Hollow iron oxide nanoshells are active and selective catalysts for the partial oxidation of styrene with molecular oxygen

Article abstract of DOI:10.1039/c4cc04749d

Well defined hollow iron oxide nanoshells are active, selective and recyclable catalysts for the oxidation of styrene into benzaldehyde using difficult-to-activate molecular oxygen as the sole oxidant. Using no noble metals, unprecedented conversion of 90% was maintained while high selectivity of 73% was able to be achieved. This journal is

Full text of DOI:10.1039/c4cc04749d

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Hollow iron oxide nanoshells are active and
selective catalysts for the partial oxidation of
styrene with molecular oxygen†
Cite this: Chem. Commun., 2014,
0, 12482
5
Received 23rd June 2014,
Accepted 29th August 2014
Monika Joanna Rak, Michael Lerro and Audrey Moores*
DOI: 10.1039/c4cc04749d
www.rsc.org/chemcomm
1
0
11
Well defined hollow iron oxide nanoshells are active, selective and More recently, Pt@Fe O3 and gold nanowires also demon-
2
recyclable catalysts for the oxidation of styrene into benzaldehyde strated oxidative activity under mild conditions. These catalysts
using difficult-to-activate molecular oxygen as the sole oxidant. Using feature good selectivity for benzaldehyde (94%) with poor yields
no noble metals, unprecedented conversion of 90% was maintained (o10%). When the conversions are pushed, selectivity drops
while high selectivity of 73% was able to be achieved.
dramatically. These heterogeneous catalysts unfortunately suffered
from poor recyclability with both activity and selectivity falling
Outdone only by vanillin (vanilla aroma), benzaldehyde (almond upon reuse of the catalyst.
1
aroma) is the second most used flavouring agent. In 2009, it was
Iron oxide, meanwhile, has long demonstrated a high efficiency
2
12–15
estimated that 90 kilotons of benzaldehyde were synthesized as an effective catalyst for a variety of oxidation reactions
through industrial processes hampered by high temperatures and because it is cheap, nontoxic, and earth abundant, making it an
3
pressures that nevertheless resulted in lackluster yields. One attractive material in the context of green chemistry. Furthermore,
industrial pathway to benzaldehyde is the hydrolysis of benzal when in the form of NPs, iron oxide can be recovered magneti-
1
6,17
chloride synthesis which generates large quantities of HCl as a cally.
Most examples are concerned with the use of hydrogen
4
12,15
by-product. The alternative, and more popular industry method peroxide to oxidize aryl double bonds and alcohols.
Notably,
of air-oxidation of toluene, results in low conversions of starting Beller and coll. demonstrated that benzaldehyde could be obtained
materials and produces benzaldehyde only as a by-product in the with high conversion and selectivity from benzyl alcohol; aryl
3
,5
production of the less valuable benzoic acid.
alkenes were also converted with high selectivities although with
12
Due to the inherent advantages of using molecular oxygen being modest conversions (o50%). More recently, Luque and coworkers
cheap, readily available, and ensuring a high atom economy, there used mesoporous silica supported iron oxide to convert styrene to
has recently been an increased interest in the oxidation of styrene benzaldehyde with high yield with H
using molecular oxygen to produce benzaldehyde. Despite these
18
2
O
2
as the oxidant.
6
We decided to study the activity of iron oxide nanoparticles,
advantages, it remains a challenge to use molecular oxygen effec- without the aid of noble metals, for the selective oxidation of
tively for partial oxidations of olefins to aldehydes while limiting styrene to benzaldehyde using molecular oxygen as the sole
the production of carboxylic acids. Work done using homogeneous oxidant (Scheme 1). Previous work has shown that a decrease in
noble metal-based systems has shown some success for this task. size of iron NPs results in an increased conversion but a decreased
Feng et al. demonstrated that a Pd(II) complex allowed excellent
7
conversions and selectivities as well as good recyclability. On the
other hand, heterogeneous supported gold nanoparticle (AuNP)
8
catalysts have garnered a lot of attention for oxidation reactions
and in 2008, Lambert and coworkers showed that titania supported
Au55NP were able to effect conversions of styrene to benzaldehyde
9
under particularly mild conditions using molecular oxygen.
