Influence of periodic nitrogen functionality on the selective oxidation of alcohols

Article abstract of DOI:10.1002/asia.201100565

An enhancement in catalytic alcohol oxidation activity is attributed to the presence of nitrogen heteroatoms on the external surface of a support material. The same Pd particles (3.1-3.2 nm) were supported on polymeric carbon-nitrogen supports and used as

Full text of DOI:10.1002/asia.201100565

DOI: 10.1002/asia.201100565
Influence of Periodic Nitrogen Functionality on the Selective Oxidation of
Alcohols
Carine E. Chan-Thaw,[a] Alberto Villa,[a] Gabriel M. Veith,[b] Kamalakannan Kailasam,[c]
Leslie A. Adamczyk,[b] Raymond R. Unocic,[b] Laura Prati,*[a] and Arne Thomas[c]
Abstract: An enhancement in catalytic
alcohol oxidation activity is attributed
to the presence of nitrogen heteroa-
toms on the external surface of a sup-
port material. The same Pd particles
(3.1–3.2 nm) were supported on poly-
meric carbon–nitrogen supports and
used as catalysts to selectively oxidize
benzyl alcohol. The polymeric carbon–
nitrogen materials include covalent tri-
azine frameworks (CTF) and carbon
nitride (C3N4) materials with nitrogen
content varying from 9 to 58 atomic
percent. With comparable metal expo-
sure, estimated by X-ray photoelectron
spectroscopy, the activity of these cata-
lysts correlates with the concentration
of nitrogen species on the surface. Be-
cause the catalysts showed comparable
acidic/basic properties, this enhance-
ment cannot be ascribed to the Lewis
basicity but most probably to the
nature of N-containing groups that
govern the adsorption sites of the Pd
nanoparticles.
Keywords: alcohols · nanoparticles ·
nitrogen functionalities · oxidation ·
palladium
Introduction
For example, the introduction of N functionalities onto the
external surfaces of carbon nanotubes resulted in a 20% en-
hancement in the catalytic activity of Pd-supported nanopar-
ticles (Pd NPs) relative to an unmodified surface.[4] In addi-
tion, covalent triazine framework (CTF) materials derived
from aromatic nitriles with a high concentration of nitrogen
functionalities[5–10] were also demonstrated to be a promising
support for the liquid-phase oxidation of glycerol and
benzyl alcohol.[11,12] Palladium-supported CTFs are indeed
more active than activated carbon-based catalysts that do
not contain any nitrogen functionalities (time of flight
(TOF) of 4917 hÀ1 versus 2678 hÀ1 for Pd/CTF and Pd/AC,
respectively, in benzyl alcohol oxidation).[12] The improved
catalytic performance is attributed to a general positive
effect of N species in the support. In this paper, we specifi-
cally ascribe the enhanced catalytic activity to the presence
of N-containing groups and to their ability to differently
adsorb the Pd NPs. Benzyl alcohol oxidation was used as a
test reaction for three different Pd catalysts that differed in
the number of nitrogen atoms and the chemical nature of
the N functionalities borne by the support. CTF supports
were obtained by a dynamic reorganization of porous poly-
mer networks by using either para-dicyanobenzene
(CTFDCB) or 2,6-dicyanopyridine (CTFDCP) as the mono-
mer.[5,7] The third support was a carbon nitride prepared by
the thermal polymerization of cyanamide in which the
amount of nitrogen incorporated was the highest stoichio-
metrically possible (58%).[13]
Selective oxidation of alcohols with oxygen by using sup-
ported metal nanoparticles (NPs) represents a well-studied
reaction because of its wide application in chemical transfor-
mations. However, up to now, scaling up a reaction that
occurs in a condensed liquid phase has been rather limited
and low reaction rates are usually obtained. The low rates
are attributed to 1) low reaction temperatures, 2) the use of
solvent(s), 3) diffusion limitations, or 4) catalyst deactiva-
tion. Catalyst deactivation is mainly due to metal leaching,
sintering, overoxidation, or irreversible adsorptions.[1] The
activity/selectivity of the catalyst is a function of several var-
iables; among those variables, the role of the support has
not yet been fully clarified.[2,3] A direct comparison of the
various catalytic performances reported in the literature is
difficult due to the different reaction conditions and materi-
als.
The introduction of heteroatoms into supports can be
beneficial for the catalytic activity of metal nanoparticles.
