CF3 radicals from triflic anhydride and collidine: Their trapping by a trimethylsilylenolether

Article abstract of DOI:10.1016/j.jfluchem.2010.02.012

The interaction of triflic anhydride with s-collidine in the presence of the (trimethylsilyl)enolether of acetophenone led to duplication products the structure of which could only be explained by the formation of CF₃ radicals.

Full text of DOI:10.1016/j.jfluchem.2010.02.012

Journal of Fluorine Chemistry 131 (2010) 738–741
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
Short communication
CF3 radicals from triflic anhydride and collidine: Their trapping by a
trimethylsilylenolether
*
´
Henri Rudler , Andree Parlier, Charline Denneval, Patrick Herson
Institut Parisien de Chimie Mole´culaire UMR CNRS 7201, Universite´ Pierre et Marie Curie, CC 47, 4 place jussieu 75252 Paris cedex 5, France
A R T I C L E I N F O
A B S T R A C T
Article history:
The interaction of triflic anhydride with s-collidine in the presence of the (trimethylsilyl)enolether of
acetophenone led to duplication products the structure of which could only be explained by the
formation of CF3 radicals.
Received 27 January 2010
Received in revised form 25 February 2010
Accepted 26 February 2010
Available online 6 March 2010
ß 2010 Elsevier B.V. All rights reserved.
Keywords:
Triflic anhydride
Collidine
CF3 radicals
(Trimethylsilyl)enolether
Dimerization
1. Introduction
transformation 5 ! 6. Indeed, Binkley and Ambrose observed the
formation of two unexpected products 10 and 11, yet in low yield,
Improving or finding new methods for the introduction of
fluorine in organic compounds, and especially in heterocycles, is an
important target since many of the selling drugs appearing on the
market contain at least one fluorine atom [1]. During our
investigations directed towards the synthesis of new hetero-
cycle-fused lactones via halolactonization reactions, [2] we were
intrigued by a surprising yet possibly important, if improved,
reaction. Indeed, we observed that the interaction of pyridine 1
with bis(trimethylsilyl) ketene acetals 2 (trimethylsilyl = TMS) in
the presence of an excess of triflic anhydride Tf2O 3 (Tf2O = (CF3-
SO2)2O) led directly to the trifluoromethyllactone 6 (Scheme 1) [3].
During this transformation, triflic anhydride not only activates
as expected the pyridine nucleus towards bis(trimethylsilyl)
ketene acetal nucleophiles 2 via the pyridinium triflate 4 to give
dihydropyridines 5, but delivers also an electrophilic CF3 group
which induces the lactonization reaction of 5. A mechanism for the
last step 5 ! 6 involving pseudocationic CF3 species was tentatively
suggested [3]. Such species might originate upon electron transfers
from the dihydropyridines, known reducing agents, [4] to triflic
anhydride, also known as an oxidant [5] in some special cases [6,7].
This result prompted us to re-examine the first and sole other
transformation in which triflic anhydride appeared to be the
source of CF3 moieties in order to try to get more insight into the
from triflic anhydride 3 reacting with s-collidine 7 [8]. According to
these authors, the expected yet unstable trifyl triflate 8 might lead
to the anhydrobase 9 upon deprotonation and then, in the case of
either a non-concerted cationic A or radical B pathway, to 10, and
with loss of SO2 to 11 (Scheme 2). Negative results from CIDNP
experiments led them however to eliminate the second pathway,
the formation of radical pairs and to conclude for the involvement
of carbocationic CF3 species.
The purpose of this communication is to show that although not
detectable by the CIDNP technique, CF3 radicals are in fact formed
at least to some extend during the interaction of collidine with
triflic anhydride. They could however only be directly trapped by
the (trimethylsilyl)enolether 12 leading to duplication products 13
which were fully characterized by X-ray crystallography, and also
to the corresponding a-trifluoromethylketone 15. Besides, the low
yield of formation of these products might be assigned to a fast,
competitive, direct, triflic anhydride induced transformation of
(trimethylsilyl)enolethers into vinyl triflates.
2. Results and discussion
Many ways to trap radical species are known. Among them,
their interaction with enol ethers or esters appeared to be useful
both on a mechanistic and on a synthetic point of view since they
can lead to
choose, as
a
a
-substituted ketones [7b,9]. For that purpose, we
first example, (trimethylsilyl)enolethers which
* Corresponding author. Tel.: +33 0 1 44275092; fax: +33 0 1 44273787.
