(6fd) First-Principals Modeling of Methanol Fuel Cells: Kinetics and Catalyst Design

First-Principals Modeling of Methanol Fuel Cells: Kinetics and Catalyst Design

Glen R. Jenness

The decreasing supplies of petroleum resources coupled with an increase in energy demand has created a need for an alternative power supply. Proton-exchange membrane fuel cells (PEMFCs) are an attractive prospect, due to the high energy density and low emissions. Previous PEMFC technology relied on the oxidation of H₂, but due to the difficulties of storing H₂, this route is not economically favorable. By replacing H₂ with an alternative renewable proton source we can alleviate the cost of storing and transporting H₂, as well as making the PEMFC technology more environmentally friendly. Methanol is an attractive proton source due to low storage costs, good thermal stability, high energy density, and is readily produced from biomass resources. While methanol fuel cells offer a high energy density, the efficiency remains low, limiting the use of methanol fuel cells. As a consequence, there has been an increased effort in recent years to find an efficient catalyst for the oxidation of methanol. Previous catalysts employed a platinum nanoparticle with an oxide support, such as γ-Al₂O₃. However, recent experiments indicate that a platinum nanoparticle supported on a carbon-based support, such as carbon nanotubes (CNT), have led to an increase in the efficiency of methanol fuel cells.

Several challenges still remain to be resolved in order to increase the efficiency of methanol fuel cells. Firstly, the CNT itself can carry an intrinsic reactivity, with the reactivity being determined by the metallic or semi-conducting nature of the CNT, arising from the chirality. However, the effect of the chirality on the reactivity of the CNT support is still debated. I have shown that the choice of nanoparticle support can result in a charge state being induced on the nanoparticle, which in turn affects the bonding/anti-bonding interactions between the reactants/products and the catalyst, with the resulting charge state being related to the band gap of the support material.1,2 As the band gap of the CNT is driven by the chirality, the Pt–CNT interaction will also be driven by the chirality. Additionally, the efficiency of CNT supported catalysts can be increased via the doping of the CNT with nitrogen, which in turn changes the electronic properties of the CNT (i.e. band gap), as well as introduces a Lewis acid center into the support. As the mechanism for the oxidation of methanol on the Pt/CNT catalyst is still in question, the effect of introducing a Lewis acid center is currently unknown. I have also shown that the strength of the Lewis acid center can be related to the atomistically resolved conduction band properties of the Lewis acid center, with the reactivity being related to the strength of the Lewis acidity.3,4 Finally, it should be mentioned that one factor preventing the wide spread use of methanol fuel cells is the cost of the Pt metal, and it would desirable to use a base metal such as nickel or iron, without sacrificing efficiency. Through the use of computational tools (i.e. state-of-the-art ab-initio methods such as density functional theory), I seek to resolve these issues. In particular, a) how the metallic vs. semi-conducting nature of the carbon nanotube affects the methanol oxidation process; b) how the effect of doping the CNT affects the electronic properties of the catalytic system; c) how the introduction of a Lewis acid center into the support affects the mechanism and kinetics of the methanol oxidation process. The investigation of these effects from an atomistic level will allow us to understand the origin of these interactions, which enable a rationale and systematic improvement of the efficiency of methanol fuel cells.

(1) Jenness, G. R. et al. ACS Catal. 2013, 2881–2890.
(2) Hermes, E. D. et al. Mol. Simul. 2014, 1–11.
(3) Jenness, G. R. et al. J. Phys. Chem. C 2014, 118, 12899–12907.
(4) Jenness, G. R. et al. J. Phys. Chem. C 2015, 119, 5938–5945.