Peroxo-dimolybdate catalyst for the oxygenation of organic sulfides by hydrogen peroxide

Article abstract of DOI:10.1016/j.ica.2015.08.016

The reaction between Na₂MoO₄ and hydrogen peroxide at pH around 6.5 resulted in a novel peroxo-dimolybdate anion [{MoO(O₂)₂}₂(μ-O)]^2-, which was isolated as a tetrabutylammonium salt (1). X-ray single crystal diffraction study of 1 revealed that there are two η²-peroxo ligands for each Mo center and a single μ-oxo bridge linking two Mo(VI) centers. Compound 1 can function as a stoichiometric oxidant for the oxygenation of organic sulfides and donate up two active oxygen atoms. Furthermore, 1 catalyzes organic sulfide oxygenation by hydrogen peroxide with 100% utility. The catalytic proficiency of 1 was examined with the oxygenation reactions of substrates including benzyl phenyl sulfide, methyl phenyl sulfide and 4-bromo thioanisole with TOF of 240, 550 and 740 h^-1, respectively. Based on the initial rate study of several para-substituted thioanisoles, a reactivity constant (ρ) of -1.06 was obtained, which implies an electron deficient transition state. The initial rate study using thioanisole as the substrate also revealed that the reaction is first order in catalyst but zero order in hydrogen peroxide. Based on these results, a mechanism for the activation of hydrogen peroxide is proposed. The DFT calculations of model catalyst and substrate resulted in an activation energy of 75 kJ mol^-1 for the transition state, and a reaction free energy of -69 kJ mol^-1.

Full text of DOI:10.1016/j.ica.2015.08.016

Inorganica Chimica Acta 437 (2015) 103–109
Contents lists available at ScienceDirect
Inorganica Chimica Acta
journal homepage: www.elsevier.com/locate/ica
Peroxo-dimolybdate catalyst for the oxygenation of organic sulfides by
hydrogen peroxide
⇑
Dylan J. Thompson, Zhi Cao, Eileen C. Judkins, Phillip E. Fanwick, Tong Ren
Department of Chemistry, Purdue University, West Lafayette, IN 47907, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 26 June 2015
Received in revised form 20 August 2015
Accepted 25 August 2015
Available online 1 September 2015
The reaction between Na2MoO4 and hydrogen peroxide at pH around 6.5 resulted in
a novel
peroxo-dimolybdate anion [{MoO(O2)2}2(
(1). X-ray single crystal diffraction study of 1 revealed that there are two
Mo center and a single -oxo bridge linking two Mo(VI) centers. Compound 1 can function as a
l
-O)]2ꢀ, which was isolated as a tetrabutylammonium salt
g
2-peroxo ligands for each
l
stoichiometric oxidant for the oxygenation of organic sulfides and donate up two active oxygen atoms.
Furthermore, 1 catalyzes organic sulfide oxygenation by hydrogen peroxide with 100% utility. The
catalytic proficiency of 1 was examined with the oxygenation reactions of substrates including benzyl
Keywords:
Peroxo-molybdate
Hydrogen peroxide
Sulfide oxygenation
DFT calculations
phenyl sulfide, methyl phenyl sulfide and 4-bromo thioanisole with TOF of 240, 550 and 740 hꢀ1
,
respectively. Based on the initial rate study of several para-substituted thioanisoles, a reactivity constant
(
q
) of ꢀ1.06 was obtained, which implies an electron deficient transition state. The initial rate study using
thioanisole as the substrate also revealed that the reaction is first order in catalyst but zero order in
hydrogen peroxide. Based on these results, a mechanism for the activation of hydrogen peroxide is
proposed. The DFT calculations of model catalyst and substrate resulted in an activation energy of
75 kJ molꢀ1 for the transition state, and a reaction free energy of ꢀ69 kJ molꢀ1
.
Ó 2015 Published by Elsevier B.V.
1. Introduction
to both the deep desulfurization of fossil fuels [21] and degradation
of nerve agents [22–25]. Catalysts identified this far include Mn
(III)-Me3TACN [26,27], diruthenium tetracarboxylate (or tetramid-
Polyoxometallates (POM) have been widely used as catalysts for
various oxygenation reactions, especially olefin epoxidation and
sulfide oxygenation reactions with environmentally friendly oxi-
dants such as H2O2 and O2 [1,2]. One of the most frequently used
inate) [28–32] and [c
-SiW10O34(H2O)2]4ꢀ [32,33]. Because of our
work on the latter, we have become interested in the aforemen-
tioned dinuclear W/Mo peroxo compounds. While most of the
efforts on dinuclear W/Mo peroxo species focus on olefin oxygena-
tion reactions, a selenate bridged ditungsten peroxo compound,
[SeO4{WO(O2)2}2]2ꢀ, facilitates efficient H2O2 oxygenation of vari-
ous organic sulfides, including the quantitative conversion of
dibenzothiophene to the corresponding sulfone [34]. Inspired by
the high activity of [SeO4{WO(O2)2}2]2ꢀ, we set out to prepare its
Mo analog, [SeO4{MoO(O2)2}2]2ꢀ. Surprisingly, the dinuclear spe-
cies isolated under conditions similar to those used for [SeO4{WO
POM catalysts is the Keggin ion, [XM12O40]
Zꢀ, with M as Mo or
W, and heteroatom X as P or Si. It has been established by the lab-
oratories of Brégeault and Hill that the real active species in the
‘‘[PW12O40] l-O2)
3ꢀ/H2O2” combination is most likely HPO4[WO(
(O2)]2 and its dimerized form {PO4[WO(O2)2]4}3ꢀ [3,4]. Since the
reports of Brégeault and Hill, a number of related dinuclear
W/Mo peroxo compounds have been structurally identified with
or without a hetero-atom [5–13], many of which are also active
catalysts for H2O2 oxygenation reactions. It is worth mentioning
that mono-nuclear peroxo/oxo complexes of Mo and W are also
highly active in both olefin epoxidation and sulfur oxygenation
reactions [14–20].
