Peroxidative oxidation of cyclic and linear hexane catalysed by supported iron complexes under mild and sustainable conditions

Article abstract of DOI:10.1016/j.molcata.2013.10.020

The development of economical heterogeneous catalysts based on first row transition metal complexes that work under mild and sustainable reaction conditions in the activation of CH bonds is very important. Herein, commercially available iron(II) and iron(III) acetylacetonate were immobilised onto amine functionalised economical and environmentally acceptable solid porous supports: hexagonal mesoporous silica and activated carbon. The materials prepared by this very simple and straightforward methodology were characterised by elemental analysis, ICP-AES, FTIR, isotherms of adsorption at 77 K and thermogravimetry. It showed that both porous materials were conveniently functionalised with amine groups and that the iron metal complexes could be effectively anchored onto these materials. The immobilised iron salts were active as heterogeneous catalysts in the oxidation of cyclohexane and n-hexane at room temperature with hydrogen peroxide, giving the respective alcohols and ketones. They could also be recycled at least two times without loss of catalytic activity in the oxidation of cyclohexane. In the oxidation of n-hexane this was only true for the anchored Fe(II) salts. In the oxidation of cyclohexane, both iron salts are more active heterogeneous catalysts anchored onto the activated carbons than onto the hexagonal mesoporous silicas. The opposite is observed in the oxidation of n-hexane. The heterogeneous catalysts reported herein are economical and work under mild reaction conditions and thus could be valuable for the improvement of the sustainability and environmental impact of processes currently used in industry.

Full text of DOI:10.1016/j.molcata.2013.10.020

Journal of Molecular Catalysis A: Chemical 381 (2014) 171–178
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Peroxidative oxidation of cyclic and linear hexane catalysed by
supported iron complexes under mild and sustainable conditions
∗
Ana Rosa Silva , João Botelho
CICECO, Departamento de Química, Universidade de Aveiro, 3810-193 Aveiro, Portugal
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 13 August 2013
Received in revised form 18 October 2013
Accepted 21 October 2013
Available online 29 October 2013
The development of economical heterogeneous catalysts based on first row transition metal complexes
that work under mild and sustainable reaction conditions in the activation of C H bonds is very important.
Herein, commercially available iron(II) and iron(III) acetylacetonate were immobilised onto amine func-
tionalised economical and environmentally acceptable solid porous supports: hexagonal mesoporous
silica and activated carbon. The materials prepared by this very simple and straightforward methodology
were characterised by elemental analysis, ICP-AES, FTIR, isotherms of adsorption at 77 K and thermo-
gravimetry. It showed that both porous materials were conveniently functionalised with amine groups
and that the iron metal complexes could be effectively anchored onto these materials. The immobilised
iron salts were active as heterogeneous catalysts in the oxidation of cyclohexane and n-hexane at room
temperature with hydrogen peroxide, giving the respective alcohols and ketones. They could also be
recycled at least two times without loss of catalytic activity in the oxidation of cyclohexane. In the oxi-
dation of n-hexane this was only true for the anchored Fe(II) salts. In the oxidation of cyclohexane, both
iron salts are more active heterogeneous catalysts anchored onto the activated carbons than onto the
hexagonal mesoporous silicas. The opposite is observed in the oxidation of n-hexane. The heterogeneous
catalysts reported herein are economical and work under mild reaction conditions and thus could be
valuable for the improvement of the sustainability and environmental impact of processes currently
used in industry.
Keywords:
Oxidation reactions
Iron catalysis
Hydrogen peroxide
Heterogeneous catalysis
Schiff condensation
©
2013 Elsevier B.V. All rights reserved.
1
. Introduction
the substrate, using green oxidants, like oxygen or hydrogen perox-
ide (H O ), non-toxic solvents, like acetonitrile, and highly active
2
2
The main oxidised products of cyclohexane are of great
and reusable catalyst, in mild conditions. H O2 simply produces
2
industrial significance. A large amount of cyclohexanol and cyclo-
hexanone is produced worldwide, mostly to be used as precursors
to the manufacture of added-value products, such as Nylon-6 and
Nylon-6,6 [1,2]. Industrially, the reaction is performed using a
cobalt(II) naphthenate catalyst at 160 C and 15 bar, yielding 4%
conversion into oxidised products with 80% selectivity towards
cyclohexanol and cyclohexanone [2,3].
Terminally oxidised hydrocarbons are also interesting feed-
stocks for the chemical and pharmaceutical industry. The
regioselective oxidation of only the terminal methyl groups in
alkanes is however a challenging task [4,5]. The synthesis of these
functionalised compounds by C H oxidation requires severe con-
ditions and non-toxic oxidants and metals. An ideal method would
yield only the desired oxidised products, with 100% conversion of
water as by-product and the controlled partial oxidation is easier
to achieve than with molecular oxygen and air [4].
Enzymes are the state-of-the-art in alkane oxidation.
Cytochrome P-450 and many other metallo-enzymes are capable
of performing the hydroxylation of alkanes [6]. Therefore, in the
past years there has been an interest in modelling such enzyme
active sites and manufacture catalysts that mimic this oxidation.
Catalysis by transition-metal complexes is however a favourable
route to the selective functionalisation of alkanes under mild con-
ditions, since they are cheap and easy to produce, particularly the
Schiff base ones, and enzymes are expensive and easily deactivated
[6].
The active centre of Cytochrome P-450 is iron which forms com-
plexes with molecular oxygen in haemoglobin and myoglobin [5].
Hence iron-containing catalysts are of high potential interest, espe-
cially because iron is cheap, widespread in nature and essential for
the human body. Chemoselectivity plays however a major role in
these reactions, thanks to potential non-selective Fenton or Gif-
type chemistry [2]. We have found that Fe(III) complexes with
N2O2 Schiff base complexes were the most efficient homogeneous
◦
∗ Corresponding author at: CICECO, Departamento de Química, Universidade de
Aveiro, Campus Universitário de Santiago, 3810-193 Aveiro, Portugal.
Tel.: +351 234 370 604; fax: +351 234 401 470.
