Solvent-Free Aerobic Oxidation of Ethylbenzene Promoted by NHPI/Co(II) Catalytic System: The Key Role of Ionic Liquids

Article abstract of DOI:10.1002/cctc.201901737

The synergistic action between imidazolium based ionic liquid (IL) [bmim][OcOSO₃] and Co(II)/N-hydroxyphthalimide (NHPI) systems in the catalytic aerobic oxidation of ethylbenzene under solvent-free conditions have been here demonstrated by reaching a 35 % conversion with 83 % of selectivity in acetophenone at 80 °C. This highly performing catalytic system have been selected after screening several different IL and Co(II) salt combinations, and making sure that the complete solubilization of the polar NHPI in the lipophilic medium, without thus requiring any chemical modification of the organic catalyst, could be attained. This solubilizing effect can be ascribed to a direct interaction between [bmim][OcOSO₃] IL and NHPI as revealed by a detailed NMR investigation which also allowed to exclude the formation of higher IL aggregates in the form of micelles or vesicles.

Full text of DOI:10.1002/cctc.201901737

Accepted Article
Title: Solvent-free Aerobic Oxidation of Ethylbenzene Promoted by
NHPI/Co(II) Catalytic System: The Key Role of Ionic Liquids
Authors: Gabriela Dobras, Magdalena Sitko, Manuel Petroselli,
Manfredi Caruso, Massimo Cametti, Carlo Punta, and Beata
Orlinska
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201901737
Link to VoR: http://dx.doi.org/10.1002/cctc.201901737
A Journal of
www.chemcatchem.org

10.1002/cctc.201901737
ChemCatChem
FULL PAPER
Solvent-free Aerobic Oxidation of Ethylbenzene Promoted by
NHPI/Co(II) Catalytic System: The Key Role of Ionic Liquids
Gabriela Dobras,[a] Magdalena Sitko,[a] Manuel Petroselli,[b] Manfredi Caruso,[b] Massimo Cametti,[b]
Carlo Punta,*[b] and Beata Orlińska*[a]
Abstract: The synergistic action between imidazolium based ionic
liquid (IL) [bmim][OcOSO3] and Co(II)/N-hydroxyphthalimide (NHPI)
systems in the catalytic aerobic oxidation of ethylbenzene under
solvent-free conditions have been here demonstrated by reaching a
35% conversion with 83% of selectivity in acetophenone at 80°C.
This highly performing catalytic system have been selected after
screening several different IL and Co(II) salt combinations, and
making sure that the complete solubilization of the polar NHPI in the
lipophilic medium, without thus requiring any chemical modification
of the organic catalyst, could be attained. This solubilizing effect can
be ascribed to a direct interaction between [bmim][OcOSO3] IL and
NHPI as revealed by a detailed NMR investigation which also
allowed to exclude the formation of higher IL aggregates in the form
of micelles or vesicles.
reaction of initially formed hydroperoxides (Fig. 1, path v).
Therefore, from a given alkylaromatic substrate, the main
products obtained by employing a NHPI/Co(II) catalytic system
would be the corresponding alcohol (Fig. 1, path v), and the
products obtained by further oxidation or termination processes
(aldehyde or ketone, and carboxylic acid). For example,
oxidation of ethylbenzene (EB) in the presence of 10 mol%
NHPI and 0.5 mol% Co(II) in acetic acid leads to acetophenone
(AP) and 1-phenylethanol (PEOH).[7]
Introduction
Alkyl C-H bond activation by hydrogen atom transfer (HAT)
processes is recently attracting a renewed interest. One of the
most versatile and effective strategies to reach this goal is
undoubtedly the oxidation of C-H bonds to selectively obtain the
desired products. In this context, over the past 20 years, the role
of N-hydroxyphthalimide (NHPI) in the aerobic oxidation of
hydrocarbons, and in particular of alkylaromatics, has been
highlighted by several studies. The reaction proceeds via a
well-established free radical mechanism, where NHPI acts as
precursor of the phthalimide–N-oxyl radical (PINO) (Fig. 1).[1-6]
Compared with the peroxyl radicals involved in autocatalytic
reactions, PINO abstracts the substrate’s hydrogen at a faster
rate (from 2 to 20 times faster)[2] and, at the same time, NHPI
favors the propagation of the radical chain at the expense of
termination reactions; both factors result in an increase in the
oxidation rate of the hydrocarbon. NHPI-catalyzed oxidations of
benzylic hydrocarbons are often carried out in the presence of
transition metal salts, most commonly of Co(II), as co-catalysts,
which play the role of activating molecular oxygen by electron
transfer.[2] These species participate in the generation of the
PINO radical (Fig. 1, path i), but also catalyze the decomposition
Figure 1. Mechanism of NHPI-catalyzed hydrocarbons oxidation[6]
NHPI-catalyzed oxidation reactions are most often carried out in
polar solvents, such as acetic acid, acetonitrile (MeCN), and
benzonitrile (PhCN), due to its poor solubility in non-polar
hydrocarbons.[7-10] The complete solubilization of NHPI catalyst
is essential, as NHPI can exploit its hydrogen atom donating
action in the HAT processes shown in Fig. 1 only provided it is
under homogeneous conditions. However, the use of polar
solvents limits the implementation of the catalytic system in real
oxidative processes.[11] As
a
consequence, given the
advantages of operating under solvent-free conditions,
especially from an economic point of view, several attempts
have been made in order to increase NHPI solubility in apolar or
weakly polar hydrocarbons. Main research avenues comprise: i)
the structural modification of the NHPI unit to increase its
lipophilicity, ii) the formation of solubilizing micelles or higher
aggregates; iii) the use of additives into the reaction mixture.