E-mail address: henri.rudler@upmc.fr (H. Rudler).
appeared to survive partially under such harsh conditions. Thus
0022-1139/$ – see front matter ß 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2010.02.012

H. Rudler et al. / Journal of Fluorine Chemistry 131 (2010) 738–741
739
Scheme 1.
when triflic anhydride 3 (1.5 eq.) was added to a dichloromethane
solution of collidine 7 (2 eq.) and 1-phenyl-1-trimethylsiloxyethy-
lene 12 (1 eq.) at room temperature, the mixture turned rapidly
deep red. After 12 h, water was added and the organic layer
washed with aqueous potassium hydroxide. After evaporation of
the volatiles, the residue was chromatographed on silica gel,
leading, besides unreacted collidine, to five new compounds
(Scheme 3). Elution with light petroleum ether/dichloromethane
(97/3) led successively to two crystalline, difficult to separate
compounds in almost equal amounts in a low 5% yield. Extended
NMR experiments allowed to establish the structures of these two
compounds. The 1H NMR spectrum of the less polar product 13a,
Fig. 1. X-ray structure of compound 13a.
signals at d 5.60 (d, J = 4 Hz) and 5.37 (d, J = 4 Hz)ppm, each for one
proton, then a low-melting solid (5%), the NMR data of which
agreed with those of the known trifluoromethyl acetophenone 15,
disclosing in the 1H NMR spectrum a signal at
d
3.78 ppm for the
two CH2CF3 protons, as a quartet (J = 10 Hz) and in the 19F NMR
spectrum, a triplet (J = 10 Hz), at
d
ꢀ61.98 ppm [6f]. Finally, a more
m.p. 168 8C, disclosed a signal at
d
0.09 ppm, corresponding to a
polar product (10%), eluted with PE/dichloromethane (60/40), as a
yellow oil. Its spectroscopic data as well as its mass spectrum
agreed with structure 16. Indeed, the 1H NMR spectrum disclosed a
TMS group, at 1.89 ppm and
d
d 3.30 ppm as two doublets of
quartets (J = 11 and 16 Hz), each for one proton, and at
d 7.36 to
7.49 ppm for five aromatic protons.
singlet, at d 7.02 ppm for one proton, and d 2.51, 2.62 and 2.84 ppm
Both the 19F NMR and 13C NMR spectra confirmed the presence
of fluorine, hence of a CH2CF3 group, giving in the 19F spectrum a
for three methyl groups. Both the 19F and the 13C NMR spectra
confirmed the presence of a –SO2CF3 group on the collidine ring
triplet at
d
ꢀ54.88, (J = 11 Hz) and in the 13C spectrum a quartet at
d
d
with a singlet at
d
ꢀ79.33 ppm and a quartet at
d 120.38 ppm,
126.33 ppm (J = 276 Hz) for the CF3 group, and a quartet at
J = 325 Hz. Deprotection to the corresponding diols 17a,b was
achieved upon treatment of 13a,b with NBu4F, H2O [12]. The NMR
data of the diols fully agreed with such structures with the typical
series of doublets of quartets for the CH2CF3 groups and the
disappearance of the signals for the SiMe3 groups (Scheme 4).
This confirms thus that CF3 species can be formed from triflic
anhydride reacting with collidine, but that these species are in fact
free radicals escaping from the solvent cage. (Scheme 2, route B)
The intermediate stabilized radicals 18, obtained upon their
39.28 ppm (J = 26 Hz) for the CH2 linked to the CF3 group. Crystals
suitable for an X-ray structure determination were grown from
heptane solutions at low temperature. As can be seen on the ORTEP
view (Fig. 1), compound 13a results from the combination of two
C(Ph)(OSiMe3)(CH2CF3) units. Having a plane of symmetry, it is a
meso compound.
Such a combination can however give rise besides the meso
form to a pair of d,l enantiomers: [10] the slightly more polar
compound 13b corresponds clearly to the d,l isomer. Indeed, its 1H
NMR spectrum is only slightly different from that of 13a, showing
interaction with the enol ether 12 can either undergo
dimerization reaction to give the observed pinacol ethers 13a,b
or might undergo further oxidation to 19 to afford the
a
up signals at
J = 16 and 11 Hz) for the two diastereomeric hydrogens of the
methylene groups, and at 7.13 to 7.20 ppm signals for five
d
0.28 for the TMS group, at
d
2.86 ppm and 3.12 (dq,
a
trifluoromethylketone 15 via 19 (Scheme 5). As far as the product
16 is concerned, its structure is not unexpected and is the result of
the sulfination of the collidine ring.