E-mail address: ana.rosa.silva@ua.pt (A.R. Silva).
1
381-1169/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.molcata.2013.10.020

1
72
A.R. Silva, J. Botelho / Journal of Molecular Catalysis A: Chemical 381 (2014) 171–178
catalysts, among other studied first-row transition metal com-
2.2. Synthesis of hexagonal mesoporous silica (HMS)
plexes, in the oxidation of cyclohexane and n-hexane using H O
2
2
at room temperature [7]. As previously reported [8] nitric acid (in
0 mol%) was found to have a promoting effect in the reaction
7].
Heterogeneous catalysts however prevent the obstacles that
commonly occur in homogeneously catalysed systems. Anchoring
a homogeneous catalyst onto solid supports makes the separation
from the reaction media easy, preventing therefore the use of labo-
rious and inefficient extraction processes, thus allowing the reuse
of the catalyst in more catalytic cycles [9–12]. The high selectivity
of transition metal homogeneous catalysts is hence combined with
the heterogeneous catalysts advantages.
The HMS was synthesised according to procedures described
in the literature [9,13]; tetraethoxysilane (37.0 mL, 0.166 mol) was
added to a stirred solution of ethanol (88.1 mL, 1.51 mol), water
(88.5 mL, 4.91 mol) and 1-octylamine (7.4 mL, 0.0448 mol). The
mixture was stirred at room temperature for 24 h. The obtained
precipitate was vacuum filtered, washed with deionised water
(100 mL) and ethanol (100 mL). In order to remove the template
1
[
◦
(1-octylamine), the precipitate was calcined at 600 C for 24 h.
HMS: elemental analysis (%) C 0.81 H 0.00 N 0.02.
AC: elemental analysis (%) C 88.68 H 0.24 N 0.47.
In terms of industrial application the use of economical
porous supports is also very important. Relatively to conventional
mesoporous silicas prepared by electrostatic assembly pathways,
hexagonal mesoporous silicas (HMS) are prepared by using eco-
nomic alkylamines as templates [13]. Consequently, they present
more extensively cross-linked frameworks, besides of thicker
framework walls, which contribute to superior thermal stability
upon calcination in air. HMS also possesses smaller particles sizes,
alongwith defects in channelpacking, which yields complementary
textural mesoporosity providing better access to the framework-
confined mesopores in liquid phase catalysis [14]. Their surface
silanol groups can be functionalised using derivatizing agents, such
as (3-aminopropyl)triethoxysilane (APTES), allowing the chemical
attachment of the homogeneous catalyst through Schiff condensa-
tion of the free amine group [9,10].
2
.3. Functionalisation of the HMS or AC with APTES
The functionalisation of both porous materials was performed
according to procedures described in the literature [9–12]; a mix-
ture of calcined HMS or dried AC (5.0 g) in dry toluene (50.0 mL) and
(3-aminopropyl)triethoxysilane (15.0 mmol, 3.5 mL) was refluxed
for 24 h (pH of the APTES in toluene solution was 10). The mate-
rial was vacuum filtered, washed with toluene (3× 50 mL) and
refluxed in toluene for 6 h. The material was dried overnight in an
oven at 100 C. These materials will be referred as APTES@HMS or
APTES@AC.
◦
APTES@HMS: elemental analysis (%) C 7.52 H 1.45 N 1.93, loading
APTES 1.38 mmol g
−1
.
Activated carbons (AC) are porous carbon materials that possess
several advantages over the inorganic porous solids, such as
hydrophobic nature of their surfaces, high specific surface area,
large pore volumes, chemical inertness and good mechanical sta-
bility [15]. Its tunable surface chemistry rich in oxygen functional
groups also allows for wider strategies of homogeneous catalyst
anchoring than the inorganic solids [16].
APTES@AC: elemental analysis (%) C 85.77 H 1.66 N 1.19, load-
−1
ing APTES 0.85 mmol g
and ICP-AES Si 2.27%, loading APTES
−1
0
.81 mmol g
.
2.4. Anchoring of iron(III) acetylacetonate onto
amine-functionalised HMS or AC
In this work, we anchored [Fe(acac)n] (n = 2, 3) complexes onto
solid supports (AC and HMS) functionalised with APTES, through
Schiff condensation between the carbonyl groups of the acetylace-
tonate ligands and the APTES amine. These materials were used as
heterogeneous catalysts for the oxidation of cyclohexane and n-
hexane under mild conditions using hydrogen peroxide as oxidant
and nitric acid as co-oxidant.
The HMS or AC functionalised with APTES (2.0 g) was added
to 100 mL of a solution of [Fe(acac) ] (0.352 g, 996 ␮mol) in
3
dichloromethane, and the mixture was refluxed for 24 h. The
resulting material was extensively washed with dichloromethane
and then refluxed with dichloromethane for 6 h and dried
◦
overnight in an oven at 60 C. These materials will be referred as
[
Fe(acac) ]APTES@HMS or [Fe(acac) ]APTES@AC.
3
3
[
1
Fe(acac) ]APTES@HMS: elemental analysis (%) C 9.49 H 2.04 N
3
2
. Experimental
−
1
.84; ICP-AES Fe 1.2%, loading Fe 215 ␮mol g
.
[
1
Fe(acac) ]APTES@AC: elemental analysis (%) C 85.62 H 1.46 N
3
2
.1. Materials and solvents
−
1
.19, loading APTES 0.85 mmol g ; ICP-AES (%) Fe 0.18 Si 2.11,
−
1
−1
loading Fe 32 ␮mol g and APTES 0.75 mmol g
.
The reagents were used as received; tetraethoxysilane, 1-
octylamine, (3-aminopropyl)triethoxysilane (APTES), dry toluene,
iron(II) acetylacetonate, iron(III) acetylacetonate, cyclohexane,
chlorobenzene, hydrogen peroxide 30 wt% in water, nitric acid
and triphenylphosphine were purchased from Sigma–Aldrich.
Dichloromethane and acetonitrile were HPLC grade and from
Romil company. n-Hexane was from Fisher Scientific. For the
FTIR potassium bromide was used spectroscopic grade and from
Sigma–Aldrich.
