Mechanistic insights into the microwave-assisted cinnamyl alcohol oxidation using supported iron and palladium catalysts

Article abstract of DOI:10.1016/j.mcat.2019.110409

The efficient oxidation of alcohols remains a challenge, especially when using solid heterogeneous catalysts. Herein, the catalytic activity of mechanochemically prepared iron and palladium particles on various supports are compared for the oxidation of c

Full text of DOI:10.1016/j.mcat.2019.110409

Molecular Catalysis 474 (2019) 110409
Contents lists available at ScienceDirect
Molecular Catalysis
journal homepage: www.elsevier.com/locate/mcat
Mechanistic insights into the microwave-assisted cinnamyl alcohol
oxidation using supported iron and palladium catalysts
a
b
a
b
b
Yantao Wang , Pepijn Prinsen , Floriane Mangin , Alfonso Yepez , Antonio Pineda ,
c
d
e
Enrique Rodríguez-Castellón , Muhammad Rehan Hasan Shah Gilani , Guobao Xu ,
Christophe Len , Rafael Luque
Sorbonne Universités, Université de Technologie de Compiègne, Centre de Recherche Royallieu, CS 60319, F-60203 Compiègne cedex, France
Departamento de Química Orgánica, Universidad de Córdoba, Campus de Rabanales, Edificio Marie Curie (C-3), Ctra Nnal IV-A, km. 396, E-14014 Cordoba, Spain
Departamento de Química Inorgánica, Cristalografía y Mineralogía (Unidad Asociada al ICP-CSIC), Facultad de Ciencias, Universidad de Málaga, Campus de Teatinos,
9071 Málaga, Spain
Department of Chemistry, Government College University of Faisalabad, Layyah campus, Layyah, Pakistan
The State Key Laboratory of Electroanalytical Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, 5625 Renmin Street, Changchun, Jilin
30022, China
a,f
b,e,g,⁎
a
b
c
2
d
e
1
f
PSL Research University, Chimie ParisTech, CNRS, Institut de Recherche de Chimie Paris, 11 rue Pierre et Marie Curie, F-75231, Paris Cedex 05, France
Peoples Friendship University of Russia (RUDN University), 6 Miklukho-Maklaya str., 117198, Moscow, Russia
g
A R T I C L E I N F O
A B S T R A C T
Keywords:
Cinnamyl alcohol
Oxidation
Hydrogen peroxide
Microwave irradiation
Fe and Pd
The efficient oxidation of alcohols remains a challenge, especially when using solid heterogeneous catalysts.
Herein, the catalytic activity of mechanochemically prepared iron and palladium particles on various supports
are compared for the oxidation of cinnamyl alcohol to cinnamaldehyde and benzaldehyde in acetonitrile using
aqueous hydrogen peroxide as the oxidant assisted by microwave irradiation. Important insights into the me-
chanism of the reaction towards the differently observed products have been provided.
Introduction
frequencies, optionally aided by co-catalysts, the latter enable an effi-
cient catalyst separation and reduce waste streams. Solid catalysts for
Aldehydes and ketones represent an important commercial volume
in the market, both as bulk chemicals as well as synthetic intermediates
selective oxidations are mainly based on supported metal particles [11],
such as Pd nanoparticles supported on mesostructured silica (SBA-15)
[
1] or as end-products, including compounds with biological activity,
2 3
[12], iron oxide (Fe O ) nanoparticles [13], (bimetallic) Au particles
olfactory properties, etc. [2,3]. The shortest pathway for the synthesis of
carbonyls is the oxidation of the corresponding alcohol [4], both in
liquid and gas phase. The preparation of aldehydes and ketones through
selective oxidations is an important transformation step in organic
synthesis [5]. In the past, stoichiometric oxidants based on chrome,
magnesium permanganate, manganese or other metal derivatives were
used, [6,7] leading to the formation of toxic waste and costly waste
disposal/treatment [8]. Research in this field has mainly focused at
increasing the atom and energy efficiency, the use of less hazardous
solvents, the use of renewable feedstocks and the development of effi-
cient oxidation methods that employ safer oxidants like hydrogen
supported on metal oxides [14,15], Ru supported on alumina [16],
zeolites [17], etc. The use of recoverable heterogeneous catalysts for
selective oxidations remains a challenging subject, especially when
using H
[18]. Solid catalysts for H
2
O
2
as the oxidant, as reviewed recently by Sheldon (2015)
directed oxidations generally involve
2 2
O
early transition elements, mostly tungsten, molybdenum and vanadium,
in high oxidation states and peroxometal complexes. Recently, efficient
oxidation of benzyl alcohol (primary alcohol) to benzaldehyde using
2 2
H O was also demonstrated for small amounts of iron supported on
furfuryl alcohol based resin support as alternative to the previously
mentioned expensive noble metals (Pd, Au, Ru) [19]. The oxidation of
cinnamyl alcohol is way more challenging, at least when one aims the
selective production of cinnamaldehyde. Alcohol feedstocks can be
classified in aliphatic, allylic, benzylic and heterocyclic alcohols, as well
2 2 2
peroxide (H O ) or molecular oxygen (aerobic oxidation with O or air)
[
9,10]. Both homogeneous and heterogeneous catalysts have been de-
veloped; whereas the former generally provide much faster turnover
⁎
Corresponding author at: Departamento de Química Orgánica, Universidad de Córdoba, Campus de Rabanales, Edificio Marie Curie (C-3), Ctra Nnal IV-A, km.
396, E-14014 Cordoba, Spain.
E-mail address: rafael.luque@uco.es (R. Luque).
https://doi.org/10.1016/j.mcat.2019.110409
Received 18 February 2019; Received in revised form 15 May 2019; Accepted 16 May 2019
2468-8231/ © 2019 Elsevier B.V. All rights reserved.

Y. Wang, et al.
Molecular Catalysis 474 (2019) 110409
as in primary, secondary and tertiary alcohols. In general, primary
benzylic alcohols can be converted efficiently, while other alcohols
remain more challenging, especially in terms of selectivity [20].
Cinnamaldehyde can be used in a wide variety of applications.
