Proton exchange membrane fuel cells, PEMFC’s, offer a clean, flexible mode of
energy generation, though efficiency improvements are required before they can
become commercially viable. While platinum is currently the most commonly used
cathode catalyst, a large overpotential exists for the oxygen reduction reaction, ORR,
limiting the effective potential of a fuel cell to approximately 0.9 V.
Efforts have been made within the literature to develop a low cost, efficient and
durable alternative cathode catalyst for use within a PEMFC though at present no viable
alternative has been found. However, M-Nx/C active sites, consisting of a central metal
atom co-ordinated to two or four nitrogen atoms embedded within a carbon support,
show great promise as possible replacements to platinum. While advances have been
made in the identification of possible active sites, no study yet exists that examines the
influence of each active site component on the overall activity.
In this thesis the influence of each component is examined, and in doing so a highly
active site is predicted. The efficacy of the model is first proven by a simultaneous
computational and experimental study of the activity of both platinum and Fe/Coporphyrins,
which are commonly used as precursors in the development of highly active
M-Nx/C catalysts. 16 active sites are modelled using graphene and amorphous carbonlike
ligands embedded with two or four nitrogen atoms and co-ordinated to either a
cobalt or iron centre. The activity of each active site towards the ORR is assessed by the
calculation of redox potentials, and by modelling 20 different elementary reactions
which collectively form a comprehensive reduction mechanism.
By direct comparison between each active site, utilising natural population, bond
orbital and localised molecular orbital analysis along with electrostatic potential maps,
the influence of each constituent part is quantified. Finally, it is predicted that active
sites consisting of a disrupted graphene ligand, embedded with four nitrogen atoms coordinated
to either an iron or cobalt centre, would demonstrate the highest activity
towards the ORR, and that such an active site is responsible for the activities reported
within the literature.
