A practical source of chlorodifluoromethyl radicals. Convergent routes to gem -difluoroalkenes and -dienes and (2,2-Difluoroethyl)-indoles, -azaindoles, and -naphthols

Article abstract of DOI:10.1021/ol501063a

The preparation of O-octadecyl-S-chlorodifluoromethyl xanthate from chlorodifluoroacetic acid and its use as a convenient source of chlordifluoromethyl radicals is described. This reagent may be used to access gem-difluoroalkenes and -dienes, as well as (2,2-difluoroethyl)indolines, -indoles, and -naphthols.

Full text of DOI:10.1021/ol501063a

Letter
pubs.acs.org/OrgLett
A Practical Source of Chlorodifluoromethyl Radicals. Convergent
Routes to gem-Difluoroalkenes and -dienes and (2,2-Difluoroethyl)-
indoles, -azaindoles, and -naphthols
Pierre Salomon* and Samir Z. Zard*
̀
Laboratoire de Synthese Organique, CNRS UMR 7652, Ecole Polytechnique, 91128 Palaiseau, Cedex, France
S
* Supporting Information
ABSTRACT: The preparation of O-octadecyl-S-chlorodi-
fluoromethyl xanthate from chlorodifluoroacetic acid and its
use as a convenient source of chlordifluoromethyl radicals is
described. This reagent may be used to access gem-
difluoroalkenes and -dienes, as well as (2,2-difluoroethyl)-
indolines, -indoles, and -naphthols.
he incorporation of fluorine atoms into organic scaffolds
fluoromethyl xanthate 3 in 81% yield (Scheme 1). This
expedient preparation of xanthate 3 was reproducibly
T
has been the subject of intense investigations over the past
decades.1 The ability of fluorine to deeply modify the chemical
and physicochemical properties of organic molecules has
proven to be of much importance to the pharmaceutical,
agrochemical, and materials industries. In medicinal chemistry,
the use of fluorinated bioisosteres to impact electronic and
conformational properties or to modify lipophilicity and
metabolic stability has become a widespread tactic in the “hit
to lead” approach.2 Of the numerous reactions used in the
synthesis of organofluorine derivatives, radicals offer unique
possibilities. Indeed, radical-based methods have enjoyed
increased importance in recent years.3
Scheme 1. Synthesis of O-Octadecyl-S-chlorodifluoromethyl
Xanthate
The reversible addition-transfer process based on xanthates
and related thiocarbonyl derivatives4 has been exploited to
introduce various fluorinated motifs into highly functionalized
molecules.5 However, with respect to the chlorodifluoromethyl
radicals, the only precursor so far described is bromochlorodi-
fluoromethane, and its use remains very limited.6 We therefore
envisioned preparing xanthate 3 as a convenient source of these
radicals, allowing perhaps a facile incorporation of the
chlorodifluoromethyl motif into complex substrates.
performed on up to 20 g scale. With the desired reagent in
hand, we first explored the simple radical addition to various
alkenes. The results are displayed in Scheme 2. To simplify
characterization, the xanthate group in the adducts 4a−i was
reductively removed using (Me3Si)3SiH to give products 5a−i
(Scheme 2).9 Olefins derived from pyrazole, D-glucose, several
substituted aromatic rings, and natural products such as sclareol
and an androstanol derivative were successfully transformed.
Functions such as carbamates, acetals, or free alcohols were well
tolerated under these mild conditions, and the overall yield for
the two steps was generally good (up to 90% over two steps;
see Scheme 2). Moreover, the reduction proved totally selective
for the xanthate group, and absolutely no reduction of the
chlorine atom was observed. Reductive dechlorination of 1-
chloro-1,1-difluoro derivatives has been reported to proceed
readily with Bu3SnH.10
By analogy with our earlier work,7 we considered using
commercially available and cheap acid 1 as a precursor for
xanthate 3, since a more classical approach involving direct
substitution does not proceed well in the fluorinated series.
Thus, treatment of chlorodifluoroacetic acid with thionyl
chloride produced chlorodifluoroacetic chloride as a gas,
which was condensed into a stirred solution of sodium O-
octadecyl xanthate in cyclohexane, resulting in the formation of
the corresponding S-acyl xanthate 2. For practical reasons
related to the extreme sensitivity to hydrolysis of S-acyl
xanthates and possible volatility of the product, the commonly
used potassium O-ethyl xanthate salt was replaced by the more
hydrophobic and bulkier sodium O-octadecyl xanthate.
In comparison with bromochlorodifluoromethane, xanthate 3
appears to be a superior precursor of chlorodifluoromethane
radicals, as far as yields and functional group tolerance are
concerned.6 In cases where the xanthate must be used in slight
excess (typically 1.5 equiv), the unreacted reagent is easily
Without isolation, S-acyl xanthate 2 was directly irradiated
with a 250 W tungsten halogen lamp, which induced a radical
chain decarbonylation8 and furnished the desired chlorodi-
Received: April 11, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ol501063a | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
the alcohol into its 2-fluoropyridyl derivative (Scheme 4). Even
though the homolysis is relatively slow, the degenerate nature
Scheme 2. Formation of 1-Chloro-1,1-difluoro-Substituted
Derivatives
Scheme 4. Synthesis of 1,1-Difluorodienes
of the xanthate exchange provides the intermediate carbon
radical 7 with sufficient lifetime to overcome this kinetic barrier.
Furthermore, the base-induced elimination of HCl from
allylated derivatives 9a−e should be significantly easier because
of the increased acidity of the allylic proton and the formation
of a conjugated difluorodiene. Since the starting allylic alcohol
is often prepared from a ketone or an aldehyde by addition of a
vinyl metal reagent, the process would correspond to an overall
difluoro-olefination of a carbonyl substrate. In the event,
exposure of fluoropyridine derivative 8a to xanthate 3 in the
presence of a stoichiometric amount of lauroyl peroxide (DLP)
gave the desired adduct 9a in a good 68% yield (Scheme 5
entry 1).
recovered at the end of the reaction. Product 5i is the result of a
double radical addition to vinyl acetate, accomplished by having
2 equiv of vinyl acetate present in the medium. Terminal,
unhindered monosubstituted alkenes are the best substrates,
but previous work on xanthates has shown that 1,1-
disubstituted alkenes and strained olefins, such norbornenes
and cyclobutenes, are also suitable.4
A known synthetic modification of the -ClF2 motif is
dehydrochlorination by treatment with an organic11 or
inorganic12 base, leading to a terminal gem-difluoroalkene. In
the present case, heating products 5a−h with large excess of
DBU at 120 °C provided the desired difluoro-olefins 6a−h in
very good yield (Scheme 3).
Scheme 5. Radical Addition−Fragmentation on Activated
Allylic Alcohols
Scheme 3. Synthesis of gem-Difluorovinyl Compounds
The difluorovinyl group is useful as an isostere of a
carbonyl,1,2 and while some of the compounds described
above could have been made using more traditional sources of
chlorodifluoromethyl radicals, those obtained in the following
section are much less accessible by previous routes.
Difluorodienes, for example, are relatively rare, and the few
approaches reported in the literature rely on either a Wittig
reaction on enones or on organometallic coupling of halogen-
substituted gem-difluoro-olefins.13 By using 2-fluoropyridyl
derivatives14 of allylic alcohols as radical allylating agents,
access to functionalized difluorodienes becomes particularly
easy.
Other examples are collected in Scheme 5. A particularly
important feature is the possibility of generating difluoroenol
ethers such as 9d and 9e by this approach (Scheme 5, entries 4
and 5). Such chlorodifluoroenol ethers are extremely scarce in
the literature, and their chemistry has remained largely
unexplored. We found, for example, that they were completely
resistant to acid hydrolysis under conditions where the starting
enol ether is totally cleaved. This is obviously a manifestation of
the electron-withdrawing effect of the chlorodifluoro motif.
The dehydrochlorination indeed occurred more easily, as was
expected. The reaction took place at 50 °C using an excess of
DBU as the base and was complete within 2 h. The gem-
difluorodienes 10a−d were obtained in generally good yields
(Scheme 6). Unfortunately, in the case of 9e the reaction gave a
We had shown recently that homolysis of the otherwise
strong C−O bond in alcohols becomes possible by converting
B
dx.doi.org/10.1021/ol501063a | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
addition to the expected difluoroalkene 13a, another product
appeared, which turned out to be indole 14a. This compound
became almost the exclusive product when the temperature was
raised to 55 °C and the reaction time was extended to 14 h
(Scheme 8).
Scheme 6. Synthesis of 1,1-Difluorodienes and 1,1-
Difluorodienol Ethers
Scheme 8. From Indolines to Indoles, an Unexpected
Double Bond Migration
complex mixture, even at room temperature. The glucose-
derived structure does not tolerate neat DBU and decomposes
rapidly, possibly through base-induced β-elimination of
alkoxides and ultimate opening of the tetrahydrofuran ring.
Traces of the desired product were nevertheless observed and
isolated, proving that the elimination occurred anyway.
Attempts with milder conditions, such as replacing DBU with
TEA or pyridine at different temperatures, were unsuccessful,
and no clean elimination could be accomplished.
Another unique feature of the present approach is the ability
of using the xanthate group in the adduct to accomplish a ring
closure on a nearby aromatic ring, thus opening access to
valuable fluorinated polycyclic aromatic and heteroaromatic
derivatives of importance to the pharmaceutical and agro-
chemical industries.4 The reaction of xanthate 3 with N-allyl-N-
mesyl anilines initiated by a small amount of DLP gave the
corresponding adducts 11a−e, which when exposed to
stoichiometric amounts of peroxide proceeded to the
corresponding indolines 12a−e in good to excellent yield,
with the exception of the pyridine derivatives 11f,g where the
cyclization step proved disappointingly poor for reasons that
are not yet clear (Scheme 7).
An unexpectedly facile double bond migration of the olefinic
bond appears to be taking place to give ultimately the
thermodynamically more stable indole structure. A further
increase in the temperature to 90 °C for 14 h resulted in the
formation of another new compound, which proved to be the
demesylated indole 15a, isolated in a good 77% yield. By an
appropriate choice of conditions, it is therefore possible to
access either 13a, 14a, or 15a as summarized in Scheme 8 (the
yields in parentheses are for isolated products; the others were
determined by NMR spectroscopy).
Scheme 7. Synthesis of Indolines with Fluorinated Side
Chains
The other indolines behaved in the same manner (Scheme
8), except for compound 12e, where no migration was
observed, and 13b could be isolated as the only elimination
product. The presence of the two electron-donating groups
presumably deactivates the difluorovinyl indoline toward base-
induced migration of the alkene. In contradistinction, the
formation of azaindoles 15f,g occurred particularly readily,
within 2 h, because of the electron-withdrawing nature of the
pyridine ring.
Access to such difluoroethyl indolines is not generally easy
but sometimes highly desirable. Thus, in the case of BACE1
inhibitor 16 (Figure 1), the presence of a difluoroethyl side
chain on the indole portion has a dramatic effect on the
potency, since a 4-fold increase in activity was observed in
comparison with the nonfluorinated analogue.15
This strategy appears to be applicable to the synthesis of 4-
(2,2-difluoroehtyl)-1-naphthols. Indeed, the addition of reagent
3 to 1-(4-methoxyphenyl)-4-buten-1-one followed by peroxide-
mediated ring closure onto the aromatic ring furnished
difluoroethyltetralone 17 in 50% yield (Scheme 9). Exposure
Other heteroaromatic rings could in principle act as viable
substrates. We have previously shown, for example, that
addition−cyclization sequences involving xanthates can be
performed efficiently on pyrimidines.4e Furthermore, the use of
a mesyl instead of an acyl group in the present case was
motivated by the desire to avoid unnecessary complications due
to amide rotamers.