Cinnamaldehyde has been reported to exhibit antibacterial and anti-
fungal activity [21,22], as well as anti-cancer-like activity [23], being
also employed as a cross-linker in bioplastics [22] and as flavor/odor
agent [24]. At one hand, it can be isolated and purified from natural
resources such as cinnamon bark and leaves [25,26]. On the other
hand, it can be produced by the selective oxidation of cinnamyl alcohol.
Previous studies on catalytic cinnamyl alcohol oxidation proposed ru-
Commercial MFI zeolites ZSM-5-30 and ZSM-5-50 (ammonium
ZSM-5 CBV3024E and CBV5524GE) and FAU zeolites Y-5.2 and Y-60
(ammonium H-60 and H-5.2) were purchased from Zeolytes
International (USA). Zeolite materials were calcined at 600 °C during
24 h prior to use to remove ammonia. NC (where appropriate) stands
for uncalcined samples.
Catalyst synthesis and characterization
Fe/Al-SBA-15, Fe/P420, Fe/H-ZSM-5 and Fe/H-Y catalysts were
prepared mechanochemically in dry conditions: 1.0 g solid support and
0.020/0.040 g FeCl .4H O (equivalent to 0.5/1.0 wt% Fe to reach a
2 2
thenium complexes (using N-methylmorpholine-N-oxide, H
butyl hydrogen peroxide as the oxidants) in dichloromethane (reflux)
27] and copper catalysts using 2,2,6,6-tetramethylpiperidine 1-oxyl
TEMP) in acetonitrile (20 °C) [28]. Methods which employed only
(1–5 eq.) were reported using Ag PO photocatalysts in water,
2 2
O and t-
[
(
theoretical amount of 0.25 and 0.50 mol% Fe with respect to the cin-
namyl alcohol load) were milled together in a planetary ball mill
(Retsch 100) under previously reported optimized conditions (350 rpm,
10 min.) [37]. The solids were recovered and heated under air atmo-
sphere during 4 h at 400 °C (Fe/Al-SBA-15). For Fe/P420, the obtained
solids were calcined during 48 h at 120 °C. In the case of Fe/H-ZSM-5
and Fe/H-Y, the obtained solids were calcined at low temperature
(48 h, 120 °C) or at high temperature (2 h, 800 °C). For Pd/Al-SBA-15
and Pd/MAGSCN catalysts, the same protocol was followed as for Fe/
H
2
O
2
3
4
[
[
29] Pd(II) complexes in ionic liquids, [30] Pd supported on silica
31,32] and Pt black (20 °C) [33]. The latter gave the highest yields, ca.
9
0% cinnamaldehyde after 2–3 h at 80 °C. To our knowledge, only Pillai
et al. (2003) reported an iron-based catalyst for this reaction, with 19%
yield after 4 h at 65 °C using 4.0 eq. of H [34]. The present work
2 2
O
explores the catalytic activity of iron and palladium supported on
various supports for the oxidation of cinnamyl alcohol in acetonitrile
2
Al-SBA-15, but instead 0.03/0.06 g Pd(acac) was used as metal pre-
assisted by microwave irradiation using 1.6–2.8 eq. of aqueous H
a green oxidant. Fe (transition metal) was selected as compared to Pd
noble metal) to investigate the possibilities of the use of a cheap, en-
vironmentally friendly and largely available transition metal for selec-
tive oxidations. In addition to this, several supports available from
previous research from our group including aluminosilicates, zeolites
and a magnetically separable support were also tested in order to in-
vestigate the effect of the support in the metal deposition as well as in
the catalytic activity in the systems.
2
O
2
as
cursor (to reach a theoretical amount of 0.25 and 0.50 mol% Pd with
respect to the cinnamyl alcohol load).
Powder X-ray diffraction patterns (XRD) were recorded on a Bruker
D8-advance X-ray diffractometer with Cu Kα radiation (0.154 nm) over
the 2Ө range of 10-80°. For low angle experiments, XRD patterns were
acquired at a step size of 0.02° with a counting time per step of 20 s.
(
N
2
adsorption/desorption isotherms of were determined in the
Micromeritics automatic analyzer ASAP 2000 at -196 °C. Samples were
−2
previously degassed overnight at 130 °C under vacuum (P < 10 Pa).
The linear determination of the BET equation was carried out to obtain
the specific surface area.
Experimental section
Scanning electron microscopy images were recorded with a JEOL
JSM-6300 scanning microscope (JEOL Ltd., Peabody, MA, USA)
equipped with Energy-dispersive X-ray spectroscopy (EDX) of 15 kV at
the Research Support Service Center (SCAI) from University of
Cordoba.
Elemental analysis in the Fe loaded zeolites was performed by
Inductively Coupled Plasma Mass Spectrometry (ICP-MS) using Perkin
Elmer Nexion 300X equipment. Prior to analysis, the samples were
digested in hydrofluoric acid.
Catalyst supports
The synthesis of mesoporous aluminosilicate Al-SBA-15 was carried
out following the procedure described by Zhao et al. (1998) [35]. The
triblock co-polymer Pluronic P123 (0.41 mmol) was used as directing
agent and dissolved in aqueous HCl (2 M, pH 1.5) for 2 h at room
temperature. After complete dissolution, tetraethyl orthosilicate
(
0.025 mol) was used as the silica source and the corresponding amount
of aluminum isopropoxide (0.01 mol) was slowly added. The mixture
was stirred for 24 h at room temperature, followed by hydrothermal
treatment in an oven at 100 °C for 24 h. The gel obtained was filtered
off, dried and subsequently calcined under nitrogen atmosphere at
X-ray photoelectron spectra (XPS) were acquired using a Physical
Electronics PHI 5700 spectrometer with non-monochromatic MgαKα
radiation (300 W, 15 kV, and 12,565 eV) at Universidad de Malaga. C1s
line at 284.6 eV from adventitious carbon was used as reference.
HR-TEM micrographs were recorded in a JEOL JEM 2010 micro-
scope (operated at 200 kV) with a resolution of 0.194 nm.
600 °C for 2 h, and then under air atmosphere for 4 h. The obtained SBA
solids (Si/Al molar ratio 20) were recovered and stored.
