76
BIRCHEM ET AL.
==
molecules, and hence the least hindered C O unsaturation
Ourresults, inparticulartheincreaseinselectivitytoward
==
is preferentially hydrogenated. Second, at the interface be- the preferential hydrogenation of the C O bond of prenal,
tween tin clusters and platinum atoms, a positive electronic and in perfect agreement with these literature results.
effect due to partially oxidized Snꢆ+ species activates the
polar carbonyl group of the prenal molecule. Such an ex-
planation was previously proposed by Poltarzewski et al. to
5. CONCLUSION
account for the promoter effect of tin during the hydrogena-
tion of acrolein and cinnamaldehyde on Sn–Pt/nylon (19).
The gas-phase hydrogenation of 3-methyl-crotonal-
dehyde (prenal) was investigated on a well-defined stepped
Pt(553) surface at pressures near 1 atm. This surface is twice
as active as the (111) plane, which underlines the higher
activity of the low-coordination step atoms for the dissoci-
ation and diffusion of hydrogen.
“2” ML of tin. Activity does not decrease any further,
confirming an island-type growth mode of tin adatoms,
which leaves roughly the same free platinum area. The se-
==
lective hydrogenation of the C O bond of prenal is even
The influence on selectivity of a preadsorbed submono-
layer of tin depends on the concentration and local arrange-
ment of this metallic additive. On Pt(111), at low coverage,
the influence of tin is, as expected, in favor of the unsatu-
rated alcohol and is ascribed to a mainly electronic effect.
On Pt(553), at low coverage (ꢃSn = 0.3), Sn atoms are likely
to be located in the vicinity of the steps and induce a local
reconstruction of the platinum surface. Under these con-
ditions, the surface is not saturated in organic species but
is enriched in hydrogen, and the reaction leads mainly to
the formation of saturated alcohol. At higher coverage, 1 or
2ML, tingrowsinislandsonthe(111)terracesofthesurface
and these clusters induce a change in the activation mode
of the molecule due to steric and electronic effects. The
metallic additive then recovers its well-known promoter
effect, leading to a higher selectivity toward unsaturated
alcohol.
more favored than in the presence of 1 ML of tin, due to
an enhanced electronic effect. Note that the fraction of
unsaturated alcohol reaches 70%, slightly higher than on
clean Pt(553).
In comparison with the literature, the observed influ-
ence of tin, when it is not restricted to a perturbation of
the steps, i.e., for ꢃSn > 0.3, can be related to two recent
studies, one dealing with CO and H2 adsorption (39) and
the other dealing with isobutane hydrogenolysis (40) on
Sn–Pt(111). In the former study, the chemisorption of CO
on Sn–Pt(111) surface alloys exhibiting the characteristic
√
√
p(2 ꢂ 2) and ( 3 ꢂ 3)R30ꢃ LEED patterns has been in-
vestigated. The CO uptake is larger than expected if Sn only
acted as a 1 per 1 site blocking atom [only 5% decrease in
CO coverage on the p(2 ꢂ 2) alloy]. Moreover, let us note
thattheauthorsdonotconsiderapossibleadsorptionofCO
on tin atoms. In contrast, tin is a drastic poison for hydrogen
adsorption. The alloy surfaces are practically inert toward
H2 dissociation at 150 K, assigned to the removal of active
ensembles of adjacent Pt atoms; similar behavior was ob-
served for supported Pt–Sn alloy catalysts (19). This result
appears contradictory to our finding concerning hydrogen
coverage, at least at ꢃSn = 0.3, or simply confirms that our
catalytic surface is a nonalloyed Pt–Sn bimetallic. Opposite
behavior of alloyed or nonalloyed surfaces have already
been evidenced in oxygen chemisorption (39): tin totally
blocks oxygen chemisorption on Pt–Sn alloys, whereas the
authors showed an increase in oxygen uptake on a bimetal-
lic nonalloyed Pt–Sn(111) surface compared with Pt(111);
this was attributed to the oxidation of tin atoms. Similarly,
strong hydrogen–tin interactions may enable hydrogen dis-
sociation despite the reduction in the number of free plat-
inum atom ensembles. We cannot exclude either that, in
our experiments, when a prenal molecule approaches the
surface, the oxygen terminal atom interacts with tin sur-
face atoms and that this also contributes to improving the
reactivity of the CO bond.
This study of the influence of a metallic additive on a
stepped surface has enabled us to distinguish steric or struc-
tural effects, prevailing at low coverage, from electronic ef-
fects, which are determining at high coverage.
ACKNOWLEDGMENTS
The authors are indebted to E. Margot for her competence and thank
Professor J. Oudar for fruitful discussions. The financial support of Rhoˆne-
Poulenc is gratefully acknowledged.
REFERENCES
1. Bauer, K., and Garbe, D., in “Common Fragrance and Flavour Mate-
rials,” VCH, New York, 1985.
2. Bauer, K., and Garbe, D., “Ullmann’s Encyclopedia,” 3rd Ed.,
Vol. A11 p. 141. VCH, New York, 1988.
3. Jenck, J., and Germain, G. E., J. Catal. 65, 141 (1980).
4. Noller, H., and Lin, W. M., J. Catal. 85, 25 (1984).
5. Hubault, R., React. Kinet. Catal. Lett. 46, 25 (1992).
6. Vannice, M. A., and Sen, B., J. Catal. 115, 65 (1989).
7. Raab, C. G., and Lercher, J. A., J. Mol. Catal. 75, 71 (1992).
8. Nitta, Y., Ueno, K., and Imanaka, T., Appl. Catal. 56, 9 (1989).
9. Giroir-Fendler, A., Richard, D., and Gallezot, P., Catal. Lett. 5, 175
(1990).
In the latter study (40), Xu et al. explained the decrease in
hydrogenolysis on n-butane for the Sn–Pt(111) alloy com-
pared with Pt(111), by a reduction in equilibrium coverage;
tin-induced selectivity was ascribed to the large reduction
in the heats of adsorption of olefins (41).
10. Raab, C. G., and Lercher, J. A., Catal. Lett. 18, 99 (1993).
11. Galvagno, S., Milone, C., Donato, A., Neri, G., and Pietropaolo, R.,
Catal. Lett. 17, 55 (1993).