Antimony Tin Oxide: From Computational Research to MEA Manufacturing Challenges in the Value Chain
Résumé
One of the main objectives in SMARTCat is to synthetize conductive tin oxide (SnO2) as a support material for
the electrocatalyst. The task is to systematically improve specific properties such as electronic conductivity (σ),
surface area (SA) and stability of SnO2-based support materials. Through such improvements, the goal is to
provide a PEM fuel cell cathode catalyst with enhance durability and stability compared to the state of the art
carbon-based support.
The properties of base SnO2 were tailored by (i) different dopants (niobium and antimony) to achieve sufficient
electronic conductivity, and (ii) different synthesis techniques to influence the particle size, surface area, and
pore size distribution. To synthesize this material, two different techniques (co-precipitation and flame spray
pyrolysis), with preferable potential to be up-scaled was used. In parallel, density functional theory (DFT)
calculations were performed to (i) establish the correct description of the SnO2/Pt catalyst-support interface in
order to obtain an accurate description of its electronic structure, (ii) model migration and localization of
antimony (Sb) and niobium (Nb) dopants in Sb- and Nb-doped tin oxide (ATO/NTO) at the Pt(111)-SnO2(110) as
a function of doping levels, (iii) segregation of the doping elements (Sb and Nb) through the interface and
catalyst layer, and (iv) to calculate the electronic transport across the interface as a function of doping levels
and segregation using DFT calculations and non-equilibrium Green’s function (NEGF) theory.
Pure tin oxide has shown poor electronic conductivity (~2⋅10-6 S/cm) as compressed powder (5.6 MPa). Thus,
doping of SnO2 has been done with different atomic level of Sb and Nb. Based on the calculations and
conductivity results, it has determined that doping with 7 or 15 at.% of Sb are the best option, increasing the
conductivity around 104 to 106 times, where the electronic transport across the interface as a function of
doping levels played a important role. Modified co-precipitation method allowed us to produce samples with
SA in of 50 – 90 m2/g, whilst FSP-synthetized SnO2-based oxides showed a SA of 90 – 115 m2/g. With both
synthesis methods, mono- and bimodal pore size distribution in the range of 20 to 100 nm was obtained. In
order to investigate the influence of dopant-segregation (related to its stability) on the transport properties of
the whole system, modeling and experimental approaches have been performed. In general, experimental and
modeling results (both in good agreement) confirmed that in the initial segregation processes, the
conductance of the system increases with the increase of the Sb concentration trough an improvement of the
electron transport properties of the system. After reaching a saturation point at a particular concentration
(Sb/Sn = 0.33 within the first 4 layers from calculations), the transport current will then decrease. In addition,
Sb-enrichment of the outermost layer have shown to induce stabilization of the electric current and thus of
the conductance, especially at voltages higher than 0.6 V.
Deposition of catalyst has been done using the well-known microwave assisted polyol-method. However, this
well established method for carbon-based catalyst is not straight forward applicable to metal-oxide based
supports. Issues concerning particle agglomeration during the synthesis are remarkable. Alternative deposition
methods have been also considered. The fact that metal-oxides are itself very reactive materials (contrary to
carbon) makes the challenge more difficult when inks from those materials are intended for the catalyst layer
fabrication. Alternative methods (e.g. magnetron sputtering) of catalyst layer manufacturing are explored.
Thus, parallel to modified/optimized metal-oxide-ink-recipes, studies focusing on improving and optimizing
target manufacturing are in progress. With those studies the goal is to produce MEA's based on porous
catalyst layers utilizing antimony tin oxide as support.