2
.5. Anchoring of iron(II) acetylacetonate onto
amine-functionalised HMS or AC
A solution of iron(II) acetylacetonate (0.127 g. 498 ␮mol) in
dichloromethane (100 mL) was refluxed with amine-functionalised
HMS or AC (1.0 g) for 24 h. The solid was vacuum filtered,
washed with dichloromethane (3× 20 mL), and then refluxed
with dichloromethane for 6 h. Finally, the solid was dried
The starting carbon material was a NORIT ROX 0.8 activated
carbon (rodlike pellets with 0.8 mm diameter and 5 mm length).
◦
overnight in an oven at 60 C. These materials will be referred as
[Fe(acac)2]APTES@HMS or [Fe(acac)2]APTES@AC.
3
−1
This material has a pore volume of 0.695 cm g , as determined
by porosimetry (corresponding to meso- and macropores), an ash
content of 2.6% (w/w), an iodine number of 1000 and mercury and
[
Fe(acac) ]APTES@HMS: elemental analysis (%) C 9.74 H 2.60 N
2
−
1
−
3
2.86; ICP-AES Fe 1.38%, loading Fe 247 ␮mol g
[
1
.
helium densities of 0.666 and 2.11 g cm , respectively. The acti-
vated carbon was purified by Soxhlet extraction with 2 M HCl for
Fe(acac) ]APTES@AC: elemental analysis (%) C 82.21 H 0.96 N
2
−
1
.32, loading APTES 0.94 mmol g ; ICP-AES (%) Fe 0.77 Si 2.38,
6
h, washed with deionised water until pH 6–7 and then dried in
−
1
−1
.
◦
loading Fe 138 ␮mol g and APTES 0.85 mmol g
an oven at 150 C for 13 h under vacuum.

A.R. Silva, J. Botelho / Journal of Molecular Catalysis A: Chemical 381 (2014) 171–178
173
2.6. Catalysis experiments
3. Results and discussion
The oxidation of cyclohexane and n-hexane was carried out
In order to effectively immobilise the iron acetylacetonate
salts onto the hexagonal mesoporous silica (HMS) or activated
carbon (AC) surfaces they were firstly functionalised using (3-
aminopropyl)triethoxysilane (APTES); according to Scheme 1 the
anchoring occurs by condensation between the free amine groups
of (3-aminopropyl)triethoxysilane (APTES) and the ligand oxygen
atoms. This procedure has already resulted in efficient and reusable
heterogeneous catalysts for several oxidation reactions using these
porous materials and first-row transition metal acetylacetonates
[9–12].
in acetonitrile (ACN), at room temperature, with constant stir-
ring. The composition of the reaction medium was 0.0125 mmol
of heterogeneous catalyst, 1.25 mmol of substrate (cyclohexane
or n-hexane) and 0.125 mmol of nitric acid in 5.0 mL of ACN.
The oxidant, hydrogen peroxide (6.25 mmol), was progressively
added to the reaction medium using a syringe pump at a rate of
1
.28 mL/h. The molar ratio of metal catalyst/alkane/HNO /H O
3 2 2
was 1:100:10:500 [7]. Aliquots (100.0 ␮L) were periodically with-
drawn from the reaction mixture, the alkylhydroperoxides were
reduced with excess of triphenylphosphine [7] and filtered through
0
.2 ␮m PTFE syringe filters and 20.0 ␮L of chlorobenzene (external
−
3
3.1. Characterisation of the porous materials and amine
functionalisation
standard; 0.272 mol dm
in acetonitrile) was added. Cyclohex-
anol and cyclohexanone were identified by comparison of their
retention times with commercially available samples and their
quantification by interpolation into multipoint calibration curves.
Other oxygenated products were identified by a Finnigan Trace
GC–MS. It is noteworthy that no alkane oxidation products were
detected in control experiments using the same experimental
conditions, but without addition of homogeneous catalyst. The cat-
alysts were then filtered and washed sequentially with 100 mL of
3.1.1. HMS
The nitrogen adsorption isotherm at 77 K of the calcined HMS
material is of type IV (Fig. 1A) and therefore is typical of mesoporous
2
−1
materials [17]. The material has a BET surface area of 829 m g
3
−1
and a total pore volume of 0.973 cm g (Table 1). The nitrogen
adsorption–desorption isotherm at 77 K is similar to the one origi-
nally reported by Pinnavaia and co-worker for the HMS material
prepared using the 1-octylamine as surfactant [14]; the well-
defined hysteresis loop between p/p0 of 0.5 and 1.0 in the nitrogen
adsorption–desorption, clearly indicates the presence of comple-
mentary textural mesoporosity arising from the non crystalline
intra-aggregates voids and spaces formed by the inter-particle con-
tacts [14]. This is due to the HMS small particles sizes and better
access to the framework-confined mesopores in liquid phase catal-
ysis is thus provided by this textural mesopority [14].
◦
ACN for 3 h, and dried overnight in an oven at 60 C, under vac-
uum. The catalysts were then reused using the same experimental
conditions. Control experiments in homogeneous phase using the
iron acetylacetonate salts in identical reaction conditions were also
performed.
2.7. Physical measurements
Elemental analysis was performed in duplicate by “Servi-
The FTIR spectrum of the calcined HMS shows the typical Si
lattice vibrations: two broad bands between 2100 and 1750 cm
O
cio de Análisis Instrumental”, CACTI Vigo, Universidade de Vigo,
Spain. The iron ICP-AES was performed at “Laboratório Central
de Análises” of the University of Aveiro, Portugal. FTIR spectra
were obtained as KBr pellets in a FT Mattson 7000 galaxy series in
the range 400–4000 cm . Thermogravimetry (TG) was performed
under air flux, except for the carbon materials where a nitrogen flux
−1
,
with medium intensity, a strong and broad band with two peaks
−1
in the region 1450–900 cm , and two strong bands between 900
−1
−1
,
and 500 cm (Fig. 2A.a). A very broad band centred at 3450 cm
−1
due to the O H stretching vibrations can be observed, as well as
−1
the H
Fig. 2A.a).