When however the enol ether of cyclohexanone was used,
neither of the expected addition products was observed. Instead, a
d
aromatic protons. Three more polar compounds could although be
isolated: first a liquid the NMR data of which were in all respect
identical with those of the known vinyl triflate 14, [11] with typical
Scheme 2.

740
H. Rudler et al. / Journal of Fluorine Chemistry 131 (2010) 738–741
towards triflic anhydride: a fast, undesired reaction led only to 23,
1,3,5-triphenylbenzene, a triflic anhydride induced elimination-
trimerization product [16] (Scheme 7).
3. Conclusion
The results reported herein provide clear-cut evidence for the
formation of electrophilic CF3 radicals upon the interaction of
methyl-substituted pyridines with triflic anhydride. In spite of our
efforts, there remains however a considerable drawback to these
quenching reactions: the high reactivity of triflic anhydride
towards all of the unsaturated scavengers used so far. Work is
progressing to the use of such substrates in the transformation
depicted in the Scheme 1.
Scheme 3.
4. Experimental
All commercially available reagents were used without further
purification. 1H, 13C, 19F NMR spectra were recorded on a Bruker
Avance 400 spectrometer. CH2Cl2 was distillated on CaH2 before
use.
Scheme 4.
4.1. Reaction of (trimethylsilyl)enolether 12 with triflic anhydride in
the presence of collidine 7
To a solution of collidine (1 g, 8.32 mmol, 1.1 mL) and 1-phenyl-
1-trimethylsiloxyethylene (0.8 g, 4.16 mmol, 852
mL) in dichlor-
omethane (50 mL) was added slowly at room temperature triflic
anhydride (1.7 g, 6.24 mmol, 1.1 mL) with syringe. The mixture
turned rapidly deep red. After12 h, waterwas added and the organic
layer washed with aqueous potassium hydroxide. After evaporation
of the volatiles, the residue was chromatographed on silica gel.
Elution with light petroleum ether/dichloromethane (97/3) led
successively to two crystalline, difficult to separate compounds in
almost equal amounts 13a and 13b (56 mg, 5% yield).
Scheme 5.
13a: white solid, m.p. 168 8C. 1H NMR (400 MHz, CDCl3):
d 7.49
(m, 2H), 7.36 (m, 3H), 3.30 (dq, J = 16, 11 Hz, 1H, CH2), 1.89 (dq,
J = 16, 11 Hz, 1H, CH2), 0.09 (s, 9H, SiMe3); 13C NMR (100 MHz,
CDCl3)
(q, J = 276 Hz, CF3), 83.82 (C–O), 39.28 (q, J = 26 Hz, CH2), 2.51
(SiMe3); 19F NMR (376 MHz, CDCl3)
ꢀ54.88 (t, J = 11 Hz); Analysis
for C24H32F6O2Si2: calcd, C, 55.15; H, 6.17; found, C, 55.26; H, 6.11.
13b: white solid. 1H NMR (400 MHz, CDCl3):
7.20 (m, 2H), 7.13
d 138.62 (q, arom), 128.84, 127.96, 127.02 (arom), 126.33
d
Scheme 6.
d
(m, 2H), 6.70 (m, 2H), 3.12 (dq, J = 16, 11 Hz, 1H, CH2), 2.86 (dq,
J = 16, 11 Hz, 1H, CH2), 0.28 (s, 9H, SiMe3); 13C NMR (100 MHz,
quantitative transformation of the TMS enol ether 20 into the
corresponding vinyl triflate 21 took place (Scheme 6).