Centre for Green Chemistry and Catalysis, Department of Chemistry,
McGill University, 801 Sherbrooke St West, Montr ´e al, QC H3A 0B8, Canada.
E-mail: audrey.moores@mcgill.ca; Fax: +1 (514) 398 3797; Tel: +1 (514) 398 4654
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
Scheme 1 Styrene oxidation using molecular oxygen as oxidant. Reac-
tion conditions: 0.5 mol% catalyst, 4 bar O , 24 h, acetonitrile.
c4cc04749d
2
1
2482 | Chem. Commun., 2014, 50, 12482--12485
This journal is ©The Royal Society of Chemistry 2014

View Article Online
Communication
ChemComm
a
Table 1 Styrene oxidation: catalyst screening
b
b
b,c
Catalyst
Conv. (%)
Benzaldehyde
Benzoic acid
Other
Fe
Fe
2
O
O
3
4
NP
NP
49
53
56
84
90
81
78
81
65
73
11
8
7
22
13
8
14
12
13
14
3
Fe/FeOC/S
Fe
eFe
3
O
4
NS
NS
Scheme 2 Catalysts tested in this work.
3
O
4
a
2
Reaction conditions: 0.5 mol% catalyst, 4 bar O , 24 h, 90 1C, 1.8 M
12,13
b
c
acetonitrile. Selectivity%. Sum of acetophenone, styrene oxide, and
oligomers.
selectivity.
We chose to investigate the effect of an interesting
feature of NPs, namely their ability to produce hollow nanoshells,
on the conversion and selectivity for oxidation of styrene to
benzaldehyde. Although several syntheses exist for making hollow 49 and 56% in all cases. Interestingly, Fe O NS afforded a much
3
4
19
iron oxide nanoparticle structures, their application in catalysis improved conversion of 84%, while the selectivity was eroded but
has been little studied. Hollow iron oxide structures for oxidations maintained at a high level of 65%. These results clearly evidenced
have been attempted in 2011, by Zhang et al. who studied a-Fe O
3 4
that Fe O NS are more active for the reaction of oxidation and the
2
3
nanotubes in a photocatalytic oxidation of naphthalene to produce slight drop in selectivity compared to non-hollow counterparts
20
CO
compared hollow ferrite nanoshells (Fe
Fe NS, to solid ferrite and ferrous oxide NPs and reduced iron reaction neat or in water resulted in styrene polymerizing.
NPs for the partial oxidation of styrene (Scheme 2). The results Toluene showed similar trends to acetonitrile, but with reduced
presented herein show that Fe NS and their etched counterparts conversions (Table S1, ESI†). Temperature optimization reac-
2
, CO, H
2
O, and phthalic anhydride. In the present work, we could be attributed to conversion to benzoic acid.
3 4
O
NS), as well as etched
Various conditions were tested with Fe O NS. Running the
3
4
3 4
O
3 4
O
successfully combined high selectivity to high activity for the tions were attempted with this solvent at two pressures; one at
very attractive synthesis of benzaldehyde from styrene and O2. 4 bar (Table S2, ESI†) and one at 30 bar (Table S3, ESI†) of O in
2
Fe
and used as received. Despite much research being done on the on the low end gave no reaction and on the high end resulted
synthesis of Fe–Fe core–shell NP (Fe/FeOC/S) through the in polymerization. Increasing the pressure resulted in the
reduction of Fe(SO)
using thermal decomposition of Fe(CO)
comparison with the nanoshell catalysts which were made NS, the complete characterization of this nanomaterial was
2 3 3 4
O and Fe O NPs were purchased from Sigma Aldrich toluene. The reaction was book-ended by temperatures which
x y
O
21,22
4
in water,
our Fe/FeOC/S NPs were made undesirable production of more benzoic acid.
to allow for clearer
In order to better understand the unique reactivity of Fe O4
5
3
2
3
using the Fe/FeOC/S NPs. Iron oxide nanoshells (Fe O NS) undertaken. Fe O NS are obtained through rapid oxidation of
3
4
3 4
were synthesised through oxidation of Fe(0) NPs by air at Fe(0) NPs obtained by thermal decomposition of Fe(CO)5.