Elemental analysis of the APTES@HMS material shows an
O H bending vibrations of physisorbed water at 1622 cm
◦
was used, with a ramp of 5 C/min in a TG apparatus, model TGA-50
(
◦
Shimadzu. Nitrogen adsorption isotherms at −196 C were mea-
sured in an automatic apparatus (Gemini; Micromeritics). Before
the adsorption experiments the samples were outgassed under
increase in the material nitrogen content (Sections 2.2 and 2.3), cor-
−1
responding to 1.38 mmol g of APTES anchored onto the surface
◦
vacuum during 2.5 h at 150 C.
of the HMS material.
Gas chromatography experiments (GC) were performed with a
Varian 450 GC chromatograph equipped with a FID detector and
using helium as carrier gas and a fused silica Varian Chrompack
capillary column VF-5ms (15 m, 0.15 mm id; 15 mm film thickness).
Upon functionalisation of the surface of the HMS material with
APTES there are no changes in the nitrogen adsorption isotherm
indicating that the structure of the material is maintained (Fig. 1A).
But as expected there is a decrease in the surface area (Table 1).
Hence comparatively to the original calcined HMS, the FTIR
◦
◦
The chromatographic conditions were: 60 C (3 min), 5 C/min,
◦ ◦ ◦
1
2
70 C (2 min), 20 C/min, 200 C (5 min); injector temperature,
spectrum of the APTES@HMS material also shows N H and C
H
◦
◦
30 C; detector temperature, 250 C. After reduction with triph-
−1
stretching vibrations of APTES in the region 3400–3000 cm and
enylphosphine (PPh ) of 150 ␮L aliquots they were injected directly
−1
3
around 2950 cm , respectively. The corresponding H
N H and
into the injector using a 1 ␮L Hamilton syringe.
H
C
H symmetric bending vibrations can also be observed in
CH3
H C
3
O
(
i) [Fe(acac) ], CH Cl , reflux, 24 h
O
(
i) APTES, toluene, reflux (N ), 24 h
2
2
2
2
Fe
N
OH
OH
(
ii) toluene, reflux, 6 h
O
(ii) CH Cl , reflux, 6 h
2
2
O
O
O Si
NH2
O Si
CH3
OEt
OEt
H C
3
[
Fe(acac) ]APTES@HMS
HMS
APTES@HMS
or
APTES@AC
2
or
or
AC
[Fe(acac) ]APTES@AC
2
Scheme 1. Anchoring of the [Fe(acac)2] onto the amine functionalised porous supports.

1
74
A.R. Silva, J. Botelho / Journal of Molecular Catalysis A: Chemical 381 (2014) 171–178
(A)
3.1.2. AC
The nitrogen adsorption isotherm at 77 K of the activated carbon
30
28
26
24
22
20
18
16
14
12
10
8
6
4
2
0
(a)
is of type I (Fig. 1A) and therefore is typical of microporous mate-
rials [17]. The inclined adsorption branch of the isotherm suggests
that there is a significant external area in the material. The type H4
hysteresis in the desorption branch of the isotherm suggests the
presence of slit shape pores. The material has a BET surface area of
[Fe(acac) ]APTES@HMS
3
[Fe(acac) ]APTES@HMS
2
APTES@HMS
HMS
2
−1
3
−1
(b)
(c)
894 m g and a pore volume of 0.642 cm g (Table 1).
The FTIR spectrum of the activated carbon (Fig. 1C) presents
several broad bands due to the vibrations of the variety of oxy-
gen functional groups (phenolic, carboxylic acid, lactone, etc.)
that can coexist at the edges of the graphene layers of the
material, as reported elsewhere [16]: a broad band in the range
(d)
−
1
3
700–3000 cm corresponding to the stretching O H vibrations;
−
1
in the range 1700–1450 cm due to aromatic vibrations. H
O H
bending vibrations of physisorbed water and C O stretching vibra-
−
1
0
,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0
tions; and between 1350 and 880 cm
due to C O and C C
stretching vibrations. Additionally, lower intensity C H vibrations
p/p0
−1
can also be detected between 3000 and 2800 cm
.
(B)
As in the case of the HMS, the functionalisation of the activated
carbon with APTES increases the material amount of nitrogen (Sec-
16
14
12
10
8
6
4
2
0
(a)
−
1
tions 2.2 and 2.3), corresponding to 0.85 mmol g of this molecule.
For this material the Si ICP-AES analysis was also performed giv-
−
1
ing 0.81 mmol g , which is in agreement with the amount of
APTES calculated from the nitrogen elemental analysis. The amount
introduced at the surface of the activated carbon is however lower
than in the case of the HMS, probably due to a lower amount of reac-
tive phenol group than hydroxyls as the material was not oxidised
(
b)
(c)
(
d)
[
11,12].
There are no changes in the shape of the isotherm upon func-
[
Fe(acac) ]APTES@AC
2
AC
tionalisation of the activated carbon (Fig. 1B), unless for a decrease
in the surface area of the material (Table 1).
[Fe(acac) ]APTES@AC
3
APTES@AC
The FTIR spectrum (Fig. 1C) shows some change in the inten-
−
1
sity of the bands in the region 1700–1450 cm
probably due
to the presence of the N H asymmetric bending of the amine
0
,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0
−
1
−1
groups at 1575 cm , besides of a new band at 730 cm due to the
N
p/p0
−1
H bending vibration. The bands in the region 1200–1000 cm
became also sharper probably due to the reaction of some of the
oxygen containing groups and the presence of the new C N stretch-
ing vibration of the APTES.
Fig. 1. Nitrogen adsorption isotherms at 77 K for the (A) hexagonal meso-
porous silica and (B) activated carbon based materials: (a) HMS or AC,
(
[
b) [Fe(acac)3]APTES@HMS or [Fe(acac)3]APTES@AC, (c) APTES@HMS or
Fe(acac)2]APTES@AC and (d) [Fe(acac)2]APTES@HMS or APTES@AC.
