Olefination with Sulfonyl Halides and Esters: Scope, Limitations, and Mechanistic Studies of the Hawkins Reaction

Article abstract of DOI:10.1021/acs.orglett.7b00517

Carbanions of alkanesulfonyl halides and esters react with nonenolizable carbonyl compounds to give olefins. Mechanistic studies reveal that initial aldol-type addition of the carbanions is followed by cyclization-fragmentation to alkenes, and the leaving group on the sulfonyl moiety (RSO₂X) controls carbanion stability and rate of the olefin formation.

Full text of DOI:10.1021/acs.orglett.7b00517

Letter
pubs.acs.org/OrgLett
Olefination with Sulfonyl Halides and Esters: Scope, Limitations, and
Mechanistic Studies of the Hawkins Reaction
Bartosz Gorski, Alicja Talko, Tymoteusz Basak, and Michał Barbasiewicz*
́
Faculty of Chemistry, University of Warsaw, Pasteura 1, 02-093 Warsaw, Poland
S
* Supporting Information
ABSTRACT: Carbanions of alkanesulfonyl halides and esters
react with nonenolizable carbonyl compounds to give olefins.
Mechanistic studies reveal that initial aldol-type addition of the
carbanions is followed by cyclization−fragmentation to
alkenes, and the leaving group on the sulfonyl moiety
(RSO2X) controls carbanion stability and rate of the olefin
formation.
ormation of the CC bonds plays a fundamental role in
alkanesulfonates as olefinating reagents was described for the
first time by Hawkins,3 who tested a series of mesylates. Indeed,
some of them, namely 2,2,2-trifluoroethyl and aryl mesylates,
deprotonated with t-BuLi at low temperature, underwent the
expected transformation with benzophenone to give 1,1-
diphenylethylene in moderate to goods yields. However,
reaction of benzaldehyde with 2,2,2-trifluoroethyl ethanesulfo-
nate (even used in excess) was much less efficient, giving alkene
in only 23% of yield. Hawkins also performed preliminary
mechanistic studies, but the limited scope of substrates,
moderate yields of products, and superficial stereochemical
studies made the report a proof of concept, with little practical
application. Shortly thereafter, Kagabu4 described reactions of
semistabilized carbanions of arylmethanesulfonyl fluorides,
generated with potassium carbonate under mild conditions.
With substituted benzaldehydes, he synthesized a set of
stilbenes in moderate to good yields and observed a correlation
between electrophilicity of the carbonyl group and isomer ratio
of the products. Unfortunately, under these conditions
nonstabilized precursors, as mesyl fluoride, condensed only
slightly with benzaldehyde. Interestingly, in a similar process,
Nader5 applied mesyl chloride in a reaction with potassium
fluoride as base and as a source of fluoride ion for the
conversion to mesyl fluoride in a one-pot process. The
olefination was demonstrated only on electrophilic ketones
and required very harsh reaction conditions (110−170 °C). To
the best of our knowledge, further studies of the olefination
reaction with precursors of nonstabilized carbanions were not
reported in the literature.
F
organic synthesis, and numerous reactions of carbonyl
compounds with derivatives of phosphorus, sulfur, and silicon
were described in the literature. For sulfur-based olefinations
the most common protocol, known as Julia olefination, involves
addition of sulfone carbanion, acylation of so-formed aldol-type
adduct, and reduction. The methodology was further improved
with heteroaromatic sulfones, which add to carbonyl com-
pounds and spontaneously undergo Smiles rearrangement,
giving olefins in a one-pot transformation (Scheme 1, top).1
Scheme 1. Sulfur-Based Olefination Reactions1
In our report, we present a new convenient protocol, which
makes the Hawkins olefination a useful methodology for the
synthesis of selected alkenes.
Our studies began from preliminary experiments, in which 1-
octanesulfonyl chloride (1a) was deprotonated with LiHMDS
in THF at −78 °C (Scheme 2, top). So-formed carbanion
Interestingly, both mechanistic schemes differ substantially
from reactivity of phosphorus reagents (e.g., organic
phosphonates in Horner−Wadsworth−Emmons reaction),
which form intermediate 4-membered oxaphosphetanes.2
A
similar mechanism for sulfur-based olefinations is much less
common, although important precedents were described by
Hawkins,3 Kagabu,4 and Nader.5 Process utilizing carbanions of
Received: February 20, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.7b00517
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 2. Preliminary Experiments and Optimization7
a
LiHMDS (1.3 mmol), PhCHO (1.5 mmol); with LiHMDS (1.0
mmol), PhCHO (1.0 mmol) the yield decreased to 57% (48:52).
appeared to be very unstable, and most likely eliminated to
sulfene,6 giving an ill-defined mixture of polar products.
However, when the base was added to a mixture of 1a and
benzaldehyde (Barbier conditions), the expected 1-phenylnon-
1-ene (2a) was isolated in 13% yield. The result proved the
E1cb mechanism of the elimination, where the short-living
carbanion was trapped in situ to form alkene. Then, under
similar conditions we tested 2,2,2-trifluoroethyl 1-octanesulfo-
nate (1b). The poor leaving group ability of trifluoroethoxide,
as compared with chloride, increased the stability of the
carbanion, and after 30 min at −78 °C we were able to recover
unreacted ester 1b with 90% of yield (GC). Interestingly, the
same carbanion, trapped with benzaldehyde, led only to an
aldol-type adduct 3, with essentially no traces of alkene formed
at low temperature. The data clearly demonstrated that the
character of the leaving group influences both carbanion
stability and rate of the olefin formation, and thus, the latter
process may run at higher temperature. Indeed, when the
reaction mixture was kept at rt for 16 h, alkene 2a was isolated
in 70% (procedure A). Attempts of application of sodium or
potassium bases (t-BuOK, NaHMDS, etc.) under similar
conditions led mainly to polar byproducts. However, after a
short optimization, we recognized lithium tert-butoxide as an
optimal base. With an equimolar proportion of substrates and a
2-fold excess of the base in anhydrous THF under argon,
combination of the reagents at rt for 16 h led to alkene 2a
isolated in 71% of yield as a mixture of two isomers (E/Z =
47:53) (procedure B; Scheme 2, bottom). With the optimized
procedure, we tested various aromatic aldehydes to obtain
substituted alkenes (Figure 1).8 Those substituted with
electron-donating groups (entries 2c−g) and a naphthalene
core (2h,i) led to the corresponding alkenes in high yields
Figure 1. Alkenes 2a−z synthesized form 2,2,2-trifluoroethyl
alkanesulfonates under the conditions of procedure B.7 Key: (a)
procedure A, 39% (89:11); (b) procedure A, 30% (70:30); (c)
procedure B at 65 °C, 71%; (d) byproducts were formed.
(65−86%), whereas the presence of acceptors as o-Br (2l,
31%), p-CF3 (2m, 19%), and alkyl tert-butyl group (2k, 16%)
decreased the yields. The reaction of benzophenone with 1b
was only moderately efficient (2n, 33%), but a higher yield was
obtained in the reaction of 2-naphthaldehyde with 2-propane-
sulfonate (2t, 68%). Evidently, the combination of electrophilic
aldehyde with secondary sulfonate appeared to be more
efficient for the synthesis of trisubstituted olefin, as compared
with ketone as a substrate. Finally, we also tested reactions of
various 2,2,2-trifluoroethyl alkanesulfonates with 2-naphthalde-
hyde. Ethyl, propyl, isobutyl, and 2-phenylethylsulfonates led to
the desired products in good yields, with the only exception of
semistabilized phenylmethanesulfonate, which in reaction with
benzaldehyde gave stilbene 2z as a pure E isomer in 52% of
yield. Limiting cases of the protocol were reactions of 1b with
p-(dimethylamino)benzaldehyde, pentafluorobenzaldehyde,
and enolizable hydrocinnamaldehyde (3-phenylpropionalde-
hyde), which failed to give the expected alkenes. Besides, an
attempt of a reaction between benzophenone and secondary
B
DOI: 10.1021/acs.orglett.7b00517
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
carbanion precursor 2,2,2-trifluoroethyl 2-propanesulfonate at
65 °C led only to traces of tetrasubstituted olefin.
Scheme 4. General Mechanistic Scheme of the Olefination
Reaction (Top) and Reactions of 1-Octanesulfonyl Halides
and Esters with Benzaldehyde under the Conditions of
Procedure B (Bottom)
To better understand these factors, which control the
reaction yield and stereoselectivity, we focused on the
mechanism of the transformation. First, intermediate aldol-
type adducts, isolated as single isomers 3a and 3b (for
stereochemistry assignments, see the Supporting Information),9
were subjected to conditions of procedure B with equimolar
amounts of 1-naphthaldehyde (Scheme 3). The diasteroisomers
Scheme 3. Reactions of Aldol-Type Adducts 3 with 1-
Naphthaldehyde under the Conditions of Procedure B7
and subjected it to conditions of procedure B.7 Only traces of
alkene 2a were detected in the reaction mixture, supporting
that sultone 4 is not an intermediate of the olefination reaction,
at least for model alkene 2a. Then, the formulated mechanism
was justified by a series of reactions of 1-octanesulfonyl halides
and esters with benzaldehyde under conditions of procedure B
(Scheme 4, bottom). At rt, 1-octanesulfonyl chloride 1a with t-
BuOLi easily eliminated to sulfene, giving only traces of olefin
2a (≤1%). 1-Octanesulfonyl fluoride, with a less nucleofugal
fluoride group, gave 13% of alkene, whereas 1b gave as much as
71%. In turn, under the same conditions, neopentyl 1-
octanesulfonate led exclusively to a mixture of diastereomeric
aldol adducts (55%) and unreacted substrate (42%). In this
case, very stable carbanion adds to benzaldehyde (although
equilibrium of the aldol-type addition seems to be less favored
at rt), but further transformation displays a substantial barrier,
and in effect, alkenes are not formed.
Finally, we carried out follow-up studies to demonstrate the
potential of the method. Preparation of 2a from 1b and
benzaldehyde was performed on a 100 mmol scale with
nondried, commercially available THF and concentrated
lithium tert-amoxide (2-methyl-2-butoxide), a cheaper alter-
native to t-BuOLi. After aqueous workup and filtration of the
mixture through a pad of silica gel in cyclohexane, alkene 2a of
excellent purity was isolated in 79% yield (Scheme 5, top).
An interesting multistep process was performed on 2,2,2-
trifluoroethyl 3-iodopropanesulfonate, 5 (Scheme 5, bottom).
Reaction with a 2-fold excess of benzaldehyde and a
corresponding amount of base led to intermediate aldol adduct,
which cyclized to the tetrahydrofuran ring (path a,12 instead of
attacking the sulfonyl group, path b). In the following step,
olefination with a second equivalent of benzaldehyde present in
the reaction mixture completed the transformation to give
functionalized product 6 as a single E isomer in 64% yield.
differed in reactivity in numerous points. A less polar, major
isomer 3a reacted slowly (ca. 1.5 h), giving a mixture of alkenes
with predominant Z isomer of 2a but also a substantial amount
of E-2a and cross-products E- and Z-2h. In turn, reaction of 3b
was very fast (completed after 2 min at rt, as determined by 1H
NMR)7 and led almost exclusively to E-2a. Most likely in the
first case, the system had enough time for excessive
equilibration of the aldol addition that it diminished overall
selectivity, whereas in the latter fast stereospecific cyclization−
fragmentation led to one isomer of the olefin. Consideration of
the general mechanistic scheme of the reaction led to the
following conclusions (Scheme 4, top). First, base deprotonates
the precursor, giving a relatively stable carbanion (step 1).
Then, two competitive processes can occur: intramolecular
elimination to sulfene (step 2) and intermolecular addition to
the carbonyl compound, giving a mixture of diastereoisomeric
aldol-type adducts (as O-anions, step 3). As the first process
seems to be irreversible (at least for 1a), the second one is
reversible, as demonstrated by the formation of cross-product
with 1-naphthaldehyde (2h). Then, the diastereomeric adducts
cyclize to pentacoordinated intermediates, which spontaneously
fragment to alkene (step 4). Although the exact structure of the
intermediates remains speculative, it is known that electro-
negative ligands present in apical positions tend to stabilize
pentacoordinated sulfur compounds.3,10 To support the
hypothesis, we synthesized sultone 4 (R1 = C7H15; R2 = Ph)
from 2a and chlorosulfonic acid in CD2Cl2/dioxane-d8,
according to the procedure described by Cerfontain.11
Although the product appeared to be too unstable for the
isolation, we characterized it in solution with 1H and 13C NMR
C
DOI: 10.1021/acs.orglett.7b00517
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 5. Follow-up Studies: Large-Scale Olefination under
Conditions of Procedure B (Top) and One-Pot Synthesis of
Functionalized Alkene 6 (Bottom)7
REFERENCES
■
(1) For reviews, see: (a) Blakemore, P. R. J. Chem. Soc., Perkin Trans.
1 2002, 2563−2585. (b) Aïssa, C. Eur. J. Org. Chem. 2009, 2009,
1831−1844.
(2) (a) Maryanoff, B. E.; Reitz, A. B. Chem. Rev. 1989, 89, 863−927.
(b) Still, W. C.; Gennari, C. Tetrahedron Lett. 1983, 24, 4405−4408.
(3) Hawkins, J. M.; Lewis, T. A.; Raw, A. S. Tetrahedron Lett. 1990,
31, 981−984.
(4) Kagabu, S.; Hara, K.; Takahashi, J. J. Chem. Soc., Chem. Commun.
1991, 408−410.
(5) Nader, B. S.; Cordova, J. A.; Reese, K. E.; Powell, C. N. J. Org.
Chem. 1994, 59, 2898−2901.
(6) (a) King, J. F. Acc. Chem. Res. 1975, 8, 10−17. (b) Caddick, S.;
Wilden, J. D.; Judd, D. B. Chem. Commun. 2005, 2727−2728.
(7) See the Supporting Information for details.
(8) In most of the olefin syntheses presented in Figure 1, in the
reaction mixture we detected some amount of unreacted ester, while
aldehyde was consumed. It is likely that low yields of alkenes with
acceptor-substituted aldehydes may arise from competitive reaction of
aldehydes with t-BuOLi (e.g., the Cannizzaro reaction).
(9) For the assignment of stereochemistry of diastereoisomeric aldol
adducts 3a,b we synthesized diastereoisomeric adducts of benzalde-
hyde with neopentyl phenylmethanesulfonate, one of which (10b) was
characterized by X-ray studies (see the Supporting Information for
details; CCDC 1530545).
(10) Perkins, C. W.; Wilson, S. R.; Martin, J. C. J. Am. Chem. Soc.
1985, 107, 3209−3218.
In conclusion, we presented a convenient protocol, utilizing
lithium tert-butoxide as a base, to perform olefination of
carbonyl compounds with 2,2,2-trifluoroethyl alkanesulfonates.
The process substantially differs in mechanism from the well-
known methodology of the one-pot Julia olefination with
heteroaromatic sulfones and mimics the reactivity of organic
phosphonates. In addition, we demonstrated that the structure
of the leaving group of the precursor displays a substantial
effect on carbanion stability and rate of fragmentation of the
sulfonyl group. The search of novel reactivities of the species is
presently ongoing in our laboratory.
(11) Bakker, B. H.; Cerfontain, H. Eur. J. Org. Chem. 1999, 1999, 91−
96.
(12) For review of intermolecular reactions of γ-halocarbanions, see:
Barbasiewicz, M.; Mąkosza, M. Helv. Chim. Acta 2012, 95, 1871−1890.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.7b00517.
Experimental procedures, stereochemistry assignments,
characterization data, and NMR spectra of the
synthesized compounds (PDF)
Structural data of X-ray studies (CIF)
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: barbasiewicz@chem.uw.edu.pl, www.aromaticity.pl.
ORCID
Michał Barbasiewicz: 0000-0002-0907-7034
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was financed by the SONATA BIS program of the
National Science Centre, Poland (NCN, Grant No. DEC-
2013/10/E/ST5/00030). T.B. thanks Warsaw Consortium of
Academic Chemistry for a personal KNOW scholarship.
D
DOI: 10.1021/acs.orglett.7b00517
Org. Lett. XXXX, XXX, XXX−XXX