(O2)2}2]2ꢀ is indentified as (Bu4N)2[{MoO(O2)2}2(
l-O)] (1) on the
basis of X-ray diffraction study. Described in this contribution are
the synthetic details of 1, and its reactivity in sulfide oxygenation
both as a stoichiometric oxygen donor and as a catalyst for hydro-
gen peroxide oxygenation. A plausible mechanism is proposed
based on the kinetics study, and DFT calculations were employed
to probe the transition state as well.
Our laboratory has been interested in homogeneous catalysts
for the oxygenation of organic sulfides, which is highly relevant
⇑
Corresponding author. Tel.: +1 765 494 5466.
E-mail address: tren@purdue.edu (T. Ren).
http://dx.doi.org/10.1016/j.ica.2015.08.016
0020-1693/Ó 2015 Published by Elsevier B.V.

104
D.J. Thompson et al. / Inorganica Chimica Acta 437 (2015) 103–109
Mizuno [9] and Cavaleiro [11]. The coordination environment of
1, especially the arrangement of four peroxo moieties, is nearly
2. Results and discussion
identical to that observed for [{WO(O2)2}2(l .
-O)]2ꢀ
2.1. Synthesis and characterization of (Bu4N)2[{MoO(O2)2}2(l-O)] (1)
The title compound was initially prepared from the reaction of
sodium molybdate with hydrogen peroxide in the presence of sele-
nic acid. Addition of tetrabutylammonium bromide in excess to the
reaction mixture resulted in the precipitation of 1 as a yellow
crystalline solid in a modest yield of 34%. The role of selenic acid
is only related to its effect on the pH of the reaction mixture, as
compound 1 can also be prepared in a yield of 35% from sodium
molybdate and hydrogen peroxide using HNO3 to adjust the pH
to ca. 6.0. Compound 1 has been characterized with FT-IR and
UV–Vis techniques. Specifically, the ESI-MS of compound 1 in
acetonitrile (see the Supplementary data) features a prominent
peak at 610 (m/z) corresponding to the monoanion {Bu4N[(MoO
2.3. Catalytic sulfide oxygenation by 1
Initial assessment of the activity of 1 was based on the H2O2
oxygenation of methyl phenyl sulfide (MPS) (Scheme 1), along
with control experiments of both the reaction without 1 and the
reaction with sodium molybdate. These reactions were all carried
out with 2.5 mmol MPS and 1 equiv of hydrogen peroxide at room
temperature in methanol, in which sodium molybdate has appre-
ciable solubility. The reaction without 1 resulted in no detectable
amount of products in 10 min and only 52% of sulfoxide in 24 h
[37], while the reaction with sodium molybdate (0.2 mol%) pro-
duced 86% of sulfoxide in 24 h. In contrast, the reaction with 1 at
0.2 mol% loading produced 40% sulfoxide 10 min, illustrating the
remarkable efficiency of 1 in activating H2O2.
The ability of 1 in catalyzing hydrogen peroxide was further
investigated through the reaction between MPS and 2 equiv of
hydrogen peroxide with 0.2 mol% of 1 in acetonitrile. The product
distribution determined with gas chromatography is shown in
Fig. 2. Catalyst 1 exhibited a short induction period, evident from
sampling during the first 5 min. At 30 min the conversion of sulfox-
ide to sulfone became concurrent with the conversion of sulfide to
sulfoxide. After the sulfide was almost entirely consumed (55 min),
the turn over frequency (TOF) was calculated to be 570 hꢀ1. Finally,
the sulfide to sulfone conversion was completed at ca. 7 h,
indicating a remarkable 100% utility of the hydrogen peroxide with
1. In comparison, the selenate-bridged ditungstate, [SeO4{WO
(O2)2}2]2ꢀ, facilitates complete oxygenation of MPS by H2O2 at
0.2 mol% loading in 30 min, and is slightly faster than 1. Compara-
ble dimolybdenum peroxo species [(Ph2PO2)(MoO(O2)2)2]ꢀ and
{(PhPO3)[(MoO(O2)2)2][Mo(O2)2(H2O)]}2ꢀ also catalyzed H2O2
oxygenation of MPS with TON of ca. 200 at 2 h [38].
Encouraged by the proficiency of 1 in the oxygenation of MPS,
the catalytic performance of 1 was examined with additional sub-
strates listed in Chart 1 under the same conditions as those for the
MPS reaction, and the results are collected in Table 1. Using one
equiv of H2O2, the reaction of BPS produced 53% sulfoxide and
23% sulfone at 3 h, which indicates both a 100% utility of H2O2
and lack of selectivity for sulfoxide. With 2.0 equiv of H2O2, BPS
was converted to the corresponding sulfone at 24 h. Compared to
the reactions of MPS and BPS, the reaction with PPS is significantly
slower, yielding 50% sulfoxide and 30% sulfone at 4 h. The reaction
with 4BT was slightly faster than that of MPS and resulted in 81%
sulfoxide and 5% sulfone in 0.5 h (TOF: 910 hꢀ1). Lastly, PTE
proceeded to 66% sulfoxide and 9% sulfone in 0.5 h for a TOF of
660 hꢀ1. At 24 h, the reaction with PTE still contained 10% sulfox-
ide along with 86% sulfone. This is possibly due to the conversion
of PTE to the corresponding aldehyde as a minor side product.