The first approach was introduced by Ishii and co-workers[12] and
more recently widely explored by our research group.[13-15] It
simply consists into the introduction of lipophilic chains on the
aromatic ring of NHPI unit trying, at the same time, to avoid any
increase of the bond dissociation energy (BDE) of the NO-H
group which would consequently negatively affect the HAT
process. The second approach was mainly addressed by
Sasson and co-workers, and consisted in attaining partial phase
transfer of NHPI into the alkylaromatic organic medium by
means of formation of multilamellar vesicles made of lipophilic
quaternary ammonium salts.[16,17] As to the last alternative, ionic
[a]
[b]
G. Dobras, Dr. M. Sitko, Prof. B. Orlińska
Department of Organic Chemical Technology and Petrochemistry
Silesian University of Technology, Krzywoustego 4, 44-100 Gliwice,
Poland
E-mail: beata.orlinska@polsl.pl
Dr. Manuel Petroselli, Mr. Manfredi Caruso, Prof. M. Cametti, Prof.
C. Punta.
Department of Chemistry, Materials and Chemical Engineering “G.
Natta” Politecnico di Milano, Via Luigi Mancinelli 7, 20131 Milano,
Italy
E-mail: carlo.punta@polimi.it
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201901737
ChemCatChem
FULL PAPER
liquids (ILs) have attracted our interest. Considered as
alternative green solvents, ILs’ advantages include low volatility
as well as high polarity and chemical and thermal stability.[18]
Several examples have been reported on the use of ILs as
solvents for NHPI-mediated oxidations of a wide range of
substrates[19,20], including the conversion of alkyl aromatic
hydrocarbons, such as p-xylene, toluene, EB, cumene and their
derivatives, to the corresponding carboxylic acids.[21-23] Among
the wide range of ILs commercially available, it was
demonstrated that the imidazolium-based ones are particularly
suitable for promoting aerobic oxidations, as they can accelerate
the electron transfer from metal ions to O2 by stabilization of the
resulting superoxyl radical.[24-27] Recently, some of us reported
the oxidation of cumene and other alkylaromatic hydrocarbons
to hydroperoxides in the presence of various N-hydroxyimides,
including NHPI, in a broad range of ILs composed of 1-alkyl-3-
methylimidazolium cations and bis(trifluoromethylsulfonyl)imide
([NTf2]), [PF6], [BF4], n-octylsulfate ([OcOSO3]), and [CH3OSO3]
anions.[28] However, the replacement of standard polar solvents,
such as MeCN, with ILs might not be an improvement in terms
of costs and process implementation. Interestingly, ILs have
been also used as catalytic additives capable to promote
oxidation reactions.[29] In this context, very recently, Mahmood et
al. reported the beneficial effect of catalytic amounts of different
imidazolium-based ILs (5% mol) on the aerobic oxidation of
toluene to the corresponding benzoic acid, in the presence of
NHPI (10% mol) and Co(OAc)2 (0.5% mol).[30] In this case, the
authors observed a behavior quite similar to that reported by
Sasson and co-workers[16,17] with lipophilic ammonium salts, that
is, the formation of a biphasic water/organic medium with the ILs
acting as surfactant and/or phase transfer catalysts.
Here, we report on the aerobic oxidation of EB with NHPI/Co(II)
system in the presence of catalytic amounts (0.25-1%) of the
imidazolium-based ILs previously investigated as solvents.[28]
The full set of experiments here described point to a marked
effect of alkylsulfate ILs in the solubilization of NHPI in EB,
which in turn allowed to perform substrate oxidation under
solvent free conditions, reaching conversion up to 35% with
selectivity higher than 82% in AP. Detailed analysis of the
outcome of oxidation reactions performed by varying IL species,
Co(II) counterions, with and without NHPI, presents a complex
scenario with multiple concomitant equilibria which should be
taken into account and points to the presence of a specific
interaction between IL and NHPI.
ketone. It is quite pertinent to note that while [bmim][OcOSO3]
and [bmim][C12H25OSO3] ILs dissolve in the system at 80°C
within the initial stages of the reaction (ca. 15 minutes),
[bmim][CF3SO3] and [bmim][CH3COO] ILs do not. By using
[bmim][BF4], [bmim][PF6] and [bmim][NTf2], instead, a brown
sediment formed in the flask. Qualitative analysis confirmed the
presence of Co(II) in the powder. Keeping aside the data where
not optimal dissolution occurred, a positive influence of the
addition of ILs [bmim][OcOSO3] and [bmim][C12H25OSO3] (Table
1, entries 4 and 5) is evident. Conversion (α) doubles (from ca. 7
to ca. 12), while the AP selectivity slightly diminishes in favor of
higher amounts of PEOH or EBOOH.
The reason for this improvement in conversion cannot be simply
attributed to a polar effect. Solvent polarity changes would
require addition of more significant quantities of polar co-solvent
in order to have an observable effect, and this is not the case as
demonstrated by entry 6 where PhCN was added (0.5 mol%)
and no effect was observed.
Table 1. EB oxidation with oxygen in the presence of Co(acac)2/IL
Entry
Additive
α
[%]
SAP
[%]
SPEOH
[%]
SEBOOH
[%]
1[a]
2[a]
3
-
0
-
-
-
[bmim][OcOSO3]
-
1.4
7.8
12.3
10.2
5.3
7.5
6.2
4.1
4.1
3.6
68.5
71.8
61.1
64.9
75.0
72.2
73.1
69.0
62.0
65.7
12.8
23.5
35.8
21.2
20.3
23.3
8.7
18.7
3.7
2.2
11.8
4.2
1.9
16.2
5.5
10.5
9.2
4
[bmim][OcOSO3]
[bmim][C12H25OSO3]
PhCN
5
6
7
[bmim][BF4]
[bmim][PF6]
[bmim][CF3SO3]
[bmim][CH3COO]
[bmim][NTf2]
8
9
25.3
26.5
18.8
10
11
EB 2 ml, Co(acac)2 0.1 mol%, additive 0.5 mol%, 0.1 MPa O2, 80°C, 6 h,
1200 rpm [a] without Co(acac)2
In the following step, we determined the influence of the addition
of NHPI to the Co(II)/IL mixture. The results are presented in
Table 2. ILs [bmim][NTf2], [bmim][OcOSO3], [bmim][BF4], and, to
a minor extent, [bmim][PF6], led to an increase in conversion in
the presence of NHPI (Table 2, entries 5,6,7, and 8,
respectively). Again, it must be noted that NHPI (1 mol%) is not
soluble by itself in EB at 80 °C, unless significant amounts of co-
solvent are added (23% of MeCN in volume was reported to be
necessary at 70 °C).[13] In the presence of both [bmim][OcOSO3]
and [bmim][C12H25OSO3], instead, NHPI dissolved in the
reaction mixture within the initial stages of the reaction (within 15
min). On the contrary, in the presence of [bmim][BF4],
[bmim][PF6], or [bmim][NTf2], a brown sediment formed similarly
Results and Discussion
Our investigation started by considering the effect on the EB
oxidation catalyzed by Co(acac)2 of the addition of catalytic
amounts of a series of [bmim] ILs (0.5 mol%) under solvent-free
conditions (Table 1). All the reactions were conducted at 80°C.
As expected, the main products of the oxidation process were
AP and PEOH, while ethylbenzene hydroperoxide (EBOOH)
was rarely present and only in traces, due to the well-known
Co(II)-induced hydroperoxide decomposition into alcohol and
This article is protected by copyright. All rights reserved.

10.1002/cctc.201901737
ChemCatChem
FULL PAPER
to that previously described, and involving Co(II). Nevertheless,
the high conversions measured also with these additives
suggests a positive effect on NHPI solubilization (vide infra).
(Table 2, entry 15). As to the selectivity of the oxidation process
by the NHPI/Co(acac)2/IL system, GC-FID and GC-MS analyses
showed that the main products were AP and PEOH, with low
content of EBOOH (2.0 - 7.8%). GC-MS analysis also showed
the presence of (among others) benzoic acid, benzaldehyde,
di(1-phenylethyl) ether, and 2.3-diphenylbutane.
Table 2. EB oxidation with oxygen in the presence of NHPI/Co(acac)2/IL
Entry
Additive
α
[%]
SAP
[%]
SPEOH
[%]
SEBOOH
[%]
Finally, as the oxidation of EB using ILs having [NTf2], [PF6],
and [BF4] counteranions, despite the above mentioned
„complicating” feature of precipitate formation, contributed to
reach a conversion up to 21.2% ([bmim][NTf2]), we decided to
explore the effect of differently substituted imidazolium
countercations, among those commercially available, aiming at
finding a more optimal solution (Table 3). It was shown that, in
the case of ILs with [NTf2] and [PF6] anions, the conversion of
EB improves along with increasing alkyl chain length, up to a
value of 25% (Table 3, entry 5). It seems plausible to assume
that the increase in length of the alkyl chain in the IL cation could
positively influence the solubility of IL in EB. On the contrary,
the increase in length of alkyl group in the IL with the [BF4] anion
caused a slight decrease in conversion from 18.6 to 13.3%. GC-
FID analysis showed that the presence of [omim][PF6] and
[omim][BF4] liquids intensified the side reactions, which led to
the production of benzoic acid, di(1-phenylethyl) ether, 2.3-
diphenylbutane, and possibly other high-boiling products that are
undetectable by gas chromatography.
1[a]
2[a]
3
-
3.9
58.2
17.0
74.9
70.5
75.5
77.7
73.2
70.9
77.1
63.8
62.0
78.9
4.1
36.9
73.7
3.3
2.0
3.1
3.3
3.1
7.8
4.4
2.1
2.1
2.5
[bmim][OcOSO3]
-
5.5
8.5
14.9
14.1
21.2
19.2
18.6
15.6
13.5
7.0
18.9
22.9
16.7
15.6
17.4
12.5
18.3
32.3
20.1
15.8
4
PhCN
5
[bmim][NTf2]
[bmim][OcOSO3]
[bmim][BF4]
[bmim][PF6]
[bmim][C12H25OSO3]
[bmim][CH3COO]
[bmim][CF3SO3]
6
7
8
9
10
11
12
9.4
[bmim][OcOSO3]
0.25mol%
18.5
Table 3. Influence of the length of alkyl group in cation of ILs on EB oxidation
with oxygen of in the presence of NHPI/Co(acac)2
13
14
15
[bmim][OcOSO3]
1mol%
24.5
11.7
13.9
81.1
40.3
72.6
13.4
8.8
2.3
Entry
Additive
α
[%]
SAP
[%]
SPEOH
[%]
SEBOOH
[%]
[bmim][OcOSO3]
2 mol%
49.0
2.3
1
2
3
4
5
6
7
8
9
-
14.9
19.3
21.2
22.8
25.0
15.6
23.7
18.6
13.3
74.9
70.6
75.5
77.3
81.9
70.9
71.6
73.2
68.0
18.9
16.2
16.7
15.2
11.9
12.5
12.1
17.4
14.8
3.3
3.2
3.1
3.5
2.6
7.8
4.3
3.1
4.6
[emim][NTf2]
[bmim][NTf2]
[hmim][NTf2]
[omim][NTf2]
[bmim][PF6]
[omim][PF6]
[bmim][BF4]
[omim][BF4]
[bmim][NTf2]
1 mol%
20.5
EB 2 ml, NHPI 1 mol%, Co(acac)2 0.1 mol%, additive 0.5 mol%, 0.1 MPa O2,
80°C, 6 h, 1200 rpm [a] without Co(acac)2
Oxidation experiments with [bmim][OcOSO3] were conducted
also at 0.25, 1, and 2% of IL (Table 2, entries 12,13, and 14).
Results showed that the highest conversions were obtained with
1% mol of IL, while the use of 2% mol [bmim][OcOSO3] resulted
in the formation of a separate phase, along with a decrease in
EB conversion. On the other hand, lower IL content, i.e., 0.25 %
mol also decreased the conversion. This indicates once more
the crucial role of [bmim][OcOSO3] in dissolving the NHPI/Co(II)
system in the medium. Notably, a biphasic system associated to
a detrimental effect is also observed with [bmim][NTf2] at 1% mol
EB 2 ml, NHPI 1 mol%, Co(acac)2 0.1 mol %, additive 0.5 mol%, 0.1 MPa O2,
80°C, 6 h, 1200 rpm
This article is protected by copyright. All rights reserved.

10.1002/cctc.201901737
ChemCatChem
FULL PAPER
At this point, our working hypothesis for the interpretation of
these data relies on the effect of ILs on the solubility of Co(II)
salts, but more importantly, on the NHPI. As said, the effect of
alkylammonium salts to NHPI catalytic performances has been
already reported.[16,17,31] However, it was concluded that better
solubility of NHPI under the tested conditions were due to the
formation of vesicles, incorporating and solubilizing the catalyst.
In order to better understand the role of ILs in the solubilization
of NHPI, we embarked on a more detailed NMR study of the
NHPI – IL interaction. Thus, we set out a series of 1H-NMR
experiments in toluene-d8, and we selected [bmim][OcOSO3] as
IL of choice for the following tests. We first verified that NHPI
becomes indeed soluble at room temperature up to 0.094 M in
toluene-d8 containing 2% IL, up to 0.047 M with IL at 1% and up
to 0.023 M in 0.5 % IL conditions. Solubility of NHPI in toluene-
d8 without the addition of ILs is quite low (lower than 0.003 M).
Given the narrow set of conditions, in terms of NHPI : IL ratios,
which warrant a complete solubilization of both species, a
classic titration experiment is impractical. This notwithstanding, it
is possible to observe changes in the 1H-NMR spectrum of the IL
(2%w/w) upon addition of sub-stoichiometric quantities of NHPI (0
– 0.5 equivalents). Indeed, as shown in Fig. 2 (spectra a-f),
signals attributed to the imidazolium ion present a progressive
upfield shift (of ca. 0.1 ppm for H2 and ca. 0.08 and 0.11 ppm for
H4 and H5, respectively).
consistent with the formation of higher aggregates involving IL
and NHPI. A question naturally arises, and it is related to the
initial aggregation state of the IL before the addition of NHPI. In
other words, are micelle or other aggregates already present in
solution, which then further assemble upon addition of NHPI? Or
is the IL completely solvated? DOSY-NMR experiments
performed with IL samples at 0.25, 1 and 2% w/w concentrations
indicate the occurrence of a somewhat intermediate situation.
Figure 2. Portions of the 1H-NMR spectra of [bmim][OcOSO3] IL (2%) alone a)
and upon addition of increasing quantities of NHPI (from 0.1 to 0.7 equivalents
b-h) at 300 K in toluene-d8.
Figure 3. DOSY 1H-NMR spectra of a) 0.25% (w/w) IL solution; b) a mixture of
IL (1% w/w) and NHPI (0.5 eq.) and c) a mixture of IL (2% w/w) and NHPI (0.7
eq.), corresponding to the sample described before (see Fig. 2h). Colored
dashed lines represent: black=toluene; green=IL; blue=NHPI; red=IL
aggregates.
These changes can be attributed to a direct interaction between
the IL and the NHPI. Similar behavior was also observed in
CDCl3 (Fig. S1). Interestingly, upon further increase of the NHPI
added, new, broad signals appear (see spectra g-h), which are
The diffusion coefficient measured for IL within the above range
of concentrations change little and is found to be within 1.3 – 2.1
This article is protected by copyright. All rights reserved.

10.1002/cctc.201901737
ChemCatChem
FULL PAPER
x 10-6 mol/cm2s. No discrimination between cation and anion is
observed Figure 3a.[32] Within the rough approximation given by
the use of the Stokes-Einstein equation, it can be estimated a
dynamic radius in the 20-30 Å range, which obviously does not
represent a good match with isolated solvated ions, but rather
with small ion-pairs aggregates.
After the addition of more than about 20 mol% [bmim][OcOSO3]
(vs NHPI) (Fig. 4, after first portion), the peak height increased
significantly. A clear solution was obtained after the reaching of
approximately 170 mol% [bmim][OcOSO3] to NHPI (Fig. 4, after
7th portion). The results obtained, in line with the solubility data
obtained by NMR, confirmed that the mixture of NHPI and
[bmim][OcOSO3] dissolves in a non-polar hydrocarbon much
better than NHPI alone. For comparison, the same experiments
were carried out with [bmim][NTf2] and PhCN. Adding over 5000
mol% of [bmim][NTf2] and 11000 mol% of PhCN to the
suspension of NHPI in benzene did not cause the NHPI to
dissolve.
However, it must be reminded that in all cases, there is no
indication of higher aggregates in the NMR spectra. Thus, the IL
in at 0.25 - 2% w/w concentration can be considered to be
forming homogeneous solutions in toluene-d8 with no phase
separation or pseudo phases (micelle, vesicles, etc…). The
addition of NHPI to IL (1% w/w) does not change much the
situation (Fig. 3b) as the IL diffusion coefficient D is unaltered (D
= 1.8 x 10-6 mol/cm2s). Under these conditions, NHPI can be
discriminated, although it features a very similar diffusion (D =
2.1 x 10-6 mol/cm2s). A safe assumption would be to consider
the two species closely interacting and forming an IL/NHPI
associated species. DOSY analysis on the sample IL 2% w/w
with NHPI, corresponding to spectrum h in Fig. 2, confirms
instead the NHPI induced formation of higher IL aggregates (Fig.
3c). The broad signals appearing indeed correspond to novel
diffusing species, whose diffusion coefficient is significantly
lower (D = 6.3 x 10-7 mol/cm2s), and whose dynamic radius
could be estimated around 60 Å.
The above NMR study therefore confirms that NHPI interacts
with [bmim][OcOSO3] ionic liquid with the latter most probably in
the form of small ion pair aggregates. The system remains
homogeneous at r.t. and thus it can be extrapolated that under
the oxidation conditions, no phase separation or vesicle
formation occurs.
Semi-quantitative information of the efficiency of a given IL in
making NHPI soluble in benzene were gathered by ATR FT-IR
technique. The height of the peak corresponding to the carbonyl
group (1732 cm-1) was measured upon addition of increasing
portions of [bmim][OcOSO3] at room temperature (Fig. 4 and Fig.
S2).
Table 4. EB oxidation with oxygen in the presence of Co(II) or NHPI/Co(II) and
[bmim][OcOSO3]
Entry
Co(II)
[bmim]
[OcOSO3]
[mol%]
α
[%]
SAP
[%]
SPEOH
[%]
SEBOOH
[%]
1[a]
2[a]
3[a]
4[a]
5[a]
6[a]
7[a]
8[a]
9[a]
10[a]
11
12
13
14
15
16
17
18
19
20
21
-
-
0
-
-
-
-
0.5
-
1.4
68.5
nd
12.8
nd
18.7
81.4
73.9
4.5
CoCl2
2.8
CoCl2
0.5
-
1.8
nd
nd
Co(C17H35COO)2
Co(C17H35COO)2
Co(acac)2
Co(acac)2
Co(OAc)2
Co(OAc)2
-
6.4
69.6
69.8
71.8
61.1
34.2
64.8
58.2
17.0
nd
22.6
19.7
23.5
35.8
9.3
0.5
-
11.0
7.8
8.4
3.7
0.5
-
12.3
3.4
2.2
53.0
9.0
0.5
-
8.4
23.1
4.1
3.9
36.9
73.7
87.5
41.7
3.6
-
0.5
-
5.5
8.5
CoCl2
6.5
nd
CoCl2
0.5
-
19.0
13.8
14.0
14.9
19.2
8.2
39.7
67.1
73.3
74.9
77.7
15.7
79.7
82.7
10.0
24.8
19.5
18.9
15.6
6.4
Co(C17H35COO)2
Co(C17H35COO)2
Co(acac)2
Co(acac)2
Co(OAc)2
Co(OAc)2
Co(OAc)2
0.5
-
2.3
3.3
0.5
-
3.3
72.6
5.0
0.5
1
30.1
35.2
11.3
7.4
2.9
Figure 4. Plot of the intensity of the IR band centered at 1732 cm-1 (NHPI
carbonyl group) versus time during the addition of [bmim][OcOSO3] IL in
portions (black dots represent time of each addition; black rhombus represent
clear homogenous solution). Benzene 10 g (128 mmol), NHPI 0.2 g (1.2 mmol,
1 mol% vs benzene), each portion of [bmim][OcOSO3] of about 0.3 mmol
EB 2 ml, NHPI 1 mol%, Co(II) 0.1 mol%, 0.1 MPa O2, 80°C, 6 h, 1200 rpm; [a]
without NHPI; nd – not determined
Finally, the recognition of the need to consider the present
catalytic mixture as a quite complex, multiple equilibria system
This article is protected by copyright. All rights reserved.

10.1002/cctc.201901737
ChemCatChem
FULL PAPER
led us to study the effect of Co(II) counteranion, seeking the
optimal condition for the EB oxidation process. Therefore, the
influence of the addition of [bmim][OcOSO3] on the oxidation of
in the system in the initial stages of the reaction (up to the 15th
minute). A negative influence was instead noticed when using
ILs
composed
of
the
lipophilic
cation
1-octyl-3-
EB in the presence of various Co(II) salts, with and without NHPI, methylimidazolium and [BF4], [PF6], and [NTf2] anions, despite
was studied. The results are presented in Table 4. Blank
experiments testing the activity of various Co(II) salts alone
show that conversion remains in the range 2.8 – 6.4 %, probably
following the different solubility of the metal salts in EB. Not
surprisingly, Co(C17H35COO)2 presents the highest conversion.
Confirming the overall picture resulting from the previously
described data, adding [bmim][OcOSO3] into the reaction
mixture caused an increase of EB conversion (Table 4, entries
3-10). The largest enhancement was observed with Co(OAc)2,
where this effect was accompanied by a decrease in selectivity
towards EBOOH, from 53.0% to 9.0%. Again, we interpret these
results in term of increased solubility of Co(II) ions in the
presence of IL. Interestingly, the addition of NHPI to the Co(II)/IL
systems caused a further increase in the conversion of EB
(Table 4, entries 13-20). The biggest effect was observed in the
CoCl2 and Co(OAc)2 cases. Notably, highest conversion and
the fact that 1 as well as the ILs were dissolved in the system,
however, this effect is easily interpreted as due to the interaction
of these counteranion with the Co(II) salts, as evidenced in the
previously described set of experiment. As the solubilizing effect
is now negligible, these ILs have only a detrimental effect.
Table 5. EB oxidation with oxygen in the presence of lipophilic derivatives of
NHPI, Co(acac)2 and lipophilic ILs
Entry
Additive
Catalyst
α
[%]
SAP
[%]
SPEOH
[%]
SEBOOH
[%]
1
2
3
4
5
6
7
8
-
1
1
2
2
1
1
1
1
29.2
34.9
11.6
17.8
29.3
26.9
24.9
22.5
73.9
87.1
27.5
66.8
84.6
78.5.
