New insights into the titanium-mediated enantioselective oxidation of fluorinated aryl benzyl sulfides and aryl phenacyl sulfides

Article abstract of DOI:10.1016/j.tet.2019.03.016

A fruitful switch from tert-butyl to cumene hydroperoxide was able to overcome a difficulty arose in the enantioselective oxidation of fluorinated aryl benzyl sulfide with hydroperoxides in the presence of a titanium/(S, S)-hydrobenzoin catalyst. New expe

Full text of DOI:10.1016/j.tet.2019.03.016

Accepted Manuscript
New insights into the titanium-mediated enantioselective oxidation of fluorinated aryl
benzyl sulfides and aryl phenacyl sulfides
Maria Annunziata M. Capozzi, Vanni Frascaro, Gennaro Pescitelli, Cosimo
Cardellicchio
PII:
S0040-4020(19)30291-1
https://doi.org/10.1016/j.tet.2019.03.016
TET 30202
DOI:
Reference:
To appear in: Tetrahedron
Received Date: 14 January 2019
Revised Date: 8 March 2019
Accepted Date: 9 March 2019
Please cite this article as: Capozzi MAM, Frascaro V, Pescitelli G, Cardellicchio C, New insights into
the titanium-mediated enantioselective oxidation of fluorinated aryl benzyl sulfides and aryl phenacyl
sulfides, Tetrahedron (2019), doi: https://doi.org/10.1016/j.tet.2019.03.016.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
ox, cat*
-hexane
Ar
Ar
S
R
S
R
n
O
93 > 98% ee
S, S)-hydrobenzoin
cat* = 5% Ti(O-i-Pr)4/(
Ox= TBHP or CHP
R= C6F5, C6H5CH2C(O)

ACCEPTED MANUSCRIPT
New insights into the titanium-mediated enantioselective oxidation of
fluorinated aryl benzyl sulfides and aryl phenacyl sulfides.
Maria Annunziata M. Capozzi,[b] Vanni Frascaro,[b] Gennaro Pescitelli,[c] Cosimo Cardellicchio*[a]
[a] Dr. C. Cardellicchio, Corresponding Author
CNR ICCOM - Dipartimento di Chimica
Università di Bari
via Orabona 4. 70125 Bari, Italy.
E-mail: cardellicchio@ba.iccom.cnr.it
[b] Dipartimento di Chimica. Università di Bari
via Orabona 4. 70125 Bari- Italy
[c] Dipartimento di Chimica e Chimica Industriale. Università di Pisa
via Moruzzi 13, 56124 Pisa- Italy
Keywords:
Catalysis; Sulfoxide; Circular Dichroism; Oxidation mechanism; Hydrobenzoin.
Abstract.
A fruitful switch from tert-butyl to cumene hydroperoxide was able to overcome a difficulty arose
in the enantioselective oxidation of fluorinated aryl benzyl sulfide with hydroperoxides in the
presence of a titanium/(S, S)-hydrobenzoin catalyst. New experiments show the complementarity of
the old and the new protocols and indicate unequivocally the right choice leading to the
corresponding highly enantioenriched sulfoxides. Moreover, in a totally unexpected way, the new
protocol was able to overcome another difficulty arose in another field of research, that is the
enantioselective oxidation of a fluorinated aryl phenacyl sulfide. Also in this case, the
complementarity of behavior is acting. Finally, this investigation gives new support to the
attribution of configuration of sulfoxides with ECD techniques, but only if the protocol outlined in
our past research was followed thoroughly.
Introduction
The synthesis of enantiopure sulfoxides1-3 is a milestone in asymmetric synthesis, because these
molecules have been employed as starting materials in many synthetic strategies.1-5 Moreover, some
of them are bioactive6 and two of them, (S)-omeprazole and (R)-modafinil, are currently sold in
1

ACCEPTED MANUSCRIPT
large amounts as drugs.6-9
The progress in the asymmetric synthesis of sulfoxides1-3 started at beginning of the '60's, when the
Andersen-Mislow procedure, and all the methods deriving from it, provided for many years a secure
source of some classes of enantiopure sulfoxides.1-2 They were obtained by an enantiospecific
displacement with organometallic reagents (e.g. a Grignard reagent) of suitable leaving groups on
an S-resolved sulfinyl compound, thus achieving the formation of a new carbon-sulfur bond.
In a different approach, since the pioneering work of Modena and Kagan in 1984,1-2 pro-chiral
sulfides were asymmetrically oxidized to chiral sulfoxides by suitable oxidants in the presence of a
chiral metal complex, in this case with the formation of a new sulfur-oxygen bond.3
These two strategies have represented the classical methods, until a third and different approach of
forming a new carbon-sulfur bond has achieved only recently the stage of a mature technology.10-13
In this strategy, sulfenate anions are asymmetrically arylated10-11 in the presence of chiral metal
complexes. Metal-free asymmetric arylations were also reported under phase transfer conditions12
or in the presence of iodonium salts.13
Other strategies, based upon biocatalyzed transformation,1-2 metal-free oxidation procedures1-3 and
carbanionic leaving groups displacements,14 received less attention.
Notwithstanding this large variety of methods, the production on an industrial scale of the two most
relevant enantiopure sulfoxides (the blockbuster (S)-omeprazole and (R)-modafinil)7-9 is achieved
by an enantioselective oxidation of the corresponding sulfide with hydroperoxides in the presence
of the “classical” Kagan-Modena titanium/diethyl tartrate complex, according to the useful
modifications introduced by the Astra-Zeneca company.7-8
After these modifications, this approach becomes well suited for an industrial production because:
(i) titanium is a cheap and non-toxic metal; (ii) diethyl tartrate is a very cheap stereodefinite ligand;
(iii) the crude reaction mixture can be treated without resorting to chromatography; (iv) non-toxic
solvents can be used; (v) the reaction is performed at room temperature, or with only a moderate
2