The magnetic support MAGSCN was prepared according to the
procedure reported by Ojeda et al. (2014): [36] 0.5 g solid SBA-15 silica
support was ground with 1.34 g solid Fe(NO
00 planetary ball mill in a 125 mL vessel using eighteen 10 mm
stainless steel balls. Optimized milling conditions were 10 min. at
50 rpm [37]. Fe-containing SBA-15 was subsequently reacted with
propionic acid (10 mL, ca. almost 1:10 material:propionic acid ratio) at
3
)
3
.9H
2
O in a Retsch PM-
Catalytic experiments
1
The experiments were carried out in a closed pressure-controlled
vessel under continuous stirring, assisted by microwave irradiation
using a CEM-DISCOVER model with PC control. The microwave irra-
diation of the reaction mixture was power-controlled (300 W), reaching
temperatures in the 84–112 °C range (average temperature 98 °C), as
measured by an infra-red probe. The pressure in the closed system
ranged from 34 to 170 psi, with an average of 100 psi. In a typical re-
action, 0.247 mL of cinnamyl alcohol (Sigma Aldrich, 98% purity) was
suspended in acetonitrile (3 mL) containing 50 mg catalyst and then
3
8
5 °C, 3 h under static vacuum to achieve the magnetic phase. The
−
1
nanocomposite was slowly heated up to 300 °C under air (1 °C min
and kept at 300 °C for an additional 30 min.
)
P420 resin material was prepared according to the procedure re-
ported by Mangin et al. (2018): [19] 0.01 g (10 μL) of pure furfuryl
alcohol was mixed with 990 μL of deionized water, followed by addition
of 10 mL 9 M HCl, causing a rapid change of color from yellow to violet.
The mixture was stirred for 1 h at 20 °C. The resin spheres were purified
with distilled water and separated by centrifugation (12,000 rpm, 4 °C,
2 2
H O (0.300 mL of 30 or 50% w/w aqueous solution) was added with
catalyst to start the reaction. After the reaction, the mixture was cooled
to room temperature and filtered off. Subsequently, the filtrate was
analyzed by GC.
40 min.). Fig. S1 shows a SEM image of the purified resin surface.
2

Y. Wang, et al.
Molecular Catalysis 474 (2019) 110409
Product analysis
the presence of any FeCl
mite (α-Fe
3 4
800 °C any phase of magnetite (γ-Fe O
2
phase (Fig. S2), neither hematite or maghe-
). Neither in the spectrum of the catalysts calcined at
) was clearly visible. It was
2
O
3
GC analysis was performed using an Agilent 6890 N GC chromato-
graph equipped with a Supelco 2-8047-U (60 m × 25 m × 25 μm) ca-
pillary column and an FID detector. The temperature of the column was
difficult to reveal the iron crystal phase as most of the XRD signals of
the zeolite are present at identical diffraction angles compared with
−
1
−1
ramped at 10 °C min to 100 °C (0 min hold time), then at 20 °C min−
pure FeCl
that, upon sublimation of FeCl
2
, α-Fe
2
O
3
and γ-Fe
3
O
4
. Marturano et al. (2001) demonstrated
in H-ZSM-5 zeolite by chemical vapor
1
to 120 °C (20 min. hold time) and finally to 180 °C at 20 °C min
3
−
1
+
(
10 min hold time). The nitrogen gas flow was 1.8 mL min . The re-
deposition at 320 °C, FeCl
2
species formed in the zeolite framework, at
tention times of benzaldehyde, cinnamaldehyde and cinnamyl alcohol
were 9.6, 17.7 and 25.1 min, respectively. Yield, conversion and se-
lectivity were determined as:
least when no washing step was applied [39]. Similar observations were
reported by Kucherov and Shelef [40]. These authors identified a wide
range of (active) iron phases depending on the exact synthesis condi-
tions with the aid of advanced characterization techniques including
EXAFS, UV–vis and AAS spectroscopy.
MolProduct
MolInitial
Yield (mol%) =
× 100
In view of the catalytic application, i.e. the oxidation of cinnamyl
alcohol, no concise data was found on the actual catalytic active phase.
Parlett et al. demonstrated, at least for Pd supported on mesoporous
silica, that the metal oxide (and not the metal) nanoparticle is the active
phase in the oxidation of cinnamyl alcohol when using O2 as the oxi-
dant [41,42]. Considerable effects of the support on the selectivity were
actually the result of varying dispersions of the metal oxide. The
identification of iron phase in the materials synthesized in the present
study by XRD was not possible due to, firstly, the low metal loading in
the catalysts as well as the high crystallinity of the zeolites whose dif-
fraction lines could overlap those corresponding with the different iron
species. XPS measurements, especially useful for these material in
which X-ray diffraction is not helpful due to the low metal loading and
high crystallinity of the support, were used to identify the metallic
species present after the different thermal treatment that these mate-
rials underwent (calcination at 120 and 800 °C). Fig. 1 shows the XPS
spectrum in the Fe 2p region of Fe/H-ZSM-5-30_120C sample, ex-
hibiting peaks corresponding with Fe(II) and Fe (III) as a result of
partial transformation of Fe(II) from the metal salt into Fe (III) while in
the respective spectrum for the material calcined at 800 °C the dis-
[
MolInitial − MolFinal]
Conversion (mol%) =
× 100
MolInitial
MolProduct
Selectivity (mol%) =
× 100
MoL,initial − MoL,final
The response factors (RF) relative to
a fixed concentration
−
1
(
9.04 mg mL ) of benzyl chloride as internal standard (IS) were ex-
perimentally determined by preparing dilution sets of cinnamyl al-
cohol, dihydrocinnamyl alcohol, cinnamaldehyde and benzaldehyde,
which were 1.207, 1.486, 1.189 and 1.047, respectively. The linear
regression coefficients of the linear fittings were high in all cases
2
(
r > 0.99). The RF of the remaining compounds identified in the
product mixtures with unknown RF were calculated by the Effective
Carbon Method (ECN) [38] relative to a compound with an experi-
mentally determined RF and with a similar structure in terms of total
carbon atom number, as the following:
ECNunknown
ECNExp. known
RFunknown = RFExp. known ×
tinctive peak at 711 eV of Fe 2p3/2 corresponding with Fe
observed (see also Fig. S3) [43,44].