CDCl3)
(q, J = 276 Hz, CF3), 85.94 (C–O), 37.26 (q, J = 26 Hz, CH2), 3.01
(SiMe3); 19F NMR (376 MHz, CDCl3)
ꢀ55.85 (t, J = 11 Hz). Further
elution gave a liquid 14 [11] (250 mg). Then 15: (40 mg, 5% yield)
deliquescent solid. 1H NMR (400 MHz), CDCl3):
7.93 (m, 2H), 7.63
(m, 1H), 7.50 (m, 2H), 3.78 (q, J = 10 Hz, 2H, CH2); 13C NMR
(100 MHz, CDCl3) 189.78 (CO), 135.94, 134.28, 129.03, 128.45
(arom), 124.08 (q, J = 275 Hz, CF3), 42.15 (q, J = 28 Hz, CH2); 19F
NMR (376 MHz, CDCl3)
ꢀ61.98 (t, J = 10 Hz). And finally 16:
(158 mg, 10% yield) yellow oil. 1H NMR (400 MHz, CDCl3):
7.02 (s,
1H, arom), 2.84 (s, 3H, CH3), 2.62 (s, 3H, CH3), 2.51 (s, 3H, CH3); 13
NMR (100 MHz, CDCl3) 164.63, 162.56, 153.28 (3 C–Me)126.18
(C–H), 123.54 (C–SO2), 120.38 (q, J = 325 Hz, CF3), 26.13 (CH3),
d 137.92 (q, arom), 128.87, 127.96, 126.52 (arom), 126.41
A similar transformation could be achieved in the absence of
collidine, by simply mixing the enol ether 20 in dichloromethane,
at room temperature, with triflic anhydride. Moreover, under such
conditions, the enol ether 12 behaved similarly: it gave, as sole
product, the vinyl triflate 14 with the exclusion of any products 13
and 15 arising from radical reactions [14,15]. This observation is
thus in agreement with the partial transformation of 12 into 14
during the transformation depicted in the Scheme 3 and with the
involvement of both picoline and triflic anhydride for the
formation of CF3 radicals [13]. Alkylenolethers such as 22 which
might also be used as radical scavengers proved even more reactive
d
d
d
d
d
C
d
24.47 (CH3), 22.58 (CH3); 19F NMR (376 MHz, CDCl3)
d
ꢀ79.33 (s,
CF3); HRMS for C9H11O2NF3 (M+H+), calcd: 254.04571; found:
254.04509.
4.2. Deprotection of 13a and 13b with Bu4NF
To a solution of a mixture (50/50) of 13a and 13b (50 mg,
Scheme 7.
0.17 mmol) in THF (3 mL) was added a solution of Bu4NF in THF

H. Rudler et al. / Journal of Fluorine Chemistry 131 (2010) 738–741
741
[3] H. Rudler, A. Parlier, C. Sandoval-Chavez, P. Herson, J.C. Daran, Angew. Chem. Int.
Ed. 47 (2008) 6843–6846.
(1 M, 0.2 mL). After 17 h at room temperature water was added
and the organic layer washed several times with water. After
evaporation of the volatiles, the residue was chromatographed on
silica gel. Elution with light petroleum ether/ethyl acetate (95/5)
led successively to two oily, difficult to separate compounds in
equal amounts 17a (18 mg, 30% yield))and 17b (19 mg, 30% yield)
[4] R. Kumar, R. Chandra, in: R. Katritzky (Ed.), Advances in Heterocyclic Chemistry,
vol.78, University of Florida, Gainesville, 2001, pp. 259–313.
[5] (a) G. Maas, P. Stang, J. Org. Chem. 46 (1981) 1606–1610;
(b) I.L. Baraznenok, V.G. Nenajdenko, E.S. Balenkova, Tetrahedron 56 (2000)
3077–3119.
[6] (a) For the introduction of CF3 see for example K. Sato, M. Higashinagata, T. Yuki,
A. Tarui, M. Omote, I. Kumadaki, A. Ando, J. Fluorine Chem. 129 (2008) 51–55;
(b) T. Billard, B.R. Langlois, Eur. J. Org. Chem. (2007) 891–897;
(c) E. Magnier, J.C. Blazejewski, M. Tordeux, C. Wakselman, Angew. Chem. Int. Ed.
45 (2006) 1279–1282;
17a: 1H NMR (400 MHz, CDCl3):
3.19 (dq, J = 16, 11 Hz, 1H, CH2), 2.50 (q, J = 2 Hz, 1H, OH), 2.09 (dq,
J = 16, 11 Hz, 1H, CH2); 13C NMR (100 MHz, CDCl3)
138.82 (q
d 7.59 (m, 2H), 7.40 (m, 3H),
d
(d) I. Kieltsch, P. Eisenberger, A. Togni, Angew. Chem. Int. Ed. 46 (2007) 754–757;
(e) N. Shibata, S. Mizuta, H. Kawai, Tetrahedron: Asymmetry 19 (2008) 2633–
2644;
(f) K. Sato, M. Higashinagata, T. Yuki, A. Tarui, I. Kumadaki, A. Ando, J. Fluorine
Chem. 129 (2008) 51–55;
arom), 128.92, 127.83, 127.78 (arom), 126.45 (q, J = 276 Hz, CF3),
127.56 (arom), 78.14 (C–OH), 39.36 (q, J = 25 Hz, CH2); 19F NMR
(376 MHz, CDCl3)
d
ꢀ58.03 (dt, J = 11, 2 Hz); HRMS calcd for
C
18H16F6O2Na: 401.09467, found 401.09418. 17b:1H NMR
(g) J. Boivin, L.E. Kaim, S.Z. Zard, Tetrahedron Lett. 33 (1992) 1285–1288;
(h) T. Umemoto, S. Ishihara, Tetrahedron Lett. 25 (1990) 3579–3582;
(i) M. Me´debielle, W.R. Dolbier Jr., J. Fluorine Chem. 129 (2008) 930–942.
