Ionic liquids vs conventional solvents: A comparative study in the selective catalytic oxidations promoted by oxovanadium(IV) complexes

Article abstract of DOI:10.1016/j.apcata.2020.117622

Two oxovanadium(IV) complexes containing 4-acyl-5-pyrazolonate-κ²-O,O' bidentate ligands with different chain lengths in the acyl moiety, namely HQ^C6 (complex I) and HQ^C17 (complex II), have been synthetized and full characterized, to study their catalytic activity toward the mild and selective oxidation of olefins or model organosulphur substrates, promoted by H₂O₂ or tert-butyl hydroperoxide (TBHP). The influence due to the type of solvent, likewise ionic liquids (ILs) or conventional medium, was analysed. H₂O₂ has proven less efficient, in comparison to TBHP. Quantum-chemical calculations have shown that, the key catalytic species involved and, consequently, the actual mechanism might be slightly different according to the oxidant used, i.e. the peroxo VO(Q^Me)(O₂) adduct in the case of H₂O₂ and the tert-butylperoxo VO(Q^Me)₂(OOt-Bu) adduct in the case of TBHP. Preliminary calculations suggested that, in ionic liquids, the VO(Q^Me)₂ complex might reveal relatively unstable, hence qualitatively explaining the moderate efficiency observed in these media.

Full text of DOI:10.1016/j.apcata.2020.117622

Applied Catalysis A, General 599 (2020) 117622
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
Ionic liquids vs conventional solvents: A comparative study in the selective
catalytic oxidations promoted by oxovanadium(IV) complexes
Patrizio Campitellia,b, Massimiliano Aschi , Corrado Di Nicola , Fabio Marchetti *,
a
b
b,
c
a,
Riccardo Pettinari , Marcello Crucianelli *
a
Department of Physical and Chemical Sciences, University of L’Aquila, Via Vetoio, I-67100 L’Aquila, AQ, Italy
School of Science and Technology, Chemistry Section, University of Camerino, Via S. Agostino 1, 62032, Camerino, MC, Italy
School of Pharmacy, Chemistry Section, University of Camerino, Via S. Agostino 1, 62032, Camerino, MC, Italy
b
c
A R T I C L E I N F O
A B S T R A C T
Two oxovanadium(IV) complexes containing 4-acyl-5-pyrazolonate-κ -O,O' bidentate ligands with different
2
Keywords:
Oxovanadium complexes
Olefin oxidations
Oxidative desulfurization
Acylpyrazolonate ligands
DFT calculations
C6
C17
chain lengths in the acyl moiety, namely HQ (complex I) and HQ
full characterized, to study their catalytic activity toward the mild and selective oxidation of olefins or model
organosulphur substrates, promoted by H or tert-butyl hydroperoxide (TBHP). The influence due to the type
of solvent, likewise ionic liquids (ILs) or conventional medium, was analysed. H has proven less efficient, in
(complex II), have been synthetized and
2 2
O
2 2
O
comparison to TBHP. Quantum-chemical calculations have shown that, the key catalytic species involved and,
consequently, the actual mechanism might be slightly different according to the oxidant used, i.e. the peroxo VO
Me
Me
(
Q
)(O
Preliminary calculations suggested that, in ionic liquids, the VO(Q
hence qualitatively explaining the moderate efficiency observed in these media.
2
) adduct in the case of H
2
O
2
and the tert-butylperoxo VO(Q
)
2
(OOt-Bu) adduct in the case of TBHP.
Me
)
2
complex might reveal relatively unstable,
1. Introduction
stereoselectivity, if proper chiral ligands are presents in the starting
complexes [17]. Often, Schiff base ligands are employed to coordinate
the vanadyl (VO)2+ group in the oxovanadium complexes but, fre-
quently, 4-acyl-5-pyrazolone ligands too have been investigated as
versatile coordinating systems [18], also because of biological interest
due to their analgesic and anti-inflammatory activity [19–21]. Mostly,
catalytic oxidation with oxovanadium compounds is performed using
The aptitude of high valent metal complexes to catalyze oxygen
transfer reactions is well known from several years [1–3]. Metal com-
plexes of Ti, V, Re, Mn or Cr are able to activate hydrogen peroxide as
well as organic peroxides such as t-butyl- or cumyl-hydroperoxide,
usually affording peroxy-, peroxo-, and oxo-type metal oxidants, thus
entailing an oxygen transfer mechanism [4]. As well, the increasing
interest observed in the last decades on the synthesis and reactivity of
several vanadium compounds has been justified also for their involve-
ment in biological processes [5,6] and medical applications [7], due to
the dimensional and structural analogy between vanadate and phos-
phate ions with the possibility to mimic the phosphate group in en-
zymes like phosphatases and kinases. In addition, the interest toward
oxovanadium containing derivatives has seen a remarkable growth also
for their proven activity as oxidant promoters in oxidative catalytic
reactions. In fact, oxovanadium (IV) and (V) complexes acting as oxo-
transfer agents [8], are able to catalyse several oxyfunctionalization
reactions such as epoxidation of alkenes [9,10], selective oxidation of
organic sulphides [11], hydroxylation of phenols [12,13], oxidations of
alcohols to aldehydes and ketones [14], hydro and oxidative amination
2 2
H O as main oxidant [22] either for olefin or sulphur oxidation [23],
because it is cheap and produces only water as by-product, but several
studies have been done also with tert-butyl hydroperoxide as oxidant
[24,25], showing high conversion and very good selectivity values. The
oxidation of organosulphides to their respective sulphoxides and sul-
phones catalysed by oxovanadium complexes has been investigated for
its potential application in the oxidative desulphurization (ODS) pro-
cess [26,27], also with heterogeneous polymer bound vanadium com-
plexes [28].
Since several years, ionic liquids (ILs) have attracted rising interest
with a diversified range of applications for several types of catalytic
processes [29] including those promoted by oxovanadium complexes
[30]. ILs present tuneable properties depending on the nature of cation
or anion and their very low vapour pressure and high thermal stability
make them versatile green solvents [31], with reduced environmental
[
15] and oxidation of bromide [16], eventually with some degrees of
⁎
Corresponding authors.
E-mail addresses: fabio.marchetti@unicam.it (F. Marchetti), marcello.crucianelli@univaq.it (M. Crucianelli).
https://doi.org/10.1016/j.apcata.2020.117622
Received 26 March 2020; Received in revised form 22 April 2020; Accepted 5 May 2020
Available online 08 May 2020
0926-860X/ © 2020 Elsevier B.V. All rights reserved.

