Asymmetric sulfoxidation of thioanisole by helical Ti(IV) salan catalysts

Article abstract of DOI:10.1246/cl.2012.974

A new helix-directing salan ligand (R,R)-2 with the (1R,2R)- diaminocyclohexyl backbone and benz[a]anthryl sidearms was synthesized by borohydride reduction of the corresponding salen ligand. Formylation and reduction of (R,R)-2 yielded the N-Me counterpart, (R,R)-3. Complexation of these flexible ligands to TiCl₄ produced the [TiCl ₂(salan)] complexes (R,R)-4 and (R,R)-5, which were characterized by NMR, IR, and HRMS techniques. The complexes were tested as catalysts for the asymmetric sulfoxidation of thioanisole with cumene hydroperoxide and hydrogen peroxide as the oxidants. Modest selectivity was observed, with the N-H salan complex showing somewhat better chiral induction.

Full text of DOI:10.1246/cl.2012.974

doi:10.1246/cl.2012.974
9
74
Published on the web September 8, 2012
Asymmetric Sulfoxidation of Thioanisole by Helical Ti(IV) Salan Catalysts
Sanmitra Barman, Smita Patil, and Christopher J. Levy*
Department of Chemistry, Kansas State University, 213 CBC Building, Manhattan, KS, 66506, USA
(
Received May 29, 2012; CL-120457; E-mail: clevy@ksu.edu)
A new helix-directing salan ligand (R,R)-2 with the (1R,2R)-
diaminocyclohexyl backbone and benz[a]anthryl sidearms was
synthesized by borohydride reduction of the corresponding salen
ligand. Formylation and reduction of (R,R)-2 yielded the N-Me
counterpart, (R,R)-3. Complexation of these flexible ligands to
TiCl4 produced the [TiCl2(salan)] complexes (R,R)-4 and (R,R)-
5
, which were characterized by NMR, IR, and HRMS
techniques. The complexes were tested as catalysts for the
asymmetric sulfoxidation of thioanisole with cumene hydro-
peroxide and hydrogen peroxide as the oxidants. Modest
selectivity was observed, with the N-H salan complex showing
somewhat better chiral induction.
Chiral sulfoxides are important chiral auxiliaries in organic
1
synthesis and many are in direct use as pharmaceuticals.
Prochiral sulfides are challenging targets for catalytic asymmet-
ric sulfoxidations and some recent progress has been achieved
using metal salan complexes.2 Salan ligands are reduced
versions of salens, where the imine groups have been hydro-
genated to give secondary amines. The amine nitrogens can
subsequently be alkylated to give salans with tertiary amine
Scheme 1. Synthesis of the ligands and Ti(IV) complexes: (R,R)-2­
(R,R)-5. i: a) 37% aq. formaldehyde, CH3CO2H b) NaBH4; c) NaOH.
3
13
donors. When compared to salens, the absence of the C=N
bond in salan ligands leads to more stable complexes that are not
4
susceptible to hydrolysis. The reduced conjugation and addition
3
of sp centers imparts added flexibility to the ligand framework
and octahedral complexes can readily adopt inherently chiral
¡
-cis- or ¢-cis-geometries which have stepped or twisted struc-
2
tures. The majority of mononuclear octahedral salan complexes
characterized have had the ¡-cis ( fac­fac)-coordination mode,
while bridged di- and tetranuclear complexes generally adopt the
5
¢
-cis ( fac­mer) mode. Equilibration between ¡-cis- and ¢-cis-
6
isomers has been observed in some cases. By contrast, more
rigid salen ligands normally favor the planar trans-coordinated
octahedral complexes.7 The helical chirality associated with
twisted metal salan complexes may have an advantage in chiral
Figure 1. Exchange between diastereomeric ¦ and ª helical ¡-cis-
coordinated ligands.
1
induction compared to their metal salen counterparts, but
they have seen considerably less study as oxidation catalysts,
partially due to the challenges of controlling the stereochemis-
try. In this communication we report the synthesis of two new
salan ligands with extended benz[a]anthryl sidearms and explore
the catalytic activity of Ti(IV) complexes for the asymmetric
oxidation of thioanisole.
+
spectroscopy. The [M + H] ion peak with the expected isotope
8
pattern was seen in each case. Complexes (R,R)-4 and (R,R)-5
were synthesized in 71% and 73% yields, respectively, as dark
purple solids, by treating a solution of the ligands in methylene
chloride with TiCl4 at room temperature (Scheme 1).
Synthesis and characterization of the salen ligand (R,R)-1,
which has benz[a]anthryl sidearms, was previously reported
by our group. Herein, we detail the synthesis of two versions
The two titanium complexes showed comparatively broad
1
H NMR spectra suggesting fluxional behavior, such as between
9
10
diastereomeric ¡-cis-conformers (Figure 1). The ¹30 °C spec-
(
N-H and N-Me) of salan ligands produced from this starting
compound. Reduction of (R,R)-1 with NaBH4 in THF gave
R,R)-2 in 79% yield (Scheme 1). The N-Me version (R,R)-3 was
trum (CD2Cl2) of complex (R,R)-4 sharpens and is consistent
with a 60:40 mixture of C -symmetric diastereomers based on
2
(
integrations. For (R,R)-4 the ¹30 °C spectrum indicates a 70:30
mixture. ESI-MS spectrometry was carried out in alcoholic
solvents (MeOH and EtOH), resulting in Ti­Cl bond hydrolysis
synthesized from (R,R)-2 in 77% yield by formylation followed
by reduction with NaBH (Scheme 1). Both (R,R)-2 and (R,R)-3
were characterized by NMR, high-resolution ESI-MS, and IR
4
+
and the observance of monoalkoxo cations: [TiL(OR)]
Chem. Lett. 2012, 41, 974­975
© 2012 The Chemical Society of Japan
www.csj.jp/journals/chem-lett/

9
75
Table 1. Sulfoxidation of thioanisole catalyzed by (R,R)-4 and
(
R,R)-5a
Catalyst Oxidant T/°C conv/% SO/%b ee/%c SO2/%d
4
4
4
5
5
5
H2O2
H O
CHP
H2O2
H2O2
CHP
20
0
0
20
0
0
90
93
92
91
93
91
90
85
95
93
89
98
39
43
31
25
30
20
10
15
5
7
11
2
2
2
Scheme 2. Asymmetric sulfoxidation of thioanisole with titanium
catalysts (major enantiomer R).
(R = Me and Et). The observed isotope patterns match well with
a1% cat., 16 h in CH2Cl2. bSO: PhSOMe. Enantiomeric excess
measured by HPLC with DAICEL OD-H column. SO2: PhSO2Me.
c
the theoretical distributions.
d
We have carried out preliminary studies utilizing these
helical titanium salan complexes as asymmetric catalysts for the
oxidation of thioanisole with hydrogen peroxide or cumene
hydroperoxide (Scheme 2). Inital catalytic runs with (R,R)-4 and
30% aqueous hydrogen peroxide indicated that the enantio-
selectivity was highest in methylene chloride (biphasic system),
and this was selected as the solvent for subsequent reactions.
Aliquots taken at different reaction times revealed more rapid
conversion at the beginning of the reaction, with 70%
conversion after 5 h. Conversion to sulfoxide increased up to
ca. 16 h. The yield decreased at longer reaction times due to
conversion of sulfoxide to sulfone and the conversion of catalyst
to oxo-bridged dimers11 with different activity/selectivity. The
Figure 2. N­H£O hydrogen bonding in the peroxo complex in the
catalytic cycle of (R,R)-4.
observed %ee of the product decreased only slightly (1­2%)
over the course of the reactions, suggesting that the stereo-
chemical course of the reactions are not affected significantly
by this process. The oxidant:substrate ratio was varied, and
the effect on product formation, overoxidation, and %ee was
determined. There was a distinct increase in oxidation to sulfone
and decrease in %ee when more than 1.5 equivalents of oxidant
were used. We chose to use the more conservative 1.1
equivalents to minimize overoxidation while maintaining high
conversion. The temperature of the reaction had a modest effect
on enantioselectivity, with higher %ee values seen at 0 °C
compared to 20 °C (43% vs. 39% with (R,R)-4 as catalyst).
The results of the optimized catalytic runs are presented in
Table 1. In all cases the major product was R-methyl phenyl
sulfoxide, consistent with other studies that show the absolute
configuration of the sulfoxide to be the same as that of the
benz[a]anthryl sidearms do not appear to have been sufficient for
high diastereoselection in coordination to the metal. Currently,
we are working on incorporating more twisted chiral backbones
that can prevent fluxionality and provide well-defined salan
catalysts for asymmetric sulfoxidation.
This work was supported by the National Science Founda-
tion (No. CHE-0349258) and Kansas State University.
References and Notes
1
2
3
4
K. P. Bryliakov, E. P. Talsi, Eur. J. Org. Chem. 2008, 3369.
K. Matsumoto, B. Saito, T. Katsuki, Chem. Commun. 2007, 3619.
J. Balsells, P. J. Carroll, P. J. Walsh, Inorg. Chem. 2001, 40, 5568.
P. Adão, F. Avecilla, M. Bonchio, M. Carraro, J. C. Pessoa, I. Correia,
Eur. J. Inorg. Chem. 2010, 5568.
1
5
a) K. Matsumoto, Y. Sawada, T. Katsuki, Pure Appl. Chem. 2008, 80,
cyclohexyl backbone for titanium salan complexes. Somewhat
1071. b) D. Xiong, M. Wu, S. Wang, F. Li, C. Xia, W. Sun, Tetrahedron:
better enantioselection was seen with H2O2 for both catalysts,
while conversions were somewhat better with cumene hydro-
peroxide, consistent with what has been observed in other
Asymmetry 2010, 21, 374. c) C. M. Manna, E. Y. Tshuva, Dalton Trans.
2
010, 39, 1182. d) C. M. Manna, G. Armony, E. Y. Tshuva, Chem.®Eur.
J. 2011, 17, 14094. e) C. M. Manna, G. Armony, E. Y. Tshuva, Inorg.
Chem. 2011, 50, 10284.
4
studies with these oxidants. The sulfoxide yields for (R,R)-4
6
7
a) D. A. Atwood, Coord. Chem. Rev. 1997, 165, 267. b) B. Saito, T.
Katsuki, Tetrahedron Lett. 2001, 42, 3873.
M. Kojima, H. Taguchi, M. Tsuchimoto, K. Nakajima, Coord. Chem.
Rev. 2003, 237, 183.
and (R,R)-5 are quite comparable, but the enantioselection is
notably higher for (R,R)-4. Better selectivity for titanium salan
oxidation catalysts with N-H groups has been previously
explained by Katsuki et al.12 The key feature is hydrogen
8
9
M. L. Kuznetsov, J. C. Pessoa, Dalton Trans. 2009, 5460.
a) A. V. Wiznycia, J. Desper, C. J. Levy, Dalton Trans. 2007, 1520.
b) A. V. Wiznycia, J. Desper, C. J. Levy, Inorg. Chem. 2006, 45, 10034.
The ¦/ª equilibrium indicated in Figure 1 is only one of possible
explanation for the fluxional behavior. trans- and cis-¢-Complexes may
also be involved, but we are unable to make a definitive determination at
this time.
bonding between the N-H group and an oxygen from a peroxo
ligand in a ¢-cis-intermediate. In sulfoxidation this likely assists
in the transfer of a peroxo oxygen to the sulfide substrate
1
0
(Figure 2).
In conclusion, we were able to synthesize and characterize
11
Y. Sawada, K. Matsumoto, S. Kondo, H. Watanabe, T. Ozawa, K.
Suzuki, B. Saito, T. Katsuki, Angew. Chem., Int. Ed. 2006, 45, 3478.
helical N-H and N-Me salan ligands. The titanium complexes
formed can interconvert between diastereomeric forms at room
temperature. The resulting distribution is likely a primary reason
for the relatively low enantioselection seen during sulfoxida-
tions, since the asymmetric environment varies significantly
with helical diastereomers. Steric interactions of the extended
12 S. Kondo, K. Saruhashi, K. Seki, K. Matsubara, K. Miyaji, T. Kubo, K.
Matsumoto, T. Katsuki, Angew. Chem., Int. Ed. 2008, 47, 10195.
13
Experimental details, NMR and IR spectra, and MS data for the
complexes is provided as Supporting Information which is available
electronically on the CSJ-Journal Web site, http://www.csj.jp/journals/
chem-lett/index.html.
Chem. Lett. 2012, 41, 974­975
© 2012 The Chemical Society of Japan
www.csj.jp/journals/chem-lett/