374
I. Cano and P. W. N. M. van Leeuwen
(vs. only 5% for Pt@SBA). Decomposition of citral leads to CO, which reduces the
rate of hydrogenation.
FeO x -doped Pt catalysts supported on 15 wt% Al 2 O 3 @SBA-15 composites were
studied by Li and co-workers in the hydrogenation of CAL [67]. The added iron
oxides raised the positive charge on Pt and on nearby Fe sites, which was held responsible for the increase in the selectivity for COL, from 40% for Pt@SBA-15 to 77%,
at almost full conversion (isopropanol/water, 20 bar H 2 , 90 °C). In a related publication, Al 2 O 3 was replaced by 15% TiO 2 that was deposited onto mesoporous silica,
SBA-15 [68]. The best result obtained in this instance with the three-component
catalyst under the same conditions was 86% selectivity to COL at 90% conversion;
at higher conversion somewhat more HCOL was produced and the selectivity for the
desired COL product dropped.
Somorjai et al. studied the influence of thin films (30 nm) of Co 3 O 4 support on
the hydrogenation of crotonaldehyde catalyzed by Pt NPs and compared these with
SiO 2 and TiO 2 [69]. SiO 2 was considered the support with least interaction with Pt
NPs (4.6 nm size), and this catalyst showed the lowest activity and zero selectivity
for crotyl alcohol. Pt@Co 3 O 4 showed the highest activity, TiO 2 being in between,
but the selectivity for the unsaturated alcohol was modest at 20%, butanol 20%, the
remainder 60% is butyraldehyde (gas-phase, 1 Torr of crotonaldehyde, 100 Torr of
hydrogen, 120 °C). Under the reaction conditions, the Pt particles are fully reduced
to the metallic state and the cobalt oxide surface is partially reduced, likely to Co
2+ .
It was thought that atomic hydrogen spills over onto the oxide to carry out the lowtemperature reduction. This reduced surface, it was stated, contains sites for the
adsorption and selective reaction of crotonaldehyde to the alcohol products. Thus,
more in general, reducible oxide supports enhance enol selectivity and velocity.
Li and co-workers also studied Pt NPs in the order of 5 nm size that interact
strongly with Co oxides [70]. Although the support used is amorphous carbon, the
topic fits better in the discussion of Pt/Co x O than under the heading carbon supports
(4.8). First-layered double hydroxides were made (Mg, Al and variable Co content)
containing intercalated chloroplatinic acid which were calcined in the presence of
glucose. Cobalt oxides were observed at the surface of the Pt NPs, which according to
XPS contained 30–50% divalent Pt. Four catalysts were prepared with a Co content
relative to Pt (4.8% wt) of 0, 0.3, 0.6 and 0.9 (Table 11.23). The amount of oxides
in the catalysts was not reported. A remarkable difference between the catalysts was
found in the hydrogenation of CAL (Table 11.23).
The presence of CoO improves the catalyst enormously, both in terms of selectivity
and activity. Catalysts Co 0.3 and Co 0.9 perform about the same, but Co 0.6 really stands
out as it is 10 times more active than Pt only and 99% selectivity to the desired allylic
alcohol product. Aliphatic enals reacted equally well when Co 0.6 Pt/C was used as
the hydrogenation catalyst. The authors assign this result to the higher Pt(0) content
of catalyst Co 0.6 (68% vs. 54% for Co 0 Pt).
Related to the last publications is the work by Xiang, Qin and co-workers [71].
They decorated Pt nanoparticles with Fe oxide by atomic layer deposition in a precise
way. First, Pt was deposited (2.7 nm) from an organometallic precursor and ozone
on Al 2 O 3 (60–100 nm), followed by ferrocene and ozone, both in several cycles.
Précédent

- 381/460

Suivant