2 3
O can be
Table S1 (see Supporting Information) shows the calculated ECN
factors and the corresponding calculated RF. Some compounds (those
present in very small amounts) which did not elute in the GC-FID
analysis were quantified by comparison of their relative peak area in
the GC/MS analysis with the peak area of a compound (both relative to
the IS peak area) which did elute in the GC-FID analysis.
The nominal metal loading in synthesized catalysts was confirmed
by ICP-MS for selected samples (Table S3). An iron content ca. 0.4 wt.%
was measured for zeolite samples with 0.5 wt% nominal loading.
Chloride anions from the metal salt precursor were present in the
samples treated at 120 °C for 48 h while in the sample calcined at 800 °C
for 2 h no Cl was not detected. Additionally, TEM micrographs were
recorded of the Fe containing zeolites to search for Fe nanoparticles as
well as to evaluate the morphology of the metal functionalized
A selected set of product mixtures obtained after filtration were also
analyzed by GC–MS analysis on a gas chromatograph coupled with a
triple quadrupole mass detector (Bruker, GC-436, Scion TQ) working
with electron impact ionization and ion detection in the scan range of
m/z 35–450 (0.5 s per scan). The chromatograph was equipped with a
high-temperature capillary 5% phenylsilicon column (DB-5HT, 30 m
×
0.25 mm i.d., 0.25 μm film thickness). Helium was used as carrier gas
−1
at a rate of 1 mL min . The injector temperature was set at 265 °C and
the split ratio was 50. The oven was heated from 80 °C to 200 °C at
−
1
−1
3
°C min
and finally from 200 to 280 °C at 35 °C min
(3.5 min.).
The identification of the compounds was performed by comparison
with pure standards, with MS data in the NIST library.
Results and discussion
Low amounts of iron and palladium (0.5–1.0 wt%) were supported
on various well-defined supports via mechanochemical synthesis in dry
conditions (ball-milling). Table S2 (see Supporting Information) shows
the Si/Al ratio, specific surface area, pore size and pore volume of the
aluminosilicate supports used in the present study, as determined ex-
perimentally by nitrogen adsorption analysis at −196 °C. The in-
corporation of such low amounts of iron and palladium onto the dif-
ferent supports did not change significantly the textural properties in
the final material as compared with their respective support.
Comparison of the XRD spectra of the H-Y-5.2 and Fe/H-Y-5.2 materials
Fig. 1. XPS spectra in the Fe 2p region for the materials: a) Fe/H-ZSM-5-
30_120C and b) Fe/H-ZSM-5-30_800C (calcined at 120 °C and 800 °C, respec-
tively).
(
iron on zeolites obtained after calcination at 120 °C) failed in revealing
3

Y. Wang, et al.
Molecular Catalysis 474 (2019) 110409
Table 1
Cinnamyl alcohol oxidation after 5 min. at 300 W microwave irradiation (ca. 100 °C) with Fe supported on Al-SBA-15 as compared to the corresponding dissolved
counterparts FeCl
2
and compared to Pd/Al-SBA-15.
Entry
Catalyst (% Fe/Pd)a
H
2
O
2
(eq.)a
Conversion (%)a
Yield (selectivity, %)a
Cinnamaldehyde
Benzaldehyde
1a
1b
2a
2b
3a
3b
4a
4b
5a
5b
6a
6b
7a
7b
–
–
1.6
2.8
1.6
2.8
1.6
2.8
1.6
2.8
1.6
2.8
1.6
2.8
1.6
2.8
5
9
1 (34)
3 (33)
3 (29)
5 (28)
8 (30)
11 (28)
7 (25)
9 (24)
10 (17)
10 (13)
11 (20)
10 (12)
8 (34)
10 (45)
1 (26)
2 (27)
4 (31)
5 (32)
13 (45)
19 (47)
14 (49)
19 (50)
32 (57)
48 (61)
32 (55)
51 (62)
6 (27)
Al-SBA-15
Al-SBA-15
FeCl
FeCl
FeCl
FeCl
Fe/Al-SBA-15 (0.25)
Fe/Al-SBA-15 (0.25)
Fe/Al-SBA-15 (0.50)
Fe/Al-SBA-15 (0.50)
Pd/Al-SBA-15 (0.25)
Pd/Al-SBA-15 (0.25)
12
16
28
41
29
38
57
79
58
82
24
22
2
2
2
2
(0.80)
(0.80)
(1.60)
(1.60)
8 (38)
a
In molar percentages on cinnamyl alcohol load.
materials after the milling process. The general trend was that samples
of the zeolite materials calcined at 800 °C showed smaller particle sizes
compared to those calcined at 120 °C (Fig. S4). Slight dark spots in the
TEM images apparently showed the presence of metal nanoparticles.
As-synthesized catalysts were tested for the oxidation of 0.5 M cin-
namyl alcohol in acetonitrile using microwave irradiation as a tool to
reduce the reaction times drastically. To compare the catalyst activity,
reaction times were set at 5 min. The amounts of supported iron (Fe)
and palladium (Pd) used in the catalytic experiments corresponded
with 0.25–0.50% (molar percentage with respect to cinnamyl alcohol
load). Table 1 shows that in the present conditions benzaldehyde is the
furfuryl alcohol derived resin (P420), aluminosilicate (Al-SBA-15),
zeolite (H-ZSM-5-30) and a magnetic nanocomposite (MAGSNC). At
lower oxidant load (1.6 eq., Fig. 2a), the conversions varied between 11
and 33%, except for Fe/Al-SBA-15 (57%). Higher oxidant loads (2.8 eq.,
Fig. 2b) were required to reach high conversion levels. Interestingly,
using small amounts of Fe (0.1%) supported on P420 resin, 52% con-
version was achieved, the highest value after Fe/Al-SBA-15 (79%). The
use of comparable quantities (ca. 0.4-0.5% Fe) on P420 did not sig-
nificantly increase the conversion in the systems (ca. 60% maximum
conversion achieved for almost identical Fe content). The selectivity to
cinnamaldehyde in turn dropped from 41 to 23%. The highest se-
lectivity was observed for Fe/H-ZSM-5-30 and Pd/Al-SBA-15. Their
conversions were 42 and 21%, respectively. This result, as the best
compromise between conversion and selectivity to cinnamaldehyde,
conducted us to further explore H-ZSM-5-30 and other zeolites as
support for Fe particles. Fig. 3 compares the catalytic activity of MFI
and FAU zeolites (H-ZSM-5 and H-Y), and the activity of the same
zeolites with Fe supported on it, obtained after calcination at low
temperature (120 °C, 48 h) and at high temperature (800 °C, 2 h).
