Towards the “Pressure and Materials Gap”: Hydrogenation of Acrolein . . .
421
The dependence of the selectivities on hydrogen partial pressure is more
difficult to understand. Calorimetric measurements indicate, that the hydrogen
interaction with silver catalysts is very weak, but exhibits a structure sensitivity.
◦
At 150 C, reversible adsorption with relatively fast desorption has been found
over the IW catalyst (small particles), whereas only very slow desorption of hy-
drogen has been observed over the P-catalyst (larger particles), together with
a very small amount of irreversibly adsorbed hydrogen. This means, that on the
P-catalyst, surface sites must exist, which are able to stabilise the adsorbed hy-
drogen. These sites do not exist in the IW catalyst. Additionally, the above TAP
results (see Sect. 3.3) have shown, that hydrogen adsorption depends strongly
on catalytic materials. Most likely, the different structural properties, i.e. size,
edge to plane ratio of the different catalysts are responsible for this behaviour.
It may be, that hydrogen activation may only take place at defect sites or kinks,
since it is known that no hydrogen dissociation should occur on silver single
crystals [10]. Hydrogen atoms may than diffuse into more stable adsorption
sites, probably on flat surfaces. Interestingly, the interaction of hydrogen with
silver catalyst could be observed at very low pressures (TAP reactor) as well
as at ambient pressure (calorimetry). However, all this can not explain the de-
pendence of selectivity on H partial pressure. But, since there is, in opposite
2
to the very weak interaction of hydrogen with single crystal silver surfaces,
a somewhat stronger interaction of hydrogen with silver nanoparticles, it may
be concluded that the hydrogen adsorption including the surface coverage is
pressure dependent. This pressure dependent surface coverage influences the
catalytic properties of the silver surfaces and leads to a pressure dependent re-
activity and selectivity. Also, it is possible that diffusion of hydrogen into the
bulk of the silver nanoparticles can, in opposite to the behaviour at single crys-
tals, take place at higher hydrogen pressures, modifying the catalytic properties
of the silver [22].
However, as pointed out in part 3.1, the selectivity is not only influenced
by the pressure but also by the catalyst preparation method, which most likely
means by the size, shape and the surface structure of the catalyst particles.
Different preparation methods lead to different particle sizes and size distri-
bution, as can be seen from the TEM result. Mohr et al. have, in a study on
supported gold catalysts, concluded, that crystal planes lead to the formation
of propanal, whereas edges favour the formation of allyl alcohol [23]. In this
context, larger particles, having relatively more atoms on planes than on edges,
should favour the formation of propanal. This is consistent with the results
obtained in this study, where the catalyst with the smaller particles (9Ag/SiO2-
IW) produced the higher amount of allyl alcohol, whereas the larger silver
particles (9Ag/SiO -P) produce lower amounts of allyl alcohol. Indeed, for
2
catalysts with even larger particles, prepared via sputtering of a silver target,
the selectivity to allyl alcohol has been found to be even lower than for the
P catalyst (ca. 20 %). These results would also be in agreement with results
obtained by Fujii et al., which concluded from IR measurements on evapo-