P. Campitelli, et al.
Applied Catalysis A, General 599 (2020) 117622
R
O
V
R
N
O
N
O
O
H O/CH OH
2
3
+
VOSO4
N
O
O
R
OH
N
NaOCH , reflux
N
N
3
HQC6, R = C H
VO(QC6) (I), R = C H
VO(Q ) (II), R = C H
17 35
6
13
6
13
C17
C17
HQ , R = C H
17
35
Scheme 1. Synthesis of new oxovanadium complexes VO(QC6 (I) and VO(QC17
) ) (II).
2 2
impact. An important property of ILs is their recyclability, being in
theory possible to catch in their phase suitably selected organometallic
complexes, thus allowing the simple recovery and reuse of precious
catalysts [32]. Among others, a simple way to increase the entrapment
of a metal complex, acting as a catalyst, within the IL phase is that of
properly functionalize the ligands, for example with different lengths of
the alkyl side-chains, thus varying the resulting hydrophobicity/hy-
drophilicity of the catalyst backbone, to discriminate its affinity for the
chosen IL. Potentially, the role of this tag is to magnify the catalyst’s
partition coefficient in an appropriately selected IL. For this reason, we
prepared initially two novel 4-acyl-5-pyrazolone based ligands having
70 % w/w aqueous solution (TBHPaq) or a 5.5 M solution in n-decane of
tert-butyl hydroperoxide (TBHP), and a 50 % w/w aqueous solution of
H O , were used as primary oxidants.
2 2
2.1.2. Analytical methods
Elemental analyses (C, H, N, S) were performed in-house with a
Fisons Instruments 1108 CHNS-O Elemental Analyzer. The samples for
microanalyses were dried in vacuo to constant weight (20 °C, ca.
0
.1 Torr). Solid-state infrared spectra were recorded on a Perkin Elmer
1
13
SpectrumOne FT-IR spectrometer. H and C NMR spectra were re-
corded, at 25 °C, with a Bruker Avance III 400 MHz instruments. When
necessary, H and C NMR analyses of products were performed after
C6
C17
different chain lengths in the acyl moiety, namely HQ and HQ
1
13
(
containing 6 and 17 carbon atom, respectively), and then their cor-
flash-chromatographic purification on columns packed with silica gel
C6
responding oxovanadium(IV) complexes, i.e. VO(Q
2
) (I) and VO
(
230–400 mesh), and compared with authentical sample. Microwave
C17
(
Q
)
2
(II), as described in the Scheme 1.
+
studies were performed with the Biotage® Initiator microwave syn-
thesizer. The magnetic susceptibilities were measured at room tem-
perature (20 °C) by the Gouy method, with a Sherwood Scientific
C17
The proligand HQ
has been recently used by some of us to afford
Zn(II) complexes with antiproliferative activity against human breast
MCF‐7 cancer cells [33] and (arene)Ru(II) complexes with very potent
cytotoxicity against to human ovarian A2780 carcinoma cells and the
A2780cisR variant with acquired resistance to cisplatin [34], to tether
the same (arene)Ru(II) complexes to polystyrene surfaces providing
potent antibacterial plastics [35] and also to immobilize lanthanide
complexes on the surface of pre-modified silica via hydrophobic inter-
actions affording hybrid luminescent films [36]. Following our ongoing
interest in the oxovanadium chemistry [37–39], and considering the
well-established aptitude of oxovanadium(IV) complexes of promoting
the activation of oxidants, especially of peroxidic nature [40], the
catalytic activity of the new oxovanadium complexes (I) and (II) has
been investigated in the selective oxidation of conjugated (styrene) or
unconjugated (cis-cyclooctene) olefins and of model organosulphur
derivatives as dibenzothiophene (DBT) and 4,6-dimethyldibenzothio-
phene (DMDBT), having in mind to compare their activity and se-
lectivity between traditional ILs and conventional solvents. Quantum
chemical DFT calculations have also been carried out to shed light on
the catalytic mechanism potentially active in our experimental condi-
tions.
Magnetic Balance MSB-Auto, using HgCo(NCS)
4
as calibrant and cor-
recting for diamagnetism with the appropriate Pascal constants. The
magnetic moments (in BM) were calculated from the equation μeff
=
corr ―1/2
corr
2.8428 (χ
m
T)
(χ
m
is the molar susceptibility corrected for
diamagnetism and T is the temperature in kelvin degrees). Gas chro-
matographic (GC) analyses were performed by means of
a
Hewlett–Packard 6890 series instrument equipped with a flame ioni-
zation detector (FID), using a 30 m ×0.32 mm × 0.25 μm film thick-
ness (crosslinked 5% phenyl-methylsiloxane) column and chromato-
graphy grade helium, as carrier gas. In GC calculations, all peaks
amounting to at least 0.5 % of the total products were considered.
UV–vis analyses were performed with a Varian Cary 50 UV–vis
Spectrophotometer, working with a thermostated UV quartz cell, at
34 °C. The 9.5 GHz Electron Paramagnetic Resonance (EPR) X-band
spectrometer was a conventional assembly of Bruker units (Bruker
Spectrospin, Milano, Italy) constituted by: a magnet B-M8, a resonant
cavity 4108 TMH/9101, and a microwave bridge ER040XR, equipped
with the field controller BH15. TGA-DTA spectra were obtained with a
STA 6000 Simultaneous Thermal Analyzer Perkin-Elmer. Inductively
coupled plasma atomic emission spectrometry (ICP-AES) analyses were
performed using a Varian Liberty 200 spectrophotometer, to determine
the vanadium content.
2. Experimental section
2
2
.1. Materials and methods
2
.2. Synthesis of complex VO(QC6
2
) (I)
.1.1. Materials
All reagents (DBT, DMDBT, styrene, cys-ciclooctene) and solvents
NaOCH
(0.8 mmol, 229 mg) in MeOH (10 mL) in a 50 mL flask. Then VOSO
(0.4 mmol, 65 mg) was solubilized in 5 mL of water and the solution
3
(0.8 mmol, 43 mg) was added to a solution of HQC6
were purchased from Sigma-Aldrich (Italy) in the highest purity grade
available and were used as such. All ionic liquids, namely 1-butyl-3-
4
methylimidazolium tetrafluoroborate BMIM-BF
dazoliumhexafluorophosphate BMIM-PF 1-butyl-3-methylimidazo-
lium trifluoromethansulfonate BMIM-OTf, were purchased from Iolitec
4
, 1-butyl-3-methylimi-
was added drop by drop into the flask. The reaction was hold under
6
,
magnetic stirring for 4 h in reflux conditions. After filtration, a pale
C6
yellow solid of VO(Q
2
) compound I, was obtained, being it partially
C6
C17
(
https://iolitec.de). The HQ and HQ
proligands have been syn-
soluble in chlorinated solvent, acetonitrile and poorly soluble in MeOH.
thesized according to previously reported methods [19,20,36]. Either a
Yield: 77 %; M.p. 255−257 °C. Elemental Analysis calcd (%) for [VO
2

P. Campitelli, et al.
Applied Catalysis A, General 599 (2020) 117622
(
(
+
QC6
KBr, cm ): 1633 m, 1611s ν(C = O), 1581vs, 1540vs ν(C = N
C = C); 898vs ν(V = O); 690 s, 475 m ν(V-O); 659 s δ(C-Me) + ν(V-
O); 627vs π(C-Me). μeff (296 K): 1.79 μ . ESI-MS (MeOH/CH Cl ) (+):
m/z (%) = 287 (15) [H(HQ )] , 309 (25) [Na(HQC )] , 360 (100)
)
2
]: C, 64.04; H, 6.64; N, 8.79. Found: C, 64.38, H, 6.51; N, 8.84. IR
5 min. After that, 2 equiv. of oxidants are inserted into the vial. The
reaction was hold under heating and magnetic stirring up to 6 h, adding
further equiv. of oxidant at regular interval of time (1 h) up to 4.0
equiv. and monitoring its progress through GC-FID analysis, with-
drawing each time an aliquot (20 μL) of the hydrocarbon phase and
adding 5 μL of n-hexadecane (internal standard). At the end of the re-
action, the n-octane phase was separated, and the IL phase was washed
with 3 mL of diethyl ether. The two organic phases were afterwards
−
1
B
2
2
C6
+
6
+
6
++
[
(VO)
2
O(QC )
2
]
. TGA-DTA (mg% vs. °C): heating from 30 to 800 °C
with a speed of 10 °C/min; at 255 °C onset fusion (ΔHfusion =68.29 kJ/
mol); from 290 to 490 °C progressive decomposition, with a final black
residual of 10 % weight.
2
joined together, quenched with MnO as above described, and the final
solution was concentrated using a rotary evaporator, with gentle
heating (40 °C), and then analysed with GC-FID.
2
.3. Synthesis of complex VO(QC17
2
) (II)
NaOCH
3
(0.5 mmol, 27 mg) was added to a solution of HQC17
2.7. Microwave-assisted Catalytic oxidation
(
(
0.5 mmol, 220 mg) in MeOH (8 mL) in a 50 mL flask. Then VOSO
0.25 mmol, 41 mg) was solubilized in 2.5 mL of water and the solution
4
The sample was prepared as above described and was inserted into
the appropriate MW glass vials (0.5–2.0 mL) with the stirring bar,
added with 2.0 equiv. of oxidant and then sealed. The experiment was
performed under control of the temperature (70 °C), by setting the
following instrumental parameters for each cycle: time =20 min, ab-
sorption level = normal, stirring rate =900 rpm. The progress of the
reaction was monitored through GC-FID analyses. At the end of the
reaction, we followed the same work-up procedure as above described.
was added drop by drop into the flask. The reaction was hold under
magnetic stirring for 4 h in reflux conditions. After filtration, a pale
yellow solid of VO(Q
chlorinated solvents and acetonitrile while only partially soluble in
methanol. Yield: 84 %; M.p. 182−184 °C. Elemental Analysis calcd (%)
for [VO(Q
5
C17
2
) compound II was obtained, being it soluble in
C17
2
) ]: C, 71.08; H, 9.16; N, 5.92. Found: C, 71.45, H, 9.07; N,
−1
.87. IR (KBr, cm ): 1635 m, 1614s ν(C = O), 1575vs, 1537s ν(C = N
+
C = C); 900vs ν(V = O); 690 s, 482vs ν(V-O); 666 m δ(C-Me) + ν(V-
O); 626vs π(C-Me). μeff (296 K): 1.91 μ
B
. ESI-MS (MeOH/CH Cl ) (+):
2
2
2.8. Recycling tests of catalyst I
C17
+
C17
+
m/z (%) = 441 (90) [H(HQ )] , 570 (60) [VO(Q )(MeOH)
2
] , 946
H] . TGA- DTA (mg% vs. °C): heating from 30 to
00 °C with a speed of 10 °C/min; at 182 °C onset fusion (ΔHfusion
40.41 kJ/mol); from 300 to 520 °C progressive decomposition, with a
final black residual of 18 % weight.
C17
+
(
100) [(VO)(Q
)
2
At the end of the first reaction cycle, using DBT as substrate and
8
=
TBHP as oxidant, in a n-octane/BMIM-PF
otherwise working under the same conditions described in 2.6 section,
the upper organic phase was recovered, quenched with MnO as above
6
(5:1) solvent mixture,
2
described, and its residual sulphur content analysed by GC-FID. Next,
new fresh DBT (0.1 mmol) portion in 1.0 mL of n-octane was added to
the recycled IL phase and a new reaction cycle, started. After the fifth
cycle, all the previously recovered organic phases were collected and
analysed by ICP-AES, in order to evaluate the residual amount of va-
nadium metal, eventually due to its partial leaching.
2
.4. EPR study
Under typical conditions, the spectrometer operated at a central
magnetic field of 355 m T (3550 G), scan range 110 m T (1100 G),
sweep time 165 s, time constant 100 ms, modulation frequency
100 kHz, modulation amplitude 0.1 m T (1 G), at room temperature
(
298 K). The effective 28 mW microwave power on the samples was
2.9. Computational details
generated by a 280 mW microwave power source (10 dB attenuation).
The analyzed samples were prepared by dissolving 1.0 mg of samples in
All the calculations were carried out adopting the following pro-
tocol. The structures were first optimized in gas-phase and character-
ized as minima or first order saddle points through the calculation of
the mass-weighted Hessian. Since, at least in principle, different mag-
netic states can be involved in the present reaction, we also investigated
their possible couplings.
1
.0 mL of acetonitrile, and spectra were recorded at room temperature
298 K). After addition of single drop portions of a 35 % aqueous so-
lution of H to the previous solution, the successive EPR spectra were
recorded, at the same temperature, after 5 min.
(
2 2
O
2
.5. Catalytic oxidation in conventional solvents
It is important to remind that two surfaces with different spin
multiplicity cross along a hyperseam whose minimum, termed as
Minimum Energy Crossing Point (MECP), can be roughly considered to
act as a Transition Structure (TS) for these processes termed as non-
adiabatic. MECPs were obtained, in the gas-phase, using standard
procedures as described elsewhere [41].
The reaction and activation free-energies at 60 °C were then calcu-
lated in the ideal gas approximation using the standard condition of
1.0 mol/liter and then corrected with the solvation free energies esti-
1.5 mL of solvent, 0.1 mmol of sulphide (or 0.2 mmol of olefin),
1
.0 mol % of catalyst and 2 equiv. of oxidant were inserted sequentially
in a 5 mL flask equipped with a magnetic stirring bar, immersed in a
thermostatic oil bath, at 60 °C. Further equivalents of oxidant were
added stepwise at regular interval of time (1 h) during the reaction, up
to 4.0 equiv. The reaction was allowed to react for different intervals of
time depending on the types of used substrates. At the end of reaction,
mixture was quenched with 5.0 mg of MnO
2
and, after 15 min, filtered.
mated, for CH
Polarizable Continuum Model (C-PCM) [42,43].
For CH CN we utilized the dielectric constant of 31.0 corresponding
to the value at 60 °C. All the calculations were performed using Density
Functional Theory with the CAM-B3LYP functional [44] with the
3 2 2
CN also for CH Cl , in the framework of Conductor-like
The progress of the reaction was monitored through GC-FID analysis,
withdrawing each time an aliquot of 20 μL and adding 5 μL of n-hex-
adecane (internal standard). The oxidation products were identified by
comparison of their GC retention times with those of authentic samples.
3
6−31+G* basis set for the optimizations and the 6−311+G* for the
2.6. Catalytic oxidation in ionic liquids
C-PCM calculations. The vanadium has been always treated with the
Hay and Wadt Effective Core Potential (ECP) [45].
0
.1 mmol of sulphide (or 0.2 mmol of olefin) and 1.0 mol % of
Because of the semi-quantitative character of our computational
investigation, due to the lack of quantitative experimental data to
compare to, we did not perform any sensitivity analysis with respect to
the quality of the functional and the size of the basis set.
catalyst were added to 1.2 mL of a mixture of n-octane/ionic liquid 5:1
v/v for sulphides, or 1.0 mL of ionic liquid alone for olefin, respectively,
in a glass vial equipped with a magnetic stirring bar, immersed in a
thermostatic oil bath at 60 °C, and the biphasic mixture was stirred for
Finally we also attempted a computational investigation of the same
3

