Environmental Engineering Reference
In-Depth Information
[L ´ pez-Cudero et al., 2005]. This was related to the observation that the highest CO
coverage of 0.68 ML is obtained only if the dosing potential is below 0.25 V, i.e.,
in the hydrogen adsorption region. The pre-peak probably corresponds to the
oxidation of CO at certain defect sites, and seems to be related to the initial plateau
current observed in chronoamperometric transients (see the last paragraph of
Section 6.2.1.1). Stamenkovic and co-workers have observed that in the pre-ignition
region there is a blue shift (lowering) of the C22O stretching frequency corresponding
to the linearly bonded CO, which they ascribe to bisulfate adsorbing onto the Pt sites
that were previously occupied by CO, leading to a mild compression of the CO islands
and an associated enhanced dipole - dipole coupling [Stamenkovic et al., 2005].
6.2.1.3 Continuous CO Oxidation on Platinum The main difference
between CO stripping and continuous CO oxidation is the CO (re-)adsorption
Reaction (6.3). In contrast to CO stripping, this leads a steady-state CO oxidation
current because of the continuous supply of CO. In modeling the continuous CO
oxidation, we also need to consider the mass transport of CO from the bulk of the sol-
ution to the electrode surface. The temporal change in the CO coverage is now given by
du CO
dt ¼ k ads c CO,s (1 u CO ) k(E)u CO (1 u CO )
(6 : 12)
where c CO,s is the CO concentration in solution and k ads is the CO adsorption rate
constant. This equation must be coupled to an equation for the temporal evolution of
the surface concentration of CO in which the diffusion/mass transport is taken into
account. The equations for modeling this situation may be found in the original
reference [Koper et al., 2001]. The important point here is that, in the case that CO
adsorption is stronger than OH adsorption over a wide potential range, such that it
leads to self-poisoning as is observed experimentally, the model predicts a very peculiar
current - potential polarization curve.
The thick line in Fig. 6.8 shows the steady-state polarization curve predicted by this
model. In contrast to the usual sigmoidal shape of a steady-state polarization curve
for a first-order heterogeneous electrode reaction coupled to diffusion, the curve is
S-shaped owing to the nonlinear autocatalytic nature of the CO oxidation reaction,
as discussed in Section 6.2.1.1. The thin line in Fig. 6.8 is the predicted cyclic voltam-
metric curve obtained at a low scan rate. Starting from a low potential, the system
follows the low current branch until it reaches E ¼ E 2 . At this potential, there is a
very sharp current spike, which basically corresponds to a fast oxidation of the poisoning
CO adlayer that was essentially stable for E , E 2 , after which the current settles down on
the high current branch, which corresponds to a diffusion-limited CO oxidation. After
reversing the scan, the system stays on the high current branch until E ¼ E 1 .Atthispoten-
tial, the current quickly drops to the low current branch. The observed hysteresis between
E ¼ E 1 and E ¼ E 2 cannot be removed by lowering the scan rate. It is intrinsic to the
system, owing to the S-shaped nature of the polarization curve.
In general, the onset of CO oxidation occurs at higher potentials than those
recorded for the stripping peak in the absence of CO in solution, i.e., E 2 values are
Search WWH ::




Custom Search