Zeolites as such gave higher conversion (15–33 %) compared to Al-
SBA-15 (16%), with significantly improved selectivity to cinnamalde-
hyde (29–65 %). When incorporating low amounts of Fe (1 wt%) in the
zeolite framework through calcination at low temperature, the
2 2
main product. Blank experiments with 1.6 and 2.8 eq. H O (Table 1,
entries 1a and 1b) showed low cinnamyl alcohol conversion (5–9 %).
When using only support (Al-SBA-15), conversions increased slightly to
12–16 % (entries 2a and 2b). Fe supported on Al-SBA-15 enhanced the
conversion drastically to 57–82 %, but with poorer selectivity to cin-
namaldehyde (12–20 %).
2
Higher amounts of free dissolved iron (0.8 and 1.6% FeCl ) gave
only 28–41 % conversion at 24–30 % selectivity to cinnamaldehyde.
When changing Fe for Pd (entries 7a and 7b) in the catalyst synthesis
procedure, the conversions dropped noticeably to 21–24 %, whereas
the selectivity to cinnamaldehyde increased to 34–45 %.
Different supports were subsequently compared (Fig. 2), including a
Fig. 2. Conversion of cinnamyl alcohol and selectivity to cinnamaldehyde and benzaldehyde using Fe and Pd catalysts prepared similarly on different supports 50 mg
of catalyst (containing 0.5% Fe and Pd, except for P420 support: 0.1% Fe) at (a) 1.6 eq and (b) 2.8 eq. H O .
2 2
4

Y. Wang, et al.
Molecular Catalysis 474 (2019) 110409
Fig. 3. Conversion of cinnamyl alcohol and selectivity to cinnamaldehyde and
benzaldehyde using 2.8 eq. H O using different zeolites as compared to the
2 2
same zeolites with supported Fe (0.5%) obtained after 48 h calcination at 120 °C
and after 2 h calcination at 800 °C.
conversions increased to 42–49 %. Interestingly, the selectivity to cin-
namaldehyde was fully retained for the zeolites having low Si/Al ratio
(
H-ZSM-5-30 and H-Y-5.2) whereas the selectivity decreased for the
high Si/Al ratio zeolites (H-ZSM-5-50 and H-Y-60). When incorporating
Fe through calcination at high instead of low temperature, the con-
versions further increased with 6–10% without any significant loss of
selectivity. Both the catalyst powders obtained after calcined at 120 and
800 °C exhibited a pink color, whereas the not-calcined samples were
pale yellow. While the color of the calcined catalysts was retained after
the catalytic experiment (and recovery, washing and drying), the color
of the recovered not-calcined catalysts disappeared, leaving behind the
bare zeolites (white powder). This indicates that both calcination pro-
cedures effectively supported iron on the support, whereas without
calcination the iron was leached from the zeolite structure during the
catalytic experiment. The low angle XRD spectra (Fig. 4) clearly show
that the samples calcined at 120 and 800 °C contained a different active
iron phase. The best performing catalyst was 0.5% Fe on H-ZSM-5-30
calcined at 800 °C (16% cinnamaldehyde and 19% benzaldehyde yields,
at 50% cinnamyl alcohol conversion after 5 min reaction time).
Some effects of changing reaction conditions were also studied.
Fig. 2 already showed that reduced oxidant loadings (1.6 eq.) resulted
in higher selectivity to cinnamaldehyde. When dosing 2.4 eq. in 3 steps
(
after 0, 1 and 3 min.) instead of adding all the oxidant at the beginning,
the selectivity to cinnamaldehyde increased from 31 to 42% (using Fe/
H-Y-60_120 °C as catalyst) but at the expense of a decreased conversion
(
dropping from 44 to 30%). Another way to increase the selectivity to
cinnamaldehyde (from 33 to 52%) was decreasing the microwave ir-
radiation power (Fig. 5a), once again conversion with a reduction in
conversion from 42 to 26%. When monitoring the course of the reaction
between 5 and 30 min (Fig. 5b) for Pd supported on Al-SBA-15, the
conversion increased from 21 to 69%, with a however significantly
decreased selectivity to both cinnamaldehyde (from 45 to 19%) and
benzaldehyde (from 38 to 30%). This means that significant amounts of
other products started to form at reaction times over 5 min as discussed
hereafter.
Fig. 4. Low angle XRD spectra of Fe (0.5 wt%) supported on H-Y-5.2 after
calcination during 2 h at 800 °C (top image), Fe (0.5 wt%) supported on H-Y-5.2
after calcination during 48 h at 120 °C (middle image) as compared to the
parent H-Y-5.2 zeolite.
mass balance (3–4%) was not covered. This is not surprising, as the
An in-depth study of the reaction product composition by GC/MS
analysis (Fig. 6a) elucidated the presence of side products, which were
not all observable in GC-FID analysis. A product mixture was obtained
containing cinnamaldehyde and benzaldehyde (14 and 18% yields,
respectively) as the main products using 0.5% Fe supported on H-ZSM-
oxidative disruption of the propyl chain in cinnamyl alcohol would lead
to the release of CO, CO , formaldehyde and/or formic acid, which are
2
not accounted in the mass balance. Interestingly, when repeating the
experiments but starting from dihydrocinnamyl alcohol (without in-
saturation, Fig. 6c) instead of cinnamyl alcohol, the mass balance
dropped to 76–85 %. Here, formic acid 3-phenylpropyl ester (com-
pound 12) was present among the side products (2.0 and 0.2% yields
for Fe/H-ZSM-5-30 and H-ZSM-5-30, respectively), which supports the
hypothesis of formaldehyde or formic acid release upon oxidative
5-30, while ca. 6% side products were formed. The same zeolite (H-
ZSM-5-30), but without Fe, resulted in 10% cinnamaldehyde and 8%
benzaldehyde yields with ca. 5% side products. At conversion levels of
42 and 26%, respectively, this means that a slight percentage of the
5

Y. Wang, et al.
Molecular Catalysis 474 (2019) 110409
Fig. 5. Conversion of cinnamyl alcohol and selectivity to cinnamaldehyde and benzaldehyde: (a) after 5 min using 0.5% Fe on H-ZSM-5-30_120C in function of the
microwave irradiation power and (b) between 5 and 30 min. at 300 W using 0.5% Pd supported on Al-SBA-15.