P. Campitelli, et al.
Applied Catalysis A, General 599 (2020) 117622
reactions in ionic liquids. However our efforts were frustrated by the
high complexity of the environmental effects which are very difficult to
model with a mean field approximation and would require a deep in-
vestigation well beyond the scope of the present study [46].
of a series of oxovanadium(IV) 4-acyl-5-pyrazolone complexes, pre-
viously reported [37]. The analysis of UV–vis spectra of complex I
treated with TBHP, showed a clear shift toward shorter wavelength
values (from 282 to ca. 270 nm).
In this respect we have only addressed some preliminary in-
vestigation on some species conceivable present in these conditions.
An attempt to rationalize the results implies that, being the most
intense band (π-π* transition) moving from the complex alone to the
complex activated in the presence of the oxidant (TBHP), which implies
the coordination of deprotonated TBHP to the metal thus expanding the
coordination sphere of vanadium(V) species (confirmed by calculations
too, see Fig. 8 onward). This necessarily causes a partial weakening of
the interaction between the metal and the carbonyl moiety of the li-
gand, thus tentatively justifying the increasing of the energy of the π-π*
transition, and consequently the shift of the observed absorption. It
should be noted that the UV–vis spectra of titrated solution of the ox-
ovanadium complex I, remained unchanged after 24 h (Fig. 1). Inter-
estingly, in the UV–vis solid phase spectra of both complexes I and II,
3. Results and discussion
3
.1. Synthesis of complexes VO(QC6
)
2
and VO(QC17
Complexes VO(QC6 (I) and VO(QC17
) ) (II) were synthesized by
2 2
)
2
reacting vanadyl sulphate with the acylpyrazolones in basic methanol/
water at room temperature (Scheme 1). They are pale-yellow solids
stable to air with medium-to-good solubility in chlorinated solvents and
acetonitrile. Complexes I and II display sharp and very different melting
points to one another, in fact complex II melts at a temperature of
all expected transitions can be assigned, namely: ca. 280 nm (π
→
π*),
1
82−184 °C, much lower than that of complex I (255−257 °C), clearly
ca. 327 nm (n
→
π*), ca. 421 nm (LMCT), ca. 560 nm (d-d) (see Figures
C17
due to the long aliphatic chain of Q
crystalline packing.
ligand in II that decreases the
S.1 and S.2 in the Supplementary Material section).
3.5. Thermogravimetric analyses
3.2. FT-IR characterization
2
Thermogravimetric analyses (TGAs), under N atmosphere, were
Solid-state FT-IR spectra have been recorded for the free ligands
performed on I and II revealing a significative difference in their onset
melting points, at 255 °C and 182 °C for I and II, respectively. This is
likely due to a superior crystalline packing aptitude of complex I
compared to II, in which the long aliphatic chain counteracts crystalline
packing. After melting, both I and II show an initial decomposition
around 280 °C and then a rapid weight loss owed to a complete thermal
decomposition, leaving a small black residual (10–18 %).
HQC6 and HQ
C17
and their complexes VO(Q
C6
)
2
(I) and VO(Q
C17
2
) (II).
The disappearance in the spectra of I and II of the broad ν(OeH) ab-
−1
sorption between 2800 and 2300 cm
and the shift of the ν(C]O)
toward lower wavenumbers with respect to free ligands [from 1638 to
−
1
C6
C6
1
1
611 cm
614 cm
comparing HQ
and VO(Q
)
2
and from 1626 to
] is in accordance with
-chelating
−1
C17
C17
comparing HQ
and VO(Q
)
2
deprotonation of HQ and bidentate coordination of the O
2
face to vanadium. Moreover a structural information arises from the
3.6. EPR measurements for complexes I and II
stretching vibration of the ν(V]O), which typically looks like a strong
−
1
signal observed in the range 880–1000 cm
[47]: in fact, when a
In order to confirm, by X band EPR analyses, the rapid oxidation of
molecule of water is coordinated in trans position to the oxo ligand of
vanadium from +4 to +5 oxidation state, a well-known process oc-
vanadyl the ν(V]O) stretching band falls at frequencies higher than
2 2
curring for the oxovanadium(IV) complexes after H O addition, we
−
1
9
8
60 cm , whereas in the anhydrous form this band falls in the range
focused our attention on both complexes I (Fig. 2) and II (see Figure
S.3, in the Supplementary Material section), analysed in a 1.0 M acet-
onitrile solution. As expected, EPR spectrum of I and II, at 298 K, ex-
hibited an expected hyperfine structure (HFS) derived from the inter-
−
1
90–920 cm
[37,48]. Complexes I and II show the ν(V]O) at 898
−
1
and 900 cm respectively, indicating the absence of coordinated water
and in agreement with the thermogravimetric analysis where any
weight loss corresponding to water molecules was detected. The low-
ering of the ν(V]O) value may arise from the molecular stacking in the
solid state which can favour the interaction between an oxygen of the
vanadyl cation with the vanadium center of another neighbour mole-
cule, thus allowing the saturation of the sixth coordinative valence of
the metal [37,49].
1
51
action of free electron (3d ) with the magnetic nuclear moment of
V
(I = 7/2, 99.76 % of natural abundance), thus confirming the presence
IV
of isolated V species as applying to a mononuclear complex. In this
case, the EPR signal splits into eight-fold lines of all anisotropic com-
ponents (Fig. 2) with values of giso = 1.97 and Aiso = 109.5 G. This
2+
values could correspond to vanadyl ions (VO ) in octahedral distorted
symmetry [51]. Furthermore, the g-value lies within the 1.9–1.99 range
and agrees with the fact that when the d-shell is less than half filled, the
spin–orbit effect reduces the g-value in comparison with that of a free
electron (2.0023) [52]. By the same way, similar measurements were
carried out on the same complex I after dropwise titration with a 35 %
3.3. Magnetic measurements
The room-temperature magnetic susceptibility obtained for I and II
is 1.79 BM and 1.91 BM, respectively, a typical paramagnetic behavior
1
within the normal range for the spin only contribution of d systems,
2 2
aqueous solution of H O , in order to mimic catalytic reaction medium.
where the orbital contribution is completely quenched [50].
The new EPR spectrum showed the disappearance of previous well
defined HFS, as expected when the oxidation toward oxovanadium(V)
complex derivatives, occurred (see Fig. 2, red line). Results fully com-
parable were obtained in the analysis of the complex.
3.4. Study of UV–vis electronic spectra
The UV–vis electronic spectra of the oxovanadium complex I and
C6
that of the corresponding proligand HQ , have been performed in di-
chloromethane. For the ligand one main band having the maximum at a
wavelength of 271 nm has been observed, within the range of π–π*
transition of the phenyl group, being the peak of the n–π* transition of
the carbonyl group in the acyl fragment (ca. 230−233 nm) not de-
tectable inside the windows we used (Fig. 1). Upon coordination, this
absorption band is red shifted to form a widened band with the max-
imum at 282 nm, indicating the involvement of the carbonyl group in
the complex formation with the metal ion, according to the behaviour
3.7. Study of the catalytic activity
3.7.1. Selective oxidation of styrene and cis-cyclooctene (CO)
As model substrates for the study of the catalytic oxidation of con-
jugated or unconjugated olefins, we focused our attention on styrene
and cis-cyclooctene (CO), as examples of electron poor or more electron
rich olefins, respectively. With the aim of finding the best experimental
parameters, provided that in accordance with our aims to work under
mild conditions, we took in consideration the results previously
4