60 1C, following a method reported by Sun and co-workers Transmission electron micrographs (TEM) reveal that the Fe O4
2
3
2
3
(Fig. 1). In a typical reaction, 5 mg of the catalyst were loaded NS are very monodisperse with a size of 16.7 + 2.0 nm and a shell
with 4.75 g of styrene in acetonitrile (1.8 M solution) at 90 1C of 3.9 + 0.5 nm of thickness. These results are consistent with
23
with 4 bar of O
2
for 24 hours (Table 1).‡ This reaction afforded previous reports. Importantly, the interior hole of the Fe O NS
3 4
4
products: benzaldehyde, benzoic acid, acetophenone and is clearly apparent from the images. Some NSs do feature at their
styrene oxide (Scheme 1). The catalysts were probed for their centre a small bead of remaining Fe(0) NP (Fig. 1, left and Fig. S1,
conversion and selectivity towards benzaldehyde formation. ESI†). The hollow structure is created by Kirkendall effect which
The latter criterion was high (78 to 81%) for Fe O , Fe O NPs forms voids that coalesce in the center of the nanoshell. ICP-MS
2
3
3 4
and Fe/FeOC/S, but associated with moderate conversions between measurements showed that the heated solution contained
less than 0.01 ppm of dissolved iron, thereby indicating hetero-
genous catalysis which may be attributed to the surface. To
understand whether the catalytic activity is occurring on the
exterior or interior surface of the NSs, Brunauer–Emmett–Teller
(BET) measurements were performed to determine the presence
and potential size of any pores (Fig. S2, ESI†). Surface area
measurements show that iron oxide nanoshells to have a specific
2
À1
area of 140 m g . This value is too high to correspond solely to
the exterior surface and corresponds to the calculated area of
both the exterior and interior surface. This indicates that hollow
iron oxide NSs possess a unique chemistry affecting the catalytic
properties. The trend of increased catalytic activity for hollow and
porous nanoshells relative to solid nanoparticles, has also been
2
4
reported for gold nanocages, nanoboxes, nanoparticles.
In order to determine how exposing the interior of the iron
NS modifies the catalytic behaviour, the NS were etched to
3 4 3 4
Fig. 1 TEM images of Fe O NS (left) and eFe O NS (right).
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 12482--12485 | 12483

View Article Online
ChemComm
Communication
The cleaved off carbon formed the highly volatile and polymer-
izable formaldehyde which at lower temperatures could be
found as polyacetal solid at the bottom of the reactor. FTIR
tests of this undissolvable material revealed a clean spectrum of
polyacetal (Fig. S3, ESI†).
3 4
eFe O NS magnetic properties were exploited to improve
the ease and efficiency of removing the catalyst from solution.
Discontinuing the stirring allows the NS in solution to settle
onto the magnetic stirring bar allowing for the separation of the
catalyst from the liquid mixture and its recycling. Throughout 4
consecutive runs, conversion remained high while selectivity
dropped off and plateaued. Interestingly, in comparable
Fig. 2 Conversions (clustered columns) and benzaldehyde selectivity systems usually both selectivity and activity dropped upon
9
–11
(
lines) as a function of temperatures with Fe
eFe NS (red, dashed) as catalysts. Reaction conditions: 0.5 mol%
catalyst, 4 bar O , 24 h, 1.8 M acetonitrile.