disruption. Independent from the starting compound, mass balances
were slightly lower when using Fe supported on zeolite as compared to
the bare zeolite, in accordance with their higher benzaldehyde yield.
Thus, the presence of iron increases the disruption activity in sacrifice
of higher selectivity to cinnamaldehyde. This is demonstrated in
Fig. 6b, were cinnamaldehyde was oxidized to benzaldehyde and cin-
namic acid among other side products. Importantly, when starting both
from cinnamyl alcohol and cinnamaldehyde, the corresponding epoxide
compounds (compounds 9 and 16) were detected in significant
amounts, while it was completely absent when starting from dihy-
drocinnamyl alcohol. This shows that the epoxidation occurs through
interaction of peroxy radicals with the double bound, independent from
the functionality of the carbon adjacent to it.
spectrum. It could be phenylaceytaldehyde but also the corresponding
epoxide (2-phenyloxirane) and consequently we could not fully exclude
the formation of epoxides in the reaction with the corresponding sa-
turated compound (dihydrocinnamyl alcohol). Anyway, this result
suggests that when starting from dihydrocinnamyl alcohol (Fig. 4c), less
benzaldehyde would be produced, which was experimentally evi-
denced. However, it did not lead to
a higher yield of dihy-
drocinnamaldehyde (compound 6). Instead, more other side products
were produced due to an increased release of formaldehyde and/or
formic acid from the terminal carbon in the propyl chain (which leads
to higher amounts of compounds 2 and 3) and due to the coupling of
radical fragments leading to products such as benzoin and 3,4-dihydro-
1-benzopyran (compounds 7 and 8) among others (which were not
detected here but which are plausible to be present as seen from the
Still, the nature of compound 2 was not fully clear based on its mass
Fig. 6. GC/MS chromatograms, cinnamyl alcohol conversion and selectivity to main and minor oxidation products and the corresponding mass balance obtained in
catalytic experiments (2.8 eq. H , 300 W, 5 min) using H-ZSM-5-30 zeolite and 0.5% Fe supported on H-ZSM-5-30 (calcined at 120 °C) starting from 0.5 M (a)
cinnamyl alcohol, (b) cinnamaldehyde and (c) dihydrocinnamaldehyde.
2 2
O
6

Y. Wang, et al.
Molecular Catalysis 474 (2019) 110409
Scheme 1. Selective oxidation of cinnamyl alcohol to cinnamaldehyde and disruptive oxidation to benzaldehyde through formation of epoxides as reaction inter-
mediates.
lower mass balance). Recently, the formation of formaldehyde was
experimentally evidenced in the gas phase isolated from batch experi-
ments with cinnamyl alcohol by Santos Costa and co-workers (2015).45
To further improve the selectivity to cinnamaldehyde using Fe based
heterogeneous catalysts, the disruptive pathway through epoxidation
needs to be suppressed. Scheme 1 shows a tentative illustration of the
selectivity. Higher activity lead in most of the cases to lower selectivity
to cinnamaldehyde as the higher activity was especially noticed in the
disruptive oxidation pathway through the formation of the epoxide
compounds through addition of oxygen radical species at the double
bound of the allylic alcohol, but also through the release of for-
maldehyde, which lead to benzaldehyde and minor amount of side
products.
2 2
cinnamyl alcohol oxidation reaction mechanism using aqueous H O , in
accordance with the results reported by Santos Costa et al. (2015), [45]
who proposed that the oxidation to cinnamaldehyde occurs on the
catalyst surface while the disruptive oxidation to benzaldehyde takes
place in bulk phase. Accordingly, they succeeded in suppressing the
disruptive oxidation by adding a stoichiometric amount of 2,6-di-tert-
butyl-4-methylphenol (HBT) as radical trap to the reaction mixture,
which strongly suggests that the disruptive oxidation to benzaldehyde
indeed is a radical mediated reaction. Similarly, when adding 1 eq. HBT
to the reaction medium using the Fe/HZSM-5-30_800C catalyst in the
present study, the conversion after 5 min. dropped from 53 to 30%,
whereas the selectivity to cinnamaldehyde increased from 26 to 39%
and for benzaldehyde decreased from 39 to 26%. This result confirms
that various reaction mechanisms take place, still it remains unclear
how effective HBT can suppress the radical mediated pathway. Our
results suggest that the supported metals (either as chloride salts or as
oxides) catalyze the formation of reactive peroxy radicals which may
act as the radical chain initiators in bulk phase (not only the cinna-
maldehyde but also the benzaldehyde yields increased upon in-
corporation of metals into the support). Whereas Santos Costa and co-
workers did not detect the formation of any epoxides in the product
mixture, [45] the present study does confirm the formation of epoxides.
Acknowledgements
The authors wish to thank the Mass Spectrometry unit of the Central
Service for Research Support of the University of Córdoba. Rafael Luque
gratefull acknowledges funding from MINECO under project CTQ2016-
78289-P, co-financed with FEDER funds. Y.W. would like to thank the
China Scholarship Council (CSC) for the financial support. Antonio
Pineda thanks Plan Propio de Investigación” from Universidad de
Córdoba (Spain) and “Programa Operativo” FEDER funds from Junta de
Andalucía. This work was performed, in partnership with the European
COST Action FP 1306 ‘Valorisation of Ligno-cellulosic Biomass Streams
for Sustainable Production of Chemicals, Materials and Fuels using Low
Environmental Impact Technologies’. The publication has been pre-
pared with support from RUDN University Program 5-100. The authors
gratefully acknowledge support from K.C. Wong Education Foundation.
