Selective and practical oxidation of sulfides to diastereopure sulfoxides: A combined experimental and computational investigation

Article abstract of DOI:10.1002/adsc.201200459

We describe an effective oxidation of diltiazem (DTZ)-like molecules (a class of prochiral sulfides with potential pharmacological properties) using m-chloroperbenzoic acid (MCPBA) as oxidant either in dichloromethane or methanol. An excellent diastereomeric excess of one sulfoxide has been observed in the absence of any chiral auxiliary . The stereochemistry of the two diastereomeric sulfoxides has been determined by TDDFT simulations of the experimental electronic circular dichroism (ECD) spectra. A computational DFT study of the reaction mechanism shows that the attack of MCPBA on the two sulfide enantiotopic faces affords two preliminary complexes M1 and M1′. M1 is more stable than M1′ by 3.3 and 3.5kcal mol^-1 in dichloromethane and methanol, respectively, and after equilibration its population must be dominant. Two diastereomeric pathways originate from M1 and M1′ and give two diastereomeric sulfoxides with R and S configurations at the new chiral sulfur, respectively. Since TS (the transition state originating from M1) is more stable than TS′ (the energy gap is 0.7kcal mol ^-1 in dichloromethane or methanol), following the Curtin-Hammett principle, the favoured path is the pro-R channel (M1→TS→M2) affording the (R_c,R_s)-2a′ product species in agreement with the observed diastereoselectivity. The M1-M1′ and TS-TS′ energy gaps are actually determined by the difference in the hydrogen bond network that features the two species even if the approaching orientation of the two molecules is governed by the interactions between the π systems of oxidant and substrate aromatic rings. The diastereomeric ratio computed on the basis of the energy difference between TS and TS′ (0.7kcal mol^-1) is 63:37, which must be compared to the experimental value 9:1. When we consider free energy differences (2.4kcal mol^-1 in vacuum and 2.9kcal mol ^-1 in solution) this theoretical ratio becomes 85:15 and 89:11, respectively, in excellent agreement with the experimental value 9:1. Copyright

Full text of DOI:10.1002/adsc.201200459

FULL PAPERS
DOI: 10.1002/adsc.201200459
Selective and Practical Oxidation of Sulfides to Diastereopure
Sulfoxides: A Combined Experimental and Computational
Investigation
Andrea Bottoni,a Matteo Calvaresi,a,* Alessia Ciogli,b Barbara Cosimelli,c,*
Giuseppe Mazzeo,d Laura Pisani,d Elda Severi,c Domenico Spinelli,a
and Stefano Superchid
a
Dipartimento di Chimica “G. Ciamician”, Universitꢀ di Bologna, Via Selmi 2, 40126 Bologna, Italy
Fax : (+39)-051-209-9456; phone: (+39)-051-209-9478; e-mail: matteo.calvaresi3@unibo.it
Dipartimento di Chimica e Tecnologie del Farmaco, Sapienza Universitꢀ di Roma, Piazzale A. Moro 5, 00185 Roma, Italy
Dipartimento di Chimica Farmaceutica e Tossicologica, Universitꢀ degli Studi di Napoli “Federico II”, Via Montesano 49,
b
c
80131 Napoli, Italy
Fax : (+39)-081-678-630; phone, (+39)-081-678-614; e-mail: barbara.cosimelli@unina.it
Dipartimento di Scienze, Universitꢀ della Basilicata, Viale dell’Ateneo Lucano 10, 85100 Potenza, Italy
d
Received: June 24, 2012; Revised: October 7, 2012; Published online: January 4, 2013
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201200459.
Abstract: We describe an effective oxidation of dil- chloromethane or methanol), following the Curtin–
tiazem (DTZ)-like molecules (a class of prochiral Hammett principle, the favoured path is the pro-R
sulfides with potential pharmacological properties) channel (M1!TS!M2) affording the (Rc,Rs)-2a’
using m-chloroperbenzoic acid (MCPBA) as oxidant product species in agreement with the observed dia-
either in dichloromethane or methanol. An excellent stereoselectivity. The M1–M1’ and TS–TS’ energy
diastereomeric excess of one sulfoxide has been ob- gaps are actually determined by the difference in the
served “in the absence of any chiral auxiliary”. The hydrogen bond network that features the two species
stereochemistry of the two diastereomeric sulfoxides even if the approaching orientation of the two mole-
has been determined by TDDFT simulations of the cules is governed by the interactions between the p
experimental electronic circular dichroism (ECD) systems of oxidant and substrate aromatic rings. The
spectra. A computational DFT study of the reaction diastereomeric ratio computed on the basis of the
mechanism shows that the attack of MCPBA on the energy difference between TS and TS’ (0.7 kcal
two sulfide enantiotopic faces affords two prelimina- molꢀ1) is 63:37, which must be compared to the ex-
ry complexes M1 and M1’. M1 is more stable than perimental value 9:1. When we consider free energy
M1’ by 3.3 and 3.5 kcalmolꢀ1 in dichloromethane differences (2.4 kcalmolꢀ1 in vacuum and 2.9 kcal
and methanol, respectively, and after equilibration molꢀ1 in solution) this theoretical ratio becomes
its population must be dominant. Two diastereomeric 85:15 and 89:11, respectively, in excellent agreement
pathways originate from M1 and M1’ and give two with the experimental value 9:1.
diastereomeric sulfoxides with R and S configura-
tions at the new chiral sulfur, respectively. Since TS
(the transition state originating from M1) is more Keywords: diastereoselection; H-bonds; p-p interac-
stable than TS’ (the energy gap is 0.7 kcalmolꢀ1 in di- tions; QM calculation; sulfoxidation
Introduction
nation capability, they are extensively used in organic
synthesis as chiral auxiliaries and intermediates.[1–8] In
Chiral sulfoxides are precious and valuable reagents addition to their synthetic use, non-racemic or stereo-
that play a key role in asymmetric synthesis. Since defined sulfoxides are important compounds in me-
these species can exert a high asymmetric induction dicinal chemistry and pharmacology. A notable exam-
and are characterized by a significant configurational ple is represented by sulfinyl-substituted benzimida-
stability of the sulfinyl moiety joined to a high coordi- zoles which are proton pump inhibitors and are
Adv. Synth. Catal. 2013, 355, 191 – 202
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
191

Andrea Bottoni et al.
FULL PAPERS
widely employed as powerful anti-ulcer drugs. The
leading compound of this type has been for many
years omeprazole. Even if this compound has been
sold for a long period as a racemic mixture, the S-
(ꢀ)-enantiomer, known as esomeprazole, has been
demonstrated to have more effective activity and is
now available on the market in enantiopure form,
representing one of the most sold pharmaceuticals in
the world.[9–13] Thus, since the development of new
synthetic strategies to obtain enantiomerically and/or
diastereomerically pure sulfoxides is of great interest
to chemists and pharmacologists, during the last two
Figure 1.
decades the efforts of many research groups have MCPBA sulfoxidation[45] is still lacking. Some studies
been devoted to the preparation of stereodefined sulf- describe an active participation of the substrate hy-
oxides.
droxy moieties in determining the direction of stereo-
Enantiomerically enriched sulfoxides can be ob- induction and the presence of H-bonds between sub-
tained using different synthetic approaches.[4,14] Chiral strate and peracid has been invoked.[28,29,34–36,38,41,42]
starting substrates can be stereospecifically converted
In the present paper we describe a diastereoselec-
to chiral sulfoxides and the Andersen reaction can be tive oxidation of diltiazem (DTZ)-like molecules of
considered the prototype example of this proce- type 1 (see Figure 1), which are the starting prochiral
dure.[15] A different approach can be the resolution of sulfide substrates. These species (1) belong to a class
racemic mixtures. Alternatively, chiral non-racemic of compounds characterized by the ability to act as l-
sulfoxides can be obtained by direct oxidation of the type calcium channels (LTCC) blockers.[46,47] Since
prochiral corresponding sulfides via biotransforma- this feature makes them potential drugs for the treat-
tions, by chiral organic oxidant like oxaziridines and ment of several heart diseases,[48,49] we aim to examine
peroxides, or using achiral organic oxidants in combi- structural changes (e.g., oxidation of the sulfide func-
nation with chiral transition metal complexes.[4,14] In tionality) that can modify the pharmacological profile.
particular, the last approach has been extensively in- Compounds 1 possess a stable chiral center adjacent
vestigated, being rather general, powerful, and suita- to the sulfur of the 1,4-thiazino ring and, thus, are op-
ble for scale-up.[16–19] The most common catalysts of timum candidates to test synthetic methods for the
this type are based on titanium complexes, but effec- chiral oxidation of the prochiral sulfur center without
tive methods involving other transition metals (in par- the need of asymmetric auxiliaries. We want to show
ticular vanadium) are well documented.[14,16–26] Diaste- that the success of such a diastereoselective oxidation
reomerically pure sulfoxides can be obtained by not only depends on the presence of a pre-existing
either submitting chiral sulfoxides to diastereoselec- chiral element on the substrate, but also on the specif-
tive processes[1–4] (which generates novel substrate ste- ic nature of the groups on substrate and oxidant: for
reogenic centers) or oxidizing chiral sulfides diaste- a given substrate only the use of a “suitable” oxidant
reoselectively.[4,14] In the latter approach, achiral oxi- can originate intermolecular interactions able to
dants can be used and the basic source of asymmetric “drive” the asymmetric induction and, thus, afford an
induction is the presence of a stereogenic center close effective diastereoselective oxidation.
to the pro-chiral sulfur on the substrate molecule.
To examine the effect on asymmetric induction of
Steric or neighbouring group participation[27] explains these hypothesized intermolecular interactions (spe-
the good diastereomeric excess obtained in many cifically, weak p-p interactions and hydrogen bonds)
cases. Many examples of diastereoselective sulfoxida- between oxidant and substrate, we have employed dif-
tions on different types of chiral sulfides are now ferent oxidative systems, i.e., MCPBA in different sto-
available in the literature.[28–44]
ichiometric amounts and solvents and 1,1,1-trifluoroa-
Different organic and inorganic oxidizing agents cetone/H2O2. Some preliminary experiments had
have been employed to this purpose, showing that shown that yields and, more importantly, the stereo-
high stereoselectivity is achieved only in the presence chemical outcome of the oxidation reaction were
of a specific oxidant/substrate matching.[34–44] One of strictly related to the nature of the employed oxidant
the most frequently used oxidants in these reactions is system. Since the substrate involves aromatic p sys-
m-chloroperbenzoic acid (MCPBA) which provides tems, interactions able to differentiate the two enan-
high levels of diastereoselectivity on a wide range of tiotopic faces of sulfides 1 are expected to be active
substrates.[28–44] In spite of the large number of exam- with MCPBA and not with trifluoroacetone/H2O2.
ples so far reported in the literature, a detailed theo- Thus, to understand the most important mechanistic
retical analysis at the molecular level of the factors features and detect the key reaction parameters re-
determining the stereochemical outcome of the sponsible for the observed stereoselectivity, after the
192
asc.wiley-vch.de
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 191 – 202

Selective and Practical Oxidation of Sulfides to Diastereopure Sulfoxides
experimental synthetic study and the determination lective oxidation of the sulfide and, partially, by a ki-
of relative and absolute configurations by electronic netic resolution process in the subsequent oxidation
circular dichroism (ECD) spectra simulation, we have of sulfoxide. To gain a better insight about the origin
carried out a computational investigation of the of the process stereoselectivity, we repeated the oxi-
mechanism of this oxidation reaction. To this purpose, dation of 1a using only 1 equiv. of MCPBA, thus
following the lines of a recently published work on avoiding the following oxidation to sulfone (entry 4).
this subject,[50] we employed a model system formed In these conditions a high yield (96%) of 2a was ob-
by compound 1a and the MCPBA oxidant.
tained, no sulfone was detected and, worthy of note,
the diastereoselectivity remained substantially con-
stant. This result is important evidence that the ob-
served stereoselectivity was determined only by the
sulfide oxidation process and not by a kinetic resolu-
tion.
Results and Discussion
Oxidation of Sulfides 1: Searching for the Best
Experimental Conditions
To elucidate the role of steric hindrance and elec-
tronic factors in determining the face selectivity that
We describe herein our experimental procedure characterizes the oxidant attack, we performed the
which was aimed to define the best conditions for the oxidation of 1a using the 1,1,1-trifluoroacetone/H2O2
oxidation of 1a and obtain high stereo- and chemose- system, which is characterized by a much smaller size
lectivity avoiding the overoxidation of sulfides to sul- of the oxidant species and the absence of p-interac-
fones. The results obtained in the oxidations of race- tions (entry 5).[51] Under these conditions the reaction
mic sulfide 1a are reported in Table 1.
was significantly slower than with MCPBA: at room
Initially 1a was treated with an excess (3 equiv.) of temperature for 6 h using one equivalent of H2O2,
MCPBA in dichloromethane at room temperature for only 50% of 2a was obtained. Furthermore, a lower
15 min (entry 1). Under these conditions a complete 6:4 diastereoisomeric ratio in the formation of 2a was
conversion of 1a was observed and the corresponding observed.
sulfoxide 2a was obtained in 94% yield and 9:1 diaste-
In a further step of our study we examined the
reoisomeric ratio. Only a 6% yield of sulfone 3a was effect of steric hindrance on rate and/or diastereose-
measured as a result of overoxidation of the sulfoxide lection. To this purpose we used MCPBA to oxidize
group. When the reaction temperature was lowered to sulfide 1b which is characterized by a bulkier isopro-
08C (entry 2) no significant changes in the reaction poxy moiety at the stereogenic carbon C-8. In the
yield and diastereoselectivity were observed, apart presence of either an excess or equimolar quantities
from an increase of the sulfone amount. Finally, after of oxidant (entries 7 and 8) we approximately ob-
a further decrease of temperature (ꢀ788C), the reac- tained the same diastereoisomeric ratio (9:1ꢁ10:1)
tion of 1a (entry 3) became much slower and was not found in the previous experiments. Thus, a change of
completed after 24 h. In this case a moderate increase steric hindrance around the sulfur atom does not ap-
of diastereoselection could be observed, even if the parently influence the stereoselectivity of the
variation of this value can probably be considered MCPBA oxidation and only moderately affects the
within the experimental uncertainty (see Experimen- absolute reactivity.
tal Section). The high diastereomeric ratio obtained
As reported in the introduction, in some cases the
in the MCPBA oxidation was particularly promising preferred attack of MCPBA on one of the two sulfur
and, in principle, could be caused both by a stereose- faces seems to be controlled by hydrogen bonding be-
Table 1. Stereoselective oxidation of sulfides 1.
Entry Sulfide Temp. [8C]/time Oxidant
Solvent Oxidant/sulfide ratio Sulfoxide yield[a] (dr)[b] Sulfone yield[a]
1
2
3
4
5
6
7
8
1a
1a
1a
1a
1a
1a
1b
1b
r.t./15 min
0/20 min
ꢀ78/24 h
r.t./15 min
r.t./6 h
r.t./15 min
r.t./20 min
r.t./20 min
MCPBA
MCPBA
MCPBA
MCPBA
CH2Cl2 3/1
CH2Cl2 3/1
CH2Cl2 3/1
CH2Cl2 1/1
94% (9:1)
87% (9:1)
56% (10:1)
96% (10:1)
50% (6:4)
92% (9:1)
70% (9:1)
50% (10:1)
6%
10%
2%
/
CF3COCH3/H2O2 CHCl3 0.1/1/1[c]
/
MCPBA
MCPBA
MCPBA
MeOH 1/1
CH2Cl2 3/1
CH2Cl2 1/1
6%
10%
/
[a]
[b]
[c]
Isolated yields.
Diastereoisomeric ratio determined by H NMR analysis.
CF3COCH3/H2O2/sulfide ratio.
1
Adv. Synth. Catal. 2013, 355, 191 – 202
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
193

Andrea Bottoni et al.
FULL PAPERS
tween the substrate and the acid carboxyl
moiety.[28,29,34–36,38,41,42] This hypothesis was supported
by the fact that some loss of stereoselectivity in oxida-
tion was observed in protic solvents. To ascertain if
even in the present case such effect could influence
the course of reaction, we oxidized sulfide 1a with
MCPBA in methanol (entry 6). Surprisingly, the usual
9:1 diastereoselectivity ratio was observed, which evi-
denced that in the present reacting system the stereo-
chemical outcome is not influenced by the presence
of a protic solvent such as methanol.
In conclusion, the above discussed experimental re-
sults show that the high stereoselectivity found in the
oxidation of sulfides 1 with MCPBA is affected nei-
ther by steric factors nor by the presence of protic sol-
vents.
Figure 2.
Determination of Relative and Absolute
Configurations by ECD Spectra Simulation
the M-2a and m-2a enantiomeric pairs are shown in
The determination of the relative stereochemistry of Figure S2 and Figure S3 in the Supporting Informa-
the major and minor sulfoxide diastereoisomers was tion, respectively. In both cases, as expected, mirror-
an essential piece of information to obtain a more image ECD spectra were obtained.
complete picture of the reaction and validate the
To determine the absolute and relative configura-
mechanistic scenario obtained from the subsequent tions of the eluted enantiomers we carried out a quan-
computational investigation (vide infra). Unfortunate- tum-mechanical simulation of the ECD spectra of
ly, the study of the reaction stereochemistry was not both 2a’ and 2a’’ diastereoisomers. It is well known
obvious because sulfoxide 2a is lacking in protons that this approach provides reliable information on
suitable to be used as probes to determine its relative the molecular stereochemistry: it allows us to assign
configuration by 1H NMR nOe analysis. Therefore, absolute configuration and, in some cases, when the
we carried out a computational analysis of the ECD ECD spectra of diastereoisomers significantly differ,
spectra to establish the relative stereochemistry of the can also give a reliable estimate of the relative stereo-
sulfoxide products 2a. The oxidation of sulfide 1a gen- chemistry.[52] Thus, we used a time-dependent density
erates two pairs of diastereomeric sulfoxides 2a’ and functional theory (TDDFT) approach[53–55] to compute
2a’’: the former is characterized by a syn relative ste- the ECD spectra of one of the two possible enantio-
reochemistry of the sulfoxide and alkoxy oxygen mers for both 2a’ and 2a’’ pair and, then, we com-
atoms, the latter by an anti relative stereochemistry. pared them to the experimental spectra of the eluted
The syn and anti stereoisomers exist as a pair of Rc,Rs enantiomers.
and Sc,Ss enantiomers and a pair of Rc,Ss, Sc,Rs enan-
tiomers, respectively (see Figure 2).
We give here a short description of the procedure
followed for the syn 2a’ and anti 2a’’ species, but
a more detailed account can be found in the Experi-
mental Section.
Analysis of Sulfoxide 2a Product
After having arbitrarily chosen the (Rc,Rs) absolute
configuration for the syn 2a’ species, a molecular me-
To solve the problem of the relative stereochemistry chanics conformational analysis was carried out. Con-
of the sulfoxide product, we first isolated the two dia- formers obtained from the conformational analysis
stereoisomers 2a, then we separated the correspond- were fully optimized at the DFT level (DFT/B3LYP//
ing racemic mixtures by HPLC on a chiral stationary 6-311+ +G** computations), which evidenced the
phase. After chiral chromatographic separation of the existence of five significantly populated conformers
major and minor diastereoisomer of sulfoxide 2a (99.9% of the overall population). Three of the five
(named here as M-2a and m-2a, each corresponding most populated conformers (79.5% of the overall
to one of the two enantiomeric pairs of Figure 2), the population) are characterized by an axial orientation
corresponding ECD spectra were online directly re- of the S=O and aryl moieties while in the other two
corded (see Figure S1 in the Supporting Information conformers these moieties are equatorial (see Fig-
and the detailed description of the procedure report- ure S4 in the Supporting Information). A similar con-
ed in the Experimental Section). The ECD spectra of formational analysis was carried out on the anti dia-
194
asc.wiley-vch.de
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 191 – 202

Selective and Practical Oxidation of Sulfides to Diastereopure Sulfoxides
stereoisomer 2a’’, where the (Rc,Ss) absolute configu-
ration was arbitrarily chosen. After DFT optimization
of the structures provided by the conformational anal-
ysis, only two significantly populated conformers were
found, representing 99.9% of the overall population
and differing in the relative (axial/equatorial) confor-
mation of S=O and aryl groups (see Figure S5 in the
Supporting Information).
Time-dependent DFT computations with the TZVP
basis were carried out on the DFT/B3LYP/6-311+ +
G** structures. In these computations we used the re-
cently introduced long-range corrected CAM-B3LYP
functional,[56] which is known to provide more accu-
rate values of the relative intensities of ECD bands
with respect to the more common B3LYP function-
al.[57]
The theoretical ECD spectra for (Rc,Rs)-2a’
(Figure 3, a) and (Rc,Ss)-2a’’ (Figure 3, b) were com-
pared to the experimental ones obtained for the enan-
tiomers of the 2a diastereoisomers after chiral HPLC
separation. Since the velocity/length calculated spec-
tra were almost coincident, indicating a satisfactory
computational level, only the velocity-form predicted
spectra are reported in the figures (see Experimental
Section for further details).[58]
The calculated ECD spectrum of (Rc,Rs)-2a’ fea-
tured a strong positive Cotton effect at 296 nm and
two negative effects at 266 and 235 nm, thus showing
a good agreement in Cotton effects shape, sequence,
and sign with the experimental ECD spectrum of the
second eluted peak of M-2a. On the other hand, the
calculated ECD spectrum of (Rc,Ss)-2a’’ stereoisomer
appeared more similar to the spectrum of the second
eluted enantiomer of m-2a (see Figure 3, b). Even if
some differences between the two spectra are evident
(which can be probably ascribed to the low quality of
the online recorded ECD spectrum), the band se-
quence is correctly reproduced and, on the whole, the
agreement can be considered reasonable.
Figure 3. (a) Comparison between the calculated (TDDFT/
CAM-B3LYP/TZVP) ECD spectrum of (Rc,Rs)-2a’ (dashed
line) and the experimental ECD spectrum (solid line) of the
second eluted enantiomer of M-2a. (b) Comparison between
the calculated (TDDFT/CAM-B3LYP/TZVP) ECD spec-
trum of (Rc,Ss)-2a’’ (dashed line) and the experimental ECD
spectrum (solid line) of second eluted enantiomer of m-2a.
For a more accurate simulation calculated spectra are red-
shifted by 40 nm.
Since the ECD spectra of the two diastereoisomers
were quite different, it was possible to assign on the
basis of the computed spectra both relative and abso-
lute configurations. Then, the above analysis has evi-
denced that the major M-2a diastereoisomer is char- online ECD spectra were recorded (Figure S7 in the
acterized by a syn configuration and corresponds to Supporting Information). Again we carried out a simu-
the pair (Rc,Rs)-2a’/ACHTUGNTREN(NUNG Sc,Ss)-2a’. Conversely, the minor lation of the ECD spectra. We arbitrarily fixed the
m-2a diastereoisomer has anti relative configuration (S) absolute configuration at the stereogenic carbon
and is formed by the pair (Rc,Ss)-2a’’/
(Sc,Rs)-2a’’.
and we performed a molecular mechanics conforma-
tional analysis. A further optimization at the DFT/
B3LYP/6-311+ +G** level showed that two conform-
ers, differing in the orientation of the ethoxy and aryl
moieties, were significantly populated (99.9% of the
Analysis of Sulfone 3a Product
A similar analysis was carried out on sulfone 3a to es- overall population, see Figure S8 in the Supporting
tablish the absolute configuration at the carbon ste- Information). comparison of the calculated
A
reogenic centre of the corresponding enantiomers. (TDDFT/CAM-B3LYP/TZVP level) ECD spectra on
After HPLC separation of the enantiomers of sulfone these structures to the experimental ones obtained
3a (Figure S6 in the Supporting Information), their after chromatographic separation, showed a good
Adv. Synth. Catal. 2013, 355, 191 – 202
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
195

Andrea Bottoni et al.
FULL PAPERS
agreement in the case of the first eluted fraction (Fig-
ure S9 in the Supporting Information), supporting the
assignment of the (S) absolute configuration to the
first eluted enantiomer of compound 3a.
Computational Investigation of the Reaction
Mechanism
The energy profile computed for the oxidation reac-
tion of sulfide 1a with MCPBA is reported in
Figure 4, while schematic representations of the struc-
tures of the critical points are shown in Figure 5 and
Figure S10 in the Supporting Information. Sulfide 1a
and the oxidant MCPBA at infinite distance (reac-
tants) represent the asymptotic limit (AL) of the reac-
tion profile indicated in Figure 4. To build the model
system we have chosen for sulfide 1a the enantiomer
with configuration R at the chiral centre adjacent to
sulfur (Rc).
The structure and energy of all critical points have
been computed both in gas-phase and in the presence
of solvent effects using a PCM approach. The corre-
sponding data (absolute energy values and Cartesian
coordinates) are collected in the Supporting Informa-
tion. In Figure 4, in addition to the gas-phase values,
we have reported in parenthesis the energies obtained
in our PCM computations emulating dichloromethane
and methanol. Even if in both cases the activation
barriers and the energies relative to AL do not signifi-
cantly differ from those obtained for the gas-phase
model, we can observe an increase of the energy gap
between the two initial complexes M1 and M1’ and
between the two transition states TS and TS’.
Figure 5. A schematic representation of the structure of the
two intermediates M1 and M1’ and the two corresponding
transition states TS and TS’. Bond lengths are in ꢂngstroms.
The energy values (kcalmolꢀ1) are relative to the asymptotic
limit AL. Values in round parenthesis refer to dichlorome-
thane and values in square brackets to methanol.
Since the solvent has negligible effects on the struc-
ture of the various critical points, in the following dis-
cussion we shall refer to the geometrical parameters
obtained for the gas-phase model. The MCPBA mole-
cule adopts two different orientations to attack each
of the two enantiotopic faces of sulfide 1a. The two
approaching directions lead to the formation of two
complexes M1 and M1’ without overcoming any barri-
er. Since M1 and M1’ afford two diastereomeric prod-
uct molecules differing in the configuration at the
sulfur atom (R and S, respectively), they identify the
starting point of two separate reaction channels that
we denote as pro-R and pro-S channels. The energy
difference between these two species is 2.3 kcalmolꢀ1
in gas-phase and becomes 3.3 and 3.5 kcalmolꢀ1 in di-
chloromethane and methanol, respectively. The lower
energy complex is the pro-R species M1, which is
13.8 kcalmolꢀ1 below AL in gas-phase.
To understand the origin of the M1–M1’ energy dif-
ference we must carefully analyse the non-bonding in-
teractions that keep together the two reactant mole-
cules. In both M1 and M1’ (see Figure 5) the two
Figure 4. The energy profiles for the two pro-R and pro-S
channels computed for the diastereoselective oxidation of
1a using MCPBA. Values in round parenthesis refer to di-
chloromethane (e=8.93) and values in square brackets to
methanol (e=32.61). Energy values are given in kcalmolꢀ1. phenyl rings belonging to substrate (ring A) and oxi-
196
asc.wiley-vch.de
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 191 – 202

Selective and Practical Oxidation of Sulfides to Diastereopure Sulfoxides
dant (ring B) lie on approximately parallel planes, but direct role since they fix the relative orientation of
they do not form the well-known “sandwich” struc- oxidant and substrate: the two molecules are “forced”
ture usually expected in these cases.
to assume in each intermediate the best relative posi-
The two planes are displaced and originate a struc- tion that activates the p-p interactions. The different
tural arrangement which has been demonstrated to hydrogen bond networks observed in the two cases
provide a more effective p-p stabilization.[59–61]
are a consequence of the “forced” relative arrange-
Typical distances between the most effectively in- ment of the two rings. In other words, the two rings
teracting carbon atoms of the two rings are in the on oxidant and substrate work like a “molecular
range 3.2–3.5 ꢃ, as reported in Figure 5. The displace- glue” in the approach of the two molecules.
ment of the two rings is slightly more pronounced in
The two transition states TS and TS’ are almost de-
M1’. In this case the peroxide group must interact generate in gas-phase (20.3 and 20.5 kcalmolꢀ1 above
with the opposite face of the sulfide, which forces ring AL, respectively), but their energy difference increas-
B to drift away from ring A. This motion activates the es in the presence of solvent effects (their energies
interaction of A with the planar peroxide p system are 19.9 and 20.6 kcalmolꢀ1 above AL both in di-
[CO1
ACHTUNGTRENNUNG(O2O3H) group] of B and the interaction of the chloromethane and methanol). The length of the new
ꢀ
bromine atom on A with the p electron cloud of B. forming bond S O3 is similar in the two structures
The O2 C(A) and O3 C(A) distances (3.12 and (2.01 and 2.07 ꢃ, respectively) and in both cases the
3.10 ꢃ, respectively) found in M1’ are comparable to hydrogen atom is still strongly bonded to the oxygen
the C(B) C(A) distance (3.15 ꢃ) computed for M1. O3 (H O3=1.01 ꢃ). The length of the new forming
Thus, on the whole the overall set of p interactions in- OH bond (O1 H in TS and O2 H in TS’) is identical
volving ring A bearing the bromine atom and ring B in the two cases (1.81 ꢃ).
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
bearing the peroxide group, are similar in the two
It is not obvious to figure out why in the gas-phase
cases. What evidently differentiates the two com- the energy difference between the two transition
plexes and makes M1 more stable than M1’ are the structures is only 0.2 kcalmolꢀ1 and this difference be-
hydrogen bonds involving O1 and O4. Because of the comes 0.7 kcalmolꢀ1 with the increasing polarity of
different orientation of the CO1
pattern of these interactions is rather different in the general trends of specific interactions. In TS the two
two species. rings A and B are rather far away and their relative
ACHTUNGERTN(NUNG O2O3H) group, the the medium. However it is possible to identify some
ꢀ
In M1 the O3 H group is involved in two hydrogen positions make unfeasible the activation of whatever
bonds: a strong stabilizing interaction with the oxygen type of p-p interactions (displaced or T-shaped inter-
ꢀ
O4 (O3 H···O4 distance=1.87 ꢃ) and a less impor- actions). The relative position of A and B is rather
tant one with O1 (O3 H···O1=2.53 ꢃ, where the different in TS’ where the two rings are parallel-dis-
ꢀ
three atoms are not colinear). Two additional hydro- placed and characteristic distances between the most
gen interactions again involve the two oxygen atoms effectively interacting atoms of A (including Br) and
ꢀ
ꢀ
O1 and O4 and one aromatic C H bond (O1···H C- B (including the peroxide system) are in the range
ꢀ
A
E
ꢀ
ꢀ
Only the O3 H···O1 and O4···H C H-bonds survive TS’ appears to be more stabilized than TS. The situa-
ꢀ
in M1’. In particular, the strong stabilizing O3 H···O4 tion is reversed when we compare the hydrogen bond
interaction disappears because of the different orien- networks in the two cases. In the bottom part of
tation of MCPBA. It is conceivable that this change Figure 5 we have collected for both TS and TS’ a list
in the hydrogen bond network is mainly responsible of the hydrogen bonds active in each structures. For
for the energy difference between M1 and M1’. As each interaction, to have a rough idea of its strength,
previously noted this energy difference increases in we have reported the H···X distance (X=O1, O2,
the presence of solvent effects. This trend can be O4). From a comparison of the two structures it is
easily understood if we take into account that the par- evident that the number of stabilizing hydrogen
tial charges on oxygen and hydrogen featuring each bonds is larger in TS than TS’: at least three H-bonds
H-bond, interact with the solvent molecules (when which are active in TS, disappear in TS’. Thus the H-
these are implicitly considered in PCM computations) bond network appears to be more stabilizing in the
and enhance the importance of the hydrogen interac- former case than in the latter and seems to compen-
tions in stabilizing M1 with respect to M1’.
sate the opposite effect of the p-p interactions leading
Thus, apparently, the lower energy of the pro-R to a slightly larger stabilization of TS. The effect of
starting intermediate (and, thus, the initial discrimina- the H-bond network in favouring TS with respect to
tion between the two enantiotopic faces of the sul- TS’ becomes more important in the presence of
fide) is not determined by the interactions between a polar medium such as methanol: in this case (as pre-
the p systems of rings A and B (that are similar in viously outlined for M1 and M1’) the partial charges
M1 and M1’), but by the different hydrogen bond pat- on the oxygen and hydrogen atoms interact with the
tern. However, p interactions play a fundamental in-
Adv. Synth. Catal. 2013, 355, 191 – 202
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
197

Andrea Bottoni et al.
FULL PAPERS
solvent molecules, thus increasing the TS–TS’ energy Conclusions
difference.
The potential energy curves obtained for the two In this paper we have described an efficient diastereo-
reaction channels explain the observed diastereoselec- selective oxidation of diltiazem (DTZ)-like molecules
tivity. In the first phase of the reaction (formation of (1a and 1b in Figure 1) used as starting prochiral sul-
the initial adducts) the system equilibrates to form fide substrates. The presence of a stable chiral centre
the more stable species M1 which becomes highly adjacent to a prochiral sulfur makes compounds 1 ex-
dominant. Furthermore, the lower energy of TS with cellent candidates for investigating effective stereose-
respect to TS’ mainly “drives” the system along the lective oxidation procedures that do not require the
pro-R channel to afford M2, that is, the product char- aid of chiral auxiliaries.
acterized by the R configuration of the chiral sulfur
and corresponding to the (Rc,Rs)-2a’ species. This ki-
Our results can be summarized as follows:
netic behaviour is in some way related to the Curtin– (i) The use of MCPBA as oxidant in dichlorome-
Hammett principle. M1 and M1’ can be considered
just like two interconverting conformational isomers
and the preferred path is determined by the relative
energy of the two corresponding transition states TS
thane or in methanol at room temperature af-
fords sulfoxides 2 with high diastereoselection
(observed ratio 9:1) in the absence of any chiral
auxiliary.
and TS’. Interestingly, the diastereomeric ratio com- (ii) The comparison of the experimental ECD spec-
puted on the basis of the energy difference between
the two transition states (0.7 kcalmolꢀ1) is 63:37,
a result close enough (although underestimated) to
the experimental data 9:1. This theoretical ratio sig-
nificantly improves when, in place of potential ener-
gies, we consider free energy changes in vacuum and
solution. In the former case the free energy difference
between TS and TS’ is 2.4 kcalmolꢀ1, which leads to
a diastereomeric ratio of 85:15. In the latter case the
free energy difference is 2.9 kcalmolꢀ1 corresponding
to a diastereomeric ratio of 89:11 which coincides
with the experimental value.
tra of sulfoxides to those obtained by TDDFT
computational simulations for 2a, allowed the as-
signment of the (Rc,Rs)-2a’/ACHTUNTRGNEUNG(Sc,Ss)-2a’ structure to
the pair of prevalent diastereoisomers and the
assignment of absolute configuration to the enan-
tiomers separated by HPLC on chiral stationary
phases.
AHCTUNGTRENNUNG
Furthermore, since M2 (and M2’) are much lower
in energy than reactants (46.0 and 47.7 kcalmolꢀ1 in
the gas-phase, respectively), the reverse reaction is
quite unlikely, which clearly evidences the kinetic
control of diastereoselectivity in this process. A sche-
matic representation of the product complexes M2
and M2’ is given in Figure S10 in the Supporting In-
formation.
Finally, in the light of the previous mechanistic pic-
ture, it is worthwhile to evidence again the diastereo-
meric ratio (9:1) found in methanol (entry 6 in
Table 1), a solvent which often causes a loss in diaste-
reoselectivity[34] because of its ability to interfere with (iv) Since the process is highly exergonic (M2 is
hydrogen bond formation. We have proposed that p-
p interactions (formally reversible) holding together
1a and MCPBA, favour the hydrogen bond formation
46.0 kcalmolꢀ1 lower than reactants), the reverse
reaction is quite unlikely, which implies a kinetic
control of diastereoselectivity.
between the two molecules. The absence of any ef- (v) The M1–M1’ and TS–TS’ energy gaps and, thus,
fects on the diastereomeric ratio when methanol is
used, suggests that the nature of the hydrogen bonds
in M1 and in M1’ , as well as in TS and TS’, must be
similar to that of “intramolecular hydrogen bonds”
which “cannot” be affected by polar protic solvents
such as methanol, via the formation of stable ad-
ducts.[35]
the discrimination between the two enantiotopic
faces of the sulfide, is actually determined by the
variation in the hydrogen bond networks. Appa-
rently, the energy difference between the two
preliminary complexes is not governed by the in-
teractions between the p systems of ring A and
B (similar in the two cases). However, the role
of these interactions is not negligible because the
different hydrogen bond pattern featuring M1
and M1’ is a consequence of the relative arrange-
198
asc.wiley-vch.de
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 191 – 202

Selective and Practical Oxidation of Sulfides to Diastereopure Sulfoxides
organic phase was separated, dried over anhydrous Na2SO4
ment adopted by oxidant and substrate to acti-
vate and optimize these p-p interactions. We
have compared these interaction to a sort of
“molecular glue” that controls the approaching
orientation of the two molecules. This finding
suggests that, for a given substrate, complemen-
tary structural features with the oxidant should
be carefully examined to take advantage of spe-
cific interactions that can lead to the best stereo-
selectivity.
and evaporated. The crude was purified by column chroma-
tography (SiO2; petroleum ether/AcOEt 2:1) affording 2a as
a white solid; yield: 6.7 mg (50%, dr 6:4).
8-(4-Bromophenyl)-8-ethoxy-5-methyl-8H-[1,2,4]oxadi-
azoloACTHNUGTRENNUG[3,4-c]ACHTUGNTREN[NUGN 1,4]thiazin-3-one 7-oxide (2a): Major stereoiso-
mer (M-2a): mp 156.0–157.08C; 1H NMR (400 MHz,
CDCl3): d=(ppm): 1.25 (t, J=6.8 Hz, 3H, CH3), 2.59 (s,
3H, 5-CH3), 3.39 (dq, J=6.8 Hz, J=7.2 Hz, 1H, CH2), 3.74
(dq, J=6.8 Hz, J=7.2 Hz, 1H, CH2), 6.33 (s, 1H, H-6), 7.55
(2H, part AA’ of the system AA’BB’, HAr), 7.65 (2H, part
BB’ of the system AA’BB’, HAr); 13C NMR (100 MHz,
CDCl3): d=15.18, 18.28, 63.95, 89.24, 113.52, 125.89, 127.94,
130.43, 132.35, 137.92, 151.40, 153.98; ESI-MS: m/z=407
(M+23); HM-RS (79Br isotope): m/z=383.9777, calcd. for
C14H13BrN2O4S: 383.9779. Minor stereoisomer (m-2a):
foamy solid; 1H NMR (400 MHz, CDCl3): d=1.24 (t, J=
7.2 Hz, 3H, CH3), 2.33 (s, 3H, 5-CH3), 3.64 (dq, J=6.8 Hz,
J=7.2 Hz, 1H, CH2), 3.91 (dq, J=6.8 Hz, J=7.2 Hz, 1H,
CH2), 5.92 (s, 1H, H-6), 7.10 (2H, part AA’ of the system
AA’XX’, HAr), 7.53 (2H, part XX’ of the system AA’XX’,
HAr); 13C NMR (100 MHz, CDCl3): d=15.36, 17.84, 65.35,
93.00, 113.11, 125.96, 128.55, 131.95, 132.94, 139.35, 152.76,
153.65.
Experimental Section
General Procedures
Melting points were determined using a Bꢄchi apparatus B
1
540 and are uncorrected. H (400 MHz) and 13C (100 MHz)
NMR spectra were recorded in CDCl3 on a Varian INOVA
400 MHz spectrometer using TMS as internal standard. ESI-
mass spectra were obtained on a micromass ZMD Waters
instrument (30 V, 3.2 kV, isotope observed 79Br). EI-mass
spectra were recorded on a VG70 70E apparatus. Column
chromatography was carried out using Macherey–Nagel
Silica Gel 60 (70–230 mesh). Analytical thin-layer chroma-
tography (TLC) was performed on Macherey–Nagel alumi-
num sheets precoated with silica gel (0.25 mm). CH2Cl2 was
freshly distilled over CaH2 prior to its use. MCPBA, hydro-
gen peroxide solution (30% wt) and 1,1,1-trifluoroacetone
were used as purchased (Aldrich). MCPBA was titrated (by
Na2S2O3) before each trial (the titre was always about 76–
77%), to be sure every time about the actual amount of oxi-
dant employed in the reactions. The diastereoisomeric ratio
of sulfoxides 2a and 2b was determined directly on crude
8-(4-Bromophenyl)-8-ethoxy-5-methyl-8H-[1,2,4]oxadi-
azoloACTHNURGTENN[UG 3,4-c]ACTHUNGTREN[NUNG 1,4]thiazin-3-one 7,7-dioxide (3a): mp 213.1–
1
215.08C. H NMR (400 MHz, CDCl3): d=1.29 (t, J=6.8 Hz,
3H, CH3), 2.57 (s, 3H, 5-CH3), 3.35 (dq, J=6.8 Hz, J=
8.4 Hz, 1H, CH2), 3.71 (dq, J=6.8 Hz, J=8.4 Hz, 1H, CH2),
6.24 (s, 1H, H-6), 7.61 (2H, part AA’ of the system AA’BB’,
HAr), 7.66 (2H, part BB’ of the system AA’BB’, HAr);
13C NMR (100 MHz, CDCl3): d=14.96, 18.12, 63.94, 90.35,
112.82, 121.33, 126.42, 131.86, 131.99, 140.24, 152.46, 153.09;
ESI-MS: m/z=423 (M+23); HM-RS (79Br isotope): m/z=
399.9727, calcd. for C14H13BrN2O5S: 399.9729.
1
products by integration of the corresponding H NMR sig-
Oxidation of Sulfide 1b with MCPBA
nals. An uncertainty of about 1% can be then attributed to
these measurements due to the intrinsic error allied to the
integral measurement. Sulfides 1a and 1b were prepared ac-
cording to known procedures.[46]
To a solution of sulfide 1b (20 mg, 0.05 mmol) in dry CH2Cl2
(1 mL) under a nitrogen atmosphere at room temperature
was added MCPBA (8.6 mg, 0.05 mmol). The mixture was
stirred for 1 h and then the solution was diluted with CH2Cl2
and washed first with 5% aqueous Na2SO3 and then with
5% aqueous NaHCO3. The organic phase was dried over an-
hydrous Na2SO4. After evaporation of the solvent, the crude
was purified by column chromatography (SiO2; petroleum
ether/AcOEt 2:1) obtaining 2b as a waxy solid; yield: 10 mg
Oxidation of Sulfide 1a with MCPBA
To a solution of sulfide 1a (45 mg, 0.12 mmol) in dry CH2Cl2
(2.0 mL) under a nitrogen atmosphere at room temperature
was added MCPBA (20 mg, 0.12 mmol). The mixture was
stirred for 15 min, then the solution was diluted with CH2Cl2
and washed with 5% aqueous Na2SO3 and then with 5%
aqueous NaHCO3. The organic phase was separated and
dried over anhydrous Na2SO4. After evaporation of the sol-
vent, the crude material, containing a mixture of two diaste-
reoisomers (96%, 44 mg, dr 10:1), was purified by column
chromatography (SiO2; petroleum ether/AcOEt 2:1) afford-
ing the major sulfoxide of 2a as a white solid (yield: 40 mg)
and the minor one as a foamy solid (yield: 3 mg).
1
(50%, dr 10:1). H NMR (400 MHz, CDCl3): d=0.95 (d, J=
6.0 Hz, 3H, CH3), 1.19 (d, J=6.0 Hz, 3H, CH3), 2.51 (s, 3H,
5-CH3), 3.83–3.95 (m, 1H, CHMe2), 6.28 (s, 1H, H-6); 7.15
(2H, part AA’ of the system AA’XX’, HAr), 7.50 (2H, part
XX’ of the system AA’XX’, HAr).
Enantiomers Separation: Apparatus and
Chromatographic Procedures
The enantiomers of M-2a, m-2a and 3a couples were re-
solved on an analytical chromatograph equipped with
a Rheodyne model 7725i 20 mL loop injector, a PU-980
Jasco HPLC pumps and a dual detection with a Jasco-975
UV detector and a Jasco 995-CD circular dichroism detec-
tor. The online ECD spectra, ranging on 220–420 nm at
4 cmꢀ1 s scan rate, were recorded in stopped-flow mode.
Oxidation of Sulfide 1a with H2O2/CH3COCF3
A solution of sulfide 1a (13 mg, 0.035 mmol) in CHCl3
(26 mL) was added to a solution of H2O2 (30% wt), 3.0 mL,
0.037 mmol)
and
1,1,1-trifluoroacetone
(0.35 mg,
0.004 mmol). The mixture was stirred for 5 h and then the
Adv. Synth. Catal. 2013, 355, 191 – 202
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
199

Andrea Bottoni et al.
FULL PAPERS
Chromatographic data were collected and processed using
Borwin software (Jasco Europe, Italy). All HPLC-grade sol-
vents were purchased from Sigma–Aldrich, and the mobile
phases were 0.45 mm filtered before use. The chiral analyti-
cal column employed was in all cases the (R,R) Whelk-O1
(250 mmꢅ4.6 mm i.d.) operating at room temperature
(258C). Two different eluents were used at flow rate of
1.0 mLminꢀ1: hexane/ethanol/methanol (60/30/10) for m-2a
couple and hexane/dichloromethane (70/30)+2% MeOH
for M-2a and 3a samples. Single UV and CD traces were
carried out at 254 nm.
stable conformer. The minimum energy conformers found
by molecular mechanics, have been further fully optimized
at DFT/B3LYP/6–311+ +G** level as implemented in
Gaussian09 package.[63] All conformers are real minima, no
imaginary vibrational frequencies have been found and the
free energy values have been calculated and used to get the
Boltzmann population of conformers at 298.15 K. Calcula-
tions of ECD spectra were carried out at TDDFT/CAM-
B3LYP/TZVP in the gas phase. The theoretical ECD spec-
tra were obtained as averages weighted on the Boltzmann
populations. The ECD spectra were obtained from calculat-
ed excitation energies and rotational strengths, as a sum of
Gaussian functions centred at the wavelength of each transi-
tion, with a parameter s (half-width of the band) of 0.3 eV.
Computational details for the study of the reaction sur-
face: All DFT computations reported in the paper were per-
formed with the Gaussian09[63] series of programs. Since aryl
groups and extended p systems are present on both sub-
strate and oxidant molecules, a functional capable of de-
scribing interactions involving p system was required. It is
well known that this class of interactions (where medium-
range correlation effects are dominant) are not properly de-
scribed by the most popular DFT functionals, such as, for in-
stance, B3LYP. However, during the last decade new func-
tionals have been recommended which are capable of treat-
ing medium-range correlation effects.[64–67] Within this family
of innovative functionals we have chosen that recently pro-
posed by Truhlar and Zhao, known as M06–2X,[66] which has
been demonstrated to provide a good estimate of p-p inter-
actions and reaction energetics. The 6–31 g* basis set has
been used in all computations.[63]
The geometry of the various critical points on the poten-
tial surface has been fully optimized with the gradient
method available in Gaussian 09 and harmonic vibrational
frequencies have been computed to evaluate the nature of
all critical points. Since the oxidation process was carried
out either in dichloromethane or methanol, the solvent ef-
fects have been taken into account during optimization
using the polarizable continuum model (PCM) in the inte-
gral equation formalism variant (IEFPCM).[68] The solute
cavity was built up using the UFF radii.[69] This approach
places a sphere around each solute atom, with the radii
scaled by a factor of 1.1. Values of 8.93 (dichloromethane)
and 32.61 (methanol) have been employed for the dielectric
constant e.
Chiral Chromatographic Separation of Sulfoxide 2a
and Sulfone 3a
Enantiomers of the major and minor diastereoisomer of 2a
(named M-2a and m-2a), obtained from MCPBA oxidation,
were separated by HPLC on a (R,R)-Whelk-O1 chiral sta-
tionary phase (Figure S1 in the Supporting Information)
and their ECD spectra were online recorded directly after
chromatographic separation by an ECD detector. The best
eluent for the separation of M-2a enantiomers (Figure S1a
in the Supporting Information) achieved a 3.93 a value,
while in the same conditions the m-2a enantiomers were not
resolved. In the latter case a stronger eluent mobile phase
was necessary to obtain the separation of enantiomers with
a=1.09, a value large enough to record the ECD online
spectra. In the separation of M-2a the first eluted enantio-
mer (k=2.42) showed in the ECD spectrum (Figure S2 in
the Supporting Information) an intense negative Cotton
effect at 300–320 nm, followed by two positive CD bands at
264 and 228 nm, respectively. The second eluted enantiomer
(k=9.51) showed, as expected, a mirror-image ECD spec-
trum. On the contrary, the two enantiomers of m-2a
(Figure S1b in the Supporting Information) showed an
ECD spectrum (Figure S3 in the Supporting Information)
characterized by a single intense band centred around
265 nm. The first eluted enantiomer (k=1.62) gave a nega-
tive Cotton effect, while the second eluted one (k=1.76)
showed a single positive ECD band.
Enantiomers of sulfone of 3a (obtained from MCPBA
overoxidation of sulfide 1a) were well separated by HPLC
on (R,R)-Whelk-O1 chiral stationary phase (Figure S6 in
the Supporting Information) and their online ECD spectra
were recorded (Figure S7 in the Supporting Information).
The spectrum of the first eluted enantiomer (k=1.54) fea-
tured an intense negative Cotton effect around 276 nm, fol-
lowed by a second positive signal at 250 nm. As expected,
an oppositely signed ECD spectrum was found for the
second eluted enantiomer (k=1.94).
Acknowledgements
Financial support from Universitꢀ della Basilicata, Universitꢀ
degli Studi di Napoli Federico II, Alma Mater Studiorum
Universitꢀ di Bologna and MIUR (Rome) [PRIN project
2008LYSESR, PRIN project 2009 “Progettazione e sviluppo
di sistemi catalitici innovativi” (PRIN 20099PKPHH 004)] is
gratefully acknowledged.
Computational Details
Computational analysis of UV and ECD spectra: All calcu-
lations have been carried out on a desktop computer endow-
ed with a Intelꢁ Xeon Quadcore E5420 at 2.50 GHz. The
preliminary conformational analysis has been performed by
Spartan02 package[62] using the MMFF94s molecular me-
chanics force field and Monte Carlo searching, fixing arbi-
trary absolute configurations for compounds 2a and 3a. All References
possible conformers were searched, considering the degrees
of freedom of system and retaining only the structures
within an energy range of 4 kcal/mole in respect to the most
[1] G. Solladiꢆ, Synthesis 1981, 185–196.
[2] M. C. CarreÇo, Chem. Rev. 1995, 95, 1717–1760.
200
asc.wiley-vch.de
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 191 – 202

Selective and Practical Oxidation of Sulfides to Diastereopure Sulfoxides
[3] A. J. Walker, Tetrahedron: Asymmetry 1992, 3, 961–
998.
[4] I. Fernandez, N. Khiar, Chem. Rev. 2003, 103, 3651–
3705.
[31] J.-F. Lohier, F. Foucoin, P.-A. Jaffrꢊs, J. I. Garcia, J.
Sopkovꢉ-de Oliveira Santos, S. Perrio, P. Metzner, Org.
Lett. 2008, 10, 1271–1274.
[32] J. Clayden, D. Mitjans, L. H. Youssef, J. Am. Chem.
[5] P. I. Dalko, L. Moisan, Angew. Chem. 2004, 116, 5248–
5286; Angew. Chem. Int. Ed. 2004, 43, 5138–5175.
[6] H. Pellissier, Tetrahedron 2006, 62, 5559–5601.
[7] H. Pellissier, Tetrahedron 2007, 63, 1297–1330.
[8] M. Mellah, A. Voituriez, E. Schultz, Chem. Rev. 2007,
107, 5133–5209.
[9] E. Fellenius, T. Berglindh, G. Sachs, L. Olbe, B. Elan-
der, S.-E. Sjostrand, B. Wallmark, Nature 1981, 290,
159–161.
Soc. 2002, 124, 5266–5267.
[33] M. Kissane, S. L. Lawrence, A. R. Maguire, Tetrahe-
dron: Asymmetry 2010, 21, 871–884.
[34] O. De Lucchi, V. Lucchini, C. Marchioro, G. Valle, G.
Modena, J. Org. Chem. 1986, 51, 1457–1466.
[35] R. Annunziata, M. Cinquini, F. Cozzi, S. Farina, V.
Montanari, Tetrahedron 1987, 43, 1013–1018.
[36] B. M. Escher, R. K. Haynes, S. Kremmydas, D. D.
Ridley, J. Chem. Soc. Chem. Commun. 1988, 137–138.
[37] M. Shimazaki, A. Ohta, Synthesis 1992, 10, 957–958.
[38] J. F. Bower, J. M. J. Williams, Tetrahedron Lett. 1994,
35, 7111–7114.
[10] P. Lindberg, A. Brꢇndstrçm, B. Wallmark, H. Mattsson,
L. Rikner, K.-J. Hoffman, Med. Res. Rev. 1990, 10, 1–
54.
[11] H. Cotton, T. Elebring, M. Larsson, L. Li, H. Sçrensen,
S. von Unge, Tetrahedron: Asymmetry 2000, 11, 3819–
3825.
[12] J. M. Shin, Y. M. Cho, G. Sachs, J. Am. Chem. Soc.
2004, 126, 7800–7811.
[39] K. M. Poss, S. T. Chao, E. M. Gordon, P. J. McCann,
D. P. Santafianos, S. C. Traeger, R. K. Varma, W. N.
Washburn, Tetrahedron Lett. 1994, 35, 3461–3464.
[40] T. Sato, J. Otera, Synlett 1995, 4, 365–366.
[41] J. F. Bower, C. J. Martin, D. J. Rawson, A. M. Z.
Slawin, J. M. J. Williams, J. Chem. Soc. Perkin Trans.
1 1996, 333–342.
[13] J. Legros, J. R. Dehli, C. Bolm, Adv. Synth. Catal. 2005,
347, 19–31.
[14] E. Wojaczynska, J. Wojaczynsy, Chem. Rev. 2010, 110,
4303–4356.
[15] K. K. Andersen, W. Gaffield, N. E. Papanikolaou, J. W.
Foley, R. I. Perkins, J. Am. Chem. Soc. 1964, 86, 5637–
5646.
[16] P. Pitchen, C. J. France, I. M. McFarlane, C. G. Newton,
D. M. Thompson, Tetrahedron Lett. 1994, 35, 485–488.
[17] P. Pitchen, Chem. Ind. 1994, 16, 636–639.
[18] P. Pitchen, D. Thompson, PCT Int. Appl. WO 94-
GB1472 19940707, 1994; Chem. Abstr. 1995, 122,
213941.
[19] P. Pitchen, Asymmetric synthesis of sulfoxides: two case
studies, in: Chirality in Industry II, (Eds.: A. N. Collins,
G. N. Sheldrake, J. Crosby), Wiley, New York, 1997,
p 381.
[42] T.-K. Yang, H.-Y. Chu, D.-S. Lee, Y.-Z. Jiang, T.-S.
Chou, Tetrahedron Lett. 1996, 37, 4537–4540.
[43] N. Maezaki, A. Sakamoto, N. Nagahashi, M. Soejima,
Y.-X. Li, T. Imamura, N. Kojima, H. Ohishi, K. Sakagu-
chi, C. Iwata, T. Tanaka, J. Org. Chem. 2000, 65, 3284–
3291.
[44] S. Singh, S. Kumar, S. S. Chimni, Tetrahedron: Asymme-
try 2001, 12, 2457–2462.
[45] C. G. Overberger, R. W. Cummins, J. Am. Chem. Soc.
1953, 75, 4250–4254.
[46] R. Budriesi, E. Carosati, A. Chiarini, B. Cosimelli, G.
Cruciani, P. Ioan, D. Spinelli, R. Spisani, J. Med. Chem.
2005, 48, 2445–2456.
[47] E. Carosati, G. Cruciani, A. Chiarini, R. Budriesi, P.
Ioan, R. Spisani, D. Spinelli, B. Cosimelli, F. Fusi, M.
Frosini, R. Matucci, F. Gasparrini, A. Ciogli, P. J. Ste-
phens, F. J. Devlin, J. Med. Chem. 2006, 49, 5206–5216.
[48] R. Budriesi, B. Cosimelli, P. Ioan, M. P. Ugenti, R. Spi-
sani, Curr. Med. Chem. 2007, 14, 279–287.
[20] F. Di Furia, G. Modena, R. Seraglia, Synthesis 1984,
325–326.
[21] P. Pitchen, H. B. Kagan, Tetrahedron Lett. 1984, 25,
1049–1052.
[22] P. Pitchen, E. DuÇach, M. N. Desmukh, H. B. Kagan, J.
Am. Chem. Soc. 1984, 106, 8188–8193.
[23] N. Komatsu, Y. Nishibayashi, T. Sugita, S. Uemura, Tet-
[49] F. J. Sharom, Pharmacogenomics 2008, 9, 105–127.
[50] F. Naso, M. A. M. Capozzi, A. Bottoni, M. Calvaresi,
V. Bertolasi, F. Capitelli, C. Cardellicchio, Chem. Eur.
J. 2009, 15, 13417–13426.
rahedron Lett. 1992, 33, 5391–5394.
[24] N. Komatsu, M. Hashizume, T. Sugita, S. Uemura, J.
Org. Chem. 1993, 58, 4529–4533.
[25] M. I. Donnoli, S. Superchi, C. Rosini, J. Org. Chem.
[51] P. Lupattelli, R. Ruzziconi, P. Scafato, A. Degl’Innocen-
ti, A. Belli Paolobelli, Synth. Commun. 1997, 27, 441–
446.
1998, 63, 9392–9395.
[26] S. Superchi, P. Scafato, L. Restaino, C. Rosini, Chirality
2008, 20, 592–596.
[27] A. H. Hoveyda, D. A. Evans, G. Fu, Chem. Rev. 1993,
93, 1307–1370.
[52] J. Autschbach, L. Nitsch-Velasquez, M. Rudolph, Top.
Curr. Chem. 2011, 298, 1–98.
[53] A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.
[54] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37,
785–789.
[28] N. Khiar, I. Alonso, N. Rodrꢈguez, A. Fernꢉndez-May-
oralas, J. Jimꢆnez-Barbero, O. Nieto, F. Cano, C. Foces-
Foces, M. Martꢈn-Lomas, Tetrahedron Lett. 1997, 38,
8267–8270.
[55] P. J. Stephens, F. J. Devlin, C. F. Chabalowski, M. J.
Frisch, J. Phys. Chem. 1994, 98, 11623–11627.
[56] T. Yanai, D. P. Tew, N. C. Handy, Chem. Phys. Lett.
2004, 393, 51–57.
[29] N. Khiar, Tetrahedron Lett. 2000, 41, 9059–9063.
[30] P. B. Hitchcock, G. J. Rowlands, R. J. Seacome, Org.
Biomol. Chem. 2005, 3, 3873–3876.
[57] A. Evidente, S. Superchi, A. Cimmino, G. Mazzeo, L.
Mugnai, D. Rubiales, A. Andolfi, A. M. Villegas-
Fernꢉndez, Eur. J. Org. Chem. 2011, 5564–5570.
Adv. Synth. Catal. 2013, 355, 191 – 202
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
201

Andrea Bottoni et al.
FULL PAPERS
[58] A. Moscowitz, in: Modern Quantum Chemistry, (Ed.:
O. Sinanoglu), Academic Press, London, 1965; p 31.
Part III.
[59] E. A. Meyer, R. K. Castellano, F. Diederich, Angew.
Chem. 2003, 115, 1244–1287; Angew. Chem. Int. Ed.
2003, 42, 1210–1250.
Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi,
N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts,
R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C.
Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma,
V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannen-
berg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Fores-
man, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian 09,
Revision A.02, Gaussian, Inc., Wallingford, CT, 2009.
[64] Y. Zhao, D. G. Truhlar, J. Phys. Chem. A, 2004, 108,
6908–6918.
[60] T. Janowsky, P. Pulay, Chem. Phys. Lett. 2007, 447, 27–
32.
ˇ
ˇ
ˇ
[61] M. Pitonak, P. Neogrꢀdy, J. R. Zꢀc, P. Jurecka, M.
Urban, P. Hobza J. Chem. Theory Comput. 2008, 4,
1829–1834.
[62] SPARTAN ꢋ02; Wavefunction Inc.: Irvine, CA; http://
www.wavefunction.com.
[65] Y. Zhao, D. G. Truhlar, J. Phys. Chem. A, 2005, 109,
5656–5667.
[63] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuse-
ria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V.
Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji,
M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J.
Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M.
Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida,
T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven,
J. A. Montgomery Jr, J. E. Peralta, F. Ogliaro, M. Bear-
park, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staro-
verov, R. Kobayashi, J. Normand, K. Raghavachari, A.
[66] Y. Zhao, D. G. Truhlar, J. Phys. Chem. B 2005, 109,
19046–19051.
[67] Y. Zhao, D. G. Truhlar, Acc. Chem. Res. 2008, 41, 157–
167.
[68] J. Tomasi, B. Mennucci, R. Cammi, Chem. Rev. 2005,
105, 2999–3093.
[69] A. K. Rappꢊ, C. J. Casewit, K. S. Colwell, W. A. God-
dard III, W. M. Skid, J. Am. Chem. Soc. 1992, 114,
10024–10039.
202
asc.wiley-vch.de
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 191 – 202