Environmental Engineering Reference
In-Depth Information
chloroplatinic acid, a large number of nanometer-size Pt metal nuclei on the carbon
support, while slowing down the rate of excessive Pt crystallite growth. Alloying of
the Pt nanoparticles is typically achieved by reaction of the carbon-supported Pt crystal-
lites with added oxide of the electropositive metal, e.g., Co or Ni, and results in some
increase in size of the catalyst particles, typically from around 2 nm in unalloyed form to
4 - 5 nm in the alloyed form. These high electrode surface area amplification factors, of
the order of 10
2
-10
3
, have clearly been key for achieving the specific activity of hydro-
gen/air PEFC electrodes, enabling lowering of the loading required in fuel cell stacks to
only 0.2 g Pt catalyst per kilowatt of power generated [Gasteiger, 2005].
Particle size effect: Since the total surface area of the dispersed catalyst available at
some given mass loading (in milligrams per geometric square centimeter, mg/cm
2
geo.), is inversely proportional to the particle diameter, further reduction of the Pt par-
ticle size to less than 2 nm has been pursued by several groups, with associated efforts
to modify the carbon support so as to reduce inter-Pt particle distance (see Chapter 10).
However, this route to further increase of the Pt catalyst surface area has encountered
difficulties, interpreted by most researchers by the lower intrinsic activity obtainable
per cm
2
Pt at such ultrasmall particle sizes. This lower catalyst surface activity reported
for such small particle sizes, particularly in the oxygen reduction process, has been
understood to be the result of a larger fraction of Pt atoms located in edge site,
rather than in terraces of catalytically preferred Pt crystal surfaces, or in steps that
could enable multisite interaction. Strong evidence for the benefit of trading off ultra-
high surface area for a “smoother” catalyst surface morphology has been provided by
the unique development at 3M Company of electrocatalyst layers based on sputter
coating by Pt of an array of micrometer-long inert dendrites that are subsequently
embedded into the ionomeric membrane [Debe et al., 2003]. Activities per cm
2
Pt
obtained with this structure of a supported catalyst are about five times higher than
those recorded for 2 nm size Pt particles supported on carbon, fully compensating
for the similar loss in Pt surface area per unit mass compared with the case of 2 nm
supported Pt particles. It should be noted here that a dissenting opinion in this
regard has been expressed through the years by Watanabe and co-workers, namely
that the loss of activity with drop in Pt particle size has only to do with exacerbated,
localized mass transport limitations within the structure of the catalyst layer, caused by
low interparticle distance (see Chapter 10). Whereas this dissenting opinion suggests
that further optimization of catalyst layer composition and structure could enable
the use of Pt particles of diameters smaller than 2 nm, i.e., enable higher surface
area per unit mass of Pt, most researchers in the field have accepted the conclusion
that reducing particle size to below 2 - 3 nm does not enhance the activity per unit
mass of Pt, because of intrinsic surface catalysis reasons, apparently to do with the
surface atomic structure of such very small Pt particles.
Catalyst layer architecture: As a consequence of the diminishing returns from ever
higher dispersion, the effort to increase the active catalyst surface area per unit mass of
Pt has centered in recent years primarily on optimization of catalyst layer properties,
aiming to maximize “catalyst utilization” in fuel cell electrodes based on Pt catalyst
particle sizes of 2 - 5 nm. High catalyst utilization is conditioned on access to the
largest possible percentage of the total catalyst surface area embedded in a catalyst
Search WWH ::
Custom Search