77.7
74.3
10.1
8.4
12.4
1.8
[bmim][OcOSO3]
-
3.3
59.0
11.5
1.6
selectivity
were
obtained
with
the
[bmim][OcOSO3]
[bmim][C12H25OSO3]
[omim][BF4]
[omim][NTf2]
[omim][PF6]
14.1
8.9
NHPI/Co(OAc)2/[bmim][OcOSO3] system and, to the best of our
knowledge, this represents the highest EB conversion ever
achieved for the oxidation of EB under mild solvent-free
conditions.
10.9
5.4
4.7
7.2
As we felt the need to confirm our thesis which attributes the
observed improvement on reaction conversion obtained by using
catalytic quantities of ILs to their capability to render NHPI
soluble in EB, we decided to test lipophilic NHPI derivatives 1
and 2 (Fig. 5) under our experimental conditions (viz. with
catalytic ILs present). We reasoned that these derivatives, much
more soluble in neat EB regardless the presence of an additive,
would respond to the presence of the IL in a way inversely
proportional to their lipophilic character. According to previously
reported data,[13] 2 is characterized by a much better solubility in
non-polar hydrocarbons than NHPI, but lower than that of 1.
14.0
8.7
EB 2 ml, 1 1 mol%, 2 0.5 mol%, Co(acac)2 0.1 mol%, additive 0.5 mol%, 0.1
MPa O2, 80°C, 6 h, 1200 rpm
Conclusions
In this study, we investigated the role and effect of the addition
of tiny amounts of imidazolium based ILs in solvent-free
oxidation of EB with oxygen in the presence of NHPI/Co(II)
catalytic systems. A positive influence of the IL which is not
ascribable to polar effect have been observed. The largest
increase in conversion was obtained when the ionic liquid with a
lipophilic anion [bmim][OcOSO3] was used. The IL resulted to be
able to promote the direct solubilization of NHPI in EB, without
thus requiring its chemical modification by introduction of
lipophilic chains on the aromatic ring. Among the cobalt(II) salts
the highest conversion of EB was attained when the system
composed of NHPI (1 mol%), Co(OAc)2 (0.1 mol%), and
[bmim][OcOSO3] (1 mol%) was applied: the conversion was
equal to 35% (only 8.2% without IL). Such catalytic efficiency
can be considered the highest reported to date in terms of
conversions of EB under mild conditions. The crucial role of
[bmim][OcOSO3] IL in promoting NHPI solubilization in the
Figure 5. Lipophilic NHPI derivatives 1 and 2.
In complete agreement with our expectations, the addition of
lipophilic [bmim][OcOSO3] to 1 caused a slight increase in EB
conversion from 29 to 35% (Table 5, entries 1 and 2), which
however is not present with [bmim][C12H25OSO3] (compare 29.2
vs. 29.3, entries 1 and 5). On the other hand, the IL has a more
marked effect when the catalyst used is 2. In this case,
conversion increases from 11.6 to 17.8 % (entries 3 and 4). It
was observed that after IL addition, the catalysts were dissolved
1
apolar medium was revealed by FT-IR and H-NMR analysis. In
particular, the latter technique confirmed the occurrence of a
direct interaction between the organocatalyst and small ion-pairs
aggregates of IL. The effect of the addition of IL in the reaction
This article is protected by copyright. All rights reserved.

10.1002/cctc.201901737
ChemCatChem
FULL PAPER
method[36], and the result was used for the calculation of the EBOOH
selectivity (SEBOOH).
medium significantly decreased when operating in the presence
of lipophilic catalysts 1 and 2, which are significantly more
soluble in EB than NHPI and for which thus the solubilization
effect is neglible. Furthermore, the choice of [bmim][OcOSO3] is
particularly beneficial as it is considered a “green” IL,[33] being
halogen-free, hydrolysis-stable and all the more so alkylsulfate
anions are known to exhibit good rates of biodegradation.[34]
푉푂 ∙ 273 ∙ 푝
2
푛푂
=
[푚표푙]
2
101325 ∙ 22.415 ∙ 푇
푛푂
훼 = 2 ∙ 100%
푛
푛푂푂퐻
푆퐸퐵푂푂퐻
=
∙ 100%
푛푂
2
nO2 –amount of oxygen consumed at normal conditions [mol], VO2
–
volume of oxygen consumed [dm3], p - pressure [Pa], T – room
temperature [K], α – conversion [%], n – amount of hydrocarbon used
[mol], SEBOOH – hydroperoxide selectivity [%], nOOH – amount of EBOOH
formed [mol]. EBOOH is thermally unstable and easily decomposes to
PEOH and AP. Thus, before GC analysis, the hydroperoxide was
quantitatively reduced to PEOH by addition of triethyl phosphite (EtO)3P,
which is oxidized to triethyl phosphate (EtO)3PO.[37] The amount of PEOH
determined by GC was the sum of the alcohol and hydroperoxide formed.
In order to calculate the PEOH selectivity, the amount of EBOOH
determined by iodometric analysis was subtracted. The amount of AP
was determined based on the GC analysis.
Experimental Section
Ethylbenzene (EB) (Acros 99.8%) was purified by washing with H2SO4
and vacuum distillation. Cobalt(II) acetylacetonate (Co(acac)2), cobalt(II)
chloride, cobalt(II) acetate tetrahydrate (Co(OAc)2·4H2O), cobalt(II)
stearate, N-hydroxyphthalimide
(NHPI), sodium lauryl sulfate and
benzonitrile (PhCN) were purchased from commercial sources and used
without purification. 4-Dodecyloxycarbonyl-N-hydroxyphthalimide (1) and
4,4’-(4,4’-isopropylidenediphenoxy)bis(N-hydroxyphthalimide) (2) were
synthesized according to a previously described procedure.[13,15] Ionic
liquids: 1-butyl-3-methylimidazolium chloride [bmim][Cl], 1-ethyl-3-
methylimidazolium bis(trifluoromethanesulfonyl)imide ([emim][NTf2]), 1-
GC analysis was performed using an Agilent Technologies 7890C gas
chromatograph (Zebron ZB-5HT capillary column) with an FID detector
and p-methoxytoluene as the internal standard. GC-MS analysis of the
products was performed using an Agilent 7890C gas chromatograph
(HP-5ms capillary column, 30 m x 0.25 mm x 0.25 µm, helium 1 ml/min)
coupled with an Agilent5975C mass spectrometer with EI (70 eV). The
products were identified using the NIST/EPA/NIH Mass Spectral Library.
