Metalloporphyrin Catalysts of Oxygen Reduction - Fuel Cell Catalysis: A Surface Science Approach - page 641

conditions using a catalyst whose n value is reliably established, for example graphite
(n ¼ 2) or thoroughly cleaned Pt (n ¼ 4, Fig. 18.7b). The latter method may provide
more precise estimates of n av , as the values of D, n, and [O 2 ] bulk under given exper-
imental conditions are usually not known with sufficient precision (for example, the
literature values of D vary by a factor of 2 [Opekar and Beran, 1976] and solutions
oversaturated in O 2 are easily obtained if O 2 is continuously bubbled in).
Reliable estimates of k typically cannot be obtained from a single Koutecky -
Levich intercept, often referred to as i k 21 , because of uncertainty in the amount of
catalytically active porphyrin at the electrode. To obtain reliable estimates of k,
Koutecky - Levich intercepts acquired at electrodes with varying surface coverage
of the catalyst, G cat , have to be plotted against G cat . It is typical to observe a direct pro-
portionality between i k 21 and G cat at low surface coverages of the catalyst ( provided
that O 2 reduction at the electrode material does not interfere) that plateaus as the cover-
age increases and may even decrease with a further increase in coverage (Fig. 18.8),
signifying a change in the rate-determining step from catalytic reaction to mass or
charge transport in the catalytic film. Only at the catalyst coverages that yield intercepts
directly proportional to G cat can the Koutecky - Levich equation be applied.
One also needs to be careful when using the slope of the Koutecky - Levich plot to
determine n av of the catalytic film. Examples of metalloporphyrin-catalyzed ORR
have been reported where, above a certain value of the electrode rotational frequency,
the catalytic currents became independent of v, indicative of a breakdown of the
Koutecky - Levich model, either because the rate of charge or substrate transfer
within the film became rate-limiting or the catalyst became partially saturated with
O 2 [Boulatov et al., 2002; Song et al., 1998; Collman et al., 1980]. In other cases,
the i 1
cat

versus v 21/2

graphs may remain mostly linear within the experimental

Figure 18.8 The inverse of the intercept of the Koutecky - Levich plots normalized to the
geometric area of the electrode, j k , versus surface coverage of two biomimetic Fe porphyrin cat-
alysts deposited on an edge-plane graphite electrode [Boulatov et al., 2002]. At surface cov-
erages ,1 - 2 nmol/cm 2 (depending on the catalyst), j k is directly proportional to G cat .
Depositing more catalyst leads to a drop in j k , suggesting a change in the turnover-determining
step. The surface coverages are per geometric area; because of the high roughness of the edge-
plane graphite, 1 - 2 nmol/cm 2 may correspond to only few monolayers of the catalyst.