[a] Dr. C. E. Chan-Thaw, Dr. A. Villa, Prof. L. Prati
Department of Inorganic Chemistry L. Malatesta
Universitꢀ degli Studi di Milano
via Venezian 21, 20133 Milano (Italy)
Fax : (+39)02503-14405
E-mail: Laura.Prati@unimi.it
[b] Dr. G. M. Veith, Dr. L. A. Adamczyk, Dr. R. R. Unocic
Materials Science and Technology Division
Oak Ridge National Laboratory
Oak Ridge, TN 37831 (USA)
Results and Discussion
[c] Dr. K. Kailasam, Prof. A. Thomas
Department of Chemistry, Functional Materials
Technische Universitꢁt Berlin
Elemental analysis and surface area data for the support
materials are summarized in Table 1.
Englische Strasse 20, 10587 Berlin (Germany)
Chem. Asian J. 2012, 7, 387 – 393
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
387

FULL PAPERS
Table 1. Characterization of the different CTF and C3N4 supports by nitrogen sorption
measurements, combustion elemental analysis, and XPS.
the native support material. Analysis of the C1s
XPS data for the CTFDCB and CTFDCP support ma-
terials showed the presence of a high concentration
Monomer Monomer/ Surface area
Pore volume XPS Elemental
ZnCl2
[m2 gÀ1
]
[cm3 gÀ1
]
analysis
[wt%]
À
À
of C C and C H species (ꢀ284.8 eV) and smaller
À
À
concentrations of C O (ꢀ286.5 eV); C=O and C
%
C
N
H
N[a]
À
N ( 288.3 eV); and C=C species (ꢀ290.8 eV). Sim-
ilar analysis of the C1s data obtained for C3N4
CTFDCB DCB
1:5
1:5
2814
10
1738
1.79
8
55
14
73
33 58
64 18
9
1
2
1
[b]
À
showed a high concentration of C N moieties
C3N4
cyanamide
CTFDCP DCP
1.05
À
(288.0 eV) with a much smaller concentration of C
À
C and C H species (ꢀ4% at 284.8 eV). N1s data
[a] Atomic% based on C/N/H. [b] Under detection limit.
collected for the support materials showed the pres-
ence of C2NH (ꢀ398.3 eV), C3N (ꢀ400.1 eV), and
Figure 1 reports the pore-size distribution of the two CTF
supports estimated by the nonlocal density functional theory
(NLDFT) method. The samples have both micro- and meso-
NH (402 eV) species on the surface of the support materials.
Analysis of the relative fraction of C3N and C2NH species
showed an increase in C2NH species on the surface of the
C3N4 material, which is consistent with a higher concentra-
tion of the species based on the molecular structure with
cross-linked tri-s-triazine units.
Comparison of carbon and nitrogen elemental analysis
data with surface concentrations estimated by XPS were in
excellent agreement (less than 6% difference) with the ex-
ception of the surface carbon content measured for CTFDCP
,
which showed an increase in the concentration of carbon.
Elemental analysis clearly shows an increase in N concentra-
tion that is consistent with the addition of pyridine to the cy-
anobenzene building block for the CTFDCB and CTFDCP sup-
port materials (9 and 18%, respectively). The N content in
the C3N4 was determined to be 58 atomic percent (Table 1).
The N content determined by XPS in the case of C3N4 is
consistent with theoretically predicted N concentrations in
this compound (C/N 42:57%); however, bulk N/C ratios
were lower due to total oxygen content. Analysis of O1s
XPS data showed that the surface of the C3N4 contained a
Figure 1. NLDFT pore-size distribution for the samples CTFDCB and
CTFDCP
.
pores with pore diameters less than 5.5 nm for CTFDCB and
4 nm for CTFDCP as shown in earlier studies.[6,8] The surface
area of CTFDCB is higher (2814 m2 gÀ1) than that of CTFDCP
(1738 m2 gÀ1). C3N4 has a very low surface area (10 m2 gÀ1)
with nontextured pores. In this latter case, we expected that
the lack of porosity would confine the Pd particles to the ex-
ternal surface of the support and avoid diffusion limitations.