Occasionally, a radical reaction mechanism has been invoked
for aerobic oxygenation facilitated by polyoxomolybdates
[39–41]. The MPS oxygenation reaction carried out in the presence
of TEMPO (2,2,6,6-tetramethylpiperidine 1-oxyl) resulted in a
speciation curve (Fig. S3 in Supplementary data) very similar to
that of Fig. 2, eliminating the possibility of a radical reaction
pathway.
(O2)2)2(
[(MoO(O2)2)2(
compound with an oxo/peroxo coordination sphere identical to 1,
namely [{WO(O2)2}2(
-O)]2+ (yellow), was converted to a colorless
peroxo-bridged derivative [{WO(O2)2}2(
-O2)]2+ upon addition of
18 equiv of H2O2 [9]. In the case of 1, the yellow color persists upon
l
-O)]}1ꢀ (1-Bu4N+), which confirms the integrity of the
-O)]2ꢀ core in solution. Previously, a ditungsten
l
l
l
addition of excessive H2O2, indicating the absence of
peroxo-bridged Mo analog.
a
2.2. Molecular structure of 1
Single crystals suitable for X-ray diffraction were grown from
an acetonitrile-toluene solution of 1 at 4 °C. The structural plot of
the anion, [{MoO(O2)2}2(
selected bond lengths and angles are provided in the caption.
The dinuclear core is held together by the -oxo (O6) ligand along
with weak Mo—
2-peroxo dative bonding indicated by the dashed
l
-O)]2ꢀ, is shown in Fig. 1, while the
l
g
lines (Mo1–O7 and Mo2–O4). The Mo1–Mo2 distance (3.12 Å) is
significantly longer than the sum of the covalent radii of two Mo
centers (2.58 Å), which is indicative of the absence of covalent
bonding between the two Mo(VI) centers. In addition to the
l
two
-oxo ligand, each Mo(VI) center has a terminal oxo ligand and
g
2-peroxo ligands. The Mo1–O5 (1.668(3) Å) and Mo2–O11
(1.691(3) Å) bond lengths both fall into a typical range for Mo@O
bond [35]. All of the peroxo OAO bond lengths are similar, ranging
between 1.47 and 1.48 Å. The structure of 1 is distinctive from
many dimolybdenum peroxo species described in a recent review
[36], and hence represents a new motif for polyoxomolybdates. It
should be noted that the W analog, [{WO(O2)2}2(l
-O)]2ꢀ, has been
structurally characterized independently by the laboratories of
O
S
O
O
Fig. 1. Structural plot of the anionic core of 1. Selected bond lengths (Å): Mo1–O1
1.924(3); Mo1–O2 1.919(3); Mo1–O3, 1.928(3); Mo1–O4, 1.977(2); Mo1–O5 1.668
(3), Mo1–O6 1.940(3); Mo2–O6, 1.937(3); Mo2–O7, 1.952(3); Mo2–O8, 1.927(3);
Mo2–O9, 1.947(3) Mo2–O10, 1.937(3), Mo2–O11, 1.691(3); O1–O2, 1.469(5); O3–
O4, 1.475 (4); O7–O8, 1.472(4); O9–O10, 1.479(4) and angles (deg) Mo1–O6–Mo2,
107.4(1).
S
S
H2O2
+
catalyst
Scheme 1. Oxygenation of MPS to sulfoxide and sulfone.

D.J. Thompson et al. / Inorganica Chimica Acta 437 (2015) 103–109
105
1
100
80
60
40
20
0
MPS
MPSO
0.8
0.6
0.4
0.2
0
-NH2
MPSO2
-OMe
-H
-Br
-0.2
-0.4
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0
50
100
150
200
250
σp
Time (min)
Fig. 3. Hammett correlation for the oxygenation of p-substituted thioanisoles
catalyzed by 1.
Fig. 2. Reaction profile of H2O2-oxygenation of MPS catalyzed by 1. The reaction
was performed with 2.5 mmol MPS, 5.0 mmol H2O2, 5.0
and sampled at the indicated times.
lmol of 1 in 5.0 mL CH3CN,
several para-substituted thioanisoles, which were obtained by
monitoring UV absorbance at 290 nm as previously described
[31,42]. The Hammett plot (log(kX/kH) versus
r) for the oxygena-
S
tion reaction is shown in Fig. 3. The satisfactory linear fit
(R2 = 0.98) suggests that the oxygenation of these thioanisoles pro-
ceeds by the same mechanism. The negative reactivity constant
S
(
q
= ꢀ1.06) implies an electron deficient transition state, and
Benzyl phenyl sulfide (BPS)
Phenyl sulfide (PPS)
S
hence the electrophilic nature of 1 as an oxo-transfer catalyst.
Comparable reactivity constants have been reported for oxygena-
tion of para-substituted thioanisoles catalyzed by [SeO4{WO
(O2)2}2]2ꢀ (ꢀ0.62) [34] and olefin expoxidation using H2O2 cat-
alyzed by [(PhPO3){WO(O2)2}2{WO(O2)2H2O}]2ꢀ (ꢀ0.9) [43].
The presence of four peroxo groups in 1 begs the question how
many of them, if any, are active oxygen (‘‘O”) donors. To address
this issue, a stoichiometric reaction between MPS (0.50 mmol)
and 1 (0.06 mmol) was carried out at ambient temperature, which
resulted in 20% sulfoxide (0.10 mmol ‘‘O”) and 2% sulfone
(0.02 mmol ‘‘O”) at 24 h, and no further product formation over
the next six days. This suggests strongly the transfer of two equiv
of active oxygen from 1 to the substrates. Upon the completion of
the aforementioned stoichiometric reaction, 2 equiv of H2O2 were
added to the mixture resulting in a complete conversion to sulfone
within 45 min, demonstrating the regeneration of the active cat-
alytic species after the loss of active oxygen from 1. Furthermore,
the absorption shoulder for 1 at 320 nm was entirely ‘‘bleached”
at the end of the stoichiometric MPS oxygenation, and it was
restored immediately upon the addition of H2O2 (Fig. 4). Clearly,
the shoulder at 320 nm is a strong indicator for the presence of
the active peroxo species in the reaction system, and its fast regen-
eration indicates that the rate of the peroxo species formation is
significantly faster than the rate of the oxo-transfer step.
S
OH
Br
Phenyl thio ethanol(PTE) 4-Bromo thioanisole(4BT)
Chart 1. Other sulfide substrates.
Table 1
Results of sulfide oxygenation with 1.
Substrate equiv
H2O2
Entry
h
%
RSR0
%
%
TOFb
RSOR0
RSO2R0
BPS
1.0
1
2
1.0
3.0
53
23
41
53
7
275
–
23
BPS
2.0
3
4
5
1.0
3.0
24
55
4
0
39
46
0
4
47
235
230
–
51c
MPS
PPS
1.0
6
7
0.75 25
2.5
65
69
9
550
–
16
13
8
9
10
1.0
4.0
24
83
18
0
16
50
4
0
30
95
80
140
–
2.0
2.0
4BT
PTE
11
12
13
14
0.17 72
0.5
1.0
25
81
77
0
0
5
735
910
595
–
To further the mechanistic understanding, kinetic experiments
under catalytic conditions were carried out. The elemental steps
of the oxygenation reaction are similar to those described for reac-
tions catalyzed by the complex [SeO4{WO(O2)2}2]2ꢀ [34], shown in
Eq. (1) and (2) (1b is defined in Scheme 2).
14
0
0
21
99
24
15
16
17
0.5
2.0
24
22
0
0
66
66
10
9
30
86
660
315
–
2.0
k1
a For each of the substrates, the reaction was carried out in acetonitrile with 2 equiv
of H2O2. For benzyl phenyl sulfide (BPS) and MPS the reactions were also carried out
with 1 equiv of H2O2; 1 at 0.2 mol% loading in all reactions.
MPS þ 1
MPSO þ 1b
ð1Þ
ð2Þ
ƒƒƒƒ!
b
TOF = Turn over frequency (hꢀ1) = {[RR0SO] + 2[RR0SO2]}/{[cat] * time (h)}.
k2
ƒƒƒƒ!
H O þ 1b
H O þ 1
2
ƒƒƒƒ
2
2
c
Part of BPS sulfone (45%) was recovered as solid, and not included in the 51% GC
k
ꢀ2
yield.
Applying steady state approximation and the mass balance of
To further understand the nature of the transition state of the
reactions catalyzed by 1, a linear free energy relationship was
established by investigating the initial rates (k) of oxygenation of
both forms of 1, the rate law can be shown to be Eq. (3), which
is the same as that describing the oxygenation by [SeO4{WO
(O2)2}2]2 [34].

106
D.J. Thompson et al. / Inorganica Chimica Acta 437 (2015) 103–109
0.005
2500
2000
1500
1000
500
(a)
(b)
(c)
0.004
0.003
0.002
0.001
0
ε
0
320
340
360
380
400
0
20
40
60
80
100
120
λ (nm)
Equiv. H2O2
Fig. 4. Absorption spectra of (a) 1; (b) reaction mixture at the end of stoichiometric
reaction; and (c) after the addition of H2O2 to (b).
Fig. 5. Rate dependence on [H2O2] for the oxygenation of MPS with 1.
stoichiometric reactions and the lack of dependence on [H2O2]
under catalytic conditions, it is inferred that k2 ꢂ k1 and kꢀ2, effec-
tively reducing Eq. (3) to Eq. (4).
O
O
O
O
O
O
O
O
Mo
O
o
M
O
O
1
S
R0 ꢄ k1½MPSꢁ½1ꢁt
ð4Þ
H2O
The kobs extracted was plotted against [1] (Fig. 6), where a linear
relationship with [1] up to 5.1 ꢃ 10ꢀ4 M (10% loading) is clear.
Linear dependence on the initial rate was established for 1 with
a kcat (=k1[MPS]) of 11 Mꢀ1 minꢀ1. The first order dependence is
consistent with a single molecule of 1 being involved in the rate
determining step.
Step 1
Step 3
H2O2
H2O
S
O
H2O2
O
O
Step 5
O
O
Mo
O
O
o
M
Mo
O
O
O
2.4. Mechanistic discussion
O
O
O
1b O
O
O
Mo
O
Mo
Step 2
O
O
On the basis of the above-mentioned results, the oxygenation of
organic sulfides catalyzed by 1 probably proceeds in the sequence
described in Scheme 2. First, the sulfide (or later in the reaction a
O
O
1c
RSR'
Step 4
O
S
sulfoxide) attacks a g
2-peroxo (Step 1), which is followed by the
formation of the sulfoxide and a Mo@O group (1b) (Step 2). At this
point, 1b can either be converted back to 1 upon reacting with
hydrogen peroxide to close the catalytic cycle (Step 3) or react with
another sulfide (or sulfoxide) to generate sulfoxide (or sulfone) and
1c. The stoichiometric reaction indicates that this step does not
lead to catalyst deactivation, and 1c can indeed be converted back
to 1b and subsequently 1 with the addition of two equiv of hydro-
gen peroxide (Steps 5 and 3). However, due to the rapidity of the
RSOR'
Scheme 2. Proposd mechanism for sulfide oxygenation with 1.