3.2. Anchoring of iron acetylacetonate onto the
amine-functionalised porous materials
−
1
−1
the regions 1546 cm and 1500–1300 cm , respectively, as weak
intensity bands (Fig. 2A.b). Furthermore, an N H bending band at
6
3.2.1. HMS
−
1
91 cm can also be seen [18].
The thermogram under air atmosphere was also performed for
the APTES@HMS material (Fig. 3A). It shows an extended weight
loss between 200 and 600 C corresponding to 9.0% of the ini-
tial sample weight. This must be due to the decomposition of the
grafted APTES organic molecule. The value is close to the sum of
the nitrogen and carbon content determined by elemental analysis
for this material (Sections 2.2 and 2.3).
Two dichloromethane solutions containing [Fe(acac)2] or
[
Fe(acac) ] were refluxed with amine functionalised HMS. Iron
3
ICP-AES analysis of the resulting materials shows that an higher
◦
−1
amount of the Fe(II) salt (247 ␮mol g ) than of the Fe(III) salt
−1
(
215 ␮mol g ) was anchored onto the material.
No change in the isotherm profile was observed upon complex
immobilisation (Fig. 1A), suggesting that the functionalised HMS
structure was unaffected by complex anchoring.
Table 1
Chemical and textural characteristics of the materials.
N (mmol g 1
−
)
Si (mmol g
−1
)
Fe (␮mol g
−1
)
ABET (m2
g
−1
)
Vtotala (cm3
−1
g )
Material
HMS
APTES@HMS
829
119
111
286
894
525
562
632
0.973
0.390
0.360
0.515
0.642
0.423
0.454
0.435
1.38
1.38
1.31
[
[
Fe(acac)2]APTES@HMS
Fe(acac)3]APTES@HMS
247
215
AC
APTES@AC
0.85
0.94
0.85
0.81
0.85
0.75
[
[
Fe(acac)2]APTES@AC
Fe(acac)3]APTES@AC
138
32
a
Calculated from the nitrogen adsorption isotherm at p/p0 = 0.95.

A.R. Silva, J. Botelho / Journal of Molecular Catalysis A: Chemical 381 (2014) 171–178
175
(A)
(A)
9,0
8,5
8,0
7,5
7,0
6,5
(
a)
(
c)
(b)
APTES@HMS
(d)
[
Fe(acac) ]APTES@HMS
3
HMS
6,0
,5
[Fe(acac) ]APTES@HMS
2
APTES@HMS
[Fe(acac) ]APTES@HMS
3
5
[
Fe(acac)3]
100
200 300 400 500 600 700 800 900
t (C)
4
000 3500 3000 2500 2000 1500 1000
500
500
500
-
1
(B)
wavenumber (cm )
1
2
1
0
9
8
7
(B)
1
(
b)
(e)
1
AC
(f)
APTES@AC
Fe(acac) ]APTES@AC
[
3
1
00 200 300 400 500 600 700 800 900
t (C)
[
[
Fe(acac)2]
Fe(acac)]APTES@HMS
2
APTES@HMS
Fig. 3. Thermograms of the (A) hexagonal mesoporous silica and (B) acti-
vated carbon materials: (a) HMS or AC, (b) APTES@HMS or APTES@AC, (c)
[Fe(acac) ]APTES@HMS or [Fe(acac) ]APTES@AC and (d) [Fe(acac) ]APTES@HMS or
4000 3500 3000 2500 2000 1500 1000
3
3
2
-
1
[Fe(acac)2]APTES@AC.
wavenumber (cm )
Comparatively to the FTIR spectrum of the parent APTES@HMS
material, the [Fe(acac) ]APTES@HMS and [Fe(acac) ]APTES@HMS
3
2
(C)
−
1
spectra show two new sharp bands at 1608 and 1554 cm , which
do not exist in the original iron acetylacetonate FTIR spectra (Fig. 1A
and B). The first band is typical of C N stretching vibrations thus
indicating that there was Schiff condensation between the amine
groups of APTES and the acetylacetonate ligand coordinated to
the iron(III). Besides that there are more vibrations in the region
(
a)
(
b)
(
c)
(
d)
e)
−
1
−1
and some changes at 3100 cm in the N
(
1500–1400 cm
stretching vibrations.
H
The thermograms for the [Fe(acac) ]APTES@HMS and
3
(f)
[
Fe(acac) ]APTES@HMS materials are shown in Fig. 3A. Com-
2
[
Fe(acac) ]APTES@AC
3
paratively to the parent APTES@HMS, they show sharper weight
APTES@AC
AC
Fe(acac)3]
◦
losses in the range 250–600 C corresponding, respectively, to
[
[
[
7.7% and 12.7% of the initial sample weight. This confirms that
reaction between the amine of the APTES group grafted onto the
surface of the HMS took place with the iron acetylacetonates,
Fe(acac) ]APTES@AC
2
Fe(acac)2]
further confirming the presence of higher amount of [Fe(acac)2]
4000 3500 3000 2500 2000 1500 1000
than [Fe(acac) ] in the materials.
3
-
1
wavenumber (cm )
3.2.2. AC
Fig. 2. FTIR spectra of the anchoring onto the hexagonal mesoporous silica
Dichloromethane solutions of [Fe(acac) ] or [Fe(acac) ] were
2
3
of the (A) [Fe(acac)3] and (B) [Fe(acac)2] complex: (a) HMS, (b) APTES@HMS,
(
(
c) [Fe(acac)3]APTES@HMS, (d) [Fe(acac)3], (e) [Fe(acac)2]APTES@HMS and
f) [Fe(acac)2]. (C) FTIR spectra of the anchoring onto the activated car-
refluxed with amine functionalised AC. Iron ICP-AES analysis of
the resulting materials shows that a much higher amount of the
Fe(II) salt (138 ␮mol g ) than of the Fe(III) salt (32 ␮mol g
were anchored onto the material. This was already observed for
bon (a) [Fe(acac)3]APTES@AC, (b) APTES@AC, (c) AC, (d) [Fe(acac)3], (e)
Fe(acac)2]APTES@AC and (f) [Fe(acac)2].
−1
−1
)
[

1
76
A.R. Silva, J. Botelho / Journal of Molecular Catalysis A: Chemical 381 (2014) 171–178
cyclohexanone and cyclohexanol were obtained than with 1 mol%
of Fe(III) acetylacetonate. Both homogeneous catalysts gave higher
yield of cyclohexanone than cyclohexanol, showing similar OH/
ratios. Comparing with a Fe(II) complex with a N4 Schiff base in
mol%, reported by us tested using comparable experimental
O
1
conditions, higher cyclohexanol and cyclohexanone were obtained
with the Fe(II) acetylacetonate. The latter homogeneous catalyst
was however much more selective towards cyclohexanol, than the
Scheme 2. Oxidation of cyclohexane to cyclohexanol and cyclohexanone.
former. On the other hand, a Fe(III) complex with a N O2 Schiff
2
base gave higher cyclohexanol and cyclohexanone yield than
the Fe(III) acetylacetonate complex reported herein. The former
complex was also more selective towards cyclohexanol, whereas
the latter was selective for cyclohexanone.
Upon anchoring of the Fe(II) acetylacetonate onto the amine-
functionalised AC comparable cyclohexanol yield was obtained
and slightly lower amount of cyclohexanone, resulting in a slight
decrease in the overall oxygenate yield and increase in the OH/ O
ratio. Whereas anchoring onto the amine-functionalised HMS
significantly lower cyclohexanol and cyclohexanone yield was
the amine functionalised HMS materials, thus indicating that the
reaction with the grafted amine is easier for the square planar
[
Fe(acac) ] complex than for the octahedral [Fe(acac) ].
2
3
No change in the isotherm profile was observed upon com-
plex immobilisation (Fig. 1A), suggesting that the functionalised
AC structure was unaffected by complex anchoring.
Probably due to the low content in [Fe(acac) ] the FTIR spec-
3
trum of the [Fe(acac) ]APTES@HMS material is similar to its
3
parent material. On the other hand, the FTIR spectrum of the
[
1
Fe(acac) ]APTES@HMS material shows a new intense band at
630 cm typical of C N stretching vibrations. This indicates that
2
obtained, despite containing the double amount of [Fe(acac) ].
2
−
1
A significant oxygenate yield increase is however surprisingly
there was Schiff condensation between the amine groups of APTES
and the acetylacetonate ligand coordinated to the iron(II).
observed upon reutilisation of the [Fe(acac) ]APTES@HMS hetero-
2
geneous catalyst in the subsequent 3 cycles. This may be due to the
inefficient washing of the heterogeneous catalyst after the anchor-
ing process and thus partial pore blocking. On the other hand,
3.3. Catalytic experiments
reutilisation of the [Fe(acac) ]APTES@AC only yields cyclohexanol
2
As can be confirmed in Tables 2 and 3 both Fe(II) and Fe(III)
acetylacetonate commercial salts work as homogeneous cata-
upon the following 2 reuses and thus the yield on oxygenates
decreases.
lysts in the oxidation of cyclohexane (Scheme 2) and n-hexane
Anchoring of the Fe(III) acetylacetonate onto the amine-
functionalised activated carbon yields lower amount of
cyclohexanol and cyclohexanone than the homogeneously catal-
(Scheme 3) at room temperature using hydrogen peroxide. This is in
accordance to the literature [2], although the results are not directly
comparable as different oxidation methodologies were used.
ysed reaction probably due to the its lower content in [Fe(acac) ].
3
This heterogeneous catalyst is however more selective towards
cyclohexanol than the corresponding [Fe(acac)3]APTES@AC.
Although the [Fe(acac)3]APTES@HMS heterogeneous catalyst
possesses higher OH/ O ratio, it gives comparable oxygenate
yield than the corresponding homogeneous phase reaction and
[Fe(acac)2]APTES@HMS heterogeneous catalyst. Furthermore,
3
.3.1. Oxidation of cyclohexane
The homogeneous and heterogeneous phase results for the
oxidation of cyclohexane (Scheme 2) are collected in Table 2. In
the homogeneous phase oxidation of cyclohexane with 1 mol%
of Fe(II) acetylacetonate complex considerably higher yield of
Table 2
a
Homogeneous and heterogeneous room temperature peroxidative oxidation of cyclohexane.
Cycle
Fe (mol%)
% yieldb
OHc
OH/
O
TON
Od
Total
1
1
1.1
2.1
8.9
6.5
13.1
11.4
22.0
17.9
0.7
0.6
20.8
8.3
[
[
Fe(acac)2]
1
2
3
1.1
1.1
1.0
8.3
8.5
10.0
9.5
–
–
18.3
8.5
10.0
0.8
–
–
16.1
8.4
10.4
Fe(acac)2]@AC
1
2
3
4
2.1
2.0
1.7
1.5
2.8
7.6
5.5
9.1
2.8
11.4
9.2
5.6
19.0
14.7
23.2
1.0
0.7
0.6
0.7
2.7
9.7
8.8
[Fe(acac)2]@HMS
14.0
15.7
1
1
0.3
1.0
12.4
5.2
10.8
7.8
22.9
13.0
1.2
0.6
78.6
12.6
[
[
Fe(acac)3]
1
2
3
0.3
0.3
0.2
3.8
2.9
2.3
3.5
2.6
2.1
7.3
5.6
4.4
1.1
1.1
1.1
26.3
21.6
6.9
Fe(acac)3]@AC
1
2
3
4
1.7
1.6
1.4
1.3
6.7
14.7
7.9
6.9
18.3
10.2
9.1
13.6
33.0
18.1
14.7
1.0
0.8
0.8
0.6
7.8
20.3
13.3
11.6
[Fe(acac)3]@HMS
5.6
a
1
05 mg heterogeneous catalyst, 1.25 mmol of alkane, 0.125 mmol of nitric acid and hydrogen peroxide (6.25 mmol) in 5.00 mL of ACN in batch reactors at room temperature
and under constant stirring (700 rpm) for 5 h of reaction: ratios catalyst/alkane/HNO3/H2O2 = 1:100:10:500.
b
Yield based on the alkane.