The 1H NMR or 13C NMR spectra were recorded in deuterated chloroform
(chloroform-d1) or dimethyl sulfoxide (DMSO-d6) with a Varian 600 MHz
spectrometer or Agilent 400 MHz NMR. 1H-NMR titration data in Figure 2
and DOSY spectra were recorded with a Bruker Avance 400 MHz
spectrometer. DOSY pulse sequence ledbpgp2s was used (LED bipolar
gradient), NS=14, DS=4, D1 = 3 s, D16 = 0.2 ms; D20 = 60 ms; D21 =
5ms, P30 = 2.59 ms; data processing were performed with both Topspin
and Mestre Nova software. High-resolution electrospray ionisation mass
spectroscopy (ESI-HRMS) experiments were performed using a Waters
Xevo G2 QTOF instrument equipped with an injection system (cone
voltage 50 V; source 120 °C).
butyl-3-methylimidazolium
([bmim][NTf2]),
bis(trifluoromethanesulfonyl)imide
bis(trifluoromethanesulfonyl)imide
1-hexyl-3-methylimidazolium
([hmim][NTf2]),
([bmim][OcOSO3]),
([bmim][CH3COO]),
1-butyl-3-
1-butyl-3-
1-butyl-3-
methylimidazolium
methylimidazolium
octyl
acetate
sulfate
methylimidazolium trifluoromethanesulfonate ([bmim][CF3SO3]), 1-butyl-
3-methylimidazolium hexafluorophosphate ([bmim][PF6]), 1-butyl-3-
methylimidazolium
tetrafluoroborate
([bmim][BF4]),
1-octyl-3-
methylimidazolium hexafluorophosphate ([omim][PF6]), and 1-octyl-3-
methylimidazolium tetrafluoroborate ([omim][BF4]) were commercial
materials and were dried under vacuum before use (50°C, 0.1 bar). 1-
Octyl-3-methylimidazolium
([omim][NTf2]) was prepared according known procedure.[35] 1-Butyl-3-
bis(trifluoromethanesulfonyl)imide
methylimidazolium lauryl sulfate ([bmim][C12H25OSO3]) was prepared
according procedure described in literature
[33]
with small modifications.
[Bmim][Cl] (5.00 g, 28.6 mmol) and sodium lauryl sulfate (6.47 g, 22.40
mmol) were dissolved in water (20 ml) and mixed for 6 h at 60°C. The
water was slowly removed under vacuum and a white solid precipitated.
Dichloromethane (15 ml) was added to extract the product from the
mixture and the white solid filtered off. To the clear, slightly yellow filtrate
water (20 ml) was added to remove chloride impurities (according
procedure described in [33]). After a second addition of 20 ml water, it was
observed that the product started to foam and formed a white thick oil.
The ionic liquid was then extracted with large portions of CH2Cl2 (6x150
ml) which was removed in vacuum to give a colorless oil which
crystallized at room temperature. Yield 77%. 1H NMR (400 MHz, DMSO-
d6) δ ppm: 9.11 (s, 1 H), 7.77 (s, 1 H), 7.70 (s, 1 H), 4.16 (t, 2 H, J=8.0
Hz), 3.85 (s, 3 H), 3.67 (t, 2 H, 6.0 Hz), 1.76 (m, 2 H), 1.47 (m, 2 H),
1.33-1.19 (m, 20 H), 0.81-0.94 (m, 6 H) (Fig. S3). 13C NMR (400 MHz,
DMSO-d6) δ ppm: 136.50, 123.59, 122.25, 65.42, 48.47, 35.71, 31.33,
31.26, 29.14-28.90, 28.75, 28.68, 25.50, 22.06, 18.74, 13.91, 13.23 (Figs.
S4 and S5). ESI-HRMS: [bmim]+ found 139.1238 calcd: 139.1237,
[C12H25OSO3]- found 265.1471 calcd: 265.1502.
The solubility tests for the NHPI were performed on a Mettler-Toledo
iC10 FT-IR spectrometer equipped with an ATR probe. The ATR probe
was immersed into a suspension of benzene (10 g or 20 g) and NHPI
(0.2 or 0.4 g; 1 mol% vs benzene). Then, IL or PhCN was added in
portions, and the height of the peak corresponding to the carbonyl groups
(1732 cm-1) of NHPI was recorded in real time.
Acknowledgements
Financial support from the National Science Centre of Poland
(UMO-2014/13/B/ST8/04256) is gratefully acknowledged. M.C.
and C.P. thank MIUR for finacial support (Funds for fundamental
research, ANVUR call n. 20/2017, 15-06-2017).