k1k2½MPSꢁ½1ꢁt½H2O2ꢁ
R0
¼
ð3Þ
k1½MPSꢁ þ k2½H2O2ꢁ þ k ½H2Oꢁ
ꢀ2
The kinetic data are based on the initial rate method, which
monitors the disappearance of MPS at 290 nm as in the
above-mentioned LFER analysis. The rate dependence on [H2O2]
was analyzed in the presence of 5 mM MPS and 0.25 mM 1. The
disappearance of MPS during the first 40 min was monitored for
a set of reactions with [H2O2] ranging from 60 mM (12 equiv) to
500 mM (100 equiv), and the kobss were extracted from the plot
of ln(Abs) versus time and plotted against the equivalence of
H2O2 in Fig. 5. The lack of dependence on [H2O2] for the catalytic
reactions shown in Fig. 5 indicates that k2 ꢂ k1, and is consistent
with the observed rapidity of the regeneration of the active peroxo
species noted in the stoichiometric reactions. In contrast, a similar
kinetic study of oxygenation with [SeO4{WO(O2)2}2]2ꢀ as the cata-
lyst revealed that k1 and k2 are comparable [34].
The rate dependence on [1] was similarly analyzed with [1]
ranging from 0.077 mM to 0.51 mM, while [MPS] and [H2O2] were
held at 5 mM and 125 mM, respectively. The kobs obtained range
from 1.0 ꢃ 10ꢀ3 minꢀ1 for [1] = 0.077 mM to 5.8 ꢃ 10ꢀ3 minꢀ1 [1]
= 0.51 mM. Higher concentrations of 1 were not investigated due
to the increasing absorbance contribution from the catalyst. Based
on both the fast regeneration of the catalytic species after the
0.007
0.006
0.51 mM
0.005
0.004
0.26 mM
0.003
0.002
0.001
0
0.15 mM
0.10 mM
0.077 mM
0
0.0001
0.0002
0.0003
0.0004
0.0005 0.0006
[1]
Fig. 6. Reaction rate dependence on [1] for the oxygenation of MPS with hydrogen
peroxide.

D.J. Thompson et al. / Inorganica Chimica Acta 437 (2015) 103–109
107
regeneration of 1 from 1b with hydrogen peroxide (Step 3), the
involvement of 1c is insignificant when hydrogen peroxide is in
large excess.
dimethyl sulfide (Me2S) instead of MPS (shown in Fig. S4 of
Supplementary data). The calculated activation barriers of sulfide
attacks at O3–O8 (76–83 kJ molꢀ1) were lower than those of at
O4–O7 (92–94 kJ molꢀ1), O1–O9 (82–86 kJ molꢀ1), and O2–O10
(83–87 kJ molꢀ1). The intermediate structures and corresponding
activation energies are provided in Supplementary data (Fig. S5).
Our finding that the sulfide attack at O3 is the most favorable is
consistent with the computational studies of sulfide oxidation
[34] and olefin epoxidation [44,45] catalyzed by [SeO4{WO
2.5. DFT calculations
In order to verify the proposed reaction mechanism, the density
functional theory calculations were performed on the B3LYP/
LanL2DZ level. The energies and peroxo intermediate in the
oxidation of methyl phenyl sulfide to methyl phenyl sulfoxide
are provided in Fig. 7. Full geometric optimizations were
performed on 10, 1b0 and 1c0, the models for compounds 1, 1b
and 1c, respectively. The geometry of 10 was optimized from the
crystal structure of 1. The structures of 1b0 and 1c0 were initially
(O2)2}2]2ꢀ
, where similar transition state geometries were
observed. For the assessment of both the reaction free energy
and activation barrier, only the MPS attack at O3 pathway was cal-
culated, and the formation of 1b0 and methyl phenyl sulfoxide from
the reaction of methyl phenyl sulfide with 10 was found to be
exothermic by ꢀ69 kJ molꢀ1. The transition-state structure was
found to exhibit spiro orientations in which the sulfur group is
nearly perpendicular to the plane defined by the peroxo group
and the Mo center. The activation energy for the oxygen transfer
constructed by replacement of one and two Mo–(g
2–O2) with
Mo@O in 10, respectively, and followed by full optimizations. The
bond lengths and angles around the Mo core in the optimized
geometries match well with those from the crystal structures,
and are given in the Supplementary data.
reaction (TS1) was calculated to be 75 kJ molꢀ1
.
The optimized Mo–O6 bond lengths of 10 (1.934 Å) are compa-
rable with the X-ray experimental values of the anionic core of 1
(1.937 and 1.940 Å). The remaining Mo–O bond lengths of 10 are
elongated with respect to X-ray determined values. Both Mo@O
bonds, Mo1–O5 (1.668 Å) and Mo2–O11 (1.691 Å), are relaxed to
1.728 and 1.729 Å, respectively. The optimized peroxo O–O bond
lengths fall between 1.526 and 1.528 Å (experimental range:
2.6. Conclusions
This work discloses the ability of 1 in promoting rapid H2O2
oxygenation of various organic sulfides with high TOFs. The study
of 1 as a stoichiometric oxygen donor reveals the existence of two
active oxygens, and that the dimeric structure remains intact after
the transfer of both active oxygens. It is also shown that the regen-
eration of the first active oxygen in 1 is significantly faster than the
oxygenation step, effectively keeping the second active oxygen out
of the catalytic cycle in the presence of excess hydrogen peroxide.