Appendix A. Supplementary data
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.mcat.2019.110409.
References
Conclusions
[
1] M. Billamboz, C. Len, First pinacol coupling in emulsified water: key role of sur-
Iron supported on various aluminosilicates showed much higher
activity for the oxidation of cinnamyl alcohol using hydrogen peroxide
as the oxidant as compared to their bulk counterparts (dissolved iron
chloride), however, with lower selectivity to cinnamaldehyde.
Palladium supported on a magnetic silicate and iron supported on a
furfuryl alcohol derived resin showed higher selectivity to cinna-
maldehyde in the range of bulk iron particles, but at significant lower
conversion levels. Better results were obtained with 0.5% iron sup-
ported on MFI and FAU zeolites, which exhibited conversions between
factant and impact of alternative activation technologies, ChemSusChem 8 (2015)
1
664–1675.
[2] C. Lee, J.E. Penkala, K.G. Wunch, Process for Preventing or Mitigating Biofouling,
Patent US2015038471 (A1), February 5 (2015).
[
3] T. Scavone, M. Aviles, B. Gray, P. Ellingson, D. Duval, Absorbent Article Comprising
Complexed or Encapsulated Reactive Compounds, Patent WO2014205047 (A1),
December 24 (2014).
4] J.-E. Bäckvall, Modern Oxidation Methods, second edition, Wiley-VCH Verlag
GmbH & Co. KGaA, Weinheim, Germany, 2019.
[
[
5] B.R. McDonald, K.A. Scheidt, Pyranone natural products as inspirations for catalytic
reaction discovery and development, Acc. Chem. Res. 48 (2015) 1172–1183.
4
2 and 49% at 26–33 % selectivity to cinnamaldehyde and 26–42 %
[6] M.M. Heravi, D. Ajami, K. Tabar-Hydar, M. Ghassemzadeh, Remarkable fast mi-
crowave-assisted zeolite HZSM-5 catalyzed oxidation of alcohols with chromium
trioxide under solvent-free conditions, J. Chem. Res. Synop. 5 (1999) 334–335.
selectivity to benzaldehyde. Higher conversions (50–55 %) were ob-
tained through fast calcination at 800 °C of the iron/zeolite catalysts
instead of slow calcination at 120 °C, without any significant loss in
[
7] A. Shaabani, P. Mirzaei, D. Lee, The beneficial effect of manganese dioxide on the
oxidation of organic compounds by potassium permanganate, Catal. Lett. 97 (2004)
7

Y. Wang, et al.
Molecular Catalysis 474 (2019) 110409
separation and determination using solid-phase extraction and gas chromato-
graphy, Procedia Eng. 42 (2012) 247–260.
1
19–123.
[
[
8] R. Noyori, M. Aoki, K. Sato, Green oxidation with aqueous hydrogen peroxide,
Chem. Commun. 16 (2003) 1977–1986.
9] T. Mallat, A. Baiker, Oxidation of alcohols with molecular oxygen on solid catalysts,
Chem. Rev. 104 (2004) 3037–3058.
[28] M. Qamar, R.B. Elsayed, K.R. Alhooshani, M.I. Ahmed, D.W. Bahnemann,
Chemoselective and highly efficient conversion of aromatic alcohols into aldehydes
photo-catalyzed by Ag
3 4
PO in aqueous suspension under simulated sunlight, Catal.
[
[
[
10] I. Hermans, E.S. Spier, U. Neuenschwander, N. Turrà, A. Baiker, Selective oxidation
catalysis: opportunities and challenges, Top. Catal. 52 (2009) 1162–1174.
11] N. Zheng, G.D. Stucky, A general synthetic strategy for oxide-supported metal na-
noparticle catalysts, J. Am. Chem. Soc. 128 (2006) 14278–14280.
12] B. Karimi, S. Abedi, J.H. Clark, V. Budarin, Highly efficient aerobic oxidation of
alcohols using a recoverable catalyst: the role of mesoporous channels of SBA-15 in
stabilizing palladium nanoparticles, Angew. Chem. Int. Ed. 45 (2006) 4776–4779.
13] F. Shi, M.K. Tse, M.-M. Pohl, J. Radnik, A. Brückner, S. Zhang, M. Beller, Nano-iron
oxide-catalyzed selective oxidations of alcohols and olefins with hydrogen peroxide,
J. Mol. Catal. Chem. 292 (2008) 28–35.
Commun. 58 (2015) 34–39.
[29] R. Dileep, B.R. Bhat, T.H.S. Kumara, Palladium complex in a room temperature
ionic liquid: a convenient recyclable reagent for catalytic oxidation, Green Chem.
Lett. Rev. 7 (2014) 32–36.
[30] S.K. Sarkar, M.S. Jana, T.K. Mondal, C. Sinha, Alcohol oxidation reactions catalyzed
by ruthenium–carbonyl complexes of thioarylazoimidazoles, Appl. Organomet.
Chem. 28 (2014) 641–651.
[31] Y. Liu, A. Xie, J. Li, X. Xu, W. Dong, B. Wang, Heterocycle-substituted tetrazole
ligands for copper-catalysed aerobic oxidation of alcohols, Tetrahedron 70 (2014)
9791–9796.
[32] D. Sahu, C. Sarmah, P. Das, A highly efficient and recyclable silica-supported pal-
ladium catalyst for alcohol oxidation reaction, Tetrahedr. Lett. 55 (2014)
3422–3425.
[
[
[
[
[
14] C. González-Arellano, J.M. Campelo, D.J. Macquarrie, J.M. Marinas, A.A. Romero,
R. Luque, Efficient microwave oxidation of alcohols using low‐loaded supported
metallic iron nanoparticles, ChemSusChem 1 (2008) 746–750.