ACCEPTED MANUSCRIPT
heating.
On the other hand, nowadays it looks unlikely that methods that employ organometallic reagents
(such as the Andersen procedure1-2) or expensive chiral ligands of precious metals (such as the
enantioselective arylation of sulfenate anions10-11) can have a comparable success in industry.
Having in mind this framework, our research in the last years15-28 has dealt with the enantioselective
oxidation of sulfides with hydroperoxides in the presence of a catalytic amount of the 1:2 complex
between titanium i-propoxide and (S, S)- or (R, R)-hydrobenzoin.29 Hydrobenzoin is a cheap chiral
ligand depicted in Figure 1.
HO
OH
Figure 1. (S, S)-hydrobenzoin
This procedure was first reported some decades ago,30 but systematically investigated in its many
features only by our work.15-28 Beyond the case of the asymmetric synthesis of Sulindac esters,15 we
were struck by the highly valuable results obtained studying the asymmetric synthesis of aryl
phenacyl sulfoxides16 and aryl benzyl sulfoxides with this protocol.17-24
The latter case is of particular relevance. Aryl benzyl sulfides were oxidized with high
enantioselectivity by the oxidation system formed by tert-butyl hydroperoxide in the presence of a
chiral titanium/hydrobenzoin complex catalyst.17-24
The reaction proceeds with good isolated yields (57-96%), because only traces of the corresponding
sulfones were produced, if any.17-22 This procedure faces satisfactory with the industrial oxidation
based upon hydroperoxides in the presence of a titanium/diethyl tartrate protocol, because the only
difference lies in the fact that enantiopure hydrobenzoin is a little more expensive than the very
cheap diethyl tartrate. Similar to the titanium/diethyl tartrate procedure, our protocol also uses a
cheap titanium compound as a metal catalyst; in the cases in which a large scale synthesis was
investigated,17, 22 our protocol was also optimized for a chromatography-free purification; our
3

ACCEPTED MANUSCRIPT
protocol too does not use highly toxic solvents and is performed at room temperature; finally, our
protocol has a simple work-up, just a mix-and-wait procedure.17-23
Tens of enantiopure (>98% ee) aryl benzyl sulfoxides were obtained by using our strategy, thus
yielding a rich chemical library of chiral nonracemic molecules.23 Some of them were also
transformed into other different enantiopure sulfoxides.17, 23 Some exceptions21, 23 of a lower
enantioselectivity were connected only to the presence on the starting sulfide of free hydroxy- or
amino-groups, that should be able to alter the productive coordination of the sulfide to the titanium.
However, it must be underlined that these few cases are easily predictable on the basis of our
theoretical and mechanistic investigations.17-23
An unexpected and singular behavior of this oxidation system emerged when we tried to oxidize
2,3,4,5,6-pentafluorobenzyl 2,3,4,5,6-pentafluorophenyl sulfide, after having successfully oxidized
many sulfides bearing only one pentafluorophenyl moiety.20, 22-24 An “unusual” 61% ee (unusual, if
compared with the many enantiopure products obtained during the years) was observed, together
with a 19% isolated yield. This result remained a serious issue of our oxidation system, until we
found that the counter-intuitive substitution of the tert-butyl hydroperoxide (TBHP) with cumene
hydroperoxide (CHP), usually a less performing oxidant in this type of oxidation,15, 24 yielded the
enantiopure (>98% ee) pentafluorobenzyl pentafluorophenyl sulfoxide in good yields (76%)
without resorting to a chromatographic separation.24
The whole reactivity framework was analyzed by DFT calculations.18, 20, 24 A general mechanism,
that follows the development of the experimental work-up, was designed. Five cases of different
fluorinated and non-fluorinated aryl benzyl sulfides were chosen as representative items of
successful and troublesome reactions, and the reaction mechanism was tested in all these situations.
18, 20, 24 DFT calculations were able to account for all the experimental results obtained so far. In the
case of the successful oxidation with TBHP of non-fluorinated aryl benzyl sulfides,18 the
stereochemical pathway towards the preferred (R)-sulfoxide, when (S, S)-hydrobenzoin was used as
4

ACCEPTED MANUSCRIPT
a chirality inducer, was due to an intermediate stabilized by CH···π interactions.31 On the other
hand, in the case of aryl benzyl sulfides bearing only one pentafluorophenyl moiety, our calculation
showed20 that, when the same chiral ligand was used, the stereochemical pathway towards the
preferred (R)-sulfoxide was due to an intermediate stabilized by π−π interactions.32 The case of the
successful oxidation of the pentafluorobenzyl pentafluorophenyl sulfide with CHP was peculiar,24
because our calculations proposed that the route to the (R)-sulfoxide was stabilized by the
interactions involving the two π-systems of the pentafluorophenyl groups of the sulfide, the first
interacting with one phenyl group of one hydrobenzoin, and the other with the phenyl group of the
CHP. In Figure 2, the crucial oxygen transfer from CHP to the pentafluorobenzyl pentafluorophenyl
sulfide with the formation of the (R)-sulfoxide is sketched.
Figure 2. DFT calculated reaction pathway for the oxygen transfer around the chiral titanium
complex from CHP to the pentafluorobenzyl pentafluorophenyl sulfide 1a with the formation of
sulfoxide 1b.
At this point, with the support of our calculations for the reaction mechanism, we considered of
interest to examine in more details the new CHP-based oxidation to evaluate new opportunities that
this protocol can provide.
5

ACCEPTED MANUSCRIPT
Results and Discussion
Having in mind the mechanism outlined in Figure 2, we organized the following experimental work
on the enantioselective oxidation of aryl sulfides with hydroperoxides, in the presence of catalytic
amounts (5%) of a complex between titanium and (S, S)-hydrobenzoin, as shown in Table 1.
6

ACCEPTED MANUSCRIPT
Table 1. Enantioselective oxidation of aryl benzyl or phenacyl sulfides.
Ar
ox, cat*
Ar
S
R
S
R
n-hexane
O
cat* = 5% Ti(O-i-Pr)4/(S, S)-hydrobenzoin
Yield (%)a
19
ee (%)b
61
Entry
1 c
2 c
3
Ar
C6F5
R
Sulfide Oxidant Product
C6F5
C6F5
TBHP
CHP
(R)-1b
(R)-1b
(R)-2b
(R)-2b
(R)-3b
(R)-3b
(R)-4b
(R)-5b
(R)-5b
(R)-6b
(R)-6b
(R)-7b
(R)-7b
(R)-8b
(R)-8b
1a
1a
2a
2a
3a e
3a e
4a
5a
5a
6a
6a
7a
7a
8a
8a
C6F5
76 d
16
>98
36
4-H-C6F4
4-H-C6F4
C6Cl5
C6F5
TBHP
CHP
4
C6F5
79
>98
5
C6F5
TBHP
CHP
-- f
6
C6Cl5
C6F5
51
93 (>98) g
>98
7 h
2,4-Cl2C6H3
2,6-Cl2C6H3
2,6-Cl2C6H3
C6F5
C6F5
TBHP
TBHP
CHP
91
8
C6F5
-- f
9
C6F5
33
89
60
10 i
11
12 i
13
14
15
C6H5-C(O)
C6H5-C(O)
C6H5-C(O)
C6H5-C(O)
TBHP
CHP
20
C6F5
58 d
>98
>98
33
2-F-C6H4
2-F-C6H4
TBHP
CHP
88
84
2,4-Cl2C6H3 C6H5-C(O)
2,4-Cl2C6H3 C6H5-C(O)
TBHP
CHP
80 d
83 d
>98
57
a Yields refer to pure isolated products. b Determined by HPLC (see text). c Data already reported in
ref. 24. d Yield refer only to the precipitated product. Further batch of sulfoxide can be obtained by
chromatography of the mother liquor. e 10% of carbon tetrachloride was added to the n-hexane to
improve the solubility of the mixture. f Analysis of the crude reaction mixture revealed too low
amounts of sulfoxide (<10%) to undertake a worthy separation. The reaction was not investigated
further. g After crystallization. h Data already reported in ref. 22. i Data already reported in ref. 28.
For the sake of the clarity, in Table 1, some previous results were added to those obtained in the
present work. The present research moved from the observation that the 2,3,4,5,6-pentafluorobenzyl
2,3,4,5,6-pentafluorophenyl sulfide 1a was oxidized (Table 1, entry 1) by TBHP in the presence of
7

ACCEPTED MANUSCRIPT
the complex between titanium and (S, S)-hydrobenzoin with lower yields (19%) and with lower ee
values (61%) in comparison with other fluorinated aryl benzyl sulfoxides that were obtained in an
enantiopure form.20, 24 On the other hand, the fruitful switch to CHP as the oxidant agent with the
same chiral complex in the same reaction conditions yielded the enantiopure (R)-1b in good isolated
yields (76%, Table 1, entry 2).24
The DFT calculated mechanism of the successful CHP oxidation of sulfides in the presence of the
titanium/hydrobenzoin complex requires the presence of two pentafluorophenyl groups on the
sulfide to be oxidized.24 Thus, we reasoned that the analogous 2,3,4,5,6-pentafluorobenzyl 2,3,5,6-
tetrafluorophenyl sulfide 2a, in which there is only a substitution of the fluorine in the para-position
of the pentafluorophenyl group with one hydrogen atom, should behave in the same manner of
sulfide 1a. Actually, a low yield and a low ee value were obtained when sulfide 2a was oxidized
with TBHP (Table 1, entry 3) in the presence of the usual chiral catalyst. On the other hand (Table
1, entry 4), a decisive improvement was obtained by using CHP as the oxidant agent. A good yield
(79%) and an enantiopure compound (>98% ee) were observed, in strict analogy with the results
reported for sulfide 1a (Table 1, entries 3 and 4).
At this point, we decided to investigate the oxidation of 2,3,4,5,6-pentafluorobenzyl 2,3,4,5,6-
pentachlorophenyl sulfide 3a, deriving from a substitution of the pentafluorophenyl with a
pentachlorophenyl group (Table 1, entries 5 and 6). In principle, the nature of electron-poor aryl
group was maintained after this substitution. However, in our calculated mechanism,24 some
fluorine atoms contribute to establish further weak stabilizing interactions, and their substitution, in
principle, could not be innocent. From an experimental point of view, when sulfide 3a was oxidized
with TBHP in the presence of the titanium/hydrobenzoin complex (Table 1, entry 5), low amounts
of sulfoxide 3b were obtained (<10%), and thus we decided to avoid the purification of the reaction
mixture, as an unworthy operation. On the other hand, the reaction with CHP (table 1, entry 6)
yielded the corresponding sulfoxide 3b in high enantiopurity (93% ee; >98 after the crystallization)
8

ACCEPTED MANUSCRIPT
and in a 51% isolated yield. Even if satisfactory, this yield is lower than the values (76-79%)
obtained for the fluorinated sulfoxide 1b and 2b. Attempts to improve this value by increasing the
reaction time were not satisfactory. Actually, the substitution of the pentafluorophenyl with the
pentachlorophenyl moiety causes a decrease in the performance of the oxidation system.
Along the lines of the investigation of the chlorinated sulfides, we recalled that the asymmetric
catalyzed oxidation of 2,4-dichlorophenyl pentafluorobenzyl sulfide 4a was successful with TBHP
(Table 1, entry 7),22 and thus new reaction conditions were not required. However, we decided to
synthesize the 2,6-dichlorophenyl pentafluorobenzyl sulfide 5a that has only two chlorine atoms, as
sulfide 4a, but both close to the sulfur atom to be oxidized, as in the case of the pentachlorophenyl
sulfide 3a.
Sulfide 5a was oxidized first with TBHP in the presence of the usual titanium/hydrobenzoin catalyst
(Table 1, entry 8). The behavior was similar to the case of the oxidation of sulfide 3a, that is a very
low yield of sulfoxide 5b (that was not isolated). On the other hand, the switch of the oxidant from
TBHP to CHP was successful also in this case (table 1, entry 9). The sulfoxide 5b was obtained in
high enantiomeric purity (89% ee), even if the isolated yield was not so high (33%), due to the large
amounts of unreacted sulfide 5a that were collected after the reaction.
In summary, the dichotomy of behavior (successful with an oxidant; unsuccessful with the other
one) in the asymmetric oxidation in the presence of the titanium/hydrobenzoin complex was
maintained also in the cases shown in Table 1, entries 1-9. CHP was a better oxidant in the
prototype case of pentafluorobenzyl pentafluorophenyl sulfide 1a, in the case of the analogous 2a
and, to a certain extent, also in the case of the chlorinated 3a. When only two chlorine atoms are
present, as in the case of sulfides 4a and 5a, only when these atoms are close to the sulfur atom, as
in 5a, CHP is a better oxidant. Otherwise, as it occurs in 4a, the usual highly successful TBHP-
oxidation must be recommended.
9

ACCEPTED MANUSCRIPT
At this point, in the analysis of results that were not satisfactory when TBHP was used as the
oxidant, we were struck by a special and recent case in the oxidation of a fluorinated aryl phenacyl
sulfide.28 In a previous work,16 we reported that TBHP in the presence of a complex between
titanium and (S, S)-hydrobenzoin is highly successful also in the asymmetric oxidation of aryl
phenacyl sulfides (91->98% ee values). However, when pentafluophenyl phenacyl sulfide 6a was
oxidized, a 20% yield with “only” a 60% ee value was obtained for the sulfoxide 6b (Table 1, entry
10).28 These values are singularly similar with the results reported in Table 1, entry 1, for the
TBHP-oxidation of the pentafluorobenzyl pentafluorophenyl sulfide 1a.
This result looks to be a mere coincidence. Even if a mechanism for the enantioselective oxidation
of aryl phenacyl sulfides with hydroperoxides in the presence of a titanium/hydrobenzoin catalyst
has not been designed, it is reasonable to believe that the presence of the carbonyl oxygen atom in
these substrates should arrange the acting molecules around the chiral titanium catalyst in a way
that is different with respect to aryl benzyl sulfides, that lacks the coordinating carbonyl moiety.
However, we decided to test also the enantioselective oxidation of the pentafluorophenyl phenacyl
sulfide 6a with CHP in the presence of the usual titanium/hydrobenzoin catalyst (Table 1, entry 11).
With our gratification, also this reaction provided the enantiopure phenacyl sulfoxide 6b (>98% ee)
in satisfactory yields (58%). Thus, in an unexpected way, we added another item to the list of
sulfides for which it is possible to return to the usual trend of high enantioselectivity by switching
from the most successful TBHP to the less usual CHP.
We investigated other aryl phenacyl sulfides to gain more insight into this oxidation. The
enantioselective oxidation of 2-fluorophenyl phenacyl sulfide 7a with TBHP in the presence of the
titanium/hydrobenzoin complex (Table 1, entry 12) yielded the enantiopure sulfoxide 7b in good
isolated yields (88%), as reported previously.28 At this stage, we decided to perform the same
reaction using CHP as the oxidant agent. We observed a slight decrease of the yield (Table 1, entry
13), but a critical decrease of the ee value (33%). This system behaves as the large majority of the
10

ACCEPTED MANUSCRIPT
aryl benzyl sulfides, for which TBHP was the high performing oxidant, and CHP was successful
only in the cases underlined in this paper.
Along these lines, we decided to complete our investigation by analyzing the case of 2,4-
diclorophenyl phenacyl sulfide 8a, in analogy with the results reported in entry 7. The analogous
sulfide, as shown in Table 1, is successfully oxidized with TBHP in the presence of the
titanium/hydrobenzoin complex and the same holds true also in this case.22 The enantiopure (>98%
ee) 2,4-dichlorophenyl phenacyl sulfoxide 8b was obtained with TBHP in good yields (Table 1,
entry 14). It was interesting to compare the behavior of CHP as the oxidant agent in this type of
reaction (Table 1, entry 15). The isolated yield of the sulfoxide 8b was good (83%) and similar to
the values obtained with TBHP, but the ee values were lower (57%, table 1, entry 15). The
combined values of yield and enantioselectivity points towards a reaction that is not so negative, as
it occurs in the CHP-reaction reported in entry 13.
Before proceeding with other reactions, a theoretical mechanism that takes into account the past and
the present experimental data about the enantioselective oxidation of aryl phenacyl sulfides with
hydroperoxides should be designed and calculated, in parallel with the satisfactory DFT
calculations that we performed with the different cases of aryl benzyl sulfides.
Absolute Configuration of sulfoxides. CD study
The absolute configuration of sulfoxides 1b,20, 24 4b,22 6b,28 and 7b28 was already established. The
configuration of sulfoxides 2b, 3b, 5b and 8b was expected to be (R) when (S, S)-hydrobenzoin was
employed as chiral ligand, in analogy with all previous cases. The configuration was confirmed by
means of electronic circular dichroism (ECD) spectra, according to the procedure reported in our
past work.25-28 The ECD spectra of these compounds, measured in acetonitrile, are shown in Figure
3. They all feature a positive band above 250 nm and one negative band with a shoulder, or two
negative bands, below 250 nm. The former band can in principle be employed to assign the absolute
11

ACCEPTED MANUSCRIPT
configuration using the so-called empirical Mislow’s rule.33-34 We have however previously
discouraged the use of this rule for perfluoro-substituted aryl sulfoxides,25, 27 for a twofold reason:
1) the nature of the diagnostic band can be very different from a sulfoxide-centered n-π* transition
which is covered by the rule;33-34 2) the presence of polar substituents may strongly alter the
conformational distribution of these compounds which, together with the fact that various
conformers are associated with very different ECD spectra, makes the overall spectra very
conformation-dependent. This latter aspect is entirely neglected in the application of empirical
rules. As an efficient alternative to assign absolute configurations, we found that DFT-based
calculations of the ECD spectra of fluorinated aryl sulfoxides can reproduce very accurately the
experimental spectra, provided that a proper functional and a continuum solvent models are
employed both in the geometry optimizations and in excited-state calculations.25-28 Computational
details are reported in the Computational Section. Briefly, after a conformational search with
molecular mechanics, all conformers were optimized with DFT at ωB97X-D/6-311+G(d,p)
including PCM solvent model for acetonitrile, then excited state calculations were run at the CAM-
B3LYP/def2-TZVP/PCM level. Boltzmann averaging of conformer spectra was accomplished using
internal energies computed at MP2/6-311+G(d,p)/PCM level, which led to more satisfying results
with respect to ωB97X-D energies. In fact, the agreement between experimental and calculated
spectra is excellent in all cases (Figure 3). A partial exception is compound 3b, for which the
splitting of the short-wavelength negative band was not reproduced by calculations; this is likely
due to the underestimation of the population of one specific conformer. In all cases, the similarity
factor (SF)35-36 between the experimental spectrum and that calculated for the (R)-enantiomer was >
0.9, leaving no doubt on the absolute configuration. This latter is definitely established as (R) for all
investigated compounds. It is interesting to notice that the quality of the calculated spectra is not
hampered at all by molecular flexibility. Compound 2b has only 3 possible conformers and its SF is
0.99; conversely, compound 8b is much more flexible and had 12 initial conformers, but the SF for
12

ACCEPTED MANUSCRIPT
the final calculated spectrum remains as high as 0.93. This finding confirms that DFT-based
calculations are completely adequate to reproduce ECD spectra of perfluorinated aromatic
sulfoxides, despite the presence of significant non-bonded interactions among the various polar
groups and of charge-transfer type electronic transitions.27
Figure 3. Comparison of experimental and calculated ECD spectra for (R)-sulfoxides 2b, 3b, 5b
and 8b. Experimental spectra were recorded in acetonitrile solutions with concentrations ≈1-3 mM
and quartz cells with 0.01 cm path length, and are not scaled for the enantiomeric excess.
Calculated spectra were obtained at CAM-B3LYP/def2-TZVP/PCM//ωB97X-D/6-311+G(d,p)/
PCM level (PCM for acetonitrile), as Boltzmann averages at 300K of spectra calculated for the
relevant conformers with populations estimated at MP2/6-311+G(d,p)/PCM level. Calculated
spectra plotted as sums of Gaussians with exponential band-width σ, wavelength shift and scaling
factor listed on each spectrum.
Conclusions
Notwithstanding the many different protocols that were reported in almost half a century for the
synthesis of enantiopure sulfoxides, nowadays asymmetric oxidation of prochiral sulfides are still
preferred in the industrial productions of bioactive molecules, and hardly new research based upon
13

ACCEPTED MANUSCRIPT
expensive reagents will change this situation.
In this context, our oxidation protocol based upon hydroperoxides in the presence of a titanium/(S,
S)-hydrobenzoin complex can become, in our opinion, an efficient and concrete alternative, because
yields a large number of enantiopure sulfoxides with a simple procedure based upon inexpensive
reagents.
As far as the present investigation is concerned, the mutually exclusive behavior of CHP and TBHP
when employed as the oxidant agents of our asymmetric oxidation of sulfides, is confirmed. TBHP
is almost always the successful choice, whereas CHP is usually a modest alternative. However, CHP
can effectively substitute TBHP in those cases of a lower performance connected to the asymmetric
oxidation of some perfluorinated sulfides, thus enlarging the already wide set of sulfoxides that can
be obtained in high enantiomeric purity with this asymmetric oxidation.
The most intriguing aspect of the present investigation remains the analogy of behavior between
fluorinated and not fluorinated aryl benzyl and aryl phenacyl sulfides, substrates that are clearly
different, but share a common behavior when subjected to the hydroperoxide oxidation in the
presence of the titanium/ hydrobenzoin catalyst.
Experimental Section
Chemicals were used as received. Elemental analyses were performed on a CHNS-O Elemental
Analyzer. High resolution Mass Spectra were determined with a HPLC-QTOF spectrometer via
direct infusion of the samples, using methanol as the elution solvent. NMR spectra were recorded
1
on a H-500 MHz, 13C-125 MHz and 19F-470 MHz spectrometer. Only absolute values of the
coupling constants were reported. As far as the 19F spectra are concerned, chemical shifts and
coupling constant patterns were measured to be in accord with the reported values.37 As for the 13C-
1
spectra of pentafluorophenyl bearing compounds are concerned, the JCF coupling constant is
always reported to discriminate among overlapping signals. ECD spectra were recorded using a
JASCO J-715 spectrometer in acetonitrile using a quartz cell with 0.01 cm path length.
14

ACCEPTED MANUSCRIPT
Sulfides 1a-9a were synthesized on a 6 mmol scale by adding the commercially available thiols to
an ethanol solution of the corresponding benzyl or phenacyl bromides in the presence of potassium
carbonate as the basic reagent.28, 38 The mixture was reacted for 2 hours at room temperature. Usual
work up gave a crude mixture that was purified by distillation. Usually, distilled sulfides solidify on
standing.
Sulfides 1a,20 4a,22 6a,28 7a,28 8a39 were already reported.
2,3,4,5,6-Pentafluorobenzyl 2,3,5,6-tetrafluorophenyl sulfide (2a). Kugelrohr oven temp 90-95
°C, p=0.1 torr. Mp 69-70 °C. 1H-NMR (500 MHz, CDCl3) 7.13 (tt, J= 9.5 Hz, J= 7.3 Hz, 1 H), 4.12
(t, J= 1.2 Hz, 2 H). 13C-NMR (125 MHz, CDCl3) 147.2 (dddd, 1JCF = 247 Hz, J= 13.1 Hz, J= 4.2 Hz,
J= 2.1 Hz), 145.8 (dddd, 1JCF = 251 Hz, J= 15.3 Hz, J= 10.4 Hz, J= 4.8 Hz), 144.9 (dddm, 1JCF = 249
1
1
Hz, J= 15.3 Hz, J= 7.6 Hz), 140.9 (dm, JCF = 255 Hz), 137.4 (dm, JCF = 253 Hz), 112.9 (tm, J=
20.4 Hz), 111.5 (tm, J= 17.3 Hz), 107.7 (tm, J= 22.9 Hz), 25.9. 19F-NMR (470 MHz, CDCl3) -132.9
_-133.0 (m, 2 F), -137.3 _-137.4 (m, 2 F), -143.2 _-143.3 (m, 2 F), -153.8 (t, J= 21.6 Hz, 2 F), -
161.2 _-161.4 (m, 2 F). Anal. Calcd for C13H3F9S : C 43.11; H 0.83. Found C 43.15; H 0.68.
2,3,4,5,6-Pentachlorophenyl 2,3,4,5,6-pentafluorobenzyl sulfide (3a). Kugelrohr oven temp 162-
168 °C, p=0.1 torr. Mp 143-144 °C. 1H-NMR (500 MHz, CDCl3) 4.15 (t, J= 1.4 Hz, 2 H). 13C-NMR
1
(125 MHz, CDCl3) 145.0 (dm, JCF = 246 Hz), 139.3, 140.8 (dm, 1JCF = 255 Hz), 137.4 (dm, 1JCF =
252 Hz), 135.6, 132.4, 132.3, 111.5-111.1 (m), 26.1. 19F-NMR (470 MHz, CDCl3) -143.1 (dd,
J=21.6 Hz, J=8.3 Hz, 2 F), -154.0 (dd, J=21.6 Hz, J=19.9 Hz, 1 F), 161.4 (ddd, J=21.6 Hz, J=19.9
Hz, J=8.3 Hz, 2 F). Anal. Calcd for C13H2Cl5F5S : C 33.76; H 0.44. Found C 33.78; H 0.59.
2,6-Dichlorophenyl 2,3,4,5,6-pentafluorobenzyl sulfide (5a). Kugelröhr oven temp 130-135 °C,
p=0.1 mbar. 1H-NMR (500 MHz, CDCl3) 7.37 (d, J= 7.8 Hz, 2 H), 7.22 (t, J= 7.8 Hz, 1 H), 4.11 (t,
1
1
J= 1.2 Hz, 2 H). 13C-NMR (125 MHz, CDCl3) 145.0 (dm, JCF = 249 Hz), 142.2, 140.5 (dm, JCF =
1
254 Hz), 137.2 (dm, JCF = 250 Hz), 131.0, 130.4, 128.6, 112.1-111.7 (m), 25.7. 19F-NMR (470
MHz, CDCl3) -143.4 (dd, J=21.6 Hz, J=8.3 Hz, 2 F), -155.3 (dd, J=21.6 Hz, J=19.9 Hz, 1 F), 162.4
15

ACCEPTED MANUSCRIPT
(ddd, J=21.6 Hz, J=19.9 Hz, J=8.3 Hz, 2 F) Anal. Calcd for C13H5Cl2F5S : C 43.48; H 1.40. Found
C 43.32; H 1.37.
Racemic sulfoxides 1b-8b were synthesized by standard mCPBA oxidation of the corresponding
sulfides and were used as standard in the HPLC separation of enantiomers.
Enantioselective oxidation of sulfide (1a)-(8a) with hydroperoxides in the presence of a
titanium/(S, S)-hydrobenzoin catalyst.
The enantioselective oxidation reactions in which TBHP was used as the oxidant follow the
procedure already reported. When CHP was used as the oxidant, the following procedure is
representative. A solution of Ti(O-i-Pr)4 99.999% (14 mg, 0.05 mmol) in 4 mL of n-hexane was
added to a solution of (S, S)-hydrobenzoin (21 mg, 0.1 mmol) in 8 mL of n-hexane under a nitrogen
atmosphere. The mixture was stirred for 1 hour at room temperature. A solution of the
corresponding sulfide (1 mmol) in 8 mL of n-hexane was then added and the mixture was stirred
for 30 minutes. After this time, 0.2 mL of a commercial solution of cumene hydroperoxide 80% was
added and the stirring was continued for one day. During this time, the desired sulfoxide
precipitated as a white solid. Further batch of sulfoxide could be obtained by purifying the mother
liquor with chromatography. For sulfoxides 1b and 3b, the separation is facilitated if the residual
cumyl alcohol was first distilled with a low-pressure kugelrohr apparatus.
Sulfoxides 1b,20, 24 4b,22 6b28 and 7b28 were already reported.
2,3,4,5,6-Pentafluorobenzyl 2,3,5,6-tetrafluorophenyl sulfoxide (2b). Mp 112-114 °C (ethanol).
25
1
[α]D = + 41.3 (c= 0.9, CHCl3). H-NMR (500 MHz, CDCl3) 7.33-7.26 (m, 1 H), 4.72-4.65 (m, 2
H). 13C-NMR (125 MHz, CDCl3) 145.9 (dm, 1JCF = 254 Hz), 145.7 (dm, 1JCF = 252 Hz), 144.7 (dm,
1JCF = 256 Hz), 141.8 (dm, 1JCF = 252 Hz), 137.7 (dm, 1JCF = 255 Hz), 122.0 (tm, J= 16.0 Hz), 110.6
19
(tm, J= 22.2 Hz), 103.6 (tm, J= 17.8 Hz), 47.2. F-NMR (470 MHz, CDCl3) -135.0 _-135.2 (m, 2
F), -139.6 _-139.7 (m, 2 F), -140.4 _-140.6 (m, 2 F), -150.2 (dd, J= 21.6 Hz, J= 19.9 Hz, 1 F), -
159.8 _-160.0 (m, 2 F). LCMS-QTOF m/z: [M+H]+ calculated for C13H4F9OS 378.9834; found
16

ACCEPTED MANUSCRIPT
378.9822. The ee value was measured by HPLC (Column: Chiralcel OD-H. Eluent: n-hexane/i-
propanol 7:3; flow rate 0.5 ml/min; tR=16.7; tS = 20.6; separation factor α =1.38).
2,3,4,5,6-Pentachlorophenyl 2,3,4,5,6-pentafluorobenzyl sulfoxide (3b). mp 172-174 °C (n-
25
hexane/acetone 4:1). [α]D = +109.5 (c= 0.5, CHCl3).. 1H-NMR (500 MHz, CDCl3) 4.74 (dm, J=
12.9 Hz, 1 H), 4.60 (dm, J= 12.9 Hz, 1 H). 13C-NMR (125 MHz, CDCl3) 146.0 (dm, 1JCF = 247 Hz),
141.8 (dm, 1JCF = 257 Hz), 137.7 (dm, 1JCF = 254 Hz), 137.5, 137.1, 134.0, 132.7, 104.4-104.1 (m),
19
44.8. F-NMR (470 MHz, CDCl3) -140.3 (dd, J=21.6 Hz, J=6.6 Hz, 2 F), -150.8 (dd, J= 21.6 Hz,
J= 19.9 Hz, 1 F), -160.1 _-160.3 (m, 2 F). LCMS-QTOF m/z: [M+Na]+ calculated for
C13H2Cl5F5OS-Na 498.8087; found 498.8070 (multiplet due to the Cl(37) atom at 500.8047,
502.8013 and 504.7985). The ee value was measured by HPLC (Column: Chiralcel OD-H. Eluent:
hexane/i-propanol 7:3; flow rate 0.5 ml/min; tR=15.5; tS = 20.7; separation factor α =1.55).
(R)-2,6-Dichlorophenyl 2,3,4,5,6-pentafluorobenzyl sulfoxide (5b). mp 160-162 °C (tert-
25
1
BuOMe/MeCN 1:1. [α]D = + 38.8 (c=0.3, CHCl3) for a sulfoxide having a 89% ee value. H-
NMR (500 MHz, CDCl3) 7.38-7.34 (m, 3 H), 4.75 (d, J= 12.7 Hz, 1 H), 4.64 (d, J= 12.7 Hz, 1 H).
13C-NMR (125 MHz, CDCl3) 145.8 (dm, 1JCF = 250 Hz), 141.5 (dm, 1JCF = 256 Hz), 137.4 (dm, 1JCF
19
= 250 Hz), 135.9, 135.3, 133.1, 130.3, 104.7-104.3 (m), 44.6. F-NMR (470 MHz, CDCl3) -140.6
(dd, J=21.6 Hz, J=8.3 Hz, 2 F), -151.7_-151.9 (m, 1 F), -160.8 _-161.0 (m, 2 F). Anal. Calcd for
C13H5Cl2F5OS : C 41.62; H 1.34. Found C 41.83; H 1.32. The ee value was measured by HPLC
(Column: Chiralcel OD-H. Eluent: hexane/i-propanol 7:3; flow rate 0.5 ml/min;tR=11.5; tS = 14.7;
separation factor α =1.57).
2,4-Dichlorophenyl phenacyl sulfoxide (8b). Mp 135-137 °C (i-Pr2O/MeCN 4:1). [α]D25 = + 400.5
1
(c= 1.1, CHCl3). H-NMR (500 MHz, CDCl3) 7.93-7.89 (m, 2 H), 7.78 (d, J= 8.3 Hz, 1 H), 7.63-
7.59 (m, 1 H), 7.49-7.45 (m, 3 H), 7.43 (d, J= 1.5 Hz, 1 H), 4.60 (d, J=14.2 Hz, 1 H), 4.31 (d,
J=14.2 Hz, 1 H). 13C-NMR (125 MHz, CDCl3) 190.6, 139.6, 138.2, 136.2, 134.2, 130.7, 129.7,
128.9, 128.5, 127.5, 61.9. Anal. Calcd for C14H10Cl2O2S: C 53.69; H 3.22. Found C 53.56; H 3.21.
17

ACCEPTED MANUSCRIPT
The ee value was measured by HPLC (Column: Chiralpak IA. Eluent: hexane/i-propanol 7:3; flow
rate 0.5 ml/min; tS=20.0; tR = 24.5; separation factor α =1.32).
Computational Section
Molecular mechanics and preliminary DFT calculations were run with Spartan’16 (Wavefunction,
Irvine CA), with standard parameters and convergence criteria. DFT and TDDFT calculations were
run with Gaussian16,40 with default grids and convergence criteria.
Conformational searches were run with the Monte Carlo algorithm implemented in Spartan’16
using the Merck Molecular Force Field (MMFF). Preliminary DFT calculations were run first at the
ωB97X-D/6-31G(d) level, and then at the ωB97X-D/6-311+G(d,p) level in vacuo. All selected
structures were then re-optimized at the ωB97X-D/6-311+G(d,p) level with PCM solvent model for
CH3CN. Single-point calculations were then run at the MP2/6-311+G(d,p) level with PCM solvent
model for CH3CN to estimate internal energies.
ECD calculations were run at TDDFT level with the CAM-B3LYP functional and the def2-TZVP
basis set including the PCM model for CH3CN. Average ECD spectra were computed by weighting
the spectra of individual conformers using Boltzmann factors at 300 K estimated from MP2 internal
energies. All conformers having population ≥0.1% at 298K were taken into consideration. The
final spectra were generated using the program SpecDis ver. 1.70. The plotting parameters were
decided on a best-fitting basis and are reported in each Figure. Similarity factors (SF) were also
estimated using SpecDis.
Acknowledgements
Rebizi Nadjib is thanked for his help in CD measurements. Italian Ministero dell'Istruzione,
dell'Università e della Ricerca (MIUR) is gratefully acknowledged for support. The authors declare
no conflict of interest.
18

ACCEPTED MANUSCRIPT
References
1. Han J, Soloshonok VA, Klika KD, Drabowicz J, Wzorek A, Chem. Soc. Rev. 2018; 47: 1307.
2. Wojaczyńska E, Wojaczyński J, Chem. Rev. 2010; 110: 4303.
3. O' Mahony GE, Ford A, Maguire AR, J. Sulfur Chem. 2013; 34: 301.
4. Otocka S, Kwiatkowska M, Madalińska L, Kiełbasiński P, Chem Rev. 2017; 117: 4147.
Pellissier, H, Tetrahedron 2006; 62: 5559.
5. Trost BM, Rao, M, Angew. Chem. Int. Ed. Engl. 2015; 54: 5026.
6. Bentley R, Chem. Soc. Rev. 2005; 34: 609.
7. Cotton H, Elebring T, Larsson M, Li L, Sorensen H, von Unge S, Tetrahedron: Asymmetry
2000; 11: 3819.
8. Seenivasaperumal M, Federsel HJ, Szabó KJ, Adv. Synth. Catal. 2009; 351: 903.
9. Courvoisier, L.; Duret, G.; Graf, S.; Prat-Lacondemine, L; Rebiere, F.; Sabourault, N.
Process for Enantioselective Synthesis of Single Enantiomers of Modafinil by Asymmetric
Oxidation, US Patent US2013066111 – 2013-03-14.
10. Jia T, Zhang M, McCollom SP, Bellomo A, Montel S, Mao J, Dreher SD, Welch CJ,
Regalado EL, Williamson RT, Manor BC, Tomson NC, Walsh PJ, J. Amer. Chem. Soc. 2017;
139: 8337.
11. Wang L, Chen M, Zhang P, Li W, Zang J, J. Amer. Chem. Soc. 2018; 140: 3467. Wu C,
Berritt S, Liang X, Walsh PJ, Org. Lett. 2019; 21: 960.
12. Zong L, Tan CH, Acc. Chem. Res. 2017; 50: 842. Gelat F, Jayashankaran J, Lohier JF,
Gaumont AC, Perrio S, Org. Lett. 2011; 13: 3170.
13. Yu H, Li Z, Bolm C, Org. Lett. 2018; 20: 7104. Wang L, Chen M, Zhang J, Org. Chem.
Front. 2019; 6: 32.
14. Capozzi MAM, Cardellicchio C, Naso F, Eur. J. Org. Chem. 2004: 1855.
19

ACCEPTED MANUSCRIPT
15. Naso F, Cardellicchio C, Affortunato F, Capozzi MAM, Tetrahedron: Asymmetry 2006; 17:
3226.
16. Cardellicchio C, Hassan Omar O, Naso F, Capozzi MAM, Capitelli F, Bertolasi V,
Tetrahedron: Asymmetry 2006; 17: 223.
17. Capozzi MAM, Cardellicchio C, Naso F, Rosito V, J. Org. Chem. 2002; 67: 7289.
18. Naso F, Capozzi MAM, Bottoni A, Calvaresi M, Bertolasi V, Capitelli F, Cardellicchio C,
Chem. Eur. J. 2009; 15: 13417.
19. Capozzi MAM, Centrone C, Fracchiolla G, Naso F, Cardellicchio C, Eur. J. Org. Chem.
2011: 4327.
20. Capozzi MAM, Capitelli F, Bottoni A, Calvaresi M, Cardellicchio C, ChemCatChem 2013;
5: 210.
21. Capozzi MAM, Fracchiolla G, Cardellicchio C, J. Sulfur Chem. 2013; 34: 646.
22. Capozzi MAM, Capitelli F, Cardellicchio C, Cryst. Growth Des. 2014; 14: 5442.
23. Capozzi MAM, Terraneo G, Cavallo G, Cardellicchio C, Tetrahedron 2015; 71: 4810.
24. Capozzi MAM, Bottoni A, Calvaresi M, Cardellicchio C, Tetrahedron 2018; 74: 2041.
25. Pescitelli G, Di Pietro S, Cardellicchio C, Capozzi MAM, Di Bari L, J. Org. Chem. 2010;
75: 1143.
26. Padula D, Di Pietro S, Capozzi MAM, Cardellicchio C, Pescitelli G, Chirality 2014; 26:
462.
27. Berardozzi R, Guido CA, Capozzi MAM, Cardellicchio C, Di Bari L, Pescitelli G, Eur. J.
Org. Chem. 2015; 5554.
28. Górecki M, Capozzi MAM, Albano G, Cardellicchio C, Di Bari L, Pescitelli G, Chirality,
2018; 30: 29.
29. Okano K, Tetrahedron 2011; 67, 2483.
20

ACCEPTED MANUSCRIPT
30. Yamamoto K, Ando H, Shuetake T, Chikamatsu H, J. Chem. Soc. Chem. Commun, 1989:
754. Donnoli MI, Superchi S, Rosini C, J. Org. Chem. 1998; 63: 9392.
31. Desiraju GR, Vittal JJ, Ramanan A, Crystal Engineering. A Textbook. World Scientific:
Singapore, 2011. Nishio M, Umezawa Y, Fantini J, Weiss MS, Chakrabarti P, Phys. Chem.
Chem. Phys. 2014; 16: 12648. Nishio M, Phys. Chem. Chem. Phys. 2011; 13: 13873.
Cardellicchio C, Capozzi MAM, Alvarez-Larena A, Piniella JF, Capitelli F, CrystEngComm,
2012; 14: 3972.
32. Salonen LM; Ellermann M, Diederich F, Angew Chem. Int. Ed. Engl. 2011; 50:4808.
33. Mislow K, Green MM, Laur P, Melillo JT, Simmons T, Ternay Jr AL, J. Am. Chem. Soc.
1965, 87, 1958.
34. Mislow K, Ternay Jr A L, Melillo JT, J. Am. Chem. Soc. 1963, 85, 2329.
35. Bruhn, T.; Schaumlöffel, A.; Hemberger, Y.; Bringmann, G. Chirality 2013, 25, 243.
36. Bruhn, T.; Schaumlöffel, A.; Hemberger, Y.; Pescitelli, G.: Berlin, Germany, 2017,
https://specdis-software.jimdo.com/.
37. Hogben MG, Graham WAG, J. Amer. Chem. Soc. 1969; 91: 283.
38. Seahadri R, Pegg WJ, Israel M, J. Org. Chem. 1981; 46: 2596.
39. Menon NK, Nargund KS, Indian J. Chem. 1967; 5: 478.
40. Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR, Scalmani G,
Barone V, Petersson GA, Nakatsuji H, Li X, Caricato M, Marenich AV, Bloino J, Janesko
BG, Gomperts R, Mennucci B, Hratchian HP, Ortiz JV, Izmaylov AF, Sonnemberg JL,
Williams-Young D, Ding F, Lipparini F, Egidi F, Goings J, Peng B, Petrone, A, Henderson T,
Ranasinghe D, Zakrzewski VG, Gao J, Rega N, Zheng G, Liang W, Hada M, Ehara M,
Toyota K, Fukuda R, Hasegawa J, Ishida M, Nakajima T, Honda Y, Kitao O, Nakai H,
Vreven T, Throssell K, Montgomery Jr JA, Peralta JE, Ogliaro F, Bearpark M, Heyd JJ,
Brothers E, Kudin KN, Staroverov VN, Keith TA, Kobayashi R, Normand J, Raghavachari
21

ACCEPTED MANUSCRIPT
K, Rendell AP, Burant JC, Iyengar SS, Tomasi J, Cossi M, Millam JM, Klene M, Adamo C,
Cammi R, Ochterski JW, Martin RL, Morokuma K, Farkas O, Foresman JB, Fox DJ,
Gaussian, Inc.: Wallingford CT, 2016.
22