The N1s and C1s X-ray photoelectron spectroscopy
(XPS) data are shown in Figures 2 and 3, respectively, for
À
small concentration of O (1.5 at%, primarily in C O moiet-
ies at 532 eV).[14] The CTFDCB and CTFDCP supports had
much higher concentrations of surface oxygen (11 and 8%,
À
respectively) primarily due to C O (532 eV) and C=O spe-
cies (530.7 eV) for the CTFDCP
support. However, studies on
unwashed, freshly prepared
CTFDCB (i.e., which still con-
tained Zn2+ catalyst) had low
oxygen content (<1.2%), thus
indicating that the addition of
oxygen to these support materi-
als occurs during storage in air
or the removal of zinc.
TEM data collected for all
the as-prepared Pd-loaded sup-
port materials (Figure 4a (Pd/
CTFDCP) and Figure 4b (Pd/
C3N4)) reveal the homogeneous
distribution of Pd particles on
the support surfaces with the
same average particle sizes and
deviations (3.09–3.18 nm). This
particle size is slightly larger
Figure 2. XPS measurements on the three catalysts (N1s XPS data).
388
www.chemasianj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2012, 7, 387 – 393

Periodic Nitrogen Functionality
35%) and Pd0 on each sample
(Figure 5, Table 2). On Pd/C3N4
and Pd/CTFDCP, PdO was also
detected in the O1s spectra.
These species are likely gener-
ated by exposure of the metal
to air and often occur in Pd cat-
alysts that progressively deacti-
vate (O2 poisoning effect).
However, in the catalytic tests
we did not observe any deacti-
vation.
Despite the similar Pd parti-
cle sizes and oxidation states,
significant differences in cata-
lytic activity were observed for
the oxidation of benzyl alcohol
over
the
three
catalysts
Figure 3. XPS measurements on the three catalysts (C1s XPS data).
(Table 4). Pd/C3N4 is more
Table 3. Median size of Pd nanoparticles immobilized on different sup-
ports before reaction.
Statistical median [nm]
Standard deviation [nm]
Pd NPs
Pd/CTFDCB
Pd/C3N4
2.81
3.18
3.16
3.09
0.5
0.6
1.1
0.9
Pd/CTFDCP
active than Pd/CTFDCP, which is more active than Pd/
CTFDCB (converted mol (molPd)À1 hÀ1 =14131, 8025, and
4917 hÀ1, respectively). Blank experiments using bare sup-
ports showed total inactivity, thereby proving the compulso-
ry need of Pd for catalysis. It should be noted that C3N4 has
been recently reported to be an active metal-free catalyst
under UV-assisted conditions for benzyl alcohol oxida-
tion;[15] however, since we are not using a UV source, this
should not be an issue.
It is known that microporosity could limit the activity of
metal NPs in liquid-phase oxidation.[16,17] Therefore a com-
parison of the catalyst activities was carried out by using
XPS and surface-area data, which provided a direct mea-
surement of exposed active metal-site accessibility. This ena-
Figure 4. TEM image of a) Pd/CTFDCP and b) Pd/C3N4.
than the size of the unsupported Pd particles (2.81 nm),
which indicates a slight coarsening during the immobiliza-
tion on the support material (Table 3). The large concentra-
tion of Pd NPs on C3N4 is because of its low surface area
and same total Pd concentration versus the other catalysts.
Due to the low surface area and variation in support mor-
phology, which will affect the fraction of material sampled
in the TEM, there is an apparent higher areal concentration
of Pd nanoparticles. Pd XPS data collected on the catalysts
revealed the presence of similar concentrations of Pd2+ (24–
Table 2. Atomic % and Pd oxidation state data as determined by XPS.
Sample
Atomic %
Binding energy [eV] (fraction Pd [%])
Pd0
Pd2+
À
Pd O
CTFDCB
Pd/CTFDCB
C 72.1, N 16.5, O 11.4
C 63.8, N 7.0, O 26.2
Pd 3.0
C 77.8, N 13.8, O 8.4
C 41.1, N 29.6, O 22.2
Pd 7.1
–
–
–
–
–
–
335.25
(64.8)
–
335.56
(76.4)
337.37
ACHTUNGTRENNUNG(35.2)
–
337.90
(23.6)
G
CTFDCP
Pd/CTFDCP
537.90
(7.3)
G
R
ACHTUNGTRENNUNG
C3N4
Pd/C3N4
C 41.6, N 56.9, O 1.5
C 38.8, N 27.8, O 26.3
Pd 7.1
–
–
–
537.80
(7.9)
335.18
(65)
337.20
(35)
AHCTUNGTRENNUNG
Chem. Asian J. 2012, 7, 387 – 393
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemasianj.org
389

L. Prati et al.
FULL PAPERS
XPS elemental analysis data reveals that the nanometers
at the top of the C3N4 and CTFDCP support material contain
the same concentration of Pd (7.1%; Table 2). By consider-
ing the similar Pd particle sizes, weight loading, and non-
porosity of C3N4, we can estimate that all of the Pd of the
CTFDCP support materials are located on the external sur-
face in an eggshell-like configuration. This location of Pd on
the external surface of the CTFDCP material is consistent
with the small pore size of the CTFDCP material estimated
by N2 physisorption (Figure 1), which would be too small for
the 3.1–3.2 nm Pd particles to condense into, particularly if
one considers the remaining polyvinyl alcohol (PVA) on the
Pd particles, evidenced by the XPS data (described below),
which probably enlarged their effective diameters. On the
contrary, the Pd concentration determined on CTFDCB
showed a much lower Pd content (3%) than the CTFDCP
-
and C3N4-based materials (Table 2). This reduction in Pd
trapped on the top nanometer of support material is proba-
bly due to the larger pore diameters (Figure 1) that are
large enough to trap 3 nm-sized Pd particles from the exter-
nal surface of the support and therefore not sampled by
XPS.
Therefore, a realistic comparison of the catalytic activity
could be made only between Pd on C3N4 and CTFDCP, which
presented the same Pd particle size and the same Pd expo-
sure. Due to the similar concentration and the similar size
of Pd particles on the external surface of the support materi-
al, we can assume that all the catalytic reactions in the
liquid phase of the Pd/CTFDCP and Pd/C3N4 catalysts are oc-
curring on the external surface of the support materials. Fur-
thermore, we can assume that the large micropore volume
of the Pd/CTFDCP catalyst will be inconsequential toward
the diffusion of reactants under reaction conditions. This is
due to the large difference in reactant volume (30 mL)
versus pore volume of the catalyst (1.05 mLgÀ1, 0.0985 g cat-
alyst) used in the reaction, which means most of the reac-
tants are flowing around the outside of the support where
the Pd clusters are located. Therefore the difference in sur-
face area between the two supports did not play a significant
role. However, the large difference in activity between Pd/
CTFDCP and Pd/C3N4 indicates that other factors are influ-
encing catalytic activity. To exclude factors related
Figure 5. Pd XPS data collected on the catalysts; concentrations of Pd2+
(24–35%) and Pd0 on each sample
Table 4. Oxidation of benzyl alcohol by using supported Pd catalysts.
to Pd exposure mediated by residual PVA, we per-
Catalyst[a] Converted mol
(molPd)À1 hÀ1
Selectivity[b]
formed TGA measurements. PVA decomposes in a
temperature range between 300 and 3508C. In the
same range, Pd/CTFDCP and Pd/C3N4 lose 0.15 and
0.18 wt%, respectively, thus indicating that a com-
parable and rather low amount of PVA (about
10 wt% of the total employed) is still present on
the catalysts. Thus the possible shielding of the Pd
particles by residual PVA should be similar in both
cases. The TGA, however, did not provide any in-
formation about the location of the PVA. As the
catalytic tests are performed in the presence of a
solvent, which can play an important role in reac-
Toluene Benzaldehyde Benzoic
acid
Benzyl
benzoate
Unknown
Pd/
4917
31
66
1
1
1
CTFDCB
Pd/C3N4
Pd/
14131
8025
35
26
62
71
0.5
1.5
2
1
0.5
0.5
CTFDCP
[a] Reaction conditions. Alcohol/metal: 5000, T=808C, pO2 =2 atm, benzyl alcohol/cyclohex-
ane ratio 50:50. [b] Selectivity calculated at 90% conversion as mol of products (mol convert-
ed)À1 ꢃ100.
bled us to identify trends on materials with different surface
areas and structures.
tion kinetics, we evaluated the active surface of Pd by chem-
isorption. The samples were treated with H2, purged with
390
www.chemasianj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2012, 7, 387 – 393

Periodic Nitrogen Functionality
He, and then dosed with O2. O2 was chemisorbed in a simi-
lar range (0.45–0.43 mol O2 per mol Pd) in both cases. These
experiments definitely demonstrated that Pd NPs on
CTFDCP and C3N4 presented a similar Pd exposure, thus con-
firming the findings of XPS measurements above.
lectivity of all the catalysts toward toluene. Therefore, be-
cause the tests were carried out in the same solvent, we
were able to conclude that in the present case a negligible
influence of the support is present in determining the prefer-
ential reaction pathway.
Based on the logic described above, the result strongly in-
dicates that nitrogen surface chemistry of the catalytic sup-
port mediates catalytic activity. In fact, the increase in cata-
lytic activity closely parallels the increase in N content mea-
sured by XPS and elemental analysis (Figure 6). This trend
is further confirmed by comparison of the activity of similar-
ly sized Pd particles (4.0 nm) supported on nitrogen-free ac-
tivated carbon (converted mol (mol Pd)À1 hÀ1 =2678 hÀ1).[12]
Moreover, O1s and C1s XPS data collected for the sam-
À
ples show an increase in C O species due to the alcohol on
the PVA (Table 5). The surface PVA reduces the surface
concentrations of C and N significantly due to blocking of
these species. Indeed, after deposition of Pd–PVA NPs, we
detected by XPS almost the same C/N/O distribution
(41:30:22 for Pd/CTFDCP and 39:28:26 for Pd/C3N4; Table 2).
Furthermore, after the deposition of Pd NPs, the generated
pH decreased (4.27 for CTFDCP and 5.35 for C3N4). There-
fore the huge difference in the catalytic activity between Pd/
CTFDCP and Pd/C3N4 cannot merely be ascribed to the con-
siderable amount of nitrogen heteroatoms, but most proba-
bly also to the nature of N-containing groups.
An indication of the impact of nitrogen-containing groups
has been revealed by studying the changes in the nitrogen
functionality of the catalyst compared to the original sup-
port material by means of XPS (Table 5). For both CTFDCP
and C3N4 support materials, there is a decrease in C3N func-
tionality with the addition of Pd. This strongly indicates that
the Pd associates with these sites. This point can be also sup-
ported by looking at XPS results for the less active catalyst,
Figure 6. Activity of the catalyst as a function of the nitrogen concentra-
tion.[12]
À
CTFDCB. In that case, there is an increase in N H and C2NH
functionality, thus indicating that during the liquid-phase
Table 5. Summary of binding energies [eV] and moiety fraction [%] data as deter-
mined by XPS.
Recent theoretical calculations by Zope et al.
have shown that the support can play an active role
in liquid-phase alcohol oxidation to improve the
Sample
C 1s
À
C N,
C+
À
C C,
À
C O
C=C
p–p
À
formation of alcoholate by means of O H scis-
À
C H
C=O
C=N
sion.[18] Thus, a possible explanation of the correla-
tion between the N content and the activity could
involve the basic properties of N-containing groups.
However, the two catalysts showed only a slight dif-
ference in the generated pH (5.45 for CTFDCP and
6.0 for C3N4), thus highlighting a very low basicity.
This finding was confirmed by a negligible amount
of adsorbed CO2 revealed by attenuated total re-
flectance (ATR)-IR measurements. The low differ-
ence in the basicity of the two catalysts could also
explain the negligible difference in terms of selec-
tivity to benzaldehyde found (Table 4). In fact, it
should be expected that an increase in basicity
would increase the overoxidation to benzoic acid.
As shown in Table 4, in all cases the main by-
product of the reaction was toluene. Toluene has
been shown to derive from a disproportionation
pathway beside the oxidation pathway that produ-
ces benzaldehyde.[19,20] It has also been reported
that the support and the reaction media can play an
important role in the ratio of products between the
two different pathways. However, in the present
cases we observed very small differences in the se-
CTFDCB [eV]
[%]
Pd/CTFDCB [eV]
[%]
CTFDCP
[%]
Pd/CTFDCP [eV]
[%]
C3N4 [eV]
[%]
Pd/C3N4 [eV]
[%]
284.78
66.0
284.89
60.1
284.66
61.3
284.78
26.7
284.80
4.3
284.84
33.7
286.47
19.7
286.56
24.6
286.18
19.1
286.95
36.3
–
288.47
8.3
288.52
10.0
288.10
12.0
288.40
37.0
288.00
93.0
288.00
64.8
291.22
6.0
291.18
5.3
290.65
5.6
–
–
–
–
–
–
–
292.94
2.0
–
–
293.36
2.7
293.36
1.5
–
–
–
–
–
–
Sample
N1s
O1s
C O
À
C2NH
C3N
NH
NOx
C=O
O=N
CTFDCB [eV]
[%]
Pd/CTFDCB [eV]
[%]
CTFDCP
[%]
Pd/CTFDCP [eV]
[%]
C3N4 [eV]
[%]
Pd/C3N4 [eV]
[%]
398.41
44.4
398.50
27.9
398.17
44.3
398.42
88.4
398.58
78
398.69
92.5
400.16
49.7
400.15
56.2
399.99
41.5
400.92
9.9
400.61
18.4
400.78
6.4
402.71
5.9
402.71
15.9
401.82
10.3
–
–
–
–
–
–
–
–
–
532.0
100
532.43
88.7
532.33
54.0
532.60
58.5
531.45
100
535.51
11.3
535.85
6.5
535.22
10.3
–
–
530.51
39.5
530.99
23.9
–
404.71
3.9
404.49
1.7
404.11
3.6
404.60
1.1
–
–
–
–
–
–
531.8
67.0
533.80
25.1
Chem. Asian J. 2012, 7, 387 – 393
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemasianj.org
391

L. Prati et al.
FULL PAPERS
uptake was followed by a mass flow controller connected to a PC
through an A/D board. The oxidation experiments were carried out in
the presence of a solvent (0.0463 mol substrate, substrate/metal=5000
(mol/mol), benzyl alcohol/cyclohexane: 50:50 (vol%), 808C, pO2 =
2 atm). Periodically samples were removed from the reactor for analysis.
synthesis some of the C3N sites are hydrolyzed to make
active sites with which the Pd can associate. In addition, by
comparing C/N/O distribution before and after Pd deposi-
tion, we noted that in the case of CTFDCP there was an in-
crease in N (from 14 to 30%), whereas in the case of C3N4
there was a decrease (from 57 to 28%). This could mean
that in the case of C3N4, the N sites bound Pd NPs, thereby
preventing the migration of Pd over the C3N sites. Thus an
important difference between Pd/CTFDCP and Pd/C3N4 is the
adsorption site of the Pd NPs that could heavily modify the
electronic structure of Pd NPs.
Nitrogen sorption analyses were carried out with an Autosorb-1 instru-
ment after evacuating the samples at 1508C overnight. The surface areas
were determined by applying the Brunauer–Emmett–Teller (BET)
method in the relative pressure range of 0.05–0.25. The pore volume was
calculated at a relative pressure of 0.99. Nonlocal density functional
theory pore-size distribution was determined by using the carbon/slit–cy-
lindrical pore model.
X-ray photoelectron spectroscopy (XPS) data were collected with a PHI
3056 spectrometer with an Al anode source operated at 15 kV and an ap-
plied power of 350 W. Samples were manually pressed between two
pieces of indium foil; the piece of In foil with the sample on it was then
mounted onto the sample holder with a piece of carbon tape (Nisshin
E.M. Co., Ltd.). Adventitious carbon was used to calibrate the binding
energy shifts of the sample (C1s=284.8 eV).[26] High-resolution data was
collected at a pass energy of 5.85 eV with 0.05 eV step sizes and a mini-
mum of 200 scans to improve the signal-to-noise ratio; lower-resolution
survey scans were collected at a pass energy of 93.5 eV with 0.5 eV step
sizes and a minimum of 25 scans. Transmission electron microscopy
(TEM) data were collected on the Pd/C3N4 catalysts with a Hitachi
HF3300 TEM/STEM equipped with a cold field-emission gun and oper-
ated at 300 kV. TEM images were collected for Pd/CTFDCB and Pd/
CTFDCP with a FEI Tecnai G2 20 S-TWIN CM200 LaB6 electron micro-
scope operating at 200 kV. The powder samples of the catalysts were di-
rectly mounted onto copper grids covered with a carbon film.
Conclusion
We have demonstrated the effect of nitrogen heteroatoms
on the catalytic liquid-phase oxidation of alcohols. As a base
level catalytic activity, we took one of the Pd particles
grown on carbon supports without nitrogen heteroatoms.
Similarly sized Pd particles on carbon-based supports in
which nitrogen heteroatoms are incorporated show an in-
crease in activity (up to converted mol (molPd)À1 hÀ1
=
14131 hÀ1 with 57% nitrogen). This increase in activity
varies linearly with nitrogen content of the support. The ac-
cepted mechanism of the reaction could correlate the im-
proved catalytic activity to the enhanced basicity of the sup-
port. However, the basic properties of Pd on CTFDCP and
C3N4 were comparable. A determining factor appeared to
be the Pd coordination site: the most active catalyst seemed
to be the one in which a more relevant coordination of Pd
on N groups was revealed. This in turn led to a linear corre-
spondence of the increased catalytic activity with the
amount of N groups. Such a correlation with nitrogen and
activity introduces a new way to tune the activity of support-
ed metal particles through the introduction of molecules
that contain nitrogen functionality.
Thermogravimetric analyses (TGA) were performed with
a Perkin–
Elmer TGA7 using a heating rate of 108CminÀ1 from 50 to 5008C.
Adsorption of CO2 was followed in situ in the attenuated total reflection
mode (ATR-IR) with
a Vector 22 spectrometer (Bruker Optics)
equipped with a commercial ATR mirror unit and a liquid-nitrogen-
cooled MCT detector. A thin film was produced on a ZnSe internal re-
flection element (IRE, 458, 50ꢃ20ꢃ2 mm; Crystran) by drying in air an
aqueous suspension that contained 10 mg of the sample. After mounting
the sample in a homemade cell, pure CO2 (50 mLminÀ1) was allowed to
flow through the cell while ATR-IR spectra were measured by co-adding
200 scans at 4 cmÀ1 resolution for 1 h. Then the cell was purged with Ar
and desorption was followed for 30 min.
Adsorption of O2 was performed with a Thermo Electron Corporation
TPD/R/O 1100 catalytic surface analyzer. The sample was treated with a
hydrogen flow at 1008C for 1 h, then H2 was removed by an Ar flow at
1008C. The sample was cooled at room temperature. Oxygen chemisorp-
tions were performed by pulsing O2 (50 mLminÀ1) at room temperature.
Experimental Section
CTFs were synthesized as reported elsewhere.[11] Carbon nitride (C3N4)
powder was prepared by the thermal polymerization of cyanamide (99%,
Aldrich) at 5508C in an alumina crucible for 4 h in air according to the
procedure reported elsewhere.[13,21–23] After preparation, the C3N4 was
yellow and showed a graphite-like structure (2q=13.3 and 27.58) with an
interplanar spacing of approximately 0.33 nm. The average grain size was
estimated to be 7 nm (Debye–Sherrer equation).[24] Palladium nanoparti-
Acknowledgements
Financial support by the UniCat cluster of excellence (Unifying Concepts
in Catalysis, Berlin) and Fondazione Cariplo are gratefully acknowl-
edged. A portion of the research was supported by Oak Ridge National
Laboratoryꢄs SHaRE User Facility (TEM), which is sponsored by the Sci-
entific User Facilities Division, Office of Basic Energy Sciences, U.S. De-
partment of Energy, and by the U.S. Department of Energy, Office of
Basic Energy Sciences, Materials Sciences and Engineering Division
(GMV, RRU), and the FIRST Energy Frontier Research Center (LAA)
under contract with UT-Battelle, LLC.
À
cles were deposited onto the C N polymeric supports through the sol im-
mobilization technique generated by NaBH4 reduction of Na
N
in the presence of a protective agent (polyvinyl alcohol, PVA).[25] The
three catalysts prepared in this study are labeled Pd/CTFDCB, Pd/CTFDCP
,
and Pd/C3N4. The total Pd loading was 1 wt%. Inductively coupled
plasma (ICP) analyses were performed with a Jobin Yvon JV24 instru-
ment to verify the quantitative metal loading on the support by checking
the residual metal in the filtrate. Carbon, hydrogen, and nitrogen analysis
for CTF support was performed with a Vario Micro setup. Carbon, hy-
drogen, and nitrogen analysis for C3N4 was performed by Gailbrath Lab-
oratories, Knoxville, Tennessee, USA.
[1] T. Mallat, A. Baiker, Catal. Today 1994, 19, 247–284.
[2] W. Fang, J. Chen, Q. Zhang, W. Deng, Y. Wang, Chem. Eur. J. 2010,
DOI: 10.1002/chem.201002469.
[3] T. Ishida, M. Nagaoka, T. Akita, M. Haruta, Chem. Eur. J. 2008, 14,
8456–8460.
The reactions were carried out in a thermostatted glass reactor (30 mL)
provided with an electronically controlled magnetic stirrer connected to
a large reservoir (5000 mL) that contained oxygen at 2 atm. The oxygen
392
www.chemasianj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2012, 7, 387 – 393

Periodic Nitrogen Functionality
[4] A. Villa, D. Wang, P. Spontoni, R. Arrigo, D. Su, L. Prati, Catal.
Today 2010, 157, 89–93.
[16] C. Bianchi, F. Porta, L. Prati, M. Rossi, Top. Catal. 2000, 13, 231–
236.
[5] P. Kuhn, M. Antonietti, A. Thomas Angew. Chem. 2008, 120, 3499–
3502; Angew. Chem. Int. Ed. 2008, 47, 3450–3453; Angew. Chem.
Int. Ed. 2008, 47, 3450–3453.
[17] S. Biella, F. Porta, L. Prati, M. Rossi, Catal. Lett. 2003, 90, 23–29.
[18] B. N. Zope, D. D. Hibbits, M. Neurock, R. J. Davis, Science 2010,
330, 74–78.
[6] P. Kuhn, A. Forget, D. S. Su A. Thomas, M. Antonietti, J. Am.
Chem. Soc. 2008, 130, 13333–13337.
[19] R. Sumathi, K. Johnson, B. Viswanathan, T. K. Varadarajan, Appl.
Catal. A 1998, 172, 15–22.
[7] P. Kuhn, A. Thomas, M. Antonietti, Macromolecules 2009, 42, 319–
326.
[8] P. Kuhn, A. Forget, J. Hartmann, A. Thomas, M. Antonietti, Adv.
Mater. 2009, 21, 897–901.
[9] M. J. Bojdys, J. Jeromenok, A. Thomas, M. Antonietti, Adv. Mater.
2010, 22, 2202–2205.
[10] A. Thomas, Angew. Chem. 2010, 122, 8506–8523; Angew. Chem. Int.
Ed. 2010, 49, 8328–8344.
[11] C. E. Chan-Thaw, A. Villa, P. Katekomol, D. Su, A. Thomas, L.
Prati, Nano Lett. 2010, 10, 537–541.
[12] C. E. Chan-Thaw, A. Villa, L. Prati, A. Thomas, Chem. Eur. J. 2011,
17, 1052–1057.
[13] A. Thomas, A. Fischer, F. Goettmann, M. Antonietti, J. O. Mꢅller,
R. Schlçgl, J. M. Carlsson, J. Mater. Chem. 2008, 18, 4893–4908.
[14] S. Kundu, Y. Wang, W. Xia, M. Muller, J. Phys. Chem. C 2008, 112,
16869–16878.
[20] M. Sankar, E. Nowicka, R. Tiruvalam, Q. He, S. H. Taylor, C. J.
Kiely, D. Bethell, D. W. Knight, G. J. Hutchings, Chem. Eur. J. DOI:
10.1002/chem.201003484.
[21] K. Maeda, K. Domen, J. Phys. Chem. C 2007, 111, 7851–7861.
[22] X. C. Wang, K. Maeda, A. Thomas, K. Takanabe, G. Xin, J. M.
Carlsson, K. Domen, M. Antonietti, Nat. Mater. 2009, 8, 76–80.
[23] M. Groenewolt, M. Antonietti, Adv. Mater. 2005, 17, 1789–1792.
[24] B. D. Cullity, Elements of X-ray Diffraction, 2nd ed., Addison-
Wesley Publishing Company, Reading, 1978.
[25] D. Wang, A. Villa, F. Porta, D. Su, L. Prati, Chem. Commun. 2006,
1956–1958.
[26] J. F. Moulder, W. F. Stickle, P. E. Solbol, K. D. Bomben, Handbook
of X-ray Photoelectron Spectroscopy, Perkin–Elmer Corp., 1992.
Received: June 24, 2011
Published online: December 23, 2011
[15] F. Su, S. C. Mathew, G. Lipner, X. Fu, M. Antonietti, S. Blechert, X.
Wang, J. Am. Chem. Soc. 2010, 132, 16299–16301.
Chem. Asian J. 2012, 7, 387 – 393
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemasianj.org
393