15] D.S. Mannel, S.S. Stahl, T.W. Root, Continuous flow aerobic alcohol oxidation re-
[33] Y. Kon, Y. Usui, K. Sato, Oxidation of allylic alcohols to α,β-unsaturated carbonyl
compounds with aqueous hydrogen peroxide under organic solvent-free conditions,
Chem. Commun. 0 (2007) 4399–4400.
actions using a heterogeneous Ru(OH)
2014) 1503–1508.
16] W. Song, P. Liu, E.J.M. Hensen, A mechanism of gas-phase alcohol oxidation at the
interface of Au nanoparticles and a MgCuCr spinel support, Catal. Sci. Technol.
(2014) 2997–3003.
17] M. Ojeda, A. Grau-Atienza, R. Campos, A.A. Romero, E. Serrano, J.M. Marinas,
J. García Martínez, R. Luque, Hierarchical zeolites and their catalytic performance
in selective oxidative processes, ChemSusChem 8 (2015) 1328–1333.
18] A. Sheldon, Recent advances in green catalytic oxidations of alcohols in aqueous
media, Catal. Today 247 (2015) 4–13.
x 2 3
/Al O catalyst, Org. Process Res. Dev. 18
(
[34] U.R. Pillai, E. Sahle-Demessie, Oxidation of alcohols over Fe3+/montmorillonite-
K10 using hydrogen peroxide, Appl. Catal. A 245 (2003) 103–109.
[35] D. Zhao, J. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D. Stucky,
Triblock copolymer syntheses of mesoporous silica with periodic 50 to 300
Angstrom pores, Science 279 (1998) 548-542.
2 4
O
4
[36] M. Ojeda, A.M. Balu, V. Barrón, A. Pineda, Á.G. Coleto, A.Á. Romero, R. Luque,
Solventless mechanochemical synthesis of magnetic functionalized catalytically
active mesoporous SBA-15 nanocomposites, J. Mater. Chem. A 2 (2014) 387-39.
[37] A. Pineda, A.M. Balu, J.M. Campelo, A.A. Romero, D. Carmona, F. Balas,
J. Santamaria, R. Luque, A dry milling approach for the synthesis of highly active
nanoparticles supported on porous materials, ChemSusChem 4 (2011) 1561–1565.
[38] R.L. Grob, E.F. Barry (Eds.), Modern Practice of Gas Chromatography, fourth edi-
tion, John Wiley & Sons, Inc., Hoboken, NJ, USA, 2004.
[
[
19] F. Mangin, P. Prinsen, A. Yepez, M.R.H. Gilani, G. Xu, C. Len, R. Luque, Microwave
assisted benzyl alcohol oxidation using iron particles on furfuryl alcohol derived
supports, Catal. Commun. 104 (2018) 67–70.
[
[
[
20] C. Zhu, Z. Zhang, W. Ding, J. Xie, Y. Chen, J. Wu, X. Chen, H. Ying, A mild and
highly efficient laccase-mediator system for aerobic oxidation of alcohols, Green
Chem. 16 (2014) 1131–1138.
21] N. Sanla-Ead, A. Jangchud, V. Chonhenchob, P. Suppakul, Antimicrobial Activity of
cinnamaldehyde and eugenol and their activity after incorporation into cellulo-
se‐based packaging films, Packag. Technol. Sci. 25 (2012) 7–17.
[39] P. Marturano, L. Drozdová, G.D. Pirngruber, A. Kogelbauer, R. Prins, The me-
chanism of formation of the Fe species in Fe/ZSM-5 prepared by CVD, Phys. Chem.
Chem. Phys. 3 (2001) 5585–5595.
3
+
[40] A.V. Kucherov, M. Shelef, Quantitative determination of isolated Fe
cations in
22] S.-S. Cheng, J.-Y. Liu, E.-H. Chang, S.-T. Chang, Antifungal activity of cinna-
maldehyde and eugenol congeners against wood-rot fungi, Bioresour. Technol. 99
FeHZSM-5 catalysts by ESR, J. Catal. 195 (2000) 106–112.
[41] C.M.A. Parlett, D.W. Bruce, N.S. Hondow, M.A. Newton, A.F. Lee, K. Wilson,
Mesoporous silicas as versatile supports to tune the palladium‐catalyzed selective
aerobic oxidation of allylic alcohols, ChemCatChem 5 (2013) 939–950.
[42] C.M.A. Parlett, P. Keshwalla, S.G. Wainwright, D.W. Bruce, N.S. Hondow, K. Wilson,
A.F. Lee, ACS Catal. 3 (2013) 2122–2129.
(2008) 5145–5149.
[23] R.V. Potineni, D.G. Peterson, Mechanisms of flavor release in chewing gum: cin-
namaldehyde, J. Agric. Food Chem. 56 (2008) 3260–3267.
[
24] M.P. Balaguer, J. Gómez-Estaca, R. Gavara, P. Hernandez-Munoz, Functional
properties of bioplastics made from wheat gliadins modified with cinnamaldehyde,
J. Agric. Food Chem. 59 (2011) 6689–6695.
[43] C.D. Wagner, W.N. Riggs, L.E.G. Davis, F. Moulder, G.E. Muilenberg, Perkin Elmer,
Eden Prairie, Handbook of X-Ray Photoelectron Spectroscopy, (1979).
[25] J. Zhang, L. Liu, Y. He, W. Kong, S. Huang, Cytotoxic effect of trans-cinnamalde-
hyde on human leukemia K562 cells, Acta Pharmacol. Sin. 31 (2010) 861–866.
[26] Y.C. Wong, M.Y. Ahmad-Mudzaqqir, W.A. Wan-Nurdiyana, Extraction of essential
oil from cinnamon (Cinnamomum zeylanicum), Orient. J. Chem. 30 (2014) 37–47.
[27] F. Golmohammad, M.H. Eikani, H.M. Maymandi, Cinnamon bark volatile oils
[44] X. Deng, J. Lee, D. Matranga, Preparation and characterization of Fe
3 4
O (111) na-
noparticles and thin films on Au(111), Surf. Sci. 604 (2010) 627–632.
[45] J.C. Santos Costa, P. Corio, L.M. Rossi, Catalytic oxidation of cinnamyl alcohol
using Au–Ag nanotubes investigated by surface-enhanced Raman spectroscopy,
Nanoscale 7 (2015) 8536–8543.
8