P. Campitelli, et al.
Applied Catalysis A, General 599 (2020) 117622
Fig. 1. Electronic spectra of HQC6 and VO(QC6
)
(I). The reactivity of I with TBHP, is also showed. Spectra were recorded after the successive addition (1.0 μL) of
2
2 2
5.5 M TBHP solution in n-decane, to 10 mL of CH Cl solution of I (0.4 mg).
2 2
in the H O promoted oxidation of styrene with high-valent oxo-metal
based catalysts containing Re(VII) [54] or V(IV) [37]. Scheme 2 de-
scribes the known reaction mechanisms generally operative under the
catalytic oxidation of styrene, promoted by hydroperoxides. Frequently,
depending on the experimental conditions (oxidant, solvent, etc.) the
styrene epoxide C is converted to the corresponding benzaldehyde A
due to the H O addition over the initially formed epoxide and suc-
2 2
cessive cleavage of the hydroxyl-hydroperoxystyrene intermediate (not
shown) [55]. In addition, for the two conceivable competing channels
(
epoxidation and oxidative cleavage, Scheme 2), previous theoretical
calculations showed that, under kinetic control and with H excess,
2 2
O
the epoxidation route is the less efficient process [56]. The excess of
oxidant is responsible for the overoxidation of benzaldehyde A to
benzoic acid B, while traces of by-products D and E come from the
benzylic oxygen CH insertion/isomerization on styrene and the partial
hydrolysis of the epoxide C, respectively. For what it concerns both the
catalytic activity and selectivity, the two oxovanadium(IV) complexes
Fig. 2. EPR spectrum of oxovanadium complex I in acetonitrile, at 298 K.
obtained in the study of the catalytic oxidation of the challenging
styrene, with oxovanadium(IV) complexes, promoted by hydrogen
C6
C17
VO(Q
)
2
or VO(Q
2
) acted in a similar way, with TON ranging from
peroxide (H
2
O
2
) as main oxidant, working at room temperature (r.t.)
80 to 85 (Table 1, entries 1–2), being these values comparable with
[
37]. Thus, after the starting optimization study, we opted for the use of
results previously described in literature [55]. The disappointing results
observed when chlorinated solvents (dichloromethane or chloroform)
are used with H O as oxidant [25,57], have been confirmed also under
2 2
our experimental conditions (Table 1, entries 3 and 5): an attempt to
force the conversion of s.m. with the aid of microwave (MW) assisted
heating, provided equally negative results (Table 1, entry 4). Quantum
chemical calculations, carried out in the framework of Density Func-
tional Theory (DFT), furnished a valid rational for these experimental
results (see onward). On the other hand, the use of t-butyl hydroper-
oxide (TBHP) as main oxidant, either in decane or aqueous solution
aqueous H
2
O
2
and either, aqueous or decane solutions of tert-butyl
hydroperoxide (TBHPaq or TBHP, respectively), as main oxidants, and
acetonitrile or CH Cl and CHCl , as conventional solvents. According
to our objective, we decided to work with moderate temperature (up to
0 °C) and moderate amounts of both oxidant (up to 4.0 equiv.) and
2
2
3
6
catalyst (1.0 mol%), within up to 6 h of reaction time. In all experi-
ments, the oxidant has been added portionwise to the reaction mixture
1.0 equiv. per hour, until the maximum amount described) to avoid its
(
decomposition, as already reported in the case of H under vanadium
2 2
O
promoted catalysis [22,53]. The experimental results obtained in the
oxidation of styrene with the main observed products (Scheme 2), are
summarized on Table 1.
(
TBHPaq), allowed to increase the selectivity and yield of styrene oxi-
dation toward benzaldehyde A, both in acetonitrile and chloroform,
reaching in the former solvent quantitative conversions of s.m. (Table 1,
entry 6). In terms of selectivity, interesting differences between acet-
onitrile and chloroform have been observed. In acetonitrile, the for-
mation of epoxide C (or the corresponding 1-phenylethane-1,2-diol E,
when using TBHPaq due to the epoxide hydrolysis) has been detected
with comparable yield and selectivity with both catalysts I and II
The starting catalytic experiments, carried out with H
2 2
O in acet-
onitrile, gave good conversion (80–85 %) of the starting material (s.m.),
C6
C17
irrespective to the oxovanadium complex VO(Q
II) used, being benzaldehyde A and benzoic acid B, the main products,
obtained with comparable yields (33–39 %) and selectivities (41–46
), along with traces of acetophenone D and 1-phenylethane-1,2-diol E
)
2
(I) or VO(Q
)
2
(
%
(
(
Table 1, entries 7–8, vs. entry 6). Differently, in chloroform aside the
Table 1, entries 1–2, and Scheme 2).
The selectivity was in accordance with results previously obtained
main aldehyde A, a fairly good amount of acetophenone D was
5

P. Campitelli, et al.
Applied Catalysis A, General 599 (2020) 117622
O
O
oxidation
H
OH
oxidative
cleavage
A
B
1
2
. oxidant addition
. cleavage
epoxidation
OH
OH
O
H O
2
C
E
CH insertion
and isomerization
O
CH3
D
Scheme 2. Main products obtained from styrene catalytic oxidation.
Table 1
Styrene oxidation catalysed by VO(QC6
)
2
and VO(QC17
Solvent
)
2
complexes.
Conv. (%)
Entrya
Catalyst
Oxidant
TONb
Products (% Y)c [% S]c
A
B
C
D
E
(see Scheme 2)
1
2
3
VO(QC6
)
2
H
H
H
H
H
2
2
2
2
2
O
O
O
O
O
2
2
2
2
2
MeCN
MeCN
85
80
3
10
5
98
90
93
76
78
90
94
80
86
90
87
85
80
3
10
5
98
90
93
76
78
90
94
80
86
90
87
(36) [42]
(33) [41]
(3) [100]
(10) [100]
(5) [100]
(60) [61]
(60) [67]
(61) [66]
(46) [60]
(43) [55]
(29) [32]
(23) [24]
(15) [19]
(20) [23]
(30) [33]
(28) [32]
(39) [46]
(34) [43]
–
–
–
(13) [13]
(10) [11]
(12) [13]
(5) [7]
(7) [9]
(55) [61]
(66) [70]
(62) [77]
(54) [63]
(60) [67]
(59) [68]
–
–
–
–
(3) [4]
(5) [6]
–
–
–
–
–
–
(7) [8]
(8) [10]
–
–
–
–
C17
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
)
2
C6
C6
C6
C6
C6
)
2
)
2
)
2
)
2
)
2
CH
CH
2
Cl
Cl
2
d
4
2
2
5
6
7
8
9
CHCl
3
–
TBHP
MeCN
MeCN
MeCN
(25) [26]
–
–
–
TBHPaq
TBHPaq
TBHP
(20) [22]
(20) [21]
C17
C6
)
2
)
2
CHCl
CHCl
3
(25) [33]
(28) [36]
–
–
–
–
–
–
–
–
C17
C6
10
11
12
13
14
15
16
)
2
TBHP
3
–
)
2
)
2
)
2
)
2
)
2
H
H
2
O
O
2
BMIM-BF
BMIM-OTf
BMIM-BF
BMIM-OTf
BMIM-PF
BMIM-PF
4
(6) [7]
(5) [6]
(3) [4]
(12) [14]
–
–
–
–
–
–
–
C6
C6
C6
C6
2
2
TBHP
TBHP
TBHP
TBHP
4
6
C17
)
2
6
–
a
Experimental conditions: styrene 0.2 mmol; catalyst =1.0 mol%; oxidant = 4.0 equiv.; reaction time: 6 h; temperature: 60 °C; volume: 1.5 mL for conventional
solvents and 1.0 mL for ILs.
b
TON (turnover number): mmoles of converted substrate per mmole of catalyst.
Y = yield; S = selectivity; yields are calculated on converted substrate.
MW-promoted reaction: 2 cycles of 20 min, 70 °C, absorption level = normal, stirring rate =900 rpm.
c
d
detected, irrespective to the type of used catalyst I or II, being the
conversion of s.m. lower than that in acetonitrile (Table 1, entries 9–10
vs. 6–8). In the first instance, this different outcome may be attributable
to the different polarity and donor ability properties between acetoni-
trile (higher) and chloroform (lower), which could affect the reactivity
of active species produced in consequence of the interaction among
oxidant and oxovanadium complex I or II (see onward).
2 2 4
H O and complex I in hydrophilic ILs like BMIM-BF and BMIM-OTf,
gave the highest conversions of styrene (90–94 %), with a different
selectivity between aldehyde A and benzoic acid B, the latter being
obtained in major yield (55–66 %), in comparison with acetonitrile,
aside traces of epoxide C (Table 1, compare entries 11–12 vs. 1–2). The
study performed with TBHP showed the following order of activity
BMIM-PF
s.m., although with minor differences (ranging from 80 % to 90 %), the
selectivity being the same as that in the case of H (Table 1, compare
entries 13–15 vs. 11–12). Meaningfully, no substantial differences have
6
been observed by using the catalyst II with hydrophobic BMIM-PF ,
6 4
> BMIM-OTf > BMIM-BF [58], in terms of conversion of
Having in mind to evaluate the discriminating role, if any, of
common alkylimidazolium ionic liquids (ILs) like BMIM-BF
and BMIM-OTf, in terms of their hydrophilic (BMIM-OTf and BMIM-
BF ) or hydrophobic (BMIM-PF ) properties and potential coordinating
ability of their anions toward the vanadium centre
) [58], we started to study the oxidation of
styrene in these solvents, with catalyst I. Then, the system formed by
4
, BMIM-PF
6
2 2
O
4
6
concerning both the conversion of substrate and the selectivity of the
oxidation (Table 1, entry 16 vs. entry 15). This could confirm that,
within the oxovanadium complex, the alkyl side-chain in the acyl
−
−
−
(
BF
4
> TfO > PF
6
6

