34
S. Ghosh et al. / Catalysis Communications 72 (2015) 33–37
the reaction, the catalyst was separated by filtration and the products
were analyzed by GC (FID) and GCMS.
3. Results and discussion
3.1. Catalyst characterization
The Ag/WO3 nanocomposite was characterized by various tech-
niques. The X-ray diffraction (XRD) pattern of the Ag–W catalyst
(Fig. 1) showed the peaks at 2θ values of 23.1°, 23.7°, 24.4°, 33.3° and
34.0°, confirms the formation of monoclinic WO3 (JCPDS No. 43-1035,
space group: P21/n) (Fig. 1). The very small peaks at 2θ values at
44.5°and 64.3° correspond to metallic Ag crystal faces (200) and (220)
respectively [JCPDS no. 04-0783]. No other peak due to other phase of
tungsten-oxide was observed. After reaction, the oxidation state of me-
tallic silver Ag(0), remained intact, which is confirmed from the XRD
(Fig. 1b). X-ray photoelectron spectroscopy (XPS) analyses confirmed
the presence of metallic silver in the fresh sample from the correspond-
ing Ag 3d5/2 and Ag 3d3/2 binding energy values of 368.2 eV and
374.2 eV respectively (Fig. S1, Supporting Information) [6–11,12]. The
W 4f5/2 and 4f7/2 spectra attributed to the binding energies 37.9 eV
and 35.8 eV respectively suggesting that the tungsten in the tungsten
oxide sample exists as W+6 (Fig. S2, Supporting Information) [6–11,13].
For further investigation of the surface property and to detect subtle
phase information of the composite, Raman (Fig. 2) and FTIR (Fig. S5,
Supporting Information) analyses were conducted. Generally, the
950–1050 cm−1 Raman wave numbers of the transition metal oxide
are assigned to be the symmetric stretching modes of metal and ox-
ygen bonds [short terminal W = O, υs(W = O) terminal bands],
and 750–950 cm−1 bands were either the antisymmetric stretching
of W–O–W bonds [υs(W–O–W)] or symmetric stretching of –O–W–
O– bonds [υs(–O–W–O–)] [14]. The Raman spectrum of the Ag/WO3
NP catalyst detected vibrational peaks at 808, 724, 330, 300, 274, 136,
Fig. 2. Raman spectra of (a) fresh and (b) spent Ag/WO3 nano-bars catalyst.
analysis of the composite (Fig. 3b, Supporting Information) revealed
that, there appeared a distribution of Ag, W and O only, and no sort of
C, N or Br, which are visualized from the corresponding SEM-EDX
image of the uncalcined sample (Fig. S4, Supporting Information). This
observation indicated the complete removal of the structure-directing
template upon calcination. This experimental finding was further sup-
ported from FTIR and TGA/DTG analysis of the uncalcined catalyst
(Fig. S5 & S6, Supporting Information). However, transmission electron
microscopy (TEM) image (Fig. 3) revealed that the sample is comprised
of several bar-like particles with 350–450 nm in length with 60–100 nm
width in average. More interestingly, the reactive species (i.e. Ag NPs)
were 2–7 nm in size and spherical in shape and were anchored upon
these nano-bars. From the HRTEM (Fig. 3c) the interplanar spacing of
the lattice fringe distance of 0.38 nm, corresponded to (020) lattice
spacing of monoclinic WO3, was clearly discriminated from 0.23 nm
corresponded to (111) plane of Ag (Fig. 6d). Typically, the dispersion
of Ag, W and O atoms in the catalyst was also analyzed by STEM-
elemental mapping (Fig. S15). It indicated that each of Ag, W and O spe-
cies was homogeneously dispersed. Moreover, the fact that the catalyst
retained its structure was interpreted in terms of TEM diagram (Fig. 3d),
TEM–EDX (Fig. S7, Supporting Information) and particle size distribu-
tion (histogram, Fig. S8, Supporting Information) analyses of the spent
catalyst.
86 and 60 cm−1. The two main intense peaks at 808 and 724 cm−1
,
and the shoulder at 686 cm−1, are typical Raman peaks of crystalline
WO3, which correspond to the stretching and bending vibrations of
the bridging tungsten and oxygen atoms [15]. They are assigned to be
the W–O stretching (υ), W–O bending (δ) and O–W–O deformation
(γ) modes, respectively. Two peaks at 326 and 274 cm−1 are assigned
to be the bending δ(O–W–O) vibrations [14]. Those below 200 cm−1
modes were attributed to the lattice vibrations [14]. After the reaction,
the Raman spectrum of the spent catalyst was unchanged (Fig. 2b),
reflected the structural stability of the catalyst under the reaction
condition.
The topology of the catalyst was studied by scanning electron mi-
croscopy (SEM, Fig. S3a, Supporting Information) which showed a typ-
ical sample composed of nano-bars with width 80–150 nm. SEM–EDX
3.2. Catalytic activities
The activity of the catalyst (AgWNB) in selective oxidation of cyclo-
hexanone has been shown in Table 1. No caprolactone was detected
during neat reaction, reflecting the necessity of the catalyst (Entry 1,
Table 1). Temperature played a crucial role in the oxidation reaction of
cyclohexanone (Fig. S9, Supporting Information). Increment in temper-
ature increased the yield of caprolactone and reached maximum (97%)
at 80 °C. But above 80 °C, yield of caprolactone decreased rapidly due to
the formation of cyclohexanone-2-ol, cyclohexanone-2-one etc. We
also noticed that, optimum molar ratio of cyclohexanone: H2O2 was
1:4; probably, excess H2O2 was needed, since major amount of H2O2
decomposed over the catalyst at the reaction temperature. Maintaining
all the optimum conditions, when the reaction was allowed to run for
hours, we did not notice any marked effect in the yield of caprolactone
after 9 h, probably due to the decomposition of H2O2 present in the
reaction medium (Fig. S12, Supporting Information).
Notably, commercial catalysts, employed separately (Entry 2–4,
Table 1), and even conventional Ag/WO3 catalyst prepared in
Fig. 1. XRD diffractogram of the (a) fresh, (b) spent Ag/WO3 nano-bars catalyst,
(c) commercial WO3 and (d) metallic Ag.