Full Papers
The in situ reduction of the precursor 2.6 wt%[NH4ReO4]-Pd
on P25 and DT51 and prepared by SI gave a Re/Pd atomic
ratio of 0.9 and 1.07, respectively (Table 1, entries 6 and 11). Ex-
posure to air increased this ratio to 1.32 and 1.37, respectively
(Table 1, entries 7 and 12), which suggests some migration of
Re on the surface. A subsequent reactivation in H2 at 4508C
gave ratios of 1.38 and 2.18, respectively (Table 1, entries 8 and
13), which are different from those observed after the direct re-
duction of the precursors.
terface. However, the addition of Re to 1.9 wt%PdKCl/DT51 by
CR decreased the Pd/Ti ratio from 0.038 to 0.027 (Table 1, en-
tries 5 and 9). This implies that on this parent solid, Re might
have been deposited on Pd by CR. However, another factor to
consider for this catalyst is the experimental problem men-
tioned previously using this support.
TPR
The TPR of samples preoxidized under O2 at 3008C was per-
formed up to 7008C. TPR profiles of the series of catalysts pre-
pared on PdCl/P25 are shown in Figure 6, and the profiles for
a series prepared on DT51 are shown in Figure 7. The results
are summarized in Table 2.
Bimetallic Re-Pd catalysts prepared by CR
Some bimetallic catalysts prepared by the CR method were an-
alyzed similarly after storage in air and/or after reduction in
the XPS cell at 4508C (Table 1).
The “as-prepared” 1.7 wt%Re-PdCl/P25-CR catalyst contained
Re as a mixture of Re4+, Re6+, and Re7+; the highest oxidation
state species Re7+ was the most abundant (63%, Table 1,
entry 19). Conversely, Pd was 66% in the zero-valent state, and
the remainder was Pd2+ (18%) and Pd4+ (16%). Compared to
the “as-prepared” Pd monometallic parent that contained only
34% Pd0, the presence of Re in interaction with Pd seems to
limit the reoxidation of Pd. However, as in the SI-prepared cat-
alysts, the shift of the BEs for Pd0 and Pd2+ to lower energies
on one part and of that of Re4+ and Re6+ to higher energies on
the other part compared to the “as-prepared” 2 wt%PdCl/P25
and 1.9 wt%RepH1/P25 was small (0.2–0.3 eV), within the experi-
mental errors. Therefore, there is no evidence of an electronic
transfer from Re to Pd in this catalyst. Although the other “as-
prepared” catalysts were not characterized, one may expect
the same results.
XPS data showed that after O2 treatment at 3008C, Re was
totally oxidized to Re7+, for instance in 0.9 wt%Re-PdCl/P25-CR,
3.6 wt%Re-PdCl/P25-CR, and 2.7 wt%Re-PdKCl/P25-SI (not
shown). Pd was oxidized to Pd2+ (62–80%) and Pd4+.
During the reduction of the oxidized 2 wt%PdCl/P25 solid
only one H2-consumption zone was clearly observed at room
temperature (Figure 6a), which was ascribed to the reduction
of oxidized Pd (nH =2.710À4 molgcatÀ1; Table 2).[20] In the case
2
of 1.9 wt%RepH1/P25, a wide consumption peak was exhibited
between 230 and 3408C, with a broad shoulder up to 4508C
(Figure 6b). According to the literature,[25,29] H2 uptake in this
temperature range is attributed to the reduction of ReOx oxi-
dized Re species and the reduction of the bulk titania support.
Quantitative analysis reveals that the measured H2 consump-
tion (2.310À4 molgcatÀ1) was lower than the theoretical con-
sumption calculated for the complete reduction of the Re7+
present in the calcined solid to Re0 (3.610À4 molgcatÀ1) and
confirmed that Re could not be reduced totally to the metal
state, as shown previously by the XPS results. The reduction of
xwt%Re-PdCl/P25-CR bimetallic catalysts displayed a large re-
duction peak at a low temperature[20] (Figure 6c and d). The
total amounts of H2 consumed were higher than that exhibited
by the Pd oxide in the monometallic catalyst, which suggests
that contact between Re and Pd was established (Table 2).
There is clearly a strong effect of Pd on the reduction tempera-
ture of Re7+. The amount of H2 consumed increased with the
addition of Re (0.9 to 3.6 wt%). Furthermore, a comparison of
3.6 wt%Re-PdCl/P25-CR and 3.5 wt%Re-PdCl/P25-SI prepared by
CR and SI, respectively, which contain very similar Re contents,
evidenced a lower H2 consumption at low temperature for the
SI-prepared solid, which nevertheless remained higher than
the consumption for the reduction of the Pd monometallic cat-
alysts (Table 2). This means that a smaller fraction of Re is in
contact with Pd in the SI-prepared catalyst. Accordingly,
a second reduction peak was detected at higher temperatures
(250–3008C) for the bimetallic SI catalysts (Figure 6e), which
could be attributed to the reduction of Re clusters isolated on
the TiO2 support.
Following in situ reduction in the XPS cell, Pd was in the
zero-valent state (>92%) whatever the nature of the support
and the precursor salt, whereas the Re species were affected
significantly by the method of preparation. If PdCl2 was used
as the precursor, most of the Re was present in the Re3+ state;
the higher the amount of Re, the higher the proportion of
Re3+ (Table 1, entries 18 and 20). However, if we start from
K2PdCl4, the nature of the support influenced the final reduc-
tion state of Re: 74% Re0 and 26%Re3+, and 30% Re0 and 70%
Re3+ for 1.5 wt%Re-PdKCl/P25-CR (entry 15) and 1.9 wt%Re-
PdKCl/DT51-CR (entry 9), respectively. Similar to that of the cata-
lysts prepared by SI (except for 1.5 wt%Re-PdKCl/P25-CR), the
zero-valent state of Re was absent or represented only a minor
fraction and the remainder is Re3+. This probably means that
under reaction conditions in the aqueous phase under H2 pres-
sure at 1608C most of the Re is in an oxidized state.
The surface Re/Pd, Re/Ti, and Pd/Ti compositions were calcu-
lated from the XPS analysis (Table 1). In the PdCl/P25 series that
contains Re deposited by CR, the determined atomic ratios of
Re/Ti and Re/Pd increased as the loading of Re increased from
0.9 to 1.7 wt% (Table 1, entries 18 and 20), which shows that
the Re loading increased at the surface. However, regardless of
the treatment and the loading of Re, the Pd/Ti ratio did not
change and was close to 0.036, which is the value in the
parent catalyst (Table 1, entry 17). This indicates that little Re is
located on the Pd, and Re was deposited at the Pd–support in-
The reduction of 2.7 wt%Re-PdKCl/P25-SI (not shown) also
gave a large reduction peak that was complete at around
408C (nH =4.510À4 molgcatÀ1). In addition, a weak consump-
2
À1
tion of 0.610À4 molgcat around 3258C was found for this
ChemCatChem 2015, 7, 2161 – 2178
2167
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim