Ready synthetic access to enantiopure allylic α(F)- branched fluoroalkenes

Article abstract of DOI:10.1021/ol400917j

Convenient access to homochiral fluoroalkenes is described via a Julia-Kocienski olefination reaction. The required homochiral fluorosulfone is synthesized by a Mitsunobu reaction from readily available enantiopure secondary alcohols.

Full text of DOI:10.1021/ol400917j

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Ready Synthetic Access to Enantiopure
Allylic r(F)-Branched Fluoroalkenes
Florent Larnaud,†,‡ Julien Malassis,†,‡ Emmanuel Pfund,† Bruno Linclau,*,‡ and
Thierry Lequeux*,†
Laboratoire de Chimie Moleculaire et Thioorganique, UMR CNRS 6507 & FR3038,
ENSICAEN, Universite de Caen Basse-Normandie, 6 Boulevard du Marechal Juin,
14050 Caen, France, and Chemistry, University of Southampton, Highfield,
Southampton SO17 1BJ, U.K.
bruno.linclau@soton.ac.uk; thierry.lequeux@ensicaen.fr
Received April 4, 2013
ABSTRACT
Convenient access to homochiral fluoroalkenes is described via a JuliaꢀKocienski olefination reaction. The required homochiral fluorosulfone is
synthesized by a Mitsunobu reaction from readily available enantiopure secondary alcohols.
Fluorination of bioactive compounds is now a routine
consideration in the drug development process.1 Apart from
preventing metabolic degradation, fluorine is often intro-
Scheme 1. Synthetic Approaches towards Enantioenriched
Allylic R(F)-Chiral Fluoroalkenes
duced to alter particular properties of adjacent functional
groups (e.g., pKa/pKb, and hydrogen bonding properties),
which then typically also has an effect on molecular properties
including conformation and lipophilicity.2 In addition, fluor-
ine can replace a functional group (e.g., alcohols), or fluorine-
containing groups can be used as functional group isosteres.
Perhaps the most known example is fluoroalkenes as amide
† Universite de Caen Basse-Normandie.
‡ University of Southampton.
(1) (a) O’Hagan, D. J. Fluorine Chem. 2010, 131, 1071–1081. (b)
ꢀ
ꢀ
Begue, J.-P.; Bonnet-Delpon, D. J. Fluorine Chem. 2006, 127, 992–1012.
(c) Kirk, K. L. J. Fluorine Chem. 2006, 127, 1013–1029. (d) Ismail,
F. M. D. J. Fluorine Chem. 2002, 118, 27–33.
(2) (a) O’Hagan, D. Chem. Soc. Rev. 2008, 37, 308–319. (b) Hunter,
L. Beilstein J. Org. Chem. 2010, 6, No 38. (c) Purser, S.; Moore, P. R.;
Swallow, S.; Gouverneur, V. Chem. Soc. Rev. 2008, 37, 320–330. (d)
Graton, J.; Wang, Z.; Brossard, A.-M.; Goncalves Monteiro, D.; Le
Questel, J.-Y.; Linclau, B. Angew. Chem., Int. Ed. 2012, 51, 6176–6180.
(3) (a) Couve-Bonnaire, S.; Cahard, D.; Pannecoucke, X. Org. Biomol.
Chem. 2007, 5, 1151–1157. (b) Choudhary, A.; Raines, R. T. ChemBioChem
2011, 12, 1801–1807. (c) Welch, J. T. Fluoroolefin Dipeptide Isosteres:
Structure, Synthesis, and Applications. In Fluorine and Health: Molecular
Imaging, Biomedical Materials, and Pharmaceuticals; Tressaud, A., Haufe,
G., Eds.; Elsevier: Amsterdam, 2008; pp 639ꢀ736.
isosteres, mimicking their dipolar nature while being hydro-
lytically stable.3 Fluoroalkenes are isoelectronic with enolates,
leading to additional applications as enolate isosteres.4
As a consequence, there are many methods for the syn-
thesis of fluoroalkenes.5 Nevertheless, the enantioselective
(4) (a) Pirrung, M. C.; Chen, J.; Rowley, E. G.; McPhail, A. T. J. Am.
Chem. Soc. 1993, 115, 7103–7110. (b) Frederickson, M.; Coggins, J. R.;
Abell, C. Chem. Commun. 2002, 1886–1887. (c) Burkhart, J. P.; Weintraub,
P. M.; Gates, C. A.; Resvick, R. J.; Vaz, R. J.; Friedrich, D.; Angelastro,
M. R.; Bey, P.; Peet, N. P. Bioorg. Med. Chem. 2002, 10, 929–934.
(5) Recent reviews: (a) Landelle, G.; Bergeron, M.; Turcotte-Savard,
M.-O.; Paquin, J.-F. Chem. Soc. Rev. 2011, 40, 2867–2908. (b) Yanai, H.;
Taguchi, T. Eur. J. Org. Chem. 2011, 5939–5954. (c) Zajc, B.; Kumar, R.
Synthesis 2010, 1822–1836.
r
10.1021/ol400917j
XXXX American Chemical Society