OH = alcohol.
O = ketone.
c
d

A.R. Silva, J. Botelho / Journal of Molecular Catalysis A: Chemical 381 (2014) 171–178
177
Table 3
Homogeneous and heterogeneous room temperature peroxidative oxidation of n-hexane.
a
Cycle
Fe (mol%)
Product distribution (mol%)
% yieldb
OH/
O
TON
3
O
2
O
1
O
3
OH
2
OH
1
OH
OHc
Od
4.8
Total
14.3
1
1
1.1
2.1
16
17
17
16
11
11
56
56
9.6
19.4
2.0
2.1
13.2
13.5
[
[
Fe(acac)2]
9.4
28.8
1
2
3
1.2
1.1
1.0
–
46
24
–
35
–
–
100
19
15
–
–
61
–
–
1.2
0.6
4.6
–
2.6
1.4
1.2
3.2
6.0
–
0.2
3.2
1.0
2.8
6.2
Fe(acac)2]@AC
1
2
3
4
2.0
1.8
1.8
1.6
–
44
45
43
–
39
39
36
–
–
–
–
–
16
16
21
–
–
–
–
–
–
–
–
–
–
–
–
–
1.1
1.2
1.0
5.6
6.1
3.7
6.6
7.2
4.6
0.2
0.2
0.3
3.8
4.0
2.9
[Fe(acac)2]@HMS
[
[
Fe(acac)3]
1
1.1
28
22
–
12
39
–
10.3
10.0
20.4
1.0
18.7
1
2
3
0.3
0.3
0.2
32
–
–
24
–
–
12
–
–
32
–
–
2.0
–
–
2.5
–
–
4.5
–
–
0.8
–
–
2.6
–
–
Fe(acac)3]@AC
–
–
–
1
2
3
4
1.7
1.6
1.4
1.3
22
44
42
41
19
39
38
36
–
–
–
–
16
16
20
23
44
–
–
–
–
–
–
7.2
1.1
1.3
1.4
4.9
5.7
5.1
4.5
12.1
6.8
6.4
1.5
0.2
0.3
0.3
7.0
4.1
4.6
4.4
[Fe(acac)3]@HMS
–
5.9
a
1
05 mg heterogeneous catalyst, 1.25 mmol of alkane, 0.125 mmol of nitric acid and hydrogen peroxide (6.25 mmol) in 5.00 mL of ACN in batch reactors at room temperature
and under constant stirring (700 rpm) for 5 h of reaction: ratios catalyst/alkane/HNO3/H2O2 = 1:100:10:500.
b
Yield based on the alkane.
c
3
OH = alcohol.
O = ketone.
d
upon reuse it significantly increases the oxygenate yield, dropping
in the following third and fourth reuses, but still maintaining higher
heterogeneous catalysts gain catalytic activity and thus higher
yields are obtained of oxygenate products. In contrast to the
homogeneous phase results with the Fe(II) acetylacetonate, more
ketones than alcohols are obtained in heterogeneous phase with
than the first cycle. The reutilisation of the [Fe(acac) ]APTES@AC
3
leads to a slight drop in the oxygenate yield and to a significant
decrease in the third utilisation of the heterogeneous catalyst.
As the OH/ O ratios are always close to 1 and the pres-
ence of cyclohexyl hydroperoxide was detected and confirmed by
GC–MS in the reaction media and upon its reduction with PPh3
their concentration decreases, to decompose into cyclohexanol and
cyclohexanone, it may be concluded that a radical mechanism is
mainly involved [7,8]. Thus the main product of the peroxidative
oxidation of cyclohexane is first cyclohexyl hydroperoxide, which
is formed from the reaction of the alkane with a hydroxyl radical
generated from the reaction of hydrogen peroxide with the iron
centre. However competitive non-radical pathways through a high
valent iron species cannot be ruled out.
the exception of the third cycle of the [Fe(acac) ]@AC cata-
2
lyst.
On the other hand, the Fe(III) complex immobilised on AC and
HMS are both active in the peroxidative oxidation of n-hexane, but
less than the homogeneous phase reaction. The [Fe(acac) ]@HMS
3
catalyst is however more active than the [Fe(acac) ]@AC. The first
3
is also more selective towards the alcohols, whereas the second is
for the ketones. Contrary to what was found in the oxidation of
cyclohexane (Table 2), these two heterogeneous catalysts based on
the Fe(III) complex deactivate upon reuse. No activity is found for
the [Fe(acac) ]@AC in the subsequent two more cycles, whereas
3
for the [Fe(acac) ]@HMS the activity drops to half but maintains
3
constant for the next three cycles.
Likewise to the oxidation of cyclohexane with hydrogen perox-
ide, in the oxidation of n-hexane the OH/ O ratios are always close
to 1 and the presence of the various n-hexyl hydroperoxides were
detected and confirmed by GC–MS in the reaction media. After-
wards, their concentration decreased upon reduction with PPh3
and thus once more it may be concluded that a radical mechanism
is mainly involved in the oxidation of n-hexane by hydrogen per-
oxide [7,8]. However competitive non-radical pathways through a
high valent iron species cannot be ruled out [7,8].
3
.3.2. Oxidation of n-hexane
The homogeneous and heterogeneous phase results for the
oxidation of n-hexane (Scheme 3) are collected in Table 3. The
anchored Fe(II) complex onto activated carbon is significantly
less active than in homogeneous phase in the more demanding
oxidation of linear hexane molecule, but with 100% selectivity
towards the oxidation of the wanted terminal position. For the
Fe(II) complex immobilised onto the HMS no oxidation products
are observed (Table 3). However, in consequent reuses these two
[
Fe(acac)n]
or
OH
[
Fe(acac)n]APTES@HMS
HO
O
or
[
Fe(acac) ]APTES@AC
OH
n
O
ACN, H2O2, rt
O
Scheme 3. Possible products of oxidation of n-hexane.

