A combined theoretical and experimental investigation on the enantioselective oxidation of aryl benzyl sulfides in the presence of a chiral titanium catalyst

Article abstract of DOI:10.1002/chem.200902110

An experimental investigation of the enantioselective oxidation of aryl benzyl sulfides by tert-butyl hydroperoxide in the presence of a titanium/hydrobenzoin catalyst has shown that these sulfides are ideal substrates for this catalytic system, with negl

Full text of DOI:10.1002/chem.200902110

FULL PAPER
DOI: 10.1002/chem.200902110
A Combined Theoretical and Experimental Investigation on the
Enantioselective Oxidation of Aryl Benzyl Sulfides in the Presence of a
Chiral Titanium Catalyst
Francesco Naso,[a] Maria Annunziata M. Capozzi,[b] Andrea Bottoni,*[c]
Matteo Calvaresi,[c] Valerio Bertolasi,[d] Francesco Capitelli,[e] and
Cosimo Cardellicchio*[a]
Abstract: An experimental investiga-
tion of the enantioselective oxidation
of aryl benzyl sulfides by tert-butyl hy-
droperoxide in the presence of a titani-
um/hydrobenzoin catalyst has shown
that these sulfides are ideal substrates
for this catalytic system, with negligible
interference by the substituents on the
proposed. The DFT MPWB1K func-
tional was used in the theoretical inves-
tigation to account for weak hydrogen-
bonding and p interactions. The com-
puted reaction profile explains the ex-
perimentally observed enantioselectivi-
ty, which is determined by the thermo-
dynamics of the first phase of the reac-
tion. A detailed discussion of the hy-
drogen-bonding and p interactions that
drive the reaction along the observed
stereochemical path is given.
Keywords:
density
functional
calculations
·
enantioselectivity ·
oxidation · reaction mechanisms ·
titanium
aryl groups.
A reaction mechanism
based on DFT computations has been
Introduction
and bioactive compounds.[1,2] Chiral non-racemic sulfoxides
are prepared by two main strategies based upon the use of
organometallic reagents[2–5] or the enantioselective oxidation
of sulfides,[4–23] or by a suitable combination of these two ap-
proaches.[4,5] Because bioactive sulfoxides often bear moiet-
ies that are not compatible with organometallic procedures,
chiral sulfinyl drugs are prepared industrially mostly by
enantioselective oxidation,[7] as in the case of esomeprazole,
one of the best-selling drugs.[8] Thus, recent interest in more
efficient sulfide oxidation protocols is growing.[7] Although
several enantioselective oxidation reactions have been re-
ported,[6–23] the synthetic developments have not been ac-
companied by an increased understanding of the reaction
mechanisms.
The most common oxidation protocols employ an achiral
oxidant (hydroperoxides or peroxides) in the presence of
chiral metal complexes. Many metals have been investigat-
ed, as reported in the pioneering work[9] of Modena and
Kagan and their co-workers on titanium/diethyl tartrate
complexes[8,9] and in other papers on different titanium com-
plexes,[10–18] as well as on vanadium,[19–21] iron[22] or alumini-
um[23] catalysts, to cite the most frequently used metals.
Two main behaviour patterns were observed in these pro-
cesses. 1) A sulfoxide with a high ee value was produced di-
rectly without the formation of significant amounts of sul-
Enantiopure sulfoxides have been employed many times as
starting materials in several asymmetric syntheses of natural
[a] Prof. F. Naso, Dr. C. Cardellicchio
Dipartimento di Chimica and CNR ICCOM, Universitꢀ di Bari
via Orabon 4, 70125 Bari (Italy)
Fax : (+39)0805539596
E-mail: cardellicchio@ba.iccom.cnr.it
[b] Dr. M. A. M. Capozzi
Dipartimento di Scienze Agroambientali, Chimica e Difesa Vegetale
Universitꢀ di Foggia, Via Napoli 25, 71100 Foggia (Italy)
[c] Prof. A. Bottoni, Dr. M. Calvaresi
Dipartimento di Chimica G. Ciamician, Universitꢀ di Bologna
via Selmi 2, 40126 Bologna (Italy)
Fax : (+39)0512099456
E-mail: andrea.bottoni@unibo.it
[d] Prof. V. Bertolasi
Dipartimento di Chimica and Centro di
Strutturistica Diffrattometrica
Universitꢀ di Ferrara, via Borsari 46, 44100 Ferrara (Italy)
[e] Dr. F. Capitelli
CNR Istituto di Cristallografia
via Amendola 122/o, 70126 Bari (Italy)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200902110.
Chem. Eur. J. 2009, 15, 13417 – 13426
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
13417

fones.[8,9,12,14,15,17–19,21–23] 2) A sulfoxide with a lower ee value
was obtained, but the enantiopurity could be increased if
overoxidation to sulfone was allowed.[10–11,16,20] In fact, the
acting catalyst can enhance the ee of the product sulfoxide
by oxidising the minor enantiomer selectively.
In connection with our work on the application of carba-
nionic leaving groups to the synthesis of chiral sulfoxides,[4]
we reported that enantiopure benzyl p-bromophenyl sulfox-
ide (1b) is a general and viable precursor of chiral non-race-
mic dialkyl sulfoxides provided that two subsequent substi-
tution steps with alkyl Grignard reagents are carried out.[14]
The precursor can be easily obtained in very high ee (>
98%) and in high isolated yield (85%) with the aid of a
straightforward catalytic enantioselective oxidation of
benzyl p-bromophenyl sulfide (1a) with tert-butyl hydroper-
oxide (TBHP) in the presence of a titanium/hydrobenzoin
catalyst (Scheme 1).[14]
lower temperature (08C). Similar conditions, able to inhibit
the formation of sulfone,[11] had been reported by Rosini an
co-workers,[13] who oxidised various sulfides in the presence
of a titanium/hydrobenzoin catalyst with the formation of
the corresponding sulfoxides (55–73% yields, 22–80% ees
for aryl alkyl sulfoxides, 92 to >99% ees for a few aryl
benzyl sulfoxides). Going into detail, water was present in
their protocol and 2 equiv of a different solution of TBHP
were employed. Seto et al. used the titanium/hydrobenzoin
catalyst to prepare a sulfoxide that was a precursor of an
anti-HIV agent.[17] Note that a high ee (91%) and a satisfac-
tory yield (71%) were obtained when a molecule similar to
an aryl benzyl sulfide was oxidised at room temperature
without the formation of appreciable quantities of sulfone.
On the other hand, a slight structural variation of the sulfide
caused a reduction in the yield and the ee (18%) of the sulf-
oxide product. Very recently, Jiang et al.[18] reported the oxi-
dation of omeprazole-like sulfides with hydroperoxides in
the presence of complexes formed between titanium and hy-
drobenzoins bearing different substituents on the aryl
groups. Water was present. High yields (up to 92%) and ee
values (up to 96% ee) were obtained for these heteroaryl
molecules that resemble aryl benzyl sulfides.
From an overview of these results obtained under a varie-
ty of reaction conditions, the oxidation of aryl benzyl sul-
fides with hydroperoxides in the presence of a titanium/hy-
drobenzoin catalyst seemed to us to represent an ideal sub-
strate/reagent/catalyst system for a detailed stereochemical
and mechanistic investigation. With this aim, we decided to
perform a systematic experimental investigation of the oxi-
dation of a variety of aryl benzyl sulfides bearing different
substituents on the aryl groups under the same reaction con-
ditions. We carried out our experiments without inhibiting
the possible formation of the sulfone to test if and when ki-
netic resolution was present. In addition, because the mech-
anistic details of titanium-catalysed oxidation have not been
elucidated yet and, as far as we know, no accurate theoreti-
cal study is available in the literature, we investigated the
potential energy surface of the reaction by using a reliable
quantum mechanical DFT approach. To this purpose we em-
ployed a simple model system formed by a benzyl phenyl
sulfide molecule reacting with tert-butyl hydroperoxide
(TBHP) in the presence of a titanium/hydrobenzoin com-
plex.
Scheme 1. Enantioselective oxidation of benzyl p-bromophenyl sulfide in
the presence of a titanium/hydrobenzoin catalyst.
Both enantiomers of the hydrobenzoin ligand are cheap
and easily available. The (R)-sulfoxide was obtained when
(S,S)-hydrobenzoin was used, whereas the reaction afforded
the (S)-sulfoxide in the presence of (R,R)-hydrobenzoin.
Furthermore, the experimental procedure was insensitive to
the presence of water, was effective at room temperature
and could be performed in n-hexane,[14] a more environmen-
tally benign solvent than the chlorinated solvents commonly
used in other procedures. Moreover, with n-hexane the sulf-
oxide precipitated from the reaction medium and could
easily be collected. A simple crystallisation step yielded the
desired enantiopure material, thus allowing easy scale-up.[14]
Because this oxidation protocol is particularly efficient,
we compared our results with other work[13,17,18] and looked
for new applications.[15,16] High ee values and satisfactory
yields were obtained in the oxidation of aryl b-ketosulfides
(70–94% yields, 92 to >98% ees).[15] Also with these sulfinyl
compounds, the reaction gave (R)-sulfoxides when (S,S)-hy-
drobenzoin was used. The same protocol was applied to the
oxidation of two alkyl esters of Sulindac sulfide, structurally
related to an aryl methyl sulfide (95–96% ee values),[16] and
an identical stereochemical outcome was again observed.[16]
On the other hand, in this oxidation reaction, we observed a
sulfone, a lower yield of the sulfoxide and an increase in the
ee value due to a kinetic resolution process.[16] This pattern
was confirmed by lower ee values when the formation of sul-
fone was inhibited by a shorter reaction time (4–6 h) or
Results and Discussion
Reagents and reaction protocol: Aryl benzyl sulfides 2a–
15a[24–30] bearing different substituents were synthesised
starting from the commercially available aryl thiols and the
corresponding benzyl halides. As reported in Table 1, we
chose sulfides bearing electron-withdrawing or -releasing
substituents on the aryl groups. Furthermore, in principle,
these groups could supply a further point of coordination
for the titanium centre (for example, in the case of the me-
thoxy moiety) and exert steric hindrance.
13418
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 13417 – 13426

Enantioselective Oxidation of Aryl Benzyl Sulfides
FULL PAPER
Table 1. Enantioselective oxidation of aryl benzyl sulfides by TBHP in the presence of catalytic amounts of a
titanium/ACHTUNGTRENNUNG(S,S)- or (R,R)-hydrobenzoin complex.
Experimental results: After
screening
the
structures
(Table 1, entries 1–17) we were
in a position to confirm that the
enantioselective oxidation of
aryl benzyl sulfides with tert-
butyl hydroperoxide in the
presence of a titanium/hydro-
benzoin catalyst is extremely ef-
ficient from a synthetic point of
view. In fact, very often it was
possible to obtain in an easy
and straightforward manner the
corresponding sulfoxides 2b–
15b with very high enantiomer-
ic purity (>98% after crystalli-
sation). The yields were satis-
factory (57–92% with two ex-
ceptions), even when n-hexane
was used as the solvent.
Most of the reactions were
performed by using (S,S)-hy-
drobenzoin as the ligand for ti-
tanium, but two of them
(Table 1, entries 4 and 8) were
repeated with the (R,R)-hydro-
benzoin to obtain the enantio-
meric sulfoxide. In the cases of
sulfoxides 2b, 4b–6b, 8b–11b,
13b and 14b, the absolute con-
Entry Sulfide
R’
R’’
Ligand configuration Sulfoxide Yield [%][a] ee [%][b] eecryst [%][c]
1
2
2a
3a
2-COOMe
2-Cl
H
H
H
H
H
H
H
H
2-NO2
4-NO2
2-OMe
3-OMe
4-OMe
3-Cl
2,4-Cl2
2,4-Cl2
4-OMe
A
(R)-2b[d]
(R)-3b[e]
77
92
91
95
>98
>98
88
>98
ACHTUNGTRENNUNG
3
4a
4-NO2
4-NO2
2-OMe
3-OMe
4-OMe
4-OMe
4-Br
4-Br
4-Br
4-Br
4-Br
R
(R)-4b[d] 65[f]
>98
>98
>98
>98
>98
>98
>98
>98
>98
>98
>98
97
4
4a
G
(S)-4b[26]
(R)-5b[d]
(R)-6b[d]
(R)-7b[33]
(S)-7b[33]
(R)-8b[d]
(R)-9b[d]
(R)-10b[d]
(R)-11b[d]
(R)-12b[e]
(R)-13b[d]
38
89
57
67
65
72
71
68
37
65
66
5
5a
ACHTUNGTRENNUNG
6
6a
N
84
7
7a
R
>98
>98
85
>98
>98
81
>98
97
>98
85
8
7a
ACHTUNGTRENNUNG
9
8a
ACHTUNGTRENNUNG
10
11
12
13
14
15
16
17
9a
N
10a
11a
12a
13a
14a
14a
15a
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
G
4-Br
4-Br
4-Br
4-OMe
ACHTUNGTRENNUNG
G
(R)-14b[d] 82[g]
>98
ACHTUNGTRENNUNG
(R)-14b[d] 47[h]
A
(R)-15b[e]
72
>98
>98
[a] Isolated yields. [b] The ee values were measured by HPLC (see text) before crystallization. [c] The eecryst
values were measured by HPLC after crystallization. [d] The absolute configurations of the sulfoxides were es-
tablished by X-Ray analysis (see text). [e] This configuration was attributed on the basis of our empirical rule
(see text). [f] The reaction was performed in a 20:3 mixture of n-hexane/CCl4 due to the high insolubility of
sulfide 3a in pure n-hexane. [g] The corresponding sulfone was also present in a yield of 14%. [h] This reac-
tion was performed at 08C for 7 h.
The oxidation reactions of 2a–15a to obtain sulfoxides
2b–15b[26,31–33] were performed in n-hexane (Table 1). In
fact, the enantioselectivities of these reactions are higher in
apolar solvents[13–18] such as carbon tetrachloride or n-
hexane, but we preferred to avoid the use of toxic chlorinat-
ed solvents. Sometimes, this environmental choice had a
drawback because the reaction was slower in n-hexane. As a
consequence, a certain amount of the starting sulfide re-
mained in the mixture after two days and the yield de-
creased in comparison with the same reaction performed in
carbon tetrachloride.
The reactions were performed on a 1 mmol scale with
1.1 mmol of oxidant at room temperature for two days in
the presence of 0.05 mmol of a 1:2 titanium/hydrobenzoin
catalyst. Many of the chiral aryl benzyl sulfoxides that were
produced precipitated from the reaction medium but, in-
stead of collecting them as in our previous work,[14] we pre-
ferred to subject the whole reaction mixture to chromatog-
raphy in order to measure the ee values of the sulfoxides
before crystallisation (81 to >98%). In fact, as reported in
Table 1, the enantiomeric purity of these sulfoxides were
usually higher after crystallisation (ee >98%). In principle,
this protocol could be subjected to experimental improve-
ments, but we were interested in performing all the reac-
tions under the same conditions.
figuration at the sulfur atom was established by X-ray analy-
sis. When (S,S)-hydrobenzoin was used as the titanium
ligand, the sulfur atom of these sulfoxides had the R config-
uration, thus confirming our empirical rule.[14–16] Note that
the four molecules that were not found suitable for X-ray
analysis were compound 3b, which is an oil, and compounds
7b, 12b and 15b, in which a para-methoxy group is present
in one or both of the aryl groups of the molecule.
The tridimensional crystal structure is the result of many
interactions, the most important being hydrogen-bonding
between the benzyl methylene group and the sulfinyl
oxygen atom. Furthermore, other contributions from the
heteroatoms (for example, halogen-bonding[34]) were pres-
ent. In the crystal structure, the aryl groups of the sulfoxides
2b, 5b, 9b, 11b and 13b are arranged in a gauche conforma-
tion (values of the dihedral angle (aryl carbon)-(methylene
carbon)-sulfur-(aryl’ carbon) are in the range of 52–658). On
the other hand, in compounds 4b, 6b, 8b, 10b and 14b, the
aryl groups are arranged in an antiperiplanar conformation
(dihedral angle values in the range 166–1798). More details
are presented in the Supporting Information.
The stereochemical picture that emerged from this oxida-
tion is extremely interesting. As reported in Table 1, very
high ee values (>95%) were obtained for nine out of the
fourteen sulfoxides (Table 1, entries 2–4, 7, 8, 10, 11, 13–15
and 17), and ee values in the range 81–91% were measured
for the remaining ones.
Chem. Eur. J. 2009, 15, 13417 – 13426
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13419

A. Bottoni, C. Cardellicchio et al.
In particular, as far as the substitution on the aryl group
bonded to the sulfur atom is concerned (Table 1, entries 1–
8), a slight but significant decrease in the ee values (84–
88%) was observed when a methoxy group was present in
the ortho or meta position (entries 5 and 6). Because a
higher ee value (95%) was obtained after the introduction
of a non-coordinating ortho-chloro substituent (Table 1,
entry 2), this effect could be due to the interference of a co-
ordinating group close to the metal centre. On the other
hand, the same moiety in a different position (Table 1, en-
tries 7 and 8) did not particularly affect the enantioselectivi-
ty (>98% ee).
As far as the substitution on the benzyl aryl group is con-
cerned (Table 1, entries 9–17), excellent enantioselectivities
were obtained (97 to >98% ee values) with two exceptions.
The first is with the ortho-nitro moiety on the benzyl aryl
group (Table 1, entry 9, ee 85%). The second is with the
meta-methoxy group (Table 1, entry 12). In this case, not
only the ee value decreased (81% ee, which is the lowest
value of the whole investigation), but the yield also dropped
to 37% with a very large amount of sulfide 11a remaining
unreacted. The decrease in both yield and enantioselectivity
could be ascribed to the presence of a coordinating group
because the oxidation of sulfide 13a (Table 1, entry 14), in
which the chlorine atom is present in the same meta-posi-
tion, yielded an usually high ee (97%) and a satisfactory
yield (66%).
Another exception was the formation of a certain amount
(14%) of the corresponding sulfone from the oxidation of
sulfide 14a (Table 1, entry 15). The reaction was repeated at
08C for 7 h to test whether kinetic resolution was present
(Table 1, entry 16). Under these conditions, a 53:47 mixture
of sulfide 14a/sulfoxide 14b was obtained with no trace of
the sulfone, the ee of 14b being 82%. Because sulfoxide
14b was produced in an enantiopure form (entry 15), only
in this particular case is a certain amount of kinetic resolu-
tion present.
From the results presented in Table 1 and in a previous
paper,[14] with some exceptions, aryl benzyl sulfides seem to
be appropriate substrates for enantioselective oxidation in
the presence of a titanium/hydrobenzoin catalyst because
high ee values were always measured. Slight decreases in
enantioselectivity could be ascribed to the local interference
of some groups placed in certain positions.
The proposal of a plausible reaction mechanism encoun-
tered serious problems,[35–37] the most significant being the
experimental evidence for a fast exchange of titanium li-
gands in solution[35] and the formation of supramolecular as-
semblies. It has been demonstrated that the complexes
formed between titanium and tartrates[36] and between tita-
nium and BINOL[14,37] yield intricate patterns when ob-
served by spectroscopic techniques. However, when we re-
corded the 1H NMR spectrum[14] of the complex obtained
from a 1:2 mixture of titanium isopropoxide and hydroben-
zoin, we observed a very simple pattern arising from a tita-
nium complex in which only the hydrobenzoin was bound to
the metal with the original isopropoxide moiety free in solu-
tion. The different behaviour characterising the titanium/hy-
drobenzoin complex with respect to the other catalysts led
us to hypothesise that a less complicated reaction mecha-
nism could be operating.
Following our hypothesis, the starting tetrahedral Ti(hy-
drobenzoin)2 species, which originates from the initial 1:2
mixture of titanium isopropoxide and hydrobenzoin, first
undergoes oxidative addition of the sulfide and then of the
oxidant. This would lead to the formation of an octahedral
assembly that is expected to be stable enough to transfer the
oxygen atom from the hydroperoxide moiety to the sulfur
atom, in accord with the experimental results, that is, lead to
the formation of the (R)- or (S)-sulfoxide when (S,S)- or
(R,R)-hydrobenzoin is used, respectively. Furthermore,
within our mechanistic hypothesis, aryl benzyl sulfides
should be the best fitting substrates.
Computational details and choice of model: Because most
of the experimentally investigated aryl substituents do not
significantly affect the enantioselectivity of the reaction, we
have used in our calculations the simplest (prototype)
model system, a benzyl phenyl sulfide molecule, which
reacts with tert-butyl hydroperoxide (TBHP) in the presence
of a titanium/hydrobenzoin complex. In particular, we chose
(S,S)-hydrobenzoin and considered the two possible reaction
pathways leading to the (S)- or (R)-sulfoxide, this last con-
figuration being in agreement with the experimental results.
Owing to the presence of aryl groups in the ligands and in
the substrate, it is reasonable to believe that interactions in-
volving the aryl p systems should be taken into account, as
was assumed by other workers.[13] Thus, an appropriate
method for their description must be chosen. It is well
known that this class of interactions cannot be correctly
treated at the DFT level of theory because the most popular
functionals (for instance, B3LYP) are inaccurate for interac-
tions in which medium-range correlation effects are domi-
nant.[38]
Mechanistic investigations: With these results in hand, we
tried to envisage a reaction mechanism that could explain
the observed reaction selectivity. As outlined in the Intro-
duction section, the mechanism for the metal-catalysed
enantioselective oxidation of sulfides is still far from being
understood. A mechanistic scheme was suggested for vana-
dium catalysts,[21] but nothing similar has been reported for
active titanium species, even though titanium was the first
metal that was proposed for these reactions[9] and a huge
amount of literature relating to its use has been published in
more than 25 years.[6,7]
In fact, p–p interactions are satisfactorily described by the
MP2 method but, given the size of the model system used
here (47 heavy atoms and 36 hydrogen atoms), MP2 compu-
tations require too much CPU time for practical and exten-
sive use. However, a new hybrid functional (denoted as
MPWB1K), which is capable of treating medium-range cor-
relation effects, has recently been proposed.[39] This func-
13420
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 13417 – 13426

Enantioselective Oxidation of Aryl Benzyl Sulfides
FULL PAPER
tional has been demonstrated to provide a good estimate of
p–p interactions and reaction energies[40,41] using reasonable
amounts of CPU time.
and M1’ without overcoming any barrier. Because M1 and
M1’ afford final products with the S and R configuration at
the sulfur atom, respectively, they can be denoted as pro-S
and pro-R compounds (see Figure 2). The energy difference
between these two species is 1.7 kcalmolꢀ1 and the one with
lower energy is the pro-S species M1, which is 10.5 kcal
molꢀ1 below the AL.
Thus, all the DFT computations reported in this paper
were carried out with the Gaussian 03 software package[42]
using the MPWB1K[39] functional and the DZVP basis
set.[43] The DZVP basis set is a local spin density (LSD) op-
timised basis set of double-zeta quality that includes polari-
sation functions and is suitable for describing weak hydro-
gen bonds and p interactions such as those occurring in the
system investigated herein. The transition vector of the vari-
ous transition states has been analysed by frequency compu-
tations. Furthermore, because an apolar solvent (n-hexane)
was used in our experiments, on the basis of our experience,
we believe that gas-phase computations can satisfactorily
emulate the experimental conditions.
Computational results: The corresponding energy profile for
the oxidation of the benzyl phenyl sulfide molecule by tert-
butyl hydroperoxide (TBHP) in the presence of a titanium-
(hydrobenzoin)2 complex is reported in Figure 1 and sche-
Figure 2. Schematic representation of the structures of the two diastereo-
meric titanium pentacoordinate adducts M1’ and M1 (bond lengths in
ꢂ). Energy values (kcalmolꢀ1) are relative to the AL.
Inspection of Figure 2 shows that as the sulfide enters into
the metal coordination sphere the original Ti(hydroben-
zoin)2 tetrahedral structure becomes distorted and the two
five-membered rings (CCO1TiO2 and CCO1’TiO2’) are now
bent in the opposite direction with respect to the approach-
ꢀ
ing line of the sulfide. The Ti O bond lengths are quite sim-
ꢀ
ilar in the two intermediates and the newly formed Ti S
bond is shorter in M1 (2.61 ꢂ) than in M1’ (2.68 ꢂ).
The two pentacoordinated titanium complexes are charac-
terised by a number of interactions involving the phenyl
rings of the substrate and those of the two hydrobenzoin li-
gands. However, the nature of these interactions is rather
different in the two cases. In M1’ p-stacking interactions are
evident between each phenyl ring of the sulfide (B3 and B4)
and one of the two phenyl rings of each hydrobenzoin
moiety (B1 and B1’, respectively). The two rings of each
pair lie in approximately parallel planes, but do not form
the well-known “sandwich” structure. The two planes are
displaced to give a structural arrangement that has been
demonstrated to provide more effective p–p stabilisa-
tion.[44–46] Typical distances between the most effectively in-
teracting carbon atoms of the two rings are in the range 3.6–
3.9 ꢂ and are reported in Figure 2.
Figure 1. Computed energy profile for the enantioselective TBHP oxida-
tion of benzyl phenyl sulfide in the presence of a titanium/
benzoin catalyst.
ACHTUNGTREN(NUNG S,S)-hydro-
matic representations of the structures of the most impor-
tant critical points are reported in Figures 2–4 and Fig-
ACHTUNGTRENNUNG
In M1 the orientation of the substrate is reversed com-
pared to M1’, and this leads in the final product to the oppo-
site configuration at the sulfur atom (the S configuration).
The result of the different substrate arrangement is a shorter
G
a
benzyl phenyl sulfide substrate molecule and a TBHP mole-
cule at infinite distance (that is, the reactants).
ꢀ
The substrate molecule can adopt two different orienta-
tions to interact with the tetrahedral complex involving the
(S,S)-hydrobenzoin enantiomer. The two approaches lead to
the formation of two pentacoordinated titanium species M1
Ti S distance and the appearance in M1 of a T-shaped con-
figuration[45–47] involving the ring-pair B1’/B4. In the T-
shaped structure the plane of the hydrobenzoin phenyl ring
B1’ is approximately orthogonal (but tilted) with respect to
Chem. Eur. J. 2009, 15, 13417 – 13426
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13421

A. Bottoni, C. Cardellicchio et al.
the plane of the substrate ring B4. This particular arrange-
ment (tilted T-shaped) is known to be more stabilising than
“parallel-displaced” and “sandwich” structures (p stacking)
because, in addition to dispersion forces and quadrupole–
quadrupole electrostatic effects, further stabilisation occurs
due to hydrogen-bonding contributions.[47] In this case, one
ꢀ
C H bond of B1’ interacts with the p-electron cloud of B4:
ꢀ
the (phenyl-B1’)C H···C(phenyl-B4) distances that feature
the tilted T-shaped interaction are in the range of 2.9–3.6 ꢂ
(specific values are reported in Figure 2).
It is interesting to examine the relative arrangements of
B1 and B3 in M1. The methylene group placed between the
sulfur and benzene makes it difficult in this case to achieve
a T-shaped structure such as that observed for the B1’/B4
Figure 3. Schematic representation of the structures of the two diastereo-
meric titanium octahedral adducts M2’ and M2 (bond lengths in ꢂ).
Energy values (kcalmolꢀ1) are relative to the AL.
ꢀ
pair. However, one of the C H bonds of the methylene
group points towards B1 and gives rise to a non-negligible
hydrogen interaction. This is evidenced by the (methyle-
ꢀ
ne)C H···C(phenyl-B1) distances: the shortest computed
value is 2.53 ꢂ, which indicates a non-negligible interaction
H···C(phenyl-B1) (similar to that described for M1) can be
ꢀ
ꢀ
of the C H bond with the p-electron cloud of B1. Even if
recognised (a (methylene)C H···C(phenyl-B1) distance of
2.64 ꢂ reflects the relative positions of methylene and B1).
Furthermore, there is a significant hydrogen bond involving
ꢀ
this C H···p interaction is less stabilising than a full benzene
dimer interaction (1.5 versus 2.5 kcalmolꢀ1),[47] the two rings
remain close enough to believe that some stabilising effects
associated with p–p aromatic interactions remain active (for
instance, dispersion forces and quadrupole–quadrupole ef-
ꢀ
ꢀ
the O4 H group and the O1’ oxygen, the O4 H···O1’ dis-
tance being 1.84 ꢂ.
Because in M2’ B3 and B1’ are tilted compared with their
positions in M1’, p-stacking interactions become impossible.
However, the relative arrangement of the B3 and B1’ planes
(these are oblique and displaced one to the other) leads to a
sort of weakly bound T-shaped structure characterised by
the interaction of one of the C H bonds of B1’ with the p-
electron cloud of B3: the shortest (phenyl-B1’)C
H···C(phenyl-B3) distance is 3.22 ꢂ.
ꢀ
fects) and must contribute to the C H···p stabilisation. Fur-
thermore, a comparison between the two structures of
Figure 2 shows that in the two species the relative positions
and distances between B1’, B2’ and B2 do not change signif-
icantly and thus the interactions involving these three rings
should not affect the relative stabilities of M1 and M1’.
In conclusion, the shorter Ti S distance observed in M1
and the stabilisation associated with the T-shaped structure
and C H···p interactions explain the lower energy of M1
compared with M1’.
The subsequent step involves the reaction of tert-butyl hy-
droperoxide with the pentacoordinated complexes. In princi-
ple, we could consider a deprotonated hydroperoxide re-
agent. However, in this case the charged hydroperoxide that
results from the deprotonation forms a h2 complex with the
titanium atom and determines the expulsion of the sulfide
from the metal coordination sphere. The system obtained in
this way is stable and cannot lead to any oxidation product.
Thus, it seems reasonable to consider the addition of neutral
TBHP. We found that this leads to the formation of two new
species (M2 from M1 and M2’ from M1’, see Figure 3) in
which the metal adopts an octahedral arrangement. The dif-
ference in energy between M2 and M2’ is 5.1 kcalmolꢀ1 and
the favoured structure is now M2’, that is, the complex lead-
ing to the experimentally observed R configuration of the
final product.
ꢀ
ꢀ
ꢀ
The B1–B4 p-stacking interactions found in M1’ also dis-
appear in M2’. However, in this case, B1 is approximately
orthogonal to B4 and originates a stabilising tilted T-shaped
ꢀ
configuration.[45–47] The interaction of one of the C H bonds
ꢀ
of B1 with the p-electron cloud of B4 is evident from
ꢀ
Figure 3
(left): the shortest (phenyl-B1)C H···C(phenyl-B4)
distances that reveal this tilted T-shaped interaction are 2.84
and 2.95 ꢂ (as shown in Figure 3left). A further interesting
aspect found in M2’ is the O4 H···O1’ hydrogen bond,
which is stronger than in M2, the O4 H···O1’ distance now
ꢀ
ꢀ
being 1.77 ꢂ.
ꢀ
ꢀ
Because the Ti S and Ti O3 distances are almost identi-
cal in M2 and M2’, it is reasonable to believe that the B1–
B4 T-shaped structure, the additional (even weak) interac-
ꢀ
tions involving B1’ and B3 and the O4 H···O1’ hydrogen
bond are the key factors that determine the lower energy of
M2’ compared with M2.
The computed energy difference between M2 and M2’ in-
dicates that under thermodynamic conditions such as those
governing the first part of the reaction, the latter species
should form almost exclusively. The thermodynamic equili-
bration requires a barrier of 30.2 kcalmolꢀ1 to be overcome
(needed by M2 to break the Ti O and Ti S bonds and
allow the substrate to reorientate to afford M2’). This
amount of energy is almost identical to the activation barri-
It is evident from Figure 3 that the nature of the interac-
tions significantly changes on passing from M1 to M2. Al-
though in M1 T-shaped interactions characterise the B1’/B4
pair, in M2 these two rings are too far away for an effective
stabilisation. More interesting is the arrangement of the two
ꢀ
ꢀ
ꢀ
rings B1 and B3. A stabilising interaction (methylene)C
13422
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 13417 – 13426

Enantioselective Oxidation of Aryl Benzyl Sulfides
FULL PAPER
er needed to achieve the subsequent transition state and is,
thus, available under the experimental conditions used for
the reaction.
The two transition states TS and TS’ (see Figure 4) lead-
ing to the final sulfoxide products from M2 and M2’, respec-
tively, were located. These are almost degenerate, their
energy difference being 0.5 kcalmolꢀ1, with TS’ more stable
than TS.
structure. All these structural features detected in TS’ and
TS explain their almost negligible energy difference.
The two transition states lead to the final product com-
plexes M3 (from TS along the pro-S pathway) and M3’
(from TS’ along the pro-R pathway), which are the precur-
sors of the enantiomer benzyl phenyl sulfoxides. A schemat-
ic representation of the two complexes is given in Figure 2S
of the Supporting Information. As found for TS’ and TS,
M3’ is more stable than M3, the energy difference being
1.1 kcalmolꢀ1. Again, a complex interplay of many factors,
ꢀ
such as p–p aromatic and C H···p interactions determine
the relative energies of the two product adducts. A shorter
ꢀ
Ti O4 distance in M3’ (2.20 ꢂ) compared with M3 (2.25 ꢂ)
certainly contributes to the stabilisation of the former spe-
cies.
However, this aspect is not particularly important because
the final product complexes are very low in energy com-
pared with the previously discussed critical points: M3’ is
42.5 kcalmolꢀ1 lower than M2’ and 77.8 kcalmolꢀ1 below the
AL. Thus, the relative populations of M3 and M3’ (the two
products with opposite configuration at the sulfur) cannot
be determined by their relative energies. Under the experi-
mental conditions used for this reaction, it is very unlikely
that thermodynamic equilibration occurs because the con-
version of M3 to M3’ would require a barrier of 76.5 kcal
molꢀ1 to be overcome (this amount of energy is required to
achieve the transition state TS).
It is reasonable to believe that the observed ee (which re-
quires that M3’ is much more populated than M3) is deter-
mined by the energetics of the first phase of the reaction
and the corresponding kinetics. Because, as previously dis-
cussed, M2’ is formed almost exclusively during the first re-
action phase and the two transition states are almost degen-
erate, the most likely pathway to complete the process is
that involving TS’ leading to M3’ (the adduct with the R
configuration at the sulfur atom). The required activation
barrier is 34.6 kcalmolꢀ1, a value that is very similar to the
amount of energy necessary for the reverse reaction to yield
the reactants. Finally, the decomposition of M3’ will produce
the desired sulfoxide.
Figure 4. Schematic representation of the structures of the two transition
states TS’ and TS (bond lengths in ꢂ). Energy values (kcalmolꢀ1) are rel-
ative to the AL.
In TS and TS’ a new bond is forming between the sulfur
atom and the oxygen atom O4 of the hydroperoxide mole-
ꢀ
cule (the incipient S O4 bond is 2.20 and 2.19 ꢂ in TS’ and
TS, respectively). Simultaneously, the hydrogen atom mi-
ꢀ
grates from O4 to O1’ causing a lengthening of the Ti O1’
ꢀ
ꢀ
bond (the (O4)H O1’ and Ti O1’ distances are 1.81 and
1.95 ꢂ in TS’ and 1.74 and 1.94 ꢂ in TS, respectively, where-
ꢀ
as the corresponding value for Ti O2’ is 1.81 ꢂ in both
cases). In both TS’ and TS B1’ interacts with the adjacent
phenyl ring of the substrate (B3 in TS’ and B4 in TS) in a
parallel-displaced arrangement (this corresponds to a weak
p-stacking interaction in which the C···C bond lengths of the
two rings are larger than 3.4 ꢂ). Also, B1’ interacts with B2’
to form a T-shaped structure that is almost identical in the
two transition states. A further interesting feature is the
H(O4)···S hydrogen bond which assists the proton transfer.
The strength of this interaction is almost identical in TS’ and
TS, the H(O4)···S distances being 2.37 and 2.38 ꢂ, respec-
tively. A hydrogen bond between the methylene group of
the sulfide and the O1’ atom can be identified in TS’
((CH2)H···O1’distance=2.64 ꢂ). This interaction is replaced
in TS by an equivalent hydrogen bond involving O4
((CH2)H···O4 distance=2.65 ꢂ). In fact, the two structures
TS’ and TS are very similar, except for the arrangement of
the two rings B1 and B4 in TS’ and B1and B3 in TS. B1 and
B3 are approximately planar (displaced p-stacking), whereas
the relative positions of B1 and B4 resemble a T-shaped
Conclusions
In the work presented in this paper, it has been shown ex-
perimentally that high ee values and satisfactory yields are
obtained in the enantioselective preparation of synthetically
important aryl benzyl sulfoxides by enantioselective oxida-
tion by TBHP in the presence of a titanium/hydrobenzoin
catalyst. The origin of this high enantioselectivity has been
explained by a mechanistic scheme that has been studied
theoretically at the DFT level of theory using the recently
proposed MPWB1K functional. This functional can provide
a reliable description of medium-range correlation effects
that play a key role in determining the relative stabilities of
the various critical points on the potential surface. A simple
(prototype) model system, formed by a benzyl phenyl sul-
Chem. Eur. J. 2009, 15, 13417 – 13426
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13423

A. Bottoni, C. Cardellicchio et al.
4-Bromophenyl 4-nitrobenzyl sulfide (9a): M.p. 102–1048C (ethanol)
(lit.:[28] m.p. 104–1058C).
fide molecule reacting with tert-butyl hydroperoxide
(TBHP) in the presence of a titanium/(S,S)-hydrobenzoin
AHCTUNGTRENNUNG
4-Bromophenyl 2-methoxybenzyl sulfide (10a): Kugelrohr oven temp.
136–1388C; p=0.01 mbar; 1H NMR (500 MHz, CDCl3, 258C, TMS): d=
complex, has been used. Two reaction pathways (pro-R and
pro-S) corresponding to the two possible approaching orien-
tations of sulfide to the Ti(hydrobenzoin)2 complex and
leading to the (R)- and (S)-sulfoxides, respectively, have
been considered.
7.38–7.34 (m, 2H; HAr), 7.24 (ddd, 4J(H, H)=1.6, 3J(H, H)=7.5, 3J
ACHTUNGTRENNNUG ACHTUNREGTNNUNG ACHTUNGTRENNUNG(H,
H)=8.2 Hz, 1H; HAr), 7.19–7.16 (m, 3H; HAr), 6.89–6.85 (m, 2H; HAr),
4.12 (s, 2H; CH2), 3.82 ppm (s, 3H; OCH3); 13C NMR (125 MHz, CDCl3,
258C, TMS): d=136.0 (CAr), 131.7 (CAr), 131.5 (CAr), 130.1 (CAr), 128.7
(CAr), 125.3 (CAr), 120.4 (CAr), 120.0 (CAr), 110.5 (CAr), 55.4 (OCH3),
33.4 ppm (CH2); elemental analysis calcd (%) for C14H13BrOS: C 54.38,
H 4.24; found: C 54.40, H 4.15.
The two reaction channels (pro-S: AL!M1!M2!TS;
pro-R: AL!M1’!M2’!TS’) are characterised by a first
phase that affords the two preliminary intermediates M2
and M2’ (octahedral adducts) with M2’ more stable than M2
by 5.1 kcalmolꢀ1. Our computations demonstrated that T-
shaped structures, which are present only in M2’, are the
key factors determining the greater stability of this inter-
mediate. Because the reverse reaction from M2 to the reac-
tants requires a barrier of 30.2 kcalmolꢀ1 to be overcome
(an amount of energy almost identical to that necessary to
reach the subsequent transition state) it seems reasonable to
believe that thermodynamic equilibration occurs and M2’
forms almost exclusively. From M2’ the system “naturally”
evolves to the final product adduct M3’ (second phase),
which can release the experimentally observed sulfoxide
with the R configuration at the sulfur atom. Thus, the ther-
modynamics of the first phase of the process (and the conse-
quent dominant population of M2’) is the key factor that de-
termines the high enantioselectivity.
4-Bromophenyl 3-methoxybenzyl sulfide (11a): Kugelrohr oven temp.
138–1408C; p=0.01 mbar; 1H NMR (500 MHz, CDCl3, 258C, TMS): d=
7.38–7.35 (m, 2H; HAr), 7.20 (t, 3J
ACHTNUGTRENUNG(H, H)=7.8 Hz, 1H; HAr), 7.17–7.14
(m, 2H; HAr), 6.87–6.85 (m, 1H; HAr), 6.83–6.82 (m, 1H; HAr), 6.81–6.77
(m, 1H; HAr), 4.06 (s, 2H; CH2), 3.77 ppm (s, 3H; OCH3); 13C NMR
(125 MHz, CDCl3, 258C, TMS): d=159.7 (CAr), 138.5 (CAr), 135.4 (CAr),
131.8 (CAr), 131.4 (CAr), 129.5 (CAr), 121.1 (CAr), 114.2 (CAr), 112.9 (CAr),
55.2 (OCH3), 39.0 ppm (CH2); elemental analysis calcd (%) for
C14H13BrOS: C 54.38, H 4.24; found: C 54.05, H 4.30.
4-Bromophenyl 4-methoxybenzyl sulfide (12a): M.p. 99–1018C (ethanol)
(lit.:[29] m.p. 100.5–1018C)
4-Bromophenyl 3-chlorobenzyl sulfide (13a): Kugelrohr oven temp. 160–
1628C; p=0.01 mbar; 1H NMR (500 MHz, CDCl3, 258C, TMS): d=7.39–
7.36 (m, 2H; HAr), 7.28–7.26 (m, 1H; HAr), 7.22–7.19 (m, 2H; HAr), 7.16–
7.11 (m, 3H; HAr), 4.03 ppm (s, 2H; CH2); 13C NMR (125 MHz, CDCl3,
258C, TMS): d=139.2 (CAr), 134.7 (CAr), 134.3 (CAr), 132.0 (CAr), 131.8
(CAr), 129.7 (CAr), 128.8 (CAr), 127.5 (CAr), 126.9 (CAr), 120.8 (CAr),
38.7 ppm (CH2); elemental analysis calcd (%) for C13H10BrClS: C 49.78,
H 3.21; found: C 49.70, H 3.55.
4-Bromophenyl 2,4-dichlorobenzyl sulfide (14a): Kugelrohr oven temp.
162–1658C, p=0.01 mbar; 1H NMR (500 MHz, CDCl3, 258C, TMS): d=
7.39–7.36 (m, 3H; HAr), 7.18–7.10 (m, 4H; HAr), 4.12 ppm (s, 2H; CH2);
13C NMR (125 MHz, CDCl3, 258C, TMS): d=134.7 (CAr), 134.2 (CAr),
133.8 (CAr), 133.6 (CAr), 132.6 (CAr), 132.0 (CAr), 131.3 (CAr), 129.5 (CAr),
127.1 (CAr), 121.2 (CAr), 36.6 ppm (CH2); elemental analysis calcd (%)
for C13H9BrCl2S: C 44.86, H 2.61; found: C 44.89, H 2.93.
Experimental Section
Chemicals were purchased from Sigma–Aldrich or Alfa-Aesar and were
used as received. n-Hexane employed in the enantioselective oxidation
protocol was distilled from 4 ꢂ molecular sieves prior to use. NMR spec-
tra were recorded on a Bruker AM500 spectrometer. HPLC were per-
formed on an Agilent 1100 chromatograph, equipped with a DAD detec-
tor. Elemental analyses were performed on a Carlo Erba CHNS-O
EA1108 Elemental Analyzer. Sulfides 2a–15a were synthesised by a
standard reaction between the corresponding sodium arylthiolates and
benzyl bromides.
4-Methoxyphenyl 4-methoxybenzyl sulfide (15a): M.p. 82–848C (ethanol)
(lit.:[30] 87–888C).
Racemic sulfoxides 2b–15b (used in the setting up of the chiral HPLC
separation) were synthesised by standard m-chloroperbenzoic acid
(MCPBA) oxidation. Enantioenriched sulfoxides 2b–15b were produced
according to our protocol[14] by TBHP oxidation in n-hexane in the pres-
ence of 5mol-% of the titanium/hydrobenzoin catalyst.
Methyl (2-benzylthio)benzoate (2a): M.p. 65–668C (ethanol) (lit.:[24] m.p.
X-Ray data were collected at 293 K by using a single-crystal X-ray dif-
fractometer. Data were corrected for Lorentzian and polarisation effects
as well as for absorption effects (8b–11b, 13b and 14b).[48] The structures
were solved by direct methods using SIR97[49] and refined by a full-
matrix least-squares technique based on F2 (SHELXL-97).[50] In all the
refinements, the non-hydrogen atoms were refined with anisotropic ther-
mal parameters. The hydrogen atoms were located through difference
Fourier maps and refined isotropically with the exception of 8b, 9b, 13b
and 14b, for which the hydrogen atoms were placed in idealised positions
56–588C).
Benzyl 2-chlorophenyl sulfide (3a): Kugelrohr oven temp. 130–1328C;
p=0.01 mbar (lit.:[25] b.p. 159–1608C, p=3 torr).
Benzyl 4-nitrophenyl sulfide (4a): M.p. 118–1208C (ethanol) (lit.:[26] m.p.
120–1228C).
(R)-Benzyl 2-methoxyphenyl sulfide (5a):[27] Kugelrohr oven temp. 140–
1428C; p=0.01 mbar.
Benzyl 3-methoxyphenyl sulfide (6a):[27] Kugelrohr oven temp. 118–
1218C; p=0.01 mbar.
Benzyl 4-methoxyphenyl sulfide (7a):[27] Kugelrohr oven temp. 123–
1258C; p=0.01 mbar.
ꢀ
ꢀ
riding on their attached atoms (C HAr 0.93 ꢂ, C HAlk 0.97 ꢂ, Uiso(H)=
1.2Uiso(C)). Geometrical calculations were performed by using PARST.[51]
The unit cell parameters, data collection and refinement are reported in
the Supporting Information. CCDC-720081 (2b), ꢀ720082 (4b), ꢀ720083
(5b), ꢀ720084 (6b), ꢀ720085 (8b), ꢀ720086 (9b), ꢀ720087 (10b),
ꢀ720088 (11b), ꢀ720089 (13b) and ꢀ720090 (14b) contain the supple-
mentary crystallographic data for this paper. These data can be obtained
free of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
4-Bromophenyl 2-nitrobenzyl sulfide (8a): M.p. 50–518C (ethanol);
ACTHNUTRGNEUNG
1H NMR (500 MHz, CDCl3, 258C, TMS): d=7.99 (dd, 4J(H, H)=1.5, 3J-
N
R
ACHTUNGTRENNUNG
A
U
ACHTUNGTRENNUNG
E
ACHTUNGTRENNUNG
7.6 Hz, 1H; HAr), 7.16–7.12 (m, 2H; HAr), 4.41 ppm (s, 2H; CH2).
13C NMR (125 MHz, CDCl3, 258C, TMS): d=133.8 (CAr), 133.3 (CAr),
133.1 (CAr), 132.0 (CAr), 131.8 (CAr), 128.4 (CAr), 125.4 (CAr), 121.6 (CAr),
37.3 ppm (CH2); elemental analysis calcd (%) for C13H10BrNO2S: C
48.16, H 3.11, N 4.32; found: C 48.43, H 3.48, N 4.40.
Methyl (R)-(2-benzylsulfinyl)benzoate (2b): M.p. 151–1528C (n-hexane/
ethanol 9:1) (lit.:[31] racemic m.p. 118–1208C); [a]2D5 =+433.7 (c=0.8 in
CHCl3). The ee value was measured by HPLC (column: Chiralcel OD-H;
eluent: n-hexane/isopropanol 70:30).
13424
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 13417 – 13426

Enantioselective Oxidation of Aryl Benzyl Sulfides
FULL PAPER
(R)-Benzyl 2-chlorophenyl sulfoxide (3b):[32] [a]D25 =+397.1 (c=1 in
CHCl3). The ee value was measured by HPLC (column: Chiralcel OB-H;
eluent: n-hexane/isopropanol 90:10).
2H; HAr), 7.18 (dd, 3J
G
ACHTUNGTRENNUNG
(ddd, 4J
A
G
ACHTUNGTRENNUNG
4
4
3
(ddd, J
G
(H, H)=0.9, J
A
N
4J
(H, H)=1.4, 4J
U
ACHTUNGTRENNUNG
(R)-Benzyl 4-nitrophenyl sulfoxide (4b): M.p. 196–1988C (ethanol/ethyl
acetate 15:3); [a]D25 =+114.6 (c=0.9 in CHCl3).
AHCTUNGTRENNUNG
OCH3); 13C NMR (125 MHz, CDCl3, 258C, TMS): d=159.6 (CAr), 132.0
(CAr), 130.0 (CAr), 129.6 (CAr), 126.1 (CAr), 125.7 (CAr), 122.6 (CAr), 115.4
(CAr), 114.4 (CAr), 63.6 (CH2), 55.2 ppm (OCH3); elemental analysis calcd
(%) for C14H13BrO2S: C 51.70, H 4.03; found C 51.56, H 4.21. The ee
value was measured by HPLC (column: Chiralcel OD-H; eluent:
n-hexane/isopropanol 90:10).
(S)-Benzyl 4-nitrophenyl sulfoxide (4b): M.p. 198–1998C (ethanol/ethyl
acetate 15:3) (lit.:[26] m.p. 179–181); [a]D25 =ꢀ115.5 (c=1.1 in CHCl3) and
[a]D25 =ꢀ259.9 (c=0.5 in acetone) (lit.:[26] [a]2D5 =ꢀ163.6 (c=0.93 in ace-
tone)). The ee value was measured by HPLC (column: Chiralcel OD-H;
eluent: n-hexane/isopropanol 70:30).
(R)-Benzyl 2-methoxyphenyl sulfoxide (5b): M.p. 60–618C (n-hexane/iso-
propanol 9:1), [a]2D5 =+510.3 (c=1 in CHCl3) (lit.:[33] +351 (c=0.32 in
CHCl3) for a sulfoxide with 81% ee). The ee value was measured by
HPLC (column: Chiralcel OD-H; eluent: n-hexane/isopropanol 95:5).
(R)-4-Bromophenyl 4-methoxybenzyl sulfoxide (12b): M.p. 173–1758C
(n-hexane/ethanol 8:2); [a]2D5 =+70.1 (c=0.9 in CHCl3); 1H NMR
(500 MHz, CDCl3, 258C, TMS): d=7.58–7.54 (m, 2H; HAr), 7.23–7.19 (m,
2
2H; HAr), 6.90–6.86 (m, 2H; HAr), 6.81–6.77 (m, 2H; HAr), 4.04 (d, J
N
2
(R)-Benzyl 3-methoxyphenyl sulfoxide (6b): M.p. 86–878C (n-hexane/
acetone 19:1); [a]2D5 =+78.6 (c=0.9 in CHCl3) and [a]2D5 =+198.5 (c=0.5
in acetone) (lit.:[33] +73.5 (c=0.17 in acetone)) for a sulfoxide with 69%
ee). The ee value was measured by HPLC (column: Chiralcel OD-H;
eluent: n-hexane/isopropanol 90:10).
H)=12.7 Hz, 1H; CH2), 3.94 (d, JACHTUNTRGNENUG(H, H)=12.7 Hz, 1H; CH2), 3.79 ppm
(s, 3H; OCH3); 13C NMR (125 MHz, CDCl3, 258C, TMS): d=159.7
(CAr), 141.9 (CAr), 132.0 (CAr), 131.6 (CAr), 126.1 (CAr), 125.5 (CAr), 120.4
(CAr), 114.0 (CAr), 62.7 (CH2), 55.3 ppm (OCH3); elemental analysis calcd
(%) for C14H13BrO2S: C 51.70, H 4.03; found C 51.50, H 3.99. The ee
value was measured by HPLC (column: Chiralcel OD-H; eluent:
n-hexane/isopropanol 90:10).
(R)-Benzyl 4-methoxyphenyl sulfoxide (7b): M.p. 136–1378C (n-hexane/
ethanol 15:1); [a]2D5 =+57.8 (c=1.2 in CHCl3) (lit.:[33] +31.9 (c=0.28 in
CHCl3) for a sulfoxide with 44% ee).
(R)-4-Bromophenyl 3-chlorobenzyl sulfoxide (13b): M.p. 116–1178C
(n-hexane/acetone 8:2); [a]2D5 =+103.6 (c=0.8 in CHCl3); 1H NMR
(500 MHz, CDCl3, 258C, TMS): d=7.61–7.56 (m, 2H; HAr), 7.31–7.18 (m,
4H; HAr), 6.97–6.94 (m, 1H; HAr), 6.89–6.85 (m, 1H; HAr), 3.98 ppm (s,
2H; CH2); 13C NMR (125 MHz, CDCl3, 258C, TMS): d=141.5 (CAr),
134.4 (CAr), 132.2 (CAr), 130.6 (CAr), 130.3 (CAr), 129.7 (CAr), 128.6 (CAr),
128.5 (CAr), 125.9 (CAr), 62.7 ppm (CH2); elemental analysis calcd (%)
for C13H10BrClOS: C 47.37, H 3.06; found: C 47.66, H 3.22. The ee value
was measured by HPLC (column: Whelk O1; eluent: n-hexane/isopropa-
nol 70:30).
(S)-Benzyl 4-methoxyphenyl sulfoxide (7b): M.p. 135–1368C (n-hexane/
ethanol 15:1); [a]2D5 =ꢀ59.7 (c=0.6 in CHCl3). The ee value was mea-
sured by HPLC (column: Chiralcel OD-H; eluent: n-hexane/isopropanol
90:10).
(R)-4-Bromophenyl 2-nitrobenzyl sulfoxide (8b): M.p. 97–988C
(n-hexane/ethanol 9:1); [a]2D5 =+203.4 (c=0.9 in CHCl3); 1H NMR
(500 MHz, CDCl3, 258C, TMS): d=8.04 (dd, 4J
G
ACHTUNGTRENNUNG
8.1 Hz, 1H; HAr), 7.59–7.55 (m, 2H; HAr), 7.54 (dt, 4J
R
ACHTUNGTRENNUNG
H)=7.5 Hz, 1H; HAr), 7.47 (ddd, 4J(H, H)=1.5, 3J
G
A
ACHTUNGTRENNUNG
(R)-4-Bromophenyl 2,4-dichlorobenzyl sulfoxide (14b): M.p. 117–1188C
AHCTUNGTRENNUNG
(n-hexane/acetone 8:2); [a]2D5 =+76.2 (c=0.9 in CHCl3); 1H NMR
G
ACHTUNGTRENNUNG
4
(500 MHz, CDCl3, 258C, TMS): d=7.63–7.57 (m, 2H; HAr), 7.39 (d, J
N
ACHTUNGTRENNUNG
H)=2.0 Hz, 1H; HAr), 7.35–7.30 (m, 2H; HAr), 7.20 (dd, 4J
(H, H)=2.0,
258C, TMS): d=142.1 (CAr), 134.1 (CAr), 133.6 (CAr), 132.4 (CAr), 129.7
(CAr), 126.0 (CAr), 125.6 (CAr), 125.5 (CAr), 125.4 (CAr), 61.8 ppm (CH2);
elemental analysis calcd (%) for C13H10BrNO3S: C 45.90, H 2.96, N 4.12;
found: C 46.18, H 2.83, N 4.21. The ee value was measured by HPLC
(column: Whelk O1; eluent: n-hexane/isopropanol 70:30).
3
3J
2J
(H, H)=8.2 Hz, 1H; HAr), 7.07 (d, J
N
CH2); 13C NMR (125 MHz, CDCl3, 258C, TMS): d=135.5 (CAr), 135.4
(CAr), 133.4 (CAr), 132.3 (CAr), 129.5 (CAr), 127.3 (CAr), 126.1 (CAr), 126.0
(CAr), 125.9 (CAr), 60.6 ppm (CH2); elemental analysis calcd (%) for
C13H9BrCl2OS: C 42.89, H 2.49; found C 42.90, H 2.60. The ee value was
measured by HPLC (column: Chiralcel OD-H; eluent: n-hexane/isopro-
panol 70:30).
(R)-4-Bromophenyl 4-nitrobenzyl sulfoxide (9b): M.p. 153–1548C
(n-hexane/ethanol 7:3); [a]2D5 =+279.3 (c=0.9 in CHCl3); 1H NMR
(500 MHz, CDCl3, 258C, TMS): d=8.14–8.11 (m, 2H; HAr), 7.61–7.58 (m,
2
2H; HAr), 7.23–7.20 (m, 2H; HAr), 7.15–7.11 (m, 2H; HAr), 4.20 (d, J
(H,
H)=12.8 Hz, 1H; CH2), 3.99 ppm (d, 2J
ACTHNUTRGNE(UNG H, H)=12.8 Hz, 1H; CH2);
(R)-4-Methoxyphenyl 4-methoxybenzyl sulfoxide (15b): M.p. 159–1608C
(n-hexane/ethanol 9:1); [a]2D5 =+31.3 (c=0.4 in CHCl3); 1H NMR
13C NMR (125 MHz, CDCl3, 258C, TMS): d=147.8 (CAr), 140.9 (CAr),
135.6 (CAr), 132.4 (CAr), 131.3 (CAr), 126.1 (CAr), 125.7 (CAr), 123.4 (CAr),
61.7 ppm (CH2); elemental analysis calcd (%) for C13H10BrNO3S: C
45.90, H 2.96, N 4.12; found: C 45.71, H 2.63, N 4.45. The ee value was
measured by HPLC (column: Whelk O1; eluent: n-hexane/isopropanol/
methylene chloride 70:20:10).
(500 MHz, CDCl3, 258C, TMS): d=7.32–7.29 (m, 2H; HAr), 6.95–6.91 (m,
2
2H; HAr), 6.90–6.86 (m, 2H; HAr), 6.80–6.76 (m, 2H; HAr), 4.06 (d, J
N
H)=12.7 Hz, 1H; CH2), 3.93 (d, 2J
ACTHNUTRGNE(UNG H, H)=12.7 Hz, 1H; CH2), 3.84 (s,
3H; OCH3), 3.79 ppm (s, 3H; OCH3); 13C NMR (125 MHz, CDCl3, 258C,
TMS): d=162.0 (CAr), 159.6 (CAr), 131.6 (CAr), 130.1 (CAr), 129.9 (CAr),
126.5 (CAr), 114.3 (CAr), 113.9 (CAr), 62.9 (CH2), 55.5 (OCH3), 55.2 ppm
(OCH3); elemental analysis calcd (%) for C15H16O3S: C 65.19, H 5.84;
found: C 65.49, H 6.05. The ee value was measured by HPLC (column:
Chiralcel OD-H; eluent: n-hexane/isopropanol 90:10).
(R)-4-Bromophenyl 2-methoxybenzyl sulfoxide (10b): M.p. 64–668C
(n-hexane/ethanol 9:1); [a]2D5 =ꢀ58.0 (c=1.1 in CHCl3); 1H NMR
(500 MHz, CDCl3, 258C, TMS): d=7.55–7.52 (m, 2H; HAr), 7.29 (ddd,
4J(H, H)=1.7, 3J(H, H)=7.4, 3J
ACHTUNGTRENNUNG ACHTUNGTRENNUNG ACTUHNGTRENNNGU
2H; HAr), 6.96 (dd, 4J
N
ACHTUNGTRENNUNG
3
4
3
4J
A
G
ACHUTGTNREN(UNG H, H)=1.0, JACHTUNTGREN(NUGN H,
H)=8.2 Hz, 1H; HAr), 4.24 (d, 2J
ACHTUNGTRENNUNG
2J
ACHTUNGTRENNUNG
(125 MHz, CDCl3, 258C, TMS): d=157.5 (CAr), 132.2 (CAr), 131.7 (CAr),
131.5 (CAr), 130.1 (CAr), 126.1 (CAr), 125.4 (CAr), 120.4 (CAr), 117.3 (CAr),
110.2 (CAr), 58.4 (CH2), 55.2 ppm (OCH3); elemental analysis calcd (%)
for C14H13BrO2S: C 51.70, H 4.03; found: C 51.99, H 4.38. The ee value
was measured by HPLC (column: Chiralcel OD-H; eluent: n-hexane/iso-
propanol 95:5).
Acknowledgements
This work was financially supported, in part, by the MIUR Rome (PRIN
Project: Stereoselezione in sintesi organica. Metodologie e applicazioni)
and by the University of Bari. We thank Luca Nigro, Caterina Centrone,
Antonio Spadaro and Elena Lorusso for preliminary oxidation experi-
ments, and Prof. Marcello Colapietro (Dipartimento di Chimica, Univer-
sitꢀ “La Sapienza”, Roma, Italy) and Giuseppe Chita (CNR, Istituto di
Cristallografia, Bari, Italy) for X-ray data collection.
(R)-4-Bromophenyl 3-methoxybenzyl sulfoxide (11b): M.p. 104–1058C
(n-hexane/acetone 9:1); [a]2D5 =+50.0 (c=1 in CHCl3); 1H NMR
(500 MHz, CDCl3, 258C, TMS): d=7.58–7.55 (m, 2H; HAr), 7.26–7.23 (m,
Chem. Eur. J. 2009, 15, 13417 – 13426
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13425

A. Bottoni, C. Cardellicchio et al.
[26] M. I. Donnoli, S. Superchi, C. Rosini, Enantiomer 2000, 5, 181–188.
[27] J. Ham, I. Yang, H. Kang, J. Org. Chem. 2004, 69, 3236–3239.
[28] J. Degani, A. Tundo, Ann. Chim. 1961, 51, 543–549.
[29] C. Wang, A. S. Batsanov, M. R. Bryce, I. Sage, Synthesis 2003, 2089–
2095.
[30] A. Krawczyk, J. T. Wrobel, Rocz. Chemii 1977, 51, 285–290.
[31] H. Kwart, H. Omura, J. Am. Chem. Soc. 1971, 93, 7250–7256.
[32] R. C. Gadwood, I. M. Mallick, A. J. DeWinter, J. Org. Chem. 1987,
52, 774–782.
[33] P. Kelly, S. E. Lawrence, A. R. Maguire, Synlett 2007, 1501–1506.
[34] P. Metrangolo, F. Meyer, T. Pilati, G. Resnati, G. Terraneo, Angew.
Chem. 2008, 120, 6206–6220; Angew. Chem. Int. Ed. 2008, 47, 6114–
6127; F: Naso, C. Cardellicchio, M. A. M. Capozzi, F. Capitelli, V.
Bertolasi, New J. Chem. 2006, 30, 1782–1789.
[1] G. H. Posner in The Chemistry of Sulphones and Sulphoxides (Eds.:
S. Patai, Z. Rappoport, C. Stirling), Wiley, New York, 1988, pp. 823–
849; G. Solladiꢃ in Advanced Asymmetric Synthesis (Ed.: G. R. Ste-
phenson), Chapman and Hall, London, 1996, pp. 60–92; M. C. Car-
reÇo, Chem. Rev. 1995, 95, 1717–1760.
[2] I. Fernandez, N. Khiar, Chem. Rev. 2003, 103, 3651–3705.
[3] J. L. Garcia Ruano, C. Alemparte, M. T. Aranda, M. M. Zarzuelo,
Org. Lett. 2003, 5, 75–78; Z. Han, D. Krishnamurthy, P. Grover,
H. S. Wilkinson, Q. K. Fang, X. Su, Z.-H. Lu, D. Magiera, C. H. Sen-
anayake, Angew. Chem. 2003, 115, 2078–2081; Angew. Chem. Int.
Ed. 2003, 42, 2032–2035.
[4] M. A. M. Capozzi, C. Cardellicchio, F. Naso, Eur. J. Org. Chem.
2004, 1855–1863.
[5] D. A. Cogan, G. Liu, K. Kim, B. J. Backes, J. A. Ellman, J. Am.
Chem. Soc. 1998, 120, 8011–8019.
[35] Y.-T. Wu, Y.-C. Ho, C.-C. Lin, H.-M. Gau, Inorg. Chem. 1996, 35,
5948–5952.
[6] C. Bolm, K. MuÇiz, J. P. Hildebrand in Comprehensive Asymmetric
Catalysis, (Eds.: E. N. Jacobsen, A. Pfalz, H. Yamamoto), Springer,
Berlin, 1999, pp. 697–710.
[7] J. Legros, J. R. Dehli, C. Bolm, Adv. Synth. Catal. 2005, 347, 19–31.
[8] H. Cotton, T. Elebring, M. Larsson, L. Li, H. Sorensen, S. von Unge,
Tetrahedron: Asymmetry 2000, 11, 3819–3825.
[9] P. Pitchen, E. DuÇach, M. N. Deshmukh, H. B. Kagan, J. Am. Chem.
Soc. 1984, 106, 8188–8193; S. H. Zhao, O. Samuel, H. B. Kagan, Tet-
rahedron 1987, 43, 5135–5144; F. Di Furia, G. Modena, R. Seraglia,
Synthesis 1984, 325–327; O. Bortolini, F. Di Furia, G. Licini, G.
Modena, M. Rossi, Tetrahedron Lett. 1986, 27, 6257–6260.
[10] N. Komatsu, M. Hashizume, T. Sugita, S. Uemura, J. Org. Chem.
1993, 58, 4529–4533; N. Komatsu, M. Hashizume, T. Sugita, S.
Uemura, J. Org. Chem. 1993, 58, 7624–7626.
[11] X. Jia, X. Li, L. Xu, Y. Li, Q. Shi, T. T.-L. Au-Yeung, C. W. Yip, X.
Yao, A. S. C. Chan, Adv. Synth. Catal. 2004, 346, 723–726.
[12] M. A. M. Capozzi, C. Cardellicchio, G. Fracchiolla, F. Naso, P. Tor-
torella, J. Am. Chem. Soc. 1999, 121, 4708–4709; M. A. M. Capozzi,
C. Cardellicchio, F. Naso, Tetrahedron: Asymmetry 2004, 15, 1471–
1476.
[13] M. I. Donnoli, S. Superchi, C. Rosini, J. Org. Chem. 1998, 63, 9392–
9395; S. Superchi, P. Scafato, L. Restaino, C. Rosini, Chirality 2008,
20, 592–596.
[14] M. A. M. Capozzi, C. Cardellicchio, F. Naso, V. Rosito, J. Org.
Chem. 2002, 67, 7289–7294.
[15] C. Cardellicchio, O. Hassan Omar, F. Naso, M. A. M. Capozzi, F.
Capitelli, V. Bertolasi, Tetrahedron: Asymmetry 2006, 17, 223–229.
[16] F. Naso, C. Cardellicchio, F. Affortunato, M. A. M. Capozzi, Tetrahe-
dron: Asymmetry 2006, 17, 3226–3229.
[36] P. G. Potvin, B. G. Fieldhouse, Tetrahedron: Asymmetry 1999, 10,
1661–1672.
[37] G. Pescitelli, L. Di Bari, P. Salvadori, Organometallics 2004, 23,
4223–4229; G. Pescitelli, L. Di Bari, P. Salvadori, J. Organomet.
Chem. 2006, 691, 2311–2318.
[38] Y. Zhao, D. G. Truhlar, Acc. Chem. Res. 2008, 41, 157–167.
[39] Y. Zhao, D. G. Truhlar, J. Phys. Chem. A 2004, 108, 6908–6918.
[40] Y. Zhao, D. G. Truhlar, J. Phys. Chem. A 2005, 109, 5656–5667.
[41] Y. Zhao, O. Tishchenko, D. G. Truhlar, J. Phys. Chem. B 2005, 109,
19046–19051.
[42] Gaussian 03, Revision A.1, M. J. Frisch, G. W. Trucks, H. B. Schlegel,
G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr.,
T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J.
Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega,
G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R.
Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchin, J. B. Cross, C.
Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J.
Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Moro-
kuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski,
S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malik,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui,
A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T.
Keith, A. Al-Lahan, C. Y. Peng, A. Nanayakkara, M. Challacombe,
P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, J. A.
Pople, Gaussian Inc., Pittsburg, PA, 2003.
[43] N. Godbout, D. R. Salahub, J. Andzelm, E. Wimmer, Can. J. Chem.
1992, 70, 560–571.
[17] M. Seto, N. Miyamoto, K. Aikawa, Y. Aramaki, N. Kanzai, Y.
Iizawa, M. Baba, M. Shiraishi, Bioorg. Med. Chem. 2005, 13, 363–
386.
[18] B. Jiang, X.-L. Zhao, J.-J. Dong, W.-J. Wang, Eur. J. Org. Chem.
2009, 987–991.
[44] E. A. Meyer, R. K. Castellano, F. Diederich, Angew. Chem. 2003,
115, 1244–1287; Angew. Chem. Int. Ed. 2003, 42, 1210–1250.
[45] T. Janowski, P. Pulay, Chem. Phys. Lett. 2007, 447, 27–32.
ˇ
˘
ˇ
ˇ
[46] M. Pitonaꢆk, P. Neograꢆdy, J. Rezaꢆc, P. Jurecka, M. Urban, P.
Hobza, J. Chem. Theory Comput. 2008, 4, 1829–1834.
[19] C. Bolm, Coord. Chem. Rev. 2003, 237, 245–256.
[20] C. Drago, L. Caggiano, R. F. W. Jackson, Angew. Chem. 2005, 117,
7387–7389; Angew. Chem. Int. Ed. 2005, 44, 7221–7223.
[21] K. P. Bryliakov, E. P. Talsi, T. Kꢄhn, C. Bolm, New J. Chem. 2003,
27, 609–614; D. Balcells, F. Maseras, A. Lledꢅs, J. Org. Chem. 2003,
68, 4265–4274; D. Balcells, F. Maseras, G. Ujaque, J. Am. Chem.
Soc. 2005, 127, 3624–3634.
[47] M. O. Sinnokrot, C. D. Sherill, J. Phys. Chem. A 2006, 110, 10656–
10668.
[48] SADABS, Absorption Correction Program, G. M. Sheldrick, Univer-
sity of Gçttingen, Gçttingen, 1996; R. H. Blessing, Acta Crystallogr.
Sect. A 1995, 51, 33–38; Oxford Diffraction, CrysAlis CCD and
CrysAlis RED, Oxford Diffraction Ltd, Abingdon, 2006.
[49] A. Altomare, M. C. Burla, M. Camalli, G. L. Cascarano, C. Giaco-
vazzo, A. Guagliardi, A. G. G. Moliterni, G. Polidori, R. Spagna, J.
Appl. Crystallogr. 1999, 32, 115–119.
[22] J. Legros, C. Bolm, Chem. Eur. J. 2005, 11, 1086–1092; A. Korte, J.
Legros, C. Bolm, Synlett 2004, 2397–2399.
[23] T. Yamaguchi, K. Matsumoto, B. Saito, T. Katsuki, Angew. Chem.
2007, 119, 4813–4815; Angew. Chem. Int. Ed. 2007, 46, 4729–4731;
K. Matsumoto, T. Yamaguchi, T. Katsuki, Chem. Commun. 2008,
1704–1706.
[50] SHELXL-97. Program for the Refinement of Crystal Structures,
G. M. Sheldrick, University of Gçttingen, Gçttingen, 1997.
[51] M. Nardelli, Appl. Crystallogr. 1995, 28, 659.
[24] J.-F. Bonfanti, D. Craig, Tetrahedron Lett. 2005, 46, 3719–3723.
[25] S. V. Kessar, U. K. Nadir, P. Singh, Y. P. Gupta, Tetrahedron 1978,
34, 449–451.
Received: July 29, 2009
Published online: November 11, 2009
13426
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 13417 – 13426