P. Campitelli, et al.
Applied Catalysis A, General 599 (2020) 117622
O
epoxidation
H O
2
F
OH
OH
CH insertion
and isomerization
H
O
G
Scheme 3. Main products obtained from cis-cyclooctene catalytic oxidation.
moiety of 4-acyl-5-pyrazolone ligands, is too far from the catalytically
active metal centre to be able to somehow affect the outcome of cata-
lytic reaction, at least in terms of different steric hindrance. Overall, it
seems clear that there is no great advantage regarding the chemistry of
the catalytic styrene oxidation, in the use of ILs in place of conventional
solvents, being equally necessary, in both cases, an excess of oxidant
after only 1 h at 60 °C (Table 2, compare entry 6 vs. entry 5). The slight
increased conversion observed in chloroform with respect to acetoni-
trile may be due, at least in part, to the donor ability of acetonitrile
which could hinder the coordination of the oxidant to the vanadium
center [59]. Differently, the experiments performed in ILs, namely
BMIM-PF
6 4
and BMIM-BF , in general showed less reactivity in com-
(
up to 4.0 equiv.) and prolonged reaction time (up to 6 h), to reach
parison to acetonitrile and chloroform, irrespective to the use of H O
2 2
satisfying conversion values.
or TBHP as oxidants, the conversions of s.m. ranging from 65 to 73 %,
after 6 h. The main difference may be found on the complete selectivity
toward epoxide F observed with TBHP (Table 2, entries 9–11), while
with H O the epoxide F was obtained with lower selectivity (64–72 %)
2 2
aside lower amounts of product G and traces of 1,2-diol H (Table 2,
entries 7–8). Newly, no main differences have been found in the cata-
The oxidation of CO, catalyzed by complex I and II, was then per-
formed with the aim of continuing the investigation of their catalytic
activity, under the same experimental conditions as above described for
styrene, but with a substrate characterized by the presence of a more
2 2
electron-rich double bond. Also in this case, H O showed to be less
efficient, in comparison to TBHP, either with conventional solvents or
ILs, irrespective to the type of catalyst I or II. Indeed, in acetonitrile,
after 6 h, only moderate conversions of s.m. have been observed (65–68
lytic activity between catalysts I and II, in BMIM-PF
trend showed in the oxidation of styrene (Table 2, entry 10 vs. 11). For
the two used ILs, the following order of activity BMIM-PF > BMIM-BF
was confirmed, even if without relevant differences [58].
6
, confirming the
6
4
%), with high selectivity for the corresponding CO epoxide F, being also
detectable traces of 1,2-diol H derived from the hydrolysis of the ep-
oxide, and cyclooctanone G, derived from the selective oxygen CH in-
sertion/isomerization on the s.m. (see Scheme 3 and Table 2, entries
3
.7.2. Selective oxidation of DBT and DMDBT
The growing interest towards the chemistry of oxovanadium(IV)
1
–2).
Chloroform confirmed to be a poor suitable solvent with hydrogen
complexes and their use in environmentally relevant catalytic industrial
processes, such as Oxidative Desulfurization (ODS) [60], for the
achievement of Ultra-Low-Sulphur-Diesels (ULSDs), has led us to test
peroxide (in spite of complete selectivity for the epoxide F), and no
significant improvement has been achieved by using MW assisted
heating (Table 2, entries 3–4). Newly, TBHP behaved as the most effi-
cient oxidant, either in chloroform or in acetonitrile, being the former
the most suitable solvent, affording quantitative epoxide formation
C6
C17
the catalytic activity of our catalysts VO(Q
for the oxidations of aromatic sulphides. Dibenzothiophene (DBT) and
,6-dimethyldibenzothiophene (DMDBT) substrates have been chosen
as models of aromatic sulphur compounds contained in crude oil.
)
2
(I) and VO(Q
2
) (II),
4
Table 2
Cis-cyclooctene oxidation catalysed by VO(QC6
)
2
(I) and VO(QC17
)
2
(II).
Time (hours)
Entrya
Catalyst
Oxidant
Solvent
Conv. (%)
TONb
Products (% Y) [% S]c (see Scheme 3)
1
2
3
VO(QC6
)
2
H
H
H
H
TBHP
TBHP
H
H
TBHP
TBHP
TBHP
2
2
2
2
O
O
O
O
2
2
2
2
MeCN
MeCN
6
6
6
68
65
3
68
65
3
F: (63), [93]; G: (3), [4]; H: (2), [3]
F: (59), [91]; G: (3), [5]; H: (3), [4]
F: (3), [100]
F: (19), [100]
F: (98), [100]
C17
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
)
2
C6
C6
C6
C6
C6
C6
C6
C6
)
)
)
)
)
)
)
)
2
2
2
2
2
2
2
2
CHCl
CHCl
3
d
< 1e
4
3
19
98
> 99
70
65
69
73
71
19
98
100
70
65
69
73
71
5
6
7
8
9
MeCN
CHCl
BMIM-PF
1.5
1
6
6
6
3
F: (> 99), [100]
2
O
O
2
6
F: (45), [64]; G: (20), [29]; H: (5) [7]
F: (47), [72]; G: (14), [22]; H: (4) [6]
F: (69), [100]
F: (73), [100]
F: (71), [100]
2
2
BMIM-BF
BMIM-BF
BMIM-PF
BMIM-PF
4
4
6
6
1
1
0
1
6
6
C17
)
2
a
Experimental conditions: cis-cyclooctene 0.2 mmol; catalyst =1.0 mol%; oxidant = 2.0 equiv.; temperature: 60 °C; solvent: 1.5 mL for conventional solvents and
1
.0 mL for ILs.
b
TON (turnover number): mmoles of converted substrate per mmole of catalyst.
Y = yield; S = selectivity; yields are calculated on converted substrate.
MW-promoted reaction: 2 cycles of 20 min, 70 °C, absorption level = normal, stirring rate =900 rpm.
Time of reaction =40 min.
c
d
e
7

P. Campitelli, et al.
Applied Catalysis A, General 599 (2020) 117622
Table 3
DBT and DMDBT oxidation catalysed by VO(QC6
)
and VO(QC17
)
2
complexes.
Oxidant
2
Entrya
Substrate
Catalyst
Solvent
MeCN
Conv. (%)
TONb
(Products)c (% Y), [% S]
C6
1
2
DBT
DBT
VO(Q
VO(Q
)
2
2
H
H
2
O
O
2
2
92
70
92
70
(SO
2
) (92), [100]
C6
)
2
2 2
CH Cl
(SO) (30), [43]
(
SO
2
2
2
2
2
2
2
2
) (40), [57]
C6
C6
C6
C6
C6
3
4
5
6
7
8
9
DBT
DBT
DBT
DBT
DBT
DBT
DBT
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
VO(Q
)
)
)
)
)
2
2
2
2
2
TBHP
MeCN
> 99
65
96
90
> 99
80
100
65
96
90
100
80
(SO
(SO
(SO
(SO
(SO
(SO
(SO
) (> 99), [100]
) (65), [100]
) (96), [100]
) (90), [100]
) (> 99), [100]
) (80), [100]
) (96), [100]
H
2
O
2
BMIM-OTfd
d
TBHP
TBHP
TBHP
TBHP
TBHP
BMIM-OTf
BMIM-BF
BMIM-PF
BMIM-BF
4
BMIM-PF
MeCN
d
4
d
6
C17
C17
d
)
)
2
d
2
6
96
80
96
80
1
0
1
DMDBT
VO(QC6)
2
H
2
O
2
(SO) (30), [37]
SO ) (50), [63]
(SO) (55), [73]
(
2
1
DMDBT
VO(QC6
)
2
H
2
O
2
CH
2
Cl
2
75
75
(
SO
2
) (20), [27]
C6
1
1
2
3
DMDBT
DMDBT
VO(Q
VO(Q
)
)
2
2
TBHP
TBHP
MeCN
BMIM-OTfd
95
94
95
94
(SO
2
) (95), [100]
C6
(SO) (8), [9]
SO ) (86), [91]
(SO) (12), [14]
SO ) (71), [86]
(SO) (10), [10]
SO ) (89), [90]
(SO) (18), [19]
SO ) (78), [81]
(
2
1
1
1
4
5
6
DMDBT
DMDBT
DMDBT
VO(QC6
VO(QC6
VO(QC6
)
)
)
2
2
2
TBHP
TBHP
TBHP
BMIM-BF d
BMIM-PF d
4
83
83
(
2
6
> 99
96
100
96
(
2
BMIM-PF
6
(
2
a
Experimental conditions: substrate 0.1 mmol; catalyst = 1.0 mol%; oxidant = 4.0 equiv.; temperature: 60 °C; reaction time: 4 h.
TON (turnover number): mmoles of converted substrate per mmole of catalyst.
b
c
SO = sulfoxide; SO
2
= sulfone. Y = yield; S = selectivity; yields are calculated on converted substrate.
d
1.2 mL of n-octane/IL = 5:1 mixture.
During combustion they produce SO
x
derivatives that are able to affect
olefins oxidation, thus probably confirming the role exerted by the
coordinating ability (toward the metal centre) of the different ionic li-
quid anions in the modulation of the catalytic activity of the core va-
nadyl cation (VO)2+, within the complexes I and II, thus influencing its
ability over the coordination with the main oxidants (H O or TBHP) in
2 2
the beginning of the activation step (see onward the next theoretical
calculations paragraph).
On the whole, in the case of 4,6-dimethyl-dibenzothiophene
(DMDBT), the presence of alkyl substituents on the 4- and 6-positions of
the aromatic ring could have a double antagonistic effect, thus influ-
encing its reactivity in consequence of either, their inductive releasing
effects or their steric concerns. Indeed, with the increase of electron
density on sulphur atom, an increase of the reactivity toward sulphur
oxidation is generally observed [65]. On the other hand, the sub-
stituents steric effects could greatly reduce the tendency of sulphur to
be oxidized, especially with systems containing sterically demanding
groups. These issues are of particular interest, since the previously re-
ported catalytic systems for ODS process, containing vanadium or
tungsten metals, showed to be sensible to steric effects [66,67]. In effect
also in our study, working with the selected complex I, the oxidation of
DMDBT showed to be, in general, less selective in comparison to DBT,
that is a mixture of sulfoxide and sulfone derivatives has been always
detected, with the only exception of the TBHP/MeCN system, where the
complete selectivity toward the corresponding sulfone, was observed
the atmospheric ozone layer, increasing the formation of acid rains, and
reducing the life of the engine due to corrosion. Thus, their complete
removal due to their complete oxidation to sulfones and successive
extraction with a selected solvent (extractive catalytic oxidative de-
sulfurization, ECODS), represents one of the most demanding chal-
lenges from an environmental point of view [61,62]. Among others, one
of the most studied class of solvents, within the ECODS technique, is
represented by ILs [58,63,64].
In order to mimic the hydrophobicity of a real crude oil, the cata-
lytic activity of complexes I and II, were studied in a 5:1 mixture of n-
octane/IL, instead of pure IL phase. Moreover, for sake of comparison
with the above reported study on olefin oxidation, we tested the model
ODS reactions also in acetonitrile and dichloromethane. In the Table 3
are reported the main results we have obtained in this study. Acetoni-
trile gave the best results, affording quantitative conversion of di-
benzothiophene (DBT) and complete selectivity toward the corre-
sponding sulfone, with catalyst I (Table 3, entry 3). With the same
catalyst I, hydrogen peroxide showed a slight reduction of activity, in
acetonitrile (vs. TBHP), while in dichloromethane this oxidant gave
both lower activity and selectivity still being present, after 4 h, about 30
%
of unreacted s.m. and 30 % of sulfoxide, in the crude (Table 3, entries
1
–2). Anyway, also with a hydrophilic IL like BMIM-OTf, H con-
2 2
O
firmed to be less efficient in comparison to TBHP, only affording a
moderate 65 % of s.m. conversion, but with complete selectivity for the
sulfone (Table 3, entry 4). A detailed investigation of the activity of
complex I, in the selective DBT oxidation with TBHP by using all the
selected ILs, namely BMIM-BF , BMIM-OTf and BMIM-PF , confirmed
4 6
almost quantitative conversion of s.m. and most notably, complete se-
lectivity for the sulfone formation (Table 3, entries 5–7).
2 2
(Table 3, entry 12). With H O , the DMDBT oxidation in acetonitrile or
dichloromethane, respectively, showed comparable conversions of s.m.
(from 75 % to 80 %), being the selectivity for the sulfone higher in the
former solvent (Table 3, compare entry 10 vs. entry 11). Newly, when
TBHP and an imidazolium ILs medium are present (as n-octane/
IL = 5:1 mixture) the same order of activity, i.e. BMIM-PF
6
> BMIM-
The complex II confirmed a high catalytic activity for the oxidation
of DBT, even if with a slight reduction in the efficiency in comparison
with the complex I (Table 3, compare entries 8–9 vs. entries 6–7). The
OTf > BMIM-BF , previously observed for the DBT oxidation, was
confirmed also with DMDBT, with selectivity values for the sulfone
ranging from 86 % to 91 % (Table 3, entries 13–15). Finally, a catalytic
4
following order of activity, for ILs, BMIM-PF
6
> BMIM-OTf > BMIM-
test performed by using the most active IL, that is BMIM-PF
6
as the
4
BF , observed in DBT oxidation, is the same as that observed during
unique solvent phase, to evaluate the role of n-octane, if any, on either
8

