C-C bond-forming desulfurizations of sulfoximines

Article abstract of DOI:10.1021/ol8015995

(Figure Presented) Highly substituted, enantiomerically pure azaheterocyclic ring systems play an important role in medicinal chemistry as potential peptide mimetics. Metalated 2-alkenyl sulfoximines offer an efficient entry to this class of compounds. In this paper, we describe a new means to remove the sulfonimidoyl auxiliary with concomitant formation of a C-C double bond.

Full text of DOI:10.1021/ol8015995

ORGANIC
LETTERS
2008
Vol. 10, No. 18
4081-4084
C-C Bond-Forming Desulfurizations of
Sulfoximines
M. Reggelin,*,† S. Slavik,‡ and P. Bu¨hle§
Clemens-Scho¨pf-Insitut fu¨r Organische Chemie and Biochemie, Petersenstrasse 22,
Technische UniVersita¨t, 64287 Darmstadt, Germany, Merck KGaA,
Frankfurter Str. 250, 64293 Darmstadt, Germany, and General Motors Europe
Engineering, Peter Sander Strasse 43, 55252 Mainz-Kastel, Germany
re@punk.oc.chemie.tu-darmstadt.de
Received July 14, 2008
ABSTRACT
Highly substituted, enantiomerically pure azaheterocyclic ring systems play an important role in medicinal chemistry as potential peptide
mimetics. Metalated 2-alkenyl sulfoximines offer an efficient entry to this class of compounds. In this paper, we describe a new means to
remove the sulfonimidoyl auxiliary with concomitant formation of a C-C double bond.
Since the days of their discovery,1 sulfoximines have evolved
into an important class of compounds.2 This is true in
Scheme 1. Overview
particular for their application as chiral ligands3 as well as
chiral auxiliaries.2 Enantiomerically pure 2-alkenyl sulfox-
imines 2 derived from cyclic sulfonimidates4 such as 1 have
been introduced by us in 1994 as efficient solutions for
asymmetric d3-synthons (Scheme 1).5 Titanated sulfoximines
derived from 2 undergo γ-hydroxyalkylation reactions with
† Technische Universita¨t Darmstadt.
‡ Present affiliation: General Motors Europe Engineering.
§ Present affiliation: Merck KGaA.
(1) Bentley, H. R.; McDermott, E. E.; Pace, J.; Whitehead, J. K.; Moran,
T. Nature 1949, 163, 675–676.
(2) Gais, H. J. Heteroatom. Chem. 2007, 18, 472–481. Okamura, H.;
Bolm, C. Chem. Lett. 2004, 33, 482–487. Reggelin, M.; Zur, C. Synthesis
2000, 1–64.
(3) Lu, S.-M.; Bolm, C. AdV. Synth. Catal. 2008, 350, 1101–1105.
Lemasson, F.; Gais, H. J.; Raabe, G. Tetrahedron Lett. 2007, 48, 8752–
8756. Spohr, V.; Kaiser, J. P.; Reggelin, M. Tetrahedron: Asymmetry 2006,
17, 500–503. Reetz, M. T.; Bondarev, O. G.; Gais, H. J.; Bolm, C.
Tetrahedron Lett. 2005, 46, 5643–5646. Reggelin, M.; Weinberger, H.;
Spohr, V. AdV. Synth. Catal. 2004, 346, 1295–1306. Bolm, C.; Verrucci,
M.; Simic, O.; Cozzi Pier, G.; Raabe, G.; Okamura, H. Chem Commun.
2003, 2826–2827. Bolm, C.; Martin, M.; Simic, O.; Verrucci, M. Org. Lett.
2003, 5, 427–429. Bolm, C.; Simic, O.; Martin, M. Synlett 2001, 1878–
1880.
R- or ꢀ-heteroatom-substituted aldehydes 3 furnishing iso-
merically pure γ-hydroxy vinylsulfoximines, which in turn
cyclize to highly substituted hetero(poly)cyclic ring systems
4 (X ) O6 or X ) N7).
Based on an alternative classification scheme for ꢀ-turn
structures proposed by Ball in 1993,8 we envisaged the
(4) Reggelin, M.; Junker, B. Chem. Eur. J. 2001, 7, 1232–1239.
(5) (a) Reggelin, M.; Weinberger, H.; Gerlach, M.; Welcker, R. J. Am.
Chem. Soc. 1996, 118, 4765–4777. (b) Reggelin, M.; Weinberger, H. Angew.
Chem., Int. Ed. 1994, 33, 444–446.
(6) Reggelin, M.; Gerlach, M.; Vogt, M. Eur. J. Org. Chem. 1999, 1011–
1031. Reggelin, M.; Weinberger, H.; Heinrich, T. Liebigs Ann. Recl. 1997,
1881–1886.
10.1021/ol8015995 CCC: $40.75
Published on Web 08/20/2008
2008 American Chemical Society

development of the azapolycyclic systems (4, X ) N)
accessible via 2-alkenyl sulfoximines into type-III peptido-
mimetics.9 For this to be successful, both the constitutional
as well as the configurational scope of the reaction sequence
leading to the desired compounds has to be as broad as
possible.
Scheme 3. Deuteration of the R-Position
Furthermore, a reliable and profitable method to remove
the auxiliary has to be found. Whereas the former demand
has been shown to be fulfilled,7,10 the removal of the
sulfonimidoyl moiety clearly needed improvement. For the
oxacycles (4, X ) O) and some of the azacycles (4, X )
N), the application of Raney nickel was successful.7a
Moreover, reductive desulfurizations with lithium naphtha-
lenide (LN) proved to be an alternative for the azacycles
(Scheme 2).7
rather difficult (Scheme 3). The carbon next to the sulfon-
imidoyl moiety (R-carbon) is in a neopentyl position.
Therefore, even after electrophilic activation of the latter,
nucleophilic substitution should be hampered by steric
hindrance.
This was found to be the case indeed.11 Consequently,
we next tried to deprotonate the R-position of N-donor
substituted derivatives like 7. From deuteration experiments,
we learned that carbanion formation takes place but the
deuteron was the only electrophile able to attack this position
(no reaction occurred with allyl bromide, acrolein and
acetaldehyde!). Interestingly, 8 was isolated as a single
isomer. At this point, we realized that only a highly reactive,
sterically nondemanding reagent might be able to form a
bond to this extremely hindered position. These consider-
ations led us to the conclusion to try carbenoids as reaction
partners. Theory predicts iodomethyl lithium to be a car-
benoid of extraordinary high electrophilicity.12 On the other
hand, due to its pronounced thermal instability,13 we decided
to apply the corresponding magnesium derivative.14a-c
Following the work of Julia on the olefinating desulfurization
of sulfones,14c,d we reacted a series of sulfonimidoylmethyl-
substituted azacycles with the carbenoid generated from
diiodomethane and isopropylmagnesium iodide (Table 1).
These experiments were performed using two flasks
connected by a glass capillary (see the Supporting Informa-
tion). The whole assembly was immersed in a dry ice-acetone
bath, and one flask contained the deprotonated sulfoximine
and the other one the carbenoid solution prepared by adding
diiodomethane to the Grignard reagent. After 1 h at -78
°C, the R-deprotonated azacycle was added dropwise to the
carbenoid solution under isothermal conditions using the
Scheme 2.
Nonfunctionalizing Desulfurizationsa
aLN: Lithium naphthalenide, R ) H or TBS, X ) N or O.
Despite these successes, several shortcomings remain.
Raney nickel gives reliable results only in the case of the
oxacycles; the yields with the azacycles are variable and often
unsatisfactory. Desulfurizations with LN require electron-
donating substituents at nitrogen and are plagued by ꢀ-e-
limination leading to ring-opened products (with the oxacy-
cles this is the main reaction). The most reliable method until
now has been the reduction with SmI2.7 For this desulfur-
ization to work, it is necessary to have the nitrogen Boc-
protected and the auxiliary deprotected (R ) H, Scheme 2).
This in turn demands for the absence of ring 3, which is
obviously a constitutional drawback. A severe disadvantage
of all three variants is the generation of a methyl group which
does not allow any further functionalization. On the other
hand, in the course of our experiments with 2-oxabicyclo-
[3.3.0]octanes we developed a sequence to angular vinyl-
substituted derivatives.5a Unfortunately, this procedure suffers
from the instability of the R-methylated allylic sulfoximines
used as starting material toward elimination to the corre-
sponding dienes, and even worse, we were not able to apply
this chemistry to the azacyclic systems.
(11) Nevertheless, with N-methyl-substituted sulfoximines this has been
shown to be possible. Rajender, A.; Gais, H. J. Org. Lett. 2007, 9, 579–
582. Gais, H.-J.; Loo, R.; Roder, D.; Das, P.; Raabe, G. Eur. J. Org. Chem.
2003, 1500–1526. Kohler, F.; Gais, H. J.; Raabe, G. Org. Lett. 2007, 9,
1231–1234. Adrien, A.; Gais, H. J.; Kohler, F.; Runsink, J.; Raabe, G. Org.
Lett. 2007, 9, 2155–2158.
For these reasons, we turned our attention to postcycliza-
tion modifications. Not unexpectedly, this turned out to be
(12) (a) Zhou, Y. B.; Cao, F. L. J. Organomet. Chem. 2007, 692, 3723–
3731. (b) Boche, G.; Lohrenz, J. C. W. Chem. ReV. 2001, 101, 697–756.
(13) Ko¨brich, G. Angew. Chem., Int. Ed. 1972, 11, 473–485. Ko¨brich,
G.; Von Nagel, R. Chem.-Ztg. 1970, 94, 984–985.
(7) (a) Reggelin, M.; Junker, B.; Heinrich, T.; Slavik, S.; Bu¨hle, P. J. Am.
Chem. Soc. 2006, 128, 4023–4034. (b) Reggelin, M.; Heinrich, T. Angew.
Chem., Int. Ed. 1998, 37, 2883–2886.
(14) (a) Villieras, J. Bull. Soc. Chim. Fr. 1967, 1520–1532. (b) Hahn,
R. C.; Tompkins, J. Tetrahedron Lett. 1990, 31, 937–940. (c) De Lima, C.;
Julia, M.; Verpeaux, J. N. Synlett 1992, 133–134. (d) Charreau, P.; Julia,
M.; Verpeaux, J.-N. J. Organomet. Chem. 1989, 379, 201.
(8) Ball, J. B.; Hughes, R. A.; Alewood, P. F.; Andrews, P. R.
Tetrahedron 1993, 49, 3467–3478.
(9) Rich, D. H.; Ripka, A. S. Curr. Opin. Chem. Biol. 1998, 2, 441–
452.
(15) (a) Plietker, B.; Seng, D.; Fro¨hlich, R.; Metz, P. Eur. J. Org. Chem.
2001, 3669–3676. (b) Plietker, B.; Metz, P. Tetrahedron Lett. 1998, 39,
7827–7830. (c) Pearlman, B. A.; Putt, S. R.; Fleming, J. A. J. Org. Chem.
2006, 71, 5646–5657.
(10) Reggelin, M.; Ku¨hl, J.; Kaiser, J. P.; Bu¨hle, P. Synthesis 2006, 2224–
2232
.
4082
Org. Lett., Vol. 10, No. 18, 2008

Table 1. Desulfurization Experiment
Scheme 4.
Mechanistic Considerationsa
aThe R-methylated product 14 was never observed.
As can be seen from Table 1, N-protection by an electron-
releasing substitutent is a necessary precondition for the
reaction to work with the catalytically cleavable allyl or
benzyl groups being particularly useful. Indeed, the N-Boc-
protected azabicycle 9g was not desulfurized under standard
reaction conditions (Table 1, entry 10), but delivered the ring
opened product 16 in 60% yield (already 47% after only 1
min. reaction time, Scheme 5).
a n-BuLi (1 equiv), i-PrMgI, and CH2I2 (10 equiv each). b One-pot
protocol: n-BuLi (1 equiv), CH2I2 (10 equiv), i-PrMgI (10 equiv). c Same
as (a) but 12 h at -78 °C, quench at -78 °C. d Same as (c) but only 1 h
reaction time. e Ring-opened product 16 was formed (Scheme 5).
cooled capillary. Finally, the reaction mixture was stirred
overnight at which point it reached room temperature
(method a) in Table 1). To our delight, both the 2-azabicyclo-
[3.3.0]octanes(n)0)aswellasthe2-azabicyclo[4.4.0]decanes
(n ) 1) delivered the expected angular vinyl substituted
products 10i in reasonable to good yields, depending on the
N-protecting group.
Scheme 5
.
Retro-Michael Addition with EWG-Protected
Derivatives
These results clearly demonstrate again the enhanced
electrophilicity of R-metalated alkyl halogenides as compared
to the halogenides themselves as was demonstrated
experimentally14b and predicted by theory.12
In order to facilitate the experimental procedure we next
tried to develop a one-pot variant of the reaction. Toward
this end, we first tried to combine the deprotonation with
carbenoid formation.
Mechanistically, it seems reasonable that the R-deproto-
nated sulfoximine 12 reacts first with either iodomethylmag-
nesium iodide or the magnesium iodide complexed
methylene12b to yield carbanionic species 13 which in turn
ꢀ-eliminates to the desired olefin 10c and sulfinamide 11
(Scheme 4). By comparison with an independently synthe-
sized sample, it was shown that this amide was generated
with complete retention of the sulfur configuration.
In a first experiment at -78 °C, we added 11 equiv of
isopropylmagnesium iodide to sulfoximine 9c to deprotonate
the R-position. After that, 10 equiv of CH2I2 were added to
generate the carbenoid which was expected to react with the
carbanion. Unfortunately, after aqueous workup, the starting
material was recovered quantitatively which was also found
after replacement of the Grignard reagent by the same
amount of n-BuLi. A deuteration experiment showed that
the former was not basic enough to deprotonate the sulfox-
imine, thus explaining the failure of the first variant. The
lack of success for the n-BuLi based variant remains unclear.
This disappointing outcome of the reaction is somewhat
surprising; the more so as a slight variation in the reaction
conditions was met with at least partial success. When 1
equiv of n-BuLi was employed to deprotonate the sulfox-
imine, 10 equiv each of the Grignard and CH2I2 were added
The whole process is somewhat related to olefinating
desulfurizations of sultones15a and sulfones.15b An interesting
difference is that in our case the carbanionic intermediate
13 cannot be trapped to the R-methylated sulfoximine 14
by protonation. Instead, quenching the reaction after 1 or
16 h at -78 °C delivered only the olefin 10c in 27% or 60%
yield, respectively. Therefore, we conclude that with our
systems the carbenoid uptake is rate limiting and the
subsequent elimination is a fast process which in turn is in
accordance with our finding that the deprotonated R-position
is almost inert toward electrophilic substitution (s. above).
Org. Lett., Vol. 10, No. 18, 2008
4083

to generate the carbenoid, and this was allowed to react with
the carbanion, 14% of the olefin was isolated after all. Two
conclusions may be drawn from these results. First, the
metalated sulfoximine obviously competes for the CH2I2
whereby starting material is regenerated by protonation.
Second, iodomethylmagnesium iodide appears to react dif-
ferently with the carbanion as compared to lithiomethyl
iodide. Obviously, only the former undergoes electrophilic
substitution. Therefore, the deprotonation step and the
carbenoid generation have to be separated, entailing the
above-described two-vessel procedure to be the superior one.
Based on the successful desulfurization of the azabicyclo-
[4.4.0]decanes and azabicyclo[3.3.0]octanes depicted in Table
1, we next explored the constitutional scope of the reaction.
Beside the decahydro[1,6]naphthyridin derivative 17,10 the
tricyclic and quadricyclic sulfoximines 18 and 19 were of
particular interest (Scheme 6).
It not only permits the removal of the auxiliary, it adds
value to the systems by the installation of a vinyl group suited
for numerous further functionalizations. Moreover, as already
mentioned, the sulfonimidoyl moiety is transformed to
sulfinamide 11 with complete retention of the sulfur con-
figuration, thus allowing the recyclisation of the auxiliary.4,16
In summary, we have developed a new functionalizing
desulfurization for sulfoximines in the course of which a
sulfonimidoylmethyl moiety is replaced by a vinyl group.
The method relies on the electrophilic substitution of lithiated
sulfoximines by the carbenoid generated from diiodomethane
and isopropylmagnesium iodide followed by ꢀ-elimination.
The reaction works even with sterically highly congested
systems such as the ones described in this paper (neopentyl
position).
Future work will be devoted to the synthesis of peptide
mimetics following the strategy outlined in a previous
publication.7a Moreover, the structural proximity of the
2-azabicyclo[4.4.0]derivatives 10 to the cyclindricine alka-
loids17 offers opportunities to explore synthetic routes to
these biologically active compounds. Finally, the pyr-
rolo[3,2,1-ij]quinoline fragment embedded in the quadricycle
22 recommends this compound and derivatives thereof as
potential 5-HT2c receptor agonists18 or indole analogues of
flavonols.19
Scheme 6. Constitutional Scope
Acknowledgment. We gratefully acknowledge Solvay
Pharmaceuticals (Hannover, Germany) for generous support.
Supporting Information Available: Experimental details
and NMR spectra of compounds 10a-f, 16, and 20-22 and
experimental data for the preparation of the starting N,O-
protected sulfoximines 9a-g as well as 17-19. This material
is available free of charge via the Internet at http://pubs.acs.org.
OL8015995
These latter two compounds were resistant toward any
other desulfurization method tried so far (LN, SmI2, sodium
amalgam). To our delight, the new functionalizing desulfu-
rization worked very well in all these cases, delivering the
highly functionalized polycyclic ring systems 20-22. In
connection with our efforts to develop peptide mimetics
based on the asymmetric synthesis of azapolycyclic com-
pounds from 2-alkenyl sulfoximines,7a this olefinating des-
ulfurization represents a major breakthrough.
(16) Reggelin, M.; Weinberger, H. Tetrahedron Lett. 1992, 33, 6959–
6962. Reggelin, M.; Welcker, R. Tetrahedron Lett. 1995, 36, 5885–5886
.
(17) Li, C.; Blackman, A. J. Aust. J. Chem. 1995, 48, 955–965. Li, C.;
Blackman, A. J. Aust. J. Chem. 1994, 47, 1355–1361. Blackman, A. J.; Li,
C.; Hockless, D. C. R.; Skelton, B. W.; White, A. H. Tetrahedron 1993,
49, 8645–8656.
(18) Isaac, M.; Slassi, A.; O’Brien, A.; Edwards, L.; MacLean, N.;
Bueschkens, D.; Lee, D. K. H.; McCallum, K.; De Lannoy, I.; Demchyshyn,
L.; Kamboj, R. Bioorg. Med. Chem. Lett. 2000, 10, 919–921.
(19) Black, D. S.; Kumar, N.; Mitchell, P. S. R. J. Org. Chem. 2002,
67, 2464–2473.
4084
Org. Lett., Vol. 10, No. 18, 2008