[7] (a) For the loss of SO2 from CF3SO2 see C.-M. Hu, F.-L. Qing, W.-Y. Huang, J. Org.
Chem. 56 (1991) 2801–2804;
(b) B. Langlois, B. Laurent, N. Roidot, Tetrahedron Lett. 33 (1992) 1291–1292.
[8] (a) R.W. Binkley, M.G. Ambrose, J. Org. Chem. 48 (1983) 1777–1779;
(b) T. Netscher, P. Bohrer, Tetrahedron Lett. 46 (1996) 8359–8362 (see also).
[9] I. Al Adel, B. Adeoti Salami, J. Levisalles, H. Rudler, Bull. Chem. Soc. (1976) 930–
933.
(400 MHz, CDCl3):
d 7.32 (m, 5H), 3.30 (q, J = 2 Hz, 1H, OH), 3.13
(dq, J = 16, 11 Hz, 1H, CH2), 2.43 (dq, J = 16, 11 Hz, 1H, CH2); 13C
NMR (100 MHz, CDCl3)
(q, J = 276 Hz, CF3), 78.94 (C–OH), 37.91 (q, J = 26 Hz, CH2); 19F
NMR (376 MHz, CDCl3)
ꢀ57.81 (t, J = 11 Hz).
d 137.02 (q arom), 128.49 (arom), 126.41
d
Supplementary material
[10] R.E. Balsells, A.R. Frasca, Tetrahedron 38 (1982) 245–255 (see for example).
[11] P.J. Stang, A.G. Anderson, J. Am. Chem. Soc. 100 (1978) 1520–1528.
[12] M.A. Brook, Silicon in Organic, Organometallic and Polymer Chemistry, J. Wiley,
New-York, 2000.
[13] One of the referees suggested the involvement of trifluoromethyltrifluorometha-
nesulfonate CF3OSO2CF3, which might form during the preparation of triflic
anhydride, as a source of CF3 radicals. Although such a hypothesis cannot be fully
eliminated at the present stage, we did not observe the formation of dimers
during the interaction of triflic anhydride with the enol ether 12 in the absence of
collidine. See also S.L. Taylor, J.C. Martin, J. Org. Chem. 52 (1987) 4147–4159 (and
references therein).
Crystallographic data (excluding structure factors) for the
structural analysis of 13a have been deposited with the Cambridge
Crystallographic Data Centre: CCDC No 738546. Copies of the
crystallographic data may be obtained free of charge from: The
Director, CCDC, 12 Union Road, Cambridge CB2 1Z, UK (fax: +44 123
336033; E-mail: deposit@ccdc.ac.uk or http://www.ccdc.cm.ac.uk).
References
[14] Vinyl triflates have been prepared starting either from the corresponding ketones
upon their interaction with triflic anhydride in the presence of an amine, in rather
low yields (45%) or from the (TMS) enol ethers, via their lithium enolates [11].
Alkyl triflates were similarly obtained from the corresponding alkyl ethers and
triflic anhydride [15]. The transformations of enol ethers into vinyl triflates
observed herein will be described in a forthcoming paper.
´
´
´
[1] (a) J.P. Begue, D. Bonnet-Delpon, Chimie Bioorganique et Medicinale du Fluor,
EDP Sciences/CNRS, Paris, France, 2005;
(b) K.L. Kirk, Org. Process Res. Dev. 12 (2008) 305–321;
(c) D. O’Hagan, Chem. Soc. Rev. 37 (2008) 308–319.
[2] (a) H. Rudler, A. Parlier, L. Hamon, P. Herson, P. Chaquin, J.C. Daran, Tetrahedron
65 (2009) 5552–5562;
´
´
[15] C. Aubert, J.P. Begue, Synthesis (1985) 759–760.
[16] For related trimerization reactions see for example F. Ono, Y. Ishikura, Y. Tada, M.
Endo, T. Sato, Synlett, (2008) 2365–2367.
(b) A. Parlier, C. Kadouri-Puchot, S. Beaupierre, N. Jarosz, H. Rudler, L. Hamon, P.
Herson, J.C. Daran, Tetrahedron Lett. 50 (2009) 7274–7279.