1.469–1.479 Å). The optimized bond lengths for Mo–O (g
2–O2)
range from 1.977 to 2.009 Å, which are about 0.05 Å longer than
the experimental values (1.927–1.952 Å). Upon the full optimiza-
tion, the Mo1–O6–Mo2 angle has been further increased from
107.4° (X-ray) to 167.1°.
In 1b0, Mo1–O6 bond length has been shortened to 1.920 Å
(1.937 Å in 10) due to the formation of the new Mo@O. The two
Mo@O bonds, Mo1–O5 and Mo1–O3, are further elongated to
1.759 Å from the 1.728 Å in 10. Upon formation of another Mo@O
group in 1c0, the model compound becomes symmetric with the
two Mo-units related by an inversion center at O6. The
Mo1–O6–Mo2 angle is further enlarged to 180.0° from 171.8° of 1c0.
To save the computational cost, the reactivity of each of the four
peroxo groups in 1 toward organic sulfide was examined using
3. Experimental
3.1. Materials and instrumentation
Acetonitrile and materials for the synthesis of 1, sodium
molybdate and selenic acid, were purchased from Sigma Aldrich.
Methyl phenyl sulfide, diphenyl sulfide, benzyl phenyl sulfide,
4-bromothioanisole, and phenylthioethanol were purchased from
Fig. 7. Calculated energy diagram of the oxidation of methyl phenyl sulfide to methyl phenyl sulfoxide by 1 in the gas phase (energies in kJ molꢀ1). Light blue, red, yellow,
black, and light gray balls represent molybdenum, oxygen, sulfur, carbon, and hydrogen atoms, respectively. (For interpretation of the references to color in this figure legend,
the reader is referred to the web version of this article.)

108
D.J. Thompson et al. / Inorganica Chimica Acta 437 (2015) 103–109
ACROS Organics. Hydrogen peroxide (30%) was purchased from
Macron Fine Chemicals and standardized via iodometric titration.
Oxygenation reaction samples were analyzed using an Agilent
7890A GC system equipped with a flame ionization detector. The
separation of substrate and products was achieved using an
Agilent HP-5 column with dimensions of 30 m ꢃ 0.320 mm with
25 micron film thickness. Reaction progress was monitored at
the addition of 2.5 mmol of TEMPO before the reaction was
initiated.
Stoichiometric reaction was carried out by first dissolving 1
(0.06 mmol or 0.03 mmol) into 5 mL acetonitrile followed by
1,2-dichlorobenzene (0.4 mmol) as internal standard. The reaction
was initiated by the addition of MPS (0.5 mmol) and monitored by
GC analysis. The reaction completed upon 2 equiv of oxygen atoms
being transferred from 1 to MPS. Addition of hydrogen peroxide
(1.0 mmol) to the mixture restarted the oxygenation reaction and
the reaction progress was monitored as described above for the
catalytic reactions.
290 nm via UV–Vis spectroscopy on
Spectrophotometer.
a
JASCO V-670
3.2. Synthesis of 1
3.5. Kinetic dependence on hydrogen peroxide concentration
3.2.1. 2a. Using selenic acid
To a solution of sodium molybdate (1.7 g in 5 mL water) was
added 0.52 mL (2.0 mmol) of 40% selenic acid, which was followed
by slow addition of 4.25 mL of 30% H2O2 during which the solution
turned dark red and bubbled. Upon the completion of H2O2 addi-
tion, the solution changed from red to light yellow and ceased to
bubble. Addition of tetrabutylammonium bromide (2.7 g) resulted
in the precipitation of 1 as a canary crystalline material, which was
collected by filtration and washed with de-ionized water. Yield:
Standard solutions with MPS and 1 were made by dissolving
17.5 mg of 1 in 10.0 mL acetonitrile followed by the addition of
47 lL MPS. This yielded Solution I which was 0.040 M in MPS
and 5% (2.0 ꢃ 10ꢀ3 M) 1. Hydrogen peroxide solution was prepared
by diluting 2.05 mL of 30% hydrogen peroxide solution to 5.0 mL
total volume with acetonitrile to yield Solution II. Kinetics reac-
tions were prepared by adding 400 lL of Solution I to a cuvette
1.20 g (34% based on Mo); FT-IR: 945, 885, 854, 740, 630 cmꢀ1
UV–Vis (kmax, nm (
, Mꢀ1 cmꢀ1)): 203 (19,600), 218 (19,600), 250
(6200), 320 (1600).
;
with enough acetonitrile that upon initiation with Solution II the
total volume was 3.2 mL. The volume of Solution II was varied to
give different concentrations of hydrogen peroxide, ranging from
12 equiv to 100 equiv hydrogen peroxide with respect to MPS.
The absorbance at 290 nm was measured every 30 seconds for 1 h.
e
3.2.2. 2b. Using nitric acid
the ph of an aqueous solution of sodium molybdate (1.7 g in
15 mL de-ionized water) was adjusted to pH ꢄ6.0 using 2 M
HNO3, which was followed by the drop wise addition of 4.5 mL
of 30% H2O2. The product was precipitated with addition of
tetrabutylammonium bromide (2.7 g), collected by filtration and
washed with de-ionized water. Yield: 1.24 g (35%).
3.5.1. Kinetic dependence on [1]
Standard solutions were prepared for each concentration of cat-
alyst. For example, the 2% catalyst solution (for 1.03 mM catalyst
reaction) was made by dissolving 7 mg of 1 in 10 mL of MeCN, fol-
lowed by the addition of 47 lL MPS. This was used in the same
manner as described for Solution I above, and enough Solution II
was used to yield 25 equiv of hydrogen peroxide. These reactions
were monitored in the same fashion as described above, except
that they were monitored for 2 h.
3.3. X-ray crystallography of 1
Single crystals of 1 were grown from slow evaporation of an
acetonitrile-toluene solution of 1 at 4 °C. A yellow plate with the
approximate dimensions of 0.62 ꢃ 0.60 ꢃ 0.12 mm was mounted
3.5.2. Hammett analysis
Stock solutions of 0.020 M para-substituted thioananisoles (–Br,
–H, –OMe, –NH2) along with 5 mol% catalyst (0.0010 M) were
made in acetonitrile. A stock solution of hydrogen peroxide was
made by diluting 1.0 mL of 30% hydrogen peroxide with 4.0 mL
acetonitrile for a 1.95 M solution of hydrogen peroxide. For each
on a Nonius KappaCCD diffractometer (k(Mo Ka) = 0.71073 Å) at
150 K for data collection. The structure was solved using the direct
method and refined using SHELX 2013 [46]. Crystal Data for 1:
Mo2O11C32H72N2, FW = 852.82, monoclinic, P21/c, a = 10.5236(3),
b = 32.900(1), c = 11.7249(4) Å, b = 92.762(2)°, V = 4054.8(2) Å3;
Z = 4, Dcalc = 1.397 g cmꢀ3. Of 29348 reflections measured, 7692
were unique. Least squares refinement based on 4937 reflections
reaction, 160
acetonitrile and the reaction was initiated by the addition of
190 L hydrogen peroxide solution. The reactions were referenced
lL of sulfide-catalyst solution was added to 2.85 mL
l
with I P 2r(I) and 432 parameters led to convergence with final
R1 = 0.046 and wR2 = 0.110. CCDC Number: 1400687.
to acetonitrile and were all monitored at 290 nm until a baseline
was achieved.
3.4. Sulfide oxygenation reactions
3.6. Computation details
The catalytic sulfide oxygenation reactions were generally
carried out as follows: 2.5 mmol of indicated sulfide was added
to 4 mL acetonitrile followed by 1.0 mL of catalyst stock solution
prepared to contain 5.0 ꢃ 10ꢀ3 mmol catalyst in each 1.0 mL for
a resulting 0.2% catalyst loading. To this solution was added
2.0 mmol of 1,2-dichlorobenzene as the internal standard. The
reactions were typically initiated by the addition of 5 mmol hydro-
gen peroxide from a 30% solution whose concentration was deter-
mined via iodometric titration. Samples were taken at the
The model compound 10 was built based on the anionic part of
the crystal structure of 1, while 1b0 and 1c0 were also constructed
based on the crystal structure of 1 with one and two Mo(g
2–O2)
being replaced with Mo@O, respectively. Full optimizations on all
three models have been carried out using the DFT formalism
implemented in the GAUSSIAN 03 program [47], with the B3LYP func-
tional [48–50]. The choices of basis sets are 6-31+G(d,p) basis sets
for H, C, S, O atoms, and LANL2DZ basis set for Mo atoms. The over-
all charge of all three systems is ꢀ2. To examine the reactivity of
different peroxo oxygen atoms (O1–O4 and O7–O10) in 1 with
respect to sulfoxidation, dimethyl sulfide was used as the model
substrate. Transition-state structures were searched by numeri-
cally estimating the matrix of second-order energy derivatives at
every optimization step and by requiring exactly one eigenvalue
of this matrix to be negative. Moreover, the transition-state
indicated times by extracting 20
quenching hydrogen peroxide therein with about 5 mg MnO2.
The quenched reactions were diluted with 80 L acetonitrile before
lL of the reaction solution and
l
analysis with gas chromatography. The reaction with radical scav-
enger 2,2,6,6-tetramethylpiperidine 1-oxyl (TEMPO) was carried
out in the same fashion as the reactions described above with

D.J. Thompson et al. / Inorganica Chimica Acta 437 (2015) 103–109
109
[17] M. Abrantes, A.M. Santos, J. Mink, F.E. Kuhn, C.C. Romao, Organometallics 22
(2003) 2112.
[18] G. Wang, G. Chen, R.L. Luck, Z.Q. Wang, Z.C. Mu, D.G. Evans, X. Duan, Inorg.
Chim. Acta 357 (2004) 3223.
structure and activation barrier for the oxidation of methyl phenyl
sulfide associated with O3 of 1 was calculated.
[19] A. Jimtaisong, R.L. Luck, Inorg. Chem. 45 (2006) 10391.
[20] M. Buchweitz, R.L. Luck, S. Sreehari, A. Beganskiene, E. Urnezius, Inorg. Chim.
Acta 363 (2010) 1818.
Acknowledgement
We thank Purdue University for the financial support.
[21] I.V. Babich, J.A. Moulijn, Fuel 82 (2003) 607.
[22] B.M. Smith, Chem. Soc. Rev. 37 (2008) 470.
[23] G.W. Wagner, Y.C. Yang, Ind. Eng. Chem. Res. 41 (2002) 1925.
[24] Y.-C. Yang, Acc. Chem. Res. 32 (1998) 109.
Appendix A. Supplementary material
[25] Y.C. Yang, J.A. Baker, J.R. Ward, Chem. Rev. 92 (1992) 1729.
[26] J.E. Barker, T. Ren, Tetrahedron Lett. 45 (2004) 4681.
[27] J.E. Barker, T. Ren, Tetrahedron Lett. 46 (2005) 6805.
[28] J.E. Barker, T. Ren, Inorg. Chem. 47 (2008) 2264.
[29] H.B. Lee, T. Ren, Inorg. Chim. Acta 362 (2009) 1467.
[30] L. Villalobos, Z. Cao, P.E. Fanwick, T. Ren, Dalton Trans. 41 (2012) 644.
[31] L. Villalobos, J.E. Barker Paredes, Z. Cao, T. Ren, Inorg. Chem. 52 (2013)
12545.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.ica.2015.08.016.
References
[1] I.V. Kozhevnikov, Chem. Rev. 98 (1998) 171.
[32] D.J. Thompson, J.E.B. Paredes, L. Villalobos, M. Ciclosi, R.J. Elsby, B. Liu, P.E.
Fanwick, T. Ren, Inorg. Chim. Acta 424 (2015) 150.
[2] C.L. Hill, C.M. Prossermccartha, Coord. Chem. Rev. 143 (1995) 407.
[3] L. Salles, C. Aubry, R. Thouvenot, F. Robert, C. Doremieuxmorin, G. Chottard, H.
Ledon, Y. Jeannin, J.M. Bregeault, Inorg. Chem. 33 (1994) 871.
[4] D.C. Duncan, R.C. Chambers, E. Hecht, C.L. Hill, J. Am. Chem. Soc. 117 (1995)
681.
[5] W.P. Griffith, B.C. Parkin, A.J.P. White, D.J. Williams, J. Chem. Soc., Dalton Trans.
(1995) 3131.
[6] N.M. Gresley, W.P. Griffith, B.C. Parkin, A.J.P. White, D.J. Williams, J. Chem. Soc.,
Dalton Trans. (1996) 2039.
[7] J.Y. Piquemal, C. Bois, J.M. Bregeault, Chem. Commun. (1997) 473.
[8] J.Y. Piquemal, L. Salles, G. Chottard, P. Herson, C. Ahcine, J.M. Bregeault, Eur. J.
Inorg. Chem. (2006) 939.
[9] K. Kamata, S. Kuzuya, K. Uehara, S. Yamaguchi, N. Mizuno, Inorg. Chem. 46
(2007) 3768.
[10] L. Salles, J.Y. Piquemal, Y. Mahha, M. Gentil, P. Herson, J.M. Bregeault,
Polyhedron 26 (2007) 4786.
[11] I. Santos, F.A.A. Paz, M.M.Q. Simoes, M. Neves, J.A.S. Cavaleiro, J. Klinowski, A.
M.V. Cavaleiro, Appl. Catal., A. Gen. 351 (2008) 166.
[12] R. Ishimoto, K. Kamata, N. Mizuno, Angew. Chem., Int. Ed. 51 (2012) 4662.
[13] A.M. Al-Ajlouni, O. Saglam, T. Diafla, F.E. Kuhn, J. Mol. Catal. A: Chem. 287
(2008) 159.
[14] F.E. Kuhn, A.M. Santos, W.A. Herrmann, Dalton Trans. (2005) 2483.
[15] F.E. Kuhn, A.M. Santos, M. Abrantes, Chem. Rev. 106 (2006) 2455.
[16] F.E. Kuhn, M. Groarke, E. Bencze, E. Herdtweck, A. Prazeres, A.M. Santos, M.J.
Calhorda, C.C. Romao, I.S. Goncalves, A.D. Lopes, M. Pillinger, Chem. Eur. J. 8
(2002) 2370.
[33] T.D. Phan, M.A. Kinch, J.E. Barker, T. Ren, Tetrahedron Lett. 46 (2005) 397.
[34] K. Kamata, T. Hirano, R. Ishimoto, N. Mizuno, Dalton Trans. 39 (2010) 5509.
[35] M. Grzywa, W. Nitek, W. Lasocha, J. Mol. Struct. 888 (2008) 318.
[36] V. Conte, B. Floris, Dalton Trans. 40 (2011) 1419.
[37] Average of two independent runs that agree within 1%.
[38] N.M. Gresley, W.P. Griffith, A.C. Laemmel, H.I.S. Nogueira, B.C. Parkin, J. Mol.
Catal. A: Chem. 117 (1997) 185.
[39] A.M. Khenkin, L. Weiner, Y. Wang, R. Neumann, J. Am. Chem. Soc. 123 (2001)
8531.
[40] H. Yang, J.Q. Chen, J. Li, Y. Lv, S. Gao, Appl. Catal., A 415 (2012) 22.
[41] S. Shinachi, M. Matsushita, K. Yamaguchi, N. Mizuno, J. Catal. 233 (2005) 81.
[42] T.S. Smith, V.L. Pecoraro, Inorg. Chem. 41 (2002) 6754.
[43] A.M. Al-Ajlouni, O. Saglam, T. Diafla, F.E. Kuhn, J. Mol. Catal. A: Chem. 287
(2008) 159.
[44] K. Kamata, T. Hirano, S. Kuzuya, N. Mizuno, J. Am. Chem. Soc. 131 (2009) 6997.
[45] K. Kamata, R. Ishimoto, T. Hirano, S. Kuzuya, K. Uehara, N. Mizuno, Inorg.
Chem. 49 (2010) 2471.
[46] G.M. Sheldrick, Acta Crystallogr., Sect. A 64 (2008) 112.
[47] Gaussian 03, Revision D.02, Gaussian Inc., 2003.
[48] A.D. Becke, J. Chem. Phys. 98 (1993) 5648.
[49] A.D. Becke, Phys. Rev. A 38 (1988) 3098.
[50] C.T. Lee, W.T. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785.