3 4
O NS (blue, solid) and
recycling (Fig. 3).
1
3 4
TEM imaging of the eFe O NS after
3 4
O
and 2 recycle runs revealed they undergo slow dislocation,
possibly explaining the loss of selectivity after recycling (Fig. S5,
2
ESI†). Also oleylamine may play an active role in selectivity
5
as shown by others. A recycling catalytic test was run with
afford eFe O NS through a process delineated by Sun (Fig. S1,
3
4
2
5,26
eFe O NS and additional oleylamine, showing that a part of the
ESI†).
This material has been used as a drug delivery
3 4
selectivity could be recovered (Table S7, ESI†). Addition
of oleylamine to Fe O NPs, however, could not afford the
vehicle. It featured TEM-visible pores (Fig. 1, right), which
may allow a styrene molecule to enter the interior of the
3
4
superior activity and selectivity obtained with hollow structures
presented herein (Table S7, ESI†).
3 4 3 4
catalyst. Fe O NS and eFe O NS were compared in optimized
conditions, at 4 bar, in acetonitrile and for a range of tempera-
tures from 80 to 100 1C (Table 1 and Tables S4 and S5, ESI†).
Fig. 2 pictures well the trend of decreased selectivity with
increased temperature and increased conversion. The etched
In sum, we showed that the porous nature of the hollow iron
oxide nanoparticles allowed for a change in catalytic properties,
allowing the previously difficult oxidation of styrene with
molecular oxygen. Further compelling evidence for the use of
hollow iron oxide nanoparticles as a catalyst is that the oxida-
tion process can be controlled to stop at the aldehyde, giving
benzaldehyde as the major product.
3 4
particles eFe O NSs have marginally better selectivity and
significantly better conversions (+10–20 point%) over the range
of temperatures. The optimal reaction set is obtained for 90 1C
where conversion of 90% are combined to a high selectivity of
We thank the Natural Science and Engineering Research
Council of Canada (NSERC) Discovery Grant program, the
Canada Foundation for Innovation (CFI), the Canada Research
Chairs (CRC), the Fonds de Recherche du Qu ´e bec – Nature et
Technologies (FRQNT) Equipe program, the Centre for Green
Chemistry and Catalysis (CGCC), NSERC-Collaborative Research
and Training Experience (CREATE) in Green Chemistry and
McGill University for their financial support.
7
3% for the desired aldehyde product (TON§ = 131.4). This
result is unprecedented in terms of selectivity and activity
combined for this reaction with O . Preliminary kinetic data
suggested that eFe O NS were able to sustain high selectivities
over time in a unique fashion (Fig. S4, ESI†). Importantly, these
results were obtained in absence of non-noble metal (for compar-
ison see Table S6, ESI†).
As the product has one fewer carbons than the starting
material, the fate of the missing carbon was investigated.
2
3
4
Notes and references
‡
Octane and dodecane were both used as internal standards to
calculate conversion and selectivity percentages via GC-FID.
TON = number of moles of aldehyde produced per mole of catalyst.
13
§
1
2
3
4
U. Krings and R. G. Berger, Appl. Microbiol. Biotechnol., 1998, 49, 1–8.
S. Hughes, in Eur. Chem. Market, 2010, vol. 44.
J. A. B. Satrio and L. K. Doraiswamy, Chem. Eng. J., 2001, 82, 43–56.
F. Br u¨ hne and E. Wright, Ullmann’s Encyclopedia of Industrial
Chemistry, Wiley-VCH Verlag GmbH & Co. KGaA, 2000, DOI:
1
0.1002/14356007.a03_463.
J. Lv, Y. Shen, L. Peng, X. Guo and W. Ding, Chem. Commun., 2010,
6, 5909–5911.
5
6
7
4
K. Schr ¨o der, B. Join, A. J. Amali, K. Junge, X. Ribas, M. Costas and
M. Beller, Angew. Chem., Int. Ed., 2011, 50, 1425–1429.
B. Feng, Z. S. Hou, X. R. Wang, Y. Hu, H. Li and Y. X. Qiao, Green
Chem., 2009, 11, 1446–1452.