1
78
A.R. Silva, J. Botelho / Journal of Molecular Catalysis A: Chemical 381 (2014) 171–178
4
. Conclusions
the improvement of the sustainability and environmental impact
of oxidation processes currently used in industry.
Two commercially available iron salts were anchored onto
amine functionalised economical porous supports by a very simple
and straightforward procedure. The immobilised iron salts were
active as heterogeneous catalysts in the oxidation of cyclohexane
and n-hexane at room temperature, giving the respective alcohols
and ketones. They could also be recycled at least two times without
loss of catalytic activity in the oxidation of cyclohexane. In the oxi-
dation of n-hexane this was only true for the anchored Fe(II) salts.
In the oxidation of cyclohexane, both iron salts are more active
heterogeneous catalysts anchored onto the activated carbons than
onto the hexagonal mesoporous silicas. The opposite is observed in
the more demanding oxidation of n-hexane.
Acknowledgements
This research was supported by a Marie Curie European
Reintegration Grant within the 7th European Community Frame-
work Programme (FP7-PEOPLE-2009-RG-256509). A.R.S. thanks
FCT, Fundo Social Europeu and Programa Operacional Potencial
Humano for the contracts under the programme Ciência 2008
and Investigator FCT 2012. FCT is also acknowledged for financing
CICECO under the Projetos Estratégicos Pest-C/CTM/LA0011/2013.
References
The main products of cyclohexane oxidation, cyclohexanol and
cyclohexanone, are used as precursors to the synthesis of added-
value products, such Nylon-6 and Nylon-6,6. Higher yields in
oxygenates were obtained with the economical homogeneous and
heterogeneous catalysts based on commercially available and eco-
nomical iron acetylacetonate salts, under mild and sustainable
conditions, than in the industrial process, where 4% conversion into
oxidised products with 80% selectivity towards cyclohexanol and
[
[
1] A.E. Shilov, G.B. Shul’pin, Chem. Rev. 97 (1997) 2879–2932.
2] B. Retcher, J. Sánchez Costa, J. Tang, R. Hage, P. Gamez, J. Reedijk, J. Mol. Catal.
A: Chem. 286 (2008) 1–5.
[
[
3] A. Pokutsa, J. Le Bras, J. Muzart, Russ. Chem. Bull. Int. Ed. 54 (2005)
312–315.
4] J.M. Thomas, R. Raja, G. Sankar, R.G. Bell, Nature 398 (1999) 227–230.
[5] G.B. Shul’pin, T. Sooknoi, L.S. Shul’pina, Petrol. Chem. 48 (2008) 36–39.
[6] B. Meunier, Chem. Rev. 92 (1992) 1411–1456.
[
[
7] A.R. Silva, T. Mourão, J. Rocha, Catal. Today 203 (2013) 81–86.
8] (a) A.M. Kirillov, M.N. Kopylovich, M.V. Kirillova, E.Y. Karabach, M.
Haukka, M.F.C.G. da Silva, A.J.L. Pombeiro, Adv. Synth. Catal. 348 (2006)
159–174;
◦
cyclohexanone, are obtained at 160 C and 15 bar using a cobalt(II)
naphthenate catalyst.
(
b) M.V. Kirillova, Y.N. Kozlov, L.S. Shul’pina, O.Y. Lyakin, A.M. Kirillov, E.P. Talsi,
A.J.L. Pombeiro, G.B. Shul’pin, J. Catal. 268 (2009) 26–38;
c) A.M. Kirillov, M.N. Kopylovich, M.V. Kirillova, M. Haukka, M.F.C.G. da Silva,
With the Fe(II) acetylacetonate anchored onto activated carbon
total selectivity to the 3 position of the molecule was also obtained
in the oxidation of n-hexane.
As the OH/ O ratios with our heterogeneous catalysts are
always close to 1 and the presence of cyclohexyl hydroperox-
ide and n-hexyl hydroperoxides were detected and confirmed by
GC–MS in the reaction media, which concentration decreased upon
(
A.J.L. Pombeiro, Angew. Chem. Int. Ed. 44 (2005) 4339–4345.
9] A.R. Silva, K. Wilson, A.C. Whitwood, J.H. Clark, C. Freire, Eur. J. Inorg. Chem.
[
2006 (2006) 1275–1283.
[
[
[
10] B. Jarrais, C. Pereira, A.R. Silva, A.P. Carvalho, J. Pires, C. Freire, Polyhedron 28
(2009) 994–1000.
11] A.R. Silva, J.L. Figueiredo, C. Freire, B. de Castro, Catal. Today 102 (2005)
154–159.
reduction with PPh , we can conclude that a radical mechanism is
3
12] B. Jarrais, A.R. Silva, C. Freire, Eur. J. Inorg. Chem. 2005 (2005) 4582–4589.
mainly involved in the oxidation by hydrogen peroxide [7,8]. How-
ever competitive non-radical pathways through a high valent iron
species cannot be ruled out [7,8].
[13] P.T. Tanev, T.J. Pinnavaia, Science 267 (1995) 865–867.
[
[
14] P.T. Tanev, T.J. Pinnavaia, Chem. Mater. 8 (1996) 2068–2079.
15] V. Calvino-Casilda, A.J. Lopez-Peinado, C.J. Duran-Valle, R.M. Martin-Aranda,
Catal. Rev. Sci. Eng. 52 (2010) 325–380.
Saturated hydrocarbons are chemically inert, thus their func-
tionalisation is challenging from a basic science point of view.
Therefore, the developed homogeneous and heterogeneous cata-
lysts based on first row transition metal complexes are economical
and work under mild reaction conditions and could be valuable for
[16] J.L. Figueiredo, M.F.R. Pereira, M.M.A. Freitas, J.J.M. Órfão, Carbon 37 (1999)
1379–1389.
[
17] J. Rouquerol, F. Rouquerol, K. Sing, Adsorption by Powders and Porous Solids:
Principles, Methodology and Applications, Academic Press, London, UK, 1999.
[
18] X. Wang, K.S.K. Lin, J.C.C. Chan, S. Cheng, J. Phys. Chem.
1763–1769.
B 109 (2005)