P. Campitelli, et al.
Applied Catalysis A, General 599 (2020) 117622
3
Fig. 3. Reaction free energies at 60 °C in CH CN solution.
the activity or selectivity of the DMDBT oxidation, furnished very si-
milar results with other cases (Table 3, compare entry 16 vs. entries
derivatives, thus allowing its stabilization and therefore its recyclability
[68]. As further confirmation of this, all the collected n-octane phases,
recovered after each recycling step, were evaporated and analysed by
ICP-AES, in order to evaluate the residual amount of vanadium. The
detected amount of metal confirmed that, after five running steps, only
a slight leaching (8.5 %) of vanadium occurred, in comparison to the
starting value.
13–15). The last result confirmed that the positive role of ILs as an
extracting medium of the more polar sulfoxide and sulfones, at com-
parable conditions of catalytic activity, is well operative also in the n-
octane/IL = 5:1 mixture we used, thus taking an advantage in eco-
nomic terms for the lower consumption of IL, and being the system
more representative of the real medium encountered with an authentic
oil sample in the ODS process.
3.8. DFT computational studies
3
.7.3. Recyclability experiments
The recyclability of the catalytic system was studied for the DBT
All the results reported in this section refer to the catalyst VO(QMe
in which the alkyl chains C H13 and C17H35 (see I and II, Scheme 1)
6
were substituted with the methyl group in order to reduce the already
high complexity of the problem. The Cartesian coordinates of all the
optimized species are reported in the Supplementary Material section
)
2
6
oxidation with the catalyst I in a n-octane/BMIM-PF (5:1) mixture,
promoted by TBHP (see above the 2.8 section). It should be underlined
that the vanadium complex retained its activity thus allowing its re-
cyclability. After the first cycle, the organic phase was recovered and
analysed by GC-FID to evaluate its residual sulphur content. Then, new
fresh portions of reagent were added to the recycled IL phase containing
the entrapped catalyst, and a new cycle started. Since the first cycle,
trace of a white precipitate was detected in the mixture, which in-
creased in the course of the next recycling steps. It was removed and
analysed, by FTIR, after the fourth recycling cycle, showing the char-
acteristically absorption bands of sulfone group [65]. Interestingly, the
catalytic system retained the original activity, without a detectable
catalyst leaching, at least for the first five cycles, showing only a very
slight decreasing of activity after the third and fourth recycle (less than
(
Table S-1).
3.8.1. Catalyst characterization
In order to assess the most stable form of the catalyst present at
60 °C in acetonitrile, at least according to our model, we have pre-
liminarily evaluated the possibility of finding a solvent molecule in the
sixth coordination position of vanadium. At this purpose we have cal-
culated the free energy associated to the reaction:
VO(QMe
CN(l) → [VO(QMe
2 3
) (CH CN)](sol)
)
2(sol) + CH
3
1
0 % overall). This result may be rationalized, at least in part, as a
Which turned out to be equal to 13 kJ/mol.
For this calculation we have utilized the density of CH
17.9 mol/liter at 60 °C.
consequence of the immobilizing aptitude of the ionic liquid medium
toward either the catalytic species or the corresponding active
3
CN of
9

P. Campitelli, et al.
Applied Catalysis A, General 599 (2020) 117622
Fig. 4. Free energy diagram (in kJoule/mole) at 60 °C in CH
state. In the inset (c) we also report the free energy profile for the reformation of 2. Note that the channel in the inset (c) should be read from right to left.
3
CN solution of the reactions of 32 with olefin along the singlet (inset a) and the triplet (inset b) magnetic
1
This value allowed us to approximately consider, at least within our
mechanism, with a slightly lower activation free-energy for 1TS-2−4
species, is also found for the reaction of 2 with DBT, as reported in
Fig. 5, inset (a), (b) and (c) and Fig. 6, Scheme Ib.
Me
model, VO(Q
)2(sol) as the species mainly present in solution and
hence it has been used as starting point in all the subsequent calcula-
tions.
In this case the only difference emerged, with respect to the pre-
vious case, is the absence of MECP-2 hence preventing the route – in
principle possible but practically not very efficient also in the reaction
with the olefin – directly forming the oxidation products starting from
3
.8.2. Reaction with H
2 2
O
3
First of all we have addressed the question concerning the species
2 and the olefin.
Me
actually present in CH
3
CN solution at the equilibrium when VO(Q
)
2
It is very important to remark that all the above processes involve
the dioxocomplex 6 as a key-intermediate for the conclusion of the
catalytic cycle. However this latter step requires a relatively high bar-
rier of 160 kJoule/mole. For this reason the kinetic-mechanistic role of
the processes initiated by 12 and/or 32 according to our model, might
be not very efficient.
is mixed with H at the temperature of the experiment.
2 2
O
According to the schemes reported in Fig. 3, the only species ex-
pected to be formed with a relevant yield, and containing the metal in
Me
2
the highest oxidation state (+5), are the peroxo VO(Q )(O ) (2),
3
mainly but not uniquely in the triplet state, and the hydroperoxo VO
Me
3
(
Q
)
2
(OOH) ( 4). This latter species also reveals thermodynamically
Consequently, although aware of the previously remarked intrinsic
%
%
unstable toward the formation of the HO
mechanistic hypothesis, in the case of H
2
radical (see below). Our
thermodynamic instability to form HO
2
radical, we have searched for
3
2
O
2
, is then based on these two
an alternative route initiated by the species 4.
species and the results (i.e. the free energy diagrams) are described in
Fig. 4 with the olefin (modelled using cis-2-butene), and in Fig. 5 with
DBT, as substrates.
Also for this species we localized a Sharpless-like mechanism taking
place only along the triplet magnetic state, as depicted in Fig. 7, both
using olefin (inset a) and DBT (inset c) as the substrate. Moreover, if we
compare the free energy barriers reported in Figures 4, 5 and 7, we can
For the reaction of 2 with the olefin, we localized a Sharpless-like
mechanism both involving the triplet and the singlet magnetic state
conclude that the species 34, although kinetically unstable toward the
3
Me
%
2
(
Fig. 4, inset (a) and (b), and Fig. 6). 2 can form the dioxo VO(Q )(O)
unimolecular formation of HO
radical, appears as much more efficient
species (6) and the epoxide either in the triplet state, through an
than 2.
3
adiabatic route involving TS-2, or in the singlet state through a non-
Me
adiabatic route involving MECP-2. The species 6 VO(Q )(O), further
reacting with H
2
O
2
, can finally restore 2 in the singlet magnetic state
3.8.3. Reaction with TBHP
(
Fig. 4, inset (c) and Fig. 6, Scheme Ia) which eventually either con-
2 2
When H O is substituted by TBHP, the whole reaction picture
tinues the reaction more efficiently (Fig. 4, inset (a), Fig. 6, Scheme Ia)
or non-adiabatically regenerate 2 through MECP-1. A very similar
undergoes a significant change. As a matter of the fact, as reported in
Fig. 8, we observe that the only species present in acetonitrile solution,
3
10

