Deactivation and Recovery of Pt/C with Commercial Glycol-Based Coolants in Rotating Disk Electrodes and Membrane Electrode Assemblies

Thursday, 9 October 2014: 08:40
Sunrise, 2nd Floor, Star Ballroom 8 (Moon Palace Resort)
B. D. Gould (Naval Research Laboratory), Y. Garsany (EXCET Inc.), and K. Swider-Lyons (U.S. Naval Research Laboratory)
Numerous power applications utilize a coolant loop and radiator to exhaust excess heat to the environment.  In a multi-cell stack, coolant is typically circulated through the separating plates of each cell to maintain a uniform temperature distribution across the stack and keep the coolant isolated from the electrocatalyst.  If glycol-based coolants contact the catalyst layer in the electrodes, deactivation of Pt active sites is possible leading to the loss of fuel cell performance.  Coolant can contact the fuel cell catalyst through faulty sealing, porous materials or improper use.  In principle, glycols should not pose a risk to fuel cell electrocatalysts because organic alcohols are easily oxidized from the surface during operation.  However, the additives in commercial coolants can be highly active catalyst poisons with varying degrees of reversibility. 

We present the extent to which coolant materials deactivate commercial carbon–supported Pt nanoparticles (Pt/C) in membrane electrode assemblies (MEAs) in the context of polarization behavior, cyclic voltammetry and rotating disk electrode (RDE) experiments.  We have previously developed these methodologies to investigate Pt/C deactivation by SO2 [1,2] and recently applied them to study coolants [3].  The role of different coolant components such as corrosion inhibitors and surfactants will be identified and compared to the base glycol/water coolant.  In prior work, RDE showed that azole additives are highly poisonous to the Pt surface, while polyols can be removed from the Pt surface.  However, deactivation of RDEs by coolant occurs in excess liquid electrolyte, while in a fuel cell deactivation occurs in an air-coolant mixed phase exposure.  The objective of this paper is to compare RDE results to those in a single cell fuel cell.  Recovery methods and strategies will be discussed along with the difference between deactivation in the RDE environment versus the fuel cell environment. 

Figure 1a compares the activity of a pristine MEA to one exposed to a glycol-based coolant with an azole corrosion inhibitor before and after cycling the electrode between 0.07 and 1.2 V in H2/N2, anode/cathode to clean the surface.  Figure 1b compares the activity of the pristine MEA compared to one exposed to a glycol-based coolant with a polyol corrosion inhibitor and the activity after cycling the electrode between 0.07 and 1.2 V in H2/N2, anode/cathode.  Figure 1a clearly shows that glycol-based coolants with azole corrosion inhibitors are particularly damaging to fuel cell catalytic activity.  Even after aggressive recovery procedures the mass activity is 35% less than the pristine material.  The azole caused a permanent loss of Pt surface area when examined with cyclic voltammetry.  In comparison polyol corrosion inhibitors are not nearly as potent a catalyst poison and can be fully recovered with mild recovery procedures after the 1st exposure.  The difference in behavior can be rationalized by the strong coordination between nitrogen and sulfur containing aromatic rings in the azole with the d-bands of the Pt atoms on the surface of the nanoparticles.  In contrast, the polyols tend to oxidize off the surface. 

This work is significant because it confirms that the deleterious coolant components identified by RDE experiments match the results of single cell measurements. This agreement verifies that RDE results for coolant deactivation are scalable to practical fuel cells.  Interestingly, coolant exposure to MEAs is more deleterious than for RDEs and can cause greater loss of electrochemical surface area by Pt corrosion, even with benign coolants. 


[1]        Y. Garsany, O. A. Baturina and K. E. Swider-Lyons, J. Electrochem. Soc., 154, B670 (2007).

[2]        B. D. Gould, O. A. Baturina and K. E. Swider-Lyons, J. Power Sources, 188, 89 (2009).

[3]        Y. Garsany, S. Dutta and K. E. Swider-Lyons, J. Power Sources, 216, 515 (2012).