We were surprised to find that the reaction of indoline 12a
with excess DBU occurred already at 40 °C instead of the usual
120 °C required in the purely aliphatic series. Furthermore, in
C
dx.doi.org/10.1021/ol501063a | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(3) (a) Studer, A. Angew. Chem., Int. Ed. 2012, 51, 8950. (b) Besset,
T.; Schneider, C.; Cahard, D. Angew. Chem., Int. Ed. 2012, 51, 5048.
Dolbier, W. R., Jr. Chem. Rev. 1996, 96, 1557.
(4) For reviews, see: (a) Zard, S. Z. Angew. Chem., Int. Ed. Engl. 1997,
36, 672. (b) Quiclet-Sire, B.; Zard, S. Z. Chem.Eur. J. 2006, 12, 6002.
(c) Quiclet-Sire, B.; Zard, S. Z. Top. Curr. Chem. 2006, 264, 201.
(d) Quiclet-Sire, B.; Zard, S. Z. Pure Appl. Chem. 2011, 83, 519. (e)
For an account of the discovery of this process, see: (d) Zard, S. Z.
Aust. J. Chem. 2006, 59, 663. (e) For applications to the synthesis of
heteroaromatics, see: El Qacemi, M.; Petit, L.; Quiclet-Sire, B.; Zard,
S. Z. Org. Biomol. Chem. 2012, 10, 5707.
Figure 1. Example of a 3-(2,2-difluoroethyl)indole derivative
possessing biological activity.
Scheme 9. Synthesis of 4-(2,2-Difluoroethyl)naphthol
(5) (a) Bertrand, F.; Pevere, V.; Quiclet-Sire, B.; Zard, S. Z. Org. Lett.
2001, 3, 1069. (b) Li, S.; Zard, S. Z. Org. Lett. 2013, 15, 5898.
(c) Denieul, M.-P.; Quiclet-Sire, B.; Zard, S. Z. J. Chem. Soc., Chem.
Commun. 1996, 2511. (d) Gagosz, F.; Zard, S. Z. Org. Lett. 2003, 5,
2655. (e) Gagosz, F.; Zard, S. Z. Org. Synth. 2007, 84, 32. (f) Tournier,
L.; Zard, S. Z. Tetrahedron Lett. 2005, 46, 455. (g) Jean-Baptiste, L.;
Yemets, S.; Legay, R.; Lequeux, T. J. Org. Chem. 2006, 71, 2352.
(6) (a) Tarrant, P.; Lovelace, A. M. J. Am. Chem. Soc. 1955, 77, 768.
(b) Molines, H.; Wakselman, C. J. Fluorine Chem. 1987, 37, 183.
(7) (a) Salomon, P.; Zard, S. Z. Org. Lett. 2014, 16, 1482. (b) Li, S.;
Zard, S. Z. Org. Lett. 2013, 15, 5898.
(8) This reaction proceeds by way of a radical chain decarbonylation;
see: (a) Barton, D. H. R.; George, M. V.; Tomoeda, M. J. Chem. Soc.
1962, 1967. (b) Delduc, P.; Tailhan, C.; Zard, S. Z. J. Chem. Soc.,
Chem. Commun. 1988, 308.
(9) Chatgilialoglu, C. Chem. Rev. 1995, 95, 1229.
(10) Yoshida, M.; Morinaga, Y.; Iyoda, M. J. Fluorine Chem. 1994, 68,
33.
(11) (a) Pitterna, T.; Boeger, M.; Maienfisch, P. Chimia 2004, 58,
108. (b) Archibald, T. G.; Baum, K. J. Org. Chem. 1990, 55, 3562.
(12) (a) Henne, A. L.; Waalkes, T. P. J. Am. Chem. Soc. 1946, 68, 496.