Fig. 3 Conversions (clustered columns) and benzaldehyde selectivity
8
9
M. Haruta, Nature, 2005, 437, 1098–1099.
(
lines) as a function of the number of recycle runs with eFe
catalysts. Reaction conditions: 0.5 mol% eFe NS catalyst, 4 bar O
4 h, 90 1C, 1.8 M acetonitrile.
3 4
O NS as
M. Turner, V. B. Golovko, O. P. H. Vaughan, P. Abdulkin,
A. Berenguer-Murcia, M. S. Tikhov, B. F. G. Johnson and
R. M. Lambert, Nature, 2008, 454, 981–983.
3
O
4
2
,
2
1
2484 | Chem. Commun., 2014, 50, 12482--12485
This journal is ©The Royal Society of Chemistry 2014

View Article Online
Communication
ChemComm
1
1
1
1
1
1
0 H. Y. Hong, L. Hu, M. Li, J. W. Zheng, X. H. Sun, X. H. Lu, X. Q. Cao, 18 F. Rajabi, N. Karimi, M. R. Saidi, A. Primo, R. S. Varma and
J. M. Lu and H. W. Gu, Chem. – Eur. J., 2011, 17, 8726–8730. R. Luque, Adv. Synth. Catal., 2012, 354, 1707–1711.
1 L. Hu, X. Cao, J. Yang, M. Li, H. Hong, Q. Xu, J. Ge, L. Wang, J. Lu, 19 K. Cheng, S. Peng, C. Xu and S. Sun, J. Am. Chem. Soc., 2009, 131,
L. Chen and H. Gu, Chem. Commun., 2011, 47, 1303–1305. 10637–10644.
2 F. Shi, M. K. Tse, M. M. Pohl, J. Radnik, A. Bruckner, S. M. Zhang 20 W. H. Zhang, X. Y. Li, Q. D. Zhao, Y. Hou, Y. Shen and G. H. Chen,
and M. Beller, J. Mol. Catal. A: Chem., 2008, 292, 28–35. Mater. Chem. Phys., 2011, 129, 683–687.
3 F. Shi, M. K. Tse, M. M. Pohl, A. Bruckner, S. M. Zhang and 21 C. M. Cirtiu, T. Raychoudhuty, S. Ghoshal and A. Moores, Colloids
M. Beller, Angew. Chem., Int. Ed., 2007, 46, 8866–8868. Surf., A, 2011, 390, 95–104.
4 K. Schr ¨o der, K. Junge, B. Bitterlich and M. Beller, Top. Organomet. 22 R. Hudson, A. Riviere, C. M. Cirtiu, K. L. Luska and A. Moores,
Chem., 2011, 33, 83–109. Chem. Commun., 2012, 48, 3360–3362.
5 C. Gonz ´a lez-Arellano, J. M. Campelo, D. J. Macquarrie, J. M. Marinas, 23 S. Peng and S. Sun, Angew. Chem., Int. Ed., 2007, 46, 4155–4158.
A. A. Romero and R. Luque, ChemSusChem, 2008, 1, 746–750.
6 V. Polshettiwar and R. S. Varma, Green Chem., 2010, 12, 743–754.
7 R. Hudson, Y. Feng, R. S. Varma and A. Moores, Green Chem., 2014,
DOI: 10.1039/C4GC00418C.
24 J. Zeng, Q. Zhang, J. Chen and Y. Xia, Nano Lett., 2009, 10, 30–35.
25 K. Cheng, S. Peng, C. J. Xu and S. H. Sun, J. Am. Chem. Soc., 2009,
131, 10637–10644.
1
1
26 K. Cheng and S. H. Sun, Nano Today, 2010, 5, 183–196.
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 12482--12485 | 12485