Influence of Ag on Co Behavior
J. Phys. Chem., Vol. 100, No. 10, 1996 4235
however, that the images obtained from this technique are only
TABLE 3: Percent Product Distribution for Ethylene/
Hydrogen (1:4) Decomposition over Co/Cu (98:2) and Co/Ag
(98:2) after 60 min at 600 °C
2
-dimensional in nature. Despite this limitation, it is possible
to monitor the modifications in metal particle morphology as a
function of the nature of the gas environment, and this operation
emphasizes the difficulties that one can encounter in attempting
to determine the characteristics of such particles by traditional
chemisorption methods. It is also apparent that the morphology
and adsorption properties of the catalyst particle when heated
in the presence of a reactant gas are going to be vastly different
than those in the presence of pure hydrogen at room temperature,
conditions where most chemisorption experiments are per-
formed. A further shortcoming of the chemisorption method
is that one assumes that the gaseous molecule adsorbs uniformly
on all the exposed surfaces of the metal particle. Lin and
Somorjai have shown from surface science studies that distinct
crystallographic faces of platinum can chemisorb hydrogen while
other faces are incapable of performing this function. This
aspect is even more apparent for the results presented here where
the active surface for precipitating carbon is in fact encapsulated
by the filamentous structure and is therefore unavailable to
chemisorb any gas. It is due to these problems, among others,
that we believe that the often-cited “turnover frequency” or
% yield
ethylene
catalyst
carbon
ethane
methane
Co/Cu
Co/Ag
48.47
71.21
22.59
22.74
5.08
4.30
23.84
1.74
Under these particular reaction conditions, high cobalt content
and high hydrogen partial pressure, the activity of the silver-
containing catalyst for carbon deposition is actually higher than
its copper analogue. It is significant, however, that the total
conversion of ethylene and yields of ethane are similar for both
catalysts. Surface science studies have shown that the electronic
nature of the metal can dictate the mode of adsorption of
26
2
8
ethylene on the metal. If the magnitude of the electronic
perturbation of cobalt by copper is appreciably different than
that induced by silver, this aspect could be reflected in a
modification in the ethylene adsorption characteristics on these
bimetallic surfaces. As a consequence, one would not be
surprised to find variations in the product selectivity pattern. A
further feature to be taken into consideration is that when
particles are undergoing reaction several crystal faces are
generated. Each face can possess an arrangement of atoms that
exhibit different adsorption characteristics with respect to the
reactant gas. This behavior can lead to the formation of a variety
of products, i.e., ethane or methane. The formation of an
ethylidyne intermediate may be favored on the exposed faces
of copper-modified cobalt, but the atomic arrangement of the
cobalt atoms at the surface may be different when silver is
present and the formation of this particular reactive intermediate
may not be a facile process. This argument could also be
extended to the events occurring at the metal/solid carbon
interface where the nature of the adatom could induce modifica-
tions in the crystallographic orientation of the faces responsible
for carbon precipitation. This effect would be manifested in a
difference in the structural and conformational characteristics
of carbon filaments generated by the respective cobalt bime-
tallics, as was observed experimentally.
“turnover number” calculated from chemisorption measurements
has little value when studying filamentous carbon formation
from metal/hydrocarbon interactions.
Comparison of the Behavior of Cobalt-Silver and Cobalt-
Copper. A comparison of the data presented here with those
obtained previously for the cobalt-copper-catalyzed decom-
position of ethylene to produce filamentous carbon (1) reveals
the existence of a number of differences in the behavior of the
two bimetallic systems. Cobalt-copper was found to exhibit
a high activity for the conversion of pure ethylene to solid
carbon, ∼70%, over a wide catalyst composition range. In
contrast, under the same conditions the activity of the cobalt-
silver system for this reaction was considerably lower, and the
maximum conversion of ethylene was obtained over a relatively
narrow catalyst composition range (Figure 1). This distinction
in the catalytic activity of the two bimetallic powders cannot
be accounted for by the selective formation of alloys in one of
the systems since both silver and copper exhibit only limited
bulk miscibility in cobalt.15 As previously stated, we believe
that in addition to geometric effects electronic interactions
between the two components are operative in these systems,
and the degree to which these two factors impact on the behavior
of the cobalt bimetallics will be a function of the adatom. In
this context it is important to consider the models devised by
Goodman and co-workers27 that predict the direction of electron
transfer resulting from the mixing of two metals. From this
model we would expect that the direction of any electron transfer
in either copper-cobalt or silver-cobalt to be from cobalt to
the adatom. The result of this electron flow would be a change
in the strength of the interaction of ethylene molecules with
the surface cobalt atoms, and this aspect may lead to the
observed enhancement in carbon deposition activity for both
bimetallics.
Summary
From this investigation it is evident that the addition of a
small amount of silver to cobalt has a profound effect on the
catalytic behavior of the host metal toward the formation of
filamentous carbon when the bimetallic is heated in the presence
of ethylene. While one might rationalize the modification in
catalytic action of the system in terms of a geometric effect
when significant amounts of silver are present, it is difficult to
conceive of how such an explanation can account for the
dramatic enhancement in activity when only 1% of the additive
is incorporated into the particles. It is possible that this change
in behavior is attributable to electronic perturbations in the cobalt
lattice brought about by the addition of silver to the system,
which results in an improvement in reactivity of the surface
toward ethylene decomposition. Similar effects are observed
when copper is introduced into cobalt; however, in this case
the gas phase selectivity pattern is significantly different than
that observed in the present system. In the former system,
methane was produced in similar amounts to that of ethane,
whereas when ethylene was decomposed over a cobalt-silver
catalyst the ratio of ethane to methane formed was between 10
and 30.
Methane and ethane, in approximately equal quantities, were
found to be the major gaseous products when ethylene/hydrogen
mixtures were passed over a selection of cobalt-copper
catalysts. These results are to be compared with the highly
selective formation of ethane in preference to methane found
in the current work under a variety of reaction conditions and
cobalt-silver catalyst composition ranges. A synopsis of the
total product distribution obtained from the decomposition of
an ethylene/hydrogen (1:4) mixture over cobalt doped with 2%
copper and 2% silver at 600 °C is shown in Table 3.
Acknowledgment. Financial support for this work was
provided by the United States Department of Energy, Basic
Energy Sciences, Grant DE-FG02-03ER14358.