The oxidation reactions were performed in a gasometric apparatus as
described in ref. [15]. EB, N-hydroxyimide, cobalt(II) salt, and the ionic
liquid were placed in a 10 ml flask connected to a gas burette filled with
oxygen under atmospheric pressure. The reaction was conducted at
Keywords: ethylbenzene • ionic liquids • N-hydroxyphthalimide •
oxidation • oxygen
80°C for 6 h with magnetic stirring at 1200 rpm. The oxygen uptake (nO2
)
[1]
[2]
Y. Ishii, S. Sakaguchi, in Modern Oxidation Methods (Ed. J.E.
Baeckvall), Wiley-VCH, Weinheim, 2004, pp. 119-163.
was measured, recalculated for normal conditions (273 K, 1 atm), and
this value was used to calculate the EB conversion (α). The amount of
EBOOH was determined iodometrically according to the described
F. Recupero, C. Punta, Chem. Rev. 2007, 107, 3800-3842.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201901737
ChemCatChem
FULL PAPER
[3]
[4]
[5]
[6]
S. Coseri, Catal. Rev. 2009, 51, 218-292.
L. Melone, C. Punta, Beilstein J. Org. Chem. 2013, 9, 1296-1310.
K. Chen, P. Zhang, Y. Wang, H. Li, Green Chem. 2014,16, 2344-2374.
R. Amorati, M. Lucarini, V. Mugnaini, G.F. Pedulli, F. Minisci, F.
Recupero, F. Fontana, P. Astolfi, L. Greci, J. Org. Chem. 2003, 68,
1747-1754.
[7]
[8]
Y. Ishii, S. Sakaguchi, Catal. Surv. Jpn. 1999, 3, 27-35.
Y. Tashiro, T. Iwahama, S. Sakaguchi, Y. Ishii, Adv. Synth. Catal. 2001,
343, 220-225.
[9]
B. Orlińska, J. Zawadiak, Reac. Kinet. Mech. Cat. 2013, 110, 15–30.
[10] C. Miao, H. Zhao, Q. Zhao, C. Xia, W. Sun, Catal. Sci. Technol. 2016,
6, 1378-1383.
[11] L. Melone, S. Prosperini, G. Ercole, N. Pastori, C. Punta, J. Chem.
Technol. Biotechnol. 2014, 89, 1370-1378.
[12] N. Sawatari, T. Yokota, S. Sakaguchi, Y. Ishii, J. Org. Chem. 2001, 66,
7889–7891.
[13] M. Petroselli, P. Franchi, M. Lucarini, C. Punta, L. Melone,
ChemSusChem 2014, 7, 2695-2703.
[14] M. Petroselli, L. Melone, M. Cametti, C. Punta, Chem. Eur. J. 2017, 23,
10616-10625.
[15] K. Kasperczyk, B. Orlińska, J. Zawadiak, Cent. Eur. J. Chem. 2014, 12,
1176-1182.
[16] N. Taha, Y. Sasson, Org. Proc. Res. Develop. 2010, 14, 701-704.
[17] R.D. Patil, B. Fuchs, N. Taha, Y. Sasson, ChemistrySelect, 2016, 1,
3791-3796.
[18] H. Weingärtner, Angew. Chem. Int. Ed. 2008, 47, 654-670.
[19] J.R. Wang, L. Liu, Y.F. Wang, Y. Zhang, W. Deng, Q.X. Guo,
Tetrahedron Lett. 2005, 46, 4647-4651.
[20] S. Koguchi, T. Kitazume, Tetrahedron Lett. 2006, 47, 2797-2801.
[21] I. Yavari, E. Karimi, Synth. Commun. 2009, 39, 3420–3427.
[22] T. Lu, M. Lu, W. Yu, Z. Liu, Croat. Chem. Acta 2012, 85, 277-282.
[23] Y. Wang, T. Lu, Chiang Mai J. Sci. 2014, 41, 138-147.
[24] M. Tokunaga, J.F. Larrow, F. Kakiuchi, E.N. Jacobsen, Science 1997,
277, 936-938.
[25] C.E. Song, W.H. Shim, E.J. Roh, S.-G. Lee, J.H. Choi, Chem. Commun.
2001, 1122-1123.
[26] C.E. Song, D.-U. Jung, S.Y. Choung, E.J. Roh, S.-G. Lee, Angew.
Chem. Int. Ed. 2004, 43, 6183-6185.
[27] D.S. Choi, D.H. Kim, U.S. Shin, R.R. Deshmukh, S.-G. Lee, C.E. Song,
Chem. Commun. 2007, 3467-3469.
[28] G. Dobras, B. Orlińska, Appl. Catal., A, 2018, 561, 59-67.
[29] S. Xu, C. Huang, J. Zhang, J. Liu, B. Chen, Korean J. Chem. Eng.,
2009, 26, 985-989.
[30] S. Mahmood, B.H. Xu, T.L. Ren, Z.B. Zhang, X.M. Liu, S.J. Zhang, Mol.
Catal. 2018, 447, 90-96.
[31] D. Pyszny, B. Orlińska, Reac. Kinet. Mech. Cat. 2018, 124, 123-138.
[32] M. Zanatta et al., Phys. Chem Chem. Phys. 2019, 21, 2567-2571.
[33] P. Wasserscheid, R. Hal, A. Bösmann, Green Chem. 2002, 4, 400-404.
[34] A. Jordan, N. Gathergood, Chem. Soc. Rev. 2015, 44, 8200-8237.
[35] C. Yao, J.L. Anderson, Anal. Bioanal. Chem. 2009, 395, 1491-1502.
[36] J. Zawadiak, D. Gilner, Z. Kulicki, S. Baj, Analyst 1993, 118, 1081-1083.
[37] P. Wiklund, C. Karlsson, M. Levin, Anal. Sci. 2009, 25, 431-436.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201901737
ChemCatChem
FULL PAPER
FULL PAPER
Gabriela Dobras, Magdalena Sitko,
Manuel Petroselli, Manfredi Caruso,
Massimo Cametti, Carlo Punta*, and
Beata Orlińska*
Page No. – Page No.
Solvent-free Aerobic Oxidation of
Ethylbenzene Promoted by
NHPI/Co(II) Catalytic System: The Key
Role of Ionic Liquids
The positive effect of the addition of catalytic amounts of [bmim][OcOSO3] in
NHPI/Co(II) catalyzed oxidation of ethylbenzene which is not ascribable to polar
effect have been reported. The IL promoted the direct solubilization of NHPI in
hydrocarbon by means of interaction with small ion-pairs aggregates of IL.
This article is protected by copyright. All rights reserved.