(b) Ackerstaff, E.; Humm, J.; Koutcher, J.; Suehiro, M.; Ouerfelli, O.;
Torchon, G.; Yang, G. Bioorg. Med. Chem. 2011, 19, 2287.
(13) Chelucci, G. Chem. Rev. 2012, 112, 1344.
(14) (a) Charrier, N.; Quiclet-Sire, B. a.; Zard, S. Z. J. Am. Chem. Soc.
2008, 130, 8898. (b) Braun, M.-G.; Quiclet-Sire, B.; Zard, S. Z. J. Am.
Chem. Soc. 2011, 133, 15954. (c) Debien, L.; Quiclet-Sire, B.; Zard, S.
Z. Org. Lett. 2011, 13, 5676. (d) Debien, L.; Quiclet-Sire, B.; Zard, S.
S. Org. Lett. 2012, 14, 5118. (e) Debien, L.; Braun, M.-G.; Quiclet-Sire,
B.; Zard, S. Z. Org. Lett. 2013, 15, 6250.
to DBU at 90 °C gave directly naphthol 18 in 52% yield, while
under milder conditions the intermediate difluorovinyltetralone
19 could be isolated in 58% yield. The synthesis of
naphthalenes with a specific substitution pattern is seldom
trivial, and the present route nicely complements existing
procedures.16
In conclusion, we have developed an efficient, flexible
synthesis of a new, convenient precursor of chlorodifluor-
omethyl radicals. It allows the synthesis of gem-difluoroalkenes
and -dienes with a high tolerance for functional groups.
Moreover, an unusually facile double bond migration provides a
route to an interesting class of difluoro-substituted indoles,
azaindoles, and naphthols.
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental procedures, full spectroscopic data, and copies of
1H and 13C NMR spectra for all new compounds. This material
is available free of charge via the Internet at http://pubs.acs.org.
(15) Rueeger, H.; Lueoend, R.; Rogel, O.; Rondeau, J.-M.; Mobitz,
̈
H.; Machauer, R.; Jacobson, L.; Staufenbiel, M.; Desrayaud, S.;
Neumann, U. J. Med. Chem. 2012, 55, 3364.
AUTHOR INFORMATION
Corresponding Authors
(16) (a) de Koning, C. B.; Rousseau, A. L.; van Otterlo, W. A. L.
Tetrahedron 2003, 59, 7. (b) Saito, S.; Yamamoto, Y. Chem. Rev. 2000,
100, 2901.
■
*E-mail: pierre.salomon@polytechnique.edu.
*E-mail: samir.zard@polytechnique.edu.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This article is dedicated with respect and admiration to Prof.
Tien-Yau Luh (National Taiwan University). We thank Ecole
Polytechnique for a scholarship to P.S.
REFERENCES
(1) (a) Wang, J.; San
Sorochinsky, A. E.; Fustero, S.; Soloshonok, V. A.; Liu, H. Chem. Rev.
2014, 114, 2432. (b) Ojima, I. Fluorine in Medicinal Chemistry and
■
́
chez-Rosello,
́
M.; Acena, J. L.; del Pozo, C.;
̃
́ ́
Chemical Biology; Blackwell: Oxford, 2009. (c) Begue, J.-P.; Bonnet-
Delpon, D. Bioorganic and Medicinal Chemistry of Fluorine; Wiley: New
York, 2008. (d) Hagmann, W. K. J. Med. Chem. 2008, 51, 4359.
(2) (a) Cho, E. J.; Senecal, T. D.; Kinzel, T.; Zhang, Y.; Watson, D.
A.; Buchwald, S. L. Science 2010, 328, 1679. (b) Muller, K.; Faeh, C.;
Diederich, F. Science 2007, 317, 1881. (c) Hiyama, T. Organofluorine
Compounds: Chemistry and Applications; Springer: Berlin, 2000.
D
dx.doi.org/10.1021/ol501063a | Org. Lett. XXXX, XXX, XXX−XXX