Full Papers
different energies associated with these bonding modes result
in the dependence of the amount of CO bonded on the cover-
age (that is, relative pressure of CO).[13b] In addition, the relative
numbers of different bonding sites (that is, plateau, hollow, ter-
race or steps; Figure 5a reveals some of these sites) affects the
amount of CO bonded.[13b] The electronic state of the metal
caused by modifications may also have a direct influence on
the bonding of CO.[13c] This, along with the data obtained, sug-
gest that a stoichiometry factor of 2 for CO is an accurate
value. However, as there are already many in-depth investiga-
tions into the nature of CO chemisorption on palladium cata-
lysts,[13a,b,e,f] this study will focus on the differences in chemi-
sorption of CO (seen as changes in crystallite size) between
the different catalysts as a potential measure on the catalytic
performance of the catalysts with regards to the binding and
reaction of carbonyl compounds.[13g] Interestingly, the pyridine-
and 1-methylimidazole-modified catalysts have an average
crystallite size of around 7 nm (by STEM analysis). This suggests
that these modifiers may have an effect on the size of the Pd
crystallites. H2 chemisorption analysis confirms this for the pyri-
dine-modified catalyst. However, the H2 chemisorption results
for the 1-methylimidazole-modified catalyst reveal a larger
average crystallite size, which is potentially the result of a hin-
drance in H2 adsorption because of the chemically modified
surface, and may be in part responsible for the differences in
the intrinsic activity observed between these two catalysts. A
similar observation can be seen from the CO chemisorption
analysis, in which the ratio of STEM analysis to CO chemisorp-
tion analysis is 1:2.6 and 1:2.9, respectively. If we take the crys-
tallite sizes measured directly (STEM analysis) into account, the
ratio of these and the uncoated catalyst tend to follow the
trend in intrinsic activity of octanal conversion.
not inactivity, with a slight increase in intrinsic activity for 1-
octene conversion.
The [MePsec-Bu3][MeOSO3]-modified catalyst has an average
crystallite size of approximately 10 nm (STEM analysis). This is
similar to that of the uncoated and [MMIM][MeOSO3]-modified
catalysts. As with the [MMIM][MeOSO3]-modified catalyst, the
H2 chemisorption analysis reveals a larger Pd crystallite size
and the intrinsic activity of the 1-octene conversion is lower
than that of the uncoated catalyst. As with the [MMIM]
[MeOSO3]- and [MMIM][NTf2]-modified catalysts, the CO chemi-
sorption analysis reveals a larger Pd crystallite size compared
to the uncoated catalyst, with the ratio of crystallite size to
STEM analysis of 3.6; the catalyst was inactive in the conver-
sion of octanal to octanol. These results show that an interest-
ing pattern emerges between the catalytic results and the
chemisorption analysis, and it may be possible to anticipate
the general catalytic results of ionic-liquid-modified catalysts
by chemisorption techniques together with STEM analysis.
Conclusion
The catalytic results from this study reveal that ionic-liquid-
modified catalysts have dramatic and specific effects and indi-
cate that they have a place alongside traditional organic modi-
fiers. They also have the ability to alter the catalytic per-
formance between different types of substrates selectively. Cat-
alyst characterisation studies reveal that the BET surface area
analysis of these catalysts can give an indication of a possible
loss in the content of ionic liquid but cannot be used as
a means to predict the behaviour of the intrinsic activity or re-
activity of a catalyst.[12] Diffuse reflectance infrared transmit-
tance spectroscopy, inductively coupled plasma optical emis-
sion spectroscopy, and thermogravimetric analysis with differ-
ential scanning calorimetry analyses of these catalysts reveal
that a relatively thick ionic liquid layer can survive the reaction
conditions. However, this is not a universal rule and ionic
liquid decomposition or removal can indeed occur. The de-
composition of the ionic liquid can potentially lead to a very
different matrix of surface modification, which depends on the
ionic liquid used. In addition to this, the in situ preparation of
the catalysts with different organic modifiers can have an
effect on the size of the Pd crystallites. However, chemisorp-
tion techniques coupled with scanning transmission electron
microscopy can reveal the size of the crystallites and potential-
ly give a general method for the anticipation of catalytic re-
sults.
The [MMIM][MeOSO3]-modified catalyst has an average crys-
tallite size of around 10 nm, which is similar to that of the un-
coated catalyst. However, the H2 chemisorption analysis re-
vealed an average particle size of around 14 nm. This suggests
that the H2 molecules used in the chemisorption analysis are
hindered in their chemisorption to the Pd crystallites, possibly
because of the ionic liquid molecular layer or decomposition
products. This, in general, correlates with the catalytic results
of the intrinsic activity and selectivity versus the iso-conversion
of 1-octene between the uncoated and [MMIM][MeOSO3]-
modified catalysts. The crystallite size determined by CO chem-
isorption is approximately 39 nm, and the ratio to STEM analy-
sis is 1:3.9. Interestingly, the catalyst was inactive in the conver-
sion of octanal to octanol.
The [MMIM][NTf2]-modified catalyst has an average crystallite
size of 7 nm by STEM analysis, which confirms the H2 chemi-
sorption analysis. Interestingly, this also correlates with the cat-
alytic intrinsic activity of this catalyst compared to the uncoat-
ed catalyst with regards to 1-octene conversion and crystallite
size. The CO chemisorption analysis reveals a larger crystallite
size compared to the uncoated catalyst. The ratio of STEM
analysis to CO chemisorption analysis is 1:3.6. These data fit
well with both the 1-octene and octanal conversion results, in
which there was a noticeable decrease in the intrinsic activity
of octanal conversion compared to the uncoated catalyst, but
These results highlight that the effects of ionic liquid modifi-
cation of traditional heterogeneous catalysts are many and
varied. In addition to any transport and/or modification effects,
crystallite size modification may also potentially be a factor. As
these varied effects are also dependent on the type of ionic
liquid used and the substrate catalysed, a large-scale investiga-
tion is required to appreciate the trends and modifications of
solid catalysts with an ionic liquid layer fully. This is especially
so if different metals and combinations of them are to be con-
sidered, as ionic-liquid-modified catalysts from this study and
others with different metal catalysts have shown to be effec-
ChemCatChem 2015, 7, 2628 – 2636
2634
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim