G Model
CATTOD-10139; No. of Pages7
ARTICLE IN PRESS
X. Li et al. / Catalysis Today xxx (2016) xxx–xxx
6
Fig. 7. TEM images and particle size histogram of the catalyst composit Pd-Ce NPs/functionalized Fe-MIL-101-NH2.
to 1717 and 1720 cm 1 for functional Fe-MIL-101-NH2 and Pd-Ce
−
phology seems to be changed in the modification of the reaction
time in synthesizing Fe-MIL-101-NH , the commonly used base-
line of 20 h have been used for the synthesis of materials of the
subsequent steps of this work.
NPs/functional Fe-MIL-101-NH , respectively.
In Raman spectra as shown in Fig. 4 (b), a new peak at 660 cm
appeared for Pd-Ce NPs/functional Fe-MIL-101-NH in comparison
2
2
−
1
2
It has been demonstrated that neocuproine coordinated palla-
with absorptions of the pristine and functional Fe-MIL-101-NH2.
Since the Raman spectroscopy is not able to detect pure metal-
lic phase [39]. However, Raman spectroscopy is able to detect the
incorporation of small atoms such as O and C, into the lattice of
these metallic phases. Furthermore, NaBH4 is commonly used to
2
dium complex, ((ꢀ -neocuproine)PdOAc) (OTf) , to be the most
2
2
active homogeneous catalyst in glycerol selective oxidation
towards DHA [36]. The neocuproine ligand plays a crucial role in the
catalytic procedure to control the stereo effects [36]. Therefore, in
the work, neocuproine ligand has been attached onto Fe-MIL-101-
NH2 by forming an amide bond (CO NH) and that further been
used as support to hold the transition metal nanoparticles-based
catalysts. Similar with other MOF supported metal nanoparticle-
based catalysts such as Pt NPs/MIL-101 [37] and Pd NPs/MOF [38],
transition metal nanoparticles used in this work could be immo-
◦
reduce palladium complexes to generate Pd particles. The result-
◦
ing Pd is not detectable in Raman. Therefore, the new peak is most
probably contributed from vibration of the Ce O bond in Ce O ,
2
3
which is produced from the reaction of Ce(acac)3 in our system.
The oxidation states of the supported metal nanoparticles
were identified by XPS analysis. As shown in Fig. 5(a), the Pd
3d showed peaks at 334.7 and 339.9 eV (ꢁ = 5.2 eV), which were
bilized inside the pores of the functional Fe-MIL-101-NH . During
2
◦
the functionalization of Fe-MIL-101-NH , as shown in Scheme 1,
consistent with Pd [40]. Two peaks were detected for the Ce
2
neocuproine (1) was first oxidized to dialdehyde (2) using sele-
nium dioxide [30,31]. Intermediate (2) was subsequently oxidized
3d (Fig. 5(b)), which appeared at the binding energy (902.5 and
884.5 eV) expected for Ce3+ species [41]. In XRD spectra, the func-
tionalized Fe-MIL-101-NH2 and catalyst Pd-Ce NPs/functionalized
Fe-MIL-101-NH2 show similar absorption patterns.
with nitric acid to form 1,10-phenanthroline-2,9-dicarboxylic acid
13
(
3) [32]. Products 2 and 3 were characterized by 1H, C NMR
spectra. The results are consistent with literature reports [32]. The
latter compound was in situ converted into 1,10-phenanthroline-
The functional products were also analyzed by SEM spec-
troscopy. As shown in Fig. 6, no significant differences were
observed in SEM images for Fe-MIL-101-NH2 and after modifica-
tion with neocuproine ligand. All the above results confirm that
the palladium and cerium-based nanocatalyst were supported on
the functional MOF Fe-MIL-101-NH2.
2
,9-diacyl chloride (4) by reacting with thionyl chloride [32–35].
The obtained intermediate (4) was immediately reacted with Fe-
MIL-101-NH2 without being isolated to provide the neocuproine
functionalized MOFs (5). Transition metal nanoparticles (NPs)-
based catalysts such as palladium NPs and palladium-cerium NPs
were immobilized on the supports by co-precipitation methods by
Fig. 7 shows the TEM images of the supported catalyst Pd-Ce
NPs/functionalized Fe-MIL-101-NH . It can be seen that the Pd-Ce
2
reducing corresponding precursors PdAc2 and Ce(acac) , respec-
tively with sodium borohydride. The products were characterized
by XRD, TEM, SEM, NMR, XPS and FT-IR.
NPs are small, with an average particle size of ∼5 nm (determined
from the measurement of ∼130 particles), and well dispersed on
the support. The catalysts were also analyzed by ICP and SEM-EDX
to identify metal loading amount and shown that the total metal is
around 5 wt% according to ICP analysis. The catalyst was examined
for alcohol oxidation reactions. A yield of 55% of DHA was obtained
by the catalyst, and that is slightly higher than literature results
[21–23]. For comparison, catalysts Pd NPs/functional Fe-MIL-101-
3
The FT-IR, Raman and XRD spectroscopy were used to character-
ize support precursor Fe-MIL-101-NH , neocuproine ligand, func-
2
tionalized Fe-MIL-101-NH , and catalyst Pd-Ce NPs/functionalized
2
Fe-MIL-101-NH2 as shown in Fig. 4. In FT-IR spectra, normal
absorptions for functional groups were observed. The pristine Fe-
−
1
MIL-101-NH2, the broad absorptions at 3468 and 3367 cm can
be attributed to the N H stretching vibrations. After function-
alization with neocuproine ligand, the N H stretching bands in
NH , Ce NPs/functional Fe-MIL-101-NH2 and Au NPs/functional
2
Fe-MIL-101-NH2 were prepared in the same method and exam-
ined for glycerol oxidation towards DHA. The DHA yields of 17%,
4% and 10% were obtained, respectively. Although the catalyst
−
1
the amide bonds down-shifted to 3210 and 3213 cm (weak and
broad) for functional Fe-MIL-101-NH2 and Pd-Ce NPs/functional
shows lower activity in comparison with corresponding homoge-
2
Fe-MIL-101-NH , respectively. Neocuproine ligand shows a strong
neous catalyst ((ꢀ -neocuproine)PdOAc) (OTf) , which gives a high
2
2
2
−1
absorption at 1736 cm which can be attributed to the aromatic
C stretching vibrations. After attaching to MOF, the peak shifted
ered conveniently either by filter or centrifugation. The catalyst
C
Please cite this article in press as: X. Li, et al., Pd-Ce nanoparticles supported on functional Fe-MIL-101-NH : An efficient catalyst for
2