Communications
impregnation and calcination (Figures S1 and S2, Supporting
Information). The patterns show the characteristic (100), (110),
and (200) peaks, which can be assigned to the hexagonal
pore structure of SBA-15;[20] this shows that the ordered meso-
porous structure has been maintained during formation of the
metal-oxide phase. The wide-angle XRD patterns of the Ni4/
SBA-15 and Ni/SBA-15 samples after calcination but before re-
duction are presented in Figures S3 and S4. The characteristic
reflexes for NiO at 2q=37.2, 43.1, and 62.88, corresponding to
the (111), (200), and (22 0) planes of cubic NiO, respective-
ly,[4,21] are observed only for the 5-Ni4/SBA sample even
though the actual Ni content is smaller (1.67 wt%) than that,
for example, in the 5-Ni/SBA sample (2.53 wt%). Still, the ob-
served reflexes are very weak and broad, which points to the
formation of very small NiO NPs. For samples with lower load-
ings, hardly any crystal phase line can be recognized, which
might be a sign of the formation of small, highly dispersed Ni
species. Nitrogen physisorption measurements were performed
for the calcined materials (Figure S5). Corresponding isotherms
show the characteristic type IV isotherm with H1-type hystere-
sis loop for mesoporous materials, which again shows that the
mesoporous structure of SBA-15 was maintained after impreg-
nation and calcination. Figures 2 and S6 show the TEM images
of the catalysts prepared from Ni4 cubane and nickel acetate
precursors, respectively. Formation of few isolated metal-oxide
particles can be observed for the lowest concentration in the
Ni4/SBA-15 series, whereas for Ni/SBA-15, metal-oxide NPs can
be spotted only for the highest concentration (5-Ni/SBA-15).
However, only for 5-Ni4/SBA-15 is a larger amount of NiO
nanoparticles seen, and these NPs are highly dispersed on the
silica support, which also supports the results obtained from
the XRD measurements.
Figure 1. Schematic illustration of the Ni4/SBA-15 catalyst prepared by using
a ligand-stabilized Ni4 cubane precursor.
proximity on the support surface (see Figure 1), which are then
able to form very small Ni clusters upon reduction, possibly to
yield an efficient catalyst for the dry reforming of methane. To
investigate a possible beneficial influence of the preorganiza-
tion of Ni atoms, two series of materials were prepared in this
work. For the first series, a mesoporous silica support (SBA-15)
was impregnated with an aqueous solution of the Ni4 cubane
precursor (“A-Ni4/SBA-15”). For comparison, in a second series
a conventional nickel acetate precursor was used in the same
way (“A-Ni/SBA-15”, A=initial wt% of Ni). The concentration of
the precursors in the aqueous solution was varied to investi-
gate the effect of nickel loading on the structure and activity
of the catalysts. All obtained materials were treated likewise;
thus, first dried at room temperature, then calcined at 5508C
for 4 h, and finally subjected to an in situ reducing atmosphere
(pure H2) at 5008C for 1 h before catalytic tests. The actual
metal contents of the samples were determined after the calci-
nation step by inductively coupled plasma (ICP) and are pre-
sented in Table 1. In almost all cases, the actual Ni content is
lower than the nominal one, probably as a result of a washing
All of the prepared materials were applied as cata-
lysts for the DRM reaction. Figures 3 and 4 show
Table 1. Characterization and test results of the Ni4/SBA-15 and Ni/SBA-15 catalysts.
methane conversion over Ni4/SBA-15 and Ni/SBA-15
after an in situ reduction step at temperatures of 500,
525, 550, and 6008C. The theoretical thermodynamic
equilibrium was also calculated by minimizing the
Gibbs free energy of all species in the system (Fig-
ure S7). For the Ni4/SBA-15 materials at 500 and
5258C, the activity is proportional to the amount of
nickel in the catalyst. It can be seen that 5-Ni4/SBA is
an exceptionally good catalyst for the DRM reaction,
as thermodynamic equilibrium is reached at a low
temperature (5008C), and it also shows stable per-
formance at a higher temperature (6008C) although
equilibrium effects should be noted. Moreover, the
[a]
Sample
Ni
Activity[b] [molCH molNiꢀ1 minꢀ1
]
4
[wt%]
5008C
5508C
6008C
1-Ni4/SBA-15
2-Ni4/SBA-15
5-Ni4/SBA-15
1-Ni/SBA-15
2-Ni/SBA-15
5-Ni/SBA-15
0.52
0.81
1.67
0.51
0.97
2.53
3.2
3.5
3.1
3.9
3.0
1.4
6.0
5.5
3.8
3.9
5.0
2.6
13.9
9.7
5.3
–
8.3
3.4
[a] Metal content from ICP measurements. [b] Reaction rate based on mole CH4 con-
verted per mole Ni per time on stream.
step after impregnation. Notably, variations in the actual Ni
content do not allow for straightforward comparison of the
Ni4 and Ni catalysts prepared from the same initial amount of
Ni. Therefore, in a first step the Ni4/SBA-15 and Ni/SBA-15
series are discussed separately by regarding their catalytic per-
formance in the DRM, whereas a comparison between the two
series is made only on the basis of the actual Ni amount evalu-
ated by ICP. Materials in both series were studied by X-ray dif-
fraction (XRD) measurements. The powder XRD pattern at low
angle provides information about the support structure after
other Ni4/SBA-15 catalysts show initially lower performance
and stability; however, they seem to be further activated at
higher temperatures, especially the 1-Ni4/SBA-15 catalyst. It
can, therefore, be concluded that a certain Ni cluster size has
to be reached to show sufficient catalytic performance in the
DRM and that for the catalysts with highly dispersed Ni parti-
cles a certain temperature is needed to form this critical size
by particle sintering. As seen from the TEM pictures, a consider-
able amount of particles is formed in 5-Ni4/SBA-15, which ex-
plains its high catalytic activity at low temperatures. For the
&
ChemCatChem 0000, 00, 0 – 0
2
ꢁ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!