2          (1)

where Ecell is the cell voltage; Erev is the reversible cell voltage, i is the current density; RΩ is the sum of the Ohmic resistances of proton conduction through the membrane and of electron conduction (commonly referred to as high frequency resistance or HFR); ηHOR and ηORR are the charge transfer overpotentials for the hydrogen oxidation reaction (HOR) and the oxygen reduction reaction (ORR), respectively; RH+ is the effective resistance to proton conduction in an electrode with subscripts a and c denoting anode and cathode, respectively; and lastly, ηtx,O2 is the oxygen transport loss which consists of two parts: one results from oxygen transport through gas phase, and the other is associated with the oxygen transport resistance local to the Pt surface, namely

               ηtx,O2 = ηtx,O2 (gas) + ηtx,O2 (Pt-local)               (2)

Differential cell tests with an active area of 5 cm2 were designed to study the effects of Pt location and Pt dispersion in cathode electrodes. Membrane electrode assemblies (MEAs) with different carbon supports and various Pt loadings were fabricated and tested. In particular, Pt/HSC (high-surface-area-carbon) and Pt/Vulcan® catalysts are tested to discern the impact of Pt particles embedded in a carbon particle. Electrochemical performance of the electrodes were characterized with electrochemical active area, mass activity and polarization curves. Additionally, protonic resistance of the catalyst layer and oxygen transport resistances were calculated with impedance and limiting current techniques.

An electrochemical model coupled with transport across the cell was developed to interpret the differential cell data. The novelty of the model is to consider the Pt utilization for the Pt particles inside the carbon support without ionomer coverage based on catalyst morphology. The oxygen transport resistance local to Pt surface, that is hypothesized to originate from ionomer-Pt interaction, is evaluated by model-data comparison. The model parameters perceived to be critical for accurate cell performance prediction and the voltage loss terms are examined by sensitivity analysis. Research needs for characterizing catalyst morphology and electrode structure will be discussed.