P. Campitelli, et al.
Applied Catalysis A, General 599 (2020) 117622
Fig. 5. Free energy diagram (in kJoule/mole) at 60 °C in CH
magnetic state. In the inset (c) we again report the free energy profile for the reformation of 2. Note that the channel in the inset (c) should be read from right to left.
3
CN solution of the reactions of 32 and 12 with DBT along the singlet (inset a) and the triplet (inset b)
1
at the equilibrium, before the addition of the substrate, is the adduct
methyl esters promoted by TBHP, has been previously observed by ESI-
MS study [39].
3
4
TBHP
Differently from the analogue species formed when H
2
O
2
is used
Both in the case of the olefin (Fig. 9, inset a) and DBT (Fig. 9, inset
b) the reaction turns out to take place onto the triplet magnetic surface
through a rather Sharpless-like epoxidation mechanism followed by the
3
(
i.e. the species 4) the species
3
4
TBHP doesn’t show any propensity to dissociate hence resulting
rather stable both thermodynamically and kinetically. Moreover, this
species represent the only plausible intermediate able to react as re-
ported in Fig. 9. Interestingly, the same tert-butylperoxovanadium(V)
complex 4TBHP active in the catalytic oxidation of unsaturated fatty acid
regeneration of 3 TBHP through the reoxidation of 3
4 3TBHP. Comparing
the related free-energy barriers with the ones observed in the case of
the homologous reaction of Fig. 7, we interestingly infer that the re-
action with TBHP is expected to be more efficient than that with H O
2 2
Fig. 6. Schematic and concise representation of the catalytic cycles as obtained from our calculations (see Figs. 3 and 4, respectively). The dotted arrows refer to non-
adiabatic channels.
11

P. Campitelli, et al.
Applied Catalysis A, General 599 (2020) 117622
Fig. 7. Free energy diagram (in kJoule/mole) at 60 °C in CH
3
CN solution of the reactions of 34 with the olefin (inset a) and DBT (inset c). In the inset (b) we also
report the free energy profile for the 34 dissociation reaction. Note that the free energy diagram depicted with the dotted line refers to the reaction between 3 and
3
2 2
H O .
3
Fig. 8. Reaction free energies at 60 °C in CH CN solution.
12

P. Campitelli, et al.
Applied Catalysis A, General 599 (2020) 117622
Fig. 9. Free energy diagram (in kJoule/mole) at 60 °C in CH
energy diagram depicted with the dotted line refers to the reaction between
3
CN solution of the reactions of 34TBHP with the olefin (inset a) and DBT (inset b). Note that the free
3
3TBHP and t-BuOOH.
because of the much higher kinetic stability of 3
4
TBHP which, hence,
represents the key intermediate of the whole process.
The same calculations were also carried out using CH
2 2
Cl as solvent.
The results, not reported for the sake of brevity, did not show any ap-
preciable difference with acetonitrile. As a matter of the fact, the barrier
heights as well as the relative thermodynamic stability of the involved
species only undergo variations within a few kJoule/mole.
As a consequence, we used a different viewpoint to interpret the
reduced efficiency of H
purpose we calculated the partition coefficient, log(PCH2Cl2/water) and
log(PCH3CN/water) of H and TBHP at 60 °C using the excess free en-
ergies already adopted for the calculations of the reaction energies
above shown. The obtained values of log(PCH2Cl2/water) for H and
TBHP equal to -0.64 and -0.45, respectively, clearly indicate that the
concentration of H is approximately half of the concentration of
TBHP in this solvent. On the other hand, in the case of log(PCH3CN/water
2 2 2 2
O in CH Cl experimentally observed. At this
2 2
O
2 2
O
2 2
O
4 4
Fig. 10. Schematic picture of the [VO(BF )
]2 species.
−
)
we obtained almost identical values of − 0.060, for both the oxidants.
In the light of these data, we can tentatively state that the reduced
−
BF
4
anion used in some of the experimental data reported in this
study. In Fig. 10 we report the picture of optimized structure of [VO
2 2 2 2
catalytic activity of H O observed in CH Cl might be predominantly
2−
(
BF
4
)
4
]
which revealed rather stable (in gas-phase) showing an
ascribed both to its reduced concentration and to its intrinsic reactivity.
average VeF bond energy of 580 kJoule/mole lower than the VeO
bond energy of 650 kJoule/mole.
However, such a difference in the binding energy might be not
sufficient to resist the mass effect, due to the very high concentration of
3.8.4. Preliminary investigations concerning ionic liquids
As already remarked in the section dedicated to the computational
details, the use of a mean field approximation in ionic liquids is much
more problematic than conventional liquids. In fact in this latter en-
vironment, the neglect of the particle (explicit) description of the clo-
sest solvation shells, although physically questionable, in some cases
can provide good results. This is basically impossible for ionic liquids in
which the cybotactic effects can be more important. For this reason we
limited our attention on the possible formation of species, emerging
−
4
the BF
species in the ionic liquid, leading to a partial (or even
quantitative) breakdown of the catalyst hence explaining the experi-
mentally observed moderate activity. This part of our study certainly
deserves further investigation.
2
+
from a coordination bond between vanadyl [(VO) ] cation and the
13

