Cathode catalyst development for use in PEM fuel cells, investigation of cell efficiency and performance modelling

Powered by hydrogen, proton-exchange-membrane (PEM) fuel cells generate electricity directly while leaving only a minimal environmental footprint. The anode and cathode electrodes in these cells typically employ carbon-black-supported catalysts containing precious metals such as Pt and Ru. The reduction reaction at the cathode proceeds far more slowly than hydrogen oxidation at the anode, making acceleration of the cathode catalyst critical. Consequently, much of the literature focuses on reducing Pt usage through bimetallic catalysts and alternative supports. Reduced graphene oxide (rGO) and carbon nanotubes (CNT) stand out, yet their combined use has not been explored in depth, and no studies have reported on the fuel-cell performance, cathodic activity, or bimetallic behaviour of Pt/rGO-CNT catalysts. In this study, alongside commercial carbon black, four different supports (GO, rGO, CNT, and rGO-CNT) were prepared. Catalysts carrying either Pt or Pt–Ni were synthesized on these supports via polyol and colloidal routes. For the polyol method, the effects of temperature (120 °C and 190 °C), reaction time (15 min and 35 min), CNT type (raw vs treated), and mixing protocol (single-stage magnetic vs multistage ultrasonic) were investigated. The catalysts were characterised by TGA, XRD, FT-IR, BET, SEM, and EDS, while their activities were assessed via cyclic-voltammetric ECSA measurements and PEM fuel-cell tests. CNT addition prevented rGO sheet restacking and boosted the BET surface area ten-fold. The optimum polyol conditions were 190 °C for 15 min, affording 92.5 wt % Pt loading efficiency, a Pt crystallite size of 3.55 nm, and an ECSA of 125 m² g⁻¹ Pt. Under these conditions the fuel cell delivered 292 mW cm⁻² at 70 °C—22 % higher than a commercial Pt/C catalyst. A regression model indicated optimum operating conditions of 65,9 °C and an H₂/O₂ volumetric ratio of 1.0; model predictions matched experimental values within ±4 mW cm⁻².