Stereocontrolled generation of α-metalated S,O-acetals by sulfoxide-ligand exchange from cyclic dithioorthoformate monooxides

Article abstract of DOI:10.1021/om200858v

Treatment of trans or cis 2-isopentoxy-1,3-benzodithiolane-Soxides with EtMgCl gave configurationally stable (≥2.5 h, at -78 °C) stereodefined α-magnesiated S,O-acetals that incorporated D atoms in a stereospecific manner upon reaction with CD₃OD. α-Lithiated S,O-acetals generated in the same fashion using PhLi were found to be less configurationally stable.

Full text of DOI:10.1021/om200858v

Communication
pubs.acs.org/Organometallics
Stereocontrolled Generation of α-Metalated S,O-Acetals by
Sulfoxide-Ligand Exchange from Cyclic Dithioorthoformate
Monooxides†
Adam L. Barsamian and Paul R. Blakemore*
Department of Chemistry, Oregon State University, Corvallis, Oregon 97331-4003, United States
S
* Supporting Information
ABSTRACT: Treatment of trans or cis 2-isopentoxy-1,3-benzodithiolane-S-
oxides with EtMgCl gave configurationally stable (≥2.5 h, at −78 °C)
stereodefined α-magnesiated S,O-acetals that incorporated D atoms in a
stereospecific manner upon reaction with CD3OD. α-Lithiated S,O-acetals
generated in the same fashion using PhLi were found to be less
configurationally stable.
Sp3-hybridized stereogenic carbanions present a myriad of
tantalizing possibilities for asymmetric synthesis;1 however, few
methods exist for their stereocontrolled synthesis.2 In a
significant advance, Hoffmann et al. reported the generation
of stereodefined α-haloalkyl Grignard reagents via sulfoxide-
ligand (or “sulfoxide-metal”) exchange from scalemic α-halo-
sulfoxides.3 Capitalizing on this discovery, we established that
such carbenoid reagents,4a and their lithiated counterparts,4a−c
can be used to chain extend boronic esters in a stereospecific
manner. When applied in an iterative sense, this stereospecific
reagent-controlled homologation (StReCH) process offers a
programmable “assembly line” approach to synthesis in which
the carbenoid presentation sequence uniquely determines the
constitution and stereochemical configuration of an emerging
polysubstituted alkyl chain (1 + 2 → 3 →→ 4, Scheme 1).5 In
Brown and Imai concerning the homologation of boronates by
methoxy(phenylthio)methyllithium (5)9 drew our attention to
α-metalated S,O-acetals. It was observed that ate-complexes
generated by the interception of 5 by boronic esters (e.g., 6)
rearrange to give α-alkoxyboronates upon treatment with
thiophile HgCl2 (Scheme 2). Assuming that 1,2-metalate
Scheme 2. Representative Example of Chain Extension of a
Boronic Ester with Methoxy(phenylthio)methyllithium (5)
a
As Reported by Brown and Imai9
a
hex = n-C6H13, Bpin = B(OCMe2CMe2O).
Scheme 1. Stereospecific Reagent-Controlled Homologation
(StReCH) of Organoboron Compounds 1 with
Enantioenriched Chiral Carbenoids 2
rearrangement occurs with high stereochemical fidelity,10 the
Brown−Imai process could form the basis of a new StReCH
manifold, providing it can be established that α-metalated S,O-
acetals are configurationally stable and that a means to access
them in a stereodefined form can be devised. Addressing these
particular issues, herein it is reported that α-magnesiated S,O-
acetals (and, to a lesser extent, their lithiated congeners) retain
stereochemical integrity during addition to hard electrophiles
and that these carbenoids are generated in a stereocontrolled
fashion by sulfoxide-ligand exchange from cyclic dithioortho-
formate monooxides.
Carbenoid 5 was generated in a racemic form by Brown and
Imai by s-BuLi-mediated deprotonation of MeOCH2SPh.9 It is
conceivable that metalation of such prochiral methylene S,O-
acetals could be achieved in an enantioselective manner using a
chiral base;2,11 however, to avoid an unnecessary reliance on
stereoinduction,4a we elected instead to focus on an inherently
addition to α-chloroalkylmetal species,4 the StReCH paradigm
has now also been successfully implemented with other types of
chiral carbenoids, including α-lithiocarbamates,6 α-lithio-epox-
ides/-aziridines,7 and α-lithio-N-Boc amines.8
The synthetic utility of StReCH would be further enhanced if
a viable configurationally stable oxygen-bearing carbenoid 2 (Ri
= OR′) could be identified that is capable of net stereospecific
insertion into a C−B σ-bond. The existence of such a reagent
would permit targeting of stereochemically complex oxygenated
compounds, e.g., polyketides and carbohydrates, by iterative
StReCH cycles. With this ultimate goal in mind, a report from
Received: September 12, 2011
Published: December 19, 2011
© 2011 American Chemical Society
19
dx.doi.org/10.1021/om200858v | Organometallics 2012, 31, 19−22

Organometallics
Communication
stereospecific route to the organometallic reagents of interest.
On the basis of our experiences with sulfoxide-ligand
exchange,4,12 and the suggestive work of Nell,13 it seemed
likely, albeit not certain, that dithioorthoformate monooxides 9
would provide stereodefined α-metalated S,O-acetals 10 upon
treatment with organolithium and Grignard reagents (Scheme
Scheme 5. Synthesis of 2-Isopentoxy-1,3-benzodithiolane-S-
oxides 18
Scheme 3. Proposed Route to Stereodefined α-Metalated
S,O-Acetals 10 via Sulfoxide-Ligand Exchange from
Dithioorthoformate Monooxides 9
Oxidation of dithioorthoformate 17 with m-CPBA14a afforded a
mixture of separable S-oxides 18t and 18c (dr > 20:1,
respectively), the major diastereoisomer of which was assumed
to be trans configured.18 Regardless of the fact that the
materials obtained were racemic, a legitimate means to assess
the configurational stability of α-metalated S,O-acetals was now
available to us. Sulfoxide-ligand exchange from either 18t or
18c would lead to carbanions (19t/19c) with different relative
stereochemistry (Scheme 6). Intermediates 19 could therefore
3). In addition, it was recognized that the little investigated14
and intrinsically chiral functional group cluster in 9 would
emerge naturally from the corresponding prochiral dithioor-
thoformate 8 upon oxidation. This tactic is synthetically
attractive, creating two new stereogenic centers in a single
operation; however, issues of both diastereo- and enantiose-
lectivity must eventually be managed in any optimal conversion
of 8 to 9. With this plan in mind, the preparation of acyclic
dithioorthoformates and their subsequent oxidation was initially
investigated.
Scheme 6. Experiments to Establish Configurational Stability
of α-Metalated S,O-Acetals (19t/c) Derived from Distinct
Diastereomers of 18
Eschewing the more elaborate protocols to prepare acyclic
dithioorthoformates,15 we devised two methods to access test
substrates (Scheme 4). In the first, S,O-acetal 12 was lithiated
Scheme 4. Two Simple Syntheses of Acyclic
Dithioorthoformates
be gauged to be configurationally stable (on the time scale of a
reaction with a given test electrophile) if otherwise identical
sulfoxide-ligand exchange experiments from 18t and 18c led to
distinguishable isomeric products 20t and 20c upon electro-
philic quench.19 In a scenario of configurational lability,
isomeric organometals 19t and 19c would experience a period
of equilibration prior to quench and net nonstereospecific
reactions would be noted.
with s-BuLi and then sulfenylated to give 13;16 in the second,
more concise approach, triethylorthoformate was subjected to
acid-catalyzed ligand metathesis with p-thiocresol to provide
dithioorthoformate 15 in addition to the separable trithioor-
thoformate. No examples of acyclic dithioorthoformate
monooxides have been reported,14 and the oxidation of
compounds 13/15 was approached with some trepidation.
Indeed, after surveying numerous oxidation protocols (see
Supporting Information for details), no traces of S-monooxide
adducts were ever identified. It was concluded that acyclic
dithioorthoformate monooxides likely fragment into thioxacar-
benium ion and sulfenate species upon their genesis;
decomposition pathways available to the latter presumably led
to the principle observed products, which were mixtures of
disulfide oxides (ArS(O)nS(O)mAr; m, n = 0, 1, 2). Whatever
the precise nature of the oxidative decomposition pathways
encountered, it became evident that simple acyclic dithioortho-
formate monooxides are not viable molecules.
Given the precedented sulfoxide-magnesium exchange from
dithioacetal monooxides,13 the generation of magnesiated S,O-
acetals from substrates 18t/c was investigated first (Table 1).
Deprotonation can compete with sulfoxide-ligand exchange,12
and this side-reaction was considered as a potential risk given
the high acidity of the central C−H bond in dithioorthoformate
monooxides 18. In the event, treatment of 18t with EtMgCl (1
equiv, THF, −78 °C) followed by methanolysis after 15 min
gave the ring-opened adduct 21 (R = Et) in a high yield,
consistent with the generation of the desired metalate 19t (M =
MgCl) without significant intervention by deprotonation (entry
1). Repetition of the experiment but with quenching by
CD3OD gave monodeuteride 20t (E = D) as a single
diastereoisomer (entry 2). The comparable experiment from
18c resulted in a less efficient sulfoxide exchange reaction, but,
crucially, the monodeuterides obtained showed a strong
preference for the opposite isomer (20c, E = D) (entry 3).
Allowing the putative organomagnesiums 19t/c (generated as
before) to incubate for an additional 2.25 h (at −78 °C) prior
to quench with CD3OD gave product deuterides in lower yield
Focus was next directed to 2-alkoxy-1,3-benzodithiolanes, a
well-described class of cyclic dithioorthoformates that have been
successfully converted into the corresponding S-oxides. 2-
Isopentoxy-1,3-benzodithiolane (17) was synthesized from
anthranilic acid by isoamyl nitrite-initiated benzyne formation
in the presence of CS2 and isoamyl alcohol (Scheme 5).17
20
dx.doi.org/10.1021/om200858v | Organometallics 2012, 31, 19−22

Organometallics
Communication
oxygen-bearing carbenoids for StReCH. Work to address such
issues is in progress and will be reported in due course.
Table 1. Sulfoxide-Ligand Exchange (SLE) by RM from
Dithioorthoformate Monooxides 18t/c and Trapping with
Probe Electrophiles (EX)
ASSOCIATED CONTENT
a
■
yield/dr of SLE adducts
S
* Supporting Information
reagent
quench
nontrivial quench protonat.
All synthetic procedures, characterization data, and 1H and 13
C
no.
S.M.
R−M
E−X
20t/c
20t:20c
21
NMR spectra for relevant compounds. This material is available
free of charge via the Internet at http://pubs.acs.org.
b
1
18t
18t
18c
18t
18c
18t
18c
Et−MgCl
Et−MgCl
Et−MgCl
Et−MgCl
Et−MgCl
Ph−Li
H−OCH3
D−OCD3
D−OCD3
D−OCD3
D−OCD3
D−OCD3
D−OCD3
n/a
n/a
84%
10%
26%
24%
21%
2.9%
1.3%
b
2
78%
20%
52%
9.4%
7.1%
5.7%
>99:01
07:93
96:04
15:85
66:34
19:81
b
3
AUTHOR INFORMATION
■
c
4
Corresponding Author
*E-mail: paul.blakemore@science.oregonstate.edu.
c
5
d
6
d
7
Ph−Li
ACKNOWLEDGMENTS
a
■
SLE adducts isolated from other components by SiO2 chromatog-
raphy and then yield/dr determined by H NMR spectral analysis.
1
Financial support for this study from the National Science
Foundation is gratefully acknowledged (CHE-0906409). The
National Science Foundation (CHE-0722319) and the
Murdock Charitable Trust (2005265) are also thanked for
their generous support of the OSU NMR spectroscopy facility.
b
R−M (1 equiv) added to 18 (1 equiv) in THF, −78 °C, 15 min, then
c
E−X. R−M (1 equiv) added to 18 (1 equiv) in THF, −78 °C, 2.5 h,
d
then E−X. Barbier conditions: R−M (1 equiv) added to a mixture of
18 (1 equiv) + E−X (2.5 equiv), THF, −78 °C.
(indicating a degree of chemical instability) but with only
slightly diminished stereoselectivity (entries 4 and 5). On the
basis of this series of experiments, it was concluded that α-
magnesiated S,O-acetals are configurationally stable on a
macroscopic time scale (at −78 °C) and that an ample
temporal window potentially exists for their coalescence with
electrophiles before stereofidelity is significantly eroded.
Unfortunately, our attempts to intercept putative metalate
19t (M = MgCl) with a variety of other simple electrophiles
revealed its poor nucleophilicity. Thus, no significant bond
formation resulted from treating 19t (M = MgCl) with MeI,
MeOTf, MOMCl, allyl chloride, or allyl bromide.20 It did,
however, prove possible to intercept this species with
benzaldehyde to afford the expected addition adduct in a
good yield as a mixture of two diastereoisomers (52%, dr =
2:1).21
By contrast to the Grignard series, generation of α-lithiated
S,O-acetals by PhLi-mediated sulfoxide-ligand exchange from
substrates 18t/c was problematic. Deprotonation of 18t/c was
now a dominant pathway, and any initially generated adducts
19t/c (or 21 formed from internal proton transfer) experienced
significant secondary sulfoxide exchange with PhLi, leading to
PhSCH2OAm and Ph2SO. Some useful data were obtained by
separately treating 18t and 18c with PhLi in the presence of
CD3OD (entries 6, 7). Deuterides 20t/c (E = D, R = Ph) were
formed via this “Barbier”-type protocol in a low yield; in each
case, the expected diastereoisomer predominated but both
reactions were far from stereospecific, suggesting that α-
lithiated S,O-acetals 19 have poor configurational stability.22
In summary, it has been established that stereodefined α-
magnesiated S,O-acetals are available in excellent yield by
sulfoxide-ligand exchange between a class of readily prepared
cyclic dithioorthoformate monooxides (e.g., 18t) and EtMgCl.
These species are configurationally stable on a macroscopic
time scale but possess low nucleophilicity. More reactive α-
lithiated S,O-acetals are similarly generated using PhLi, but
side-reactions accompanying the sulfoxide-metal interchange
limit the synthetic usefulness of the process. It remains to be
discovered whether the favorable attributes identified for α-
magnesiated S,O-acetals (i.e., configurational stability, ease of
stereocontrolled generation) can be reconciled against their
poor nucleophilicity in the context of potential deployment as
DEDICATION
■
†Dedicated to Prof. Philip J. Kocienski on the occasion of his
65th birthday.
REFERENCES
■
(1) (a) Basu, A.; Thayumanavan, S. Angew. Chem., Int. Ed. 2002, 41,
716. (b) Organolithiums: Selectivity for Synthesis; Clayden, J., Ed.;
Pergamon: New York, 2002.
(2) Prevalent methods involve deprotonation; see: (a) Hoppe, D.;
Hense, T. Angew. Chem., Int. Ed. 1997, 36, 2282. (b) Beak, P.; Basu, A.;
Gallagher, D. J.; Sun Park, Y.; Thayumanavan, S. Acc. Chem. Res. 1996,
29, 552.
(3) (a) Hoffmann, R. W.; Nell, P. G.; Leo, R.; Harms, K. Chem.
Eur. J. 2000, 6, 3359. (b) Hoffmann, R. W. Chem. Soc. Rev. 2003, 32,
225.
For earlier work with racemic sulfoxide precursors, see:
(c) Durst, T.; LeBelle, M. J.; van den Elzen, R.; Tin, K.-C. Can. J.
Chem. 1974, 52, 761. (d) Satoh, T.; Takano, K. Tetrahedron 1996, 52,
2349.
(4) (a) Blakemore, P. R.; Marsden, S. P.; Vater, H. D. Org. Lett. 2006,
8, 773. (b) Blakemore, P. R.; Burge, M. S. J. Am. Chem. Soc. 2007, 129,
3068. (c) Emerson, C. R.; Zakharov, L. N.; Blakemore, P. R. Org. Lett.
2011, 13, 1318.
(5) For an introduction to the related substrate-controlled chain
extension process developed extensively by Matteson, see:
(a) Matteson, D. S. Tetrahedron 1998, 54, 10555. (b) Thomas, S.
P.; French, R. M.; Jheengut, V.; Aggarwal, V. K. Chem. Rec. 2009, 9, 24.
(6) Stymiest, J. L.; Dutheuil, G.; Mahmood, A.; Aggarwal, V. K.
Angew. Chem., Int. Ed. 2007, 46, 7491.
(7) (a) Vedrenne, E.; Wallner, O. A.; Vitale, M.; Schmidt, F.;
Aggarwal, V. K. Org. Lett. 2009, 11, 165. (b) Schmidt, F.; Keller, F.;
Vedrenne, E.; Aggarwal, V. K. Angew. Chem., Int. Ed. 2009, 48, 1149.
(8) Coldham, I.; Patel, J. J.; Raimbault, S.; Whittaker, D. T. W.;
Adams, H.; Fang, G. Y.; Aggarwal, V. K. Org. Lett. 2008, 10, 141.
(9) Brown, H. C.; Imai, T. J. Am. Chem. Soc. 1983, 105, 6285. An
essentially identical chain extension of Ph(CH2)2Bpin using carbenoid
5 was successfully conducted in our laboratory.
(10) It was previously noted by Matteson et al. that ate-complexes
closely related to 6 but lacking the alkoxy group fail to rearrange (and
often fragment) under similar conditions. Thus, 1,2-metalate
rearrangement from 6 likely involves oxacarbenium ion character in
the transition state; however, this fact does not preclude inversion of
configuration at the migratory terminus upon 1,2-nucleophilic
rearrangement; see: (a) Mendoza, A.; Matteson, D. S. J. Org. Chem.
1979, 44, 1352. (b) Matteson, D. S.; Majumdar, D. Organometallics
1983, 2, 230.
21
dx.doi.org/10.1021/om200858v | Organometallics 2012, 31, 19−22

Organometallics
Communication
(11) Direct lithiation of S,O-acetal 12 with a combination of s-BuLi
and (−)-sparteine was briefly explored, but modest enantioselectivity
was observed. In the best result, p-TolSCH(Li)OMe generated with
this reagent pair (in PhMe at −78 °C) gave the expected addition
adduct upon treatment with PhCHO in 55% yield with dr = 74:26; the
major diastereoisomer exhibited 24% ee; see Supporting Information
for details.
(12) Blakemore, P. R.; Burge, M. S.; Sephton, M. A. Tetrahedron Lett.
2007, 48, 3999.
(13) Seminal work from Nell (aspects of which were later expanded
on by Satoh et al.) established that ArSCHRMgX species (R = alkyl)
offer excellent chemical and configurational stability and that such α-
thio-alkylmetals can be generated in enantiopure form by sulfoxide-
ligand exchange from dithioacetal monooxides; see: (a) Nell, P. G.
New J. Chem. 1999, 23, 973. (b) Satoh, T.; Akita, K. Chem. Pharm. Bull.
2003, 51, 181.
(14) Only five examples of dithioorthoformate monooxides are
known, all are cyclic; see: (a) Seio, K.; Sasaki, T.; Yanagida, K.; Baba,
M.; Sekine, M. J. Med. Chem. 2004, 47, 5265. (b) Bulman Page, P. C.;
Wilkes, R. D.; Namwindwa, E. S.; Witty, M. J. Tetrahedron 1996, 52,
2125. (c) Larkin, J. P.; Wyatt, J. A.; Hawkes, D. J. U.S. Patent 5026874
A 19910625, 1991.
(15) For a general method and leading references, see: Takeda, T.;
Sato, K.; Tsubouchi, A. Synthesis 2004, 1457.
(16) A related method has been previously reported: Fuji, K.; Ueda,
M.; Sumi, K.; Fujita, E. J. Org. Chem. 1985, 50, 662.
(17) (a) Nakayama, J. Synthesis 1975, 38. (b) El-Wareth Sarhan, A.;
Izumi, T. J. Chem. Res., Synop. 2002, 11.
(18) The analysis that follows, and the conclusions drawn, do not
depend on this assignment; however, given the near planarity of the
five-membered-ring system in 17, it is likely that oxidation occurred
preferentially from the face opposite the isopentyloxy substituent. In
addition, we observe that 18c rearranges to 18t under mildly acidic
conditions, bolstering the tentative assignment that 18t is the trans
isomer and that 18c is the cis isomer.
(19) For an analogous experiment to establish configurational
stability for an α-bromoalkyllithium reagent, see: Hoffmann, R. W.;
Ruhland, T.; Bewersdorf, M. J. Chem. Soc., Chem. Commun. 1991, 195.
(20) Poor nucleophilicity was also noted by Satoh and co-workers for
α-thioalkylmagnesium halides; see ref 13b.
(21) See Supporting Information for details. Addition adducts of this
type [RCH(OH)CH(SAr)OR′] have found utility in a connective
synthesis of enol ethers (RCHCHOR′) via radical elimination from
the corresponding xanthate derivatives; see: Vatele, J.-M. Tetrahedon
Lett. 1984, 25, 5997.
(22) α-Thioalkyllithiums generally lack good configurational stability
at −78 °C (e.g., a, b), but there are notable exceptions (e.g., c, d); see:
(a) Reich, H. J.; Kulicke, K. J. J. Am. Chem. Soc. 1995, 117, 6621.
(b) Hoffmann, R. W.; Dress, R. K.; Ruhland, T.; Wenzel, A. Chem. Ber.
1995, 128, 861. (c) Hoppe, D.; Kaiser, B.; Stratmann, O.; Frohlich, R.
Angew. Chem., Int. Ed. Engl. 1997, 36, 2784. (d) Hoffmann, R.;
Koberstein, R. J. Chem. Soc., Perkin Trans. 2000, 2, 595.
22
dx.doi.org/10.1021/om200858v | Organometallics 2012, 31, 19−22