P. Campitelli, et al.
Applied Catalysis A, General 599 (2020) 117622
4. Conclusions
Appendix A. Supplementary data
Two new oxovanadium(IV) complexes with 4-acyl-5-pyrazolonate-
Supplementary data associated with this article can be found, in the
online version, at https://doi.org/10.1016/j.apcata.2020.117622.
2
κ -O,O' bidentate ligands bearing aliphatic substituents have been pre-
pared and fully characterized both in the solid and solution state. Their
catalytic potentials have been thoroughly investigated aiming to eval-
uate the role, if any, of the reaction medium (imidazolium based ILs vs
acetonitrile or chlorinated solvents) on steering the activity and se-
lectivity of the oxidation reactions of olefins like styrene and cis-cy-
clooctene, or organosulphur substrates as DBT and DMDBT, promoted
References
[1] G. Strukul, Catalytic Oxidations with Hydrogen Peroxide As Oxidant, Kluwer
Academic Publishers, Dordrecht, 1992.
[
2] W. Adam, Peroxide chemistry. Mechanistic and Preparative Aspects of Oxygen
Transfer, Wiley-VCH, Weinheim, 2000.
in the presence of H
worked better than H
2
O
2
or TBHP as main oxidants. As a whole, TBHP
in the activation of catalysts I and II, affording
[3] V. Conte, O. Bortolini, Transition Metal Peroxides, Synthesis and Role in Oxidation
Reactions, Wiley Online Library, Weinheim, 2009.
2
O
2
[
4] W. Adam, W. Malisch, K.J. Roschmann, C.R. Saha-Möller, W.A. Schenk, J.
almost in all cases quantitative conversions, regardless of the type of
substrates (olefins or organosulphur derivatives) or reaction medium,
acetonitrile being the best one. No significative differences in terms of
catalytic activity or selectivity have been detected between the two
oxovanadium complexes I and II, neither in conventional solvents nor
in ILs, the former being only slightly more active. This may be ascribed,
at least partially, to the fact that the diverse lengths of the alkyl side-
chain of the ligands, being too far from the catalytically active metal
centre, are unable to someway diversify the outcome of catalytic runs.
For the reactions performed in ILs, the following order of activity,
Organomet. Chem. 661 (2002) 3–16.
[5] D.C. Crans, J.J. Smee, E. Gaidamauskas, L. Yang, Chem. Rev. 104 (2004) 849–902.
6] D. Rehder, Metallomics. 7 (2015) 730–742.
7] J.C. Pessoa, S. Etcheverry, D. Gambino, Coord. Chem. Rev. 301 (2015) 24–48.
8] A. Butler, M.J. Clague, G.E. Meister, Chem. Rev. 94 (1994) 625–638.
[9] K.A. Jørgensen, Chem. Rev. 89 (1989) 431–458.
[10] F. Farzaneh, Z. Asgharpour, Appl. Organomet. Chem. 33 (2019) e4896.
11] R. Tan, C. Li, Z. Peng, D. Yin, D. Yin, Catal. Commun. 12 (2011) 1488–1491.
12] M.R. Maurya, M. Kumar, S.J.J. Titinchi, H.S. Abbo, S. Chand, Catal. Letters 86
[
[
[
[
[
(2003) 97–105.
[13] S. Rezaei, A. Ghorbani‐Choghamarani, R. Badri, A. Nikseresht, Appl. Organomet.
Chem. 32 (2018) e3948.
[
14] H.P. Mungse, S. Verma, N. Kumar, B. Sain, O.P. Khatri, J. Mater. Chem. 22 (2012)
427–5433.
[15] M.R. Maurya, U. Kumar, I. Correia, P. Adão, J. Costa Pessoa, Eur. J. Inorg. Chem.
008 (2008) 577–587.
6 4
BMIM-PF > BMIM-OTf > BMIM-BF , observed in the oxidation of
DBT and DMDBT and also detected during olefins oxidation, is in ac-
cordance with previous results from literature. Very interestingly, the
6
system formed by complex I and BMIM-PF used for the study of DBT
5
2
[
[
16] J. Hartung, Pure Appl. Chem. 77 (2005) 1559–1574.
17] I. Lippold, J. Becher, D. Klemm, W. Plass, J. Mol. Catal. A Chem. 299 (2009) 12–17.
oxidation with TBHP showed to be stable, retaining the activity of the
catalyst caught inside the IL phase and so allowing its recyclability,
without a relevant catalyst leaching, at least for five running cycles.
Quantum-chemical calculations have shown that, in conventional sol-
vents, the species, and hence the whole mechanism, might be different
according to the diverse main oxidants used. The lower efficiency ob-
[18] S. Parihar, R.N. Jadeja, V.K. Gupta, RSC Adv. 4 (2014) 10295–10302.
[
19] F. Marchetti, C. Pettinari, C. Di Nicola, A. Tombesi, R. Pettinari, Coord. Chem. Rev.
01 (2019) 213069.
20] F. Marchetti, R. Pettinari, C. Pettinari, Coord. Chem. Rev. 303 (2015) 1–31.
4
[
[21] F. Marchetti, C. Pettinari, R. Pettinari, Coord. Chem. Rev. 249 (2005) 2909–2945.
[22] V. Conte, B. Floris, Inorganica Chim. Acta 363 (2010) 1935–1946.
[
23] G. Silva, S. Voth, P. Szymanski, E.M. Prokopchuk, Fuel Process. Technol. 92 (2011)
656–1661.
1
served with H
2
O
2
in chlorinated solvents, has been mainly ascribed to
[24] M. Sedighipoor, A.H. Kianfar, W.A.K. Mahmood, M.H. Azarian, Inorganica Chim.
the water/solvent partition coefficient particularly low in the case of
H O .
2 2
Acta 457 (2017) 116–121.
[
[
[
[
25] G. Grivani, G. Bruno, H.A. Rudbari, A.D. Khalaji, P. Pourteimouri, Inorg. Chem.
Commun. 18 (2012) 15–20.
26] A.C.V. Conte, P. Galloni, M. Centazzo, Process for Desulphurisation of Fuels, WO
2015/162162 AI, 2015.
27] A.S. Ogunlaja, W. Chidawanyika, E. Antunes, M.A. Fernandes, T. Nyokong,
N. Torto, Z.R. Tshentu, Dalton Trans. 41 (2012) 13908–13918.
28] M.R. Maurya, A. Arya, A. Kumar, M.L. Kuznetsov, F. Avecilla, J. Costa Pessoa, Inorg.
Chem. 49 (2010) 6586–6600.
Finally, preliminary calculations carried out for roughly modelling
Me
what might happen when IL are used, suggest that the VO(Q
2
) initial
adduct might undergo a partial or complete disruption essentially due
to the vanadyl-IL anion interaction. This observation might qualita-
tively explain the substantially not so relevant efficiency observed in
ILs, in comparison to conventional solvents, but it requires further in-
vestigations, at the moment rather difficult for the high complexity for
the computational IL modelling.
[29] H. Olivier-Bourbigou, L. Magna, D. Morvan, Appl. Catal. A Gen. 373 (2010) 1–56.
[30] A. Mota, N. Butenko, J.P. Hallett, I. Correia, Catal. Today 196 (2012) 119–125.
[31] S. Zhang, J. Sun, X. Zhang, J. Xin, Q. Miao, J. Wang, Chem. Soc. Rev. 43 (2014)
7838–7869.
[
[
32] M. Lombardo, C. Trombini, ChemCatChem. 2 (2010) 135–145.
33] F. Marchetti, C. Di Nicola, R. Pettinari, C. Pettinari, I. Aiello, M. La Deda,
A. Candreva, S. Morelli, L. De Bartolo, A. Crispini, Eur. J. Inorg. Chem. (2019).
34] J. Palmucci, F. Marchetti, R. Pettinari, C. Pettinari, R. Scopelliti, T. Riedel,
B. Therrien, A. Galindo, P.J. Dyson, Inorg. Chem. 55 (2016) 11770–11781.
CRediT authorship contribution statement
[
Patrizio Campitelli: Investigation, Formal analysis. Massimiliano
Aschi: Investigation. Corrado Di Nicola: Formal analysis. Fabio
Marchetti: Conceptualization, Writing - review & editing. Riccardo
Pettinari: Validation. Marcello Crucianelli: Conceptualization,
Writing - review & editing.
[35] C. Di Nicola, F. Marchetti, R. Pettinari, A. Tombesi, C. Pettinari, I. Grappasonni,
P.J. Dyson, S. Scuri, Materials (Basel). 13 (2020) 526.
[
36] C. Pettinari, F. Marchetti, R. Pettinari, Y.A. Belousov, I.V. Taydakov,
V.D. Krasnobrov, D.I. Petukhov, A.A. Drozdov, Dalton Trans. 44 (2015)
14887–14895, https://doi.org/10.1039/c5dt01964h.
[
[
[
[
37] F. Marchetti, C. Pettinari, C. Di Nicola, R. Pettinari, A. Crispini, M. Crucianelli, A. Di
Giuseppe, Appl. Catal. A Gen. 378 (2010) 211–220.
38] A. Di Giuseppe, C. Di Nicola, R. Pettinari, I. Ferino, D. Meloni, M. Passacantando,
M. Crucianelli, Catal. Sci. Technol. 3 (2013) 1972–1984.
39] M.M. Cecchini, F. De Angelis, C. Iacobucci, S. Reale, M. Crucianelli, Appl. Catal. A
Gen. 517 (2016) 120–128.
40] R.R. Langeslay, D.M. Kaphan, C.L. Marshall, P.C. Stair, A.P. Sattelberger,
M. Delferro, Chem. Rev. 119 (2018) 2128–2191.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
[41] J. Harvey, M. Aschi, H. Schwarz, W. Koch, Theor. Chem. Acc. 99 (1998) 95–99,
https://doi.org/10.1007/s002140050309.
[
[
[
[
42] S. Miertuš, E. Scrocco, J. Tomasi, Chem. Phys. 55 (1981) 117–129.
43] M. Cossi, N. Rega, G. Scalmani, V. Barone, J. Comput. Chem. 24 (2003) 669–681.
44] T. Yanai, D.P. Tew, N.C. Handy, Chem. Phys. Lett. 393 (2004) 51–57.
45] P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 299–310, https://doi.org/10.1063/
Acknowledgment
1
.448975.
[
[
[
46] C. Chiappe, C.S. Pomelli, Phys. Chem. Chem. Phys. 15 (2013) 412–423.
47] M.R. Maurya, Coord. Chem. Rev. 237 (2003) 163–181.
48] J. Selbin, L.H. Holmes Jr., S.P. McGlynn, J. Inorg. Nucl. Chem. 25 (1963)
1359–1369.
The Universities of L’Aquila and Camerino are gratefully acknowl-
edged for financial support. Authors thank Mr. Renato Di Bartolomeo
(
University of L’Aquila) for his help in the preparation of the graphics of
the manuscript.
[49] C. Pettinari, F. Marchetti, A. Drozdov, Compr. Coord. Chem. II. (2003) 97–115.
14

P. Campitelli, et al.
Applied Catalysis A, General 599 (2020) 117622
[61] R. Saladino, G. Botta, M. Crucianelli, Advances in Nanotechnology Transition Metal
Catalysts in Oxidative Desulfurization (ODS) Processes: Nanotechnology Applied to
ODS Processing, Appl. Nanotechnol. to Desulfurization Process Pet. Eng., IGI Global
(2016) 180–215.
[62] M. Crucianelli, B.M. Bizzarri, R. Saladino, Catalysts 9 (2019) 984.
[63] B. Rodríguez-Cabo, H. Rodríguez, E. Rodil, A. Arce, A. Soto, Fuel 117 (2014)
882–889.
[50] J. Selbin, Chem. Rev. 65 (1965) 153–175.
[51] S. Parihar, S. Pathan, R.N. Jadeja, A. Patel, V.K. Gupta, Inorg. Chem. 51 (2012)
1
152–1161.
[
[
[
52] K. Kustin, D.L. Toppen, Inorg. Chem. 12 (1973) 1404–1407.
53] Y. Xiao, H. Huang, D. Yin, D. Guo, L. Mao, Z. Fu, Catal. Commun. 10 (2008) 29–32.
54] S. Vezzosi, A.G. Ferré, M. Crucianelli, C. Crestini, R. Saladino, J. Catal. 257 (2008)
2
62–269.
[
[
[
[
55] P. Adão, J. Costa Pessoa, R.T. Henriques, M.L. Kuznetsov, F. Avecilla, M.R. Maurya,
U. Kumar, I. Correia, Inorg. Chem. 48 (2009) 3542–3561.
56] M. Aschi, M. Crucianelli, A. Di Giuseppe, C. Di Nicola, F. Marchetti, Catal. Today
[64] C. Wang, Z. Chen, X. Yao, W. Jiang, M. Zhang, H. Li, H. Liu, W. Zhu, H. Li, RSC Adv.
7 (2017) 39383–39390.
[65] S. Otsuki, T. Nonaka, N. Takashima, W. Qian, A. Ishihara, T. Imai, T. Kabe, Energy
Fuels 14 (2000) 1232–1239.
[66] J.M. Campos-Martin, M.C. Capel-Sanchez, J.L.G. Fierro, Green Chem. 6 (2004)
557–562.
[67] L. Cedeño-Caero, H. Gomez-Bernal, A. Fraustro-Cuevas, H.D. Guerra-Gomez,
1
92 (2012) 56–62.
57] R. Bikas, M. Ghorbanloo, S. Jafari, V. Eigner, M. Dusek, Inorganica Chim. Acta 453
2016) 78–85.
58] A. Mota, J.P. Hallett, M.L. Kuznetsov, I. Correia, Phys. Chem. Chem. Phys. 13
2011) 15094–15102.
(
(
R. Cuevas-Garcia, Catal. Today 133 (2008) 244–254.
[
[
59] S. Rayati, A. Ghaemi, N. Sadeghzadeh, Catal. Commun. 11 (2010) 792–796.
60] A.S. Ogunlaja, R.S. Walmsley, C. du Sautoy, N. Torto, Z.R. Tshentu, Energy Fuels 27
[68] J.W. Lee, J.Y. Shin, Y.S. Chun, H. Bin Jang, C.E. Song, S. Lee, Acc. Chem. Res. 43
(2010) 985–994.
(2013) 7714–7723.
15