Chemistry - An Asian Journal
10.1002/asia.202100504
FULL PAPER
graphite samples by the thermal treatment of the mixture of
platinum chloride and graphite powder samples under chlorine
atmosphere followed by the reduction with hydrogen gas.
However, the locations of platinum nanosheets in graphite
particles were different in artificial and natural graphite samples.
The platinum nanosheets were located on inner portion of artificial
graphite particles while, they were located densely at the outer
portion of natural graphite particles. The different locations of Pt
nanosheets would be due to the difference in surface structure of
two graphite samples. Artificial graphite sample has a lot of
cracks in the surface and platinum chloride would intercalate from
the cracks. Platinum nanosheets were formed by the reduction of
platinum chloride intercalated near the cracks. On the other
hand ,the surface of natural graphite sample is smooth and flat.
Only the edge of layers is the location where platinum chloride
can enter into the layers and platinum nanosheets are formed by
the reduction of metal chloride located near the edge of graphite
layers.
Hydrogenation of cinnamaldehyde over Pt-GIC and Pt/Gmix
Cinnamaldehyde (CAL) is an unsaturated compound and has
three types of unsaturated bonds: aromatic ring, carbon-carbon
double bond (C=C), and aldehyde group (C=O). The CAL
hydrogenation is a structure sensitive reaction. The selectivity to
cinnamyl alcohol (COL), which is produced by the hydrogenation
of C=O bond in CAL molecules, and to 3-phenyl propionaldehyde
(HAL), which is a product by the hydrogenation of C=C bonds,
depends on metal species and the structure of metal particles.
Supported platinum catalysts are active for the CAL
hydrogenation and form hydrogenation products as cinnamyl
alcohol, 3-phenyl propionaldehyde, and 3-phenyl-1-propanol
(HCOL), which is produced by both C=C and C=O bonds[10]. The
catalytic behavior of supported platinum catalysts for the CAL
hydrogenation, especially the difference of selectivity to COL is
closely related to the structure of platinum sites. The CAL
hydrogenation over platinum nanosheets between graphite layers
and particles on graphite surface in a supercritical carbon dioxide
solvent was studied with an expectation that different selectivities
would be observed for platinum nanosheets intercalated between
graphite layers and spherical platinum particles located on
graphite surface. Conversion and selectivity were determined as
follows, conversion (%) = (total amount of products) / ((total
amount of products) + (amount of unreacted cinnamaldehyde)) x
100, selectivity (%) = (quantity of each product) / (total quantity of
each product) x 100.
Figures 8 and 9 show the conversions and product yields of the
CAL hydrogenation in supercritical carbon dioxide with graphite
supported platinum catalyst with the same platinum metal loading
of 5wt%. All the catalysts were active for the hydrogenation and
COL, HAL, and HCOL were formed. However, COL selectivity
depended on the catalyst. The platinum nanosheets intercalated
between graphite catalysts (Pt-SGP5-GIC and Pt-SNO5-GIC)
showed high COL selectivity of 70% or more (Figure 8) regardless
the reaction time and conversion. On the other hand, the COL
selectivity for the platinum supported on graphite surface
(Pt/SGP5-Gmix and Pt/SNO-Gmix) was not so high and was
about 55% and dehydroxylation products such as -methyl
styrene and propyl benzene) were formed (Figure 9) regardless
reaction time and conversion. The higher UOL selectivity over
platinum intercalated (Pt-GIC) samples than that over Pt/Gmix
samples would be explained by the preference adsorption of C=O
bond in CAL molecules on platinum nanosheets compared with
spherical platinum particles. For the system with Pt-GICs
Platinum metal particles on artificial and natural graphite
surface
Figures 7 shows the TEM images of Pt-SGP5-Gmix and Pt-
SNO5-Gmix in which 5wt% of platinum metal particles were
supported on SGP5 or SNO5 by the impregnation of platinum
precursor followed by hydrogen reduction. Spherical platinum
metal particles less than 20 nm were loaded on the artificial (a and
b) and natural (c and d) graphite surfaces. The size of platinum
metal particle on SGP5 were a little smaller than that of SNO5.
On SGP5 surface, a number of platinum particles of less than 10
nm diameter were observed, on the other hand, several larger
particles of more than 10 nm were seen on SNO5 (Figures 7(c)
and (d)). The artificial graphite sample SGP5 has cracks and the
2
-1
surface area value is 16.2 m g . The natural graphite SNO5
2
-1
sample has plain surface and the area value (9.6 m g ) was
smaller than that of SGP5. The larger platinum metal particles
were formed on SNO5 surface because of the larger surface area
and much smooth surface than SGP5 surface.
(
platinum nanosheets between graphite layers), supercritical
carbon dioxide solvent would transport CAL molecules between
the graphite layers. CAL molecules interact with the upper and
lower graphene layers, and the benzene ring exists parallel to the
layers, resulting in the adsorption of end functional aldehyde
group of CAL molecules at the edge of platinum nanosheets. On
the other hand, in the Pt/Gmix (platinum particles on graphite
surface), platinum particles existed on the graphite surface, and
the CAL molecules dissolved in supercritical carbon dioxide are
adsorbed on the platinum particles without any spatial preference,
which led to generate both HAL and COL leading to low COL
selectivity.
Figure 7. TEM images of Pt/SGP5-Gmix (a, b) and Pt/SNO5-Gmix (c, d).
The initial conversion over Pt-SGP5-GIC (50% for 30 min) was
higher than that over Pt-SNO5-GIC (43%). Similar platinum
nanosheets were formed in SGP5 and SNO5 layers; however, the
number of active platinum sites, on which CAL molecules could
adsorb, would be larger in Pt-SGP5-GIC than in Pt-SNO5-GIC
4
This article is protected by copyright. All rights reserved.