synthesis of fluoroalkenes having a chiral allylic center is
much less well-developed, especially for chiral centers that do
not contain a heteroatom.6 We are only aware of one
enantioselective synthetic method toward acyclic fluoroalk-
enes 1 with such a chiral center next to the dCF moiety
(R(F)), recently reported by Jorgensen et al.7 Organocata-
lyzed Michael addition involving readily available 2 leads to
enantioenriched 3 in excellent ee (Scheme 1). Elegantly,
reduction of 3 induces alkene formation, with E or Z
fluoroalkenes accessible in good selectivities (∼4:1) depend-
ing on the presence or absence of a chelating agent.
The JuliaꢀKocienski olefination reaction is now a pop-
ular method for fluoroalkene synthesis.5 Nevertheless, to the
best of our knowledge this method has not yet been reported
for the synthesis of RF-branched fluoroalkenes 1 (except one
isolated example involving a cyclopropyl group).8
We report a general synthesis of R(F)-branched allylic
fluoroalkenes, including homochiral fluoroalkenes, via
the JuliaꢀKocienski reaction using a benzothiazolyl (BT)
based fluorosulfone 6. Crucially, this methodology relies
on an effective synthesis of the required homochiral fluo-
rosulfone 6, which is also described here.
There are a number of possible syntheses of fluorosul-
fones 6, such as opening of internal epoxides or conjugate
additions to 3-substituted enones, both of which have been
Figure 1. Nucleophiles used for the synthesis of fluorosulfones
related to 6.
Table 1. Synthesis of the JuliaꢀKocienski Substrate 6 by Mitsunobu and Krapcho Reactions
a Isolated yield. b Secondary alcohols lead to a mixture of diastereomers; see text and SI. c At 20 °C. d 3% of 3-aryl-2-fluoroacrylate elimination
product. e At reflux temp. f Racemic substrate.
B
Org. Lett., Vol. XX, No. XX, XXXX

reported by Hu using nucleophiles derived from 7a (n-
BuLi, HMPA) or 8 (Cs/K2CO3) (Figure 1).9 To the best of
our knowledge, alkylation reactions starting from 5,10
7aꢀb,11 and 812 using secondary halides have not been
reported.13 Our attempts starting from 5 only led to
product 10 in low yield (eq 1). Moreover, a likely in situ
halide-mediated racemization of starting material will
preclude formation of homochiral derivatives 6. Finally,
direct fluorination leading to 6 is also a possibility. Rele-
vant precedence was published by Zajc,8 which showed
however that complete fluorination was difficult to
achieve. In addition, the required starting substrate would
require similar efforts to obtain.
Table 2. The JuliaꢀKocienski Reaction Leading to Allylic
R
(F)-Branched Fluoroalkenes
We have achieved a successful alkylation approach for
the synthesis of fluorosulfone precursors 6 starting from
secondary alcohols 4 instead of halides, using a modified
Mitsunobu reaction (Scheme 1). Importantly, homochiral
fluorosulfones 6 become accessible due to the ready avail-
ability of homochiral secondary alcohols. The required
subsequent Krapcho decarboxylation reaction was re-
cently reported by us,10b and significant improvements of
this reaction are reported herein.
The Mitsunobu reaction was first optimized for primary
alcohols (Table 1, entries 1ꢀ7). It was found that an
azodicarboxamide reagent (azodicarbonyldipiperidide,
ADDP)14 was necessary to provide a basic enough inter-
mediate to allow deprotonation of sulfone 5, and reaction
in toluene at elevated temperature gave the best yields. The
reaction worked well with both benzothiazolyl (BT) and
pyrimidyl sulfones, but no reaction was obtained using a
pyridyl sulfone (see Supporting Information). All further
optimization was carried out using BT-sulfones. From 5,
the Mitsunobu process with primary (including allylic,
benzylic) alcohols gave good yields, which are comparable
with the corresponding alkylation process with halides to
give 10.10a,b Nevertheless, more sterically hindered side
chains such as isobutyl were more efficiently incorporated
via the Mitsunobu process (41% via alkylation vs 64%,
entry 2). In addition, reaction with p-nitrobenzyl alcohol
(6) Example of racemic methodology: Lin, X.; Zheng, F.; Qing, F.-L.
J. Org. Chem. 2012, 77, 8696–8704.
(7) Jacobsen, C. B.; Nielsen, M.; Worgull, D.; Zweifel, T.; Fisker, E.;
Jorgensen, K. A. J. Am. Chem. Soc. 2011, 133, 7398–7404.
(8) Ghosh, A. K.; Zajc, B. J. Org. Chem. 2009, 74, 8531–8540.
(9) (a) Ni, C.; Li, Y.; Hu, J. J. Org. Chem. 2006, 71, 6829–6833. (b) Ni,
C.; Zhang, L.; Hu, J. J. Org. Chem. 2008, 73, 5699–5713.
(10) (a) Calata, C.; Catel, J.-M.; Pfund, E.; Lequeux, T. Tetrahedron
2009, 65, 3967–3973. (b) Larnaud, F.; Pfund, E.; Linclau, B.; Lequeux,
T. J. Fluorine Chem. 2012, 134, 128–135.
(11) (a) For 7a: Asakura, N.; Usuki, Y.; Iio, H.; Tanaka, T.
J. Fluorine Chem. 2006, 127, 800–808. (b) For 7b: Zhu, L.; Ni, C.; Zhao,
Y.; Hu, J. Tetrahedron 2010, 66, 5089–5100.
(12) Surya Prakash, G. K.; Chacko, S.; Vaghoo, H.; Shao, N.;
Gurung, L.; Mathew, T.; Olah, G. A. Org. Lett. 2009, 11, 1127–1130.
(13) The alkylation reaction of 7b with hindered halides such as
isobutyl iodide has been described in 62% yield; see ref 11b.
(14) (a) Tsunoda, T.; Yamamiya, Y.; Ito, S. Tetrahedron Lett. 1993,
34, 1639–1642. (b) Lai, J.-Y.; Yu, J.; Hawkins, R. D.; Falck, J. R.
Tetrahedron Lett. 1995, 36, 5691–5694.
a Isolated yield. b Determined by 19F NMR analysis.
gave a good yield (entry 7), while the corresponding
alkylation using p-nitrobenzyl bromide gave no reaction.
Alternatively, Pd-catalyzed allylation of 5 with cinnamyl
methyl carbonate has also been described (85% yield) as a
way to introduce allylic groups.15
(15) Zhao, X.; Liu, D.; Zheng, S.; Gao, N. Tetrahedron Lett. 2011, 52,
665–667.
Org. Lett., Vol. XX, No. XX, XXXX
C

The reaction of secondary alcohols (entries 8ꢀ18) pro-
ceeded generally in good to very good yields. In all cases,
a mixture of fluoride epimers was produced (see table in
the SI), which however ultimately is inconsequential for the
JuliaꢀKocienski olefination. The clean inversion of config-
uration was unambiguously proven by chiral HPLC for the
Mitsunobu reaction with homochiral (S)-2-butanol and
(S)-1-phenyl ethanol (entries 8ꢀ9). In each case a mixture
of homochiral fluoride epimers 10hꢀi was obtained (see SI).
Benzylic and allylic secondary alcohols react in higher yields
compared to saturated secondary alcohols, though with the
latter a higher temperature can be employed to increase the
yield. Of note is the formation of 10l (entry 12), which was
only accompanied by <5% of SN20 adduct. This is in stark
contrast to the corresponding alkylation reaction of 5 with
secondary allylic bromides, which exclusively led to the
corresponding SN20 product,10b and is further illustration
of the expanded scope provided by the Mitsunobu process.
The reaction has also been demonstrated on cyclic alcohols,
with the benzylic 4o leading to a higher yield than 4n. Given
the large size of the nucleophile, the Mitsunobu reaction
with cis-4-tBu cyclohexanol 4p is higher yielding compared
with the trans stereomer 4q. Compared to the saturated
cyclohexanols, a much higher yield was obtained using
2-cyclohexenol (at 20 °C).
The next step is a decarboxylation reaction to give the
desired JuliaꢀKocienski fluorosulfone precursors6, which
was generally achieved in good to excellent yields under
Krapcho decarboxylation conditions.10b A considerable
improvement in yield and purity was achieved by replac-
ing NaCl by KBr and increasing the amount of water to
100 equiv. The less basic bromide significantly decreased
the amount of sulfone elimination, which is a side reaction
with allylic and benzylic substrates (see SI).
The JuliaꢀKocienski reaction involving fluorosulfones 6
gave low yields and incomplete conversions with tBuOK as
base (not shown). To avoid HMPA as the additive,
NaHMDS was then employed as the base (Table 2),16
leading to good yields. With aromatic aldehydes E-alkenes
were obtained as the major isomer, but aliphatic aldehydes
gave the Z-alkene as the major isomer.17 The structure of
E-1i was unambiguously determined by X-ray crystallography
(see SI). Interestingly, while the observed diastereoselectiv-
ity could result from the stereoselectivity and possible
reversibility of the initial aldehyde addition step leading to
11, the relative rates of the subsequent Smiles rearrangement
of the up to four possible diastereomers of 11 (M = Na), or
even the precise subsequent elimination mechanism,5,18 the
Jorgensen fluoroalkene E/Z ratios arise from a different
stereodefining step. In that case, the selectivity of the
borohydride β-ketosulfone reduction (Scheme 1), also lead-
ing to 11 (M = Li), appears to be translated into the
fluoroalkene E/Z ratio.
In conclusion, we report the first synthesis of homochiral
R-fluorinated BT-sulfones with a chiral center in the
β-position, based on a successful modified Mitsunobu
reaction. It was shown that the Mitsunobu process has a
much wider structural scope compared to the equivalent
alkylation starting from alkyl halides. Both enantiomeric
forms are readily accessible thanks to the wide availability of
homochiral alcohols. The application of this reaction,
together with an optimized procedure for the required
subsequent Krapcho decarboxylation, results in a conveni-
ent method for the synthesis of homochiral fluoroalkenes
with a branched alkyl group next to the dCF center, using a
JuliaꢀKocienski reaction. The procedure allows a wide
substrate scope, with regard to both substitution of the
fluorinated benzothiazole and the nature of the aldehyde
(aliphatic and aromatic aldehydes). Interestingly, the for-
mer leads to fluoroalkenes predominantly with Z-config-
uration, while the latter leads to E-alkenes as the major
isomer. Nevertheless, the observed diastereoselectivity levels
are within the range usually observed for JuliaꢀKocienski
mediated fluoroalkene formation.
Acknowledgment. The European Community (INTER-
REG IVa channel programme, IS:CE-Chem, Project 4061)
is thanked for financial support.
Supporting Information Available. General reaction
procedures (including optimization of the Krapcho
decarboxylation); characterization of all compounds;
confirmation of optical purity of 6i, 10h, and 10i; X-ray
structure of 6o; copies of NMR spectra of all compounds
(1H, 13C, 19F). This material is available free of charge via
the Internet at http://pubs.acs.org.
(16) Chevrie, D.; Lequeux, T.; Demoute, J. P.; Pazenok, S. Tetra-
hedron Lett. 2003, 44, 8127–8130.
(17) These results are consistent with observations by Zajc; see ref 8.
(18) (a) Aissa, C. Eur. J. Org. Chem. 2009, 1831–1844. (b) Robiette,
R.; Pospisil, J. Eur. J. Org. Chem. 2013, 836–840.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX